-
https://clinton.presidentiallibraries.us/files/original/d9637f54d21bd2c149858620e9ced229.pdf
8aaea61edce8be87c8d65edc1124ca9e
PDF Text
Text
FOIA Number:
2006-0469-F (2)
FOIA
MARKER
This is not a textual record. This is used as an
administrative marker by the William J. Clinton
Presidential Library Staff.
Collection/Record Group:
Clinton Presidential Records
Subgroup/Office of Origin:
Speechwriting
Series/Staff Member:
Michael Waldman
Subseries:
OA/ID Number:
14454
FolderlD:
Folder Title:
MN - 1993 Budget: The State of Working America 1992
Stack:
Row:
Section:
Shelf:
Position:
s
92
4
3
3
�\
If you want to know what happened to the economic well-being of the average
American in the past decade or so this is the book for you. It should be required
reading for Americans of all political persuasions as they look to the future of the
country.
Richard Freeman
Harvard University
The Economic Policy Institute has become the best, most interesting, most credible,
and I think most effective voice for economic reform in the country.
Robert Heilbmner
New School for Social Research
The Slate of Working America provides very valuable factual and analytic
material on the economic conditions of American workers. It is the very best source of
information on this important subject
Ray Marshall
former Secretary of Labor,
University of Texas
The State of Working America documents in great detail how middle class families have
paid dearly — in reduced wages and declining standard of living — as a result of the
Reagan and Bush economic policies of the 1980s.
The
State of
Working
America
1992-93
Congressman Richard Gephardt
House Majority Leader
By documenting the legacy of the 1980s — increased inequality and diminished
opportunities — this book has provided an important and sobering message for the
1990s.
Lawrence Mishel
Jared Bernstein
Mario Cuomo
Governor of New York
No matter what political camp you're in, this is the single most valuable book I know
of about the slate of America, period It i i the most referenced, most influcntiaJ
rcsouive book of its kind.
JeffMadrtck
Television news correspondent
Economic Policy Institute
�/
The
State of
Working
America
1992-93
Lawrence Mishel
Jared Bernstein
Economic Policy Institute
17.10 Rh<xlc Island Aviv. NWI Suitt- 2(M), Washington, IX: 2(H).16
(202) 775-HHIO
�Lawrence Mishel is the Research Director of the Kconomic Policy Institute He is the author of the previous two
eciitions of The State of Working America (1988-1989 with
Jacqueline Simon, 1990-1991 with David M. Frankel), The
Myth of the Coming l a b o r Shortage (with Ruy Tcixcira),
Manufacturing Numbers and Shortchanging Education
(with M. Fdith Rasell). He holds a Ph.D. in economics from
the University of Wisconsin and has published a variety of
academic and non-academic journals.
Jared Bernstein, a jazz bassist and former social worker, is
an economist at the Ixonnmic Policy Institute. Mr Bernstein is completing his dissertation in pursuit of the Doctor
of Social Welfare degree at Columbia University.
Copyright © 1992
KCONOMK; POLICY INSTITUTK
1730 Rh(xlc Island Avenue, NW
Suite 200
Washinnton, DC 20036
To my mother, (ieraldine Baron
I..M.
To my family and to Beth
JB.
�TABLE OF CONTENTS:
Executive Summary
1
Introduction
13
Presentation & Methodology
25
Acknowledgement
28
CHAPTER 1:
Family Income: Slow and Unequal Growth
31
Sluggish Income Growth
Young Families Hurt Most
Income Growth Among Racial/Ethnic Groups
Only Married Couples Gain
Growing Inequality of Family Income
Greater Capital Incomes, Lower Labor Incomes
The Shrinking Middle Class?
Increased Work by Wives Cushions Income Fall
The Downside of Increased Labor Market Work
Falling Behind the Earlier Generations
Income Mobility
33
37
39
41
43
55
65
69
83
87
91
CHAPTER 2:
Taxes: A Further Cause o f Worsening Ineqality
The Tax Burden: Still Light By Comparison
Recent Federal Tax Changes Have Favored The Very Rich
The Diminished Progressivity of Federal Tax Rates
What Federal Tax Changes Mean in Dollars
The Causes of Changes in the Federal Tax Burden
Changes in Corporate Taxation: The Shift to Untaxed Profits
The Shift to State and Local Taxes
CHAPTER 3:
Wages: Working Longer for Less
More Hours and Stagnant Wages
Low-Paid Occupations Fared Worse
Wage Trends by Wage Level
The Male-Female Wage Gap
The Expansion of Low-Wage Jobs
97
97
101
107
109
Ill
117
121
129
131
137
139
145
147
�Benefit Reduction
What Explains Wage Trends
The Role of Education and Experience
Rising Education Wage Differentials
Young Workers Have Been Hurt Most
The Shift to Low-Paying Industries
Trade and Wages
The Union Dimension
An Eroded Minimum Wage
Technology
Executive Pay Soars
Increased Black-White Wage Inequality
Did Lower Wage Buy More Employment?
Jobs of the Future
CHAPTER 4:
Jobs: Worsening Underemployment
Unemployment and Underemployment
The Early 1990s Recession
Job Creation: How Fast in the 1980s?
The Growth in Part-Timers
More Than One Job
The Contingent Workforce
Self-Employment
CHAPTER 5:
Wealth: Losses for Most, Gains for Few
Aggregate Household Wealth: Financial Assets Boomed,
Tangibles Failed to Grow
Wealth Inequality Exceeds Income Gap
Growing Wealth Inequality
Soaring Debt
Increased Racial Wealth Gaps
CHAPTER 6:
Poverty: H i g h Rates o f Poverty Unresponsive t o
E c o n o m i c Expansion
Background: The Failure of Economic Growth
Measuring the Extent of Poverty
Alternative Measurements of Poverty
The Poor Get Poorer
Poverty Spells. The Length of Time Spent in Poverty
153
159
161
165
169
173
181
187
195
201
203
207
209
211
215
217
219
225
229
237
243
249
251
253
255
257
266
266
271
273
275
277
281
281
The Poverty Status of Different Demographic Groups
Family Structure and Poverty
Decomposing the Role of Demographics
The Poverty of Female-Headed Families
The Changing Effects of Taxes and Transfers
Government Benefits and Family Structure
Explaining the High Poverty Rates of the 1980s
Falling Wages and Poverty in the 1980s
Vferk Efibrt
CHAPTER 7:
Regional Analysis: The Differences in Growth by Region
and Residence
Income Inequality
Poverty
Wages and Taxes
Employment
Urban versus Rural
CHAPTERS:
Losing Access to Basic Necessities: Education, Housing,
Health Care, and Child Care
287
289
291
293
299
302
307
307
309
315
317
331
333
343
345
357
Education
357
Educational Proficiency: Low Scores for Minorities and the Poor
359
Access to College: Higher Costs, Less Aid
367
Minority Education: Uneven Results
371
Minorities and Advanced Degrees
373
Educational Expenditures: Less Investment than Other Countries
and Greater Inequality Within the U S
373
Housing
379
Housing Costs.- Rising Faster than Incomes
383
Home Ownership Rates: Falling for Young Families as a Result of
Rising Costs
383
Low-Income Housing: Reduction in Housing Stock and Deteriorating
Conditions
385
Homelessness
389
Health Care: Higher Costs, Poor Health Outcomes, and Many Uninsured ...
393
U.S. Health Care Costs: Highest Among the OECD Countries
393
Are We Getting Our Money's Worth?: Lagging Health Indicators
399
Health Insurance: Incomplete and Uneven Coverage
401
Child Care: A Constraint on Working Parents
409
�Child Care Arrangements
The Costs of Child Care
CHAPTER 9:
International Comparisons: The United States is
Falling Behind
Incomes and Productivity: Sluggish Growth in the U S
Wages: Falling for Production Workers
Poverty: Highest and Most Persistent in the U S
Comparing Economic Mobility Between Countries
Job Growth and the Public Commitment to Job Training
and Unemployment Compensation
413
415
419
419
425
431
439
443
Table Notes
451
Figure Notes
475
Appendix
479
Bibliography
483
The
State of
Working
America
�Executive Summary
Using a wide variety of data on family incomes, taxes,
wages, unemployment, wealth and poverty, The State o f
Working America: I W J - l W . i closely examines the impact
of the economy on the living standards of the American
people. The story we tell is one of great disparities.
The economic recovery of the 1980s was unusual in that
the vast majority of Americans were, in many ways, worse
off at the end of the recovery in 1989 than they were at the
end of the 1970s. In 1989, for instance, many families were
working longer for less money, many fell deeper into poverty, and many lacked access to basic necessities such as
health care and housing. Meanwhile, a small group of
American families at the top of the income scale bcncfitied
extensively from the recovery. As this edition goes to press,
the economy is somewhere between a protracted recession and an anemic recovery, wherein the economic problems of the 1980s have been exacerbated in the early
1990s. At this point in time, even the wealthy and highly
educated are experiencing the income.deterioration which
has plagued the majority of Americans throughout the
1980s. The following is a summary of the economic realities that characterize the state of working America.
The economic
recovery of the
1980s was
unusual in that
the vast majority
of Americans
were, in many
ways, worse off
the end of the
recovery in 1989
than they were at
the end of the
1970s.
�Family Incomes: Slow and Unequal
Growth
Between 1979 and 1989, family income grew more
slowly than in any other business cycle since World War II.
In 1989, the median family's income was only $1,528
greater than it was ten years earlier. In the 1967-1973
periotl for example, family income rose that amount every
The slow rate of22 months (18 years) From 1989 to 1991, as the ecoincome growthnomic downturn began, the average family's inflationbetween 1979 adjusted income fell by 4.4%, a $1,640 drop.
and 1989 did not The slow rate of income growth between 1979 and 1989
affect all familiesdid not affect all families equally. The incomes of families at
the
62.9%
equally. Thefromtop of the income scale—the richest 1%—grewincome
1980 to 1989, capturing 53.2% of the total
incomes of growth among all families. The bottom 60% of families
families at the actually experienced a decline in income. Such uneven
top of the income
growth led to a substantial increase in the wealthiest famiscale—the lies' share of total income (their income share grew by 4.4
richest 1%—grewpercentage points between 1980 and 1989, to 51.8%) and
62.9% from 1980 equally dramatic losses for everyone else, reversing the
to 1989. The entire post war trend toward lessening inequality.
bottom 60% of What income growth there was over the 1979-1989
families actuallyperiod was driven primarily by increased work hours at
experienced a lower wages. In fact, the only family type to experience
positive income growth was married-couple families,
decline in income.
whose incomes grew due to the increased earnings of
wives. Among married couples with children, the average
wife was employed 32.3% more hours in 1989 than in
1979, the equivalent of nearly seven weeks of full-time
work.
In addition, families headed by someone aged 25 to 34
in 1989 had incomes $1,715 less than their counterparts
did 10 years earlier in 1979. There is strong evidence that
most families headed by someone bom after 1945 will not
achieve the same incomes in middle-age as achieved by
preceding generations.
Finally, the economic downturn beginning in 1989 saw
growth rates turn negative as the typical family experienced an income loss of 4.4% between 1989 and 1991.
The character of this recessionary decline was unusual in
that the incomes of the wealthiest families fell the most:
between 1989 and 1991, the inflation adjusted income of
the richest 5% of families fell by 9 3%, twice the income
loss of middle class families. I'art of this income decline
reflects adverse trends that developed in the laic 1980s,
preceding the onset of the recession, such as the decline
in wages for white-collar and college-educated worken and
in higher income regions such as the Northeast
Taxes: A Further Cause of Worsening
Inequality
Our tax system has not ameliorated the growing income
inequalities examined in Chapter 1. In fact, recent tax
changes have worsened the distribution of after-tax income
by taxing the middle class and the poor more heavily whilegiving large tax cuts to the richest 1% of taxpayers. The
average federal tax burden on the wealthiest families fell by
4 9 percentage points between 1980 and 1989, an average
decrease of $27,430. Over the same peritKl, the federal tax
burden on the bottom fifth of families grew by 1.2 percentage points, an addition of $100 to their federal tax bill.
These trends contributed to a 26.0% increase in the aftertax income of the average family in the top fifth, compared
to a 1.7% increase for the average family in the bottom
fifth.
Overall, taxes have increased little and the total U S tax
burden (federal, state, and local) remains one of the light
est among the Western industrialized countries While the
U.S. tax burden gradually edged up during the postwar
period, it has held relatively constant since 1973, staying
between 31% and 32% of Gross Domestic Product (GDP).
However, while the federal burden grew by only 1.4% of
GDP since 1959, the state and local burden grew by 4.5%
of GDP Since state and local taxes are more regressive than
federal taxes, the effect of this shift in the tax burden is to
widen inequality.
Wages: Working Longer for Less
Declining wage trends are arguably the primary determinant of the slow growth of family income anil the greater
income inequality we have experienced in the 1980s.
Between 1979 and 1989, the hourly wages of 80% of the
Recent tax
changes have
worsened the
distribution of
after-tax income
by taxing the
middle class and
the poor more
heavily while
giving large tax
cuts to the richest
1% of taxpayers.
�workforce declined, with the wage of the typical (median)
worker falling 4 9% between 1979 and 1989. However,
wage loss varied dramatically by education and gender.
Toward the end of the recovery, weakening demand for
white-collar workers began to place downward pressure
on the wages of workers who had previously escaped wage
erosion. White-collar workers lost 2.1% of their total compensation (wages and benefits) between 1987 and 1992;
the hourly wage of college-educated workers fell by 3.1%,
In 1991, the 1987-1991.
wages paid The most severe wage reductions, however, have been
to young male for entry level jobs for young, high school graduates, a
and female group comprising two-thirds to three-fourths of all young
high school workers. In 1991, the wages paid to young male and
graduates were,female high school graduates were, respectively, 26.5% and
respectively, 15.4%, less than the wages their counterparts received in
26.5% and 1979. Over the 1973-1991 period, the wages of entry level
15.4%, less thancollege graduates have also fallen by 9.8%. Thus, young
workers face a job market providing far lower wages than
the wages theirwhat comparably educated young people faced two
counterparts decades ago.
received in The wages of women grew faster (or fell more slowly)
1979. than those of men over the 1980s. Between 1979 and
1989, the average (median) male hourly wage fell by
11.7% while the average female wage grew by 5.3%. These
trends led to a narrowing of the gender wage gap, mostly
due to the fall in male wages (72%). However, the recession reversed the growth of the female wage, and between
1989 and 1991, the average female wage fell by 2.8%, so
that in 1991, it was only 1.1% higher than in 1973.
There has also been an expansion of jobs that, even with
full-time work effort, do not pay enough to lift a family of
four out of poverty. In 1989, 28% of the workforce earned
poverty level wages, up from 25.1% in 1979 and 21.4% in
1973. By 1991, during the recession, 312% of the workforce was employed at poverty level wages.
A significant cause of this widespread wage decline is a
shift to lower paying industries. The overall shift in the
composition of employment, between 1980-1989, led to a
3.4% decline in total compensation. Other causes of wage
decline include a lower minimum wage, increased import
competition, technological change, and fewer and weaker
unions. All of the evidence suggests continued wage erosion in the future
Jobs: Worsening Underemployment
Over the course of the 1980s recovery, a rising number
of workers experienced labor market distress, due to: unor underemployment (involuntary pan-timers and discouraged workers), low pay, or overemployment (multiple job
holders, working more than one job to compensate for
falling incomes). The unemployment rate in 1989 was a
relatively moderate 5.3%, but the total underemployment
rate was a much higher 9.8%.
The recession of the early 1990s differed from prior post
war recessions in several important ways (based on the
assumption that the recession ended in the second quaner
of 1992). The recent recession lasted at least 22 months,
twice as long as the 11-month average length of earlier
downturns, but unemployment rose by far less than in the
recessions of the mid-1970s and mid-1980s. Nevertheless,
the recent recession involved roughly the same amount of
permanent (versus temporary) job loss—a rise of 1.6%—as
in recent, heavier downturns. The recent recession was
also uncharacteristic in that it prompted an unusually large
rise in white-collar unemployment coupled with historically slow growth in white-collar employment.
As noted above, there has been a growth in underemployment due to more people working part-time who
want full-time jobs. The share of part-time jobs increased
from 16.6% in 1973 to 18.1% in 1989, an increase almost
entirely due to the increased rate at which workers arcworking pan-time involuntarily. By 1991, over one-fourth
of the part-time workforce was comprised of involuntary
part-timers. A related phenomenon is the increase in the
use of temporary workers, as the share of the workforce
employed in the personnel services industry (consisting
primarily of workers hired through or working for temporary agencies) doubled between 1982 and 1989. These
trends reflect the increased use of 'temps" and pan-timers
by employers and not the preferences of the workforce for
shorter hours.
The recent
recession
was also
uncharacteristic
in that it prompted
an unusually
large rise in
white-collar
unemployment
coupled with
historically slow
growth in
white-collar
employment.
�Wealth: Losses for Most, Gains for Few
Stagnant incomes and falling wages tell only pan of the
story of rising inequality. The distribution of wealth also
plays an important role in determining a family's living
standards.
Wealth is even more concentrated at the top than
income: in 1989, the top 1% of families earned 14.1% of
total income, yet owned 38.3% of total net worth and
The wealth 50.3% of net financial assets. The wealth distribution has
holdings of thealso become more unequal over time. The wealth holdings
richest 0.5% of of the richest 0.5% of families grew by one percentage
families grew by point over the entire 21-year period, 1962-1983, but grew
one percentageby four times as much in just six years between 1983 and
point over the 1989. Meanwhile, the bottom 60% of families had lower
entire Zl year wealth holdings in 1989 than 1983.
period, 1962- The main cause of greater wealth inequality was the
1983, but grew growth of financial assets—the stocks and bonds owned
at an annual rate of
by four times as primarily by the wealthy—which grewgrowth rate over the
2.9%, 1979-1989, about six times the
much in just six previous business cycle, 1973-1979. Tangible assets—
years between homes, cars and appliances—did not grow at all,
1983 and 1989. 1979-1989, in marked contrast to the annual growth of
Meanwhile, the tangible assets (between 4.2% and 4.3%, annually)
bottom 60% of throughout the 1950s, 1960s and 1970s. Since tangible
families had assets are spread out more evenly than financial assets,
lower wealth their stagnation mainly affected the bulk of the population
holdings in who are not wealthy.
The concentration of financial assets at the top implies
1989 than
that American businesses are owned and financed primar1983.
ily by the richest families. In 1989, for example, the
wealthiest top 0.5% owned 37.4% of all corporate stocks,
while the bottom 90% owned only 16.1%. The top 0.5%
also owned 64% of bonds, while the bottom 90% owned
only 6.0%.
The gap in average wealth between minority and white
households is even larger than the gap in income, with the
average wealth in non-white families being just 19% of that
in white families in 1989. There was, however, significant
progress in lessening racial wealth inequality in the 1962
to 1983 period as non-white mean wealth grew from 13%
to 24% of while mean wealth. In contrast, the racial wealth
gap widened between 1983 and 1989.
Poverty: High Poverty Rates
Unresponsive to Economic Expansion
Despite the growing economy between 1983 and 1989,
poverty rates were high by historic standards, averaging
13.6% over the 1980s. In fact, those in poverty in 1989
were significantly poorer than the poor in 1979 for example, 8% more poor persons had incomes at 50% of the
poverty line in 1989 than in 1979. Furthermore, these high
rates are not sensitive to different measurement choices;
poverty was a significant problem over the decade, given
any accepted definition.
Minorities and children were especially vulnerable to the
high poverty rates. The poverty rates of blacks have liccn at
least three times that of whites since 1979, reaching 32.7%
in 1991. The Hispanic rate has climbed from 21 9% in
1973 to 28 7% in 1991 Child poverty grew in all racial categories between 1979 and 1991 By 1991, more than one
out of every five children was poor. For children under sixyears-old, the rate was even higher, reaching 24 .(>% in
1991. Perhaps the most alarming statistic in the chapter
refers to young black children: in 1991 more than half of
black children (51.7%) under six were poor
The conventional explanations for high poverty rates—
the formation of mother-only families, the failure of the
poor to exploit labor market opportunities—do not hold
up under close scrutiny. The growth of families headed by
a female slowed over the 1980s compared to the previous
decade, and that growth contributed less to the rising poverty rates of the 1980s than in other periods. Regarding
work effort, the percentage of the ptxjr who worked grew
over the decade, and those who worked increased their
hours.
The reasons poverty rates remained high despite the
recovery have to do with wage decline and the failure of
the "safety net," i.e., the government's system of taxes and
transfers designed to ameliorate poverty Over the 1980s,
the already low wages of low-income workers fell 15.9%
for male and 6.8% for female workers in the bottom 20%
of the earnings' distribution.
The failure of the safety net led to an additional 11% of
the population on the poverty rolls, 1979-1989, including
an additional 8 2% of persons in single-headed families
with children.
Despite the
growing economy
between 1983
and 1989, poverty
rates were high
by historic
standards,
averaging 13.6%
over the 1980s.
The reasons
poverty rates
remained high
despite the
recovery have to
do with wage
decline and the
failure of the
"safety net."
�Regional Analysis: The Differences in
Growth by Region and Residence
Kconomic growth was quite uneven on a regional basis
over the 1980s. I he Northeast region, comprising New
England and the Mid-Atlantic divisions, did comparatively
well in terms of greater income and wage growth, and
lower unemployment. Conversely, the 1980s economy in
the Midwest and the Western regions was characterized by
While average
wages for the relatively flat or negative income growth over the 1980s,
rising poverty, and significant wage loss. The South expericollege educated
enced somewhat more mixed results over the 1980s, but
rose only slightly
of
for children)
on a national did sustain high ratesand poverty (particularlyinequality.
and unemployment,
a surge in income
basis from 1979 This pattern of uneven growth has left winners and
to 1989 (1.8%), loseis in its wake. In 28 states, the average income of fiamregional analysis in the lowest fifth fell over the 1980s, with dramatic
ilies
reveals that thislosses in the Midwest and South. On the other end of the
positive growthincome spectrum, Northeastern families in the top fifth
was almost experienced dramatic income gains over the 1980s.
exclusively Although the Northeast experienced the most growth
driven by the between 1979 and 1989, it was "hit" the hardest during the
Northeastern recent recession. Between 1989 and 1991, the median
region. family in the Northeast saw a dramatic 7.2% drop in its
income, the largest regional loss. Median family income
also fell in the Midwest (3 3%), and fell steeply in the
South and West (4.7% and 5.2%, respectively).
While average wages for the college educated rose only
slightly on a national basis from 1979 to 1989 (1.8%),
regional analysis reveals that this positive growth was
almost exclusively driven by the Northeastern region,
where the wages of college-educated workers grew by
5.9%. The only other region with positive growth for this
group was the West, where wages grew less than half as
fast.
The average hourly wages of college-educated men in
the Northeast grew by 4.6%, 1979-1989, and fell by 3 6%,
1989-1991, leaving it practically flat (0.9% growth) over the
hill 1979-1991 period. College-educated workers also saw
significant wage loss in the other regions during the recession, again denoting the end of the 1980s "white-collar"
txxmi and the white-collar nature of this recession. In the
South and West, even workers with post-graduate educa-
tions saw a slight decline overall between 1989 antl 1991
Shifting from regions to residence (uriian versus rural),
we find evidence of a downward shift in the rural incomedistribution over the 1980s, as 5.0% of rural families shifted from the middle- and upper-income class to the lowerincome class. Kural areas also experienced higher unemployment than urban areas. However, despite this rural
shift, the highest increase in poverty by residence was in
cities. By 1989, 2.4% more city than rural residents were
poor.
Losing Access to the Basic Necessities:
Education, Housing, Health Care, and
Child Care
Another dimension of the well-being of American families involves access to those basic necessities that have a
vital impact on our living standards.
Education. Americans place a high value on equal
access to quality education, yet we are far from providing
it. The United States devotes fewer public resources to
public education than most other industrialized countries,
and our expenditures are unequally distributed, contributing to wide differences in the test scores of children from
different economic backgrounds Access to higher education, an important factor in lifetime earnings, is notably
restricted for students from minority and low-income families and increasingly inaccessible overall due to rising
costs. For example, the expected family contribution to private college tuition was over three and a half limes the
average income of low-income families in 1987.
Housing: The cost of owning or renting a home rose
quickly in the early 1980s, both in real terms antl as a percentage of income. In the period between 1974 and 1979,
the typical first-time home buyer made a housing expenditure that was 29 9% of his or her income; in the period
between 1980 and 1989, that percentage was 37.2%.
Though home ownership costs fell in the latter pan of the
decade, the proportion of families owning their own
homes also fell over the decade, with young, large, antl
low-income families facing the most ditiiculty in the housing market
Access to
higher education,
an important
factor in lifetime
earnings, is
notably restricted
for students from
minority and
low-income
families and
increasingly
inaccessible
overall due to
rising costs.
�The high proponion of income that many families had
to expend on housing meant that many families' budgets
were severely constrained after paying housing costs. We
find that in 1989, one-third of families are "shelter poor,"
meaning that these families did not have enough to pay for
a minimum amount of non-housing necessities, as defined
by the U.S. Bureau of l^bor Statistics. Housing problems
became particularly severe for low-income households. In
1970, there were 200,000 more low-income units than
Even a full-timelow-income households; by 1989, there were 4.1 million
job, does not more low-income households than low-income units. One
guarantee the result has been the highly visible problem of
provision of homelessness.
health insurance. Health Care. Americans spend more on health care
In 1990 almost than any other nation in the world, over J650 billion in
one-half of the 1990. Yet we still lag behind most other countries, accorduninsured wereing to important health indicators. Our health care system
is
connected to the plagued by inflated costs and the serious problem of
full-time/lull-yearuneven health insurance coverage. Over 35 million Americans were without health insurance in 1991, a number that
workforce. has grown steadily since 1987.
In 1989, we spent 11.8% of our Gross Domestic Product
(GDP) on health care, well above the industrialized
nations' average of 7.6%. Hospital costs, the most expensive component of our health care bill, rose 355.7% on a
per capita basis, 1960-1990. The average American family
now pays close to 12% of its income on health care, up
from 9.0% in 1980.
Despite our high expenditure level, the U.S. ranks
increasingly poorly on basic health indicators. The U.S. was
situated 19 out of 23 countries in 1989 on infant mortality
rates. Regarding life expectancy, the U.S. position was 17
for females and 19 for males in 1989 (again, out of 23).
Approximately 34 6 million Americans, 13 9% of the
population, were without health insurance in 1990. In
addition, having a job, even a full-time job, does not guarantee the provision of health insurance. In fact, in 1990
almost one-half of the uninsured were connected to the
lull-time/full-year workforce, and a minority of persons in
each year were non-workers. By 1990, 49.3% of the uninsured were full-year, lull-time workers or their dependents.
One factor contributing to the high rates of those who are
employed and uninsured is the rapid job growth in industries that are less likely to provide their employees with
10
health insurance, primarily low-paying jobs in the service
sector.
Child Care. By 1991, two-thirds of women with children had entered the labor force, creating the need for an
essential corollary to mothers' workforce participation:
child care. I j c k of affordable, high-quality child care is
found to be a significant constraint, keeping mothers
either out of the workforce or working for fewer hours
than they would prefer.
Most families with children under five (57%) were paying for child care in 1990, and wealthy families were paying a smaller percentage of their income than middle- antl
low-income families. In fact, low-income families spent as
much as one-quarter of their income for child care, significantly constraining their family budgets, while iamilies
with incomes over $50,000 spent on average, only 6 2% of
their income on child care.
International Comparisons: The
United States is Falling Behind
While Americans still have the highest per capita income
as measured by relative purchasing power, the U.S. is falling behind on many important economic indicators Our
prtKluctivity growth rate was 67% of the average of the
industrialized countries on an annual basis, 1979-1989.
Our wage growth was stagnant while other countries saw
strong wage growth. U.S. production workers (non-managerial and non-supervisory) in manufacturing saw their
hourly compensation fall at a rate of 0.6% per year,
1979-1989, while hourly compensation generally grew in
other advanced countries.
The inequality in income antl wealth is more evident in
the U.S. than in other countries. Data from the mill- 1980s
show the U.S. to have the largest low-income class, the
smallest middle class, and the third largest high-income
class in a study with eight other comparable countries.
However, the relatively large size of America's lowincome class is not driven by relative differences in market
outcomes. In fact, prior to taxes and transfers, poverty
(measured as 40% of median family income, adjusted lor
family size) is higher in most other countries. The American rale, 19.9%. was lower than ihe rate lor France. <ier-
VJhile Americans
still have the
highest per
capita income
as measured
by relative
purchasing
power, the U.S.
is falling
behind
on many
important
economic
indicators.
�many, Sweden, and others. Yet after taxes and the transfer
of government benefits, American poverty, at 13.3%, is well
above all the other countries. Furthermore, over the
course of the 1980s, as all countries generated higher poverty rates, the American safety net became less effective at
reducing poverty, while other countries' system of taxes
and transfers expanded to ameliorate the increased rates of
market-generated poverty. Finally, U.S. poverty is the most
persistent. In a three-year study tracking families in povAmerican poverty, 14.4% of U.S. families were poor for the duration of
erty,
at 13.3%, is well the study, a higher percentage than any other comparable
above all the country (the German rate was 1.5%, the French rate was
other countries.
1.6%).
The American One area where U.S. economic performance has been
safety net became
highly touted is job growth. It is true that from 1979 to
less effective at1989, millions more jobs were created in the U.S. than in
other
reducing poverty, countries. However, the U.S. should be expected to
while other have the largest absolute number of jobs, as it is by far the
largest country with the greatest population growth. A betcountries' system
of taxes and ter indicator is the rate of job creation. U.S. employment in
1989 was only 18.7% greater
transfers increases of 26.5% in Australiathan in 1979, compared to
and 20.1% in Canada over
expanded to the same period. Furthermore, the U.S. is among the smallameliorate the est spenders on job training and placement, increasing the
increased rates of
likelihood that other countries will continue to out pace us
market-generated productivity and wage growth.
on
poverty.
12
Introduction
The State of Working America presents a comprehensive
statistical portrait of recent trends in the various factors
that make up the standard of living of America's working
families. The data show that the economy is failing most
Americans and not simply because of the early 1990s recession: we are in the midst of a long-term erosion of incomes
and opportunities.
Current income problems can be pictured as having two
phases. The first phase emerged in the early 1980s and
was concentrated in the wage deterioration experienced
by the 75% of the workforce lacking a college degree. Family income problems essentially flowed from this wage erosion. By 1987, prior to the onset of the recent recession, a
second phase of income problems emerged. As the whitecollar boom of the 1980s ended, the wages of white-collar
and college-educated workers fell behind inflation and
women's wages and labor force participation stalled or fell.
The peculiarly white-collar nature of the early 1990s recession has contributed to low consumer confidence and the
generalized economic fears that the younger generation
will not fare as well as their predecessors. As we show in
Chapter 1 there is fire beneath that smoke.
Unfortunately, it has taken a lengthy and painful recession for most Americans to recognize both the shortcomings of the unbalanced growth of the 1980s and the fundamental problems that are limiting their ability to improve
their standard of living. The economy is not now generating higher real wages for either blue- or white-collar work13
The economy is
failing most
Americans, and
not simply
because of the
early 1990s
recession: we are
in the midst of a
long-term erosion
of incomes and
opportunities.
�ers or for either college-educated workers or those without
a college degree. On our present trajectory, litde relief is in
sight. In the short-term, the anemic recovery from the early
1990s recession will not reestablish the wage growth necessary to achieve significant improvements in income. Our
analysis of long-term employment trends suggest that,
without major policy changes, the 1990s will be a period
of historically slow growth in skilled, high-wage jobs
(Tables 3 51 and 3 52, meaning the 51st and 52nd tables in
second Chapter 3).
In this
phase of
economic The End of the White-Collar Boom
hardship many of
the labor force In our view, a second phase of income problems
groups that hademerged in the late 1980s, around 1987, prior to the onset
previously of the recent recession. In this second phase of economic
enjoyed wage hardship many of the labor force groups that had previgrowth also ously enjoyed wage growth (white-collar, college-educated
experienced anand women workers) also experienced an erosion of
wages. These new income problems reflect a broadening
erosion of wages.
of income difficulties and not the substitution of some new
for some older problems.
The defining development in the second phase is the
post-1987 wage setbacks for white-collar and college-educated workers. Between 1987 and 1991, for instance, the
real wages of college-educated workers fell 3.1%, with a
4.9% fall among male college graduates (Table 3-19 and
3-20). There was a similar fall in white-collar hourly wages
and compensation (wages and benefits) fell, respectively,
3 5% and 2.1% from 1987 to 1992 (Table 3.4). This second
phase of income problems is the consequence of the end
of the 1980s white-collar employment boom. Starting in
1987, economic difficulties in a wide variety of white-collar
intensive industries such as retailing, finance, real estate,
insurance and banking led to a slackened growth in whitecollar jobs and the consequent cutback in wages. "Lean
and mean" strategies or "restructuring" in manufacturing,
and to a lesser degree in the service sector, have also led to
diminished white-collar employment opportunities and
wages.
This contraction in white-collar opportunities was evident in the early 1990s recession, when white-collar
employment grew much slower than in earlier recessions.
Moreover, there was a greater rise in white-collar than
14
blue-collar unemployment in the recent recession (Table
4.6) . In the prior five recessions the rise in white-collar
unemployment was at most 53% of the rise in blue-collar
unemployment. Further, the income declines in the recession have been most severe at the top of the income scale,
an unusual pattern that reflects the upward expansion of
wage and income problems.
There was also an economic reversal for women in this
period, as the 2.8% wage decline from 1989 to 1991 (Table
3.7) wiped out half of the gains of the prior ten years. The
wage reversal among women combined with a leveling off
of growth of women's labor force participation meant that
one of the main engines of family income growth in the
1980s—increased women's employment and earnings—
had stalled in the early 1990s.
The First Phase Problems
The first phase income problems, which began in the
early 1980s and are ongoing, consist of the wage reductions experienced by blue-collar workers and workers
without a college degree (particularly men and younger
workers), growing poverty and the "squeeze" on middleclass incomes. The defining development in the first phase
was that incomes fell for the bottom 60% of families, while
there was an extraordinary 63% income growth among the
best-off 1% of families (Table 19). This same disparity can
be seen in wealth trends, as the wealth (i.e., ownership of
assets) of the upper 1% of families grew 54% over the
recovery (1983-1989), while the net worth of the bottom
60% of families actually declined (Table 5.8).
During the first phase, only married-couple families
achieved income growth, while fiamilies headed by single
parents, male or female, fell. Moreover, the income growth
among married couples was almost entirely fueled by the
increased employment and hours of the wives in these
families. For instance, the average wife in a married-couple
family with children worked 32.3%, or 268 hours, more in
1989 than in 1979 (Table 1.24). Without this increased
work effort the incomes of the bottom 80% of married-couple families with children would have fallen (the bottom
60%) or not risen (the upper-middle) from 1979 to 1989
(Table 1.26).
15
The defining
development in
the first phase
was that incomes
tell for the bottom
60% of families,
while there was
an extraordinary
63% income
growth among the
best-off 1% of
families.
�These first phase problems were characterized by the
dramatic erosion of wages among noncollege-educated
men, particularly those in their twenties or thirties. The
group of noncollege-educated male workers who suffered
serious wage declines in this period comprised threefourths of male employment (Table 3.20). For instance, the
real wages of male high school graduates fell 12.7% from
1979 to 1989. Men with "some college" had a somewhat
lesser wage decline (8.3%) from 1979 to 1989 while male
The group of non- school dropouts suffered much more (18.2%). In conhigh
college-educated
trast, men with a college degree achieved a 2.8% wage gain
male workers who
from 1979 to 1987. The greatest wage gains among men
suffered seriouswere the almost 10% growth for those with at least two
wage declines inyears of schooling beyond college.
this period Wage erosion was most severe for younger workers, particularly those without a college degree. A young male
comprised threefourths of male (female) high school graduate earned 22.4% (13.5%) less
employment. in 1989 than in 1979 (Table 3 23).
Falling Wages
The cumulative effect of these two phases of income
problems has been stagnant or falling incomes in the middle class, rising poverty, and rapidly expanding incomes
and wealth for only the very richest families. In our view,
these income developments stem from the failed performance of the economy as reflected in the erosion of wages
among nearly all groups and the deterioration in the types
of jobs available. As we will discuss below, wage and job
trends have been driving both the slow growth in incomes
and the surge in income inequality. We discuss other factors that have led to greater income inequality below.
The extent of the labor market's failure to provide
decent earnings and adequate employment even at the
peak of the business cycle is illustrated in Table A, which
shows the number of workers and the percent of the labor
force that was affected by various types of "labor market
distress" in 1989. Overall, there were more than 30 million
workers, nearly a third (315%) of the workforce, experiencing at least one of these types of labor market distress in
1989 (i.e., eliminating any double counting). Thus, at the
end of the 1980s recovery and in a year of low unemployment, 1989, the economy was failing to provide adequate
16
wages and employment to a large segment of the
workforce.
For instance, some 10 3% and 5.8% of the workforce
experienced at least four weeks, respectively, of unemployment and involuntary part-time work (had only pan-time
work even though wanting full-time work) in 1989.
Another 1.8% of the workforce were discouraged workers,
people who wanted to work but felt that no jobs were
available. Additionally, a sizeable 22.2% of the prime-age
workforce earned poverty level hourly wages, a wage that
was insufficient for a year-round, full-time worker to support a four-person family beyond the poverty standard. We
view this as a conservative estimate of the number of people being failed by the poor performance of the labor market and the economy. This analysis is limited to prime-age
workers (ages 25 to 64 years old) in a year of low unemployment. Moreover, this measure of labor market distress
excludes a sizeable number of workers affected by less easily measurable types of distress: temporary or contract
workers wanting a permanent position; people who are
"self-employed" for lack of a decent job opportunity (Table
4.24); the nearly 3.0% of the workforce that work at least
two-jobs because pay on the first job was insufficient to
meet expenses (Table 4.16); workers experiencing job-
There were more
than 30 million
workers, nearly a
third (31.5%) of
the workforce,
experiencing
labor market
distress in 1989.
TABLE A
N m e and Share o Workforce Experiencing
u br
f
Labor Market Distress in 1989
Type of Labor Market Distress*
Earn Poverty Level Wages
Discouraged Workers
Involuntary Part-time at Least
Four Weeks
Unemployed at Least Four Weeks
Total with Labor Market Distress**
Number
(000)
Share of
Labor Force
21,205
1,711
22.2%
1.8
5,543
9,799
30,143
5.8
10.3
31.5%
"Based on a sample of civilian wage and salary workers ages 25-64 who worked, were unemployed or
were a discouraged worker in 1989.
* 'Counts each person once even if subject to more than one type of distress.
17
�related illnesses, injuries or death; workers who are not
covered by any employer-provided (and partially paid)
health insurance from their own or even a spouse's
employer (Chapters 3 and 8); and, workers affected by
wage and/or benefit reductions (Chapter 3). A more inclusive measure would clearly show that the labor market is
failing to provide adequately paid, safe and steady employment for a much broader group of the workforce.
Unfortunately, data limitations prevent us from consistUnderemployment
ently estimating "labor market distress" in earlier or later
grew from 1979 years. It is certain, however, that much more of the work1989 and was force was adversely affected by labor market trends in
raised further in 1991—the average number of unemployed, involuntary
the early 1990s part-time, and discouraged workers rose 24% from 1989 to
recession. The1991 (Table 4.1) and the share of the workforce earning
largest labor poverty level wages rose 11% at the same time (Table 3 9).
market failure Underemployment grew from 1979 to 1989 and was
has been the raised further in the early 1990s recession (Chapter 4).
deterioriation However, the largest labor market failure has been the
of wages. deterioriation of wages, as we discussed earlier (and extensively in Chapter 3). There are a number of factors driving
these changes in the wage structure. One was the quickened shift of employment from well-paying to low-paying
employment: a trend where most of the new jobs were created in the lowest-paying segments of the service sector.
This industrial job shift particularly affected the level of
benefits paid to the average worker. The forces driving this
adverse job shift were a rising trade deficit and slow
growth in service sector productivity (especially relative to
manufacturing).
International trade has directly, and adversely, affected
earnings as workers whose jobs were displaced by imports
or by a fall in exports were forced to accept jobs at lower
wages. But there have also been indirect, or spillover,
effects of trade that have had an even larger effect. The loss
of good jobs because of higher imports and fewer exports
has forced trade-displaced workers and young workers
(who might have obtained a good job in an earlier period)
to compete against similarly skilled workers for a more limited set of jobs, thus lowering wages for the entire workforce without a college degree. Trade pressures have also
affected wages by limiting nominal wage growth or forcing
wage concessions on workers who remained in tradesensitive industries.
18
The weakening of unions and the erosion of union
membership has also adversely affected the wages of the
union and nonunion noncollege-educated workforce
(Tables 3.34 to 3 38). The erosion of the real value of the
minimum wage since 1979 has meant lower wages for at
least 11% of the workforce (Table 3 41) and has perhaps
undercut the wages of many more. The minimum wage is
just one of the many elements of the "social wage" whose
weakening has depressed wages: overtime standards,
unemployment insurance, workers' compensation, and
anti-discrimination protections.
Technological change has also diminished the demand
for noncollege-educated workers and rewarded, in the
form of wage premiums, those using new technologies.
Finally, as discussed earlier, the economic problems in various white-collar intensive industries—finance, retail trade,
banking, real estate—and employment "restructuring" in
manufacturing and in services has diminished white-collar
employment growth, thereby leading to lower wages for
white-collar and college-educated workers.
The most common explanations of the poor wage performance for most workers are "slow productivity growth"
and that technological change and international competition are placing an increased "premium" on education and
skills. However, although slow productivity growth can
explain slow wage growth, it cannot explain falling wages.
Since 1979 there has been positive productivity growth at
the same time that wages and compensation have dramatically f a l l e n among noncollege-educated workers
(1979-1991) and among college-educated workers
(1987-1992). Only factors which affect the wage structure
(lifting one group's wage and/or depressing another's) or
cause average wages and compensation to fall can explain
recent wage trends.
Similarly, attributing higher wage inequality solely to a
rising "education wage premium" is misleading. The question is why the "education premium," the ratio of the
wages of college graduates to those of high school graduates, has been increasing.
It is clear that a higher education premium is not the
result of a benign "bidding up" of the wages of "more educated" workers. Rather, the greater wage gap between high
school- and college-educated workers is almost entirely the
result of a dramatic push downward of the wages of the
19
Although slow
productivity
growth can
explain slow
wage growth,
it cannot
explain
falling wages.
�noncollege-educated
workforce, the
result of
deunionization, industry shifts, an eroded minimum wage,
trade-induced wage reductions and other identifiable
forces. Moreover, over the last twelve years the wages of
college graduates (male and female) have hardly been bid
upwards, growing only 0.3% from 1979 to 1991, and actually falling in the South and the Midwest.
It is also commonly asserted that wages are growing
slower because fringe benefits are rising (or total compen// is also sation is growing even if wages are not). This is not supcommonly ported by the evidence. Over the 1977 to 1989 period the
asserted that value of fringe benefits fell, although not as much as wages
wages are (Table 3.2), and total compensation (wages and benefits)
growing slowerfell 7.9%.
because fringe Nor are adverse demographic trends—a rise in the numbenefits are ber of college graduates or the entry of the baby-boomrising. This is ers—reducing wages. In the late 1980s the size of the colnot supportedlege-educated workforce grew more slowly yet the wages
by die evidence.of college graduates fell. The baby-boomers have entered
the time in their careers when wages advance quickly and
the baby-busters' short supply should be raising the relative wage of young workers, yet the opposite has
happened.
Income Inequality
Income growth in the 1980s was modest by historical
standards and even slower than the slow growth of the
1970s. Equally important, lower- and middle-income families either lost ground or struggled to stay even by working
harder through longer hours and more family members
working. In contrast, the top 1% of iamilies had an extraordinary after-tax income growth of 75% from 1980 to 1989.
One way of characterizing the income trends of the 1980s
is that income growth for upper-income groups returned
to the rates obtained in the more prosperous period from
World War II to 1973, while the remainder of the population had stagnant or falling incomes. The result was a
surge in inequality in the 1980s that reversed three
decades of declining income inequality.
The rapid income growth of the top 1% is the result of
both fast wage growth and the rapid growth in capital
incomes from greater interest and dividend income and
from more realized capital gains (their greater unrealized
20
capital gains are reflected in the rapid growth of the net
worth o f the top 1%; see Chapter 5). The upper 1%
received more capital income partly because there was an
expansion of capital incomes relative to labor incomes and
partly because capital incomes became more unequally
distributed over the 1980s (see Chapter 1). The expansion
of capital incomes is primarily due to the high real interest
rates that prevailed in the 1980s. The reduced incomes at
the bottom of the income scale reflect the erosion of wages
as well as the reductions in government assistance, or
transfer payments (see Chapter 6).
In our view, the growth in inequality has been generated
by changes in the market economy and by government
policies and not in any great measure, by changing "demographic trends": the smaller size of iamilies, more femaleheaded households and more two-earner families. Analyses which take into account the smaller size of families (an
indicator of how much income is "needed") show an even
slower income growth in the 1980s relative to the 1970s
(Table 1.2). Moreover, family size-adjusted income data
show a greater polarization of income in the 1980s (Table
1.21). Any shift in the population to different family situations is at best a minimal explanation since there was a
large growth in income inequality among each type of family, be it single parent or married couples (Table 1.8).
Moreover, the pattern of increased wives' earnings prevented income declines among middle-class married couples and has meant less inequality among married-couple
families (Tables 1.23 to 1.29). So, high-earning men marrying high-eaming women did not cause greater inequality.
However, wives working has contributed slighdy to overall
family income inequality since this engine of income
growth among married couples was not available to single
persons or single-parent iamilies.
Arguments which explain income or poverty trends by
the rise of female-headed households should be greeted
with skepticism (Tables 6.13 and 6.14). At the broadest
level, we wonder why it is a "demographic" phenomena
when a woman cannot support a family without access to a
man's income. Is this not primarily a reflection of the lesser
earnings opportunities available to women and inadequate
access to child care?
It is also important to understand the demographic composition of female-headed iamilies because it is too fre21
The growth in
inequality has
been generated
by changes in
the market
economy
and by
government
policies and
not in any
great measure,
by changing
"demographic
trends.
�quently assumed that the growth of female-headed families
is an uiban, black phenomena of women with out-of-wedlock children. In fact, less than a fourth of female-headed
families are headed by "never-married" women; the vast
majority of female-family heads are divorced, separated or
widowed (Table 6.17). Moreover, this is not primarily a
black phenomena, especially in the 1980s. White and black
female-headed families accounted for, respectively, 59.1%
and 33 8% of the growth of such famUies in the 1979-1989
Economic forces period. It was in the 1970s and not in the 1980s that there
led to higher was a major increase in black female-headed families.
poverty in the
The deterioration of incomes and the surge in income
1980s, a reversal inequality in the 1980s could not have been driven by
of the trend in the more female-headed fiamilies since the growth of femaleprior two decades headed families was actually slower in the 1980s than in
when economic the 1970s (Table 1.5).
growth was a It is true that any disproportionate growth of femaleforce reducing headed families does, given their higher poverty rates, raise
poverty. the aggregate poverty rate. Nevertheless, the growth of
female-headed families has been occurring for a long time
and, as we have seen, has even slowed down. The growth
in poverty, or the failure of poverty to fall in the 1980s
recovery, is primarily due to the failure of economic
growth in the 1980s to ameliorate poverty as it had in the
prior twenty years (witness the higher poverty among married-couple families). In fact, economic forces led to higher
poverty in the 1980s, a reversal of the trend in the prior
two decades when economic growth was a force reducing
poverty (Table 6.14). Finally, it is worth noting that the
growth of poverty among female-headed families can
mosdy be attributed to an erosion of government support
and not to any reduced work effort (Tkble 6.19)The recent shifts in the tax structure were not responsible for the middle-class squeeze. Still, the large reductions
in federal taxes were an important factor in the rapid
growth of after-tax incomes among the top 1% of families:
lower taxes allowed the top 1% to obtain a 75% gain in
after-tax incomes on a 62.9% gain in pre-tax income
(Tables 1.9 and 2.3). The remaining 99% of families paid
only slightly higher federal taxes in 1989 than in 1980
(Table 2.6). Thus, changes in tax policy had a minimal
effect on middle-class incomes, but a sizeable effect on the
incomes of the veryrichestfamilies. The slow and unequal
income growth of the 1980s was primarily the product of
22
market outcomes as reflected in the pre-tax incomes of
families and was not due to changes in tax policy.
What Has Greater Inequality Bought
Us?
The recent surge in inequality of both incomes and
wages might be viewed positively had greater inequality
led to faster income or wage growth for most families or
income groups. Clearly it has not. Neither productivity, nor
investment, nor employment growth, nor wage growth,
nor our international competitiveness was significantly better (and most fared worse) in the 1980s than the 1970s.
For instance, employment growth over the 1979-1989
period or in the 1982-1989 recovery was not above average for the postwar period. The 1980s did have a long
recovery (the second longest one of the postwar period),
but it has been followed by a lengthy recession on the
heels of several years of minimal growth. By now it is clear
that the sources of consumption growth in the 1980s
(debt, more family members working, longer hours of
work) were unsustainable. It is also true that the groups
with the greatest wage problems (noncollege-educated
workers) also faced an erosion of employment prospects;
for them, lower wages did not "buy" more jobs (Table
3-50).
Conclusion
Policy discussions have increasingly, and usefully, characterized our options as whether we will follow a "highwage" or a "low-wage" growth path. The panorama of indicators of economic performance and economic well-being
that we present in this book suggest that we have already
taken the "low-wage path" without even becoming more
competitive. No one in government or business has explicidy announced a program of lowering American wages and
working conditions in order to become "competitive";
nevertheless, the cumulative impact of both government
policy and business strategies has achieved a lowering of
wages and the standard of living of most Americans.
Recent wage problems sometimes are dismissed as the
problems of "unskilled" or "less-educated" workers not fit23
The recent surge
In inequality of
both incomes and
wages might be
viewed positively
had greater
inequality led to
faster income or
wage growth for
most families or
income groups.
Clearly it has not.
�ting into a world of new technologies and the "global
marketplace." The problem, however, is not that a small
group of workers are undergoing a painful adjustment to a
new, more beneficial economic order. What has occurred is
the lowering of wages, benefits and working conditions of
the three-fourths of the workforce without a college
degree and the consequent pressure on family incomes.
Moreover, now that income problems have spread
upwards to the white-collar and college-educated groups,
The problem is it is harder to find beneficiaries of the new order.
t that a small
It is obvious that the economy is not generating higher
group of workersincomes for the vast majority of Americans. 1 herefore, the
are undergoing afundamental economic problem we face is to generate adepainful quate income growth for the majority, based on jobs payadjustment to a ing high hourly wages and benefits. Government policy
and elected leaders
to
new, more change the economic must beofjudged on their ability the
course
the country and leave
beneficial low-wage path for the higher-wage one.
economic order.
What has
occurred is the
lowering of
wages, benefits
and working
conditions of the
three-fourths of
the workforce
without a college
ee. Income
problems have
spread upwards
to the whitecollar and
college-educated
groups.
Presentation and
Methodology
Presentation
In this book we present a comprehensive portrait of
changes in incomes, taxes, wages, employment, wealth,
poverty and other indicators of economic performance and
economic well-being. We rely almost exclusively on data in
the tables and figures to describe the who, what, why, when
and where of income changes in the post-war period Consequently, the documentation of our analysis is essentially
the documentation of the tables and figures in the hook.
This allows us to omit distracting fixitnotes or citations in
the text or tables. All of the documentation is contained in
the "Table Notes" section on pages 451-473 which provides the sources of the data we use and information about
the calculations we made with the data In some selective
circumstances, however, we incorporate data in the discussion that is not in a table or figure.
Time Periods
Kconomic indicators fluctuate considerably with shortterm swings in the business cycle. For example, incomes
tend to fall in recessions and rise during expansions. Therefore, economists usually compare business cycle peaks with
peaks and troughs with troughs, so as not to mix apples and
oranges. In this hook, we examine changes Ix'twccn business cycle peaks The initial year lor most tables is 19-17 or,
when using the National Income and Product Accounts,
24
25
�1959 (the earliest year available). The intermediate years in
the analysis are 1967, 1973, 1979, and 1989, all of which
were business cycle peaks, at least in terms of having low
unemployment. We also present the latest year for which
data are available (usually 1990 or 1991) to show the effect
of the early 1990s downturn. Some information was available only for other non-peak years. If this information was
imponant enough, we included it.
Growth Rates and Rounding
Since business cycles differ in length, we usually present
the annual growth rates in each period rather than the total
growth We present compound (log) annual growth rates
rather than simple annual rates. Compound annual growth
rates are just like compound interest on a bank loan: the
rate is compounded continuously rather than yearly.
While annual growth rates may seem small, over time
they can amount to a large change. For example, the
median incomes of families headed by persons aged under
25 fell 2.5% per year between 1979 and 1989 (Table 1.3).
Over the full period, incomes declined by a considerable
21.8%.
In presenting the data we "round" the numbers, usually
to one decimal place. However, we use "unrounded" data
to compute growth rates or percentage shares and so on.
Therefore, it is not always possible to exactly replicate our
calculations by using the data in the table. In some circumstances, this leads to an appearance of errors in the tables.
For instance, we frequently present shares of the population
(or families) at different points of time (i.e., an early and a
later year) and compute changes in these shares. Because
our computations are based on the "unrounded" data the
change in shares presented in a table does not always
match the difference in the actual shares presented in the
table. Although our change in shares may appear to be in
error because of rounding, in fact, our computations based
on the "unrounded" data are more precise.
Adjusting for Inflation
In most popular discussions, the Consumer Price Index
for All Urban Consumers (CPI-U). often called simply "the
Consumer Price Index," is usetl to adjust dollar values for
26
inflation. However, some analysts hold that the CPI-U overstated inflation in the late 1970s and early 1980s by measuring housing costs inappropriately. The methodology for the
CPI-U from 1983 onward was revised to address these
objections. Not all agree that it should have been revised.
We chose not to use the CPI-U so as to avoid any impression
that this report overstates the decline in wages and understates the growth in family incomes over the last few
decades.
Instead of the CPI-U, we adjusted dollar values for inflation using the CPI-U-X1 index, an index which uses the
new methodology for housing inflation over the entire
1967-1991 period. The CP1-U-X1 however is based on small
sample, experimental indices for the 1970s antl there is
some slight variation in methods over the entire period
Nevertheless, the use of the CPI-U-X1 is becoming standard
and we employ it. Because the CPI-U-X 1 is not available for
years before 1967 we extrapolate the CPI-U-X I back to earlier years based on inflation as measured by the CPI-U.
The original data on which we draw were calculated
with a variety of price indices. Whenever possible, we readjusted the figures using the CPI-U-X 1. In our analysis of poverty in Chapter 6, however, we used the CPI-U rather than
the CP1-U-X1. This is because Chapter 6 is based
almost entirely on publications of the Census Bureau that
use the CPI-U. Moreover, the net effect of all of the criticisms of the measurement of poverty is that current methods understate poverty. Simply switching to the CPI-U-X 1
without incorporating other revisions (i.e., revising the
actual poverty standard) would lead to an even greater
understatement and would be a very selective intervention
to improve poverty measurement. A fuller discussion of
these issues appears in Chapter 6.
Household Heads
We often categorize families by the age or the race/
ethnic group of the "household head." This is the person in
whose name the home is owned or rented. If the home is
owned jointly be a married couple, either spouse may be
tlesignated the household head, livery family has a singlehousehold head
27
�Hispanics
Unless wc specify otherwise, we follow the Census
Bureau's designation of Hispanic persons. That is, Hispanics
are included in racial counts (e g , with blacks and whites),
as well as in their own separate category For instance, in
government analysis a white person of Hispanic origin is
included both in counts of whites and in counts of Hispanics. In some tables (primarily in Chapter 3 on wages) we
remove Hispanic persons from other racial (white or black)
categories. Using this technique, the person described
above would appear only in counts of Hispanics. When this
technique is applied, it is noted in the table.
1H
Acknowledgements
The preparation of this publication required the intensive work cttort of many people on the KPl staff. Jessica Burton, Miranda Martin and Stephanie Scott diligently
word processed the text and tables. Daphne Clones prepared all of the graphs and
provided research assistance, especially on the regional analysis. l.ory Camba provided computer assistance, including analyses ol wage data on computer tapes. Kdie
Rasell provided advice on health and education matters Max Sawicky carefully
reviewed our tax analysis and Robert Blecker anil lixld Schaler provided comments
on international comparisons. Bill Spriggs provided advice on a wide variety of areas,
but especially on the minimum wage and our cohort income growth analyses. Ruth
Polk reviewed every word we wrote and effectively edited, checked, made consistent
and substantially improved the text and the presentation. Carol Pott guided the production and design of the book as well as assisting with editing Typesetting by MidAtlantic/Type 2000. Nan Gibson and Roger Hickey worked to provide a large audience for the book.
This third version of The State o f Working America obviously owes much to the
efforts, ideas and spreadsheets of the previous co-authors, David Frankel and
Jacqueline Simon.
A number of experts were helpful in providing data for our use. Kd Wolff was especially kind in providing his analysis of wealth data. Others include: Frank Sammartino of the Congressional Budget Office; Wendell Primus of the House of Ways and
Means Committee; Isaac Shapiro of the Center on Budget and Policy Priorities; Albert
Schwenk, Bill Gullickson anil Joseph Meisenheimerof the Bureau of l^ibor Statistics;
Kevin Murphy; Rebecca Blank; Alan Krueger; Robert Moffitt; Kugene Steuerle;
Michael Stone; Ix-e R.Jones of the National Assessment of Fducational Progress and
Ijee Price of the Joint Kconomic Committee. We also drew heavily on the published
works of Sheldon Danziger, Richard Freeman, l-arry Kau, Timothy Smeeding, Bob
Costrell and Chris Tilly. Our intention is to thank, not implicate, any of these
researchers.
29
�Chapter 1_
Family Income: Slow and
Unequal Growth
Introduction
A person's economic well-being depends upon his/her
access lo economic resources. Since most Americans live in
families in which income is shared, family income is the
best single measure of how Americans are doing economically. This chapter examines how the performance of the
economy in generating jobs, wages and income from capital assets has affected the growth of family income and the
recent surge in income inequality.
During the recession of the early 1990s, from 1989 to
1991, there was a -i -t%, or $ 1,6-iO, fall in the median family's
inllation-adjusted income. The character of the recessionary income decline was unusual in that the incomes of the
wealthier families fell the most. Part of the income decline
during the recession reflects adverse trends that developed
in the late 1980s preceding the onset of the recession, (actors such as the decline in wages for white-collar and college-educated workers and in higher-income regions such
as the Northeast.
The post-1989 income decline, however, comes on the
heels of a period from 1979 to 1989 (both years of relatively low unemployment) when family income grew even
more slowly than it did in the 1970s anil far more slowly
than it did in the period between the end of World War II
and 1973 The income growth slowdown in the 1979-1989
perioil tlitl not aliect all families equally. Upper-income
groups, particularly the upper 1%, experienced significant
31
The income
growth slowdown
in the 1979-1989
period did not
affect all families
equally.
Upper-income
groups,
particularly the
upper 1%,
experienced
significant income
growth while the
bottom 60% ol
families actually
experienced a
decline in income
�income growth while the hot loin 60% ol families actually
cx|XTicnccil a decline in income This resulted in a dramatic rise in the income gap between high- and lowincome families, reversing the entire postwar progress in
lessening inet|uuliiy. The incomes of the top 1% grew
62.9% from 1980 to 1989, capturing '>3.2% of the total
income growth among all Iamilies
l-Yom 1979 to 1989 the income of every type of family
fell except for married-couple families Among married
Those familiescouples with children, however, there would have been no
headed by income growth had there not been an increase in the
someone bornwives' earnings anil hours of work. In fact, the bottom 80%
after 1945 may of married-couple families with children would have had
not achieve the stagnant or falling incomes from 1979 to 1989 without the
same incomes inincreased earnings of wives. Income growth over the
middle age as the1979-1989 period was driven primarily by more work at
preceding lower hourly wages.
A major
inequality was
generation accelerationfactor fueling the growingin the 1980s duethe
of capital income growth
to
achieved. high real interest rates and the stix-k market boom, both of
which primarily benefitted the richest families. In contrast,
hourly wages and fringe benefits, which provide support
for most families, grew slower than inflation. Inequality
among wage-earners also rose as real wages among highincome groups nose and wages among the broad middle
class anil lower income groups fell.
The pattern of increases in wives' hours and earnings
lessened income inequality among married couples, but
slightly increased inequality among all families.
Income growth has not only been slow, there has also
been a decline in income for younger families. A cohort, or
inter-generational, analysis of income growth shows that
the recent groups of young families have started out at
lower incomes and obtained slower income gains as they
approach middle age. Consequently, those families headed
by someone born after 1945 may not achieve the same
incomes in middle age as the preceding generation
achieved.
In the first few sections of this chapter we examine the
changes in the level of income in recent years relative to
other periods since 1947. This analysis focuses on changes
in median family income (the income which is more than
but also less than that of hall the families) for families overall as well as for families dillerentiaied by the age or race/
32
ethnicity of the household head and by family iy|X- (married couples, single parents and so on) In the final
sections of this chapter we turn our attention lo the
growth of inequality anil its causes
Sluggish Income Growth
Income growth in recent years has been slow and the
gains unequally distributed. Tables 1.1 and 1.2 show
changes in family income, adjusted tor changes in consumer prices, in various cyclical peak (or low unemployment) years since World War II. As explained in the introductory section on presentation and methodology,
examining income changes from business cycle peak to
business cycle peak eliminates the distortion caused by the
fact that incomes fall drastically in a recession and then
recover in the upswing (Figure 1A).
Income growth in
recent years has
been slow and the
gains unequally
distributed.
TABLE 11
Median Family Income, 1947-1991*
(1991 dollars)
Year
1947
1967
1973
1979
1989
1991
Median
Family
Income
$17,059
29,765
34,774
36,051
37,579
35,939
Total Increases:
1947-67
1967-73
1973-79
1979-89
1989-91
$12,706
5,009
1,276
1,528
1,640
•Income includes all wage and salary, sell employmenl. pension, interest, rent, government cash assist
ance and olhei money incomes
33
�TABLE 1.2
Annual Growth of Median Family Income, 1947-1991
(1991 Dollars)
Median Family
Income (SizeAdjusted)*
Median Family
Income Growth
Period
Percent
Dollars
Percent
1947-67
1967-73
1973-79
1979-89
1989-91
2.8
2.6
0.6
0.4
-2.2
$635
835
213
153
-820
n.a.
n.a.
1.3
0.6
-3.1
'Family income adjusted for changes in the size of families 1989-91 data is based on 1989-90
income change.
FIGURE 1A
Median Family Income,
1967-1991
40,000
1989 Peak
38.000
1979 Peak
36,000
-
1973 Peiik
(A
is
"5
Q
34,000
32.000
30.000
28.000 t _ I
1
1 1
—
_l
34
I
I
I
I
I
l_
J
I
L
J
I
I
I
L
There was a substantial increase in family income in the
t w o decades (1947-1967) immediately following World
War I I , w h e n the median family income increased by
$12,706 for an annual rate of growth o f 2 8% (Table 1.2).
Family incomes continued to grow into the early 1970s,
but since 1973 have risen very slowly. In 1989, the median
family's income was $1,528 greater than it was in 1979
This translates into a growth of just $ 153, or 0.4%, per year
from 1979 to 1989, a rate o f growth just two-thirds the
sluggish 0.6% annual growth of the 1973-1979 period and
only one-seventh the rate o f income growth of the postwar
years prior to 1973 (see Figure I B ) . In fact, the $1,528
income growth over the 10 years after 1979 equals the
amount that incomes rose every twenty-tux) munibs (1.8
years) in the 1967-1973 period.
It is c o m m o n practice to also examine measures of family income growth which adjust for changes in family size,
o n the basis that if the same total family income is shared
by fewer family members then the economic well-being of
each family member has improved. Another advantage of
family size-adjusted income measures is that they incorporate income trends o f individuals not living in families in
addition to income trends among families. Because the
average family size has been falling over the last several
decades, family income growth is being shared w i t h fewer
family members and the change in economic well-being of
family members exceeds the growth portrayed by a simpleanalysis o f income that ignores changes in family size.
However, trends in incomes "adjusted for family size"
can be misleading, since the recent decline in the average
family's size (or growth of single-person households) is
partially due to lower incomes, i.e., some families feel they
cannot afford as many children (or individuals feel they
cannot get m a r r i e d as early) as they c o u l d have, had
incomes continued to rise at postwar rates. Yet a family
deciding to have fewer children o r a person putting off
'starting a family because incomes are d o w n appears "belter o f f " in size-adjusted family income measures. It also
seems selective to adjust family incomes for changes in
family size—yielding a greater growth rale of economic
well-being in recent years—and not adjust family income
trends for other demographic trends that are lowering economic well-lx-ing such as more hours of work and the loss
of leisure.
35
The $1,528
income growth
over the 10 years
after 1979 equals
the amount that
incomes rose
every twenty-two
months (1.8
years) in the
1967-1973
period.
�FIGURE I B
Annual Growth of Median Family Income,
1947-1990
1947-1967
1967-1973
1973-1979
1979-1989
1989-1990
TABLE 1.3
Growth of Median Family Income By Age of Householder, 1967-1990
(1991 Dollars)
Year
Under 25
25-34
By Age of Household Head
35-44
45-54
55-64
Young Families Hurt Most
Over 65
Median Family Income
1967
1973
1979
1989
1990
$21,927
23,125
23,956
18,743
16,902
$30,373 $34,665
35,222 41,284
35,626 42,257
33,911
44,157
32,822 42,789
$36,305
43,927
46,660
50,637
49,150
$30,174
36,881
40,457
41,347
40,678
$14,738
18,543
20,817
25,354
26,103
Annual Growth Rate
1967-73
1973-79
1979-89
1989-90
1.0%
0.9
-2.5
-10.3
2.6%
0.5
-0.5
-3.3
36
3.0%
0.7
0.4
-3.1
3.3%
1.3
0.8
-3.0
3.5%
1.8
0.2
-1.6
Nevertheless, even when income growth is adjusted for
the shift towards smaller families (as in Table 1.2), the
income growth from 1979 to 1989 is only slighdy more
than an "unadjusted" measure (0.6% versus 0.4%). It is
noteworthy that the size-adjusted family income measure
shows a far greater slowdown in income growth between
the 1979-1989 period and the 1973-1979 period (0.6%
versus 1.3%), reflecting the slower decline in family size in
the 1980's than in the 1970s.
The recent downturn starting in 1989 has significandy
reduced incomes. For instance, the 2% fall from 1989 to
1990 meant a loss of $738 in income for the median family,
a reversal of half the income growth over the entire
1979-1989 period (Table 1.1). The 1989-1990 income
decline appears to reflect more than a normal business
cycle downturn. For example, this is a large income fall for
a period where unemployment actually grew only slighdy,
up 0.3% from 5.3% to 5.6%. In addition, the large income
decline reflects several ongoing and new structural shifts in
income growth, such as the fall off in wages among whitecollar and college-educated workers that preceded the
recession (see Chapter 3), the continuing reductions in
blue-collar wages, and a slowdown in labor force growth
(see Chapter 4).
3.9%
2.2
2.0
2.9
Table 1.3 shows that the largest drop in incomes has
occurred among the youngest families. The average
income of families headed by someone under age 25
declined at an annual rate of 2.5% from 1979 to 1989.
Theirs is also the only age group for which median family
income is lower today than it was in 1967; such young
families in 1989 had $3,184 less income to spend in real
dollars than their 1967 counterparts had when they were
starting out.
Families headed by someone between the ages of 25 and
34 years have also fared poorly relative to earlier years. The
incomes of this type of family eroded 0.5% per year from
1979 to 1989. This is in stark contrast to the six years
between 1967 and 1973, when income for this group
increased at a 2.6% annual rate, or even to the six years
between 1973 and 1979 when income grew at a 0.5%
pace. Many families in this age group are likely to be bring-
37
The largest drop
in incomes has
occurred among
the youngest
families. The
incomes of
families headed
by someone
between the ages
of 25 and 34 fell
0.5% annually
from 1979 to
1989.
�TABLE 1.4
Growth of Median Family Income By Race/Ethnic Group, 1947-1991
(1991 Dollars)
Race or Hispanic Origin
of Household Head
Year
White
Black
Ratio to White
Family Income of:
Hispanic
Black
Hispanic
Median Family Income
1947
1967
1973
1979
1989
1991
$17,768
30,895
36,344
37,619
39,514
37,783
n.a.
$18,291
20,975
21,302
22,197
21,548
n.a.
n.a.
$25,148
26,079
25,753
23,895
Annual Growth Rate
1947-67
1967-73
1973-79
1979-89
1989-91
2.8%
2.7
0.6
0.5
-2.2
2.3%
2.3
0.3
0.4
-1.5
n.a.
n.a.
0.6%
-0.1
-3.7
n.a.
59.2%
57.7
56.6
56.2
57.0
n.a.
n.a.
69.2%
69.3
65.2
63.2
up young children and try ing to buy a home ol their
o w n . I he income problems ol these youiiK Iamilies thus
represent income problems lor the nation's children. Families w i t h household heads aged 25 to 3H in 1989 had
incomes $1,715 less than their counterparts did 10 years
earlier in 1979. The deterioration o l incomes lor these
young iamilies is one o l the most signilicanl income developments over the last t w o decades.
The incomes o f the 35 to M ami 45 to 5 i age groups
have grown modestly—roughly one-half o f \% per year—
since 1979 In contrast, family income for these age groups
grew by more than 3.0% per year between 1967 and 1973
Incomes o f families headed by someone over 65 increased
at a 2% pace from 1979 l o 1989, a healthy growth, but only
half as fasi as that o f the 1967-1973 period
From 1989 to 1991 there was signilicanl slippage in
income among nearly every age group, w i t h Iamilies
headed by someone under age i 5 losing at least 5 % and
the youngest families losing 10 2 %
I n c o m e G r o w t h Among Racial/Ethnic
Groups
Sluggish income growth has affected all racial groups, as
Table 1.4 illustrates White families, w h o fared the l>csi
from 1947 to 1973, experienced o n average very modest
0.4% annual growth in real income from 1979 to 1989
Black families, w i t h a median income more than 40%
lower than that o f whiles, experienced slightly slower
income growth than whites in the 1973-1979 peritKl (0 3%
versus 0 6%) a n d had slow i n c o m e g r o w t h i n the
1979-1989 period comparable to that o f w hites (0 i% versus 0.5%). The median income o f families of Hispanic origin declined 0 1% |>cr year between 1979 and 1989
The ratio b e t w e e n w h i l e antl black m e d i a n family
income fell from 1967 l o 1973 but has remained fairly constant since 197.3, r e f l e c t i n g c o m p a r a b l y slow incomegrow i l l among both groups The decline in I lispanic family
incomes, however, has let I to a larger gap bclwecn white
antl Hispanic families. The ratio o f Hispanic l o w h i l e
median income fell l i o m 69 3% in 1979 to just 65.2% in
1989.
Sluggish income
growth has
affected all racial
groups.
�There were si/cablc income losses lor every racial/
ethnic cutegory during the l )H ) I W I downlum That
white family income fell more than black family income
in a recession is unusual and lurihcr reinforces that
this recession has a different character than prior
recessions.
TABLE 1.5
Income Growth b Type ol Family, 1947-1991
y
l
Married Couples
Year
Wife
in Paid
Labor
Force
Total
Wife Not
in Paid
Labor
Force
No Spouse Present
MaleFemaleHeaded Headed
All
Families
Median Family Income
1947
1967
1973
1979
1989
1990-91
$17,498
31,671
37,594
39,441
42,340
40,995
n.a.
$37,356
43,968
45,758
49,720
48,169
n.a.
$28,557
32,948
32,589
31,575
30,075
$16,524
25,567
30,997
30,936
30,587
28,351
$12,224
16,111
16,728
18,185
18,060
16,692
$17,059
29,765
34,774
36,051
37,579
35,939
Annual Growth Rate
1947-67
1967-73
1973-79
1979-89
1989-91
n.a.
2.7%
0.7
0.8
-1.6
n.a.
2.4%
-0.2
-0.3
-2.4
n.a.
3.2%
0.0
-0.1
-3.8
n.a.
0.6%
1.4
-0.1
-3.9
2.8%
2.6
0.6
0.4
-2.2
19.9%
31.8
33.5
40.6
45.7
3.0%
2.9
0.8
0.7
-1.6
67.1%
55.1
49.7
41.9
33.5
3.0%
2.4
2.6
2.9
4.4
10.0%
10.7
12.4
14.6
16.5
100.0%
100.0
100.0
100.0
100.0
Distribution of Families
1951*
1967
1973
1979
1989
87.0%
86.9
85.0
82.5
79.2
* Earliest year available.
i()
l
Only Married Couples Gain
I h e only type of family which experienced incomegrowth from 1979 to 1989 was married couples with a w ilein the paid labor force. Incomes among married couples
where the wife was not in the labor force and among single-parent families, whether headed by a man or a woman,
declined from 1979 to 1989. This was an especially dramatic turnaround for female-headed families since their
incomes grew 1.4% per year in the 1973-1979 period, but
fell thereafter.
This pattern of income growth suggests that it was only
among families with two adult earners that incomes grew
in the 1979 to 1989 period The data in Table 1.5 also
show a sizeable growth in the importance of working
wives. In 1979, there were more married couples without
a wife in the labor force than those with a wife in the labor
force (41.9% of all families versus 40.6% of all families). Uy
1989, married couples with two earners (assuming the
husband worked) comprised 45.7% of all families, while
one-earner married couples were proportionately lewer in
number, just 33.5% of the total. This shift towards twoearner families has been a major factor in recent incomegrowth.
Married-couple families, although still predominant—
representing 79 2% of the total in 1989—comprise a
smaller share of families than they did in the 1950s and
1960s. There has been a continuing rise in the importance
of female-headed families who, in 1989, represented
16 5% of the total. Although this phenomenon has been
the focus of increased attention in recent years, there was a
much faster growth of female-headed families in the period
from 1967 to 1979 than in the peritKl since 1979 (Chapter
6 provides a further analysis of this issue)
The downturn from 1989 1991 caused income losses
lor every type of family. However, incomes fell most among
•i I
This pattern of
income growth
suggests that it
was only among
families with two
adult earners that
incomes grew in
the 1979 to 1989
period.
�single-parent families and lor married couples with the wirenot in the labor force.
TABLE 16
Shares o Family Income Going to Various Fifths,
f
and to T p 5%, 1947-1991
o
Growing Inequality of Family Income
Income Share Going lo:
Breakdown o) Top Filth
Year
1947
1967
1973
1979
1969
1991
Lowest
Fifth
Second
Fifth
50%
55
55
52
46
45
119%
124
11.9
11.6
10.6
107
Middle
Fifth
17.0%
17.9
17.5
17.5
16.5
166
Fourth
Fifth
Top
Fifth
Total
23.1%
23.9
24.0
24 1
23.7
24 1
43 0%
404
41 1
41.7
446
44.2
100 0%
100.0
1000
100 0
100.0
1000
-0.4
0.4
29
-0 4
Top 5%
Next 15%
17 5%
15.2
15.5
158
17.9
17.1
25.5%
25.2
256
259
26.7
27.0
Point Change:
1979-89
1989-91
0.6
-0.1
- 1.0
01
•i 2
-1.0
01
0
0
2.1
-0.8
0.8
0.4
The vast majority of American families have experienced
either very modest income growth or an actual erosion in'
their standard of living in recent years I he small minorily
of upper-income families, however, had substantial income
growth. The result has been an increase in inei|ualily. The
rich have gotten richer; the ptx>r are more numerous and
are poorer than they have been in decades. This section
examines the income trends of families at different incomelevels and the dramatic growth of income inequality in the
1980s.
.Table 1.6 presents the percentage distribution of
income among the various fifths (or "quintiles") of the
population, and the top 5%. For an analysis of incomefifths the incomes of families are ordered from lowest to
highest and the 20% of families with the lowest incomes
are considered the "lowest fifth," the next best-off 20% of
families are said to be the "second fifth" and so forth until
the remaining fifth, the "top fifth," are the 20% of families
with the highest incomes. Because the size of families differ, the bottom 20% of families does not necessarily
include 20% of the population living in families. This is
why some analysts categorize fifths by the number of people so that, for example, the lowest fifth contains the lowest income families comprising 20% of the population living in families.
The upper 20% received 44.6% of all income in 1989.
The upper 5% received more of total income, 17.9%, than
the families in the bottom 40%, who received just 15.2% in
1989. As we will see in a later chapter providing international comparisons (Chapter 9), income in the U S is distributed far more unequally than in other industrialized
countries.
Income distribution has grown even more unequal since
1979, reversing the trend towards less income inequality
over the postwar period into the 1970s. Since 1979, the
iS
Income
distribution has
grown more
unequal since
1979, reversing
the trend towards
less income
inequality over
the postwar
period into the
1970s.
�FIGURE 1C
Ratio of Family Income of Top 5% to Lowest 20%,
1947-1989
1947
1967
1973
1979
1989
TABLE 1.7
Real Family Income Growth b Fifth, 1947-1991
y
(1991 Dollars)
Year
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
1947
1967
1973
1979
1989
1991
$ 4,989
9,107
10,746
10,766
10,359
9,734
$11,875
20,606
23,451
23,750
24,185
23,105
$16,964
29,619
34,456
35,871
37,571
35,851
$23,051
39,723
47,090
49,395
54,055
51,997
Top
Fifth
Top 5%
$ 42,908 $ 69,850
70,141 110,974
80,794 121,931
85,589 129,585
101,780 163,042
95,530 147,817
Annual Change
1947-67
1967-73
1973-79
1979-89
1989-91
3.0%
2.8
0.0
0.4
-3 1
2.8%
2.2
0.2
0.2
-2.3
2.8%
2.5
0.7
0.5
-2.3
2.7%
2.8
0.8
09
- 1.9
2.5%
2.4
1.0
1.7
-3.2
2.3%
16
1.0
2.3
-4.9
Ixmom 80% has lost income share, and only the lop ill"',
gained Moreover, the 1989 income share ol the upper
fifth, -M.6%, was far greater than the share it received during the entire postwar period and even higher than the
-B% received in 1947. liven among the rich, the growth in
income was skewed to the very top, as the highest 5% saw
its income share rise by 2.1 percentage points (from IS.8%
to 17.9%) between 1979 and 1989, which makes up the
bulk of the 2.9 percentage point total rise in the income
share of the upper fifth. At the other end of the income
spectrum, the 1989 share of total income in each of the
three lowest income fifths (i.e, the bottom 60% ol families)
was even smaller than in 1947.
The increase in the income gap between upper- and
lower-income groups is illustrated in Figure 1C, which
shows the ratio of the incomes of the top 5% to the
incomes of the bottom 20% from 1947 to 1989 As Figure
IC shows, the gap between the top and the bottom
incomes fell from 1947 to 1979, but grew to a historic ratio
of 15.7 by 1989, reversing three decades of lessening
inequality.
Another way of viewing this recent surge in income inequality is to compare the income growth of families by
income fifth, or quintile, as in Table 1.7. Over the early
postwar period from 1947 to 1973 there was strong
income growth across the income spectrum. From 1947 to
1967, for instance, income growth ranged from the 2.5%
annual pace obtained by the top fifth to the 3 0% annual
pace obtained by the lowest fifth Because incomes grew
more quickly for lower- and middle-income families than,
upper-income families (the top fifth or 5%) from 1947
through 1973, there was a general decline in income
inequality.
The pattern of income growih since 1973 was far more
uneven and far slower than in the earlier period. From
1973 to 1979 the fastest income growth was the 1% annual
growth among the top fifth antl top 5% of Iamilies which,
modest as it was, far exceeded the 0 2% annual income
growth among the second fifth antl the absence of growth
among the families w ith the lowest incomes Incomes continuetl lo grow slowly in the 1979-1989 pcritxl, but the
pattern of growth was even more unequal. The Iamilies
with the lowest incomes actually lost ground from 1979 to
1989 (incomes fell 0 i % annually) while al the same time
IS
Even among the
rich, the growth in
income was
skewed to the
very top, as the
highest 5% saw
its income share
rise by 2.1
percentage points
between 1979
and 1989, the
bulk of the 2.9
percentage point
total rise in the
income share of
the upper fifth.
�TABLE 1.8
Income Growth By Fifth and Family Type, 1979-1989
Families With Children
Fifth
All
Families
Highest
Fourth
Middle
Second
Lowest
13.9%
4.4
0.8
-1.3
-2.1
Childless Families
Married
Couples
All
12.3%
5.1
-0.8
-7.6
-17.7
Single
Mothers
NonElderly
Elderly
15.2%
9.2
5.3
1.3
-3.6
7.5%
-2.4
- 10.1
-17.1
-23.2
18.5%
13.0
8.1
4.0
-1.2
29.5%
20.7
18.4
15.4
13.7
TABLE 1.9
Income Growth Among Top Fifth and By Fifth, 1977-1990
(1992 Dollars)
Average Family Income in:
Income Group
1977
1980
1989
Percent Change
1977-89
1980-89
All
$ 40,065
$ 39,433
$ 43,495
8.6%
10.3%
Top Fifth:
Top 1%
Next 4%
Next 5%
Next 10%
$ 87,268
314,526
107,945
76,525
60,073
$ 89,031
343,610
109,551
75,876
60,428
$109,424
559,795
132,036
87,711
65,900
25.4%
78.0
22.3
14.6
9.7
22.9%
62.9
20.5
15.6
9.1
Bottom Four
Fifths:
Fourth
Middle
Second
Lowest
$ 27,789
$ 46,772
34,505
22,333
9,368
$ 26,886
$ 45,827
32,948
21,009
8,791
$ 27,251
$ 47,913
32,681
20,140
8,391
H6
- 1.9%
2.4%
-5.3
-9.8
- 10.4
1.4%
4.6%
-0.8
-4.1
-4.6
the top *)% o l families were able lo obtain the same 2 3%
rate o f income growth as they had in ihe earlier 19-17-1967
peritKl.
This uneven pallern of income growth from 1973 to
1989 left the bcsl-off families with llieir greatest share o l
income in the entire postwar periotl ( Iable 1 6 ) antl the
bottom 60% o f families with their lowest income shares
In contrast to ihe pallern over ihe 1970s antl 1980s, the
d o w n t u r n from 1989 to 1991 affected the highest in
comes the most The top 5 % lost 9 . 3 % while the l o p f i l t h
lost 6 1 % The income losses from 1989 to 1991 among
the broad class—the middle three t|uiniiles—were hall as
large as those o f the upper 5 % . Nevertheless, by 1991 ihe
incomes o f the b o t t o m 6 0 % o f families were less than
they were in 1979.
Inequality has widened among each type o f family in
much the same pattern as has occurred overall—incomedeclines for families at the very bottom antl rapitl incomegrowth for the best-off families (Table 1.8). For instance,
the incomes o f the bottom fifth declined for each family
type, except elderly families. The greatest income decline
was among the bottom 40% o f single-mother families, a
group whose income decline ranged from 17.1% to 23 2%
from 1979 l o 1989. In fact, the incomes o f the bottom
four-fifths o f single-mother families have all fallen in recent
years. O n the other end, elderly families in each income
fifth obtained significant income g r o w t h , although the
highest income elderly had more than double the income
growth o f the lowest fifth o f elderly families (29.5% versus
13 7%). This pattern o f income growth among the elderly
reflects growing social security benefits at all income levels.
Table 1.9 and Figure I D examine changes among various income groups using a more comprehensive pre-tax
measure o f income (it includes capital gains) The data also
provide finer detail o n upper-income groups, but unfortunaiely are only available for 1977 antl 1980—not for the
last business cycle peak in 1979. lable 1 9 shows the bott o m 60% losing income over the I98()s. These trentls
began prior l o 1980, as shown by the income declines o f
the bottom 80% o f families from 1977 it) 1980.
•i7
Inequality has
widened among
each type of
family in much
the same pattern
as has occurred
overall.
�FIGURE 1D
Income Growth Among Top Fifth and By Fifth,
1980-1989
c
o
Q.
^
^
fVP
off*
18
^
o°l°
N
Al ihe oilier eiul ol ihe income scale, however, lliere was
considerable income growth (see linure ID) Ihe highest
income lifih ol Iamilies saw their incomes rise by 22.9% in
the I98()s, live times the -I.(•>% income gain ol the next richest filth. In fact, the income guin of the average family in
the upper 20% over the l9H0s, $20,39.3, is more lli.in the
average income leivl of families in ihe Ixmom two filths.
The most spectacular income growth was among the
upper 1%, whose incomes grew by 62 9% over the 1980s
The $216,185 average income gain of the richest 1% of
families in the 1980s is nearly double the average income
leivl of the next best-off 4% of families and is roughly live
times what the average family income was in 1989
($43,495).
Income growth within the upper lilih was very unequal
over the 1980s. I amilies with income in the SOih lo 90ih
percentile range (the "next 10%" group, 'lable 1.9) experienced only a 9.1% income gain, far better than any group
in the bottom 80% but only one-seventh of the income
growth of the upper 1%.
The rapid income growth of the top fifth relative lo the
bottom four-fifths of the population led to a substantial
increase in the share of total income accruing to the rich
(Table 1.10) By 1989, the upper 20% of families received
more than half (51.8%) of all income, increasing their
share of income by 4 4 percentage points since 1980 and
5.3 percentage points from 1977. However, within the
upper fifth, the top 5% gained a greater share of income
while the remaining 15% saw their share of total incomedecline. I he unbalanced income growth of the 1980s thus
caused the bottom 95% of the population to lose incomeshare to the upper 5%. Kcllccling their spectacular incomegrowth, the upper 1% of families increased their share of
income from 9.4% in 1980 to 13.0% in 1989 As a result,
the income share of the upper 1% (13.0%) in 1989 was
equal to that of the bottom 40% (1.3 0%)
Table 1.11 uses an additional source of information—
data on adjusted gross income from personal lax
returns—to examine the pattern of income growth since
1979 lax return data have a bias, it should be noted.
Ix-cause they necessarily exclude the poorest families who
do not file tax returns (perhaps as many as 20% of all families), the result of which is to portray greater income
19
The most
spectacular
income growth
was among the
upper 1%, whose
incomes grew by
62.9% over the
1980s. The
$216,185
average income
gain of the richest
1% of families in
the 1980s is
nearly double the
average income
level of the next
best-off 4% of
families.
�TABLE 1.10
Changes in Family Income Shares, 1977-1989
Family Income Shares in:
Income
Grouup
Maximum
Income**
1977
1989
100.0%
All*
1980
100.0%
Top Fifth:
Top 1%
Next 4%
Next 5%
Next 10%
$330,107
149,103
112,847
46.5%
8.7
11.8
-10.-1 15.8
Bottom Four
Fifths:
Fourth
Middle
Second
Lowest
$ 83,437
55,109
37,228
20.496
53.8%
22.6
15.7
10.6
4.9
Share Change
1977-89
1980-89
100.0%
0.0
0.0
47.4%
9.4
12.0
10.3
15.7
51.8%
13.0
12.9
10.3
15.5
5.3
4.3
1.1
0.2
-0.3
4.4
3.6
0.9
0.0
-0.2
52.9%
22.5
15.5
10.4
4.5
49.0%
21.4
14.6
9.3
3.7
-4.8
-1.2
-1.1
-1.3
-1.2
-3.9
-1.1
-0.9
-1.1
-0.8
'Shares of income add to more than one hundred because those with negative incomes were excluded
from lowest fifth.
'•Highest income in the group. Income of family of four in 1993 dollars.
TABLE 1.11
Change in Family Income by Income
Level Using Tax Return Data, 1979-1988
(1989 Dollars)
Family
Income
Percentile
20th
40th
60th
80th
90th
95th
99th
Change
Income
1979-88
Dollars
Percent
1979
1988
$ 7,811
15,825
26,723
42,105
55,522
70,040
134,998
$ 6,681
14,593
25,041
42,192
58,109
76,511
168,725
50
$-1,130
-1,232
-1,682
87
2,587
6,471
33,727
-14.5%
-7.8
-6.3
0.2
4.7
9.2
25.0
growth than actually occurred (assuming the lowest
income families fared worst). Remarkably then, the data in
Table 1.11 show that incomes fell among the bottom sixty
percent of tax-filing families, with a 6.3% income decline
for a family at the 60th percentile (whose income is above
60%, but lower than 40% of other families). There was
essentially no income growth, a rise of just 0.2%, for a family at the 80th percentile. As we have seen with other data,
the income growth of the best-off families—here the 25%
gain of the family at the 99th percentile—far exceeded the
income gains of the rest of the population. Again, the
income gain of the richest families, 533,727 over nine
years, exceeded the total incomes of more than 60% of the
families.
Some economists express doubts about analyses of
income because the incomes of families fluctuate from year
to year in response to special circumstances—a layoff, a
one time sale of an asset and so on. As a result, a family's
income reflects transient events and does not indicate its
economic well-being over the long term. A preferable item
to study, in this view, is the consumption levels of families
because families typically gear their consumption to their
expected incomes over the long term.
Table 1.12, which presents the trend in the distribution
of consumption, permits a view of this final dimension of
growing inequality. The distribution of consumption
appears to be more equal than that of income (compare
Table 1.12 to Table 1.6) in any one year. However, the
trend towards equality in the 1960s and towards inequality
over the 1970s and 1980s is apparent in both income and
consumption measures of well-being. Moreover, using
both income and consumption data one sees rapid growth
among the best-off while the relative well-being of the bottom 60% was reduced over the 1980s.
Another way of characterizing the unequal income
growth of recent years is to calculate the increase in total
family income (omining income growth simply due to a
greater number o f families) that went to the various
income groups, as shown in Table 1.13 and Figure I E .
The income gains of the top 1% of families were so great
relative to the rest of the population that their income
growth accounted for 71.5% of total income growth from
1977 to 1989. Over the more recent time period, 1980 to
1989, the top 1% received over half (53.2%) of all of the
51
The trend towards
equality in the
1960s and
towards inequality
over the 1970s
and 1980s is
apparent in both
income and
consumption
measures of
well-being.
�TABLE 1.12
Shares of Consumption, 1960-1988
Share of Total Consumption*
Year
1960-61
1972-73
1980
1988
Change,
1980-88
Lowest
Fifth
Middle
Fifth
Second
Fifth
8.2%
9.3
8.4
7.5
14.0%
14.5
14.0
13.5
-0.9
-0.5
18.3%
18.5
18.5
18.2
-0.3
Fourth
Fifth
Top
Fifth
23.3%
23.0
23.2
23.6
36.2%
34.7
35.9
37.2
0.4
1.3
•Adjusted lor family size.
income growth. This can be contrasted to the much
smaller 9.4% of total income growth (equal to their 1980
share of income) that the top 1% of families would have
enjoyed from 1980 to 1989 if income inequality had not
grown. Other income groups within the top fifth of families also obtained sizeable shares of the total income
growth.
Reflecting the fall in their average incomes, the bonom
60% of families received none of (a negative share of) the
income growth generated from 1977 or 1980 to 1989.
Other dimensions of the recent growth in inequality are
examined in later chapters. Chapter 2 focuses on the effect
of changes in tax levels and the distribution of the tax burden on after-tax income inequality and growth. Chapter 3
examines and explains trends in wage levels and inequality.
Chapter 5 looks at growing wealth inequality while Chapter 6 examines the growth in poverty. We now turn to
some explanations of the growth in income inequality we
just documented.
From 1980 to
1989, the top 1%
received over half
(53.2%) of all of
the income
growth.
TABLE 1.13
Shares of Aggregate Income Growth, 1977-1989
Shares of Aggregate Income Growth*
Group
1977-89
71.5%
28.1
16.3
17.0
Bottom FourFifths:
Fourth
Middle
Second
Lowest
53.2%
22.1
14.6
13.5
6.7%
-10.6
-12.8
-5.7
Top Fifth:
Top 1 %
Next 4%
Next 5%
Next 10%
10.3%
-1.3
-4.3
-2.0
FIGURE 1E
Shares of Aggregate Income Growth Among Top Fifth,
1980-1989
1980-89
c
<D
•Shares sum to more than one hundred because negative incomes have been excluded from the lowest
fifth.
Top 1%
52
Next 4%
Next 5%
53
Next 10%
�Greater Capital Incomes, Lower Labor
Incomes
TABLE 1.14
Source ol Family Income for Each Fifth o Families, 1989
f
Source of Family Income
Gov't
Income Group
Labor
All
70.7%
15.3%
7.0%
5.0%
100.0
Top Fifth:
Top 1%
Next 4%
Next 5%
Next 10% -
72.7
51.4
70.9
79.6
81.0
22.9
46.4
22.2
13.3
9.9
2.3
0.6
2.0
2.5
4.0
4.1
1.6
4.9
4.6
5:1
100.0
100.0
100.0
100.0
100.0 "
Bottom Four
Fifths:
Fourth
Middle
Second
Lowest
80.4
76.4
69.8
50.9
8.3
7.4
5.7
3.6
5.8
9.9
17.7
39.1
5.5
6.4
6.8
6.4
100.0
100.0
100.0
100.0
Capital*
"
Transfer
"
Year
86.1%
84.5
82.6
Total
Shares of
Market-Based Incomes**
Labor
Capital*
Total
1977
1980
1989
Other
13.9%
15.5
17.4
100.0%
100.0
100.0
•Includes rent, dividend, interest income and realized capital gains.
"Incomes from wages (wages and salaries, self-employment earnings) and capital, excluding transfer
incomes (AFDC, social security, pensions, alimony, etc.)
54
In the 1980s, the fortunes of individual families
depended on the sources of their incomes: labor income,
capital income, or government assistance. One significant
reason for the unequal growth in family incomes is that, in
recent years, a greater share of our total income has been
in the form of capital incomes (such as rent, dividends,
interest payments, capital gains) and a smaller share has
been earned as wages and salaries. Since most families
receive litde or no capital income, this shift has had a substantial impact on income distribution. There has also
been a growing inequality in the_distribution of both wages
and capital income.
Table 1.14 presents data which show the sources of
income for families in each income group. The top fifth
gets a far larger percentage of its income from financial
assets, i.e., capital, compared to the other 80% of the population. The top 1% receives nearly half its income from
financial assets. Besides the top 1%, only the next richest
4% of families receives at least 20% of its income from capital incomes, with the bottom 90% of families relying on
capital incomes for less than 10% of its income.
Those without access to capital income depend either
on wages (the broad middle) or on government transfers
(the bottom) as their primary source of income. As a result,
the cutback in government cash assistance primarily affects
the income prospects of the lowest 40% of the population,
but particularly the bottom fifth (see Chapter 6). The
income prospects of families in the 20th to 99th percentile, on the other hand, depend primarily on the level and
distribution of wages and salaries.
Table 1.15 shows that income accruing to owners of
capital has been growing far faster than labor income. In
the 10 years between 1979 and 1989, those who owned
income-producing property such as real estate, corporate
stock, bonds, and other interest-bearing assets saw their
incomes from these sources increase by 65.5%. The fast
growth in capital income is primarily due to strong growth
in interest payments of 78.9% which was caused by the
high real (inflation-adjusted) interest rates of the 1980s.
55
Income accruing
to owners of
capital has been
growing far faster
than labor income
primarily due to
strong growth in
interest payments
of 78.9% which
was caused by
the high real
(inflation-adjusted)
interest rates of
the 1980s.
�TABLE 1.15
Real Income Growth B Type of Personal Income, 1979-1989
y
Total
Growth,
1979-89
Income Type
1) Total Capital Income
a) Rent
b) Dividends
c) Interest
65.3%
-156.1
41.9
78.9
2) Total Labor Income
a) Wages & Salaries
b) Other Labor Income
22.8%
22.9
21.8
3) Proprietor's Income*
13.9%
Total Market-Based Personal Income**
28.4%
•Business and farm owners' income.
••Total of listed income types.
TABLE 1.16
Shares of Market-Based Personal Income B Type, 1959-1991
y
Income Type*
1959
1967
1973
1979
1) Total Capital Income
a) Rent
b) Dividends
c) Interest
10.7%
3.9
3.4
6.1
14.9%
3.2
3.6
8.1
13.7%
1.7
2.7
9.3
15.3% 19.7%
0.5
-0.2
2.7
3.0
12.1
16.9
19.3%
0.3
3.0
16.7
2) Total Labor Income
a) Wages & Salaries
b) Fringe Benefits
72.6%
69.8
2.8
73.6%
70.0
3.5
74.8%
69.9
4.9
74.8%
68.1
6.7
71.6%
65.2
6.4
71.9%
65.1
6.7
3) Proprietor's Income*
13.9%
11.6%
11.5%
9.9%
8.7%
8.8%
Total Personal Market
- Based Income*
1989
1991
100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Dividend income grew by less, but by a still substantial
41.9%. The growth in capital, or property-based, income is
actually somewhat understated in Table 1.15 since the data
do not account for the appreciated value of assets in this
period. As Chapter 5 shows over the 1979-1989 period
there was a rapid gain in the value of financial assets
(bonds, stocks) and real estate and businesses. This
increased value of assets yields both realized capital gains
when the assets are sold or unrealized capital gains that
can generate income at a future time. The analysis in Table
1.15 ignores both realized and unrealized capital gains and
therefore understates the increased returns to capital.
During the same period, total labor income, which
includes the pay of executives and professionals as well as
hourly workers, rose by 22.8%, just one third as much as
capital income. (This rise in labor income primarily reflects
a larger employed population working longer hours and is
consistent with flat or declining real hourly wages.) Farmer
and small business owners' incomes, called proprietor's
income, rose 13 9%.
The post-1973 shift away from labor income and toward
capital income is unique in the postwar period (Table
1.16) and is partly responsible for the recent surge in inequality. Since the rich are the owners of income-producing
property, the fact that the assets they own have commanded an increasing share of total income automatically
leads to income growth which is concentrated at the top.
This is especially the case since, as we discuss below, there
was also a growth in the inequality of the distribution of
capital incomes among families.
Capital income was 19-7% of total income in 1989,
nearly double the 10.7% share in 1959 (the earliest year for
which there are data). Half of the increase in capital
income's share of total income occurred since 1979,
driven by high real interest rates. Labor income's share of
total income declined from 1979 to 1989, reversing the
pre-1979 trend in which labor income's share rose or held
its own over the prior three decades. The share of wage
and salary income in 1989 was the lowest in the postwar
period. There was also a shrinkage of the share of fringe
benefits in the 1980s, the first shrinkage over the last four
decades.
•See definitions In Table 1.15.
56
57
Since the rich are
the owners of
income-producing
property, the fact
that the assets
they own have
commanded an
increasing share
of total income
automatically
leads to income
growth which is
concentrated at
the top.
�TABLE 1.17
Shares of Income b Type, b Sector, 1959-1989
y
y
Income Type
1959
1967
1973
1979
1989
National Income, All Sectors
Labor
Capital
Prop. Profit
Total
68.6%
18.6
12.5
100.0
69.9% 72.6% 73.4% 73.1%
19.2
16.0
16.3
16.5
10.4
10.4
8.9
8.1
100.0
100.0
100.0
100.0
(a) Corporate and Business Sector
Labor
Capital
Total
44.3%
18.6
63.0
45.8%
19.2
64.9
48.2%
16.0
64.3
50.5%
16.3
66.8
49.4%
9.0%
0.3
12.5
21.9
6.5%
0.5
10.4
17.4
5.2%
1.0
10.4
16.6
4.7%
1.4
8.9
15.0
4.4%
2.3
8.1
14.8
15.2%
0.0
15.2
17.6%
0.0
17.6
19.2%
0.0
19.2
18.2%
0.0
18.2
19.2%
0.0
19.2
16.5
65.9
(b) Proprietors Sector
Labor
Capital
Prop. Profit
Total
(c) Government/Nonprofit Sector
Labor
Capital
Total
Addendum:
Shares of Corporate Sector Income*
Labor
Capital
Total
78.5%
21.5
100.0
78.0% 82.4% 83.2% 82.1?
16.8
18.0
22.0
17.6
100.0
100.0
100.0
100.0
•Does not include sole proprietorships, partnerships and other private noncorporate businesses. The
corporate sector, which includes both financial and nonfinancial corporations, accounted for 58% of
national income and 71.6% of private sector income in 1989.
58
It is difficult to interpret changes in proprietor's income
because it is a mixture o f both labor and capital income.
That is, the income that an owner o f a business (or farmer)
receives results from his o r her w o r k effort (labor income)
and his or her ownership (capital income) o f the business
o r farm. To the extent that the shrinkage o f proprietor's
income results from a shift o f people out o f the proprietary
sector (i.e., leaving farming) and i n t o wage and salary
e m p l o y m e n t there w i l l be a c o n c o m i t a n t increase i n
labor's share o f income (i.e., as farm income is replaced by
wage income). This shift out o f proprietor's income thus
helps to explain a rising labor share in some time periods.
It is interesting to note, therefore, that labor's share o f
income feu from 1979 to 1989 d e s p i t e a n erosion o f proprietor's income from 9 9% to 8.7% o f total income.
The data i n the bottom panel o f Table 1.14 confirm a
shift towards capital income among families. The share o f
capital income i n all market-based income (i.e., excluding
government and o t h e r transfers o f income) rose f r o m
13 9% i n 1977 to 17.4% i n 1989From the p o i n t o f view o f national income (incomes
generated by the corporate, proprietor and government
sectors) there does not appear to be a significant shift away
from labor income towards capital income (Table 1.17).
For instance, labor's share o f national income fell only
slighdy from 73 4% i n 1979 to 7 3 . 1 % in 1989. A closer
look at the underlying data, however, suggests a significant
shift away from labor income. First, note that labor's share
o f national income had steadily risen from 1959 to 1979.
One reason for the expanding share o f labor income was
the steady expansion o f the government/nonprofit sector.
When the government/nonprofit sector grows, there is a
tendency for labor's share o f income to grow because this
sector generates only labor income and n o capital income.
For example, the growth o f the government/nonprofit sector from 18.2% to 19 2% o f national income between 1979
and 1989 necessarily added 1 % to labor's share o f national
income (other things remaining equal).
Labor's share o f national income also grows as the proprietary sector (farm and non-farm unincorporated businesses) shrinks, as it d i d from 1959 to 1979, because
labor's share o f income i n that sector is relatively l o w (less
than one-third o f it i n 1979). When resources shift from a
sector with a l o w labor share o f income, such as the propri-
59
The failure of
labor's share to
grow stems
primarily from the
unique decline in
labor's share of the corporate
business sector
between 1979
and 1989,
reversing the
tendency towards
more labor
income from the
1960s to the
1970s.
�etors sector, to sectors with a higher labor share (all of the
other sectors) the share of labor income in the economy
necessarily rises. Thus, the changing composition of
income across organizational sectors (expanding government, shrinking proprietors) provides momentum for an
increase in labor's share of national income.
Why then did labor's share of national income not continue to rise in the 1979-1989 period as it had earlier?
Given the continued, albeit slow, expansion of the governThere was an ment/nonprofit sector one would have expected a continincrease in the ued expansion of labor's share of national income. The
capital income failure of labor's share to grow stems primarily from the
share of nationalunique decline in labor's share of the corporate and busiincome betweenness sector between 1979 and 1989, reversing the tend1979 and 1979 ency towards more labor income from the 1960s to the
and an even 1970s. This is most clearly seen in the bottom panel of
larger growth Table 1.17 which shows the division of incomes in the corbetween 1979 porate sector—labor's share fell from 83 2% in 1979 to
and 1989. 82.1% in 1989. These data suggest that there has been a
shift away from labor income in the private sector (both
corporate and proprietor) that does not show up in
national income because of the growth of the labor-intensive government/nonprofit sector.
The data in Table 1.17 confirm the expansion of capital
income's share of total income seen in other tables. For
instance, there was an increase in the capital income share
of national income between 1973 and 1979 and an even
larger growth between 1979 and 1989.
Nevertheless, there was a larger shift away from labor
towards capital income in our analysis of personal income
in Tables 1.15 and 1.16. One reason is that capital income
is defined differently in these tables. At the "personal
level," capital income includes interest, dividends and rent
while at the "national income" level capital income
includes rents, profits, interest and dividends. The difference is that corporate generated capital income received
by persons takes the form of dividends and interest paid by
firms. Individuals also benefit from higher corporate
retained profits (the after-tax profits not paid out in dividends), which are linked to realized, and unrealized, capital gains. The personal income data used in Table 1.15 and
1.16, however, do not include any capital gains. Capital
income in "national income" is primarily based on corporate profits and interest income.
60
One possible explanation for a faster rise in capital
income among persons than in national income is that
increased corporate indebtedness has forced a change in
the composition, or form, of capital income—more interest income and less capital gains based on corporate
retained earnings—but not the total amount of capital
income. The increased debt on corporate balance sheets
over the 1980s, for instance, meant that corporations paid
out more in interest to persons. These greater interest payments, however, reduce corporate profits and thus
retained earnings (roughly defined as profits after dividend
payments), slowing the appreciation of net worth in the
corporate sector.
To the extent that greater interest payments to persons
were the result of more corporate debt one could argue
that households were just receiving their capital income in
a different form—interest payments—rather than capital
gains (from higher stock prices) based on higher retained
corporate profits (if companies had not paid out so much
interest they would have "plowed back" that much more in
profits). In this view, since capital gains are not measured
as part of personal income (in the relevant government
data) what is actually a change in the form of capital
income (more interest, less capital gains) appears as an
expansion of capital income.
This dynamic, however, can provide only a partial explanation for the rising share of capital income in personal
income. One reason is that interest income has risen not
only because of increased corporate (and government) leverage, but also because interest rates in the 1980s, were
historically high in real (inflation-adjusted) terms. Moreover, capital gains expanded over the 1980s, so it seems
that households managed to receive both more interest
(partially at the expense of retained earnings) and more
capital gains. The increase in realized capital gains is discussed below. There was also an increase in unrealized
capital gains, reflected in the rapid growth of real financial
assets and wealth, which we discuss in Chapter 5. Thus,
owners of capital in the 1980s were able to expand both
their current flow of capital income (i.e., receive more dividends and interest income) and to augment their wealth
holdings (i.e., their stocks and other assets appreciated in
value).
61
Owners of capital
in the 1980s were
able to expand
both their current
flow of capital
income (i.e.,
receive more
dividends and
interest income)
and to augment
their wealth
holdings (i.e.,
their stocks and
other assets
appreciated in
value).
�TABLE 1.18
Distribution of Labor and Capital Incomes, 1977-1989
Shares of
Labor Income
Shares of
Income Group
1977
All
100.0% 100.0% 100.0%
1980
1977
1989
1989
1980
100.0% 100.0%
100.0%
Top Fifth:
Top 1%
Next 4%
Next 5%
Next 10%
72.6%
33.1
20.0
9.2
10.3
74.7%
34.9
19.7
9.7
10.5
77.5%
39.7
18.8
8.9
10.1
46.0%
5.8
11.5
11.1
17.6
46.4
5.9
11.6
11.1
17.8
50.3%
9.2
12.6
11.2
17.2
Bottom FourFifths:
Fourth
Middle
Second
Lowest
12.3%
7.9
5.1
1.8
12.3%
6.9
4.2
1.5
11.6%
7.1
3.5
0.9
24.8%
16.8
10.0
2.8
25.1%
16.6
9.8
2.6
23.7%
15.3
8.9
2.6
FIGURE IF
Distribution of Labor and Capital Incomes, Top 20%,
1980-1989
50%
40
-
20
Capital
-
30
10 5
10 c
'0 '
0
Top 1 %
Next 4%
Next 5%
Next 10%
Next 5%
Next 10%
50%
Labor
40
30
20
-
10
-
-£9_
0
Top 1%
Next 4%
•
62
1980
There was not only an increase in the share of total
income accruing to owners of capital in the 1980s, there
was also a growing inequality of the receipt of capital
incomes (Table 1.18 and Figure IF) Between 1980 and
1989, for instance, the top 1% of families increased their
share of capital income from 34.9% to 39 7%. All of the
other income groups (except the middle fifth) had a lesser
share of capital income in 1989 than in 1980, including a
decline in the shares of the remaining groups within the
top fifth.
There has also been a growing inequality of labor
incomes. Between 1980 and 1989, for instance, the top
10% of families expanded their shares of labor income
while the bottom 90% saw their share of labor income
decline (Table 1.18). Because labor income roughly comprises a 71% share of total income (Table 1.14), any
increased inequality of labor incomes will have a major
impact on overall income inequality. In Chapter 3 we analyze the growth of wage inequality.
When we examine the growth of incomes of the top-fifth
between 1980 and 1989 by type of income (Table 1.19), it
is evident that the increased inequality of both labor and
capital income and the shift towards greater capital income
has benefitted upper-income groups. The top 1% had a
spectacular 62.9% growth in their incomes between 1980
and 1989 because of fast growth (47.6%) in their capital
income and an even faster growth (78.5%) in their labor
incomes. As Table 1.20 shows, the growth in capital
income explains 38.8% of the total growth of the top 1%'s
income from 1980 to 1989. The growth in realized capital
gains alone account for 15.2% of the total income growth
among the top 1%, providing $32,874 more in income in
1989 than in 1980—an amount greater than the total
income growth among the next richest 4% from 1980 to
1989 an increase of $22,485. The fast growth of labor
income fueled 58.6% of the total income growth of the top
1%. Any explanation of the rising income level and share of
the top 1% must then account for the growth in both capital and labor incomes.
The remaining segments of the top fifth had much more
modest income gains. For the "next 4%" and "next 5%"
1989
63
There was not
only an increase
in the share of
total income
accruing to
owners of capital
in the 1980s,
there was also a
growing inequality
of the receipt of
capital incomes.
�(the remainder of the upper 10% of families) income
growth was based on both increased capital incomes and
labor incomes, but the importance of higher labor incomes
was more important to these groups than for the top 1%
(Table 1.20).
TABLE 1.19
Sources o Income Growth of T p Fifth, 1980-1989
f
o
Wages
and
Salaries
Self
Employment
1980
1989
Change
$121,294
209,363
$ 39.859
78,371
Dollar
Percent
$ 88.069
$ 38,513
72.6%
96.6%
$ 32,874
$126,582
$ 50,383
$ 83,817
$216,185
78.5%
59.6%
36.0%
62.9%
47.6%
1980
1989
Change
$ 68,360
81,334
$ 79,205
93,614
Dollar
Percent
$ 12,974
$
19.0%
1,543
14.4%
$ 14,408
$
18.2%
4,235
$
24.3%
1980
1989
Change
$ 57,514
65,432
4,097
4,386
$ 61,611
69,818
7,588
9,279
Dollar
Percent
$
Income
Group
All
Labor
Rent,
Dividends
& Interest
Realized
Capital
Gains
All
Capital
Total
$161,153
287,735
$ 84,528
134,911
$ 91,400
124,274
$175,928
259,745
$343,601
559,795
The Shrinking Middle Class?
Top 1 %
Next 4 %
$ 10,736
12,279
$ 17,419
21.654
$
7,340
7,658
$ 24,759
29,312
318
$
4.3%
$109,551
132,036
$ 22,485
4,553
18.4%
20.5%
Next 5%
$
7,918
$
13.8%
288
70%
$
$
8,207
$
13.3%
$
$
1,710
22.5%
2,276
2,368
$
92
$
4.0%
9,864
11.666
$ 75,876
87,711
1.802
$ 11,835
15.6%
18.3%
Another dimension of income growth is the proportion
of the population that has low, middle and high incomes.
There are two factors which determine the distribution of
the population at various income levels—the rate of
growth of average income and changes in income equality.
As average, or median, income grows there will be a
greater proportion of the population at higher-income
levels. However, if inequality grows such that the lowincome population receives an historically low proportion
of the income growth and the high-income population
obtains an usually large proportion of the income growth
then a rise in average income may not translate into a general upward movement of the population to higher-income
levels. That is, it is possible for incomes to grow on average
and have segments of the population with declining
incomes.
Another
dimension oi
income growth is
the proportion ol
the population
that has low,
middle and high
incomes.
Next 10%
1980
1989
Change
S 49,068
51,007
Dollar
Percent
$
$
1,939
$
4.0%
2,055
2,372
318
15.5%
$ 51.183
53,379
$
$
$
2,196
4.3%
4.532
5,536
$
1,004 $
22.1%
1,088
989
$
$
-99
-9.1%
5,559
6,524
$ 60,428
65,900
$
965
17.4%
5,472
-9.1%
TABLE 1.20
Shares o Income Go t b Type o Income for T p Fifth, 1980-1989
f
r wh y
f
o
Percent of Income Growth Due to:
Income
Group
Top1%
Next 4%
Next 5%
Next 10%
Wages
and
Salaries
40.7%
57.7
66.9
35.4
Self
Employment
17.8%
6.9
2.4
5.8
All
Labor
58.6%
64.1
69.3
40.1
Rent,
Dividends
& Interest
23.3%
18.8
14.4
183
Realized
Capital
Gains
15.2%
1.4
0.8
-1.8
•Includes transfer and other incomes not shown.
64
65
All
Capital
38.8%
203
15.2
176
Total*
100.0%
100.0
100.0
100.0
�TABLE 1.21
Distribution o Persons, Households and
f
Families b Income Level, 1969-1980
y
Percent of Persons at
Relative Income Levels
Income Level
1969
1979
1989
Percentage Point
Change
1969-79
1979-89
Families (Income)*
17.7%
63.9
18.6
16.3%
57.3
26.2
17.0%
51.4
31.6
-1.4
-6.6
7.6
0.7
-5.9
5.4
25.8%
58.2
16.1
Under $15,000
$15,000 to 50,000
Over $50,000
25.0%
53.8
21.1
24.1%
50.2
25.7
-0.8
-4.4
5.0
-0.9
-3.6
4.6
20.0%
68.1
11.9
22.1%
63.3
14.7
2.1
-3.0
1.0
2.1
-4.8
2.8
Households (Income)*
Under $15,000
$15,000 to 50,000
Over $50,000
Persons (Income Relative to Median)**
Less than 50%
From 50% to 200%
More than 200%
17.9%
71.1
10.9
*1990 Dollars.
"Each person assigned his or her family's per capita (size-adjusted) income
66
Tabic 1.21 shows the p r o p o r t i o n ol Iamilies. households and persons with low, middle anil high incomes in
1969, 1979 and 1989 l o r households anil families, the
definition o i middle income has been ariiilranly set as a
range from $15,000 to 50,000 Among persons, a middle
income is one which is from 50% to 200% of median family (size-adjusted) income
Over the 1969-1989 period, there was a decline in die
p r o p o n i o n o f families with middle incomes, as more families attained incomes beyond $50,000 Ihe shift upward
was smaller over the 1979-1989 p e r i o d than in ihe
1969-1979 period (5.4% versus 7 6%) Ihe later period
also saw a s h i f t i n g o f 0.7% o f families out o f m i d d l c
incomes and into the low-income category.
When one looks at households—which includes singlepersons in addition lo families—there does not appear to
be any shift downward in the 1979-1989 period, although
the upward shift is still somewhat smaller in the morerecent than the earlier period.
The analysis of persons examines the incomes of people—single and in families—according to the per capita
incomes in their family (size-adjusted), with a single person given an individual income. In this analysis, there has
been a shift out o f middle incomes since 1969, but the
shift is into the lower- as well as into the higher-income
group However, there is a stronger shift upward in the
most recent, 1979-1989, p e r i o d than in the earlier,
1969-1979, period (2.8% versus 1.0%).
Table 1.22 repeats the analysis o f persons from fable
1.21, but categorizes families anil persons by the education
level of their household head and is limited to people in
their prime earnings years (25 to 64 years old) Over the
1969-1989 period, there has been a shrinkage in the proportion of families headed by a high school graduate (or a
high school d r o p o u t ) w h o hail either a middle-class o r
high income, especially since 1979. Hven among the "some
college" category there has been a general shift to a smaller
middle-income population and a greater low-income population, with a stronger shift downward from 1969 to 1979
(and a slight uptick of high-income families since 1979)
Ihe families with college graduates as their household
heail were shilling d o w n w a r d in income from 1969 to
67
In this analysis,
there has been a
shift out of middle
incomes since
1969, but the shift
is into the loweras well as into the
higher-income
group.
�l
TABLE 122
Distribution o Prime A e Adults
f
g
b Relative Income Level, 1969-1984
y
Income Relative
to Median*
Percent of Persons at Relative
Income Levels
1969
1979
1989
Percentage Point
Change
1969-79 1979-89
Less than High School
21.7%
71.5
6.8
29.0%
65.4
5.6
38.4%
57.0
4.5
7.3
-6.1
-1.2
9.4
-8.4
-1.1
7.5%
77.8
14.6
11.0%
75.6
13.4
15.7%
71.4
12.9
3.5
-2.2
-1.2
4.7
-4.2
-0.5
5.9%
71.1
23.0
8.6%
71.7
19.7
10.2%
69.5
20.2
2.7
0.6
-3.3
1.6
-2.2
0.5
3.8%
58.8
37.4
Less than 50%
From 50 to 200%
More than 200%
5.1%
61.0
33.9
4.6%
55.0
40.4
1.3
2.2
-3.5
-0.5
-6.0
6.5
High School Graduate
Less than 50%
From 50 to 200%
More than 200%
Some College
Less than 50%
From 50 to 200%
More than 200%
College Graduate
Less than 50%
From 50 to 200%
More than 200%
•Persons from 25 to 64 years old. based on a person's per capita (size-adjusted) family income.
68
l >79, but upwards alter 1979 Muse channcs in the
income levels of families antl persons by the educational
attainment of the household head retlecls the trends in
wages for the various educational groups, with college
graduates faring well in the 1980s, hut poorly in the 1970s
(due to a surplus of them) and falling wages among the
noncollege-educated workforce Ixtween 1979 and 1989
(see Chapter 3 for an analysis of these trentls)
Increased Work by Wives Cushions
Income Fall
family earnings growth has not only been slow and
unecjual, it has also increasingly come from greater work
effort—from a rise in the number of earners per family and
in the average weeks and weekly hours worked per earner.
The primary source of the increased work effort has been
women, including many with children. As will lie detailed
in Chapter 3, this increased work effort has occurred
simultaneously with a fall in real wages lor men anil for
many women in the 1980s. The result has been increases
in annual earnings primarily through more work rather
than through higher hourly wages
This is troublesome for several reasons. This type of
earnings growth is "self-limiting," meaning it can only go
on until all adult (or even teen) family members are fulltime, hill-year workers. The slowdown in the growth in
women's labor force participation in recent years may
even, in fact, signal the near exhaustion of this type- of
income growth (see Chapter 4). Moreover, there are significant costs and problems associated with this type of
growth, one of the most significant being the lack of affordable, adequate child care (see Chapter 8)
I he problem is not that more women or mothers arcworking, but that they are tloing so because it is the only
way to maintain family income in the face of lower real
wages. It is a sign of poor pcrfbrmancc of the economy
when increased work is elicited through falling real wages
In addition, it is clearly second-best for Iamilies to obtain
higher incomes by working more hours and having more
mcml>crs in the paid lalx>r force than by earning higher
family wages.
69
Family earnings
growth has not
only been slow
and unequal, it
has also
increasingly come
from greater work
effort—from a
rise in the number
of earners per
family and in the
average weeks
and weekly hours
worked per
earner.
�TABLE 1.23
Changes in Incomes o Married-Couple Families
f
with Children b Source, 1979-1989
y
(1989 Dollars)
Year
Lowest
Fifth
Family Income
1979
$16,071
1989
15,448
Percent
-3.9
Change
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Average
$29,594
29,639
$38,903
40,626
$49,826
54,144
$77,692
89,471
$42,417
45,866
02
4.4
8.7
15.2
8.1
$ 1,723
$ 4,318
$11,779
$ 3,449
3,268
4,218
6,987
3,450
-1,374
-171
284
-184
5,563
-771
124
-125
Dollar Change by Source
Total
$-623
$45
Wives'
Earnings
851
2,070
Husbands
Earnings
-1,438
-1,952
Other Income
-36
-73
It is primarily married-couple families (79.2% o f all families in 1989, Table 1.5) w h o are able to increase income
through greater w o r k effort because there are t w o potential adult workers. Single-parent families and individuals,
however, can only increase w o r k effort by having the adult
w o r k more weekly hours and/or more weeks in a year. It is
for this reason that we initially focus o n the greater w o r k
hours o f married couples.
Table 1.23 shows the income growth in total and by
type o f income among married-couple families w i t h children. Among the b o t t o m two-fifths there was either stagnant or foiling incomes, while the middle fifth achieved
modest growth (4.4% over ten years o r 0.43% annually). In
contrast, the best-off married couples w i t h children had a
far faster (15.2%) income growth.
The data in Table 1.23 also show that among the b o n o m
60% there was a fall-off in the earnings of husbands and
other incomes (capital incomes and transfers) while the
earnings o f wives grew in each income fifth. It is apparent
that the rapid income growth among the top fifth was the
result o f a growth in both husband's and wive's earnings
and that the stagnant or falling incomes in the bottom twofifths resulted from the fact that the higher earnings o f
wives could at best offset (as i n the second fifth) the fall in
husband's earnings. O n average (see the last column), all
o f the growth in the incomes o f married-couple families
w i t h children can be attributed to the higher earnings o f
the wives in these families.
Table 1.24 presents the growth in annual hours worked
by husbands and wives and shows that it is more work,
rather than higher hourly wages, that has been fueling
income growth. The data o n the growth in the hours o f "All
husbands" and 'All Wives" reflect changes in the hours of
"Employed Husbands" and "Employed Wives" as well as
changes i n the p r o p o r t i o n o f husbands a n d wives
employed in the year. As the data show, the d r o p in husbands' annual earnings from 1979 to 1989 (Table 1.23)
occurred despite the fact that husbands were w o r k i n g
more, not less, hours (except for a slight d r o p in the bottom fifth). I n 1989, the average husband (including some
not employed) in the second- to top- fifths worked at least
frill-time and year-round (40 hours for 52 weeks is 2080
hours).
70
71
On average alt of
the growth in the
incomes of
married-couple
/amilies with
children can be
attributed to the
higher earnings of
the wives in these
families.
�TABLE 1.24
Husbands and Wives Hours o Work, 1979-1989
f
Family
Member*
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Average
2,189
2,211
2,249
2,274
2,349
2,411
2,122
2,143
Annual Hours, All Husbands*
1979
1989
Percent
Change
Absolute
Change
1,693
1,682
2,115
2,138
-0.6%
1.1%
1.0%
1.1%
2.6%
1.0%
-11
23
22
25
62
21
2,224
2,239
2,273
2,299
2,372
2,428
As we have discussed, it was the greater earnings o f the
wives i n these families which either moderated or counterbalanced the impact o f the falling wages o f husbands. The
average wife in a married-couple family w i t h children was
employed 32.3% m o r e hours i n 1989 than in 1979, a
growth o f 268 hours annually or the equivalent of nearly
The reduction in
the hourly wages
ol husbands, the
general trend in
the 1979-1989
period, led to their
having lower
annual earnings
despite an
increase in
work hours.
2,211
2,234
Annual Hours, Employed Husbands
1979
1989
Percent
Change
Absolute
Change
The reason that husbands' annual earnings fell in the
bottom 60% o f these families is the erosion o f their hourly
wages, which fell 5.5% for husbands i n the middle fifth and
11.8% for husbands in the lowest fifth (Table 1.25). There
was a growth in hourly wages only for husbands in the
best-off fifth o f famUies, a rise of 7.8% from 1979 to 1989.
The reduction in the hourly wages o f husbands, the general trend in the 1979-1989 period, led to their having
lower annual earnings despite an increase in annual w o r k
hours.
1.996
1,958
2,180
2,207
-1.9%
1.2%
-38
27
Annual Hours, All Wives
1979
450
1989
651
Percent
Change
44.7%
Absolute
Change
201
708
996
0.7%
15
852
1,193
1.1%
2.4%
26
56
23
1,078
1,311
1,068
1,341
831
1,099
40.7%
40.0%
21.6%
25.6%
32.3%
288
431
233
273
268
Annual Hours, Employed Wives
1979
1,001
1,226
1989
1,181
1.414
Percent
Change
18%
15%
Absolute
180
Change
188
"In married-couple families with children.
1,314
1,554
18%
240
TABLE 125
Change in Hourly Wages of Husbands' and Wives', 1979-1989
(1991 Dollars)
1.0%
1,478
1,624
10%
146
1,567
1,694
8%
127
1,347
1,516
13%
169
Family
Member*
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Average
$6.84
6.03
$10.95
9.91
$13.66
12.91
$15.98
15.93
$22.24
23.98
$14.38
14.30
-11.8%
$-0.81
-9.5%
$-1.03
-5.5%
$-0.76
-0.3%
$-0.05
7.8%
$1.74
-0.6%
$-0.08
$6.01
6.35
$7.25
7.92
$8.41
10.13
Lowest
Fifth
Husbands' Hourly Wage
1979
1989
Percent
Change
Dollar
Change
Wives' Hourly Wage
1979
1989
Percent
Change
Absolute
Change
$4.56
4.46
-2.2%
$-0.10
5.7%
$0.34
9.2%
$0.67
$7.97
9.17
20.5%
27.8%
15.0%
$1.72
$3.01
$1.20
*ln married-couple families with children.
72
$10.83
13.84
73
�FIGURE IG
Growth in Annual Hours Worked by All Wives,
1979-1989
<
D
0-
Lowest Fifth
Second Filith
74
Middle Fifth
Fourth Fifth
Top Fifth
Average
seven weeks o f full-time w o r k (see Figure I G ) . There was
at least a 40% growth i n w o r k hours among w o m e n i n the
b o t t o m 60% o f these fiamilies, nearly double the increase
among the wives in the bener-off families. Thus, the modest o r falling incomes among families i n the bottom threefifths occurred despite significandy more w o r k effort by
these families and, in particular, by the wives in these families. These greater w o r k i n g hours were the result both o f a
higher p r o p o r t i o n o f wives b e i n g e m p l o y e d and o f
employed wives w o r k i n g more hours (up from 15% to
18% in the b o t t o m three-fifths). In fact, 43% o f the 268
hours growth in the average wives' w o r k year was the
result o f employed wives increasing their hours by 169It is useful to note the degree to which an increase in the
hours o f wives is necessary to offset the fall in the wages of
husbands given that the m e n earn significantly higher
hourly wages. For example, the $0.76 d r o p in the hourly
wage o f husbands in the middle fifth from 1979 to 1989
(working 2,221 annual hours) created an annual loss o f
$1,688. Replacing this income w o u l d require an average
wife i n the middle fifth (earning $7.92 hourly) to w o r k 213
additional hours annually, o r more than five extra full-time
weeks.
Table 1.26 brings together the information f r o m the
prior tables to identify the role o f increased earnings of
wives' and hours o n family income growth. The essential
point is that w i t h o u t the increase i n the earnings o f wives
there w o u l d have been lower incomes in each o f the lowest three-fifths in 1989 than in 1979 and a trivial $100
income g r o w t h for the fourth-fifth (see F i g u r e 1 H ) . In
o t h e r w o r d s , o n l y the t o p fifth d i d n o t d e p e n d o n
increased earnings from wives to obtain income growth or
stem an income decline. The entire 8 . 1 % growth in the
average income o f married-couple families with children
from 1979 to 1989 can be attributed to the increased earnings o f wives (see the 0.0% growth in the last column o f
the second panel o f numbers).
Most o f the contribution to the increased earnings o f
wives came from greater hours o f work. If the average wife
i n the b o t t o m three-fifths had not w o r k e d more hours
there w o u l d have been a fall in income between 1979 and
1989 (or a greater fall for the lowest fifth).
75
The average wife
in a marriedcouple family with
children was
employed 32.3%
more hours in
1989 than in
1979, a growth of
268 hours
annually or the
equivalent of
nearly seven
weeks of full-time
work.
�TABLE 1.26
Role o Higher Wives' Earnings and Hours
f
on Family Income Growth, 1979-1989*
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Average
Change in Family Income
Percent
Dollar
-3.9%
$-623
0.2%
$45
15.2%
$11,779
8.1%
$3,449
0.2%
$100
6.2%
$4,812
0.0%
$-1
3.9%
$1,958
10.3%
$8,001
-9.2%
-6.8%
-4.0%
$-1,474 $-2,025
$-1,545
Change in Family Income
Without Higher Wives' Hours
Percent
Dollar
-9.5%
-6.0%
$-1,520 $-1,784
-2.5%
$-978
Effect on Family Income of Higher:
F I G U R E 1H
G r o w t h in Family I n c o m e , B y W i v e s ' Higher E a r n i n g s ,
(a) Wives' Earnings
Percent
Dollar
5.3%
$851
7.0%
$2,070
8.4%
8.5%
$3,268
$4,218
9.0%
$6,967
8.1%
$3,450
1979-1989
30
Without Wives'
Higher Earnings
With Wives'
Higher Earnings
(b) Wives' Hours
Percent
Dollar
// the average
wife in the bottom
three-fifths had
not worked more
hours there would
have been a fall in
income between
1979 and 1989.
2.3%
$992
8.7%
4.4%
$4,318
$1,723
Change in Family Income
Without Higher Wives' Earning
Percent
Dollar
Has the growing earnings of wives contributed to the
growth of income inequality? The notion that the growth
of "two-earner" families has contributed to growing inequality is intuitively plausible if one thinks that there has
been a growth of high-wage employed women marrying
high-wage men, a trend that would increase the monetary
gap between high- and middle- or low-income families. It
is true, in fact, that the wives in higher income families do
earn more than those in other families and that their
hourly wages have grown the quickest (Table 1.25). However, the fastest growth in work hours has been among the
wives in the bottom three-fifths (Table 1.24) and the effect
of wives' earnings on family income growth has been fairly
even, with a somewhat lower effect on the lowest two-fifths
(Table 1.26). The data discussed so far, however, only indirectly relate to whether the pattern of growth of wives'
earnings led to greater inequality and do not address
whether inequality would be higher or lower without any
earnings from wives.
5.6%
$897
6.2%
6.9%
$1,829
$2,701
4.7%
$2,360
4.9%
$3,778
5.8%
$2,457
20
15.2
10
8.7
-
8.1
6.2
•Married Couples with Children.
4.4
1
0.2
02
—I ~m
0.0
-6.8
-10 | -
.g.2
-20
Lowest Fifth
76
_l_
Second Fifth
Middle Fifth
Fourth Fifth
77
Top Fifth
Average
�TABLE 1.27
Effect o Wives' Earnings on Income Shares
f
A o g Married Couples with Children, 1979-1989
mn
Shares of Income
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Total
Family Income, 1979
Actual
Without Wives' Earnings
Effect of Wives' Earnings
7.6%
6.9
0.7
14.0%
13.3
0.7
18.3%
18.0
0.3
23.5%
23.4
0.1
36.6%
38.5
-1.9
100.0%
100.0
0.0
6.7%
5.8
0.9
12.9%
12.0
0.9
17.7%
17.0
0.7
23.6%
23.2
0.4
39.0%
41.8
-2.8
100.0%
100.0
0.0
-1.1%
-1.3
0.2
-0.6%
-1.0
0.4
0.1%
-0.2
0.3
2.4%
3.3
-0.9
0.0%
0.0
0.0
Family Income, 1989
Actual
Without Wives' Earnings
Effect of Wives' Earnings
Change in Income Shares, 1979-89
Actual
-0.9
Without Wives' Earnings
-1.1
Effect of Wives' Earnings
0.2
78
The data in Table 1.27 do allow a direct examination of
the effect of wives' earnings on income inequality in particular years (1979 and 1989) and on the growth of inequality over that time period. In 1979, the presence of wives'
earnings led to a more equal distribution of income For
instance, without wives earnings the lowest fifth would
have had a 6.9% share of total income, but actually had a
7.6% share of income. The presence of employed wives
positively increased the income shares of the bottom fourfifths in 1979 and decreased the share (although not the
level) of income of the top fifth by 1.9 percentage points
(because increased wives' earnings did not raise the
incomes in the top fifth proportionately as much as in the
bonom four-fifths). By 1989, the presence of wives' earnings had a larger effect on raising the income shares of the
lowest 80% of these families and on lessening the income
share of the top fifth.
Thus, the earnings of wives disproportionately augmented the incomes of the broad middle class and the
lowest income families in 1979 and even more so by 1989.
As a result, the increased earnings of wives had a positive
effect on the relative income growth of the bottom 80% of
families and a negative effect on the income share of the
top fifth. That is, the pattern of growth of wives' earnings
from 1979 to 1989 created a tendency for incomes of the
bottom 80% to grow faster than the incomes of the top
fifth.
Although egalitarian in nature, the effect of increased
incomes from the earnings of wives was not sufficient to
forestall the overall growth of inequality and the loss of
income shares by the lowest three-fifths. Nevertheless,
instead of a 1.0 percentage point drop in income share for
the middle-fifth between 1979 and 1989, more wives'
earnings meant a lesser 0.6% drop in income share.
Greater wives' earnings, however, did enable the fourthfifth to avoid a 0.2 percentage point loss in income share.
The pattern of earnings growth among wives, therefore,
ameliorated rather than exacerbated the growth in inequality among married-couple families with children from 1979
to 1989, at least as reflected by the lesser loss of shares of
income of the bottom 80%.
The growth in work hours in the 1979-1989 period took
place on average in a l l families and not just in marriedcouple families with children to whom all of the data previ79
AHhough
egalitarian in
nature, the effect
of increased
incomes from the
earnings of wives
was not sufficient
to forestall the
overall growth of
inequality and the
loss of income
shares by the
lowest three-fifths.
�TABLE 1.28
Changes in Hours Worked in All Families
and in Married-Couple Families, 1979-1989
Annual Hours of Work*
Family
Type
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
Total
All Families
1979
1989
820
850
Percentage
Change
4.6%
1,800
1,840
2.4%
2,480 3,010
2,550 3,020
2.6% 0.5%
3,440 2,280
3,510 2,330
2.0% 2.3%
Married Couples with Children
1979
1989
Percentage
Change
2,240
2,350
2,950
3,250
3,270
3,580
3,600
3,860
3,990
4,090
3,250
3,460
4.6% 10.0%
9.6%
Nonelderly Childless Married Couples
1979
1,980 3,110
1989
2,150
3,390
Change
8.7%
8.9%
3,590
3,890
8.3%
3,910
4,270
9.3%
4,140
3,330
4,410 3,610
6.6%
8.5%
Nonelderly Individuals
1979
1989
Change
1,820
2,030
1,940
2,090
6.5%
2.8%
2,160
1,590
2,240
1,690
4.0%
6.2%
590
1,350
640
1,540
7.0% 14.2%
7.2%
'Rounded to nearest ten, with percent changes calculated from actual numbers.
80
2.5%
6.6%
ousty presented applied. As Table 1.28 shows, there were
greater hours o f w o r k i n 1989 than in 1979 among families
i n each income fifth and 2.3% greater hours overall. The
largest growth i n hours was among non-elderly marriedcouple families, especially those w i t h o u t children. This
makes sense since families with t w o adults present (compared to single-parent families, for instance) have more
potential labor-time t o offer employers. Regardless, there
was also a significant growth in hours by n o n elderly single
people from 1979 t o 1989.
Because greater earnings from wives can contribute to
greater income g r o w t h for married-couple families b u t n o t
for single-parent families o r individuals, it is possible that
greater wives' earnings may have contributed to growing
inequality among all families (the earlier discussion was o f
inequality a m o n g m a r r i e d couples w i t h c h i l d r e n ) . For
instance, if increased wives' earnings raised the incomes o f
married couples, as it d i d , and increased w o r k effort d i d
not raise the incomes o f single-parent families o r o f individuals i n the same degree, then overall income inequality
w o u l d grow.
Table 1.29 uses t w o measures o f income inequality (the
Gini coefficient and the squared coefficient o f variation—
higher numbers indicate more inequality) to examine the
effect o f wives' earnings o n inequality among married-couple families and all families i n particular years (1968, 1978,
1988) and o n the growth o f inequality over the 1968-1988
period. These data confirm o u r earlier analysis that wives'
earnings d o lead to lesser income inequality i n any particular year and that the effect o f greater wives' earnings has
been t o ameliorate the growth o f inequality among married couples. For instance, income inequality i n 1968 was,
according t o the t w o measures, either 17.4% o r 6.2% less
than i f there had been n o wives' earnings. Greater wives
earnings had a strong egalitarian impact o n married couples' incomes from 1968 to 1978 (reducing inequality by
2 0 . 0 % i n 1978 r a t h e r than 17.4% i n 1968 u s i n g t h e
squared CV and by 8.5% i n 1978 rather than 6.2% i n 1968
using the Gini coefficient). Between 1978 and 1988, however, the pattern o f growth o f wives earnings only led to a
slight decrease o f inequality.
A l t h o u g h greater wives earnings have created m o r e
equal incomes among married couples they have made a
slight contribution to overall family inequality since 1968.
81
The growth in
work hours in the
1979-1989 period
took place on
average in all
families and not
just in marriedcouple families
with children.
�TABLE 1.29
Effect of Wives' Earnings on Income Inequality
Among All Families and Married Couples, 1968-1988
By Family Type
and Inequality Measure*
Inequality Among Families
and Married Couples
1968
1978
1988
Percent Change
1968-88
1978-88
All Families
(a) Squared CV of:
Total Family Income
Less Wives' Earnings
Percent Change
0.51
0.56
-8.9%
0.51
0.55
-7.3%
0.66
0.69
-4.3%
29.4%
23.2
29.4%
25.5
(b) Gini Coefficient of:
Total Family Income
Less Wives' Earnings
Percent Change
0.359
0.366
-1.9%
0.381
0.384
-0.8%
0.422
0.420
0.5%
17.5%
14.8
10.8%
9.8
(a) Squared CV of:
Total Family Income
Less Wives' Earnings
Percent Change
0.38
0.32
0.41
0.46
0.40
0.52
17.4% -20.0% -21.2%
7.9%
13.0
28.1%
30.0
(b) Gini Coefficient of:
Total Family Income
Less Wives' Earnings
Percent Change
0.305
0.325
-6.2%
10.2%
13.2
12.0%
12.2
Married Couples
0.300
0.328
-8.5%
0.336
0.368
-8.7%
"The squared coefficient of variation (CV) and the Gini coefficient are measures of inequality for which
higher values indicate greater inequality.
By both measures the presence of wives' earnings lessened
overall family income inequality in 1968, although only
slightly according to the Gini coefficient. However, the
egalitarian effect of wives' earnings lessened from 1968 to
1988, particularly in the 1980s. In fact, the Gini coefficient
measure shows a slight inegalitarian effect of wives' earnings on all families in 1988.
Nevertheless, the changing pattern of wives' earnings
played a very small role in the overall growth of family
income inequality in the 1980s. This can be seen by comparing the increase in income inequality that actually
occurred and that which would have occurred if there
were no earnings from wives (the last column of Table
1.29). Total family income inequality grew a great deal
from 1978 to 1988 according to both measures after growing more slowly (in the Gini measure) or nor all in the
prior ten years. The growth in inequality between all families from 1978 to 1988 would have been only slightly
smaller (9.8% versus 10.8% or 25.5% versus 29.4%) without the presence of wives' earnings. Another way of
assessing the effect of the changing pattern of wives' earnings on income inequality is to compute the hypothetical
level of inequality that would have prevailed in 1988 if
wives' earnings had as egalitarian an impact as in 1978. For
instance, if wives' earnings had still reduced inequality by
7.3% in 1988 as in 1978 (using the squared CV) then the
level of inequality would have been 0.64 and inequality
between 1978 and 1988 would have grown by 25.4% from
0.51 to 0.64 rather than by the actual 29 4% growth from
0.51 to 0.66. According to the Gini coefficient, if wives'
earnings had had as egalitarian an effect on incomes in
1988 as in 1978 then inequality would have grown 9.4%
rather than the actual 10.8% growth in inequality.
The Downside of Increased Labor
Market Work
As we have shown, the increased earnings of wives in
married-couple families with children have been responsible for nearly all of the income growth in these families,
especially for middle-income families. More than half of
the increase in earnings by wives is due to more wives
working and to more weekly hours and weeks per year
worked by employed wives. The earnings from this
82
83
Although greater
wives earnings
have created
more equal
incomes among
married couples
they have made a
slight contribution
to overall family
inequality since
1968.
�TABLE 1.30
Effect o Second Earner on Household Expenditures
f
(1990 Dollars)
Average Income
and Expenditures
1 -Earner
2-Earner
Families
Families
$12,952
10,576
$21,990
18,996
$9,038
8,420
70%
80
$4,802
3,795
63
574
$7,128
5,225
525
825
$2,326
1,450
462
251
48%
38
733
44
$13,136
$15,494
$2,358
18%
$27,876
26,192
$38,475
36,399
$10,598
10,207
38%
39
$5,933
4,503
81
817
$9,139
6,435
677
1,210
$3,206
1,932
596
392
54%
43
736
48
$7,227
Income Level*
Difference Between
2-Earner and
1-Earner Families
Dollars
Percent
$9,165
$1,939
27%
Low-Income Families
Income
Income
Annual Wage
Expenditures
Work Related
Transportation
Child Care
Food Away From
Home
Other Expenditures
Middle-Income Families
Income
Income
Annual Wage
Expenditures
Work Related
Transportation
Child Care
Food Away From
Home
Other Expenditures
•Married Couples with Children.
84
increased w o r k effort is a positive financial gain but, as we
have discussed, it comes w i t h associated costs: increased
work-related spending, child-care costs and problems, the
loss o f leisure time, and added stress. The downside o f the
increased w o r k effort is delineated in the next few tables.
A two-earner family has considerably more expenses
than a one-earner family w i t h a comparable income. Each
earner i n a family requires transportation t o w o r k and frequendy needs appropriate clothes. With two-earners there
are many household tasks that must be completed for the
family by purchasing services previously performed by a
family member. A major example is the cost o f child care.
Other examples w o u l d be home repair and some meal
preparation.
T a b l e 1.30 shows t h e differences i n e x p e n d i t u r e s
between one- and two-earner families. A m o n g low-income
families, the work-related expenses o f two-earner families
are 48% more than those o f one-earner families. I n contrast, n o n - w o r k related expenditures, " o t h e r e x p e n d i tures," were only 18% greater in two-eamer families. The
major a d d i t i o n a l expenses f o r two-earner families are
transportation and child care. Among middle-income families, two-eamer families spend 54% more o n work-related
items than one-eamer families. One interpretation o f these
higher expenditures is that the average two-eamer family
has more income than the average one-eamer family. This
explanation fails, however, because the growth o f n o n w o r k related expenditures only rise by half as much, 27%.
These added financial costs to two-eamer families are
large enough so as to significantly reduce the net gain from
increased earners. A m o n g m i d d l e - i n c o m e families, f o r
instance, the added w o r k related financial costs o f $3,206
absorb 3 1 % o f the added $10,207 o f annual wages associated with the second earner.
These expenses are averages among families which d o
and those which d o not have to purchase child care. Many
families d o n o t have to pay for child care because their children are i n school or because they can rely o n unpaid relatives for assistance. For instance, only 32% o f families w i t h
employed mothers pay for child care (Table 1.31). A much
higher percentage (from 54.7% to 6 1 3 % ) o f families w i t h
young children (ages 0 to 4) must pay for child care. For
these families, the financial cost o f child care absorbs from
6.4% to 7.4% o f their incomes.
85
These added
financial costs to
two-earner
families are large
enough so as to
significantly
reduce the net
gain from
increased
earners.
�TABLE 1.31
Child-Care Expenditures and Family income
for Employed W m n With Children
o e
(1990 Dollars)
Average
Percent
Weekly
Paying lor Child-Care
Child Care Expenses*
Average
Monthly
Family
Income*
Percent
of Income
Spent on
Child Care
Per Month**
Number
of Women
(000)
Number
Paying lor
Child Care
(000)
Total
18,244
5,831
32.0%
$53.18
$3,586
6.4%
Age of Youngest Child:
Less Than 1 Year
1&2Years
3 & 4 Years
5 Years & Older
1,473
3,451
2,602
10,718
806
2,114
1,509
1,401
54.7%
61.3
58.0
13.1
$62.00
61.29
51.76
37.18
$4,218
3,584
3,375
3,454
6.4%
74
66
4.7
Number of Children Less
Than 5 Years Old:
1 Child
6,030
2 or More Children
1,496
3,510
919
58.2%
61.4
$54.00
74.12
$3,526
4,019
6.6%
8.0
Falling Behind the Earlier Generations
•Refers only to women making child-care payments.
*'Average weekly child-care expenditures converted to a monthly average.
TABLE 1.32
Hours W r e b Employed Parents
ok d y
b Gender, 1969 and 1989
y
Total Hours
Non-Market Hours
Market Hours
Family
Member*
1969
1989
Change
1969
1989
Change
1969
1989
Change
All
Parents
1,916
1,988
72
1,034
1,101
67
2,950
3,089
139
Mothers
1,281
1,627
346
1.583
1,402
-181
2,864
3,029
165
Fathers
2,316
2,330
14
688
816
128
3,004
3,146
142
•Limited to parents without involuntary leisure due to unemployment or short hours (involuntary parttime work).
86
Another cost of having family members spend more time
in the paid labor force is that there is an erosion of leisure
time. The trends in parents' annual hours of work, broken
down into paid labor market work and unpaid household
"non-market" (shopping, child care, cooking, repairs, laundry, etc.) work, from 1969 to 1989 are presented in Table
1.32. The total amount of work has increased in this time
period, and has not been offset by a decline in non-market
hours. This suggests that parents have less leisure time.
For instance, the average parent was employed for 72
more hours in 1989 than in 1969, but also had 67 more
hours of non-market work. The average parents' work year
thus rose by 139 hours over this time period, equivalent to
three and a half full-time weeks of work. There was a rise
in total hours for both mothers and fathers. The increase in
the average mother's market hours was only partially offset
by lesser non-market hours while the average father did
not increase his market hours by much (he was already a
full-time, year-round worker with over 2,300 annual market hours) but did put in 128 more non-market hours
annually in 1989 than in 1969.
Until this point we have exclusively focused on cross-section comparisons, e.g., comparing the incomes of highincome, older, or married-couple families in one year to
those of another. Although these comparisons accurately
portray changes in various dimensions of the income distribution over time, they do not trace the incomes of particular families or individuals over time. For instance, consider
comparisons of income growth by income fifth over a tenyear period. A person or family may be in the middle fifth
in the first year, but in a higher or lower fifth ten years
later. Thus, a comparison of middle-income families over
time actually compares one set of families in the first year
to a different set of families in the later year.
It is especially important to note that the incomes of
individuals and families generally follow a life-cycle pattern. Typically, a person, after completing schooling, starts
earning income in a relatively low-paying entry level job;
sees fast income growth as job changes, accumulated experience and seniority occur over the next two decades; and
then, obtains slower income growth in his or her later
87
Another cost of
having family
members spend
more time in the
paid labor force is
that there is an
erosion of leisure
time. The average
parents' work
year thus rose by
139 hours over
this time period,
equivalent to
three and a half
full-time weeks of
work.
�FIGURE II
Family Income By Age of Householder,
1989
w o r k i n g years. As Figure I I shows, young families (headed
by someone i n his or her twenties) in 1989 had much
lower incomes than families headed by a middle-aged person (in his o r her forties o r fifties). Plus, Figure I I shows
that incomes g r o w relatively rapidly as the household head
proceeds through his or her thirties and forties and that
income growth slackens and then declines as the household head approaches retirement years.
Viewed this way, the income g r o w t h o f families as a
whole over a ten-year period depends both o n how high
incomes are for y o u n g families w h e n they start out and o n
h o w fast incomes g r o w as families progress through their
life cycle pattern. I n fact, the slow growth o f median family
incomes i n recent years is most accurately portrayed as
young families starting off w i t h lower incomes then their
predecessors and o t h e r families p r o c e e d i n g to higher
incomes at an historically slow pace. Note that it is even
possible for average family income to fall over a ten-year
period even though each family may have achieved higher
i n c o m e s o v e r the p e r i o d — t h e mechanism o f i n c o m e
decline c o u l d be m u c h l o w e r starting incomes for the
youngest families and very slow g r o w t h o f incomes for
families.
50,000
|
o
Q
40.000 -
S
30,000 -
0)
o
c
20,000 -
S
10,000 -
^
•P"
Age of Householder
TABLE 1.33
Income Growth b Cohort
y
(1989 Dollars)
Median Family Income By
15-24
Year/Cohort
By Cohort
Bom 1965-74
Born 1955-64
Born 1945-54
Bom 1935-44
Born 1925-34
Cohort
Number
5
4
3
2
1
25-34
35-44
$17,064
21,802
20,976
15,166
11,318
$30,873
32,354
29,895
21,676
$40,202
38,396
34,500
45-54
55-64
$46,101
42,395 $37,643
By Year
$17,064
21,802
20,976
15,166
11,318
1989
1979
1969
1959
1949
88
$30,873
32,354
29,895
21,676
$40,202
38,396
34,500
$46,101 $37,643
42,395
There are n o data available w h i c h can trace the income
growth o f particular people or families over their life cycle.
However, it is possible to examine the income trajectories
o f "birth cohorts" (a group o f similarly aged persons o r
families) over time as in Table 1.33 and Figure 1J. For
ease o f p r e s e n t a t i o n , w e n u m b e r each p o s t w a r b i r t h
cohort w i t h number one being the earliest cohort, the families headed by someone b o m between 1925 and 1934
and w h o m we first observe i n 1949 as families w i t h a
household head 24-years-old o r less. The latest cohort,
number five, consists o f families headed by someone b o m
between 1965 and 1974. We can only observe the incomes
o f this youngest cohort in 1989 w h e n the family head is
24-years-old o r less.
The bottom line o f Figure 1J shows the income trajectory o f cohort number one. These families had a median
income of just $11,318 in 1949 as they started out, but saw
t h e i r incomes rise by 5 9 . 2 % t o $ 2 1 , 6 7 6 by 1959, t o
$34,500 by 1969 and to $42,395 by 1979
The second c o h o r t ( b o r n b e t w e e n 1935 and 1944)
started out w i t h a higher income than cohort number one
89
The slow growth
of median family
incomes in recen
years is most
accurately
portrayed as
young families
starting off with
lower incomes
then their
predecessors and
other families
proceeding to
higher incomes at
an historically
slow pace.
�FIGURE 1J
Median Family Income By Cohort and Age,
1949-1989
50
Born 1935-1944
40
Born 1945-1954
M , '
&
S
Born 1925-1934
$
2
=
o
Q
30
O)
00
O)
Born
1955-1964
/
(Ji
T3
c
C
O
(O
Income
in ...
20
• 1989
Born
1965-1974
o
-C
•
/
/
O 1979
A 1969
10
• 1959
• 1949
i.
15 to 24
25 to 34
35 to 44
45 to 54
Age of Household Head
90
($15,166 versus $11,318) and achieved higher incomes at
each age level than cohort number one. The same pattern
holds for cohon number three which we first observe in
1969 with incomes of $20,976, a significantly higher
income than that of the two earlier cohorts when they
started out. Cohort number three, headed by someone
aged 35-44 years old in 1989, earned $40,202 and was still
ahead of the prior cohorts when they were similarly aged
($38,396 and $34,500).
The postwar pattern of upward income mobility was
broken for the fourth cohort. This cohort, which was bom
between 1955 and 1964, did start out in 1979 with higher
incomes than the earlier cohorts had at that age. However,
ten years later when the fourth cohort was in their late
twenties and early thirties their incomes were less than that
attained by the earlier cohort (number 3) at similar ages
($30,873 versus $32,354). Matters have worsened for the
most recent cohort, number five, which started out in
1989 with a median income of $17,064, significantly
behind the starting incomes of the prior two cohorts. It is
also possible that slow income growth could lead the third
cohort to fail to achieve the incomes of the second cohort
at ages 45 to 54 years old ($46,101).
The break in upward income mobility across cohorts
raises the issue of whether the two, and maybe the three,
most recent cohorts (those bom after 1945 or 1955) will
ever achieve incomes equal to that of the earlier postwar
cohorts. Given continued slow income growth and falling
real wages (see Chapter 3) in the 1990s it is very possible
that the recent (and some future) cohorts will have lifetime
incomes inferior to those of their parents' generation.
Income Mobility
_L
55 to 64
The previous section examined the growth in the
median family income of the various postwar age cohorts
and found that the recent sets of younger families were
starting out with lower incomes and achieving a slower
growth in incomes over their twenties and thirties. This
section examines the income growth of families over a
recent ten-year period using "panel" data which tracks the
same set of families over time. These panel data provide
information on the degree of family income mobility in certain periods as reflected in the ability of particular families
91
Given continued
slow income
growth and falling
real wages in the
1990s it is very
possible that the
recent (and some
future) cohorts
will have lifetime
incomes inferior
to those of their
parents'
generation.
�TABLE 1.34
Distribution o Individuals in Final Year b Family Income
f
y
Fifth in Starting Year
Family Income
Fifth in 1967*
Bottom
Second
Third
Fourth
Top
All
Family Income
Fifth in 1977*
Bottom
Second
Third
Fourth
Top
All
Family Income Fifth in 1976
Lowest
11.2%
4.1
2.5
1.3
0.9
20.0
Second
5.2%
6.0
4.2
2.9
1.8
20.0
Middle
2.0%
5.0
6.0
4.7
2.1
19.9
Fourth
1.3%
3.0
4.9
5.9
4.8
20.0
Top
0.3%
1.7
2.4
5.2
10.4
20.0
All
20.0%
19.8
20.1
20.0
20.0
100.0
Family Income Fifth in 1986
Lowest
10.6%
4.3
2.9
1.0
1.2
20.0
Second
5.0%
6.0
3.8
2.9
2.2
20.0
Middle
2.2%
5.1
5.9
4.3
2.5
20.0
•Sample limited to adults, ages 25 to 54 in starting year.
92
Fourth
1.3%
2.9
4.8
6.8
4.1
20.0
Top
0.8%
1.7
2.6
5.0
10.0
20.0
All
20.0%
20.1
20.0
20.0
20.0
100.0
to start w i t h l o w incomes and end w i t h high incomes, o r
vice versa.
T a b l e 1.34 uses p a n e l data i n t w o time periods
(1967-1976 and 1977-1986) to examine the ten-year relative income growth o f individuals w h o were from 25 to 54
years old in the starting year. Specifically, the table presents
the fraction o f the individuals in each income fifth in the
starting year by the income fifth achieved in the final year.
For instance, the t o p row o f the top panel shows the percent o f persons that were in the lowest income fifth in
1967, delineated by the income fifth they had achieved in
1976. For example, 5.2% o f the people in the sample had
incomes in the lowest fifth in 1967 but in the second fifth
i n 1976.
The data reflect t w o different dimensions o f income
growth—"within-cohort" and "across-cohort." Because the
group examined includes relatively young persons (ages
25 to 34 years old) as well as people already in their peak
e a r n i n g years (ages 45 t o 54 years o l d ) some o f the
observed income mobility is due to the different income
growth rates of the earlier versus later cohorts, i.e., the lifecycle pattern o f the various cohorts. Thus, some o f the
upward income mobility simply reflects the younger individuals experiencing faster income growth than older individuals (55 to 64 years o l d ) . This is the "across-cohort"
dimension.
The "within-cohort" dimension is that w i t h i n any age
g r o u p individuals o r families may have relatively bener o r
worse income growth than their peers. That is, someone
may have a lower (greater) than average income in the
starting year and a greater (lower) than average income at
the ending year. The degree to which individuals change
their income position relative to their peers, either up o r
d o w n , can be considered to be the degree o f "income
mobility."
There are several reasons for interest in income mobility.
One is that an individual's ability to achieve a relatively
higher income, o r move u p the relative income ladder,
reflects a more fluid and open society and greater opportunities for advancement. Moreover, if upward mobility is
more possible there may be a lesser concern about the
growing inequality o f what jobs pay, since there may be a
greater income gap between those in well-paid and poorly
paid jobs because an individual's chance o f getting a well-
93
Some of the
upward income
mobility simply
reflects the
younger
individuals
experiencing
faster income
growth than older
individuals.
�TABLE 1.35
Transitions Into or Out of Middle Income
for Families with Children
Transition*
1967-80
1980-89
8.5%
9.8%
5.8
6.8
35.1%
24.6%
30.8
27.6
Out of middle income:
Fraction of middle-income families
falling to low-income status
Fraction of middle-income families
attaining high-income status
Into middle income:
Fraction of low-income families
attaining middle-income status
Fraction of high-income families
falling into middle-income status
•Income transition of men and women over consecutive five-year periods for those aged 25-50 in the
first year. Income is defined as after-tax incomes including the value of food stamps. Middle incomes
are defined as between $18,500 and $55,000 in 1987 dollars.
94
paid job has improved. Last, it is important to note that
incomes fluctuate both up and down for particular people
and families year by year, a dimension of income change
that can not be captured in the usual cross-section
comparisons.
What does Table 1.34 tell us about income mobility?
First, the minority of families do not stay in the same
income fifth over a ten-year period. For instance, only 30%
of the famUies in the middle fifth (6 0% divided by 20.1%)
in 1967 were still in the middle fifth nine years later in
1976. Second, a large portion of the low-income families
seem to remain low-income over a long period of time—
more than half of the lowest fifth remained in the lowest
fifth over each time period, 1967-1976 and 1977-1986.
Likewise, many of the upper-income families remain upper
income. This rigidity at the top and bonom of the income
ladder, it should be noted, partly reflects the statistical fact
that poor families can only move up and not down and
well-off families can only move down and not up.
Perhaps the most important point to draw from Table
134 is that there has not been much change in income
mobility from the first period to the second period. Thus, a
greater inequality of income "positions" has not been offset by a greater chance of obtaining a bener "position."
The implication is that the widening inequality of incomes
discussed in earlier sections has led to a widening inequality of the lifetime incomes of people and families.
Panel data—tracking the same families over time—can
also be used to examine the mobility of families into and
out of particular absolute income ranges. In Table 1.35,
for instance, panel data are used to show the percentage of
families moving into and out of the "middle class." These
transitions are categorized by whether there was downward mobility (from high to middle income or from middle to low income) or upward mobility (from low to middle income or middle to high income).
The starting point of the analysis is to note, as we did in
an earlier section, that there has been a shrinkage in the
proportion of families with middle-class incomes over the
last two decades. The data in Table 135 allow us to see the
flows of families into and out of the middle class so we can
determine what underlying income trends are driving the
middle-class shrinkage, e.g., are more people moving up
or down the income scale?
95
Perhaps the most
important point to
draw from the
data is that there
has not been
much change in
income mobility
from 1967-1976
to 1977-1986.
�Chapter 2.
Looking at the transitions out of the middle (the top
panel of Table 1.35) we see that in the post-1980 period
families were both more likely to fall out of the middle
class (9.8% of them did versus 8.5% in the earlier period)
and more likely to "climb out" of the middle-class (6.8%
versus 5.8%). The middle class also declined in the recent
The positive period because far fewer low-income families were able to
trends are that inattain middle-class incomes (24.6% versus 35.1%) and a
the post-1980 lesser fraction of high-income families fell into the middle
period middle- class.
income families These data amplify the earlier findings of widening inewere better able quality and income polarization. The positive trends are
to attain high that in the post-1980 period middle-income families were
incomes and better able to attain high incomes and more high-income
families were able to sustain their high incomes. The
more high-income
downside, however, is that more middle-income families
families were ablefell into lower incomes and fewer low-income families
to sustain theirwere able to rise into the middle class.
high incomes.
The downside,Conclusion
however, is that
more middle- There was historically slow income growth over the
income families1979-1989 period and an historically large surge in income
fell into lower inequality. These trends can also be viewed as there having
incomes and been rapid income growth among upper-income families,
' low-income particularly the 62.9% income growth of the top 1%, and
families were able
stagnant or falling incomes for the bottom 80% of families.
to rise into the The remainder of the book elaborates on the themes
middle class. established in this chapter. In the next chapter we focus on
changes in the level and distribution of taxes to examine
the extent to which our tax system has exacerbated or ameliorated the slow and unequal income growth of the
1980s. The third and fourth chapters examine the labor
market trends (wages and employment) which are at the
heart of the growth in inequality and our sluggish income
growth. Chapter 5 and 6 broaden our income analysis by
examining trends in wealth and poverty. Chapter 7 examines income trends by geographic location (region/states,
urban/rural), while Chapter 9 compares trends in the U.S.
to those in other advanced countries.
96
Taxes: A Further Cause of
Worsening Inequality
In Chapter 1, we showed that the pre-tax incomes of
most families have stagnated, while the incomes of wealthy
families have grown significantly. This chapter broadens
our analysis by examining the role played by the tax system
in these developments. The major finding of this chapter is
that federal, state, and local taxes have not ameliorated the
economic problems documented in Chapter 1. In fact,
recent tax changes have worsened the distribution of aftertax income by taxing the middle class and the poor more
heavily and giving large tax cuts to the richest 1%. Since
1977, federal tax changes alone amounted to an average tax
break of *49,262 in 1989 for the richest families. The
causes of this shift include a less progressive personal
income tax, higher payroll taxes, and lower corporate taxes.
There has also been a steadily increasing reliance on state
and local taxes, which tend to be regressive.
The Tax Burden: Still Light By
Comparison
In comparative terms, taxes have increased little and the
overall U.S. tax burden remains one of the lightest among
the Western industrialized countries. Table 2.1 shows the
total tax burden in the U.S. as a percent of Gross Domestic
Product (GDP) from 1959 to 1991, broken out by federal
and state/local contributions. The overall burden has gradually edged up during the postwar period, but has held rela97
Recent tax
changes have
worsened the
distribution of
after-tax income
by taxing the
middle class and
the poor and
giving large tax
cuts to the
richest 1%.
�TABLE 2.1
Federal vs. State and Local Tax Burdens, 1959-1991
Revenue as Percent of GDP*
Federal
State/Local
18.3%
18.7
19.5
20.3
20.1
19.7
7.7%
9.6
11.6
11.0
11.9
12.2
1959
1967
1973
1979
1989
1991
'Gross Domestic Product
98
Total
26.0%
28.4
31.1
31.3
32.0
32.0
lively constant since 1973, staying between 3 1 % and 32%
of GDP. However, the state and local burden grew by 4.5%
of GDP over the total period w h i l e the federal burden grew
by only 1.4% of GDP. Since 1979, the federal burden has
been stable or slightly i n decline (although the 1991 data
may reflect the recession), w h i l e the state and local burden
grew. As we show below, the effect of shifting the tax burden to state and local governments is to widen inequality.
Table 2.2 puts the U.S. tax burden in an international
context by showing government revenues as a percent of
GDP over time for the countries i n the Organisation for
Economic Co-operation and Development (OECD), the
group of advanced industrial countries. The percentage
point changes i n the last column show that since 1967
every country other than the U.S. has had a greater increase
i n the ratio of tax revenues to GDP As a result, the total U.S.
tax burden i n 1988 (the latest year of available data) was
lower than i n any other OECD country save Turkey.
99
Every country
other than the
U.S. has had a
greater increase
in the ratio of tax
revenues to GDP.
As a result, the
total U.S. tax
burden in 1988
was lower than in
any other OECD
country save
Turkey.
�Recent Federal Tax Changes Have
Favored The Very Rich
TABLE 2.2
Tax Revenues in OECD Countries, 1967-1988
As Percent of G P
D
Tax Revenues as % of GDP*
Percentage
Point Change
Country
1967
1979
1988
1967-88
Sweden
37.1%
49.5%
55.3%
18.2
Denmark
33.1
44.5
52.1
19.0
Norway
36.6
Netherlands
35.7
45.7
45.0
46.9
48.2
10.3
12.5
Belgium
33.5
44.5
45.1
11.6
France
34.7
40.2
44.4
9.7
Luxembourg
30.8
40.2
42.8
12.0
Austria
35.3
41.0
41.9
6.6
Ireland
New Zealand
28.7
25.4
31.2
32.9
41.5
37.9
Germany
32.2
37.7
37.4
12.8
12.5
5.2
U.K.
32.9
32.8
37.3
4.4
Greece
23.3
30.1
35.9
12.6
Italy
26.2
26.6
37.1
10.9
Finland
Canada
31.6
28.2
33.3
30.6
37.9
34.0
6.3
5.8
Spain
16.8
23.4
32.8
16.0
Switzerland
21.6
31.1
32.5
10.9
Portugal
19.7
26.0
34.6
14.9
Australia
23.4
27.9
30.8
7.4
Japan
18.3
24.4
31.3
13.0
U.S.
27.3
29.0
29.8
2.5
Turkey
16.2
20.8
22.9
Within the context o f the relatively l o w U.S. tax burden,
there have been important changes i n the d i s t r i b u t i o n of
taxes. These changes have combined w i t h changes i n pretax income to enrich the highest income families—primarily the top 1% —at the expense of almost everyone else. The
redistributive effects o f the changes i n the tax system constitute the primary focus of this chapter.
Family i n c o m e i n e q u a l i t y a f t e r federal taxes has
increased just as it has before federal taxes. Table 2.3 examines this trend by looking at the changes i n after-tax
income (the years chosen for the federal tax analysis i n this
section are dictated by the data source; therefore, non-peak
years appear i n some of these tables). The average after-tax
income of the top f i f t h o f families grew by 28.1% from
1977 to 1989. But the most dramatic g r o w t h occurred
among the most wealthy families of all: the top 1% of the
income distribution. After taxes, their average income grew
by 102.2% over the period i n question. Note that this level
of income g r o w t h exceeds their pre-tax rate of growth
(final c o l u m n ) by 24.2 percentage points, showing that
reduced taxation significantly boosted their after-tax
incomes. Moreover, comparisons of pre- and after-tax
growth i n incomes of each group show the top 1% to be
the o n l y g r o u p w h o s e after-tax g r o w t h significantly
exceeded its pre-tax growth.
6.7
•Social Security Included. United States numbers differ from those in Table 2.1 because the tables
come from different sources.
100
101
It is particularly
striking, in terms
of equity, that it
was the poorest
members of
society who
posted the
biggest
percentage
declines in
after-tax
incomes.
�TABLE 2.3
Average After-tax Family Income*
Income Group
Percent Change in
After-tax Income
Percent
Percent
Change
Change
1980-89
1977-89
1977
1980
1989
All
$30,948
$30,269
$33,663
11.2%
Top Fifth:
Top 1%
14%
Next 5%
64,583
234,516
78,937
81,399
410,148
98,058
65,250
49.959
26.0
74.9
tio%
63,546
202,809
78,820
57,218
45,660
Bottom Four-Fifths:
Fourth
22,582
36,563
55,884
45,170
Middle
27,788
21,694
35,320
26,417
Second
Lowest
18,885
8,495
17,740
8,079
8.8%
Percent
Change
in Pre-tax
Income
1977-89
8.6%
24.2
28.1
102.2
24.4
25.4
78.0
23.1
16.8
10.6
14.0
9.4
14.6
9.7
22,061
37,379
1.7
5.8
-2.3
2.2
-2.8
2.4
26,350
-0.3
-4.2
-5.2
16,987
-5.3
-9.8
7,608
-5.8
-10.1
-10.4
-10.4
'Federal Taxes Only
TABLE 2.4
Shares of After tax Income for All Families, 1977-1989
After-tax income trends were different at the other end
of the income distribution, as the income of the average
family in the b o t t o m 80% of the distribution fell by 2.8%
before federal taxes and 2.3% after federal taxes (Table
2.3). It is particularly striking, i n terms of equity, that it was
the poorest members of society w h o posted the biggest
percentage declines i n after-tax incomes.
The disparity i n after-tax income g r o w t h implies substantial changes i n the distribution of after-tax incomes.
Table 2.4 shows that between 1977 and 1989, the after-tax
income g r o w t h of the most wealthy o c c u r r e d at the
expense of the bottom 90% of the income distribution, all
of w h o m lost income shares to the top 10%. By 1989, the
top 20% of families held almost 50% o f total after-tax
income, up from 44% i n 1977 and 44.9% i n 1980. Again,
most of the gain to the upper quintile was generated by
those in the top percentile, w h o gained 5.1 percentage
p o i n t s i n t h e i r share o f after-tax i n c o m e o v e r the
1977-1989 period, w i t h most of the shift occurring i n the
period between 1980 and 1985. O n the other hand, the
shares received by each of the bottom four fifths and the
second-richest tenth o f families (i.e., the bottom 9 0 % )
declined steadily over the period.
The distributional findings shown i n the previous tables
pose an important question regarding the impact of federal
tax changes: did the major income shifts occur before or
after taxes? Table 2.5 examines this question by comparing
pre- and after-tax income shares, 1977 and 1989
Percentage
Point Changes
Shares of Afler-tax Income
Income Group
1977
1980
1985
1989
All
100.0%
100.0%
100.0%
100.0%
0.0
0.0
44.0
7.3
44.9
8.4
49.7
12.4
4.8
4.0
11.3
9.9
15.6
11.3
9.9
12.3
9.9
15.2
1.0
0.0
-0.1
5.7
5.1
1.0
15.3
49.0
11.2
12.3
10.2
15.4
56.3
22.7
16.3
55.6
22.6
16.2
51.8
22.0
15.3
-4.4
-1.0
11.6
5.7
11.4
5.4
10.1
4.4
51.2
21.6
15.2
10.1
Top Fifth:
Top 1 %
Next 4%
Next 5%
Next 10%
Bottom Four-Fifths:
Fourth
Middle
Second
Lowest
102
4.3
1980-89
-1.0
-1.3
-1.1
1977-8
0.0
-0.4
-5.1
-1.1
-1.1
-1.5
-1.4
103
The after-tax
income growth of
the most wealthy
occurred at the
expense of the
bottom 90% of
the income
distribution, all
of whom lost
income shares to
the top 10%.
�TABLE 2.5
The Effects of Tax and Income Changes
on After-tax Income Shares, 1977-1989*
Change in
Shares Due to:
Pre-tax Shares
<2)
(D
After-tax Shares
(3)
(4)
Change in
After-tax
Shares
(5)
Income Group
1977
1977
1977-89
All
100.0% 100.0% 100.0% 100.0%
1989
1989
Pre-tax
Income
Shifts
(6)
Lesser
Tax
Progressivity
(7)
1977-89
1977-89
0.0
0.0
0.0
Top Fifth:
Top 1%
Next 4%
Next 5%
110%
46.6
8.7
12.0
10.1
15.7
51.1
12.6
12.8
10.3
15.4
43.9
7.3
11.3
9.8
15.5
49.0
11.9
12.2
9.8
15.1
5.1
4.6
0.9
0.0
-0.4
4.5
3.9
0.8
0.1
-0.3
0.6
0.7
0.1
-0.1
-0.1
Bottom Four Fifths:
Fourth
Middle
Second
Lowest
53.4
22.4
15.6
10.5
4.9
48.9
21.4
14.6
9.3
3.7
56.1
22.6
16.3
11.5
5.7
51.0
21.5
15.1
10.1
4.3
-5.1
-1.1
-1.2
-1.4
-1.4
-4.5
-1.0
-1.1
-1.2
-1.2
-06
-0.1
-0.1
-0.2
-0.2
'B
I on recalculation of the underlying data in Table 2.4 which force the sum of shares to be exactly
100%; due to rounding error, this is not Ihe case in the original data.
104
Column five of Table 2.5 shows the difference i n after-tax
income shares that developed over the period in question.
Columns six and seven "decompose" these differences into
pre-tax shifts (generated prior to the imposition of tax liabilities) and into differences caused by the less progressive tax
system (the remainder of the change, after accounting for
pre-tax shifts). The effect of changes i n the tax code are
derived by comparing 1989 after-tax incomes and shares
under 1989 and 1977 effective tax rates.
Column five shows that the b o t t o m 90% of families had a
lower after-tax income share in 1989 than i n 1977 (as
noted i n Table 2.4), w i t h the bottom 80% of families losing
5.1% of total after-tax income over the period. These losses
primarily reflect a lower share of pre-tax incomes. In fact,
4.5 of the 5.1 percentage point decline i n the after-tax
income share of the bottom 80% was due to a relative loss
of pre-tax income, attributable to a more unequal distribution of wages and other incomes. Thus, shifts in the tax burden, though significant, were not the primary mechanism
of the overall redistribution of income i n recent years.
Nevertheless, as shown in column seven, the less progressive tax structure d i d provide a significant 0.7%
increase i n the after-tax income share of the top 1 % . As the
next section shows, a modest increase i n the effective tax
rates among most families permitted a sizable tax reduction
for the top 1 % . However, as noted above, the largest part of
the income gains made by the most wealthy occurred
before taxes, from greater wages and capital incomes, as
seen i n Chapter 1.
105
The average
effective tax rate
on the top
percentile fell by
8.8 percentage
points between
1977 and 1989.
�Table 2.6
Elective Average Federal Tax Rates in 1977,1980, and 1989
Effective
Federal Tax Rates
1977
Income Group
1980
Change in
Tax Rates
1989
1977-89
All
22.8%
23.3%
22.7%
-0.1
Top Fifth:
Top 1 %
Next 4%
Next 5%
Next 10%
27.3
35.6
26.9
25.3
23.9
27.5
31.7
27.9
26.3
25.3
25.6
26.8
26.2
25.6
24.1
-1.7
-8.8
-0.7
0.3
0.2
Bottom Four-Fifths:
Fourth
Middle
Second
Lowest
16.3
21.9
19.5
15.4
9.3
16.5
22.9
19.8
15.6
8.1
16.6
22.0
19.3
15.6
9.3
0.3
0.1
-0.2
0.2
0.0
FIGURE 2A
Federal Tax Burden, 1977 and 1989
0)
e
o
o
c
c
a
>
o
iu
Q-
(>
/
C
O
The Diminished Progressivity of
Federal Tax Rates
Even while most of the income shifting between 1977
and 1989 occurred before taxes, regressive changes in federal taxes since 1977 have contributed to the increased inequality of after-tax family incomes. Table 2.6 lists the effective federal tax rates (effective, or average, tax rates are the
ratio of taxes paid to pre-tax income) by family income
group. The average effective tax rate changed little over the
time period, from 22.8% to 22.7%. What has changed is the
tax rate on different income groups. For example, the rate
on the top percentile fell by 8.8 percentage points, from
35.6% to 26.8%. Conversely, the effective rates on the bottom 95% of families grew slightly or remained unchanged
(with the exception of the middle quintile). Thus it appears
that slightly higher taxes on the bottom 95% paid for a substantial reduction in the tax rate for the top 1%. Figure 2A
portrays the decreased progressivity, 1977-1989, in terms
of average tax rates, at the top of the income distribution of
the 1989 federal tax system.
The structure of marginal rates on income taxes (the tax
rate on the last dollar of income) has also become less progressive at the top of the income distribution. Figure 2B
shows the percent of returns taxed at or below the various
marginal rates in 1979 and 1988 (base years arc slightly different than other figures due to data availability). Note that
in 1979, the top (approximately) 16% of income tax payers
had higher marginal rates than in 1988. Due to the highly
progressive structure of the 1979 top marginal rates (part
of which resulted from inflation), the 1979-1988 rate differential grew rapidly at the very top of the distribution.
w
O
X
to
(0
•o
^
6
^
Family Income Group
106
107
Due to the highly
progressive
structure of the
1979 top
marginal rates
the 1979 1988
rate differential
grew rapidly at
the very top of
the distribution.
�FIGURE 2B
Federal Income Tax: Cumulative Returns by Marginal Rates,
1979 and 1988
40
50
60
70
Cumulative Percent of Tax Returns
108
80
90
too
What Federal Tax Changes
Mean in Dollars
The effect of any change in a group's tax rate on their
actual, after-tax income depends on both the size of the
change i n the tax rate and o n their level of income. Table
2.7 shows the average pre-tax family income in 1989 for
each income group ( c o l u m n one, in constant 1992 dollars).
Columns t w o through four show the changes i n the average amount of taxes paid by each income group due to federal tax changes enacted i n different periods between 1977
and 1989, assuming that incomes were maintained at their
1989 levels. The final c o l u m n shows the cumulative effect
of changes i n the tax code between 1977 and 1989 on
1989 federal tax payments.
Because of federal tax changes since 1977, the top 1% of
families each received an average tax cut of 849,262 i n
1989. The next richest 4% of families received tax cuts that
averaged S924 per family i n 1989—considerably smaller
than the gains of the richest 1 % . Since their effective rates
changed little over the period, the bottom 95% of families
either lost or gained marginally. The poorest families
gained income from tax changes in the late 1980s, but
these gains were 'washed out' by regressive changes over
the total period. Clearly, the distributional impact of the
rate changes was primarily driven by the gains made by the
most wealthy. Their tax cuts came mainly from steps taken
between 1980 and 1985, though changes i n the late 1970s
also contributed. Tax changes since 1985, including the Tax
Reform Act of 1986, reduced the tax cut somewhat.
109
The poorest
families gained
income from tax
changes in the
late 1980s, but
these gains were
'washed out' over
the total period.
�The Causes of Changes in the
Federal Tax Burden
TABLE 2.7
Effect of Federal Tax Changes o Family
n
Tax Payments, 1977-1989
The Effect of Federal Tax Changes, 1977-89 on the
Income Group
1989 Pre-tax
Family Income
1977-80
1980-85
1985-89
Total
1977-89
$43,495*
$217
$-652
$391
$-43
All
Top Fifth:
109,424
219
-3,611
1,532
-1,860
Top1%
559,795
-21,272
-38,626
10,636
-49.262
Next 4%
132,036
1,452
-4,885
2,509
-924
Next 5%
87,711
965
-1,842
1,140
263
Next 10%
65,900
857
-1,120
395
132
26,841
54
54
-27
81
Bottom
Four-Fifths:
Fourth
47,913
479
-575
144
48
Middle
32,681
98
-229
65
-65
Second
20,140
40
20
-20
40
Lowest
8,391
-101
185
-84
0
'Constant 1992 Dollars.
110
In order to understand the causes of the regressive
changes i n the federal tax system. Table 2.8 decomposes
the federal effective rates into their main components: the
personal income tax, payroll taxes, the corporate income
tax, and the excise tax.
The personal income tax is progressive; the effective
rates go up consistently as income rises. However, it has
become less progressive, a point taken up i n the next table.
Since 1977, the overall payroll tax rate has risen from
6.5% to 8.5% (Table 2.8). The b o t t o m 99% have borne the
brunt of increased payroll taxes; the top 1% has been virtually unaffected. This is partly because the rich earn much of
their income from their investments (Chapter 1), w h i c h
are not subject to payroll taxes. Another reason is that payroll taxes are 'capped': they apply only to the first 555,500
(1992 dollars) of earnings i n 1992. The payroll tax was the
largest federal tax paid by the b o t t o m 80% of the distribution i n 1989- As w e discuss below, the g r o w t h of payroll
taxes has led to greater inequality.
I n this analysis, the tax incidence (i.e., w h o ultimately
bears the burden of the tax) of the payroll tax falls totally
on employees: the employer p o r t i o n is shifted to employees
i n the f o r m of lower wages. While this is a common and
valid assumption of payroll tax incidence, it does not speak
to the redistributive effects o f the social security system. In
the current context, retirement benefits are redistributed
i n a progressive manner: less wealthy retirees receive more
benefits than they contributed over their lifetime, while
the most wealthy receive fewer benefits then they put in.
This redistributive effect counteracts the regressive nature
of the tax documented i n the text.
in
Since 1977, the
overall payroll
tax rate has risen
from 6.5% to
8.5%. The
bottom 99% have
borne the brunt of
increased payroll
taxes; the top 1%
has been virtually
unaffected.
�TABLE 2.8
Effective Tax Rates for Selected Federal Taxes, 1977-1989
Effective Tax Rates
Personal
Income
Tax
Income Group
1977
1989
Payroll
Tax
1977
1989
Corporate
Income
Tax
1977
1989
Excise
Tax
1977
1989
All
11.1%
11.0%
6.5%
8.5%
3.9%
2.3%
1.3%
0.9%
Top Fifth:
Top 1 %
Next 4 %
Next 5%
16.1
25.2
16.6
5.3
1.3
4.4
7.0
1.6
6.8
6.5
7.4
9.6
10.0
5.0
8.8
5.1
3.9
3.4
2.8
4.2
2.8
2.2
2.0
0.9
0.3
0.8
13.9
12.0
15.3
20.7
16.2
13.1
11.3
0.5
0.3
0.4
0.7
0.8
4.7
9.6
4.2
8.9
7.1
7.8
9.3
10.3
2.7
3.2
6.9
3.4
-0.6
6.4
3.2
8.1
7.5
5.1
10.0
9.4
7.6
3.0
2.7
1.6
1.9
1.8
1.5
1.1
1.9
1.3
1.5
1.8
2.9
Next 10%
1.0
1.1
Bottom
Four-Fifths:
Fourth
Middle
Second
Lowest
-1.8
112
1.9
1.5
0.9
1.1
1.5
2.4
However, the age structure of the population is such that
the worker/beneficiary ratio w i l l fall in the future This
demographic phenomenon w i l l severely strain the social
insurance system's ability to maintain its progressive distributive structure. Without this distributive component,
the burden of the tax falls on wage earners in the present,
and the regressive nature of the tax is not mitigated for
future generations. Therefore, the rise i n payroll taxes is
considered a regressive change.
Corporate income taxes, w h i c h have a progressive structure, have declined considerably. For example, the average
effective rate for the richest f i f t h of families fell from 5.0%
i n 1977 to 2.8% i n 1989 (Table 2.8). Since corporations are
owned by households through shares of stock, households
also bear the burden of corporate taxes. The burden can be
transferred to households in a variety of ways. In this analysis, it is assumed that half of corporate taxes are borne by
stockholders i n the f o r m of lower dividends and slower
stock appreciation, and half are paid by consumers i n the
form of higher prices.
Excise taxes have fallen since 1977 w h i l e maintaining
their regressive structure. These are taxes levied on alcohol, gasoline, cigarettes, etc.; they are the most regressive
federal taxes (Table 2.8, F i g u r e 2C). In 1989, the poorest
families paid 2.4% of their income i n excise taxes while the
richest f i f t h paid 0.5%. This is because families w i t h higher
incomes spend smaller proportions of their incomes on
gasoline, cigarettes, alcohol, and other goods subject to
excise taxes.
113
The burden of the
tax falls on wage
earners in the
present, and the
regressive nature
of the tax is not
mitigated for
future
generations.
�FIGURE 2C
Excise Tax Burden,
1977 and 1989
Although the personal income tax is still the most progressive federal tax, it has become less so over time (Table
2.9, F i g u r e 2D). Since 1977, the average personal income
tax for the richest families (the top 1%) has fallen by 4.5
points. In contrast, reductions in personal income taxes for
groups in the bottom 99% have been much smaller
Except for the top 1 % , the reductions i n income taxation
have been more than offset by higher payroll taxes. For
instance, the effective income tax rate for the middle f i f t h
fell 0.5 percentage points between 1977 and 1989, but the
payroll tax grew by 1 9 percentage points over the same
period.
B e t w e e n 1977 and 1989, federal c o r p o r a t e taxes
declined by 1.6 percenuge points (Table 2.9). According to
our burden assumptions, corporate taxes arc progressive;
thus, lower corporate taxes have benefitted the rich more
than other groups. In particular, the top 1 % of families saw
their corporate tax burdens reduced by 4.6 percentage
points, w h i c h accounts for more than half the overall 8.8
percenuge point decline i n the effective tax rate for the
top 1 % .
The 0.4 percentage point decline i n federal excise taxes,
from 1.3% of family incomes i n 1977 to 0.9% i n 1989, has
reduced the tax burden somewhat on the bottom 99% of
families, w i t h slightly greater tax cuts going to low-income
families. Nevertheless, the average family i n the bottom
80% saw their total tax burden increase by 0.3 percentage
points.
Family Income Group
TABLE 2.9
Changes in Effective Federal Taxes, 1977-1989
Point Change, 1977-89
Personal
Income
Tax
Income Group
Payroll
Tax
Corporate
Income
Tax
Excise
Tax
Total
All
-0.1
2.0
-1.6
-0.4
-0.1
Top Fifth:
Top 1 %
Next 4%
Next 5%
Next 10%
-0.8
-4.5
-0.4
-0.8
-0.7
1.7
0.3
2.4
3.1
2.6
-2.2
-4.6
-2.3
-1.7
-1.4
-0.4
0.0
-0.4
-0.3
-0.3
-1.7
-8.8
-0.7
0.3
0.2
Bottom Four-Fifths:
Fourth
Middle
Second
Lowest
-0.5
-0.7
-0.5
-0.2
-1.2
2.2
2.5
1.9
1.9
2.5
-1.1
-1.3
-1.2
-1.2
-0.8
-0.4
-0.4
-0.4
-0.3
-0.5
0.3
0.1
-0.2
0.2
0.0
114
115
Although the
personal income
tax is still the
most progressive
federal tax, it has
become less so
over time. Since
1977, the
average personal
income tax for
the richest
families has
fallen by 4.5
points.
�FIGURE 2D
Personal Income Tax Burden,
1977 and 1989
Changes in Corporate Taxation:
The Shift to Untaxed Profits
30
25 -
1977
/
*
*
o
V)
10
E
8
c
.IT
-5
i
i
i
i
i
J
i
i
VI.
Family Income Group
TABLE 2.10
Taxed and Untaxed Corporate Profits, 1977-1991
(Nonfinancial Corporations Only)
As Percent of GDP in:
1977
1980
1986
1989
1991
8.2%
8.0
0.2
1.6
-0.5
-0.8
6.6%
6.7
-0.1
2.1
-0.7
-1.6
7.1%
3.5
3.6
2.3
1.1
0.2
7.7%
4.8
2.9
2.7
0.5
-0.3
7.8%
3.6
2.6
2.6
0.0
0.1
Corporate Profits Taxes*
3.0
2.5
1.8
1.9
1.5
Profits After Taxes
Taxable Profits Only
All Profits
5.0
5.2
4.2
4.1
1.7
5.3
2.9
5.8
Though corporace rates have dropped, this does not necessarily lead to lower corporate tax liability, since the lower
rates could be applied to greater absolute profits (i.e., an
expanded tax base). In fact, this was a goal of the 1986 tax
reform. However, contrary to the goal of the reform, corporate tax revenue did decline through the period due to the
growth o i untaxed corporate income.
Since 1980, corporate profits as a percent of GDP have
consistently risen, from 6.6% to 7.8% in 1991 (Table
2.10). However, taxed profits fell significantly from 1980 to
1986 (6.7% to 3 5%) and from 1989 to 1991 (4.8% to
3 6%). This phenomenon of rising total profits and falling
taxed profits is due to the rise in untaxed corporate profits,
which rose from 0.2% of GDP in 1977 to 2.9% in 1989
(Table 2.10, Figure 2E). Over the full period (1977 to
1989), corporate taxed profits have been cut virtually in
half, from 8.0% of GNP to 4.8%.
There are two main reasons for the shift from taxed to
untaxed profits.- increased indebtedness (along with higher
real interest rates) and a more favorable tax treatment of
depreciation. Since 1977, corporations have increasingly
raised money by selling bonds (i.e., they have borrowed)
rather than by selling stocks to raise money. As a result, corporations have been paying out more and more of their
gross profits as interest payments (on bonds) rather than as
dividend payments (on stocks). This is reflected in Table
2.10 in the steady increase in net interest payments, from
1.6% of GDP in 1977 to 2.7% in 1989.
Since interest payments are treated more favorably than
dividends by the federal tax system (they are considered an
expense for non-financial corporations and deducted from
taxable profits), and since interest rates have been high relative to inflation, the corporate emphasis on debt over
equity has lowered the corporate tax burden. Corporate liability is lowered, as corporations deduct interest payments
from their tax base.
2.1
6.3
All Profits, Taxed and Untaxed
Taxed profits
Untaxed profits
Net Interest
Depreciation Allowances
Other Deductions
"Federal, slate and local combined.
116
117
The phenomenon
ofrisingtotal
profits and falling
taxed profits in
the corporate
sector is due to
the rise in
untaxed
corporate profits,
which rose f mm
0.2% of GDP in
1977to 2.9% in
1989.
�FIGURE 2E
Taxed and Untaxed Corporate Profits as a Percent of GDP,
1946-1991
15
10
Q.
a
O
£
Untaxed Protits
^
^
^
,96* ^ 1 , 10 ^
9
jgfi
TABLE 2.11
Corporate Profits Tax Rates, 1947-1991
(Nonfinancial Corporations Only)
^
^
^
The second reason for lower corporate taxes is the federal tax system's relatively favorable treatment of capital
depreciation. With the introduction of the Accelerated Cost
Recovery System in 1981, corporations were able to take
tax deductions on purchases of new equipment and other
investments sooner. Since this allowed them to defer some
of their taxes for several years, accelerated depreciation
amounted to an interest-free loan. The main tax break of
this type came between 1980 and 1986, when untaxed
profits due to depreciation allowances went from -0.7% of
GDP to 1.1% (Table 2.10). Since 1986, Congress has taken
back part of this tax cut, reducing untaxed profits in this
category to 0.5% of GDP in 1989.
Table 2.11 shows the results of these changes. Since
1977, corporate taxes have declined from 36.8% of actual
profits (defined as the sum of taxable profits, net interest,
and the difference between allowable depreciation and
true depreciation) to 24.6% in 1989, while generally
increasing as a percent of taxed profits. Moreover, the 1989
effective tax rate, 24.6%, is barely over half of what corporate taxes were in 1957 (Table 2.11; Figure 2F). The beneficiaries of this fall in corporate tax liability relative to true
economic profits are by and large those who own and lend
to corporations: the wealthy.
Corporate Taxes*
as Percent of:
Taxable
Profits
1947
1957
1967
1977
1980
1985
1989
1991
37.1%
48.0
41.2
37.8
36.9
42.2
39.4
41.1
Actual
Profits
50.9%
48.6
34.5
36.8
37.4
22.4
24.6
24.0
'Federal, stale, and local combined.
118
119
The 1989
effective
corporate tax
rate, 24.6%, is
barely over half
of what corporate
taxes were in
1957.
�FIGURE 2F
Coporate Profits Taxes
1946-1991
The Shift to State and Local Taxes
70
60
;
\
£
As Percent ol
Taxed Profits Only «
«
1= 50 -
40 -
-
As Percent of
All Profits
20
i
i
i
^ ,S
^9 '
^
i
^
i
^
120
\
i
^
r
.
i
^
^ 1
^10 ^ 1 1 ^ 6
^19 ^
i
^
i
, «,«
9
i
^
As noted in the beginning of this chapter, the proportion
of total tax receipts contributed by state and local taxpayers has grown over time, relative to the federal tax burden (see Table 2.1). Figure 2G, total tax receipts as a percentage of GNP, by different components, makes this point
graphically. The figure shows that, starting in the post-war
period, the growth of tax receipts has been led by state and
local taxes (as well as federal payroll taxes), while the federal income tax shows a relatively flat trend. Since state and
local taxes are less progressive than federal taxes, the distribution of the overall tax burden is considerably, and
increasingly, less progressive than the distribution of federal taxes alone.
State and local governments rely predominantly on
regressive taxes for their revenues: sales taxes, property
taxes (here regressivity is sensitive to incidence assumptions, discussed below), and nontax revenues (regressive
when they apply to consumption) such as fines and fees.
Table 2.12 gives the effective rates for state and local taxes
by family income group for a family of four. The regressive
structure of the state and local burden is shown in the
table's final column: the proportion of income that a family
of four paid in state and local taxes fell as their income
increased.
121
Since state and
local taxes are
less progressive
than federal
taxes, the
distribution of the
overall tax
burden is /ess
progressive than
the distribution of
federal taxes.
�FIGURE 2G
Total Tax Receipts As a Percent of GNP,
1930-1990
TABLE 2.12
Total State and Local Taxes in 1991 (Effective Rates)
as Shares of Income for Families o Four
f
Income Group
Personal
Income
Tax
Corporate
Income
Tax
Property
Tax
Sales
Tax
Excise Total
Tax
Tax
Total After
Federal
Deductions
Top Fifth:
Top 1%
Next 4%
Next 15%
4.6%
4.1
3.5
0.4%
0.2
0.1
1.3%
2.3
2.7
1.2%
1.8
2.4
0.1% 7.6%
0.3
8.7
0.5
9.2
6.0%
6.9
7.7
Bottom Four-Fifths:
Fourth
Middle
Second
Lowest
3.0
2.6
2.0
0.7
0.1
0.1
0.1
0.1
2.9
3.2
37
5.4
29
3.3
4.0
5.7
0.7
0.8
1.1
1.9
However, as w i t h federal income and corporate taxes
(Table 2.8), state and local income and corporate taxes are
progressive. Table 2.12 shows that the average family at the
bottom of the income distribution paid 0.7% of its income
i n state and local personal income taxes, w h i l e the wealthiest 1% of families paid 4.6%. The state corporate tax is flat
throughout most of the distribution, then slightly progressive at the top.
Conversely, property, sales, and excise taxes are regressive. In the case of sales and excise taxes, their regressivity
results from the fact that the percenuge of income consumed falls as income increases. This burden was exacerbated i n 1986 w h e n federal deductibility of state sales taxes
was removed. In 1991, the poorest families i n the average
sute paid 5.7% of their income i n sales taxes and 1.9% i n
excise taxes. This proportion fell as income increased, until
the wealthiest families paid a combined 1.3%.
The calculation of p r o p e r t y tax incidence is particularly
sensitive to the assumptions upon w h i c h the analysis is
based. To the degree that owners of land and structures are
able to shift the property tax burden onto tenants, the tax is
regressive. Conversely, since ownership of land and capital
generally rises w i t h income, the property tax is progressive
to the extent that the owners of these factors are unable to
shift the tax forward onto property users.
The property tax column i n Table 2.12 assumes that
homeowners bear the f u l l burden of their property taxes.
For residential renters, half of the property tax is allocated
to renters and half to owners. These t w o assumptions lead
to the regressive structure reflected i n the table, since
poorer homeowners and renters devote a larger proportion
of income to housing than the wealthy. The property tax
liability on business is assumed to be that part of state property tax revenue n o t accounted for by residential renters
and homeowners.
8.4
9.5
10.7
13.8
122
9.5
10.0
10.9
13.8
123
In 1991, the
poorest families
in the average
state paid 5.7%
of their income in
sales taxes and
1.9% in excise
taxes. This
proportion fell as
income
increased, until
the wealthiest
families paid a
combined 1.3%.
�TABLE 2.13
Federal vs. State & Local Taxes, 1991
as Percent of Revenue at Each Level
Type of Tax
Federal
Progressive
Personal Income Tax
Corporate Income Tax
Estate/Gift Taxes
51.1%
40.9
Regressive
Excise/Customs/Sales/Other*
Contributions for Social Insurance
Property
State &
Local
19.1%
16.0
9.2
3.1
1.0
0.0
47.5
67.8
* Other taxes include vehicle licenses, severance taxes, etc.
"Fines, certain fees, rents, royalties, tuition, hospital fees, etc.
23.4
0.0
13.1
100.0
Over 67% of
state/local
revenues are
raised from
regressive taxes,
compared with
less than half
(47.5%) of
federal revenues.
8.6
100.0
Total
35.8
1.5
Nontaxes*
5.7
41.8
The shift to state and local governments' revenues asseen i n Figure 2G is of particular concern from a distributional perspective because state and local taxes are more
regressive than federal taxes. In spite of increases in social
insurance taxes, i n 1991, 5 1 1 % of federal revenues still
came from progressive taxes, but only 19.1% of state/local
revenues were raised progressively (Table 2.13, F i g u r e s
2 H a n d 21). (Payroll taxes at the state level are not strictly
comparable w i t h the federal version, since these contributions at the state level are mostly for pension funds for state
and local employees. As such, they are more like personal
assets than federal payroll taxes w h i c h arc distributed
under a pay-as-you-go system. In this regard, they are Jess
regressive than federal payroll taxes.) Over 67% of state/
local revenues arc raised from regressive taxes, compared
w i t h less than half (47.5% ) of federal revenues.
FIGURE 2H
Federal Revenue Sources,
1991
— Non-tax Revenues (1.5)
Payroll Taxes (41.8)
Excise/Customs/Sales Taxes (5.7)
Estate/Gift Taxes (1.0)
Corporate Income Taxes (9.2)
Personal Income Taxes (40.9)
124
125
�FIGURE 21
State and Local Revenue Sources,
1991
Table 2.14 shows the percent of GDP taxed by progressive and regressive taxes, together w i t h nontax revenues, at
peaks of business cycles since 1959. W h i l e progressive
taxes varied slightly since 1959, regressive taxes have
• increased their share of GDP substantially—from 12 6% to
17.0% i n 1989. This has resulted from increases at both the
federal and state/local levels. Nontax revenues have risen
but remain low, at only 1.7% of GDP i n 1989.
Non-tax Revenues (13.1)
Property Taxes (8.6)
Conclusion
Payroll Taxes (23.4)
The g r o w t h of income inequality documented i n Chapter 1 has not been ameliorated by either federal or state
taxes. I n fact, during the 1980s the tax system has generally
worked to the advantage of the wealthy, w h o received large
tax cuts at the expense of the poor and middle class. The
causes of these tax changes were a less progressive tax rate
structure, a rise i n untaxed corporate profits, and a shift to
greater dependence on regressive state and local taxes.
Excise/Customs/Sales Tax (35.8)
Corporate Income Taxes (3.1)
Personal Income Taxes (16 0)
TABLE 2.14
The Composition o Taxes, 1959-1991
f
As Percent of G P
O
Type of Tax
1959
1967
1973
1979
1989
1991
Progressive*
Federal
State and Local
13.3%
12.6
0.7
13.0%
11.9
1.1
13.5%
11.7
1.8
14.3%
12.2
2.1
13.3%
10.9
2.4
12.4%
10.1
2.3
Regressive**
Federal
State and Local
12.6
5.6
6.9
14.5
6.6
7.9
16.5
7.8
8.8
15.7
7.9
7.8
17.0
9.0
8.0
17.6
9.4
8.3
Nontaxes"*
Federal
State and Local
0.2
0.1
0.1
0.8
0.2
0.7
1.1
0.1
1.0
1.3
0.2
1.1
1.7
0.2
1.5
1.9
0.3
16
* Personal and corporate income taxes; estate and gift taxes.
" Customs, excise, sates, and other taxes; property taxes; contributions for social insurance. Other
taxes include vehicle licenses, severance taxes, etc.
• " Fines, certain fees, rents, royalties, tuition, hospital fees, etc.
126
127
During the 1980s
the tax system
has generally
worked to the
advantage of the
wealthy, who
received large
tax cuts at the
expense of the
poor and middle
class.
�Chapter 3-
Wages: Working Longer
for Less
Introduction
Wage trends have been the primary determinant of the
slow growth in income and the greater inequality of
income we have experienced in recent years. This should
not be surprising since wages and salaries comprise
roughly three-fourths of total family income and an even
higher percentage of the incomes of the broad middle class.
From 1979 to 1991, the real hourly wage of most workers severely eroded. This is true whether we examine
wages or total compensation (= both wages and benefits).
The groups experiencing the greatest fall in wages include
the three-fourths of the workforce with less than a college
degree, the bottom eighty percent of men, low-wage
women workers, younger workers and blue-collar workers. The result of this widespread wage erosion was a sizeable growth in the 1980s in the proportion of the workforce earning less than poverty level wages and
proportionately fewer workers earning mid-level wages.
Starting in the final stages of the 1980s recovery, wages
began falling for groups that had previously escaped the
: downward pressure on wages, including white-collar and
college-educated workers and middle- and high-wage
women. Meanwhile, the wages of those who lost the most
in the 1980s (high school graduates, blue collar workers
and men) have continued to fall. By the end of the 1980s
the only group seemingly exempt from falling real wages
was the 8% of the workforce with advanced or professional
degrees.
129
Starting in the
final stages
of the 1980s
recovery, wages
began falling for
groups that had
previously
escaped the
downward
pressure on
wages, including
white-collar and
college-educated
workers and
middle- and
high-wage
women.
�The recent fall-off in the wages of white-collar and college-educated workers occurred prior to the early 1990's
recession and can be expected to continue even after the
economy recovers, especially given the expected slow rate
of future growth. The wage decline among college-educated workers is due to a slower growth in employer demand,
as reflected in the historically high levels of white-collar
unemployment and the historically slow growth in whitecollar employment in recent years (see Chapter 4). ProYoung workers
jected employment trends for the 1990s suggest continued
face a job market slow growth in white-collar employment and sluggish
providing demand for college-educated workers (see Chapter 4 ).
considerably The most severe wage reductions, however, have been
lower wages thanfor entry level jobs for young, high school graduates, a
comparably group comprising two-thirds to three-fourths of all young
educated youngworkers. In 1991, the wages paid to young male and
people faced twofemale high school graduates were, respectively, 26.5% and
decades ago. 15.4% less than the wages their counterparts received in
1979. Over the 1973-1991 period, the wages of entry level
college graduates fell by 9.8% while the wages of entry
level high school graduates dropped 25.4% (29 3% among
men and 20.4% among women). Thus, young workers face
a job market providing considerably lower wages than
comparably educated young people faced two decades
ago.
There have been several consequences of falling wages.
The most significant is that in order to maintain or improve
their standard of living, workers are putting in more hours
per week as well as working more weeks per year and having more members of the family employed.
Why have these changes in the wage structure occurred?
Wage growth would have been stronger had productivity
growth been robust, but slow productivity growth cannot
wholly explain stagnant or falling real hourly pay (either
wages or compensation), nor the widening of the gap
between high- and low-wage workers and college-educated
and high school-educated workers. Part of the explanation
is the effect of technological change and other factors
which are expanding the relative demand for more educated workers. Yet, despite this expanding demand, the
wages of college-educated workers grew only a modest
1.8% from 1979 to 1989 and fell from 1987 to 1991. We
find that the forces depressing the wages of the vast majority of workers include a lower minimum wage, increased
130
import competition, fewer and weaker unions and the
employment shift to low-wage industries.
More Hours and Stagnant Wages
To understand changes in wage trends it is important to
clearly distinguish between trends in annual, weekly and
hourly wages. Trends in annual wages, for instance, are
driven by changes in both hourly wages and the amount of
time spent working (weeks worked per year and hours
worked per week). Likewise, weekly wage trends reflect
changes in hourly pay and weekly hours. In this chapter
we locus on the hourly pay levels of the workforce and its
subgroups. We do this to be able to distinguish changes in
earnings as a result of more (or less) work rather than
more (or less) pay. Chapter 4 addresses employment,
unemployment, underemployment, and other issues
related to changes in work-time and opportunities.
Table 3-1 illustrates the importance of distinguishing
between annual, weekly and hourly wage trends. The
annual wage and salary of the average worker in inflationadjusted terms was 8.2% greater in 1989 than in 1979.
However, much of this growth in annual wages was due to
longer working hours. For instance, the average worker
worked 1,785 hours in 1989, or 78 hours more than the
1,707 hours worked in 1979. This increase is equivalent to
each worker working roughly two additional full time
weeks in 1989 than in 1979. The end result was that the
average worker in 1989 worked 4.6% additional hours at
an hourly wage that was only 3 7% more than in 1979. Any
"wage" analysis which focuses on annual wages would miss
the fact that more than half the growth in annual wages
from 1979 to 1989 was due to more work rather than
higher hourly wages.
The wage and hour trends in the 1980s—more hours
and stagnant hourly wages—represent a continuation of
the trends in the 1973-1979 period, when hourly wages
were essentially flat (falling 0.1% annually) while work
hours and the number of weeks worked per year were
increasing. In contrast, real hourly wages rose 2.9% annually (and annual hours declined 0.4% annually) between
1967 and 1973. Thus, the post-1973 trend of greater work
effort with modestly rising or falling wages replaced a
131
We focus on the
hourly pay It
of the workforce
and its subgroups
to be able to
distinguish
changes in
earnings as a
result of more (or
less) work rather
than more (or
less) pay.
�TABLE 3.1
Trends in Average Wages and Average Hours, 1967-1990
Real Wage
Hours Worked Trends
Indices (1987=100)
Year
Productivity
Per Hour
(1987=100)
Annual
Wages
Weekly
Wages
Hourly
Wages
Annual
Hours
Weeks
Per
Year
Hours
Per
Week
1967
1973
1979
1989
1990
78.9
88.9
92.2
99.9
100.1
81.5
94.8
95.5
103.3
100.1
84.2
98.2
98.0
102.1
99.1
83.5
99.3
98.5
102.1
99.1
1,720
1,683
1,707
1,785
1,781
42.8
42.7
43.1
44.8
44.7
40.2
39.5
39.7
39.9
39.9
(Annual Growth Rates)
Change
1967-73
1973-79
1979-89
1989-90
2.0%
0.6
0.8
0.2
2.5%
0.1
0.8
-3.1
2.6%
-0.0
0.4
-2.9
2.9%
-0.1
0.3
-2.9
-0.4% - 0 . 1 %
0.2
0.2
0.4
0.4
-0.2
-0.2
-0.3%
0.1
0.1
-0.1
TABLE 3.2
Changes in Hourly Wages, Benefits and Compensation, 1966-1992
(1991 Dollars)
Private Sector
Employer Cost Per Hour
Year
Wages &
Salaries
Health &
Pension
Payroll
Taxes
Total
Compensation
$0.61
1.03
1.43
1.45
1.40
1.53
$0.73
0.91
1.22
1.36
1.40
1.43
$13.17
15.88
17.15
16.21
15.79
15.67
Pay Level
1966
1972
1977
1987
1989
1992
$11.83
13.94
14.50
13.41
12.98
12.70
Percent Change
1966-72
1972-77
1977-89
1989-92
17.8%
4.1
-10.5
-2.2
68.5%
38.4
-1.7
9.3
25.2%
33.5
15.5
1.7
20.5%
8.0
-7.9
-0.7
'Based on available data since peak year data are not available. Data for 1987, 1989 and 1992 are for
March.
132
trend o f strong real annual wage growth based o n higher
real hourly wages which allowed for reductions i n w o r k
time.
Other measures of wage trends presented below suggest
that hourly wages were falling over the 1979-1989 period,
especially i n the latter part of the 1980s. Thus, the picture
presented by Table 3.1 may be somewhat optimistic for the
1979-1989 period.
The onset of rising unemployment i n 1989 led to a
reduction i n both w o r k i n g time and real hourly wages by
1990, resulting i n a 3.1% d r o p i n real annual wages. As w e
w i l l show below, the decline i n hourly wages evident i n the
early 1990s recession reflects both the cyclical d o w n t u r n
and a long-term d r o p i n wage g r o w t h that began several
years before the onset of the recession.
Productivity g r o w t h between recent cyclical peaks has
been less than that of the pre-1973 economy (Table 3 1).
However, lower productivity g r o w t h is not sufficient to
explain the modest g r o w t h i n real wages from 1979-1989.
After all, hourly productivity actually increased by 0.8%
annually between 1979 and 1989 w h i l e real hourly wages
grew by only 0.3%. We w i l l examine explanations for
recent wage trends below.
The data i n Table 3.2 provide a more comprehensive
portrait of hourly pay trends showing more up-to-date
information on trends i n hourly wages, benefits and total
compensation. Unfortunately, there arc no data available
between the years of 1977 and 1987, so our analysis must
necessarily deviate from an examination of peak-to-peak
trends (although one w o u l d expect more favorable wage
trends between 1979 and 1989 than between 1977 and
1989).
Between 1977 and 1989 there was a 10.5% erosion i n
hourly wages and a 7.9% d r o p i n hourly compensation.
Both hourly wages and compensation continued to fall i n
the recession years from 1989 to 1992 d r o p p i n g 2.2% and
0.7% respectively. It is important to note that hourly wages
and benefits began falling before the onset o f the recession, so the recent decline cannot be ascribed to a cyclical
weakness i n the labor market. I n fact, hourly wages and
compensation fell faster i n the t w o years before unemploym e n t started rising i n 1989. Between 1987 and 1989,
h o u r l y wages a n d c o m p e n s a t i o n fell 3.2% a n d 2.5%,
respectively. This suggests that the post-1989 erosion o f
133
Between 1977
and 1989 there
was a 10.5%
erosion in hourly
wages and a
7.9% drop
in hourly
compensation.
�TABLE 3.3
Hourly and Weekly Earnings of
Production and Non-supervisory Workers, 1947-1991*
(1991 Dollars)
Year
Average
Hourly
Earnings
Average
Weekly
Earnings
1947
1967
1973
1979
1982
1989
1991
$6.37
10.06
11.37
11.34
10.94
10.61
10.34
$256.53
382.11
419.54
404.75
380.76
367.12
354.66
Changes (Annual Rates of Growth)
Peak to Peak
1947-67
1967-73
1973-79
1979-89
2.3%
2.0
0.0
-0.7
2.0°/
1.6
-0.6
-1.0
-1.2
-1.3
-2.0
-1.7
-0.4
-0.5
Contractions
1979-82
1989-91
Recovery
1982-89
"Production and non-supervisory workers comprise more than 80% ot wage and salary employmenl.
134
c o m p e n s a t i o n reflects l o n g - t e r m rather than cyclical
trends.
The erosion of real hourly wages and compensation since
1977 represents a dramatic reversal of their strong g r o w t h
between 1966 and 1977. From 1966 to 1972 there was a
17.8% g r o w t h i n real wages, a 2.7% annual rate of growth.
Because health and pension benefits grew by 68.5% i n this
period, hourly compensation grew even faster than wages.
Both wages and benefits grew more slowly between 1972
and 1977, but still a respectable 4 . 1 % and 8.0% increase.
Some analysts have presumed that fringe benefit g r o w t h
in recent years has balanced declining wages, leaving overall compensation g r o w i n g slowly. This reasoning seems
plausible since it is well-known that health-care costs have
been rising rapidly (see Chapter 8 ) and it is believed that
fringe benefits comprise a large share of total compensation, perhaps as high as 40 to 45%.
I n actuality, benefits are not as important in the overall
compensation package as many people believe, nor have
the benefits been rising rapidly in recent years. The data in
Table 3.2 indicate that over the 1977 to 1989 period the
cost o f fringe benefits, measured as employer pension and
insurance costs per hour, did not grow, and that they comprised less than 10% o f the total compensation package in
1989.
It is certainly true that health insurance costs have risen
quickly. Apparendy the rapid growth o f jobs w i t h litde o r
n o employer-provided health benefits and the increased
shift o f employer health-care costs o n t o employees has
meant that average fringe benefit costs d i d not rise over the
1977 to 1989 period. I n fact, fringe benefits declined m o d esdy (-1.7%) over this period.
Part of the confusion about the role of fringe benefits is
definitional. I n surveys of employers by trade associations
and by the Bureau of Labor Studies, fringe benefits are
broadly defined ( f o l l o w i n g standard corporate accounting
procedures) to include paid leave (holidays and vacations),
supplemental pay (overtime and shift premiums), and payroll taxes (employer social security and unemployment
taxes). Under this broad definition, benefits do comprise
about 28% of total compensation costs. However, wagerelated items that are received by workers in their regular
paychecks, such as paid leave and supplemental pay, are
135
Over the 1977 to
1989 period the
cost of fringe
benefits,
measured as
employer pension
and insurance
costs per hour,
did not grow.
�FIGURE 3A
Real Hourly Earnings of Production and Nonsupervisory Workers,
1967-1991
12.50
1967
1970
1973
1976
1979
1982
1985
1988
1991
FIGURE 3B
Weekly Earnings of Production and Nonsupervisory Workers,
1967-1991
380.00 -
360.00 1967
1970
1973
136
1976
1979
1982
1985
1988
1991
defined as wages by workers when they report their wages
in government surveys.
Although studies of labor market trends should examine
both wage and benefit trends, those that focus on wage
trends alone (usually because of a lack of benefit data) are not
misleading. That is, taking account of pension and insurance
costs (including both health and life insurance), given their
small size and slow growth, would not substantively alter the
picture emerging from an analysis of government wage data
frequently used to track labor market trends. It should be
kept in mind that the erosion of wages is somewhat larger
than the erosion of overall compensation. Plus, the years
since 1989 have seen a 9 3% expansion of health and
pension costs, an increase of *0.13 per hour worked, which
partially offset the 80.28 drop in houriy wages.
Further, it should be noted that the data in Table 3 2 show
an even larger drop in real hourly wages than the data in
Table 3 1 . Since the data in Table 3.2 are drawn from a survey of employers rather than one of households, they are
probably more accurate.
Low-Paid Occupations Fared Worse
We now turn our attention to the various segments of the
workforce and how they managed during the 1980s. In
general, the workers who experienced the greatest fall in
real wages were those who initially had lower wages, were
without a college degree, were in blue-collar or service
occupations, or were in the younger age brackets. Since
1987, however, the wages of white-collar and college-educated workers have also eroded in inflation-adjusted terms.
The data in Table 3.3 and Figures 3A and 3B show wage
trends for the 80% of the workforce who are "production
and non-supervisory workers." This category includes factory workers, construction workers, and a wide variety of
service sector workers, ranging from restaurant and clerical
workers to nurses and teachers. For all of these workers,
average hourly earnings fell J0.73 from 1979 to 1989, a
decline of 0.7% each year. In contrast, hourly earnings
were flat in the 1973-1979 period and grew 2.0% to 2.3%
per year from 1947 to 1973Because of reductions in weekly hours since 1979 as
well as in hourly earnings, weekly earnings fell at a rate of
1.0% each year until 1989. The earnings of production and
137
For the 80% of
the workforce
who are
"production and
non-supervisory
workers" houriy
earnings fell
$0.73 from 1979
to 1989, a decline
of 0.7% each
year.
�TABLE 3.4
Changes in Wages and Compensation, 1987-1992
(1991 Dollars)
Real
Wages and Salaries
Per Hour
Year*
BlueCollar
WhiteCollar
Real
Compensation
Per Hour
BlueCollar
WhiteCollar
Pay Level
1987
1988
1989
1990
1991
1992
$12.91
12.74
12.53
12.05
11.83
11.93
$15.82
15.56
15.37
15.46
15.12
15.27
$16.22
16.10
15.87
15.33
15.15
15.42
$18.79
18.54
18.32
18.51
18.15
18.40
-2.8%
-4.9
-7.6
-2.8%
-0.7
-3.5
-2.2%
-2.8
-4.9
-2.5%
0.4
-2.1
Percent Change
1987-89
1989-92
1987-92
'Data are (or March. Consistent data lor earlier years not available.
138
non-supervisory workers in 1989 were $367.12 per week
(in 1991 dollars), less than what they were earning in
1967.
Because of cutbacks in hours worked, the weekly wages
of production and non-supervisory workers fell 1.7% annually between 1989 and 1991, even faster than the 1.3%
annual fall of hourly earnings.
Table 3.4 presents wage and compensation trends
between 1987 and 1992 for the blue- and white-collar
workforces. Unfortunately, there are no consistent data
available for earlier years.
Blue-collar workers have experienced a severe, 7.6%
reduction in hourly wages over the 1987-1992 period.
White-collar wages also fell in this time period, declining
3 5%. Among both blue- and white-collar workers, wages
were falling in the final two years of the 1980s recovery,
from 1987 to 1989, suggesting that the forces leading to
lower pay were present before the onset of the early 1990s
recession. Both groups also saw their total compensation
per hour decline between 1987 and 1992, with a 4.9% loss
among blue-collar workers and a 2.1% loss among whitecollar workers.
Wage Trends by Wage Level
For any given trend in average wages, there will be different outcomes for particular groups of workers if wage inequality rises (or falls), as it has in recent years. Table 3.5
provides data on wage trends for workers at different points
(or levels) in the wage distribution, thus allowing us to
characterize wage growth for low-, middle-, and high-wage
earners. The data are presented for the cyclical peak years
of 1973, 1979, 1989, and for the most recent year, 1991.
These data (and Figure 3C) show that the deterioration
in real wages since 1979 was both broad and uneven. The
breadth of falling real wages is clear from the fact that over
the 1979 to 1989 period wages fell for the bottom 80% of
the workforce. The overall pattern of wage growth showed
that a group's wages fell more the lower the group's wage,
with wages falling by 11.8% at the 20th percentile and by
just 0.4% at the 80th percentile. The wage of the median
worker, who earned more than half the workforce but also
less than half the workforce, fell 4.9% from 1979 to 1989.
139
Blue-collar
workers have
experienced a
severe, 7.6%
reduction in
hourly wages ove
the 1987-1992
period. Whitecollar wages also
fell in this time
period, declining
3.5%.
�TABLE 3.5
Wages for All Workers by Wage Percentile, 1973-1991
(1991 Dollars)
Wage by Percentile
Year
20
40
Median
$6.49
6.44
5.68
5.75
$8.97
8.85
8.23
8.00
$10.30
10.10
9.61
9.48
- 0.8%
-11.8
1.3
-11.4
-1.3%
- 7.1
-2.8
-10.8
-$0.05
- 0.76
- 0.07
- 0.74
-$0.12
- 0.63
- 0.23
- 0.97
60
80
Real Hourly Wages
1973
1979
1989
1991
$11.60
11.50
10.98
10.99
$15.70
15.82
15.76
15.45
Percent Change
1973-79
1979-89
1989-91
1973-91
-1.9%
-4.9
-1.3
-8.0
-0.8%
- 4.5
-0.1
-5.3
0.8%
- 0.4
-2.0
-1.6
Only those people at the very top of the earnings scale
were exempt from wage decline (e.g., wages at the 9 0 t h
percentile—not shown i n the table—did rise 2.5% from
1979 to 1989)
During the recession years from 1989 to 1991 some o f
the largest wage declines were among the highest wage
workers: the hourly wage at the 80th percentile fell 2.0% in
these two years, far greater than the 0.4% decline in the
prior ten years. These data, along w i t h the findings o f falling white-collar wages, show that i n recent years high-wage
workers had been experiencing the wage declines that
blue-collar and low- and middle-wage workers experienced over the entire 1980s.
Over the entire 1973-1991 period the hourly wage o f the
median worker fell 10.82 from $10.30 to $9 48, a 8.0%
decline. There were also wage declines among high-wage
workers at the 80th percentile (1.6%) and among lowwage workers at the 20th percentile (11.4%), w i t h the largest wage reductions experienced by the workers w i t h the
lowest wages.
Over the entire
1973-1991 period
the hourly wage
of the median
worker fell $0.82
from $10.30 to
$9.48, a 8.0%
decline.
Dollar Change
1973-79
1979-89
1989-91
1973-91
-$0.20
- 0.50
- 0.13
- 0.82
-$0.10
- 0.52
- 0.01
- 0.61
$0.12
- 0.06
-0.31
-0.25
FIGURE 3C
Real Wage Growth By Wage Percentile,
1979-1991
20lh Percentile
140
40th Percentile
Median
60th Percentile
141
80th Percentile
�TABLE 3.6
Wages for Male Workers b W g Percentile, 1973-1991
y ae
(1991 Dollars)
Wage by Percentile
40
Median
60
80
$8.08
7.84
6.59
6.42
$10.82
11.02
9.61
9.25
$12.29
12.43
10.98
10.83
$14.09
13.80
13.17
12.50
$18.02
18.37
17.64
17.49
-3.0%
-15.9
-2.6
-20.6
1.9%
-12.9
-3.7
-14.5
1.2%
-11.7
-1.4
-11.9
-2.0%
-4.6
-5.1
-11.3
1.9%
-4.0
-0.8
-3.0
-$0.24
-1.25
-0.17
-1.66
Year
$0.20
-1.42
-0.36
-1.57
$0.14
-1.45
-0.15
-1.46
-$0.29
-0.63
-0.67
-1.59
$0.34
-0.73
-0.15
-0.53
20
Real Hourly Wages
1973
1979
1989
1991
Percent Change
1973-79
1979-89
1989-91
1973-91
Dollar Change
1973-79
1979-89
1989-91
1973-91
TABLE 3.7
Wages for Female Workers b W g Percentile, 1973-1991
y ae
(1991 Dollars)
Wage by Percentile
Year
20
40
Median
60
80
$5.74
5.70
5.32
5.15
$7.19
6.99
7.06
7.00
$7.91
7.81
8.23
8.00
$8.92
8.85
9.33
9.40
$11.54
11.49
13.17
13.00
- 0.7%
- 6.8
- 3.1
-10.3
-2.8%
1.1
-0.9
-2.6
-1.3%
5.3
-2.8
-1.1
-0.7%
5.3
-0.8
-5.4
-0.4%
14.6
-1.3
12.6
-$0.04
- 0.38
- 0.17
- 0.59
-$0.20
0.08
- 0.06
- 0.19
-$0.10
0.41
- 0.23
- 0.09
-$0.06
0.47
- 0.07
0.48
This overall picture, however, masks somewhat different
outcomes for men and women. Among men, wages have
fallen more, and at all parts of the wage distribution (Table
3.6). In the middle, the median male hourly wage fell
11.7% between 1979 and 1989 with an additional 1.4% fall
between 1989 and 1991 for a total fall of 12.9%. Even highwage men (those at the 80th percentile) experienced a significant 4.8% wage decline over the twelve-year period,
with the 1979-1989 rate of decline continuing after 1989.
Wages among low-wage men fell the most (18.1%) from
1979 to 1991. These data show significant wage deterioration for nearly all men, with the bottom 60% of men sufiering more than a 10% wage reduction since 1979- Some
groups had experienced falling wages in the 1973-1979
period as well.
The only significant wage growth between 1979 and
1989 appears to have been among higher-wage women
(Table 3.7). For instance, wages at the 80th percentile
grew 14.6%. Even at the median, wages grew by 5 3%.
Among women in the bottom 40%, however, wage growth
was either minimal or negative.
One of the surprising stories emerging from these data is
that women's wages fell across the board during the recession years from 1989 to 1991, with wage reductions at
both the 80th and 20th percentiles At the middle, the
hourly wage of the median female worker fell 2.8% from
1989 to 1991, reversing half the 5.3% wage expansion over
the prior ten years.
- $0.05
1.68
- 0.17
1.46
Real Hourly Wages
1973
1979
1989
1991
Percent Change
1973-79
1979-89
1989-91
1973-91
Dollar Change
1973-79
1979-89
1989-91
1973-91
142
143
These data show
significant wage
deterioration for
nearly all men,
with the bottom
60% of men
suffering more
than a 10%
wage reduction
since 1979.
�TABLE 3.8
Changes in the Gender W g Differential, 1979-1991
ae
(1991 Dollars)
Median Hourly Wage
Year
Male
Change, 1979-1989
Ratio
$12.59
11.12
10.83
12.59
12.59
1979
1989
1991
1989 Alt.*
1991 Alt*
Female
$7.91
8.33
8.00
8.33
8.00
62.8%
74.9
73.9
66.2
63.5
$-1.47
$0.42
12.1% pts.
Contribution to Narrower Wage Gap of Changes in:**
Male
Wage
Decline
72%
1979-1989
Female
Wage
Growth
28%
Total
100%
'Scenario il male wages had not declined in real terms between 1979 and 1989 or 1991.
"The contribution of "female wage growth" is the growth of the gender differential assuming male real
wages had not fallen, relative to the actual change in the differential.
144
The Male-Female Wage Gap
From 1979 to 1989, the median hourly wage for men
fell $1.47 or 11.7% while the median hourly wage for
women rose S0.42, or 5.3% (Table 3.8). This led to a
reduction in the hourly wage gap between men and
women by 12.1 percentage points, from 62.8% in 1979 to
74.9% in 1989. This represents a sizeable reduction in
wage inequality given that there was a 12.1 percentage
point reduction in the initial 37.2 percentage point wage
gap (100% less 62.8%). Even after this progress, however,
women still earned one-fourth less than men in 1989.
This narrowing of the male/female wage gap is the result
of both improvements in real hourly wages for women and
real wage reductions for men. Table 3-8 provides an assessment of how much the narrowing of the male-female wage
differential is due to rising real wages for women, and how
much is due to the real wage loss for men. If real wages
among men had not fallen by 1989 but had remained at
their 1979 level, the wage gap would have been 66.2% and
the wage gap would have closed by 3.4 rather than by the
actual 12.1 percentage points. Thus, falling real wages
among men can explain 72% of the closing of the gender
wage gap between 1979 and 1989. Only 28% of the narrowing of the gender wage gap was due to rising women's
real wages (3 4% divided by 12.1%). Unfortunately, the
even greater fall in women's than men's median wages
from 1989 to 1991 (4.0% versus 2.6%) meant that both
genders had falling wages and that gender inequality grew
(Table 3 6 and 3 7).
145
Falling real wages
among men can
explain 72% ot
the closing ol the
gender wage gap
between 1979
and 1989.
�TABLE 3.9
Distribution of Total Employment b W g Level, 1973-1991
y ae
Hourly Wage
Relative to
Poverty Level*
Percent Distribution
ot Employmenl:
1973
1979
1989
1991
Percentage Point Change
1973-79
1979-89
1989-91
Total
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
13.2% 12.7%
18.5
14.8
14.7
14.3
29.4
29.8
16.4
18.9
8.7
8.6
100.0
100.0
-1.7
5.4
1.0
-2.8
-0.4
-1.5
2.7%
9.6%
9.5%
3.3%
9.7
13.7
12.7
16.3
10.3
13.1
10.9
13.2
36.9
34.2
30.3
30.5
19.4
27.3
27.4
22.4
12.5
11.7
11.5
11.1
100.0
100.0
100.0
100.0
-0.6
4.1
0.6
-2.8
0.1
-1.4
6.9
-1.1
23
-3.7
-5.0
0.7
-0.1
3.7
-0.1
-0.2
-3.0
-0.3
9.4%
6.1%
17.5% 16.3%
24.4
30.7
20.9
17.3
15.7
18.3
19.3
16.4
28.4
34.4
31.7
28.9
10.4
14.9
10.1
13.1
3.1
2.1
5.6
5.0
100.0
100.0
100.0
100.0
-3.4
6.4
1.1
-2.7
-0.3
-1.1
11.4
-13.5
-3.0
-2.8
4.8
3.0
-1.2
3.7
-0.7
-05
-1.8
0.6
5.8%
4.1%
15.6
21.0
13.5
14.7
35.9
33.1
20.5
20.1
8.8
7.3
100.0
100.0
9.1
-6.2
0.2
-3.4
-1.2
1.4
-0.6
3.7
-0.4
-0.4
-2.5
0.1
Men
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
Women
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
The wage ranges are equivalent in 1991 dollars lo: Low - $4.88, $4.89 - $6.52, $6.53 - $8.13,
$8.14 - $13.02, $13.03 - $19.54, and $19.55 and more.
The Expansion of Low-Wage Jobs
A n o t h e r useful way o f characterizing changes i n the
wage structure is to examine the trend in the p r o p o r t i o n o f
workers earning low, middle or high wages. These trends
are presented i n Table 3-9 for all workers and for men and
w o m e n . The workforce is divided into six wage groups
based o n multiples o f the "poverty wage level," o r the
hourly wage a full-time, year-round worker must earn to
sustain a family o f four at the poverty threshold—which
was $6.52 i n 1991. Thus, workers are assigned t o a wage
g r o u p according t o the degree to which they earned more
(or less) than poverty level wages.
The data in Table 3 9 show that there was a significant
expansion of workers earning less than poverty level wages
between 1979 and 1989. In 1979, only 4 . 1 % of the workforce earned wages at least 25% below poverty level wages.
By 1989, 13 2% of the workforce earned such low wages, a
shift of 9.1 % of the workforce into this low-wage group. Likewise, i n 1989, 28% of the wotkforce earned poverty level
wages (adding the t w o lowest paying categories together), a
rise from 2 5 . 1 % i n 1979 and just 21.4% in 1973.
Over the 1979 to 1989 period there was a general d o w n ward shift i n the entire wage structure as there were proportionately fewer workers in the middle- and high-wage
groups. The only exception is that there was a modest
expansion o f the share o f the workforce at the very highest
earning level (exceeding three times the poverty level
wage). There was a further downward shift in the wage
structure from 1989 to 1991, such that 3 1 2 % o f the workforce in 1991 earned poverty level wages o r less.
Women are much more likely to earn low wages than
men. I n 1979, 36.8% of w o m e n (adding the t w o b o t t o m
groups) earned poverty level wages o r less, more than
double the 16.4% o f men earning such low wages. Women
are also m u c h less likely to earn very high wages. I n 1979,
only 2 . 1 % o f w o m e n , but 11.1% o f men, earned at least
three times the poverty wage level.
Among women over the 1979-1989 period there was a
larger shift downward—an additional 11.4% earned very
low wages—and a larger shift upwards—the t w o highest
wage groups grew by 7.8%. The shift downwards among
146
147
Over the 1979 to
1989 period there
was a general
downward shift
the entire wage
structure as
there were
proportionately
fewer workers
the middle- and
high-wage
groups.
�TABLE 3.10
Distribution o White Employment b W g Level, 1973-1991
f
y ae
Hourly Wage
Relative to
Poverty Level*
Percent Distribution
of Employment.
1989
1991
1973
1979
Percentage Point Change
1973-79
1979-89
1989-91
Total Whites"
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
5.4%
4.0% 12.0% 11.3%
14 7
13.4
19.6
16.8
130
14.2
14.1
13.8
358
33.5
30.3
30.2
21.5
17.9
21.0
20.3
9.6
9.8
8.0
9.9
100.0
100.0 100.0
100.0
-1.5
4.9
1.1
-2.3
-0.6
-1.6
8.0
-6.2
0.1
-3.1
-0.6
1.8
-0.6
3.4
-0.4
-0.1
-2.4
0.1
White Men
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
2.9%
2.4%
8.1%
7.8%
8.9
12.3
10.8
14.1
9.6
10.3
12.4
12.3
36.4
34.2
30.9
31.4
28.5
28.7
24.2
21.2
13.7
12.1
13.5
13.3
100.0
100.0 100.0
100.0
-0.4
3.5
0.7
-2.2
0.1
-1.6
5.6
-15
2.2
-3.2
-4.5
1.4
-0.2
3.2
-0.2
0.4
-3.0
-0.3
White Women
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
9.3%
6.1% 16.5% 15.2%
23.7
29.7
165
19.9
18.1
19.3
16.2
15.6
34.9
32.5
29.6
29.0
10.8
10.2
15.8
14.1
3.1
2.1
5.4
6.2
100.0
100.0 100.0
100.0
-3.3
6.0
1.2
-2.4
-0.5
-1.0
10.4
-13.2
-3.1
-2.9
5.6
3.3
-1.2
3.4
-0.6
-0.6
-1.7
0.8
'The wage ranges are equivalent in 1991 dollars to: Low-$4.88, $4.89 - $6.52, $6.53 - $8.13,
$8.14 -$13.02, $13.03 - $19.54, and $19.55 and more.
"Defined as white non-Hispanic.
148
women appears to be a shifting down of the workforce
earning just below poverty level wages (the second group)
to the very lowest wage category. The shift upwards was
largely a shift from the middle-wage levels. During the
recession years from 1989 to 1991, however, structure of
women's wages generally shifted downward.
Among men, the overall changes in the wage structure
between 1979 and 1989 meant proportionately fewer highwage and more low-wage male workers. For instance, there
was an increased proportion of men earning less than
125% of the poverty level wage and a shrinking proportion
of men in the second and third highest wage groups. In
addition, there was only a modest growth in the highest
wage group.
Tables 3.10, 3.11 and 3-12 (and Figure 3D) present a
analysis similar to Table 3 9 for, respectively, white, black
and Hispanic employment. For instance, Table 3-10 shows
that there was a modest shift downward in the wage structure for whites in the 1970s, followed by a larger downward shift in the 1979-1989 period, and during the early
1990s recession. As with the total workforce, there was a
modest growth of high-wage employment among white
men and a much larger shift to high-wage employment
among white women. Over the entire period from 1973 to
1991, however, there was a slight shift toward proportionately fewer men in the middle and higher wage groups and
proportionately more men in the lowest wage groups,
which nearly doubled in size (from 11.8% in 1973 to
21.9% in 1991). Among white women, there was a simultaneous shift towards very low-wage work and an expansion
of women at the highest wage levels.
Among blacks, there was a general shift out of high-wage
employment into low-wage employment from 1979 to
1989 without any significant growth at the highest wage
levels. By 1989, 37.4% of black workers were in jobs paying
less than poverty level wages, with 33.7% of black men and
41% of black women earning such low wages. The shift
towards the very lowest paying jobs—an additional 11.2%
of black men and 16.5% of black women—was much larger
than among whites. The early 1990s recession only added
to the downward shift of the wage structure among both
black men and black women.
149
Among blacks,
there was a
general shift out
of high-wage
employment
into low-wage
employment from
1979 to 1989
without any
significant growth
at the highest
wage levels.
�TABLE 3.11
Distribution of Black Employment b W g Level, 1973-1991
y ae
Hourly Wage
Poverty Level*
Total Blacks"
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
Percent Distribution
of Employmenl:
1973
1979
1989
1991
7.7%
19.6
16.3
36.9
15.5
4.1
100.0
4.7%
18.6% 17.1%
26.8
18.8
24.0
16.9
16.3
17.0
27.8
31.5
27.3
16.2
14.0
11.6
4.0
3.7
3.9
100.0
100.0
100.0
Percentage Point Change
1973-79
1979-89
1989-91
-2 9
7.2
0.7
-5.4
0.7
-0.3
13.8
-8.0
-0.1
-3.7
-2.2
0.1
-1.5
5.2
-0.6
-0.5
-2.4
-0.2
Black Men
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
5.6%
4.2% 15.4% 14.9%
23.7
14.2
20.9
18.3
14.5
14.5
17.1
16.0
40.7
33.4
29.1
28.0
20.4
21.3
15.8
13.3
4.2
4.6
5.7
4.3
100.0
100.0
100.0
100.0
-1.4
6.7
0.0
-7.3
0.8
1.2
11.2
-2.6
2.6
-4.2
-5.4
-1.4
0.5
5.4
1.1
1.2
2.5
0.0
Black Women
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
10.0%
5.2% 21.7% 19.3%
25.7
19.3
24.3
33.1
16.7
16.6
18.3
19.6
32.5
265
26.6
29.5
12.2
9.8
10.7
10.0
3.7
3.6
3.3
1.9
100.0
100.0
100.0
100.0
-4.8
7.3
1.3
-3.0
0.9
-1.8
16.5
-13.8
-2.9
-3.0
1.4
1.7
2.4
5.0
0.1
0.1
2.2
0.4
'The wage ranges are equivalent in 1991 dollars to: Low-$4.88, $4.89 - $6.52, $6.53 - $8.13,
$8.14 - $13.02, $13.03 - $19.54, and $19.55 and more.
"Defined as black non-Hispanic.
There has been a general downshifting of the Hispanic
wage structure, for men and for both men and women
combined, throughout the 1973-1991 period, coupled
with a very modest growth in the highest wage jobs.
Among Hispanic women, however, between 1979 and
1989 there was a large shift into the lowest wage jobs, but
also a significant (4.6 percentage point) growth in employment paying twice the poverty wage level. Much of this
growth in high wage jobs for Hispanic women was
reversed in the 1989-1991 downturn.
There has been
a general
downshifting of
the Hispanic
wage structure.
TABLE 3.12
Distribution of Hispanic Employment b W g Level, 1973-1991
y ae
Hourly Wage
Relative to
Poverty Level"
Percent Distribution
of Employment:
1979
1973
1989
1991
Percentage Point Change
1973-79
1979-89
1989-91
Total Hispanics
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
7.8%
5.3%
18.3% 19.2%
21.9
30.2
22.5
26.7
17.1
16.3
16.2
16.1
36.1
24.5
30.9
27.0
9.4
13.4
14.3
12.1
3.0
3.9
4.1
3.8
100.0
100.0
100.0
100.0
-2.5
6.3
-0.8
-5.2
1.0
-0.9
12.9
-7.7
-0.1
-3.9
-2.2
0.9
1.0
4.2
-0.1
-2.5
-2.7
0.2
Hispanic Men
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
3.7% 15.9% 16.8%
6.8%
15.6
22.9
22.3
26.9
14.9
15.1
15.3
166
39.7
35.1
28.1
24.3
17.9
190
13.9
10.9
4.2
4.5
5.0
4.6
100.0
100.0
100.0
100.0
-3.2
7.3
0.2
-4.6
1.1
-0.8
12.2
-0.6
0.1
-7.0
-5.1
0.4
0.9
4.6
1.3
-3.7
-3.0
-0.1
Hispanic Women
Below 75% Pov.
To below 100%
To below 125%
To below 200%
To below 300%
Above 300%
All
7.8% 21.9% 22.9%
9.3%
41 2
26.5
32.2
22.8
15.4
17.7
20.6
18.1
25.4
24.7
30.2
24.6
7.4
7.2
5.9
9.5
3.4
1.0
1.8
2.8
100.0
100.0
100.0
100.0
-1.5
9.0
-2.5
-5.6
1.4
-0.8
14.1
-18.4
-04
0.8
2.1
1.7
1.0
3.7
-2.3
-0.7
-2.3
0.6
'The wage ranges are equivalent in 1991 dollars to: Low-$4.88, $4.89 - $6.52, $6.53 - $8.13,
$8.14 - $13.02, $13.03 - $19.54, and $19.55 and more.
150
151
�FIGURE 3D
Employment at or Below Poverty Level Wages, By Race,
1973-1991
1973
1976
1979
152
1962
1965
1988
Benefit Reduction
1991
This section examines the trends in the provision of
fringe benefits. As we saw earlier in Table 3 2, the real value
of both wages and fringe benefits has fallen in recent years.
In the 1980s, employers reduced pension and health coverage and provided fewer days off with pay. These benefit
reductions have affected workers across the wage scale,
but, again, low-wage workers have experienced the greatest
fall-off in health and pension coverage.
Table 3.13 examines the trend in employers' provision
of the major benefit plans, pensions, and health insurance,
between 1966 (the earliest available data) and 1992. Over
this time period, the average employer contribution for
pension and savings plans more than doubled between
1966 and 1977, but has been falling in each period thereafter. Between 1977 and 1989 the costs of retirement benefits fell nearly 40% from $0.75 to $0.46 an hour. Measured by average inflation (the 'All Goods and Services"
column), employer provided health insurance benefits
rose rapidly over the late 1960s until 1977. Health insurance benefits continued to grow into the 1980s and
through 1992 but at a slower annual rate of growth. However, the overall rise in employer health insurance costs of
$0.26 per hour worked, between 1977 and 1989, is far
from exorbitant. Apparently, even though health insurance
costs had indeed risen quite rapidly among some employers in this period, there was such large employment
growth among employers who provided litde or no health
insurance benefits that the average costs grew only
modesdy.
However, the far faster inflation in medical services
means that the real health insurance coverage provided per
hour worked—the amount of real medical services purchasable by an employer's average benefit cost—remained
relatively stable between 1977 and 1989. That is, employers increased their spending on health insurance only
about as fast as medical service prices were rising, so the
quantity and quality of health insurance provided by
employers did not generally rise between 1977 and 1989.
There was an uptick in health insurance spending between
1989 and 1992 no matter which inflation measure is used.
As the last column in Table 313 shows (repeated from
Table 3.2) the increase in "voluntary" fringe benefits
153
Even though
health insurance
costs had indeed
risen quite rapidly
among some
employers in this
period, there was
such large
employment
growth among
employers who
provided little
or no health
insurance benefits
that the average
costs grew only
modestly.
�TABLE 3.13
Employer Hourly Benefit Costs, By Type, 1966-1992
(1991 Dollars)
Health Insurance
Adjusted for
Inflation in:
Pension
and
Savings
Year
All
Goods &
Services*
Medical
Services*
Total
Voluntary
Benefits**
$0.52
0.79
1.00
0.98
1.01
1.07
$0.61
1.03
1.43
1.45
1.40
1.53
Benefit Level
$0.34
0.55
0.75
0.58
0.46
0.45
1966
1972
1977
1987
1989
1992
$0.27
0.49
0.68
0.87
0.94
1.09
(c-xcluding payroll (axes lor social security, workers compensation. and unemployment insuranee) was r.ipul over
the I 9 6 6 I 9 7 7 period, hut then did not resume until alter
1989.
Table 3.14 examines the decline in pension and health
insurance coverage for different demographic groups
between 1979 and 1989 The percentage of the private
workforce covered by a pension plan dropped from 50.0%
in 1979 to 42 9% in 1989. This d r o p in coverage is perhaps
one of the reasons for the lessening of pensions costs lor
employers over this time peritKl. Ix)wer pension coverage
appears to be occurring primarily among men, whose coverage fell 10.1 percentage points. Women workers, however, are still less likely than men to be covered by an
employer's pension plan. Blacks and whites fared about
TABLE 3.14
Changes in Private Sector Benefit Coverage, 1979-1989
Percent Change
Pension Coverage"
58.6%
36.6
-37.8
-3.8
1966-72
1972-77
1977-89
1989-92
81.2%
40.5
37.7
15.7
52.2%
27.3
0.6
5.6
38.4
-1.7
9.3
•The two inflation-adjusted measures ot real insurance costs per hour differ because medical service
prices rose substantially faster than average prices.
•Benefits exclusive of payroll taxes, adjusted for overall inflation.
The percentage of
the private
workforce
covered by a
pension plan
dropped from
50.0% in 1979 to
42.9% in 1989.
Group'
1979
1989
All Workers
50.0%
42.9%
By Gender
Men
Women
55.9%
41.1
By Race
White
Black
Hispanic
Health Insurance Coverage'"
Change
1979-89
Change
1979-89
1979
1989
7.1
68.5%
61.1%
-7.4
45 8%
39.2
- 10.1
19
74.6%
59 1
66.3%
54 4
83
-4 7
51.5%
457
37.8
45.1%
40 5
26.1
6 4
5.2
- 11.7
69.7%
63.0
603
634%
562
460
6.3
-6.8
14.3
43.7%
499
49.6
59.2
609
27.5%
41.6
43.5
53.4
596
162
8.3
6.1
5.8
13
61 0%
684
69.2
78.5
78.5
44 6%
60.0
60.9
73.5
77.1
16.4
84
-8.3
5.0
-14
By Education
High School Dropout
High School Graduate
Some College
College
More than College
'Wage and salary workers, ages 18-64, with al least 20 weekly hours and 26 weeks ol work.
'Employee participates in employer s plan.
•Employer provides coverage plan and pays at least some ol the costs.
IS t
I'ST
�TABLE 3.15
Changes in Private Sector Benefit Coverage
by Wage Fifth, 1979-1989
By Wage Fifth
Benefit Coverage*
Lowest
Fifth
Second
Fifth
Middle
Fifth
Fourth
Fifth
Top
Fifth
All
Pension Coverage
1979
1989
Change, 1979-89
18.3%
12.8
-5.5
36.6%
28.8
-7.8
51.9%
44.1
-7.8
67.8%
59.1
-8.7
76.2%
69.8
-6.4
50.0%
42.9
-7.1
60.3%
51.5
-8.8
74.3%
67.1
-7.2
83.1%
77.4
-5.7
87.6%
83.1
-4.5
68.5%
61.1
-7.4
Health Insurance Coverage
1979
1989
Change, 1979-89
37.9%
26.4
-11.5
•See Table 3 14 lor definitions.
TABLE 3.16
Trends in Days Off With Pay, 1947-1988
Number of Days
Off With Pay Per Year*
Year
Non-farm
Business
Manufacturing
1947
1966
1972
1977
1979
1981
1989
n.a.
15.1 days
17.4
19.8
n.a.
19.8
16.1
15.9 days
17.9
20.5
22.6
23.1
22.9
20.8
equally well, experiencing a 5.2% and 6 H% d r o p in coverage, respectively. Hispanics, however, were the hardest hit,
with a 11.7% decline in pension coverage from 1979 to
1989 When coverage is examined by education level, the
data show declining pension coverage among college graduates and those with less than a college degree, and that
coverage declined more among those with the lowest education levels.
Similar trends have occurred regarding health insurance
coverage. The greatest loss ol health insurance coverage
was among men, Hispanics, and workers w i t h less than a
college degree. By I9H9, more than a third of the workers
without a college degree—even those w i t h some education
beyond high school—did not have employer provided
health insurance.
Table 3.15 shows the erosion of pension and health
insurance coverage at each wage level from 1979 to 1989
The general pattern is that the groups w i t h the lowest
wages experienced the greatest decline in benefit cover
age. The primary exception is for pension coverage among
the lowest wage group w h i c h had such low pension cover
age in 1979—just 18.3%—that there was not much room
for erosion.
Workers are also receiving less time oil for vacations and
holidays (Table 3.16). The longest historical comparison
available is for manufacturing workers, f o r these employees, the number of paid days off rose from 15.9 per year in
1947 to 23 I days i n 1979 Hut this t r e n d towards
increased time off w i t h pay was reversed in the 1980s In
1989, the average manufacturing worker enjoyed only 20 8
days off w i t h pay, 2.3 days less than in 1979 f o r the nonfarm business workforce as a whole, there was a growth in
paid time off from 15 I days in 1966 to I 9 8 d a y s i n 1981
By 1989, however, paid time off had declined lo 16.1 days, a
decline of 3.7 days for the average worker In contrast,
European workers have more paid time off and have been
obtaining more, rather than less, paid time oil since 1980.
•Assuming 2.080 hours paid per year. Days off would include, lor instance, holiday, vacation, sick, and
funeral leave. Years selected based on data availability
156
15^
By 1989, more
than a third oi
the workers
without
a college
degree—even
those with some
education beyond
high schooldid not have
employer
provided health
insurance.
�What Explains Wage Trends?
TABLE 3.17
The Joint Effect of Education
and Experience on Wages, 1973-1991
Effect of:
Years
More (Less)
Education
and Experience
Other
Wage
Growth*
Total
Changes
in Wages
3.5%
5.2
11.6
- 6.4%
-8.4
-17.7
- 3.6%
-2.2
-6.6
3.5%
4.5
10.5
-6.1%
-9.6
-19.7
-3.2%
-4.6
-9.6
4.2%
6.0
13.5
-5.8%
-1.2
-8.7
-2.7%
7.6
5.7
Total
1973-79
1979-89
1973-91
Men
1973-79
1979-89
1973-91
Women
1973-79
1979-89
1973-91
"Wage growth of workers holding education and experience constant. The effect of the interaction
between more education and experience and within-group wage trends was part of the decomposition
but is omitted in the table.
158
In this section we shift the discussion from a presentation of wage and benefit trends overall, and for subgroups,
to an examination of explanations for the pattern of recent
wage growth. The items to be explained include declining
average wages since 1973 and the dramatic growth i n
wage inequality i n the 1980s. More specifically, it is important to understand b o t h the average performance of wage
g r o w t h and w h y particular groups fared w e l l or poorly.
Table 3.17 ( i n the last c o l u m n ) shows the change in the
hourly wage of the average worker over several time periods between 1973 and 1991. As we have seen earlier, average wages have been falling since 1973. Because of differences i n measurement, data sources, and time periods, the
drop in average wages in Table 3-17 differs from estimates
presented earlier—being a somewhat larger d r o p than
shown i n Table 3 1, but a smaller drop than shown i n Table
32.
The most commonly mentioned reason for these wage
problems is slow productivity g r o w t h (i.e., changes i n output per hour worked) since 1973- As the data i n Table 3.1
show, productivity grew by only 0.6% annually from 1973
to 1979- Productivity did pick up i n the 1979-1989 period,
to a slightly higher 0.8% annual rate. The productivity
growth record since 1973, however, does not come near
the rapid g r o w t h in the earlier post-war period—productivity grew 2.4% annually from 1959 to 1973.
Slow productivity has been a major problem and it is
important to raise our productivity g r o w t h . Notwithstanding this, it is also true that slow productivity g r o w t h provides only a partial explanation for average wage trends
since productivity grew, but average wages f e l l . For example, productivity grew a total of 3-6% between 1973 and
1979 and another 8.3% from 1979 to 1989- It should also
be noted that slow productivity g r o w t h cannot explain
changes i n the wage structure, or w h y some groups experienced wage g r o w t h and others experienced wage reductions (such as men's wages falling 9 6% and women's wages
rising 5.7% from 1973 to 1991).
159
5/oiv productivity
growth provides
only a partial
explanation for
average wage
trends since
productivity grew,
but average
wages fell.
�TABLE 3.18
The Separate Effects o Education
f
and Experience on Wages, 1973-1991
Effect of:
More (Less)
Education*
Years
More (Less)
Experience*
Total
Wage
Growth
Total
1973-79
1979-89
1973-91
3.0%
3.4
7.7
0.0%
1.7
2.7
-3.6%
-2.2
-6.6
2.9%
2.8
6.5
-0.2%
1.7
2.6
- 3.2%
-4.6
-9.6
3.8%
4.5
10.6
0.4%
1.3
1.9
-2.7%
7.6
5.7
Men
1973-79
1979-89
1973-91
Women
1973-79
1979-89
1973-91
•Exclusive ol any interaction between rising education and experience and wage growth ol education/
experience categories.
160
It has been suggested that the exclusion of benefit trends
from a table such as Table 3 17 yields a misleading characterization of pay trends since wages may be falling, but
since benefits have risen, there has been an overall increase
in total compensation (wages and benefits). However, as we
saw earlier in Table 3 2, both compensation and wages have
fallen since 1973. It is true, however, that compensation has
not fallen as much as wages because, in various time periods, benefits declined less than wages or even grew slightly.
So, fringe benefit trends do not solve the puzzle of falling
pay and rising productivity.
Unfortunately, there has not been extensive research on
this issue. One explanation is that prices for national output
have grown more slowly than prices for consumer purchases. Therefore, the same growth in nominal, or current
dollar, wages and output yield a faster growth in real (inflation adjusted) output (which is adjusted for changes in the
prices of investment goods, exports, and consumer purchases) than in real wages (adjusted for changes in consumer purchases only). Another explanation is that a
greater share of productivity growth is accruing to owners
of capital via higher profits and interest rates, as shown in
Chapter 1.
The Role of Education and Experience
Economists have identified that greater education and
more work experience (from accumulated knowledge and
seniority) are associated with wage growth both for individuals and for the economy. Changes in the economic benefits of having greater education and experience can also
be an explanation for changes in the wage differences
among educational and age groups and between high- and
low-wage workers.
Since 1973 the workforce has been getting older (more
experienced) and more educated, factors which lead to
economy-wide wage increases. Table 3.17 identifies the
effect of greater education and experience on average wage
growth overall and by gender. As expected, the shift to a
more educated and older workforce has meant higher
wages, an increase of 11.6% over the 1973-1991 period.
161
Economists have
identified that
greater education
and more work
experience (from
accumulated
knowledge and
seniority) are
associated with
wage growth both
for individuals
and for the
economy.
�TABLE 3.19
Change in Real Houriy Wage by Education, 1973-1991
(1991 Dollars)
Year
High
School
Drop-out
High
School
Graduate
Some
College
College
College
2+ years
Real Hourly Wage
1973
1979
1987
1989
1991
$9.87
9.59
8.24
7.95
7.62
$11.28
10.69
9.92
9.63
9.43
$12.41
11.85
11.32
11.14
11.03
$16.45
14.72
15.24
14.99
14.77
$20.13
17.85
19.04
19.20
19.24
Percent Change
1973-79
-2.8%
1979-89
17.1
1989-91
-4.2
1973-91
22.8
-5.2%
-9.9
-2.1
-16.4
-4.5%
-6.0
-1.0
-11.1
-10.57o
1.8
-1.5
-10.2
-12.1%
7.6
0.2
-5.3
40.8%
41.8
41.3
40.5
40.0
18.0%
18.0
21.9
22.3
23.1
9.1%
10.9
13.7
14.0
14.7
4.5%
4.9
6.3
6.9
7.0
Share of Workforce*
1973
1979
1987
1989
1991
24.8%
21.9
14.2
13.7
12.6
'Since the shares of those with one year of schooling beyond college are not shown, the presented
shares do nol sum to one hundred.
162
Wages received a similar wage boost from greater education and experience (on an annual basis) over both the
1973-1979 and the 1979-1989 periods.
It is also useful to examine the column in Table 3 1 7
labeled "Other Wage G r o w t h , " w h i c h tells us what the
change i n wages w o u l d have been had the workforce not
become more educated and experienced. Or, in other
words, this column tells us the average wage trends among
workers w i t h similar education and experience. The results
show that economic forces dramatically reduced the wages
paid for particular types of workers, yielding a 17.7%
decline from 1973 to 1991. Therefore, the greater education and experience of the workforce i n recent years was
necessary to counter the strong wage reductions among
groups of workers w i t h similar education and experience.
The data also shed some light on the reasons for the closing of the gender wage gap. Men's wages received a lesser
boost from greater education and experience than did
women's wages, 10 5% versus 13 5%. Plus, the economic
forces reducing wages for workers of similar education and
experience ("Other Wage G r o w t h " ) had twice the effect on
men's wages as they d i d on women's wages, 19.7% versus
8.7%, over the entire 1973-1991 period. It was in the
1979-1989 period that the effect of economic forces particularly disadvantaged men relative to women. Other sections of this chapter detail some of the factors involved such
as international trade, the shift to low-wage industries, and
deunionization.
Table 3.18 separates out the wage effects of more education and more experience. These data show that it was educational upgrading rather than experience accumulation
that generated the strong wage effects i n Table 3.17. Educational upgrading—the shifting of the workforce to proportionately more high school and college graduates and fewer
high school dropouts—raised men's and women's wages,
respectively, by 6.5% and 10.6% between 1973 and 1991
These data also show that it was in the 1979-1989 period
that the experience levels of the workforce grew and that
educational upgrading was weaker (on an annual basis).
163
The greater
education and
experience of the
workforce in
recent years was
necessary to
counter the strong
wage reductions
among groups of
workers with
similar education
and experience.
�FIGURE 3E
Change in Real Hourly Wage By Education,
1973-1991
Rising Education Wage Differentials
22
20
IS
16
I
>.
14
12
O
I
10
S
8
6 -J
1973
1
1
1
1976
1979
1982
U
1985
._
1
1988
1
1991
As we have seen, the g r o w t h i n education achievement
affects the level of wages. Changes in the economic returns
or payoff to greater education, however, affects the structure of wages by changing the wage gaps between different
educational groupings. The g r o w t h in "education wage differentials" led to greater wage inequality in the 1980s and
helps explain the faster wage g r o w t h among high-wage
workers. This section examines wage trends among workers at different levels of education and begins the discussion of the causes of rising education wage differentials.
Table 3-19 presents the wage trends and employment
shares (percentage of the workforce) for workers at various
education levels over the 1973-1991 period. It is common
to point out that the wages of "more-educated" workers
grew faster than the wages of "less-educated" workers from
1979 to 1989, w i t h the real wages of "less-educated" workers falling sharply. This pattern of wage g r o w t h is sometimes described i n terms of a rising differential between
the wages of the college-educated and high school-educated workforces. For instance, college-educated workers
earned 37.7% more than high school graduates in 1979
(*14.72 versus 110.69), but 55.7% more i n 1989 ( F i g u r e
3E).
The usual terminology of the "less-educated" and "moreeducated" turns out to be somewhat misleading. Given that
workers w i t h some college education (from one to three
years) also experienced falling real wages, it is apparent
that the "less-educated" group w i t h falling wages comprises
more than three-fourths of the workforce. Moreover, the
"college-educated" group that did w e l l consists of those
w i t h just four years of college w h o enjoyed a minimal 1.8%
wage gain from 1979 to 1989 as w e l l as the more-educated
(college plus at least t w o more years of schooling), but
smaller, group that enjoyed better wage growth.
This increased differential between college-educated and
other workers is frequently ascribed to an increased relative employer demand for workers w i t h greater skills and
education. This interpretation follows from the fact that the
wages of college-educated workers increased relative to
others despite a sizeable increase in their relative supply
from 10.9% of the workforce i n 1979 to 14.0% in 1989.
164
165
The usual
terminology ol the
"less-educated"
and "moreeducated"
turns out to
be somewhat
misleading. Given
that workers with
some college
education also
experienced
falling real wages,
it is apparent
that the "lesseducated" group
with falling wages
comprises more
than three-fourths
of the workforce.
�TABLE 3.20
Change in Real Hourly W g for M n b Education, 1973-1991
ae
e y
(1991 Dollars)
High
School
Drop-out
High
School
Graduate
Some
College
College
College
2 + years
$11.48
11.01
9.35
9.01
8.45
$13.50
12.77
11.55
11.15
10.72
$14.08
13.80
13.01
12.65
12.49
$18.99
17.08
17.55
17.11
16.69
$21.09
19.16
20.85
21.02
21.11
-4.2%
-18.2
-6.2
-26.4
-5.4%
-12.7
-3.8
-20.6
-2.0%
-8.3
-1.3
-11.3
-10.0%
0.2
-2.5
-12.1
-9.2%
9.7
0.3
0.4
37.1%
38.0
39.0
38.7
38.5
18.6%
17.8
20.7
21.0
21.8
9.4%
11.4
14.1
14.2
14.7
5.6%
6.0
7.3
7.8
7.8
Year
Real Hourly Wage
1973
1979
1987
1989
1991
Percent Change
1973-79
1979-89
1989-91
1973-91
Share of Workforce*
1973
1979
1987
1989
1991
26.6%
24.0
16.4
15.9
14.7
•Since the shares of those with one year of schooling beyond college are not shown, the presented
shares do not sum to one hundred.
166
That is, given the increased supply of educated workers, the
fact that their wages were bid up implies a strong growth in
employer demand for more-educated workers, presumably
reflecting technological and other workplace trends.
An increased relative demand for educated workers is
only a partial explanation, especially if ascribed to a benign
process of technology or other factors leading to a higher
value of education, thus bidding up the wages of moreeducated workers. Note, for instance, that the primary reason for an increased wage gap between college-educated
and other workers is the precipitous decline of wages
among the non-college-educated workforce and not any
strong growth of the college wage. Moreover, as discussed
below, there are many other factors, such as the shift to lowwage industries, deunionization, and import competition
that are important factors.
Perhaps the most important new development is that
the real wages of college graduates have begun to fall.
From 1989 to 1991 the wages of college graduates fell
1.5% and the wages of those with education beyond college rose a minimal 0.2%. It is critical to note that this
trend pre-dates the recession. The college wage began falling in 1987, two years before unemployment started rising
in 1989. This suggests that the fall-off in the college-wage
reflects long-term rather than cyclical factors. Moreover,
the fact that the share of college-educated workers in the
workforce increased slowly from 1987 to 1989, and more
slowly than from 1979 to 1987, suggests that it was a slowdown in the growth of demand for educated workers
rather than any oversupply that was forcing wages down.
This slowdown in demand for college-educated workers
probably reflects the misfortunes of several white-collar
intensive industries in the late-1980s (finance, banking,
insurance, retail stores, and real estate) as well as the
shrinking of middle-management in manufacturing.
The pattern of wage growth in the 1970s was far different. In that period it was the wages of college educated
workers that fell sharply—a drop of 10.5%. Given the sharp
drop in the college wage from 1973 to 1979, the fact that
the college wage grew only a modest 1.8% from 1979 to
1989 and fell during the 1989-1991 recession years has
meant that the college-wage was a remarkable 10.2% less
in 1991 than in 1973. In fact, the wages of every educational group fell over this time period. This finding rein167
Perhaps the most
important new
development is
that the real
wages of college
graduates have
begun to fall.
The college wage
began falling in
1987, two
years before
unemployment
started rising
in 1989.
�TABLE 3.21
Change in Real Hourly W g for W m n b Education, 1973-1991
ae
o e y
(1991 Dollars)
High
School
Drop-out
High
School
Graduate
Some
College
College
College
2 + years
$7.10
7.07
6.44
6.24
6.29
$8.66
8.34
8.25
8.07
8.10
$9.78
9.28
9.59
9.63
9.60
$12.78
11.14
12.41
12.54
12.65
$18.01
14.63
16.14
16.48
16.57
- 0.4%
-11.7
0.8
-11.4
-3.7%
-3.2
0.4
-6.4
-5.2%
3.8
-0.3
-1.9
-12.9%
12.6
0.9
-1.0
-18.7%
12.6
0.5
-8.0
46.2%
46.6
44.0
42.6
41.8
17.2%
18.4
23.3
23.9
24.5
8.6%
10.2
13.2
13.8
14.8
2.9%
3.4
5.2
5.8
6.1
Year
Real Hourly Wage
1973
1979
1987
1989
1991
Percent Change
1973-79
1979-89
1989-91
1973-91
Share of Workforce*
1973
1979
1987
1989
1991
22.2%
19.0
11.7
11.2
10.2
forces our earlier conclusion that had there not been a significant increase in education levels since 1973 there
would have been a much more severe fall in average wages.
Table 3.20 and Table 3.21 present wage and employment share trends for the various education groups for,
respectively, men and women. Among men, the decline in
wages among non-college-educated workers was stronger
than that for women and the rise in the college wage was a
trivial 0.2% (which was wiped out between 1989 and
1991). The wage of the average high school-educated man
fell 12.7% from 1979 to 1989, with the wages of those with
some college education falling somewhat less, 8.3%.
Among men, the post-1987 drop in the college-wage was
also stronger than among women, falling 4.9% from
$17.55 to $16.69. The fact that the college-wage fell more
among men than among women while there was a slower
growth in the supply of college-educated men than of
women once again suggests a significant slowdown in
employer demand for college-educated men. The fact that
the wage among college-educated women has continued
to grow from 1987 to 1991 despite a continued fast
growth in their supply implies a continued relative growth
in the demand for more educated women.
Even though the wages of college-educated women grew
rapidly since 1979, a woman college graduate in 1989 still
earned less than a man with "some college" education
(512.54 versus $12.65) or than what a high school-educated man earned in 1979 (*12.77).
Young Workers Have Been Hurt Most
'Since the shares of those with one year of schooling beyond college are not shown, the presented
shares do not sum to one hundred.
168
Since 1973, the wages of younger workers have been falling faster than the wages of older workers (Table 3.22). As
a result, there have been significant changes in the wage
differentials between younger and older workers. The real
hourly wages of workers with from 1 to 15 years of experience fell at least 10% from 1973 to 1991. In contrast, the
wages of workers with a minimum of 26 years of experience fell by less than 5% in the same time period. This
wage shift against young workers was especially pronounced among men.
169
The wage of the
average high
school-educated
man fell 12.7%
from 1979 to
1989, with the
wages of those
with some college
education falling
somewhat less,
8.3%.
�TABLE 3.22
Change in Real Hourly Wage By Work Experience, 1973-1991
Years of Work Experience
Years
1-5
6-10
11-15
16-20
21-25
26-30
31-35
35 +
-4.7% -6.6%
-9.2
-7.2
-3.1
-1.4
-16.2
-14.6
-3.6%
-5.5
-2.0
-10.7
-7.7%
-0.6
-2.1
-10.1
-4.7%
0.0
-0.7
-5.4
-1.0%
-2.4
0.5
-2.8
-0.5%
-1.6
-2.4
-4.4
2.8%
-3.6
-1.7
-2.6
-3.4% -5.7%
-14.1
-10.4
-3.7
-3.3
-20.1
-18.3
-3.4%
-9.7
-2.6
-14.9
-5.1%
-4.5
-3.6
-12.6
-6.2%
-1.3
-3.0
-10.2
-1.2%
-2.3
-0.7
-4.2
-0.2%
-1.3
-3.5
-4.9
2.6%
-2.3
-3.3
-3.0
0.4%
8.4
-1.1
7.6
-7.5%
11.7
1.4
4.8
-0.3%
12.0
1.7
13.6
-0.9%
5.6
3.7
8.5
-0.3%
4.9
0.8
5.5
1.4%
2.0
1.4
4.9
All
1973-79
1979-89
1989-91
1973-91
Since the wages o f both younger and non-college-educated workers have fallen most rapidly, it follows that the
wages o f workers w h o are both young and non-college educated have dramatically fallen. These adverse wage trends
were strongest among men. Table 3.23 and F i g u r e s 3F
and 3G present trends i n "entry level" wages for high
school and college graduates as reflected i n the wages of
workers w i t h from one to five years of experience. The
entry level hourly wage of a young, male high school graduate i n 1989 was 22.4% less than that for the equivalent
worker i n 1979, a d r o p of S2.11 per hour. The accumulated
The entry level
hourly wage of a
young, male high
school graduate
In 1989 was
22.4% less than
that for the
equivalent worker
in 1979, a drop of
$2.11 per hour.
Men
1973-79
1979-89
1989-91
1973-91
Women
1973-79
1979-89
1989-91
1973-91
-5.9%
-1.6
-2.1
-9.5
-5.9%
2.6
1.6
-1.9
TABLE 3.23
Entry Level Wages and Employment Shares, 1973-1991
(1991 Dollars)
College Graduates
Year
All
Women
All
$9.75
9.39
7.46
7.28
6.90
$7.56
7.12
6.32
6.16
6.02
$12.52
11.32
11.70
11.72
11.30
-3.8%
-22.4
-5.2
-29.3
-5.8%
-13.5
-2.3
-20.4
38.5%
41.2
38.1
36.4
36.5
45.2%
43.9
37.1
35.1
34.0
Men
Men
Women
Entry-Level Hourly Wage*
1973
1979
1987
1989
1991
$8.69
8.32
6.91
6.74
6.48
$13.39
12.57
12.54
12.32
11.93
$11.54
10.07
10.96
11.17
10.75
- 9.6%
3.5
-3.6
-9.8
-6.1%
-2.0
-3.2
-10.9
-12.8%
11.0
-3.8
-6.9
14.1%
14.9
19.0
19.1
19.4
13.3%
13.7
17.5
17.3
17.6
15.2%
16.3
20.5
21.1
21.4
Percent Changes
1973-79
1979-89
1989-91
1973-91
-4.2%
-18.9
-3.9
-25.4
Shares of Employmenl**
1973
1979
1987
1989
1991
41.5%
42.4
37.7
35.7
35.3
"Wage of workers with 1 to 5 years of work experience.
"Share of those with 1 to 5 years of work experience by education level within gender.
170
171
�FIGURE 3F
Real Entry Level Wages for High School Graduates,
1973-1991
wage declines from 1973 f o r w a r d , left entry level wages for
male high school graduates 29 3% lower i n 1991 than i n
1973- Among women, the entry level high school wage fell
a remarkable 20.4% from 1973 t o 1991.
Entry level wages among college graduates have fallen as
well. Following the steep wage decline o f the 1970s due to
a surplus in college graduates, the entry level college wage
fell modestly over the 1980s and into the 1990s By 1991,
the entry level college wage for men and w o m e n had fallen
10.9% and 6.9%, respectively, since 1973. The fact that
entry level wages for college graduates have fallen less than
for high school graduates in recent years means that it still
makes economic sense for individuals to complete college.
Nevertheless, those w h o obtain a college degree w i l l have
a lower wage than that obtained by an earlier generation o f
college graduates.
The Shift to Low-Paying Industries
1973
1976
1979
1962
1985
1968
1991
FIGURE 3G
Real Entry Level Wages for College Graduates,
1973-1991
1973
1976
1979
172
1982
1985
1988
1991
There was a large employment shift t o low-wage sectors
i n the 1980s that was a consequence o f trade deficits and
deindustrialization, as well as o f stagnant or falling productivity g r o w t h i n service sector industries. This section
examines the significant erosion o f wages and compensation that occurred as a result.
Recent employment g r o w t h by major industry sector is
presented i n Table 3-24. The 17.2 m i l l i o n ( n e t ) jobs created between 1979 and 1989 involved a loss of roughly 1.7
m i l l i o n manufacturing and m i n i n g jobs and an increase of
18.8 m i l l i o n jobs i n the service sector. The largest amount
of job g r o w t h (13-7 m i l l i o n ) was i n the t w o lowest paying
industries—retail trade and services (business, personnel
and health). I n fact, these t w o industries accounted for
79.7% of all the ( n e t ) new jobs over the 1979-1989 period.
The extent o f the shift to low-wage industries is more
evident in an analysis o f changes in the shares o f the workforce in various sectors (Table 3.25). Several high-wage
sectors such as construction, transportation, wholesale
trade, c o m m u n i c a t i o n s , a n d g o v e r n m e n t increased
employment in the 1980s but still had a smaller o r similar
share of overall employment in 1989 as in 1979. A lower
share o f employment i n these high-wage sectors caused
the average wage t o fall. Overall, the share of the workforce
w o r k i n g i n the low-wage services and retail trade indus-
173
By 1991, the
entry level college
wage lor men
women had fallen
10.9% and 6.9%,
respectively,
since 1973.
�tries w a s 7.1 p L T i x - n t a n i : p o i n t s h i n h u r i n
TABLE 3.24
Employment Growth b Sector, 1979-1989
y
Employment (000)
979
1989
Growth
Industry Sector
1979
point
d r o p i n t h e share o f t h e w o r k f o r c e i n h i g h - p a y i n n i n d u s -
Industry
Share
of Job
Growth
Median
Weekly
Earnings
1989
tries s u c h as m a n u f a c t u r i n g , c o n s t r u c t i o n , m i n i n g , g o v e r n m e n t , transportation, c o m m u n i c a t i o n s , and utilities.
The e f f e c t o n pay levels o f t h e e m p l o y m e n t shift t o w a r d s
l o w e r p a y i n g s e c t o r s is i l l u s t r a t e d i n T a b l e 3 . 2 6 . T h e first
c o l u m n s h o w s t h e level o f pay t h a t w o u l d have p r e v a i l e d i n
Goods Producing
Mining
Construction
Manufacturing
Durable Goods
Nondurable Goods
26,461
958
4,463
21,040
12,760
8,280
24,815
678
4,741
19,396
11,448
7,948
-1,646
-280
278
-1,644
-1,312
-332
- 9.6%
-1.6
1.6
-9.6
-7.6
-1.9
$565
$431
$415
$445
$373
Service Producing
Trans, Comm, Util
Wholesale
Retail
Fin, Ins, Real Est
. Services
Government
Total
63,363
5,136
5,204
14,989
4,975
17,112
15,947
89,823
82,211
5,549
6,195
19,115
6,639
26,702
18,011
107,026
18,848
413
991
4,126
1,664
9,590
2,064
17,203
109.6%
2.4
5.8
24.0
9.7
55.7
12.0
$502
$412
$276
$406
$357
$472
TABLE 3.25
Changes in Employment Share By Sector, 1979-1989
Share of Employment
1979
1989
Industry Sector
1989 than in
The parallel t r e n d was the 7 0 pereeniaj-e
Change in
Employment
Share
1979-1989
23.2%
0.6
4.4
18.1
10.7
7.4
76.8%
5.2
5.8
17.9
6.2
24.9
16.8
- 6 . 3 % pts
-0.4
-0.5
-5.3
-3.5
-1.8
6.3% pts
-0.5
0.0
1.2
0.7
5.9
-0.9
1989 if t h e w o r k f o r c e was i n the same occupations and
i n d u s t r i e s as i n
1 9 8 0 ( t h e earliest available year l o r t h i s
a n a l y s i s ) b u t w o r k e d at 1 9 8 9 w a g e levels i n each o c c u p a t i o n a n d i n d u s t r y . The s e c o n d c o l u m n s h o w s t h e a c t u a l pay
levels i n 1 9 8 9 . T h e d i f f e r e n c e b e t w e e n t h e pay levels i n t h e
The largest
amount of job
growth was in
retail trade and
services. These
two industries
accounted for
79.7% of (net)
new jobs over
1979-1989.
TABLE 3.26
Effect o Structural Employment Shifts o Pay Levels, 1980-89
f
n
Effect of
1980-89 Employment
Composition Shift
1989 Pay Levels
With Employment
Composition of:
Occupation Group
Dollars
Percent
1980*
1989*
$12.05
1.42
14.79
$11.72
1.29
14.28
$-0.33
-0.13
-0.51
- 2.7%
-8.8
-3.4
$13.76
1.47
16.46
$13.88
1.42
16.57
$0.12
-0.05
0.11
0.8%
-3.1
0.6
$12.10
1.68
15.41
$11.31
1.50
14.35
0.79
0.18
1.06
- 6.5%
- 10 8
-6.9
All Workers
Wages
Benefits
Compensation
White Collar
29.5%
1.1
5.0
23.4
14.2
9.2
70.5%
5.7
5.8
16.7
5.5
19.1
17.8
Goods Producing
Mining
Construction
Manufacturing
Durable Goods
Nondurable Goods
Service Producing
Trans, Comm, Util
Wholesale
Retail
Fin, Ins, Real Est
Services
Government
17 i
Wages
Benefits
Compensation
Blue Collar
Wages
Benefits
Compensation
'March 1989 pay levels using 1980 Census weights corrected lor differences between household and
employer surveys
** March 1989 pay levels using current 1989 weights
I ^5
�first two columns
TABLE 3.27
Pay In Expanding and Shrinking Industries, 1948-1987
(1987 Dollars)
and
thus reflects
occupational
the first two columns
1954-62
1962-73
1973-81
1981-87
compensation.
0.91
0.78
0.78
0.80
1.09
stales
the
between
as a measure of the effect of changes
of employment
in this time period
adverse
luiwccn
difference
on wages, henctits
Since there was a shift towards
ing occu/>utions
Annual Rate of
Industry Job Shift*
iiKlusirial
ol employmenl
1980 and 1989. We use the percentage
in the composition
1948-54
the change in the
composition
effect
higher
this comparison
of the
shift
to
IIIKICT
lower-paying
nuiuslrws.
The c h a n g e i n t h e c o m p o s i t i o n o f e m p l o y m e n t
Wages and Salaries
and
pay-
from
1980 t o 1 9 8 9 led t o a 2 . 7 % f a l l i n w a g e s , an H.8", f a l l i n
Expanding Industries
Shrinking Industries
Wage Gap**
$14,938
14,204
734
$13,648
18,609
-4,961
$16,598
16,765
-167
$19,144
19,058
86
$19,154
26,194
- 7,040
b e n e f i t s , a n d a .-V-i'A', t a l l i n t o t a l c o m p e n s a t i o n . W h i t e - c o l l a r
pay levels, e x c e p t b e n e f i t s , w e r e o n l y m i l d l y a l l c c l c d by
recent compositional
however,
Benefits
shifts. A m o n g b l u e - c o l l a r
the changing
industrial
workers,
composition
of
jobs
l o w e r e d w a g e s a n d c o m p e n s a t i o n by m o r e t h a n (>%.
Expanding Industries
Shrinking Industries
Benefits Gap**
$817
951
-134
$839
2,018
-1,179
$1,603
2,091
-488
$2,778
3,672
-894
$2,829
6,194
-3,365
T h e degree to w h i c h the e m p l o y m e n l shift, f r o m high- to
l o w - p a y i n g i n d u s t r i e s , l o w e r s average pay levels d e p e n d s
o n t w o factors. O n e factor is t h e a m o u n t o f s h i l l i n g that
o c c u r s , i.e., t h e p e r c e n t a g e o f t h e w o r k f o r c e t h a i is s h i t t i n g
o u t o f h i g h - p a y i n g sectors. T h e o t h e r f a c t o r is t h e size of
Compensation
Expanding Industries
Shrinking Industries
Compensation Gap**
$15,753
15,156
597
$14,491
20,623
-6,133
$18,206
18,853
-647
$21,921 $21,983
22,732 32,387
-811 -10,404
t h e pay g a p b e t w e e n e x p a n d i n g a n d s h r i n k i n g i n d u s t r i e s .
B o t h factors significantly increased i n the
l9H(ls
3 . 2 7 ) . F o r i n s t a n c e , a m e a s u r e of t h e r a l e of
(Table
structural
s h i f t i n g i n each p e r i o d s h o w s that t h e r e was m o r e r a p i t l
s h i f t i n g i n t h e 1 9 8 1 - 1 9 8 7 p e r i o d t h a n i n any o t h e r p e r i o d .
The d e g r e e o f s h i f t i n g a c c e l e r a t e d b y 3 6 % , f r o m 0 . 8 0 t o
*The annual percentage points thai employmenl shares among industries are shifting.
"Dollar gap by which pay of expanding industries is higher or lower than pay in shrinking industries.
Expanding (shrinking) industries are those whose share of total employmenl increased (decreased)
over the particular time period
and the 1981-1987 periods.
Note: These are the only time intervals for which data are available
expanding and shrinking industries
1.09 p e r c e n t a g e p o i n t s p e r year, b e t w e e n t h e
1973-1981
T h e r e w a s also a n increase i n t h e pay g a p b e t w e e n t h e
In the
1981-1987
p e r i o d , the industries w h e r e e m p l o y m e n l contracted paid
JlO.-lO-t, o r 4 7 % , m o r e i n a n n u a l c o m p e n s a t i o n t h a n t h e
e x p a n d i n g i n d u s t r i e s ( T a b l e 3 . 2 7 , F i g u r e 3 H ) . T h i s was by
far t h e largest c o m p e n s a t i o n g a p b e t w e e n e x p a n d i n g a n d
s h r i n k i n g i n d u s t r i e s i n the p o s t - w a r e r a T h e biggest d i f f e r ential was i n bcnclils, w i t h s h r i n k i n g industries
ofiering
m o r e l h a n t w i c e as m u c h i n a n n u a l b c n c l i l s .
T h e faster pace o f j o b s h i f t i n g a n d t h e l a r g e r pay gap
b e t w e e n e x p a n d i n g a n d s h r i n k i n g s e c t o r s i n t h e 1980s l e d
to a s i g n i f i c a n t
|7(,
reduction
in
both
wages
and
benefits
In Ihe 1981-1987
period, the
industries where
employment
contracted paid
$10,404, or 47%,
more in annual
compensation
than the
expanding
industries.
�FIGURE 3H
Pay in Expanding and Shrinking Industries,
1948-1987
B
1948-1954
(Table 3.28). Industry employment shifts reduced compensation growth by 0.48% a year from 1981 to 1987, far
higher than i n any other period. The compensation of the
overall workforce was thus reduced by 3% from 1981 to
1987 because of the shift to lower paying sectors.
The expansion of low-wage industries reduced the wage
growth of production and nonsupervisory workers by
twice as much as for the average worker. As previously mentioned, production and nonsupervisory workers make up
over 80% of all wage and salary employment. I n the most
recent period for w h i c h data arc available, 1981-1986, the
wages paid by expanding industries were J 134.69 less per
week than those paid by shrinking industries (Table 3.29).
This pay gap, combined w i t h an even faster rate of job shifting, lowered wage g r o w t h by 0 . 6 1 % annually. This compares to the 0.32% wage reduction induced by industry
shifting among all workers (Table 3.28).
Expanding Industries
I I Shrinking Industries
113,646
1954-1962
18.609
1962-1973
1973-1981
The compensation
ot the overall
workforce was
thus reduced by
3% from 1981 to
1987 because of
the shift to lower
paying sectors.
1981-1987
5,000
10,000
15,000
20,000
25,000
30,000
Average Annual Compensation (1987 Dollars)
TABLE 3.29
Industry Employment Shifts and Production Worker
Weekly Wages, 1972-1986
(1987 Dollars)
TABLE 3.28
Effect of Industry Employment Shifts on Pay, 1948-1987
Effect of Industry Shifts on Annual Pay Growth
Pay
Wages
Benefits
Compensation
1948-54
.05%
.01
.04
1954-62
1962-73
1973-81
1981-87
- .22%
- .05
- .28
.01%
-.02
-.02
.00%
-.03
-.03
-.32%
-.15
-.48
Year
1972-81
1981-86
Weekly Wage
Rate of
Industry
Job Shift*
Expanding
Industries
Shrinking
Industries
Gap
1.23
1.46
$295.65
287.51
$ 61.13
422.20
$-65.48
-134.69
Shift Effect
on Annual
Wage Growth
-0.24%
-0.61
'The annual percentage points that employment shares among industries are shifting.
Note: These are the only time intervals for which data are available.
178
Note: These are the only time intervals for which data are available.
179
�Trade and Wages
TABLE 3.30
Trade Induced Changes in Labor Supply, 1967-1985*
Implicit Labor Input Represented by:
As Percent of
Total Work Hours
Year
As Percent of
Mfg. Hours
Imports
Exports
Trade
Balance
Trade
Balance
1967-69
1970-72
1973-75
1976-78
1979-81
1985
1.31%
1.64
1.82
2.06
2.31
3.49
1.57%
1.69
2.17
2.26
2.75
2.15
-0.26%
-0.05
-0.35
-0.20
-0.44
1.34
-0.87%
-0.18
-1.39
-0.80
-1.86
6.36
Change,
1979-81
to 1985
1.18%
-0.60%
1.78%
8.22%
'Equivalently, these numbers can be viewed as increases or decreases in labor demand.
180
Increased international trade and the emergence of large
trade deficits in the 1980s were important factors in slowing wage growth and in reducing the wages of non-collegeeducated workers. The mechanism by which international
trade affects wages is that increased imports (or lower
exports) cause certain types of jobs, e.g., manufacturing or
blue-collar jobs, to be lost or to grow more slowly. Displaced workers, or those not hired in manufacturing
because of trade pressures, then compete with similar
workers in other sectors, and usually accept lower wages.
The increased competition lowers the wages of many
workers not directly affected by fewer expons or greater
imports. International trade can also affect the wages of
workers in the tradeable goods sector who are not displaced, by forcing wage "restraint" or concessions in order
to maintain or improve exports or to deter imports.
This section examines the effect of international trade on
the wage structure in recent years. Because all of the ways
in which international trade affects wages have not been
examined by economists, the discussion is necessarily
incomplete and understates the effect that trade had on
wages in the 1980s. We also consider the effect of greater
immigration. Consider that impons and exports represent
a certain amount of embodied labor. For instance, an
imponed car represents the hours worked by employees
in the auto industry and its feeder industries. Likewise, an
exported airplane represents the domestic workforce in
the aerospace industry and its feeder industries. An
increase (decrease) in impons (exports) can thus be represented by an increased supply of labor competing in the
domestic labor market, a factor which can be expected to
depress wages. Or, equivalently, greater imports can be
considered to lower the demand for domestic labor and
therefore depress wages.
Table 3.30 presents such an analysis by translating
imports and exports into their equivalent supply of labor in
various periods from 1967 to 1985. Because wc had a trade
surplus (exports exceeding imports), the net effect of trade
flows until the 1980s was to lower the size of the workforce (measured in total hours), albeit slightly. For instance,
the fact that the labor embodied in exports in the
181
Increased
international trade
and the
emergence ot
large trade
deficits in the
1980s were
important factors
in slowing wage
growth and in
reducing the
wages of
non-collegeeducated workers.
�FIGURE 31
Trade Induced Changes in Labor Supply,
1967-1985
2
?
3
O
X
o>
c
c
n
t3
«
c
(0
c
2
2.
1967-1969
1970-1972
182
1973-1975
1976-1978
1979-1981
1985
1979-1981 period (2.75% of w o r k hours) was greater than
the labor embodied i n imports ( 2 31 % of work hours) had
the effect of decreasing the overall labor supply by 0.44%.
It is as if 0.44% of the workforce moved abroad to compete
w i t h foreign workers, and therefore no longer competed
w i t h U.S. workers for jobs. A l l things equal, a cut i n domestic labor supply w o u l d be expected to raise wage levels.
The g r o w t h of imports and the decline i n exports over
the early 1980s (i.e., the rise of the trade deficit) meant the
effect of our trade balance i n 1985 was to raise the effective labor force by 1.34%. Our trade failures in the early
1980s thus can be characterized as augmenting the labor
force by 1.78% (the shift from the 1979-1981 period to
1985).
O f course, trade developments d o not fall o n all sectors
o f the workforce equally but, in fact, impinge direcdy o n
the manufacturing workforce. When o u r trade balance is
translated into the equivalent share o f the manufacturing
workforce (in the last column) we see that the deteriorating trade deficit was effectively equivalent to increasing the
manufacturing workforce by 8.22% (Table 3.30, Figure
31). These data indicate that changes in trade flows can
have a sizeable effect o n the domestic workforce. This is
especially the case since these data only consider the direct
effects o f trade flows (the loss o f steel jobs to steel
imports), but not the indirect effects (the loss o f iron ore
and coal jobs w h e n domestic steel production declines).
Table 3.31 examines the size of the immigrant labor
force in total and by education level. Increased immigration
i n the 1980s increased the immigrant share of the male
(female) labor force from 7.0% to 9 9% (6.8% to 8.6%)
from 1980 to 1988. Immigrants accounted for twenty-five
percent of the increase i n the labor force over this time
period (not shown i n the Table).
Did the increase i n immigration adversely affect particular segments of the native workforce? Table 3 31 shows the
share of the immigrant and native workforces at particular
education levels. Overall, the native and immigrant workforces had similar education levels, particularly in the percentage w i t h a college degree. The only striking difference
is that immigrants are much more likely to be high school
dropouts, a disparity that grew stronger over the 1980s.
This suggests that immigration may have adversely affected
high school dropouts relative to college graduates.
183
Increased
immigration in the
1980s increased
the immigrant
share ol the n
(female) labor
force from 7.0%
to 9.9% (6.8%
to 8.6%) from
1980 to 1988.
Immigrants
accounted tor
twenty-five
percent of the
increase in the
labor force over
this time period.
�TABLE 3.31
Immigration and Labor Supply, 1980-1988
Distribution of Workers
High School
Drop-out
High School
Graduate
Some
College
College
or More
Total
Immigrant
Share of
Labor Force
Immigrant Men
1980
1988
39.6%
36.0
22.4%
22.6
16.3%
16.2
21.7%
25.1
100.0%
100.0
7.0%
9.9
35.7%
28.3
30.3%
27.6
17.9%
21.8
16.0%
22.3
100.0%
100.0
6.8%
8.6
22.7%
15.3
37.9%
36.0
19.2%
24.6
20.2%
24.0
100.0%
100.0
44.9%
40.6
20.9%
27.6
16.0%
20.7
100.0%
100.0
Immigrant Women
1980
1988
Native Men
1980
1988
Table 3-32 carries f o r w a r d the analysis of Table 3 30 by
translating the overall effect of trade into the effects on the
size of (or demand for) the high school and college-educated workforce. These data show that the deterioration of
the trade deficit from 1980 to 1985/86 raised the size of
the high school-educated workforce by 3-1% and the college-educated workforce by 0.5% ( i n effect, reducing it by
1.2% rather than 1.7%). This suggests that trade developments had a particularly adverse effect on high school-educated workers ( w h o are about two-thirds of the entire
workforce), absolutely, as w e l l as relative to college graduates. W i t h the decline of the trade deficit i n the late 1980s
the overall effect of trade lessened, but remained large.
Table 3.33 presents estimates of the effect of trade and
i m m i g r a t i o n developments o n the wage d i f f e r e n t i a l
between college- and high school educated workers The
impact on the wage differential depends on t w o factors.
One is the impact of trade and immigration on raising the
Trade
developments i
a particularly
adverse effect on
high schooleducated workers
(who are about
two-thirds of the
entire workforce),
absolutely, as i '
as relative to
college
graduates.
Native Women
1980
1988
18.2%
11.0
TABLE 3.32
Effect o Trade o Labor Supply of
f
n
High School and College Workers, 1980-1988
Trade Induced Increased (Decreased) Labor Supply*
Year
High School
Equivalents**
College
Equivalents"
Share of Workforce
1980
1985-86
1987-88
0.4%
3.5
2.5
-1.7%
-1.2
-1.5
3.1%
2.1
0.5%
-0.2
Change
1980-85/86
1980-87/88
'This analysis expresses a trade deficit as an increased supply of workers (i.e., as if those producing
imports were augmenting domestic labor supply). Equivalently, these data could be interpreted as
reduced demand (positive numbers) or increased demand (negative numbers).
"In this analysis the "high school drop-out" and "some college" workforce has been assigned to be
either a high school or college equivalent.
184
185
�TABLE 3.33
Effect of Trade and Immigration on
Education W g Differential, 1980-1988
ae
Time Period
Item
1980-85
1980-88
Change in College/High School
Wage Differential*
10.7% pts.
12.4% pts.
2.0% pts.
1.3% pts.
1.9
1.0
Change in Wage Differential
Due to Trade and Immigration
with Elasticity":
(a) 0.709
(b) 1.0
(c) 0.5
2.9
1.5
Contribution of Trade
and Immigration to
Change in Wage Differential
with Elasticity**:
(a) 0.709
(b) 1.0
(c) 0.5
As the first line of Table 3.33 shows, the wage gap
between high school- and college-educated workers grew
by 10.7 percentage points i n the early 1980s and by 12.4
percentage points by 1988. The estimates of the effects of
trade and immigration on the change in the wage differential through 1985 range from 1.5 to 2.9 percentage points.
Over the longer period the effect is from 1.0 to 1 9 percentage points. Overall, trade and immigration can account for
from 14% t o 27% of the higher differential over the early
1980s or from 8% to 15% of the higher differential over
the longer period through 1988.
The Union Dimension
19%
27
14
10%
15
8
'The differential and effects of trade and immigration are computed as changes in log wages, which
approximates percentage or percentage point changes.
"The "elasticity" gives the percentage change in the college/high school wage differential resulting from
a percentage change in the relative supply of college graduates.
186
labor supply of ( o r lowering the demand for) high schoolrelative to college-educated workers. The second factor is
the degree t o w h i c h a given increase in the relative supply
of (or decrease i n the demand f o r ) high school workers
translates into a change in the wage differential The uncertainty about the size of this second factor—the "elasticity"
of the wage differential w i t h respect to relative supplies—
motivates the need to examine the effect of trade and
immigration using a range of estimates of the "elasticity."
The percentage o f the workforce represented by unions
fell more rapidly i n the 1980s than in the previous several
decades. This falling rate o f unionization has lowered
wages, n o t only because some workers no longer receive
the higher u n i o n wage, but also because there is less pressure o n n o n u n i o n employers t o raise wages.
Table 3.34 and F i g u r e 3J show the union wage prem i u m — t h e degree to w h i c h union wages exceed nonunion wages—by sector, by occupation, and by type of pay
(benefits or wages). The union premium is larger for total
compensation, 33 8%, than for wages alone, 25 7%. This
reflects the fact that unionized workers arc paid insurance
and pension benefits w h i c h are more than double those o f
non union workers. For blue-collar workers, the union prem i u m in insurance and benefits is even larger, w i t h union
blue-collar workers receiving from 157.6% to 278.3%
more than their non union counterparts. The union wage
differential is also much higher among blue-collar workers.
The union premium is generally higher w i t h i n the servicethan the goods-producing sector.
187
Trade and
immigration can
account for from
14% to 27% of
the higher
differential over
the early 1980s
or from 8% to
15% of the higher
differential over
the longer period
through 1988.
�TABLE 3.34
Comparison of Union and Non union Houriy Wages and Benefits, 1989
(1989 Dollars)
Occupation
or Industry
Total
Compen-
Group
sation
Wage
Insurance
Pension
All Workers
Union
Nonunion
Union Premium*:
Dollars
Percent
$16.47
12.31
$14.13
11.24
$1.52
0.72
$0.76
0.35
$4.16
33.8%
$2.89
25.7%
$0.80
111.1%
$0.41
117.1%
$17.46
10.40
$4.82
9.49
$1.70
0.66
$0.87
0.23
$7.06
67.9%
$5.33
56.2%
$1.04
157.6%
$0.64
278.3%
Collar
Union
Nonunion
Union Premium':
Dollar
Percent
The effect of the erosion of unionization o n the wages of
a segment o f the workforce depends o n the degree to
which deunionization has taken place and the size o f the
union wage premium among that segment of the workforce. Table 3.35 shows both the degree to w h i c h unionization declined and the union wage premium by occupation
and education level. These data are used to calculate the
effect of deunionization over the 1978-1988 period o n the
wages o f p a r t i c u l a r groups and h o w d e u n i o n i z a t i o n
affected occupation and education wage differentials
Union representation fell dramatically among blue-collar
and high school-educated workers from 1978 to 1988.
Among the high- school graduate workforce 12% fewer
were unionized i n 1988 than i n 1978, a more than 25%
shrinkage of the unionization rate from 42% to 3 0 % .
Because unionized high school graduates earned 16% more
than equivalent non union workers, the loss o f unionization
lowered the average high school graduate's wage by 1.4%.
The net effect of deunionization was to increase the college
to high school wage differential by 1.5 percentage points.
FIGURE 3J
Union Pay Premiums,
1989 Dollars
Goods-Producing
Union
Nonunion
Union Premium":
Dollar
Percent
$17.84
14.6
$15.00
13.15
$1.85
1.04
$0.89
0.44
$3.18
21.7%
$1.85
14.1%
$0.81
77.9%
Because
unionized high
school graduates
earned 16% more
than equivalent
non union
workers, the loss
of unionization
lowered the
average high
school graduate's
wage by 1.4%.
$0.45
102.3%
Service-Producing
Union
Nonunion
Union Premium":
Dollar
Percent
$15.11
11.57
$13.25
10.62
$1.20
.62
$0.64
0.32
$3.54
30.6%
$2.63
24.8%
$0.58
93.5%
$0.32
100.0%
O
L
40 -
"The dollar and percent by which the pay of union workers exceeds that of nonunion workers.
All Workers
188
Blue Collar
Goods-Producing
189
Service-Producing
�Deunionization had a larger negative effect on blue-collar
wages—lowering them 3 6 % — a n d thus a larger effect on
the occupational than the educational wage differential.
Less unionization can account for 25% and 50% of, respectively, the higher educational and occupational wage differentials i n 1988.
TABLE 3.35
Effect of Deunionization on Male Occupation
and Education Differentials, 1978-1988
A. Effects of Union Decline on Wages*
Group
Union
Wage
Premium*
Percent Union
1978
1988
Change
1978-88
Effect of
Union Decline
On Wages
- 5
-14
Table 3.36 presents a similar analysis for workforce
groups by race, gender and education level In 1979, unionization rates were higher among blacks than whites and
-0.1%
-3.6
Deunionization
had a larger
negative effect
on blue-collar
wages—lowering
them 3.6%.
By Occupation
White-Collar
Blue-Collar
Difference
1%
26
3.5
18%
47
13%
33
TABLE 3.36
Impact of Deunionization on Changes in
Average Hourly Earnings, 1979-1989
Percent Union
By Education
College
High School
Difference
-2%
16
1.5
17%
42
14%
30
- 3
-12
0.1%
-1.4
(D
Change In
Wage Differential
1978-88*
Union Decline
Effect on
Wage Differential
White-Collar/Blue-Collar
College/High School
7%
6
1979 1989
Change
1979-89
Union
Premium
in 1979**
Effect of
Union Decline
On Wages
White Males
B. Contribution of Deunionization to Higher Wage Differentials*
(2)
Demographic/
Education Group*
(3)
Percent of
Higher Differential
Explained by
Union Decline
3.5%
1.5
50%
25
High School Drop-out
High School Graduate
College or More
48%
47
24
24%
32
17
-24
-15
- 7
24%
10
-1
-6%
-1
0
42%
56
41
29%
30
42
-13
-26
1
22%
18
-12
-3%
-5
0
27%
19
35
11%
12
25
-16
- 7
-10
14%
11
15
-2%
-1
-2
32%
43
52
18%
24
28
-14
-19
-24
17%
10
21
-2%
-2
-5
Black Males
High School Drop-out
High School Graduate
College or More
White Females
'Differentials and change in differentials computed in natural log percentage points.
"Controlling for workforce characteristics.
High School Drop-out
High School Graduate
College or More
Black Females
High School Drop-out
High School Graduate
College or More
'Workers ages 25-64.
"Controlling for workforce characteristics.
190
191
�TABLE 3.37
Effect of Unions on Wages, By Wage Fifth, 1973-1987
Lowest Second Middle
Fifth
Fifth
Fifth
Fourth
Fifth
Top
Fifth
Average
Percent Union
1973
1987
Change, 1973-87
38.9%
43.7%
23.5
30.3
-15.4
- 13.4
38.3%
33.1
-5.2
33.5%
24.7
-8.8
18.0%
6.0
0.9%
0.2
12.5% 33.7%
26.4
17.7
-7.3
7.2
Effect of Union on:
Union Wage, 1987
Average Wage, 1987
27.9%
6.6
Wage Effect of Deunionization
1973-87
-4.3%
16.2%
4.9
10.5%
2.1
-2.2% -0.9% - 0 . 1 %
15.9%
4.2
0.8% - 1 . 1 %
TABLE 3.38
Effect of Unions on Male W g Inequality
ae
Wage Inequality,
1973-87*
.235
.269
.034
.012
.007
21%
Early Year
Later Year
Change in Inequality
Change Due to Lower
Unionization
Deunionization Contribution to Total
Rise in Inequality
Wage Inequality,
1978-88*
.227
.284
.057
Item
among men than women. Among men of both races, the
decline in unionization was greatest for high school dropouts and graduates. Because of the relatively high-unionization rates of black and white w o m e n college graduates i n
1979, the extent of deunionization among women was sizeable for all educational groups. The effect of deunionization
on wages was as high as 6% for white high school dropouts
and 5% for black high school graduates. Among men, the
pattern of deunionization's impact o n wages led to a sizeable increase in the differential between less educated and
more educated workers.
Table 3.37 examines the effect o f lower unionization on
workers at various wage levels. The data show that unions
have their largest effect on the wages of lower-wage workers, as unions raise the wages of union members i n the lowest and second lowest fifths by respectively, 27 9% and
16.2%. Because workers i n the b o t t o m three-fifths have
higher unionization rates and higher union wage premiums,
the effect of unions on average wages for these groups is
largest—increasing the average wage from 4.9% to 6.6%.
Thus, unionization is a force that moderates the wage gap
between high-wage and middle- or low-wage workers.
Unionization declined more among low-wage workers
than among high-wage workers f r o m 1973 to 1987, w i t h
unionization actually increasing among the top f i f t h . The
wage impact of this deunionization was to increase the
wage gap between high- and low-wage workers.
Table 3-38 shows t w o studies of the effect o f the d r o p i n
unionization on overall wage inequality. Both studies show
that there was a sizeable g r o w t h i n wage inequality
between the 1970s and the late 1980s. Moreover, both
studies show that lower unionization can account for 2 1 %
of the higher wage inequality.
21%
'Change in variance of Ln earnings among men ages 25-64.
"Change in variance of Ln earnings among men ages 24-66.
192
193
Unionization is
a force that
moderates the
wage gap
between
high-wage and
middle- or
low-wage
workers.
�FIGURE 3K
Value of Minimum Wage,
1960-1992
An Eroded Minimum Wage
6.00
< P <P
#
^ jl* ^ ^ J>
^ ^
^> ^
N
#
The real value of the m i n i m u m wage has fallen considerably since its high point i n the late 1960s ( F i g u r e 3K). The
decline was particularly steep and steady between 1979
and 1989, when inflation whittled the real m i n i m u m wage
( i n 1991 dollars) d o w n from J5.34 to $3-68 (Table 3.39).
The legislated increases i n 1990 and 1991 raised the m i n i mum wage, but still left it 81.09 less in 1991 than in 1979.
The m i n i m u m wage's purchasing power in 1992 ( $ 4 . 1 1 )
was 20% less than the value of the m i n i m u m wage i n the
1960s or 1970s.
W h i l e a full-time, year-round job at the m i n i m u m wage
kept a family of three above the poverty line in 1979, the
same job w o u l d have placed such a family $2,300 below the
poverty line i n 1992 (Table 3.40). Even a two-person family dependent on a m i n i m u m wage worker would have
been 5606 below the poverty line i n 1992
It has been argued that the m i n i m u m wage primarily
affects teenagers and others w h o do not have family
responsibilities. Table 3-41 examines the demographic
composition of the workforce earning less than the minimum wage. In fact, 84.8% of m i n i m u m wage workers i n
The minimum
wage's
purchasing power
in 1992 ($4.11)
was 20% less
than the value of
the minimum
wage in the
1960s or 1970s.
TABLE 3.39
Value of Minimum Wage
Minimum
Minimum
Year
Wage,
Current
Dollars
Wage,
1991
Dollars
1967
1973
1979
1989
1990
1991
1992
$1.40
1.60
2.90
3.35
3.80
4.25
4.25
$5.25
4.62
5.34
3.68
3.96
4.25
4.11
Period Averages
1960s
1970s
1980s
1990-91
$1.29
2.07
3.33
4.25
$5.04
5.11
4.39
4.11
194
TABLE 3.40
Amount by Which Earnings of a Full-Time, Full-Year
Minimum Wage Worker are Above (Below) the Poverty Line
(1991 Dollars)
Dollar Difference Between Annual
Minimum Wage Earnings and Poverty
Line By Family Size
Year
2-Person
3-Person
4-Person
1967
1973
1979
1989
1990
1991
1992
$2,711
1,045
2,152
-1,510
-927
-324
-606
$1,017
-649
459
-3,204
-2,621
-2,017
-2,300
-$2,046
-3,712
-2,605
-6,268
-5,685
-5,081
-5,364
195
�TABLE 3.41
Minimum W g Workforce Demographic Composition, 1991
ae
At or Below
Current Minimum Wage
Groups
Number of
Workers
(000)
Percent of
Minimum Wage
Workforce
At or Below
Real Value of 1979
Minimum Wage
Number of
Workers
(000)
Percent of
Minimum Wage
Workforce
Gender
2,019
3,630
35.6%
64.3
6,487
10,063
39.2%
60.8
859
4,796
15.2%
84.8
2,312
14,237
14.0%
86.0
3,853
878
729
197
68.1%
15.5
12.9
3.5
11,229
2,523
2,220
577
67.9%
15.2
13.4
3.5
Full-time
Part-time
2,409
3,248
42.6%
57.4
8,199
8,351
49.5%
50.5
Total
5,657
100.0%
16,550
100.0%
Male
Female
Age
Teenagers
Adults
Race/Ethnicity
White
Black
Hispanic
Other
Hours
Minimum Wage
Workforce as
Share of Total
5.7%
196
16.8%
1991 were adults, suggesting that many m i n i m u m wage
workers do have economic responsibilities. Although the
majority of m i n i m u m wage workers do w o r k part-time (less
than 35 hours weekly), there were 2.4 million full time
m i n i m u m wage workers i n 1991, comprising 42.6% of the
total. While minorities arc disproportionately represented
among m i n i m u m wage workers, more than t w o thirds of
the 5,657,000 m i n i m u m wage workers in 1991 were white.
M i n i m u m wage workers also tend to be women (64.3% of
the total)
This analysis o f the m i n i m u m wage workforce is narrow,
however, as a higher m i n i m u m wage affects workers w h o
earn more than, but close to the m i n i m u m — t h e y receive
increases w h e n the m i n i m u m wage increases. Moreover,
any legislated m i n i m u m wage increase w o u l d affect those
earning below the new rate, as well as those somewhat
above the new rate. For these reasons, Table 3.41 also presents the demographic breakdown o f those workers w h o ,
i n 1991, earned at o r below, the m i n i m u m wage level that
prevailed i n 1979 ($5.34 i n 1991 dollars). This m o r e
b r o a d l y d e f i n e d m i n i m u m wage w o r k f o r c e i n c l u d e d
16,550,000 w o r k e r s , o r 16.8% o f the total w o r k f o r c e .
Thus, any significant change in the m i n i m u m wage w o u l d
affect a substantial p o r t i o n o f the workforce. The demographic b r e a k d o w n o f the m i n i m u m wage w o r k f o r c e
under this broader definition is more inclusive o f full-time
workers and m e n but otherwise similar to the narrower
definition.
Table 3.42 examines the degree to w h i c h m i n i m u m
wage workers live i n poor families. The analysis presents
the percentage of low-wage workers living i n poor families,
under three definitions of poverty. One definition (the
t h i r d i n the Table) is the conventional definition where a
family is considered poor if its total wages do not exceed
the poverty threshold. This definition is inadequate since
many families may have "earnings problems" but have
above poverty incomes because there are multiple workers.
The other t w o definitions of "poor" families identify those
families w i t h "earnings problems" because the family head
has either a l o w hourly wage (the first section in the Table)
or low annual wages (the middle section i n the Table). The
analysis separately examines the workforce at or below the
1989 m i n i m u m wage and those workers who w o u l d be
affected by the 1990 and 1991 legislated increases.
197
Any legislated
minimum wage
increase would
affect those
earning below the
new rate, as well
as those
somewhat above
the new rate.
�TABLE 3.42
Distribution of W g Earners by Minimum W g Status
ae
ae
and Family Income, 1989
Percent of Workers in
Wage Class in Poor Families*
Men
Women
Distribution by Definition
of Poor Family
1. Family Head Earns Poverty Level Hourly Wage
Below 1989 Minimum Wage
($3.35)
Affected by 1990 Minimum
Wage ($3.36-3.80)
Affected by 1991 Minimum
Wage ($3.81-4.25)
Above 1991 Minimum
62.7%
52.2%
54.4
40.8
50.7
38.1
11.6
16.8
2. Family Head Earns Poverty Level Annual Wage
Below 1989 Minimum Wage
($3.35)
Affected by 1990 Minimum
Wage ($3.36-3.80)
Affected by 1991 Minimum
Wage ($3.81-4.25)
Above 1991 Minimum
64.4%
49.6
56.2
46.8
16.2
22.7
TABLE 3.43
Distribution o Workers Before and After
f
Mnm m W g Increases 1990-1991
ii u
ae
All
3. Total Family Annual Wages Below Poverty Level
Below 1989 Minimum Wage
($3.35)
Affected by 1990 Minimum
Wage ($3.36-3.80)
Affected by 1991 Minimum
Wage ($3.81-4.25)
Above 1991 Minimum
When the
definition of poor
is expanded to
include families
where the head
earns poverty
level wages we
see that nearly
two-thirds of i
and over half ot
female minimum
wage workers in
1989 were
poor.
54.4%
56.1
Using the conventional definition of poverty, the percentage of minimum wage workers in poor families was
46.4% among men and 45.1% among women. Among
workers earning above the 1989 but below the 1991 minimum wage there were from 30% to 40% in poor families.
When the definition of poor is expanded to include families where the head earns poverty level wages (either
hourly or annually) we see that nearly two-thirds of male
and over half of female minimum wage workers in 1989
were "poor." Also, roughly half of the workforce in 1989
that would be affected by future minimum wage increases
were poor in 1989. These data suggest that the majority of
workers affected by the minimum wage live in either poor
families or in families where the head has low earnings.
Table 3.43 shows the share of the workforce that was
aided by the minimal increases in the minimum wage in
1990 and 1991. These increases lifted 3 6% of the workforce (and 3-1% of adults and 15 9% of teenagers) above
the new minimum.
46.4%
45.1%
38.1
37.4
34.3
31.8
6.2
9.1
'Shows percentage ol workers in particular wage ranges living in poor families.
Shares of
Workers Earning*
Less than 1990
Minimum
to 1991 Minimum
to 1979 Minimum
Above
Total
Percent of workers
shifted above new
1991 minimum**
First
Qtr.
1990
Teens (16-19)
Second
to Fourth
Qtr. 1991
5.1%
2.2%
4.4
10.2
83.2
100.0
5.1
11.1
78.7
100.0
3.6%
Second
to Fourth
Qtr. 1991
First
Qtr.
1990
21.1%
6.1%
22.0
382
33.7
100.0
22.9
30.8
25.3
100.0
15.9 /o
Adults (20+)
First
Qtr.
1990
Second
lo Fourth
Qtr. 1991
4.5%
2.0%
3.8
9.2
85.0
100.0
4.4
10.3
80.9
100.0
3.1%
'Measured in inflation adjusted wages so distribution changes because of minimum wage change or
the (very unlikely) chance that low-wage worker wage gains exceeded inflation.
"Difference between share of workforce below new 1991 minimum in last nine months of 1991 versus
first three months of 1990.
198
199
�Technology
TABLE 3.44
Effect of Computer Useage on W g Structure, 1984-1989
ae
Education Group
1984
1989
1984
1989
Computer Use
Average
Wage Effect"
Change
1984
1989
1984-89
High School Drop-out
High School Grad.
Some College
College
Post-College
All
5.0%
19.3
30.6
41.6
42.8
24.6
7.8%
29.3
45.3
58.2
59.7
37.4
15.4%
17.0
166
20.3
22.0
14.5%
17.5
20.6
23.7
27.0
0.8%
3.3
5.7
8.5
9.4
Percent Using Wage Premium for
Computer on Job Computer Use*
•Wage differential received by workers using computer on the job.
"Effect of computer useage on average wage in educational group.
200
1.1%
5.1
9.3
13.8
16.1
0.4%
18
3.6
5.4
6.7
Technological change can affect the wage structure by
displacing some types of workers and by increasing
demand for other types of workers. Given the seemingly
rapid diffusion of microelectronic technologies i n recent
years, many analysts have considered technological change
as a major factor leading to the recent increase i n wage inequality. Unfortunately, because it is difficult to measure the
extent of technological change and its overall character
(whether it is generally de-skilling or up-skilling) it is difficult to identify the role of technological change on recent
wage trends. There is some information available on the
spread of one leading type of technology—computer use—
that does suggest that technological change has played a
significant role i n creating greater wage inequality.
Table 3.44 shows that there was a considerable rise i n
the use of computers at the workplace, w i t h the percentage
of the workforce using computers at w o r k g r o w i n g from
24.6% i n 1984 to 37.4% i n 1989. This increased use of
computers has affected educational wage differentials
because the most educated groups are more intensive users
of computers on the job and receive a higher wage prem i u m for doing so. For instance, in both 1984 and 1989 a
college graduate was at least twice as likely as a highschool graduate to use a computer on the job (41.6% versus 19 3% and 58.2% versus 29 3%) Although high school
graduates received about 17% extra pay for having computer skills i n either 1984 or 1989, the wage premium for
computer use among college graduates rose from 20.3% to
23.7%.
The result of differential payment for and differential use
of computers on the job has been to w i d e n pay differences
between educational groups. This can be seen i n the last
three columns of Table 3-44 w h i c h present the effect o f
computer use on the average wage of each educational
group (based on multiplying the intensity of computer use
by the wage premium) i n 1984 and 1989 and the change
between 1984 and 1989. These data suggest that the
spread of workplace computer technology widened the
educational pay gap between 1984 and 1989 by 3 6% by
increasing the pay of the average college graduate by 5.4%
but increasing the pay of the average high school graduate
by just 1.8%.
201
The result of
differential
payment for and
differential use
computers on the
job has been to
widen pay
differences
between
educational
groups.
�Executive Pay Soars
TABLE 3.45
Executive Pay Levels, 1979 and 1989
(1989 Dollars)
United Slates
1979
1989
France
1979
West Germany
1989
1979
1989
Japan
1979
1989
United Kingdom
1979
1989
Annual CEO Pay (Thousands ot Dollars)*
Pre-TaxPay
Atter-TaxPay
$308.2 $612.8 $119.4 $181.7 $136.2 $206.7 $149.3 $371.1 $121.2 $194.9
153.9
429.1
77.5
111.7
75.3
112.4
70.9
178.4
35.0
122 2
'Includes the salaries and capital distributions (such as equities and bonds) for CEOs of firms with sales
exceeding $790 million.
TABLE 3.46
Real Growth in Executive Pay, 1979-1989
Percent Change, 1979-1989
Pre-Tax
Pay
United States
France
West Germany
Japan
United Kingdom
After-Tax
Pay
19%
13
17
22
3
66%
7
15
24
123
*ln national currencies adjusted for domestic consumer price increases. Same definition as Table 3.45.
202
Another cause of greater wage inequality has been the
enormous pay increases received by top executives and the
spill-over effects.
The 1980s was a very prosperous decade for top U.S.
executives. I n stark contrast to the declining real wages of
average workers, the pay o f chief executive officers (CEOs)
o f major U.S. companies grew by 19% from 1979 to 1989
(Tables 3.45 and 3.46). Given the poor performance o f
the U.S. economy relative t o competitor nations in the
1980s, it may be surprising to f i n d that the pay of the
United States' CEOs has far outpaced that o f the CEOs in
other countries. The 19% growth in pre-tax pay for U.S.
CEOs in the 10 years from 1979 to 1989 was greater than
the g r o w t h o f CEO pre-tax pay o f 13%, 17%, and 3%
respectively, i n France, West Germany, and the United Kingd o m (Table 3 4 6 ) .
Since tax rates on high-income residents have been
severely reduced i n the U.S. and the United Kingdom over
the last 10 years, changes i n pre-tax and after-tax pay differ
widely. I n Britain, for instance, tax rates for the rich have
been reduced so much that CEO after-tax pay increased
123% f r o m 1979 t o 1989, even t h o u g h pre-tax pay
increased by only 3% ( F i g u r e 3L). I n the United States,
after-tax pay of CEOs increased by 6 6 % , more than three
times as fast as their pre tax pay increased. I n the 10 years
between 1979 and 1989, after-tax CEO pay increased far
faster i n the U.S. than in Japan, West Germany, or France.
CEOs in the U.S. are paid roughly twice as much as the
CEOs of our major industrial competitors (Table 3-47). For
instance, the average annual compensation of a CEO heading a U.S. f i r m w i t h at least »100 m i l l i o n of sales was
$508,000 in 1988. A comparable executive i n Japan makes
62% as much w h i l e a CEO in West Germany makes 5 1 % as
much.
203
The 1980s was a
very prosperous
decade for top
U.S. executives.
In stark contrast
to the declining
real wages of
average workers,
the pay of chief
executive officers
grew by 19%
from 1979 to
1989.
�FIGURE 3L
Real Growth in Executive After-Tax Pay,
1979-1989
140
c
eg
-C
In contrast, the pay of American manufacturing workers
was more equal to that of our major competitors in Canada,
France, Italy, and Japan in 1988. However, West German
workers earn 36% more than American workers. Only British workers have significantly lower wages than U.S. workers (see Chapter 9 for further analysis).
The pay gap between CEOs and manufacturing production workers is far larger in the U.S. than in other industrialized countries (Table 3-48). In 1988, U S CEOs were paid
17.5 times as much as manufacturing production workers, a
pay gap about twice as wide as in West Germany, Italy, and
France and 50% greater than in Japan.
O
c
v
o
«
0-
United States
France
West Germany
Japan
United Kingdom
TABLE 3.47
Comparative Pay Levels of Workers and CEOs, 1988
Country
Canada
France
Germany
Italy
Japan
United Kingdom
United States
CEO
lotal Compensation
Index
Dollars
(U.S. = 100)*'
$267,000
255,000
260,000
216,000
317,000
286,000
508,000
'CEOs of firms with sales exceeding $100 million.
"Using April 1988 exchange rates.
204
53
50
51
43
62
56
100
The pay gap
between CEOs
and
manufacturing
production
workers is far
larger in the U.S.
than in other
industrialized
countries.
Index, Manufacturing
Production Worker
Hourly Compensation
(U.S. = 100)**
96
99
136
97
98
77
100
TABLE 3.48
The Pay Gap Between Executives and Production Workers, 1988
Ratio of
CEO Pay to
Production
Worker Pay*
Country
9.5
8.9
6.5
7.6
11.6
12.4
17.5
Canada
France
Germany
Italy
Japan
United Kingdom
United States
•Assumes CEO works 52 40-hour weeks.
205
�Increased Black-White Wage
Inequality
TABLE 3.49
Black-White W g Differentials A o g Y u g Workers
ae
mn on
Year
Black-White
1973
1979
1989
Total
Education Level
High
High
School
School
College
Drop-outs
Graduates
Graduates
Earnings Gap*
-10.3%
-16.4%
-10.9
-17.4
-16.4
-16.5
-10.3%
-12.5
-16.7
Region"
Midwest
NorthEast
South
- 3.7%
- 2.5
-15.5
0.8% - 0.4% - 1 3 . 9 %
- 2.5
- 6.6
-15.8
-18.9
-13.8
-16.7
-1.6%
-1.4%
Annual Growth in Earnings Gap
-0.6%
-0.2%
-0.5%
-0.8% -0.2%
Cause ot Increased Racial Earnings G a p "
Region
Urban
Industry
Occup'n
Deunionization
Lower Min.
Wage
Unexplained
Total
-4%
15
10
19
5
-95%
43
51
-9
23
-12%
21
10
35
10
7%
3
7
19
2
0%
13
32
11
9
31%
7
18
-5
14
-13%
19
-26
44
24
17
98
27
3
7
4
0
100
9
100
59
100
28
100
22
100
0
100
Education-wise,
the greatest
increase in the
black-white
earnings gap was
among college
graduates, with a
minimal 2.5%
differential in
1979 exploding to
15.5% in 1989.
75
38
100
One other dimension of increased wage inequality is that
the pay of black workers fell relative to that of white workers. As the top panel (first c o l u m n ) i n Table 3.49 shows, a
black worker i n 1989 earned 16.4% less than an equivalent
( i n terms of education, experience, region and so o n ) white
worker i n 1989. This "black-white earnings gap" jumped up
f i f t y percent from 1979 to 1989 from a 10 9% level in
1979 to its 16.4% level in 1989. Education-wise, the greatest increase i n the black-white earnings gap was among college graduates, w i t h a minimal 2.5% differential in 1979
exploding to 15.5% i n 1989 ( F i g u r e 3M). Regionally, racial
inequality rose the most i n the Midwest and in the Northeast and increased only slightly i n the South (where substantial racial pay differences persisted).
Table 3.49 also presents information ( i n the bottom
panel) on what factors explain the g r o w t h i n the overall
racial wage gap (shown in the middle panel) as w e l l as the
FIGURE 3M
Black-White Wage Differentials Among Young Workers,
1973-1989
'Based on regression of log hourly earnings on race, schooling, education, region and metropolitan
status (converted to percentage terms) of workers with less than nine years experience.
"Positive (negative) numbers imply the factor led to a larger (lower) earnings gap between black and
white hourly earnings. Each causal factor can lead to a greater earnings gap if black workers become
more concentrated in lower wage parts of the wage structure (e.g., occupations, regions, industries)
or if the parts of the wage structure where blacks are concentrated see less wage growth.
"'For high-school graduates and dropouts only.
Total
206
High School Dropouts
High School Grad.
207
College Grad.
�TABLE 3.S0
Employment Opportunity and W g Growth, 1980-1989
ae
Employment/Population
Ratio
Group*
1980
1989
Change
1980-89
Real Wage
Growth
(Annual Rate)
White Males
High School Drop-out
High School Graduate
College Graduate
75%
88
94
71%
86
93
-4
-2
-1
-1.8%
-1.1
0.3
64%
84
90
57%
76
86
-7
-8
-4
-0.3%
-1.2
0.2
40%
58
71
42%
65
79
2
7
8
-0.3%
0.3
1.4
43%
64
85
41%
68
86
-2
4
1
-0.5%
-0.2
0.5
Black Males
High School Drop-out
High School Graduate
College Graduate
White Females
High School Drop-out
High School Graduate
College Graduate
Black Females
High School Drop-out
High School Graduate
College Graduate
•Workers ages 25-64.
208
gap among workers by education and region. Not surprisingly, the causes of rising racial wage inequality differ for
the various workforce subgroups. Among the workforce as
a whole, the increase i n blacks' wage disadvantage arose
from the fact that the occupations and cities in w h i c h
blacks are concentrated had lower wage growth, that blacks
became relatively more concentrated i n low-wage industries and non-union settings, and that blacks were particularly disadvantaged by the failure of the m i n i m u m wage
threshold to keep up w i t h inflation.
The large increase i n the racial gap among college graduates is largely unexplained, except for the concentration of
blacks i n occupations w i t h l o w wage-growth. The greater
racial wage gap i n the Midwest is primarily due to an
increased concentration of blacks (relative to whites) i n
non-union low-wage industries and i n occupations w i t h
below average wage-growth.
That a sizeable p o r t i o n of the g r o w t h in the racial pay
gap remains unexplained suggests that greater, or more
effective, discrimination as w e l l as weaker government
enforcement o f anti-discrimination laws may have played
an important role.
Did Lower Wages Buy More
Employment?
This chapter has found that inequality rose between the
college- and non-college-educated and the black and white
workforces. Pay differences between men and women lessened, but primarily because men's wages fell. This analysis,
however, can not tell us what happened to differences
between groups i n their total labor market experiences
without addressing the various groups' employmenl and
unemployment trends. The reason is that a group w i t h falling relative wages (a g r o w i n g wage gap between it and
other groups) may have improved its relative opportunities
for employment (or lower unemployment) independent of,
or because of, its falling relative wage.
Table 3.50 shows what happened to various groups (by
race, gender and education) between 1980 and 1989 i n
terms of wage g r o w t h and i n changes i n the employmentpopulation ratio, a measure of employment opportunity
( w h i c h falls because of higher unemployment and rises
w i t h greater labor force participation at steady unemploy209
The greater racial
wage gap in the
Midwest is
primarily due to
an increased
concentration of
blacks (relative
to whites) in
non-union
low-wage
industries and in
occupations with
below average
wage-growth.
�TABLE 3.51
The Effect of Occupation and Industry Employment
Shifts on Skill and Education Requirements, 1970-2000
Education or
Skill Requirement
1970-79
1980-88
BLS Projections
1988-2000
(Ten-Year Rates of Change*)
Skill Indices
Handling Data
Verbal Aptitude
Intellectual Aptitude
General Education
Development (GED-L)
Handling People
Handling Things
4.02%
2.31
2.28
4.34
2.19
-1.36
4.28%
2.05
2.03
3.91
1.85
2.13
1.37%
0.80
0.75
1.52
0.69
-0.37
Education
1.41%
Mean Years Required
1.09%
0.49%
Shares of Employment Requiring: (Percentage Point Change*)
Less than High School
High School Graduate
Some College
College Graduate or More
- 1.44%
-1.33
0.56
2.22
0.88%
1.88
0.63
2.00
- 0.42%
-0.67
0.33
0.83
'To facilitate comparisons of these time periods which are of different length Ihe data have been converted to ten-year rales of change: Ihe change if the annual rate of change in these time periods had
continued for ten years.
mcnt rates). In general, the groups that tared most poorly
on the wage front also fared poorly on the einployinent
front. Among both black and white men, for iiiManee, nun
college-educated workers not only tell behind eollegeeducated workers in wage growth, they also experienced a
greater employment d r o p Among w h i t e women, the
groups w i t h the fastest wage g r o w t h — h i g h - school and
college graduates—also had a greater expansion ol employ
nient. Only among black women did high school graduates
fare worse in wage g r o w t h hut do belter in terms of
employment expansion In general, that is, the groups that
experienced wage difliculties also experienced employ
mcnt difficulties Among the demographic groups in the
l9H0s, lower wages did not "buy" more employment The
groups most adversely affected by wage trends were also
adversely affected by reduced employment opporiunities.
Jobs of the Future
The jobs of the future w i l l not be markedly different than
the jobs available today. Future jobs w i l l have somewhat
greater education and skill requirements, primarily theneed for basic literacy and numeracy, but w i l l not necessarily pay more than current jobs. Moreover, the skill and education requirements of jobs are expected to grow more
slowly than in the iy7()s and 1980s. Despite the widely
held assumption that higher-paying white-collar jobs are
the wave of the future, there is little evidence that the deterioration of job quality and wages that look place in the
1980s w i l l be reversed in the 1990s, unless current trends
change dramatically One view of future jobs can be
obtained by analyzing labor market trends anticipated by
the Bureau of Labor Statistics (ULS) in its employment projections to the year 2(MM).
Key to this analysis is the examination of both the continued expansion of white-collar, high-paying ocui/Hitiuiis and
the job shift from high-paying to low paying inilustrws.
Analyses of future jobs which focus solely on occupational
change, as many do, paint too optimislic a picture
Tables 3-51 a n d 3.52 provide data w h i c h allow us to
assess the combined effect of industry shifts (eg., the rising
importance of services) and occupation shifts (e.g.. the rising importance of white-collar professional/technical jobs)
210
211
The groups that
fared most poorly
on the wage front
also fared poorly
on the
employment front.
�TABLE 3.52
The Effect of Occupation and Employment
Shifts on Pay, 1970-2000
1970-79
Pay
1980-88
BLS Projections
1988-2000
(Ten-Year Rates of Change*)
Weekly
Wages
Compensation
2.76%
2.20
1.52%
0.64
0.20%
-0.17
Hourly
Wages
Compensation
2.47%
1.96
1.49%
0.63
0.32%
-0.03
•To facilitate comparisons of these time periods which are of different length the data have been convened to ten-year rates of change: the change if the annual rate of change in these time periods had
continued for ten years.
FIGURE 3N
Growth in Skill and Education Requirements,
1970-2000
Education
Specifically, the data show the effect of changes in the dis
tribution of jobs, among occupations and industries, on the
education levels and skills (measured by six indices)
required on the job. and on hourly compensation (wages
and fringe benefits) levels in recent years, as well as the
effects anticipated by bureau of l abor Statistics employ
nient projections.
Our analysis in lahle S SI shows that job skill require
ments have been increasing over the last t w o decades and
are projected lo increase over the lyyOs. Kather lhan the
skills explosion projected by some analysts, however, future
growth in skill and education requirements w i l l be historically slow. Kor instance, the new distribution of jobs across
industries and occupations in 2000 w i l l require worker's
education levels to grow at a 0.49% (ten-year) rate over the
I98H-2000 period, a rate less than half the 1.09% rate, and
a third the 1 4 1 % rate required by actual job shifts, respectively, in.the 1980s and 1970s ( F i g u r e 3N).
The data in lable 3.52 c o n f i r m these trends by showing
that expected shifts in the industry and occupation distribution of jobs w i l l have a slight negative effect on hourly
and weekly compensation. Over the 1970s and 1980s, in
contrast, employment compositional shifts led to higher
rates of pay.
These analyses corroborate one of our conclusions
drawn from the recent declines in the pay of white- collar
and college-educated workers—the expansion of demand
for more "skilled" and more "educated" workers is slowing
down. The ultimate conclusion to be drawn, therefore, is
that wage trends and employment projections both suggest
that only a modest skill and educational upgrading of jobs
w i l l occur in the 1990s, absent some change in government policies and employer strategies
Skill
Ten-Year Rates of Change
212
213
Rather than the
skills explosion
projected by
some analysts,
however, future
growth in skill
and education
requirements will
be historically
slow.
�Chapter 4.
Conclusion
There have been dramatic changes in the wage structure
over the last dozen years. The real hourly wages of most
workers have fallen. The group experiencing the greatest
wage decline was male, non-college-educated workers.
Between 1973 and 1991, the wages paid on entry-level jobs
for high school graduates have declined by 29 3% for
young men and 20.4% for young women. Given that threeThe vast majority
fourths of the workforce has not completed college, the
ot men are recent drastic deterioration of the wages of high school
working at far graduates (whose wages fell somewhat more than high
lower wages than
school dropouts, but somewhat less than those with "some
their counterparts
college") means that the vast majority of men are working
did a generationat far lower wages than their counterparts did a generation
earlier. earlier.
There were, however, some bright spots in the wage
trends of the 1980s. College graduates obtained a modest
growth in wages; workers with advanced or professional
degrees enjoyed substantial wage growth; and, middle- and
high-wage women made significant wage advances. Unfortunately, most of the positive news of the 1980s turned negative in the later stages of the recovery (1987-1989) and/or
during the early-1990s downturn. The wages of men and
women have fallen across the board in recent years. Moreover, there are reasons to believe that wages will not
improve as the economy rebounds.
214
Jobs: Worsening
Underemployment
There has been much attention focused on the length of
the 1980s recovery—the second longest in the post war
period—and the length of the 1990s recession—the longest of the post-war period. However, it is the character
rather than the duration of these phases of the business
cycle that is important to note. As described in Chapter 3,
the incidence of low-wage work grew over the 1980s and
expanded further in the early 1990s. This chapter
describes the growth of other types of labor market
distress.
For instance, the recent recession was characterized by
an unusually large rise in white-collar unemployment coupled with historic slow growth in white-collar employment. Although groups traditionally impacted heavily by
recessions—manufacturing, construction, and blue-collar
workers—experienced an even greater growth in their
unemployment rate than white-collar workers, the rise in
white-collar unemployment signaled important market
shifts. For example, this sluggish growth of white-collar
employment signifies the end of the service sector boom of
the 1980s and helps explain the recent reductions in the
real wages of white-collar and college-educated workers.
Further, since the 1970s, there has been a growth in parttime and temporary employment—situations where workers receive low pay, are economically insecure, and receive
few benefits or opportunities for training and advancement. The increase in these and other types of "contingent" work reflects the types of jobs being created by
215
Sluggish
white-collar
employment
growth signifies
the end of the
1980s service
sector boom. . .
helps explain the
recent reductions
in the real wages
of white-collar
and collegeeducated
workers.
�employment has decidedly shifted so that a much larger
proportion of the workforce is either underemployed,
overemployed, l o w paid, or trapped i n unfavorable job
situations.
TABLE 4.1
Unemployment Rales
Year
1947
1967
1973
1979
1989
1991
Total
Male
3.9%
3.8
4.9
5.8
5.3
6.7
Female
White
3.7%
5.2
6.0
6.8
5.4
6.3
n.a.
3.4%
4.3
5.1
4.5
6.0
4.0%
3.1
4.2
5.1
5.2
7.0
Black
n.a.
n.a.
9.4%
12.3
11.4
12.4
Hispanic
n.a.
n.a.
7.5%
8.3
8.0
9.9
TABLE 4.2
Rates ol Underemployment*, 1973-1991
1973
(000)
1979
(000)
1989
(000)
1991
(000)
89,429
4,365
689
2,343
104,962
6,137
771
3,373
123,869
6,528
859
4,894
125,303
8,426
1,025
5,767
Total Underemployed*
7,397
10,281
12,281
15,218
Rate of Underemployment**
Unemployment Rate
8.2%
4.9
9.7%
5.8
9.8%
5.3
12.0%
6.7
Civilian Labor Force
Unemployed
Discouraged Workers
Involuntary Part-Time
'Unemployed, discouraged, and involuntary part-lime.
"Total underemployed workers divided by the sum of the labor force plus discouraged workers.
employers rather than any increased preference of the
Many people workforce for these types of jobs. That is, many people are
work in w o r k i n g i n part-time or temporary jobs because they cantemporary jobs not f i n d full-time or permanent jobs.
because they There has also been an expansion of l o w paying self-emcannot find ployment plus a g r o w t h of multiple jobholding primarily
permanent jobs. fueled by the fact that the wages received o n one job are
not sufficient to meet regular expenses. The character of
216
Unemployment and
Underemployment
Table 4.1 gives a broad view of unemployment rates by
gender and race d u r i n g various peak years i n the business
cycle since World War I I and for 1991, the most recent year
for w h i c h data are available. I n peak years, the economy is
at its strongest, and therefore unemployment is at its lowest. I n 1989, the most recent cyclical peak year, unemployment i n every category was less than or roughly equal to
that i n 1979, the prior cyclical peak. Relative to 1973 and
earlier peak years, however, 1989 unemployment rates are
generally above average. The recession caused unemployment to rise to 6.7% by 1991 and to 7 5% i n the second
quarter of 1992.
Unemployment among m i n o r i t y workers continues to be
roughly double that of white workers, as has been the case
throughout the post-war period. Moreover, even at the peak
of the last business cycle i n 1989, the 11.4% unemploy
ment rate among black workers was higher than the average unemployment rate reached i n any post-war recession
(see Table 4.3 below). The major shift i n the 1980s was a
lowering of the unemployment rate for women relative to
men.
A broader measure o f the lack o f employment success i n
the labor market is the rate o f underemployment. This
alternative measure includes unemployed workers but also
includes people w o r k i n g pan-time w h o want to w o r k lulltime ("involuntary" part-timers) and those w h o want to
w o r k but have been discouraged from searching for jobs
by lack o f previous success ("discouraged workers"). The
rate o f underemployment was essentially the same at the
end of the 1980s recovery i n 1989 as it was i n 1979, 9 8%
versus 9 7% respectively, despite a somewhat lower unemployment rate (Table 4.2 and Figure 4A). The 1989 underemployment rate was substantially higher than the 5 3%
unemployment rate primarily because o f a high rate o f
involuntary part-time work, a phenomena discussed below.
217
The character of
employment has
decidedly shifted
so that a much
larger proportion
of the workforce
is either
underemployed,
overemployed,
low paid, or
trapped in
unfavorable job
situations.
�FIGURE 4A
Real Rates of Underemployment,
1973-1991
BLS Unemployed
Q
Discouraged Workers
Q
Involuntary Part-Time
15
12.0
8
I
o
below. The 1989 underemployment rate, equivalent to the
1979 rate, 9-7%, was significantly higher than the underemployment rate of 8.2% in 1973. The data in Table 4.2
also show that in the recession year of 1991, underemployment has risen from 1989 not only because of higher unemployment but also because of nearly one m i l l i o n more
involuntary part-time workers.
The Early 1990s Recession
The recession of the early 1990s differed from prior postwar recessions i n several different ways (based on the
assumption that the recession ended i n the second quarter
of 1992). For example, even though the recent recession
was longer than other recessions, unemployment showed
lesser gains. As Table 4.3 shows, the recent recession lasted
at least 22 months, twice as long as the 11 -month average
length of earlier downturns ( F i g u r e 4B). Unemployment,
however, rose by 2.2%, far less than the 3% plus rise i n the
mid-1970s and mid-1980s recessions. Nevertheless, the
10
CD
a.
1979
1973
1989
The recent
recession lasted
at least 22
months, twice as
long as the
11-month
average length
of earlier
downturns.
1991
FIGURE 4B
Duration of Recessions,
1948-1992
TABLE 4.3
Changes in Unemployment in Postwar Recessions
Unemployment
Recessions:
Beginning and
Ending Quarter
Changes in Unemployment by Cause:
Rate
Peak
Trough
7.0%
5.8
7.4
6.8
5.8
8.2
7.7
10.7
All
Permanent
Unemployment Job Loss
3.2%
3.1
3.2
1.6
2.2
3.4
1.4
3.3
Other
Reasons"
Duration of
Contraction
n.a.
n.a.
n.a.
n.a.
0.9%
1.3
0.6
1.7
n.a.
n.a.
n.a.
n.a.
1.3%
2.1
0.8
1.6
11 Months
10
8
10
11
16
6
16
—
1948:4-1949:4
1953:3-1954:2
1957:3-1958:2
1960:2-1961:1
1969:4-1970:4
1973:4-1975:1
1980:1-1980:3
1981:3-1982:4
3.8%
2.7
4.2
5.2
3.6
4.8
6.3
7.4
Average of
prior recessions
4.8
7.4
2.7
—
1990:2-1992:2
5.3
7.5
2.2
1.6
0.6
11
22
'Includes unemployment due to temporary layoff or quits and of new entrants or re-entrants who have
not yet found work.
218
1948-49 1953-54 1957-58 1960-61 1969-70 1973-75
Recession
219
1980
1981-82
1990-92
�TABLE 4.4
Changes in Labor Force Levels
and Participation in Recessions, 1969-1992
Change in Labor Force
Labor Force Growth
Recession
Total
Men
Women
Total
Men
Women
(Annual Growth)
1969:4-1970:4
1973:4-1975:1
1980:1-1980:3
1981:3-1982:4
1990:2-1992:2
2.4%
2.2
1.4
1.7
0.9
'Percentage point change annualized.
220
2.2%
1.2
0.9
1.2
1.0
2.8%
3.6
2.1
2.4
1.1
0.1%
0.1
0.0
02
0.0
-0.3%
-0.7
-0.4
-0.1
-0.1
0.4%
0.7
0.2
0.6
0.1
recent recession involved roughly the same amount of permanent job loss—a rise of 1.6%—as in recent, heavier
downturns.
What's the difference between permanent job loss and
unemployment? In government statistics unemployed people are defined as those w h o are w i t h o u t a job and have
looked for w o r k i n the last four weeks. Therefore, someone
could be unemployed because of a job loss (either a permanent or a temporary layoff) or because she/he has been
unsuccessfully looking for w o r k after completing school
(an "entrant") or after a spell out of the labor force (a
"re-entrant") or after q u i t t i n g a job (a "quit"). So, the relatively large 1.6% increase i n the unemployment of those
w h o permanently (rather than temporarily) lost their job
suggests that the recent recession involved as much job loss
as prior recessions.
Overall unemployment did not rise much in the recent
recession due to slow labor force g r o w t h (fewer unsuccessful entrants and re-entrants). The g r o w t h i n both the male
and female labor force was far slower i n the early 1990s
than i n the prior four downturns (Table 4.4). D u r i n g the
recent recession the labor force grew at a rate ( 0 . 9 % )
roughly half that ( 1 . 7 % ) during the prior recession and
only forty percent of that ( 2 2 % ) d u r i n g the mid-1970s
recession. A major reason for this slow labor force g r o w t h is
a more slowly g r o w i n g adult population because of the
lower b i r t h rates i n the 1960s.
However, the labor force also grew more slowly because
the p r o p o r t i o n o f adults choosing to w o r k (the labor force
participation rate) d i d not grow d u r i n g the recent recession, primarily reflecting an historically modest annual rise
i n women's labor force participation o f 0 . 1 % . Had the
labor force g r o w n as quickly in the recent recession as in
the prior four recessions (at a 1.9% rather than an 0.9%
rate, o r by 1.0 percentage point more per year) the unemployment rate w o u l d have risen by roughly 2 percenuge
points more (1.0 percentage point for t w o years). Thus, if
it had not been for slow population growth and the stall i n
the long-term rise o f labor force participation, unemployment in the recent recession w o u l d have reached an historically high level—over 9% (7.5% plus 2%)—because o f an
historically high 4.2% rise in unemployment (2.2% plus
2.0%).
221
Had the labor
force grown as
quickly as in the
prior four
recessions the
unemployment
rate would have
risen by roughly
2 percentage
points more to
over 9%.
�TABLE 4.5
Change in Unemployment Rate b Occupation
y
Gender, Race, and Industry, 1990-1992
Group
1990:2
5.3%
All
1992:2
Change in
Unemployment
7.5%
2.2%
Occupation
Managers, Prof
Tech, Sales, Admin
Craft
Operatives & Laborers
2.1
4.0
5.3
8.4
3.2
5.7
8.8
11.2
1.1
1.7
3.5
2.8
4.7
4.7
7.2
6.3
2.5
1.6
4.6
10.6
7.6
6.5
14.5
11.2
1.9
3.9
3.6
10.5
5.5
5.0
17.0
7.9
6.9
6.5
2.4
1.9
Gender
Adult Men
Adult Women
Race/Ethnic
White
Black
Hispanic
Industry
Construction
Manufacturing
Services
TABLE 4.6
Changes in Employment and Unemployment
in Recessions b Occupation
y
Change in
Recessions:
Beginning and
Ending Quarter
1960:2-1961:1
1969:4-1970:4
1973:4-1975:1
1980:1-1980:3
1981:3-1982:4
1990:2-1992:2
Change in
Peak Share
Employment (000)
Unemployment (000)
of Employment
WhiteCollar
BlueCollar
WhiteCollar
BlueCollar
WhiteCollar
764
695
871
721
827
643
-1,142
-737
-2,004
-1,557
-2,583
-1,500
210
516
765
253
826
1,130
592
971
1,903
1,013
2,124
1,018
43.0%
47.7
48.1
51.5
52.7
57.0
222
Was the demographic and occupational composition o f
unemployment different i n the recent recession? There has
been some discussion, for instance, of this being a "whitecollar" recession. There is some t r u t h to this characterization since the white-collar unemployment rate has risen by
an above-average amount. Still, blue-collar workers have
been more deeply hurt by the recession, w i t h unemployment among blue-collar operatives and laborers rising by
2.8%, far more than the 1.7% rise in unemployment among
technical, sales, and administrative workers (Table 4.5).
Moreover, recession levels ( 1 9 9 2 : 2 ) of unemployment
among white-collar workers are lower than the unemployment rates among blue-collar workers at the cyclical peak
( 1990:2). The same comparison could be made of workers
by race/ethnicity—recession levels of w h i t e unemployment
(6.5% ) are still far below the unemployment among blacks
(10.6% ) or Hispanics (7.6% ) at the prior cyclical peak. The
absolute rise i n unemployment from 1990 to 1992 was
greater among white-collar (1,130,000) than among bluecollar workers (1,018,000), w h i c h is unusual given that
blue-collar unemployment rose far more than white-collar
unemployment i n prior recessions (Table 4.6). The greater
rise i n w h i t e collar unemployment reflects the increased
dominance o f white-collar employment ( 5 7 % of total i n
1990), the above-average rise i n the white-collar unemployment rate, and the fact that unemployment rose nearly as
much i n services as i n manufacturing (Table 4.5).
The most unique aspect of the recession, however, was i n
the modest g r o w t h of white-collar employment. Over eight
quarters, white-collar employment grew by only 643,000,
far less than the 827,000 g r o w t h over just five quarters i n
the prior recession (Table 4.6). The slow g r o w t h i n whitecollar jobs signals the end o f the service-sector boom of the
1980s and is likely responsible for the recent fall off i n
wages among white-collar and college-educated workers
(see Chapter 3).
BlueCollar
36.9%
36.0
35.1
32.4
31.3
26.7
223
The rise in
unemployment in
the recent
recession was
greater among
white-collar than
among bluecollar workers
. . . blue-collar
unemployment
rose far more
than white-collar
unemployment in
prior recessions.
�Being unemployed was a bigger financial hardship in the
1980s and 1990s than in the 1970s. As Table 4.7 shows, in
1990 and 1991, only about 40% of the unemployed
received unemployment insurance, a substantial drop
from the financial protection available in the 1975 and
1976 recession years when 76% and 67%, respectively, of
the unemployed received unemployment insurance.
Unemployment insurance also provided less assistance in
the 1980s downturn. Changes in the laws and administration of unemployment insurance at both the federal and
state levels are primarily responsible for the lower coverage of unemployment insurance.
TABLE 4.7
Proportion of the Unemployed W o Receive Some
h
Unemployment Insurance Payment
Percent of
Unemployed
Receiving
Benefits
Year
1960
1967
1973
1975
1976
1979
1982
1983
1989
1990
1991
Unemployment
Rate
54%
43
41
76
67
42
45
44
33
37
42
5.5%
4.9
5.8
8.5
7.7
5.8
9.7
9.6
5.3
5.5
6.7
Job Creation: How Fast in the 1980s?
TABLE 4.8
Employment Growth in Expansions, 1949-1990
Expansion
Trough
Length
(Mo.'s)
Peak
Private Sector
Employment Growth
Total
Goods
Services
(Annual Rate of Growth)
Oct. 1949
May 1954
Apr. 1958
Feb. 1961
Nov. 1970
Mar. 1975
July 1980
Nov. 1982
July 1953
Aug. 1957
Apr. 1960
Dec. 1969
Nov. 1973
Jan. 1980
July 1981
July 1990
224
4.3%
2.8
3.6
3.5
3.6
4.0
2.8
3.0
5.9%
2.9
3.7
2.9
3.6
3.4
2.6
1.0
2.8%
2.8
3.5
4.0
3.6
4.3
2.9
3.8
52
Average
45
39
24
106
36
58
12
92
3.5
3.3
The fact that unemployment dropped after 1982 led to
the impression that the 1980s were a time of exceptional
job creation. Now that the 1980s recovery has ended, it is
possible to directly compare it to other recoveries. As it
turns out, the notion that the 1980s recovery was a time of
rapid job creation is factually incorrect.
Table 4.8 provides the growth rates of private sector
employment in each of the eight post-war recoveries. The
3% annual employment growth over the mid- and
late-1980s exceeded two previous recoveries and was less
than in five other recoveries. The recovery from 1982 to
1990 was the second longest recovery of the post-war
period, lasting 92 months. The mid-1980s recovery is sometimes referred to as the longest post-war "peace-time"
recovery, implicitly labeling the 1960s recovery as a "wartime" recovery. While the Vietnam buildup did occur during this period, it is also true that there was a major military buildup in the 1980s (and even larger in terms of
economic output) so the distinction between "war-time"
and "peace-time" recoveries seems manufactured and
misleading.
3.5
225
In 1990 and
1991, only about
40% of the
unemployed
received
unemployment
insurance, a
substantial drop
from 1975 and
1976.
�TABLE 4.9
Employment Growth, 1947-1991
Measures of Employment
Hours
Period
Full-time
Working
Labor Force
of
Work
Civilian
Employment
Equivalent
Employment
Age
Population
Participation
Rate
(Percentage
Point Annual
Growth)
(Annual Rates of Growth)
1947-67
1967-73
1973-79
1979-89
1989-91
1.3%
2.2
2.5
1.7
-0.2
1.7%
1.6
1.8
1.6
n.a.
1.8%
1.9
2.3
1.8
n.a.
1.2%
2.1
1.9
1.2
0.9
0.07%
0.20
0.48
0.28
-0.25
FIGURE 4C
Civilian Employment Growth,
1947-1991
co
as
Z3
C
c
<
Table 4.9 looks at the employment growth over the last
two business cycles (from peak to peak) relative to the earlier post-war period to see whether the long-term growth
in employment has been foster in recent years. Three measures of employment growth are presented: civilian
employment, total hours worked, and full-time equivalent
employment (which combines part-time and full-time
according to practices within each industry). The trend is
unmistakable: each measure of employment grew far more
slowly from 1979 to 1989 than between 1973 and 1979.
For instance, the rate at which civilian employment rose
from 1979 to 1989, 1.7% annually, was more than 30%
slower than the 2.5% annual employment creation rate
prevailing between 1973 and 1979 (Figure 4C).
The last two columns of Table 4.9 show two reasons why
unemployment has fallen despite a much slower growth in
j o b creation since 1979- First, the working-age population
has grown more slowly than it did in the period from 1967
to 1979. For instance, there were fewer new potential
young workers in 1989 than in 1979, even though the
overall population grew by over 12% (1.2% annually for
ten years). In 1989, there were 32.2 million 16- to 24-yearolds, a smaller number than the 37.0 million in 1979 or
even the 33 5 million in 1973.
Second, the long-term rise in the proportion of the working-age population seeking work (the labor force participation rate) has also leveled off in recent years. As Table 4.9
shows, the annual percentage point increase in the labor
force participation rate in the ten years between 1979 and
1989 was only sixty percent as much as in the six years
between 1973 and 1979. This primarily reflects a slower
increase in women's labor force participation since 1979
These two trends, the smaller number of potential young
workers and the slower labor force participation growth,
have meant a slower growth in the number of people seeking work. Thus, the lower unemployment of the later part
of the 1980s was not due to superior job creation. Despite
the slower growth rate of new jobs, unemployment was
reduced because of the historically slow growth of new
workers seeking employment.
-0.5
1947-67
1967-73
226
1973-79
1979-89
1989-91
227
The lower
unemployment of
the later part of
the 1980s was
not due to
superior job
creation hut to
the historically
slow growth of
new workers
seeking
employment.
�The Growth in Part-Timers
TABLE 4.10
Composition of Non-Agricultural Employment, 1973-1991
Percent Part-Time
Year
Total
1973
1979
1989
1991
16.6%
17.6
18.1
18.9
Involuntary
3.1%
3.8
4.3
5.4
Voluntary
Percent
Full-time
13.5%
13.8
13.8
13.5
83.4%
82.4
81.9
81.1
Total
100.0%
100.0
100.0
100.0
FIGURE 4D
Part-Time Employment Rates,
1973-1991
1973
1979
228
1989
As we have seen earlier, there has been a g r o w t h i n
underemployment due to more people w o r k i n g part-time
who want full-time jobs. This section examines the g r o w t h
of part-time work and the associated problems in more
detail.
The expansion of part-time w o r k is not necessarily a
problem. Many workers prefer a part-time schedule because
it allows time to pursue education, leisure, or family
responsibilities. But large numbers of part-timers w o u l d
prefer to w o r k full-time. Part-timers generally have lower
pay, lower-skilled jobs, poor chances of promotion, less job
security, inferior benefits (such as vacation, health insurance, and pension), and lower status overall w i t h i n their
places of employment.
Table 4.10 shows that the share of total employment
made up by full-timers has declined steadily from 1973 to
1989. At the same time, the share of jobs w h i c h are parttime increased from 16.6% i n 1973 to 18.1% i n 1989, at
the end of the business cycle ( F i g u r e 4D). This increase i n
the proportion of workers w o r k i n g part-time from 1973 to
1989 has been almost entirely due to the increased rate at
w h i c h workers are w o r k i n g part-time i n v o l u n t a r i l y . It thus
reflects the increased use of part-timers by employers and
not the preferences of the workforce for shorter hours.
In 1989, involuntary part-time workers comprised 4 3%
of the workforce, an increase of 0.5 and 1.2 percentage
points of the workforce over, respectively, 1979 and 1973Neariy five m i l l i o n workers (Table 4.2), on average, were
involuntary part-timers i n 1989, encompassing nearly onefourth of all part-time workers.
1991
229
This increase in
the proportion of
workers working
part-time from
1973 to 1989 has
been almost
entirely due to
the increased
rate at which
workers are
working part-time
involuntarily.
�TABLE 4.11
A e and Gender Composition of the Labor Force
g
and Rate of Part-time Employment, 1969,1979, and 1989
1969
Gender & Ages
All, 16-21
Women, 22-44
Women, 45-64
Men, 22-64
All, 65 +
Total
Share of
Employment
12.8%
17.3
13.2
53.2
3.5
100.0
1979
Percent
part-time
40.6%
22.7
22.5
3.7
41.1
Share of
Employment
14.0%
23.1
11.3
48.9
2.7
1989
Percent
part-time
41.7%
22.5
24.4
4.8
52.9
Share of
Percent
Employment part-time
10.36%
27.7
11.6
47.8
2.6
46.3%
21.9
23.8
6.7
52.4
15.5
100.0
17.6
100.0
18.1
Total, holding
wilhin-group
rates at 1969
levels
15.5
—
16.4
—
15.9
Total, holding
within-group
rates at 1979
levels
—
—
17.6
—
17.0
230
Teenagers, persons of retirement age, and adult women
are the groups most likely to work part-time (Table 4.11).
In 1989, 46.3% of employed 16- to 21-year olds and 52.4%
of workers 65 or older worked in part-time jobs. More than
one in five employed adult women was a part-time worker
in 1989.
These demographic groups—women, teenagers, and
older workers—are the ones most willing to accept lower
pay and status in order to obtain a reduced work schedule.
However, a demographic explanation for the expansion
of part-time jobs—the growth of women and younger
workers in the labor force raises the need for part-time
work—is not consistent with the available data. As shown
in Table 4.10, the rise of part-time work since 1973 is
entirely due to the increased incidence of involuntary parttime work. Moreover, for adult men, the group with the
least preference for part-time work, the rate of part time
work rose from 3 7% in 1969 to 6.7% in 1989.
In fact, the increase in part-time work over the last
twenty years can be fully explained by a rise in part-time
work within each demographic group rather than by any
change in the age-gender composition of employment.
Table 4.11 presents an analysis which allows us to assess
the impact of demographic shifts on the rise in the overall
rate of part-time employment between 1969 and 1979 and
between 1979 and 1989. If the rates of part-time work
within each group had remained at their 1969 levels while
the mix of employment, such as more women in the workforce, changed as it did from 1969 to 1979, then the overall rate of part-time work would have risen from 15.5% in
1969 to 16.4% in 1979. Thus, demographic shifts can
account for just 0.9 percenuge point (16.4% less 15.5%) of
the overall 2.1 percentage point rise (17.6% less 15.5%) in
the rate of part-time employment between 1969 and 1979
Surprisingly, demographic shifts between 1979 and 1989
were a force for less part-time work. Had demographic
shifts been the only factor affecting the overall rate of parttime work then 0.6% less of the workforce (17.0% less
17.6%) would have been part-time workers. Yet, the rate of
part-time work actually rose from 17.6% to 18.1%, indicating that non-demographic factors led to a 1.1 percentage
point rise in the rate of part-time work (the difference
between the demographically neutral 17.0% and the actual
18.1%).
231
Nearly I'm million
workers were
involuntary
part-timers in
1989,
encompassing
nearly one-fourth
of all part-time
workers.
�TABLE 4.12
Industry Composition of the Labor Force
and Rate of Part-time Employment, 1969,1979, and 1989
1969
Industry
Construction
Durable Mfg.
Nondurable Mfg.
Trans., Comm.
Ulil.
Trade
Finance
Services
Public Admin.
Mining
Share of
Employmenl
6.0%
18.2
12.3
1979
Percent
part-lime
8.6%
3.2
7.8
Share of
Empbyment
1989
Percent
part-time
Share ot
Employment
Percent
part-time
6.0%
15.3
9.9
10.5%
3.8
8.6
5.9%
12.0
8.3
10.5%
3.9
8.1
7.3
18.5
5.1
25.3
6.5
0.7
7.8
26.3
10.5
26.2
6.2
5.0
7.0
20.7
6.2
28.1
5.9
1.0
9.0
30.0
11.9
25.0
6.6
4.0
7.3
21.4
7.1
32.1
53
0.6
8.7
29.7
11.5
24.0
5.8
4.8
All industries
100.0
15.1
100.0
17.1
100.0
17.6
All industries,
holding withinindustry rates
at 1969 levels
—
15.1
16.2
—
29 7% of the retail/wholesale trade workforce and 24% of
the service industry workforce were employed part-time
(Table 4.12). As we saw in Chapter 3, roughly threefourths of the net new jobs created between 1979 and
1989 were i n these t w o industries. This shift of employment towards industries w h i c h provide low-wage, part-time
employment can explain 2.1 of the 2.5 percentage point
rise (or 84% of the increase) i n the rate of part-time
employment among non-agricultural wage and salary workers between 1969 and 1989. Shifts i n the industry composition of employment has led to an additional 1% of the
workforce i n part-time w o r k i n each of the last t w o
decades.
The rate o f voluntary part-time work in each industry
sector generally rose between 1969 and 1979, but fell
between 1979 and 1989 (Table 4.13). Involuntary parttime work, o n the other hand, has generally risen over the
entire post-1969 period, especially i n the service and trade
industries where the largest numbers of involuntary parttime workers are employed. This suggests that the rise i n
part-time w o r k i n particular sectors reflects employer strategies (using a low-wage, high-turnover workforce) rather
than workforce preferences.
The major reason
for growing
part-time wort is
that job growth
has been
concentrated in
the service
industries which
intensively use
part-timers.
17.2
TABLE 4.13
Rates of Voluntary and Involuntary
Part-time Work by Industry, 1969,1979, and 1989
All industries,
holding within
industry rates
at 1979 levels
—
17.1
—
18.1
Involuntary
1979
Voluntary
Industry
Over the entire 1969 to 1989 period, demographic shifts
can account for only 0.4 percenuge points (15.9% less
15.5%) of the 2.6 percenuge point rise (18.1% less 15 5%)
in the rate of part-time employment. This is because the
overall rate of part-time work is rising because teens, older
workers, and men are more likely to work part-time. Adult
women, on the other hand, were more likely to work fulltime in 1989 than in 1979.
The major reason for growing part-time work is that job
growth has been concentrated in the service industries
which intensively use part-timers. In 1989, for instance,
232
1969
Construction
Durable Mfg.
Nondurable Mfg.
Trans., Comm.
Util.
Trade
Finance
Services
Public Admin.
Mining
All Industries
4.4%
1.4
3.4
5.4%
1.5
3.8
6.1%
1.6
3.4
1.8
2.9
1.0
3.1
0.8
1.6
2.5
2.6
5.4
1.7
4.1
1.3
2.3
3.6
2.7
6.3
1.8
4.7
0.9
2.5
4.1
1989
1969
4.2%
1.9
4.4
6.0
23.4
9.5
23.1
5.4
3.4
12.6
233
1979
5.0%
2.3
4.8
6.4
24.5
10.2
21.0
5.3
1.7
13.5
1989
4.4°/
2.3
4.7
6.0
23.4
9.6
19.3
4.9
2.3
13.5
�TABLE 4.15
Differences in Fringe Benefits,
for Full-time and Part-time Workers,
b Gender or Marital Status and Occupation, 1987
y
TABLE 4.14
W g Differences Between Part-time and
ae
Full-time Workers b Gender and Selected Occupations, 1989
y
Percent of
Part-time
Workers
Average Hourly
Wage, 1989
(1989 Dollars)
Parttime
Fulltime
Wage Difference
Dollars
Percent
Women
All Workers
Prof., Tech.
Sales
Admin.
Service
100.0%
16.9
18.4
24.7
29.3
$7.02
11.81
5.49
7.10
5.12
$9.20
12.42
8.19
8.50
5.88
-$2.18
-0.61
-2.70
-1.40
-0.76
-23.7%
-4.9
-33.0
-16.5
-12.9
100.0%
12.6
13.0
10.4
25.6
15.4
$7.24 $12.28
13.43 16.19
6.26 12.60
6.33 10.91
4.88
8.43
5.84
8.13
$5.04
-2.76
-6.34
-4.58
-3.55
-2.29
-41.0%
-17.0
-50.3
-42.0
-42.1
-28.2
Men
All Workers
Prof., Tech.
Sales
Admin.
Service
Laborers
A major problem associated w i t h part-time employment
A ma jor problem is lower hourly pay received by part-time workers comassociated with pared to equivalent full-time workers (Table 4.14). Women
part-time w h o work part-time earn 23.7% less wages per hour than
employmentare those w o r k i n g full-time. This wage differential is not simply
lower hourly due to the concentration of part-timers i n low-wage occuwages and pations since part-timers are paid significantly less in each
benefits received occupation. The part-time wage differential is even greater
among men than among women.
by part-time
The pay difference between part-time and full-time
workers workers is even greater for benefits, however (Table 4.15).
compared to For instance, only 25.6% of part-time female family heads
equivalent are included i n their employers' health plans compared to
full-time workers. 74.1 % of full-time female family heads. Similarly, men worki n g i n part-time jobs are only about half as likely ( 3 7 . 1 %
versus 7 9 % ) to be included i n their employers' health
plans.
234
Gender or Marital
Status and
Occupation
Included in Health Plan
Included in Pension Plan
Part-time
Full-time
Part-time
Full-time
25.6%
74.1%
14.2%
50.2%
41.3
20.5
36.8
17.2
85.9
60.3
82.1
46.4
23.0
11.8
20.6
8.1
62.4
31.9
55.9
28.3
22.8
69.1
18.8
45.2
17.8
59.7
17.9
50.9
23.5
16.4
18.4
12.1
68.0
44.7
63.1
37.2
26.2
12.0
18.6
13.4
64.0
29.7
55.0
27.6
20.8
58.8
18.3
41.7
37.1
79.0
18.3
58.5
43.6
38.7
44.1
27.8
86.1
75.3
84.8
71.8
24.5
18.5
17.6
12.9
67.6
44.2
70.0
55.6
36.5
75.2
17.7
54.5
Female Family Heads
All Workers
Professional,
Managerial, and
Technical
Sales
Clerical
Service
Craft, Operative
and Labor
Wives
All Workers
Professional,
Managerial, and
Technical
Sales
Clerical
Service
Craft, Operative
and Labor
Husbands and Male Family Heads
All Workers
Professional,
Managerial, and
Technical
Sales
Clerical
Service
Craft, Operative,
and Labor
235
�FIGURE 4E
Partially or Fully Paid Medical Benefits Offered
to Employees, by Full- or Part-time Status, 1991
Q
Share Cos)
|
Paid by Employer
15
Full-lime
30+ Hrs Per Week
20-29 Hrs Per Week Below 20 Hrs Per Week
Even when part-time workers are covered by their
employer's health insurance plan, they are much more
likely to have to pay towards it—or less likely to receive
insurance coverage paid by their employer.
As Figure 4E shows, the fewer the hours of work, the
less likely an employee w i l l have employer provided health
insurance (99% of full-timers, but 58% of those working
20-29 hours per week) and the less likely the provided
insurance will be paid for by the employer. More information on uninsured workers is provided in Chapter 9.
Pay differentials are even greater when it comes to pensions. In each occupation group, workers with part-time
jobs are less likely than full-time workers to be enrolled in
their employers' pension plans (Table 4.15).
The gap between part-time and full-time workers in the
provision of health and pension benefits is almost as wide
for women who are heads of households as it is for women
in married couple families (Table 4.15). It is also just as
wide among professional women as among women in clerical or service occupations.
More Than One Job
The growth of part-time work reflects growing underemployment and deteriorating pay and opportunity in the
economy. The growth of multiple jobholding—people
working in at least two jobs—reflects the overemployment
due to the deterioration of real wages since 1979, which
persisted even during the recovery years from 1985
through 1989 This increased incidence of multiple
jobholding between 1979 and 1989 due to economic hardship reasons (adding one percentage point) was responsible for roughly 1.17 million additional multiple jobholders
(if the multiple jobholder rate was 5.2% rather than 6.2%
in 1989, there would have been 6.06 million rather than
7.23 million multiple jobholders).
236
237
The growth of
multiple
jobholding
reflects the
overemployment
due to the
deterioration of
real wages since
1979.
�TABLE 4.16
Growth of Multiple Jobholding, All Workers, 1973-1991
Percent of Workforce
Year
Number of
Multiple
Jobholders
(000)
1973
1979
1985
1989
1991
4,262
4,724
5,730
7,225
7,183
Multiple
Jobholding
Rate
5.1%
4.9
5.4
6.2
6.2
Who Hold Multiple
Jobs Because of:
Economic
Hardship*
n.a.
1.8%
2.2
2.8
2.5
Other
Reasons**
n.a.
3.1%
3.2
3.4
3.7
Change
1973-79
1979-85
1985-89
1989-91
462
1,006
1,495
-42
-0.2
0.5
0.8
0.0
n.a.
0.4
0.6
-0.3
n.a.
0.1
0.2
0.3
In 1989, 7.2 m i l l i o n workers held at least t w o jobs
(Table 4.16). The rate of multiple jobholding grew from
4.9% in 1979 t o 6.2% i n 1989, an increase of 1.3 percentage points ( F i g u r e 4F). It is especially noteworthy that
most of the g r o w t h of multiple jobholding occurred in the
recovery years from 1985 t o 1989, w h e n an additional 1.5
million workers began w o r k i n g more than one job. The
multiple jobholding rate held steady at 6.2% even as total
employment declined in the 1989-1991 d o w n t u r n .
An analysis of the reasons for multiple jobholding shows
that increased multiple jobholding over the 1979 to 1989
period reflects deteriorating economic performance, rather
than enhanced o p p o r t u n i t y . The g r o w t h o f multiple
jobholding has occurred primarily among workers w h o
w o r k at more than one job because of "economic hardship,"
the need to meet regular expenses or to pay o f f debts (Table
4.16). Multiple j o b h o l d i n g due to economic hardship
increased by 1.0 percentage points from 1979 to 1989,
accounting for more than three-fourths of the 1.3 percentage point rise in multiple jobholding. This same pattern
holds even over the recovery years from 1985 to 1989,
•To meet regular household expenses or pay off debts.
"Includes saving for the future, getting experience, helping a friend or relative, buying something
special, enjoying the work, and so on.
FIGURE 4F
Multiple Jobholding Rates,
1979-1991
1979
238
Increased
multiple
jobholding over
the 1979 to 1989
period reflects
deteriorating
economic
performance,
rather than
enhanced
opportunity.
1985
1989
239
1991
�TABLE 4.17
Growth of Multiple Jobholding B Gender, 1973-1991
y
Percent of Workforce
Year
Number of
Multiple
Jobholders
(000)
Multiple
Jobholding
Rate
Who Hold Multiple
Jobs Because of:
Economic
Hardship*
Other
Reasons**
n.a.
1.5%
2.0
2.9
2.2
n.a.
2.0%
2.7
3.0
3.7
Women
1973
1979
1985
1989
1991
Change
1973-79
1979-85
1985-89
1989-91
869
1,407
2,192
3,109
3,129
2.7%
3.5
4.7
5.9
5.9
538
785
917
20
0.8
1.2
1.2
0.0
n.a.
0.5
0.9
-0.7
n.a.
0.7
0.3
0.7
3,393
3,317
3,537
4,115
4,054
6.6
5.9
5.9
6.4
6.4
n.a.
2.1
2.3
2.6
2.4
n.a.
3.8
3.6
3.8
4.0
-76
220
78
-61
-0.7
0.0
0.5
0.0
n.a.
0.2
0.3
-0.2
n.a.
-0.2
0.2
0.2
Men
1973
1979
1985
1989
1991
w h e n increased economic hardship accounted for 0.6 percentage points o f the 0.8 percentage point rise in the multiple jobholding rate. I n contrast, the multiple jobholding
rate for "other reasons" rose only 0 3 percentage points
from 3 . 1 % to 3 4% between 1979 to 1989 (with a rise o f
0.2 percenuge points between 1985 and 1989) and was
responsible for about 350,000 more multiple jobholders
(relative to a situation where the "other reasons" rate had
held at 3 2%).
Many of those holding multiple jobs for "economic hardship" reasons may have lost their extra job (or jobs) i n the
1989 to 1991 d o w n t u r n , as t h e i r rate o f m u l t i p l e
jobholding fell from 2.8% to 2.5%. It is curious that there
was a g r o w t h i n n o n - e c o n o m i c h a r d s h i p m u l t i p l e
jobholding during a time of rising unemployment
There was a rapid increase i n the multiple jobholding
rate among women between 1979 and 1989, rising from
3 5% to 5.9% (Table 4.17). Economic hardship explains
roughly 60% of this increase ( 1.4 percentage points of the
2.4 percentage point increase). Between 1985 and 1989,
75% of the rise in women's multiple jobholding is due to
economic hardship. By 1989, a greater p r o p o r t i o n of
women ( 2 . 9 % ) than men ( 2 . 6 % ) worked more than one
job in order to meet regular expenses or to pay off debts.
The decline i n multiple jobholding for "economic reasons"
in the 1989 to 1991 d o w n t u r n occurred primarily among
women.
Change
1973-79
1979-85
1985-89
1989-91
'To meet regular household expenses or pay off debts.
"Includes saving for the future, getting experience, helping a friend or relative, buying something
special, enjoying the work, and so on.
240
241
By 1989, a
greater
proportion of
women (2.9%)
than men (2.6%)
worked more than
one job in order to
meetregular
expenses or to
pay off debts.
�TABLE 4.18
Distribution of Multiple Jobholders Experiencing
Economic Hardship, 1979-1989
Gender/Marital Status
Distribution of Multiple
Jobholders Who Have
Economic Hardship Reasons
1979
1985
1989
Total, 16 Years and Over
100.0%
100.0%
33.7
40.6
14.9
25.7
10.8
14.9
48.2
19.8
28.5
13.1
15.5
59.3
42.8
51.8
36.2
15.6
11.4
4.2
Women
Married, Spouse Present
Other
Single
11.2
22.5
n.a.
n.a.
Widowed, Divorced, or Separated
66.4
54.9
11.6
n.a.
n.a.
Men
Married, Spouse Present
Other
Single
Widowed, Divorced, or Separated
100.0%
16.5
10.2
6.4
•Workers who report they have multiple jobs in order to meet regular household expenses or pay off
debts.
TABLE 4.19
Hours Worked By Multiple Jobholders, B Gender, 1989
y
Gender
Percent of Multiple Jobholders Working:
Average
Weekly
Hours
0-40
Hours
41-48
Hours
49-69
Hours
70 +
Hours
Total
Total
52.0
17.6%
24.5%
43.9%
14.0%
100.0%
Male
Female
55.8
47.1
9.3
28.5
23.9
25.4
48.4
38.0
18.5
8.2
100.0
100.0
242
Table 4.18 provides a demographic breakdown of those
w o r k i n g more than one job for economic hardship reasons.
As suggested by Table 4.17, the share of w o m e n among
those w h o worked more than one job because of economic
hardship has rapidly increased so that by 1989 nearly half
of these workers were women. I n 1989, more than half of
multiple jobholders were married (wives, 19 8 % ; husbands,
36.2% ) and an additional f i f t h were widowed, divorced or
separated men or women. These data thus show that the
hardship associated w i t h and r e f l e c t e d by m u l t i p l e
jobholding is concentrated in w o r k i n g families and among
adult workers.
The majority of those w o r k i n g multiple jobs i n 1989
worked at least 49 hours per week (Table 4 . 1 9 ) Two-thirds
of men w o r k i n g more than one job worked at least 49 hours
per week w h i l e about 46% of women multiple jobholders
did the same. These data c o n f i r m that multiple jobholding
does not simply reflect people combining t w o part-time
jobs into the equivalent of one full-time job. Rather, multiple jobholders worked an average of 52 hours per week in
1989.
The Contingent Workforce
A significant p o r t i o n of people i n today's workforce have
become "contingent workers." Workers can be hired on a
temporary or "contingent" basis i n a variety of ways. Some
firms put workers directly on their payroll but assign them
to an internal temporary worker pool. Employers also use
temporary help agencies to obtain workers on a temporary
basis, sometimes for long periods. Some businesses hire
independent contractors to p e r f o r m work that w o u l d
otherwise be done by employees. A l l three types of contingent workers are frequently denied health insurance and
pension coverage and have little access to promotions and
better jobs.
Data drawn from surveys of major companies suggest
that the use of contingent labor is widespread and rising
(Table 4.20).
I n 1986, 36% of the surveyed f i r m s had an internal temporary worker pool. A later survey (not of the same firms)
shows that 49% of firms use their o w n temporary worker
pool. The h i r i n g of workers through temporary help agen-
243
Surveys of major
companies
suggest that
the use of
contingent labor
is widespread
and rising.
�TABLE 4.20
The Use of Various Types of Contingent Labor
Percent of Surveyed Firms That
Use Contingent Labor, By Type of Labor
Internal
Temporary
Worker Pool
Survey
1986*
1989"
Temporary
Agencies Help
77%
97
36%
49
Independent
Contracting
63%
78
'Survey of 477 major companies.
"Survey of 521 major companies.
TABLE 4.21
The Growth in Personnel Services Industry Employment, 1973-1991
cies is now nearly universal among large companies. In the
1989 survey, 97% of the firms used temporary help agencies. Independent contracting is also widespread, with 78%
of the firms using this method in 1989. In the earlier survey, 63% of the firms hired independent contractors
Unfortunately, these surveys do not have information on
how many contingent workers these firms actually use.
The only subgroup of contingent workers that is captured
by government statistics is the workforce employed
through temporary help agencies. Unfortunately, these
data are not available for the years before 1982. For a
longer perspective, however, it is possible to examine the
growth of the entire personnel services industry, which
consists primarily of workers hired through, or working
for, temporary agencies (three-founhs of total) but also
includes people working in employment agencies (Table
4.21 and Figure 4G). There has been an explosive growth
in personnel services employment from 1979 to 1989, but
especially between 1982 and 1989. Industry employment
rose by 824,000 from 1979 to 1989, with 796,000 jobs
The only
subgroup of
contingent
workers that is
captured by
government
statistics is the
workforce
employed
through
temporary help
agencies.
Personnel Industry Employment*
Number (OOP)
Year
1973
1979
1982
1989
1991
Total
256
527
555
1,351
1,522
Women
133
309
354
782
890
As Share of
Total Employment
Total
0.3%
0.6
0.6
1.2
1.4
Women
0.2%
0.3
0.4
0.7
0.8
'This industry consists of temporary help agencies (75% of total employment) and employment
agencies (25% of total employment).
244
FIGURE 4G
Temporary and Personnel Services Industry Employment,
1973-1991
�TABLE 4.22
Growth in Temporary Help Industry Employment, 1982-1991
Year
Temporary Help Industry Employment*
As Share of
Number (OOP)
Total Employment
Total
Women
Total
Women
1982*
1989
1991
401
1,032
1,268
256
618
730
0.4%
0.9
0.3%
0.7
0.7
1.2
'Earlier data not available.
TABLE 4.23
Elements of The Marginal Workforce*
Number of Workers
(000)
Temporary Help Services**
Leased Workers**
At-home Workers***
8 hours/week or more
35 hours/week or more
"Data are from different sources and reflect Ihe situation in the 1985-87 period.
"1987
•"1985
246
944
120
2.243
1,992
1,067
added since 1982. As a result, the share o f the workforce
employed i n the personnel services industry d o u b l e d ,
from 0.6% in 1982 to 1.2% i n 1989. The industry share o f
total employment in 1989, 1.2%, was four times greater
than its share i n 1973, 0 3%. The increase in the proportion o f the workforce in the personnel services industry
from 1973 to 1989 (from 0 3% to 1.2%) meant an additional 1 m i l l i o n workers in the industry. Even as total
employment fell between 1989 and 1991, the industry was
able to add roughly 171,000 jobs.
Women are more likely to be employed as temporary
workers, partly because such w o r k provides flexible hours
but also because there is a higher incidence of temporary
w o r k i n occupations dominated by women, such as clerical
work. Moreover, "temping" is sometimes the only option
available. In 1989, 782,000, or 58%, of the 1,351,000 workers i n the personnel services industry were women.
The available information o n employment in the temporary help industry since 1982 is presented in Table 4.22.
These data confirm that the growth in the overall personnel services industry is primarily due to more workers
employed through temporary help agencies. An additional
0.5% o f the workforce, or 631,000 more workers, were
employed through temporary help agencies in 1989 compared to 1982. Most of the increase was among w o m e n .
Temporary employment rose by more than 230,000 even
d u r i n g the recession years from 1989 to 1991.
Other types of marginal workers (Table 4 . 2 3 ) include
"leased employees" w h o w o r k for one f i r m but arc leased to
another, and workers employed at home (so employers can
reduce their overhead and avoid labor regulations such as
m i n i m u m wage, overtime, and safety protections). These
workers, along w i t h those employed by " t e m p " agencies,
are the statistically visible portion of this new marginal
workforce; many others—such as new immigrants—escape detection.
247
Women are more
likely lobe
employed as
temporary
workers, partly
because such
work provides
flexible hours but
also because
there is a higher
incidence of
temporary work in
occupations
dominated by
women.
�Self-Employment
TABLE 4.24
The Growth of Self-Employment
Self-Employment as Share
of Total Employment* Among
Year
All
Workers
1948
1967
1973
1979
1989
1991
12.1%
7.3
6.7
7.1
7.5
7.8
Men
n.a.
8.8%
8.2
8.8
9.0
9.4
Women
n.a.
4.9%
4.4
4.9
5.8
6.0
'Nonagricultural industries.
TABLE 4.25
Self-Employment and Paid Employment Earnings
Median
Annualized Earnings
of Full-time
Year-Round Workers
($1983)
Men
Self-Employed
Paid Employee
Self-Employed
Relative to Paid
Employee Earnings
248
Women
$13,520
20,039
$3,767
12,079
67%
31%
A significant portion of total employment consists of selfemployed workers, those whose primary job is working in
their own business, farm, craft, or profession. Individual
independent contractors, discussed above, would be considered self-employed.
In 1989, self-employment represented 7.5% of total
employment, up from 7.1% in 1979 (Table 4.24). The
greatest growth in self-employment has been among
women workers, who earn far less in self-employment than
in regular paid employment. An additional 1.4% of women
workers were considered self-employed in 1989 compared
to 1973. The rate of self-employment among men is higher
than among women but has risen more slowly.
Much of self-employment is disguised underemployment,
as can be seen from the fact that self-employed workers
earn far less than those on regular payrolls. Self-employed
women earn only 31% as much as women wage and salary
workers, the growth of self-employment means lower
earnings for millions of women (Table 4.25). Selfemployed men earn one-third less than men who are paid
as wage and salary workers. So, the growth of male selfemployment means depressed earnings for millions of men.
The rise in self-employment between 1989 and 1991
probably suggests rising unemployment rather than rising
entrepreneurialism
Conclusion
This chapter has reviewed recent trends in employment,
underemployment, overemployment and unemployment.
Even in a year of relatively low unemployment such as
1989, an average of nearly ten percent of the labor force
was either unemployed, working part-time but wanting
full-time work, or too discouraged to look for work. Millions more were holding at least two jobs to make ends
meet, making do with much lower wages in some selfemployed venture, or stuck in a temporary job. It is thus
important to examine the character of employment, including wages and benefits, as well as the growth of unemployment and underemployment when analyzing employment
trends.
249
Even in a year of
relatively low
unemployment
such as 1989, an
average of nearly
ten percent of the
labor force was
either unemployed,
working part-time
but wanting
full-time work, or
too discouraged to
look for work.
�Chapter 5-
Wealth: Losses For Most,
Gains For Few
Stagnant incomes and falling wages are only part of the
decline in well being of working Americans. A family's
wealth also affects its ability to cope with financial emergencies as well as its standard of living. For example, financial assets, such as money in a bank account or stocks or
bonds, can help a family make ends meet during periods of
disability or unemployment. Tangible assets, such as a home
or a car, can directly affect a family's quality of life and the
ease with which it meets its needs for housing and
transportation.
The distribution of wealth is even more concentrated at
the top than the distribution of income. In 1983, for
instance the richest 1% of families controlled 34% of
household net worth, but received 12.8% of household
income. During the 1983 to 1989 recovery the distribution
of wealth became even more concentrated In fact, the
period 1983-1989 witnessed an accrual of accumulated
household wealth almost exclusively among the very rich
est families, while wealth among the bottom 60% of families actually fell. There has also been a growth in the differ
ences in wealth held by white and non white families The
consequence of increased wealth inequality is a larger gap
in overall financial security between the rich, on the onehand, and the middle class and the poor, on the other
The rate of wealth creation in the l9H()s was comparable
to that of the 1970s, a growth rate of just under 2", annually. In the I98()s, however, there was only growth in the
251
The period
1983-1989
witnessed an
accrual ot
accumulated
household wealth
almost
exclusively
among the very
richest families,
while wealth
among the
bottom 60% ot
families actually
fell.
�types of assets owned primarily by the wealthy—liii.iiRi.il
assets. On the other hand, there was no growth in t.ingihle
assets—homes and consumer durables — the type of
wealth most likely to be held by middle-class families.
TABLE 5.1
Growth ot Household Wealth, 1949-1991
Annual Growth of Household*
Net Worth Per Adult
19491967
3.6%
4.2
3.3
Type of Wealth
Net Worth
Net Tangible Assets**
Net Financial Assets
19671973
1.3%
4.3
-0.1
19731979
1.9%
4.3
0.5
19791989
1.8%
0.0
2.9
19891991
-1.4%
-5.3
0.5
'Includes all households, personal trusts, and nonprofit organizations.
"Consumer durables, housing and land assets less home mortgages.
FIGURE 5A
Growth of Household Wealth Per Adult,
1949-1989
Total Net Worth
252
Net Financial Assets
Net Tangible Assets
Aggregate Household Wealth:
Financial Assets Boomed, Tangibles
Failed to Grow
We have seen in Chapter 1 that income growth in recent
years was slow by historical standards, increasingly
unequal, and characterized by more work al lower wages.
Has there been a similar trend regarding wealth crc.ition
and distribution? In this section we examine the change in
the overall growth of wealth while trends in wealth inequality are examined in the next section.
A basic measure of aggregate wealth is boHSvboUl iicl
UKirth—the total assets of all households minus their debts.
Table 5.1 traces the growth of household net worth over
the post-war period between cyclical peak years. Household net worth per adult increased 1.8% per year between
1979 and 1989 slightly less than the 1.9% rate of
1973 1979 and well under the 3 6% annual growth rate of
1949-1967. Thus, the 1980s growth in household wealth as
well as in income was slow by historical standards.
Not all household assets increased in value at the same
rate. Financial assets boomed, but tangible assets grew
much more slowly (Table 5.1 and Figure 5A). Since the
wealthy own most of the financial assets in the country,
these developments have benefitted them in particular.
Growth of net financial assets per adult was 2.9% per year
in the 1979-1989 period, close to the 3 3% growth rate of
1949-1967. The financial assets boom of the 1980s contrasts starkly with the modest 0.5% annual growth of
1973-1979 and the slight decline between 1967 and 1973.
On the other hand, the value per adult of tangible assets,
such as housing and land, automobiles, appliances, and so
on, did not grow at all in the 1979 1989 period This lack of
tangible asset growth from 1979 to 1989 is especially
impressive because of strong tangible asset growth
thniughiiut the 1950s, 1960s, and 1970s. Since tangihle
assets are spread out more evenly than financi.il assets.
253
The financial
assets boom of
the 1980s
contrasts starkly
with the modest
0.5% annual
growth of
1973-1979 . . .
On the other
hand, the value
per adult of
tangible assets
did not grow at
all in the
1979-1989
period.
�Net Worth
Net
Financial
Assets
Distribution of
Family Income
their stagnation affected mainly the bulk of the population
who are not wealthy. The changing composition of wealth
growth in Table 5.1 thus previews the growing wealth inequality shown in the next section.
Between 1989 and 1991, wealth creation suffered, as net
financial assets grew at a modest 0.5% annual rate and the
value of tangible assets fell rapidly.
100.0%
100.0%
100.0%
Wealth Inequality Exceeds Income Gap
38.3%
32.9
28.8
50.3%
35.4
14.4
14.7%
24.2
61.1
TABLE 5.2
Distribution ol Wealth, 1989
Wealth Class
All
Top 1 %
Next 9%
Bottom 90%
FIGURE 5B
Distribution of Income and Wealth,
1989
a
>
CL
I
o
£
to
5
Family Income
Net Financial Assets
Net Worth
Bonom 90%
254
Q
Next 9%
|
| Top 1%
The distribution of wealth is less equal than the distribution of income. The concentration of wealth among very
high-income families is dramatic. Table 5.2 and Figure 5B
reveal that in 1989 (the most recent year for which such
data are available), the top 1% of families earned 14.7% of
total income, yet owned 38.3% of total net worth and a
remarkable 50.3% of net financial assets. Net worth is the
value of one's assets minus one's debts. For a typical family
this might be the value of its house, car, other consumer
goods, and bank accounts, less the amount owed on its
mortgages and credit cards. Net financial assets are f i n a n cial assets minus debts. For the same family, this would be
the bank account balance, minus mortgage and credit card
debts. The value of pension plans are not included in this
analysis.
There is no information on the holding of wealth among
the bottom 90% of families in 1989. Earlier analyses for
1983, however, showed that at the bottom of the income
distribution, 20% of families had a zero or negative net
worth and fully 54% of families had zero or negative net
financial assets. Though most of the latter are not poor,
their lack of financial assets indicates that more than half of
American families are living from paycheck to paycheck
with little or nothing in the bank in case of a loss of job or
other serious financial emergency.
The concentration of financial assets at the top implies
that American businesses are owned and financed primarily by the richest families. In 1989, for example, the
wealthiest top 0.5% owned 37.4% of all corporate stocks,
255
In 1989 the top
1% ol Iamilies
earned 14.7% ot
total income, yet
owned 38.3% of
total net woi
and a remarkable
50.3% of net
financial assets.
�regular expenses. Non-wealthy families, however, also
"owe" 62.6% o f the debt, primarily mortgages o n their
homes.
TABLE 5.3
Selected Holdings of Assets b
y
Family Wealth Level, 1989
Growing Wealth Inequality
(percent o total assets held b each group)
f
y
Asset Type
Super Rich
(Top 0.5%)
Very Rich
(Next 0.5%)
Rich
(90-99%)
Rest
(0-90%)
37.1%
21.1
37.5
10.0
37.4
16.1%
6.0
9.9
18.8
18.8
100.0%
100.0
100.0
100.0
100.0
63.6%
58.7
42.2
62.6
100.0%
100.0
100.0
100.0
Total
A. Assets Held Primarily by the Wealthy
Stocks
Bonds
Trusts
Business
Non-Home Real
Estate
37.4%
64.0
33.3
56.2
33.8
9.3%
8.9
19.3
8.9
10.0
B. Assets and Liabilities Held Primarily by the Non-Wealthy
Principal Home
Life Insurance
Liquid Assets*
Total Debt
5.2%
7.7
12.3
11.8
3.3%
5.6
6.8
2.7
27.9%
28.0
38.7
22.9
As we shall see, there was a dramatic g r o w t h i n wealth
inequality over the late 1980s. It is important to note, however, that the long-term trend prior to the 1980s was
towards a lesser concentration o f wealth. Table 5.4 shows
the share of total wealth ( b o t h excluding and including
retirement wealth) over the 1922-1981 period for years i n
which data are available. I n general, wealth was more concentrated i n the 1920s and 1930s than i n any period since
World War I I . The concentration o f wealth (excluding
retirement wealth) held by the upper 1% remained fairly
steady over the 1940s, 1950s, and 1960s, ranging from a
low of 25.7% i n 1949 to a high of 3 1 9 % i n 1965 w i t h no
discernable trend up or d o w n . The data for 1976 and 1981,
TABLE 5.4
Share of Total Household Wealth Held
b Richest O e Percent of Individuals, 1922-1981
y
n
'Includes demand deposits, savings and time deposits, money market funds, certificates of deposit, and
IRA and Keogh accounts.
Year
while the bottom 90% owned only 16.1% (Table 5 . 3 ) The
The share of top 0.5% also owned 64% o f bonds and 56.2% o f private
wealth of the business assets, w h i l e the bottom 90% owned only 6.0% of
upper 0.5% grew bonds and 18.8% o f business assets. Overall, Table 5 3
by one shows that the types of wealth w h i c h generate income,
percentage pointsuch as bonds, businesses, stocks, and other financial assets,
between 1962 tend to be held almost exclusively by the richest 10%, i f not
and 1983, but the top 1 % , of families.
The types
m 90% of families
grew four times are primarily of wealth held by the b o t t o Non-wealthy famihomes and life insurance.
as much betweenlies also o w n liquid assets (primarily cash i n checking, sav1983 and 1989. ings and money market accounts), w h i c h are used to meet
256
The types of
wealth which
generate income,
such as bonds,
businesses,
stocks, and other
financial assets,
tend to be
held almost
exclusively by the
richest 10%, if
not the top 1%,
of families.
1922
1929
1933
1939
1945
1949
1953
1958
1962
1965
1969
1972
1976
1981
Excluding
Retirement
Wealth
Including
Retirement
Wealth
37.9%
36.7
28.2
33.4
22.4
20.5
21.6
20.7
22.5
23.4
21.0
20.5
13.8
n.a.
38.3%
37.2
28.9
38.1
28.9
25.7
28.1
27.0
30.1
31.9
29.0
28.6
18.9
23.6
257
�TABLE 5.5
Composition o Aggregate Household Wealth,
f
1962,1983, and 1989
Percent of Gross Assets Represented by:
Year
Gross
House
Value
Other
Real
Estate
1962
1983
1989
25.6% 6.2%
29.8
14.7
32.6
14.4
Net
Home
Equity
Business
Equity
Deposits Bonds
Stocks Trusts
Total
Total
Assets Debt
15.2%
18.6
17.9
19.5% 4.6%
8.9
2.6
7.4
2.4
100% 13.8% 18.5%
100
13.0 22.3
100
15.9 23.6
19.3%
18.7
14.7
7.6%
4.1
6.1
TABLE 5.6
Changes in Distribution o Net Worth
f
and Family Income, 1962-1989
Percent of Family Income
Percent of Net Worth
Change
1983-89
1983
1989
Change
1983-89
81.7% 81.5% 84.3%
25.2
26.2
30.3
8.2
7.8
8.0
22.1
21.6
21.6
12.4
12.1
11.3
14.3
13.3
13.1
2.8%
4.1
0.2
-0.5
-0.8
-0.2
51.9%
9.1
3.7
13.3
8.3
17.5
54.3%
11.5
3.2
13.8
10.4
15.4
2.4%
2.4
-0.5
0.5
2.1
-2.1
Four-Fifths
Fourth
Middle
Second
Lowest
18.3% 18.5% 15.7%
12.9
12.5
13.0
5.2
5.2
2.7
0.8
1.1
0.2
-0.5
-0.3
-0.2
-2.8%
0.5
-2.5
-0.9
0.1
48.1% 45.7%
21.7
22.2
14.1
12.6
8.7
7.8
3.7
3.0
-2.4%
0.5
-1.5
-0.9
-0.7
Total
100.0
Wealth
Class
1962
Top Fifth
Richest 0.5%
Next 0.5%
Next 4%
Next 5%
Next 10%
1983
1989
Bottom
100.0
258
100.0
0.0
100.0
100.0
0.0
however, suggest that wealth became less concentrated
during the 1970s. The backdrop for the recent surge i n
wealth inequality is a long-term d r o p i n wealth concentration since the early 1920s including a fall o f f i n wealth concentration i n the 1970s.
Table 5.5 shows the changing composition of wealth
from 1962 to 1983 and 1989, d r a w n f r o m a consistent set
of surveys. At first glance, it appears that housing represents a much larger percentage o f household wealth, having
risen from 25.6% o f gross assets i n 1962 to 32.6% i n 1989.
However, household mortgage debt also rose and, as the
last column shows, the net value of people's homes (asset
value less debt) rose a smaller amount.
I n general, a lesser share o f wealth is held i n financial
assets such as deposits, bonds, stocks, and trusts w h i c h i n
1962 comprised 5 1 % of all assets but fell to just 30.6% of
all assets i n 1989. Correspondingly, real estate and business
ownership has increased i n importance, g r o w i n g from
21.4% of gross assets i n 1962 to 32 3% i n 1989. Since real
estate and business ownerships as w e l l as financial assets
are disproportionately held by the very rich, the shift
towards the former does not necessarily lead to greater or
lesser inequality.
Fortunately, the 1962, 1983, and 1989 surveys allow a
direct examination o f changes i n the distribution of wealth.
Between 1962 and 1983 there were only minor changes i n
the distribution of wealth (Table 5.6). For instance, the
percentage of total wealth held by each f i f t h of families
remained comparable between 1962 and 1983, w i t h the
upper f i f t h having a slightly lower share—81.5% rather
than 8 1 . 7 % — i n 1983 than i n 1962. Although among the
upper f i f t h , there was a modest redistribution towards the
upper 5% and upper 0.5%.
Between 1983 and 1989, however, there was a major
upward redistribution o f wealth ( F i g u r e 5C). In 1989, the
richest 0.5% of families owned 30 3% of household net
w o r t h , up 4 . 1 % from the 26.2% share i n 1983 The share of
wealth o f the upper 0.5% grew by one percentage point
(from 25.2% to 26.2%) over the entire twenty-one-year
period between 1962 and 1983, but grew four times as
much i n just six years between 1983 and 1989. The only
other group to increase its share of wealth was the next
richest 0.5% of families (and the fourth-fifth). The remainder o f the population had a lesser share of wealth i n 1989
259
The backdrop for
the recent surge
in wealth
inequality is a
long-term drop
in wealth
concentration
since the early
1920s including
a fall off
in wealth
concentration in
the 1970s.
�FIGURE 5C
Distribution of Wealth,
1983-1969
40
•
|
1983
1989
30
I
20
I
o
i;
130
5
133
10
C
O
0
-tO I
1
1
1
^
1
^
1
J
L
^
+#'
L
oS**
TABLE 5.7
Sources of Wealth Growth, 1962-1989
Percentage of Wealth Growth Due to
Years
Initial
Wealth-holding
Appreciating
1962-83
1983-89
75%
70
260
Savings
25%
30
Annual
Growth
of Wealth
1.3%
2.4
^
than in 1983, including the groups other than the upper
1% among the wealthiest f i f t h of families.
Most startling is the erosion of the wealth shares of the
bottom sixty percent of families. The poorest f i f t h of families had more debt than assets i n both 1983 and 1989,
although slightly less so. The second f i f t h saw its share of
wealth almost disappear, shrinking from a negligible 1.1%
in 1983 to just 0.2% in 1989. The middle f i f t h saw its share
of wealth essentially halved, falling from 5.2% in 1983 to
2.7% in 1989.
Aside from the modest gains of the fourth fifth, the character of the wealth redistribution of the late 1980s was a
much larger share of wealth for the upper 1 % and a lower
share of wealth for everybody else. The redistribution to
the top was in fact spectacular—the upper 0.5% gained an
additional 4 . 1 % of household net w o r t h , an amount greater
than that held by the entire bottom sixty percent in 1989
We have several clues as to why this occurred. As lable
5.6 also shows, there was a redistribution of family income
d u r i n g this same period. Given that, as the incomes of the
well-off grew rapidly, it could be expected that their accumulation of wealth w o u l d also grow. In 1989, the upper
f i f t h had an additional 2.4% of total family income relative
to 1983- The bottom sixty percent of families, had a lower
share of income i n 1989 than i n 1983, paralleling the
trends in wealth.
The g r o w t h of income shares at the top, however, was
less than what occurred for wealth—the upper 0.5%
gained 2.4% of income but 4 . 1 % of wealth. In addition, the
g r o w t h of income shares was more widespread than that of
w e a l t h — t h e upper 10% gained income shares but only the
upper 1% increased their share of wealth
That income and wealth concentration should generally
move together is not a surprise. Neither should it be surprising that there were differences in the extent of the
growth of the concentration in income and wealth. As
Table 5.7 shows, most wealth g r o w t h arose from the appreciation (or capital gains) of pre-existing wealth and not
from diligent savings out of income. Over the 1962 to 1989
period roughly three-fourths of new wealth was generated
by increasing the value of initial wealth (much of it inherited). The somewhat more important role of savings in the
1983 to 1989 period reflects the unusually fast income
growth of the very rich in that time period.
261
The redistribution
to the top was in
fact
spectacular—
the upper 0.5%
gained an
additional 4.1%
of household net
worth, an amount
greater than that
held by the entire
bottom sixty
percent in 1989.
�TABLE 5.8
Change in Net W rh b Wealth Class, 1983-1989
ot y
(1989 Dollars)
Net Worth (000)
Wealth
Class
Percent of Total
Real Wealth Growth
Accruing To:
1983
1989
Change
1983-89
$596
7,491
2,457
1,445
519
211
$713
10,248
2,706
1,856
552
222
19.6%
36.8
10.1
28.5
6.4
5.2
94.2%
46.9
6.6
19.1
13.1
8.5
Four-Fifths
Fourth
Middle
Second
Bottom
$33.8
$33.2
-1.8%
5.8%
91.4
38.0
8.0
-2.2
109.8
22.8
1.7
-1.7
20.1
-40.0
-78.8
n.a.
14.8
-6.1
-3.0
0.2
Total
146.3
169.1
15.6
100.0
$4,974
1,647
982
$6,477
2,026
1,204
The importance of the growth of pre-existing wealth in
generating wealth accumulation suggests that the financial
assets boom of the 1980s was responsible for the surge in
wealth inequality (see Table 5.1). Since the richest 1% of
families owns over half and the bottom 90% owns only
14.4% of the net financial assets (Table 5.2) rapid gains in
financial assets such as stocks and bonds will generate
wealth for the very rich and rising wealth inequality.
Table 5.8 and Figure 5D show another dimension of
growing wealth inequality by presenting the pattern of percentage changes in wealth for the various wealth classes
over the 1983 to 1989 period. On average, the bottom 60%
of families not only had less wealth in 1989 than in 1983,
but also experienced substantial losses in wealth. The middle fifth of families saw a 40% drop in average wealth from
«38,000 per family to just 822,800 per family. The second
fifth of families (wealthier than the poor but less wealthy
than the middle) saw the modest wealth holdings of i8,000
in 1983 nearly wiped out by 1989, falling to just «1,700.
Top Fifth
Richest 0.5%
Next 0.5%
Next 4%
Next 5%
Next 10%
Bottom
Memo:
Top 1 %
Top 5%
Top 10%
30.2%
23.0
22.6
FIGURE 5D
Change in Net Worth By Wealth Class,
1983-1989
53.5%
72.6
85.7
at
c
to
c
<
D
u
<
u
Q.
y*
262
Since the richest
1% of families
owns over half
and the bottom
90% owns only
14.4% of the net
financial assets,
rapid gains in
financial assets
such as stocks
and bonds will
generate wealth
for the very rich
and rising wealth
inequality.
^
x**
Ntf"
263
�FIGURE 5E
Distribution of Net Worth Growth By Wealth Class,
1983-1989
<3
a
o
Percentage Share ot Wealth Growth by Wealth Class
TABLE 5.9
Change in Wealth by Income Class, 1984-1988
(1988 Dollars)
Median Net Worth
Income
Fifth
1984
1988
Change
1984-88
Top
Fourth
Middle
Second
Lowest
$98,411
49,947
29,459
21,248
5,130
$111,770
46,253
28,044
19,694
4,324
13.6%
-7.4
-4.8
-7.3
-15.7
37,012
35,752
It was only the fourth-fifth and groups in the upper 5%
that experienced a faster than average growth in wealth.
The richest 0.5% of families experienced an increase in
wealth by more than a third (36.8%) in the six years after
1983, achieving an annual growth in wealth of 5.4%. Having started with nearly J7.5 million per family in 1983, the
richest families were able to achieve an average net worth
of J10.2 million by 1989This pattern of wealth growth yielded a highly skewed
distribution of the share of real wealth growth accruing to
the various wealth classes, as shown in the last column of
Table 5.8 and in Figure 5E. As we have seen before, the
bottom 60% of families did not participate in the wealth
growth of the 1983 to 1989 period—the wealth of this
group was falling or negative. On the other hand, fully
94.2% of the wealth created from 1983 to 1989 accrued to
the wealthiest twenty percent of families The richest 1%
of families alone obtained 53.5% of the wealth created in
this period.
The data in Table 5.9 are taken from a different set of
surveys and confirm the pattern of unequal wealth growth
over the late 1980s shown in previous tables. Ranking families by income (rather than wealth), one sees that only the
highest income families saw their wealth increase in the
1984 to 1988 period. The bottom eighty percent of families had less wealth in 1988 than in 1984. The startling
truth is that during the 1980s recovery, the vast majority of
families lost wealth as the economy grew, unemployment
fell, and family income rose.
Total
264
-3.4
265
The bottom 60%
ol Iamilies did
not participate in
the wealth growth
of the 1983 to
1989 period—the
wealth of this
group was falling
or negative.
�TABLE 5.11
Ratio of Household Income and Wealth
Between Non-White and White Families, 1940-1989
TABLE 5.10
Household Debt Burden, 1949-1991
Percent Debt as of:
Personal
Total
Income
Assets*
Year
1949
1967
1973
1979
1989
1991
30.3%
61.7
61.5
66.6
82.4
86.7
6.9%
12.5
13.6
14.1
16.3
17.2
Ratio of Non-White to White.
Year
A.
Means
Household Income
63%
62
63
58%
57
59
13%
24
19
1967
1983
1989
B.
Medians
4%
9
7
Net Worth
'Financial assets (including pension funds and insurance), real estate, and consumer durables.
Soaring Debt
The gap in
average wealth Household debt has skyrocketed in recent years. Two
between minoritymeasures o f the total debt burdens o f households, debt as
and white a percent o f assets and as a percent o f personal income,
households is have each g r o w n markedly in the 1980s after a long period
even larger than o f relative stability (Table 5.10). Household debt leapt
the income gap from 66.6% of personal income in 1979 t o 82.4% in 1989,
with the average and from 14.1% o f household assets in 1979 to 16.3% in
wealth in 1989.
These higher levels of debt have left families more vulnon-white
nerable and the economy weaker. The recession of the early
families being
1990s only worsened the situation, d r i v i n g debt levels that
just 19% of the much higher.
average wealth
in white families
in 1989. Increased Racial Wealth Gaps
1962
1983
1989
C.
Homeownership Rates
a.
1940
1950
1960
1970
1980
1985
b.
1983
1989
Census of Population
52%
61
60
64
65
64
Federal Reserve Board Survey Data
62%
63
We already know that blacks and other minorities tend to
have lower incomes than whites. Yet, non-whites lag even
further behind in the accumulation of wealth. As Table
5.11 shows, the average income of a non-white household
was 63% of that of an average white household in 1963, a
ratit) that prevailed in both 1983 and 19H9 (looking at
median incomes shows the same failure to close the racial
income gap).
266
267
�TABLE 5.12
Net Worth b Race and Household Characteristics, 1989
y
(1989 Dollars)
Average Net Worth ($000)
Household
Group
Whites
NonWhites
AH
$210.9
$39.1
19%
$37.0
70.9
95.7
188.7
664.1
$5.6
11.4
42.8
48.3
236.6
15%
16
45
26
36
A. Income Class
Less than $10,000
$10,000—$19,999
20,000-29,000
30,000-49,999
50,000 and more
B.
Age Class
$117.5
159.0
211.5
425.6
331.9
176.1
$8.2
53.4
75.8
37.3
69.5
20.5
7%
34
36
9
21
12
$276.5
114.9
147.3
Under 35 years
35-44
45-54
55-64
65-74
75 and over
C.
Ratio
$72.7
22.8
18.2
The gap in average wealth between minorily anil whitehouseholds is even larger than the income gap, with the
average wealth in non-while families being just 19% ol the
average wealth in white families in 1989. There was, however, significant progress in lessening racial wealth inequality in the 1962 to 1983 pcruxl as non-white mean wealth
grew from 13% to 24% of while mean wealth In contrast,
the racial wealth gap widened between 1983 and 1989
The gap in racial wealth holding is pervasive. At every
level of income a minority family holds far less wealth than
a white family (Table 5.12) l ikewise, a racial wealth gap is
evident at every age level, but especially among the very
oldest and youngest families Racial wealth differences are
also present in each type of family, whether in married couple families or lh»)se headed by single females.
Conclusion
Trends in the growth and distribution of wealth are comparable to those for income. Both the after tax income and
the wealth of the bottom 60% of families declined over the
1980s and Ihe greatest share of both increased wealth and
income accrued to the upper 1% of families Thus, by 1989
the bulk of the families not only held a lower share of total
wealth, they had less wealth to fall back on in case of a fiscal or medical emergency or to finance large expenditures
such as a college education.
26%
20
12
Household Type
Married Couples
Single, Male Head
Single, Female Head
268
269
By 1989 the bulk
of the families
not only held a
lower share of
total wealth, they
had less wealth
to fall back on
in case of
emergency or to
f inance large
expenditures
such as a college
education.
�Chapter 6_
Poverty: High Rates of
Poverty Unresponsive to
Economic Expansion
The previous chapters have documented the winners
and losers in the highly uneven economic expansion ol the
1980s, l itis chapter locuses on the experience ol the poor
est members of society: those persons and families whose
economic situations are the most precarious.
The poor did not fare well in the l9H0s, despite the economic growth economy from 1983 to 1989 The ecoiioinic
policy strategy of the period—a rising tide will lift all
boats—proved ineffective As one analyst noted, "Appar
ently, the rising tide lifted the yachts; the rowhoats
foundered."
This chapter examines in detail a nuinhcr of the arguments proffered to explain the unresponsive poverty rates
of the decade and finds that the facts do nol support them.
Specifically, some critics have claimed thai poverty was
mis measured, and that the economic condition ot the poor
did improve over the decade However, our analysis of both
the government's official measurement and a series of more
conceptually satisfying ways to measure poverty revealed
that the general finding of high levels of poverty over the
1980s is valid no matter which form of meusurenieni is
used
Others argue that the poor themselves are lo blame,
either because the) lornied families more likely to he poor
regardless of the expansion (i.e., single-headed families), or
because they failed to exploit the labor market opportunilies that prevailed over the decade, 'lo explore these issues,
we analyze the perceived problem of female headed family
271
The rising tide
lifted the ySChtS;
the rOWbOdtS
fOUndeKd.
�FIGURE 6A
Predicted vs Actual Poverty Rates
1959-1991
24
22
20
0)
18
ro
DC
>.
16
-
s
o
0-
AUUJI Povevly Rale
formation, examine the question of whether welfare programs have created incentives that have increased poverty,
and explore the issue of whether the poor chose not to
work. Yet, the data do nol support these explanations for
the high and intractable poverty rales throughout the economic expansion of the 19H0s.
In fact, the problems analyzed in previous chapters—the
heightened inequality of the income distribution, lessened
progressivity of the tax system, and in particular, falling
wages—all conspired to keep poverty rates historically
high throughout the decade. In addition, the "safety net"
(the social provision of assistance to those in poverty) grew
less effective at providing relief to the poor.
14
Background: The Failure of
Economic Growth
12
10
v
Predicted Poverly Rale
I
I
I
I
I
I
I
I
I
272
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
'
I
The failed relationship between poverty and economic
growth in the 1980s is portrayed graphically in Figure 6A
The dotted line in the figure represents a statistical model
of changes in the poverty rate over time, based on economic variables historically correlated with poverty rates:
unemployment, inflation, income, and government trans
fers. The bold line is actual poverty rates.
What is particularly notable about Figure 6A is the diver
gence of predicted and actual values in the 1980s The
model tracks poverty extremely well up until 1983 (the
beginning of the expansion) when the model incorrectly
predicts a fall in poverty rates. In other words, a major shift
in the relationship between economic growth and poverty
occurred in the decade such that the recent economic
recovery failed to reduce poverty as much as occurred in
earlier periods of growth. This chapter begins by exploring
the extent of poverty in the I98()s, and then examines the
reasons for the failure of economic growth to ameliorate
poverty.
273
A major shift in
the relationship
between
economic growth
and poverty
occurred in the
decade such that
the recent
economic
recovery failed to
reduce poverty as
much as occurred
in earlier periods
of growth.
�Measuring the Extent of Poverty
TABLE 6.1
Percent and N m e o Persons in Poverty, 1959-1991,
u br f
with Averages Over Peak Years
Year
Actual
Poverty
Rate
Predicted
Poverty
Rate
Number in
Poverty
(000)
1959
1967
1973
1975
1979
1983
1989
1991
22.4%
14.2
11.1
12.3
11.7
15.2
12.8
14.2
22.1%*
14.0
11.0
12.8
11.4
15.2
9.3
n.a.
39,490
27,769
22,973
25,877
26,072
35,303
31,528
35,708
Averages Over Peak Years
1959-65
1967-73
1974-79
1980-89
19.8%
12.5
11.7
13.6
"From 1960 due to data availability.
274
19.4%
12.3
11.8
12.5
There are, of course, a variety of ways to measure pov
erty. The most commonly accepted measures are the official government poverty lines. These are pretax, post transfer income thresholds, adjusted for family size and for price
changes. That is, the Census Bureau categorizes as poor
those families with cash incomes below the poverty line
for that sized family. While there exist a number of conceptual problems with this definition of poverty (eg, it
ignores the value of non-cash benefits such as food and
medical care), the official poverty measures have a long history and are a widely used measure of economic despair.
Other, more conceptually satisfying measurements, are preiscnted below. However, regardless of the choice of measure,
the poverty trends they expose are similar.
Table 6.1 shows overall poverty rates at peaks in the
business cycle since 1959 and for 1991, the last year of data
availability. Poverty was exceptionally high in the late
1950s, but began falling in the 1960s in response to economic growth and more generous government transfers,
dropping from 22.4% in 1959 to 14.2% in 1967. Poverty
rates continued to fall during the early 1970s, hitting a low
of 11.1% in 1973, before climbing again. On average, poverty rates were higher throughout the 1980s expansion at
13.6% over 1980-1989, than over the growth periods from
1967 onward, when they ranged from 12.5% ( 1967-1973)
to 11.7% (1974-1979).
Table 6.1 also includes predicted poverty rates, generated
by the model used for Figure 6A. lip to 1983, the model
tracks poverty rates quite closely. For the years shown in
Table 6.1, the predicted poverty rates never diverge more
than 0.5 percentage points from the actual rates until I9K9.
These findings tell us that, based on the way economic
growth has effected poverty in the past, the rates were
expected to fall by 5.9 percentage points between 1983
and 1989, from 15.2% to 9 3%. However, the actual rates
only fell 2.4 points over this period, from 15.2% to 12.8%.
The causes of this divergence between the predicted and
actual poverty rates, unique to the growth period of the
1980s, will be examined throughout the chapter.
275
On average,
poverty rales
were higher
throughout the
1980s expansion
at 13.6% over
1980 1989, than
over the growth
periods from
1967 onward,
when they ranged
from 12.5% to
11.7%.
�Alternative
Measurements
of Poverty
In t h e analysis ot povc-rly, MR- issue arises as to w h e t h e r
TABLE 6.2
Consumption-Based Poverty Measures
p o v e r t y is b e i n g a p p r o p r i a t e ! ) i l e l ' i n e i l a m i m c . i M i i v i l
The
n e x t l e w tables e x a m i n e s o m e o l t h e w a y s t h e o l l u i.il Census lUireau's p o v e r t y m e a s u r e b o t h overstates a m i u m l e r
Poverty Rate
Consumption
Standard
1977
1988-
Ofticial Measure
Housing Consumption
Food Consumption
11.6%
20.7
18.0
Point
Change
1977-1988
13.0%
230
25.8
states p o v e r t y . O t h e r measures are p r e s e n t e d , a m i i . i t h o n e
stresses a d i l t i r e n l c o n c e p t o l w h a t c o n s t i t u t e s
poverty.
H o w e v e r , n o m a i l e r w h i c h c o n c e p t o f p o v e r t y is used, l l u
t r e n d d o c u m e n l c d i n l i g u r e 6A a n d lable d I is e v i d e n t .
1.4
2.3
7.8
A c o m m o n c r i t i c i s m o f t h e o f f i c i a l p o v e r t y lines is lli.it
they n o l o n g e r r e f l e c t e v e n m i n i m a l levels ol c o i i s i m i p i i i n i
and thus understate the extent of p o v e r t y The original con
sumption
data
were
collected
in
I95S. w h i n
n
was
assumed that p o o r f a m i l i e s spent o n e - t h i r d o l t h e i r i n c o m e
'Latest year ol data availabilty.
on food (thus assuming thai families could punltasc
all
o t h e r necessities f o r t w i c e w hat the)- spent o n l o o d )
p o v e r t y l i n e s w e r e t h e n c o n s t r u c t e d by m u l t i p l y i n g
TABLE 6.3
Poverty Rates when (Nonmedical) Non-cash Benefits
are Included
The
the
Department of Agriculture's m i n i m u m f o o d Inidgcts lor d i f
ferent sized f a m i l i e s by t h r e e . H o w e v e r , p a t t e r n s o f c o n s u m p t i o n have c h a n g e d a great d e a l s i n c e I 9 5 S For e x a m
pie, t h e p r o p o r t i o n o f i n c o m e spent o n f o o d has s h i f t e d
o v e r t i m e , w i t h t h e average f a m i l y s p e n d i n g a s m a l l e r p r o
Year
Otficial
Measure
1979
1989
1990
11.7
12.8
13.5
Plus Market
Value of
Food and
Housing
Benefits
p o r t i o n o l t h e f a m i l y b u d g e t o n t h i s necessity ( a n d . i n m m .
a l a r g e r p r o p o r t i o n o n o t h e r necessities). I l i e r e l o r e . if t h e
poverty
lines w e r e
recalculated
h i g h e r (as w o u l d p o v e r t y
today,
they
would
rales), s i n c e t h e f o o d
be
Inidgel
w o u l d be m u l t i p l i e d by a l a r g e r n u m b e r t h a n t h r e e .
T a b l e 6 . 2 presents t w o a l t e r n a t i v e measures o l p o v e r t y
9.6%
11.2
11.8
rales based o n u p d a t e d i n f o r m a t i o n o n c o n s u m p t i o n o l f o o d
a n d h o u s i n g ( h o u s i n g c o n s u m p t i o n has also s h i f t e d o v e r
time, a c c o u n t i n g for a larger percentage of family i n c o m e
7
l
t h a n i n the past) T h e available data are l o r IS)?" a n d \ ) H H
Point Increase
1979-89
W h e n t h e i n c r e a s e d real cost o f h o u s i n g (see ( l i a p t i r 8 ) is
1.1
1.6
factored into the p o v e r t y thresholds, the p o v c r t )
rate i n
I W 8 rises f r o m 13 0 " , io2.V0'.V.,a f u l l 10 p o i n t s a b o v e t h e
ofticial measure
VX'hen c o n s i d e r i n g t h e c h a n g i n g p r o p o r -
t i o n o f the l a m i l y b u d g e t d e v o t e d l o l o o d e x p e n d i t u r e s , t h e
rate is even h i g h e r ( 2S H'.V. ) l u n h c r m o r c . the g r o w t h o f
l
p o v e r t y o v e r the l ) 8 ( l s is e v e n m o r e i l r a m a l i i w h i n these
c o i i s i i m p t i o n s l i i f l s are t a k e n i n t o a c c o u n t . T h u s , i h a n g i n g
c o n s u m p l i o n p a t t e r n s i m p l y t h a i o u r o f f i c i a l p o v e r t y measures severely l i m i t rslale i h c c x t c n l a n d g r o w i h o l p o v e r t y .
As i i o l c d ahovc. [ I n o l l i c i a l m e a s u i e d o c s n o l t a i l o r i n
( h i value ol non-cash b e n e l i l s . s u c h as f o o d stamps, h o u s i n g
2^6
Changing
consumption
patterns imply
that our ofticial
poverty measures
severely
understate the
extent and
growth of
poverty.
�FIGURE 6B
Poverty Rates by Price Index,
1974-1991
16
15
/
QC
£
13
/
S
f
I
o
0.
/
/
12
t
/
/
^
I
^
I
^
I
I
^
N
/
I
^
N
N
\v
t
•
^
•
/
//
^^
CI 'I U X I
/
/
1
^
/
^
*
^
r.i'i u
/
/
11
/
/
/
-
10
/
\
1
1
p ^ ^
N<#
s<
1
1
1
1
1
? ^
^
^
^
N<#
1
1
1
^ ^
N
TABLE 6.4
Percent ol Persons with Low Relative Income, 1964-1989
Adjusted for Family Size
Relative Income Poverty Measures
Year
1969
1979
1989
Less than
1/4 the
Median
5.5%
6.7
8.3
278
1/4 to 1/2
of the
Median
12.4%
13.3
13.7
Less
than 1/2 of
the Median
17.97o
20.0
22.0
1
^
1
assistance, and medical henctits. thus undersiating the
extent of poverty Table 6.3 shows what poverty rates
would be il the market value ol lood and housing benefits
were included in recipients' income Medical benefits arcomitted since they are not a common part of everyday consumption like food and housing and their inclusion would
have the perverse effect of making the ill appear less poor.
Although the inclusion of food and housing benefits low ers
the absolute rates in both 1979 and 1989. the increase in
poverty over the period is even greater using the broader
measure.
As noted above, the official poverty lines are indexed lor
inflation. However, some analysts claim that the price index
used to adjust the poverty lines overstated inflation in the
1970s and early 1980s, and thereby overestimated real poverty rates Figure 6B tracks poverty rates, 197-1-1990,
using an alternative price index, CPI II X I , which is considered a more conservative measure of inflation. As would be
expected, the more conservative price index leads lo lower
measured poverty rates. However, Figure 68 shows that
regardless of the price index chosen to adjust the poverty
lines for price changes, there has been a rise in poverty
since 1979 despite overall income growth.
Another important indicator of poverty is reUitirv economic well being. A conceptual shortcoming of the ahso
lute poverty lines used in the above tables is that they arconly adjusted for inflation; they do not reflect overall
income growth. Ihe poverty lines are fixed levels of
income that represent a particular standard of living (level
of consumplion) at a point in time. However, as average
income grows over time and standards of living rise, the
economic "distance" between the officially poor and the
rest of society expands. While the earliest poverty lines
were close to 50% of the median family income for a given
family size, they have fallen to about 35'A,
Table 6.4 presents measures of poverty which adjust for
economic growth by measuring poverty rchiliiv to the
median family income (which changes yearly). As would be
expected (since median family income has grown taster
lhan prices), the poverty rates at one half the median are
substantially higher than the absolute rates in 'lable 6 I lor
example, whereas 12 8% of the population was poor
according lo the official, or absolute, poverty measure in
198V, 22.()''.. of all persons were in families with incomes
279
Although the
inclusion of food
and housing
benefits lowers
the absolute
rates in both
1979 and 1989,
the increase in
poverty over the
period is even
greater using the
broader
measure.
�below one hull of the median family income. Again, the
trend of worsening poverty over the last two decades is
evident.
TABLE 6.5
Poverty Gap: Aggregates and Means, 1967-1991
(1990 Dollars)
Families
Year
Aggregate
Poverty
Gap
(Millions)
1967
1973
1979
1983
1989
1991
$24,610
21,058
26,010
41,708
36,966
40,814
The Poor Get Poorer
Persons Not in Families
Mean
Poverty
Gap
Aggregate
Poverty
Gap
(Millions)
Mean
Poverty
Gap
$4,342
4,362
4,762
5,281
5,237
5,292
$13,361
12,060
15,153
19,874
20,225
22,497
$2,672
2,582
2,638
2,949
2,989
2,894
Percent Change
1967-73
1973-79
1979-89
-14.4%
23.5
42.1
0.5%
9.2
10.0
- 9.7%
25.6
33.5
TABLE 6.6
Persons Below 50% of the Poverty Level
Year
Percent
of all
Poor
1975
1979
1983
1989
1990
29.9%
30.2
38.5
38.0
39.4
280
Number of
Persons
(000)
7,733
8,340
13,590
11,983
14,059
- 3.4%
2.2
13.3
The depth of poverty at a point in time is another useful
gauge of how the poor are faring. As lable 6.5 shows, there
has not only been a growth in poverty; the poor have also
become poorer This is evident in measures of the so-called
•poverty gap," either the aggregate or the average dollar
amount persons or families are from the poverty threshold
For instance, in 1989, the average poor family had an
income J5,237 below the poverty line. The aggregate poverty gap, which was close to 137 billion in 1989, is the sum
of every poor family's income deficit.
looking again at trends, the aggregate gap grew by
42.1%, from approximately 26.0 to 37.0 billion dollars
from 1979 to 1989. The average deficit for families has
grown between 1979 and 1989 for both persons in families
(by 10% ) and not in families (by I 3 3% ) Such growth indicates that the poor have become poorer over time
Table 6.6 shows another measure of the depth of poverty: the percentage of the poor below 50% of the pov erty
line. In 1979, slightly less than one third ( 30 2%) of the
poor were in "deep poverty." By I9H3, this proportion had
approached two fifths (38 5%), where it essentially held
throughout the decade. Thus, not only have poverty rates
failed to fall in the context of the 1983 1989 recovery (as
Figure 6A and lable 6 I show), but more of the poor have
fallen into deep poverty.
Poverty Spells: The Length of Time Spent in
Poverty
All of the measures of poverty examined thus far have
been "point in time" measures. That is, we have compared
those in poverty (using different definitions) at one point
in time, to those in poverty at a different point in time I bis
approach fails to answer important c|ueslions about the
dynamics of poverty: how much mobility exists in the population of poor persons; are most poor persons in poverty
for a short or a long period; what personal characteristics
are associated with those experiencing short versus long
spells of poverty?
281
The average
poverty deficit for
families has
grown between
1979 and 1989
for both persons
in families (by
10%) and not in
families (by
13.3%)...
indicating that
the poor have
become poorer
over time.
�KcsL'urchcT.s have a n a l y z e d t h e s e i|iiesli(>ns a n d l o i m d a
TABLE 6.7
Distribution o Poverty Spells for
f
Nonelderly Persons Entering Poverty
Spell Length
(Years)
g o o d deal of t u r n o v e r i n t h e p o v e r t y p o p u l a t i o n ; c h r o n i c
p o v e r t y is t h e e x c e p t i o n , n o t t h e r u l e M o s t p o v e r t y s p e l l s
are relatively s h o r t , a l t h o u g h t h e r e d o e s e x i s t a s m a l l g r o u p
o f l o n g - t e r m p o o r (a " p o v e r t y s p e l l " is a c o n t i n u o u s p e r i o d
Percent of Persons
Completing
Poverty Spells
w h e r e i n i n c o m e is b e l o w t h e p o v e r t y , l i n e ) . U n f o r t u n a t e l y ,
t h e most r e c e n t
research o f t h i s t y p e c o v e r s t h e
time
p e r i o d f r o m t h e late 1960s t o t h e e a r l y l9H().s
T a b l e 6 . 7 s h o w s t h e d i s t r i b u t i o n o f p o v e r t y spells t o r t h e
1
2
3
4
5
6
7
8
9
>9
Total
44.5%
15.8
9.8
6.2
4.7
2.8
2.1
1.0
1.1
12.0
100.0
n o n - e l d e r l y p o p u l a t i o n , a n d f i n d s t h a t f o r t h e vast m a j o r i t y ,
p o v e r t y spells last for less t h a n t h r e e years ( t h e data c o v e r
t h e years 1 9 7 0 1 9 8 2 )
O n l y a m i n o r i t y o f spells last f o r
e i g h t o r m o r e years. The p e r c e n t a g e s i n t h e table c a n b e
i n t e r p r e t e d as a n s w e r i n g t h e f o l l o w i n g q u e s t i o n : " O f thosep e r s o n s just b e g i n n i n g a p o v e r t y s p e l l , w h a t
percentage
;
w i l l leave p o v e r t y a l t e r a g i v e n t i m e peri»Kli'" l o r e x a m p l e ,
44.VX, of p o v e r t y spells e n d e d w i t h i n o n e year, a n d a b o u t
7 0 % e n d e d w i t h i n t h r e e years. A r e l a t i v e l y s m a l l p e r c e n t
age o f p o v e r t y spells, I .V1 % , lasted o v e r e i g h t years.
T h e h i g h rates o f t u r n o v e r i n t h e p o v e r t y
population
i m p l i e d by l a b l e 6.7 suggest that f o r m o s t p e r s o n s , b e i n g
TABLE 6.8
Events Leading to Poverty Spells, and Average Spell Length
Primary Reason
for Spell Beginning
Earnings of Head Fell
Earnings of Wife Fell
Earnings of Others Fell
Transfer Income Fell
Rising Needs Standard*
Child Became Head or Wife
Wife Became Female Head
Child of Male Head Became
Child of Female Head
Child Was Born Into Poverty
Total/Average
Percent
of Beginnings
37.9%
3.7
7.7
8.0
8.2
14.7
4.7
6.4
8.6
100.0
Average
Length of
Completed Spell
(Years)
3.3
3.1
6.5
5.2
5.3
2.4
3.7
4.0
7.6
4.2
'The needs standard is a measure used in this research lo reflect an increased burden on a familys
economic situalion. generally resulting from the addition ol a new lamily member
p o o r is a t e m p o r a r y c o n d i t i o n . I n fact, o t h e r r e s e a r c h i n t o
p o v e r t y d y n a m i c s o v e r t h e 1970s f o u n d that as m u c h as
o n e - q u a r t e r o f t h e U n i t e d States p o p u l a t i o n w a s p o o r at
s o m e p o i n t w i t h i n a ten-year p e r i o d , u s u a l l y t o r a s h o r t
t i m e Recall also f r o m C h a p t e r I ( l a b l e M S ) t h a t t h e p r o b
a b i l i t y o f f a l l i n g o u t o f t h e m i d d l e class t o l o w i n c o m e sta
tus rose i n t h e 1980s Such f i n d i n g s d e b u n k t h e n o t i o n that
t h e m a j o r i t y o f t h e p o o r are c h r o n i c a l l y i m p o v e r i s h e d a n d
exhibit long-term dependency on government
assistance
(research on "welfare spells" shows similar f i n d i n g s to the
a b o v e ) F u r t h e r m o r e , as s h o w n i n t h e f o l l o w i n g tabic, t h e
events that m o s t c o m m o n l y cause a p o v e r t y s p e l l are phen o m e n a c o m m o n t o m o s t w o r k i n g p e r s o n s (e.g., j o b o r
e a r n i n g s ' loss, c h i l d b i r t h ) M o s t p e r s o n s e x p e r i e n c i n g s u c h
e v e n t s w i l l he p o o r f o r a r e l a t i v e l y s h o r t t i m e .
The e v e n t s that cause a p o v e r t y s p e l l t o l x - g i n , a n d t h e
average d u r a t i o n o f s u c h spells, are g i v e n i n T a b l e 6 . 8 , cove r i n g t h e s a m e p c r u x l as t h e p r e v i o u s t a b l e . As n o t e d , t h e
m o s t c o m m o n s p e l l - c a u s i n g e v e n t s are e a r n i n g s ' loss a n d
f a m i l y s t r u c t u r e c h a n g e s . Uy far t h e m o s t c o m m o n p o v e r t y
spells ( 3 7 . 9 % ) b e g a n w i t h a t l r o p i n e a r n i n g s by t h e l a m i l y
h e a d , a n d lasted a n average 3 3 years. Spells
beginning
w h e n a c h i l d ( p e r s o n 18 o r u n d e r ) b e c o m e s a f a m i l y h e a d
283
Researchers
have found a
good deal of
turnover in die
poverty
population;
chronic poverty is
the exception,
not the rule.
�o r a w i l e arc also r d a l i v c l y c o m m o n , a c c o u n i i n g l o r I \ . l ' k ,
TABLE 6.9
Length of Poverty Spells for Children Age 1-10*,
by Race, in Percent
of s p e l l h e g i n n i n g s Such spells are t h e shortest i n t h e table,
l a s t i n g o n average 2. i years. T h e longest average spells ( 7 (>
years) are caused by t h e b i r t h o l a c h i l d i n t o a l a m i l y
already p o o r ( t h e b e g i n n i n g s p e l l refers t o t h e n e w c h i l d )
Poverty Spell Length
White
Black
Other
T h e p h e n o m e n o n o f p e r s i s t e n t c h i l d p o v e r t y is e x a m i n e d
Total
next.
Never Poor
Poor 1 -3 Years
Poor 4-6 Years
Poor 7-10 Years
73%
19
5
3
22%
22
22
34
53%
28
10
8
T a b l e 6 . 9 e x a m i n e s t h e l e n g t h o f p o v e r t y spells l o r c h i l -
65%
20
8
8
d r e n b e t w e e n t h e ages o f o n e a n d t e n , by race ( t h e data are
f o r c h i l d r e n b o r n a r o u n d 1 9 7 0 ) . The l a b l e show s t h a i raceis an i m p o r t a n t d e t e r m i n a n t o f t h e d u r a t i o n o l p o v e r t y :
black c h i l d r e n have t h e h i g h e s t l i k e l i h o o d o f s p e n d i n g t h e i r
100
Total
100
100
c h i l d h o o d i n p o v e r t y . The last c o l u m n s h o w s that (>S% o f
100
children experienced no poverty
al a l l t h r o u g h o u t
p e r i o d . H o w e v e r , t h e results v a r y d r a m a t i c a l l y
the
by race
W h e r e a s 7 3 % o f w h i t e c h i l d r e n s p e n t n o p a r t of t h e p e r i o d
'Data are for children born around 1970.
p o o r , t h e same c o u l d b e said f o r o n l y 2 2 % of black c h i l d r e n .
W e l l o v e r h a l f ( 5 6 % ) o f b l a c k c h i l d r e n w e r e p o o r f o r al
least f o u r o f t h e f i r s t t e n years o f t h e i r l i v e s — f o r w h i t e c h i l d r e n , o n l y 8 % f e l l i n t o t h i s c a t e g o r y . Finally, o v e r a t h i r d
( 3 4 % ) o f black c h i l d r e n w e r e p o o r f r o m seven t o t e n years
FIGURE 6C
Persistent Poverty in Industrialized Countries: Households With Children,
Mid-1980s
of t h e i r first decade; the analogous f i g u r e for w h i t e child r e n w a s 3%
A f i n a l m e a s u r e o f p o v e r t y d y n a m i c s s h o w s that persis
tent p o v e r t y is m o r e c o m m o n i n t h e U.S. t h a n i n o t h e r
16
i n d u s t r i a l i z e d c o u n t r i e s ( t h i s issue is e x a m i n e d al l e n g t h i n
Chapter 9 ) F i g u r e 6 C shows the percentage of households
w i t h c h i l d r e n that w e r e p o o r i n e v e r y y e a r o f a t h r e e year
s t u d y i n t h e m i d 1980s ( p o v e r t y h e r e is m e a s u r e d i n rela
l i v e t e r m s , as 5 0 % o f t h e m e d i a n i n c o m e )
I n t h e U.S.,
14 4 % of h o u s e h o l d s w i t h c h i l d r e n w e r e p o o r t h r o u g h o u t
t h e t h r e e year study. O n l y Canada h a d a s i m i l a r l y h i g h rate,
at I 1.9%. T h e o t h e r c o u n t r i e s i n t h e f i g u r e h a d rales w e l l
b e l o w that o f t h e U.S.
H a v i n g e x a m i n e d a v a r i e t y of a l t e r n a t i v e measures o f
p o v e r t y , w e h a v e s h o w n that s o m e m e a s u r e s — t h e i n c l u s i o n o f i n k i n d benef its a n d t h e a l t e r n a t e p r i c e
index—
make the official measure appear
poverty
Other
measures—consumption
l o overstate
a d j u s t m e n t s a n d relative
m e a s u r e s — m a k e the o f f i c i a l measure appear to understate
0.4
us
Canada
West Germany
France
0 4
Luxemberg
Netherlands
p o v e r t y F u r t h e r m o r e , c h a n g e s i n t h e p o v e r t y gap. t h e i n c i d e n c e of d e e p p o v e r t y , a n d p o v e r t y d y n a m i c s are b e y o n d
t h e s c o p e o f t h e o l l i c i a l m e a s u r e H o w e v e r , n o n e o f these
alternative
284
measures
change
the fundamental
finding
285
Whereas 73% ot
white children
spent no part ot
the period poor,
the same could
be said for only
22% of black
children. Well
over hall (56%)
of black children
were poor for at
least four of the
first ten years of
their lives.
�regarding die trend toward high poverty rates in the
1980s, despite the recovery. Given the wide use of the olli
eial measure, and the tact that it adequately captures poverty trends, we will use- it throughout the rest ol this chap
ter, except where noted.
TABLE 6.10
Poverty b Race/Ethnicity
y
Poverty Rates
Year
White
1967
1973
1979
1989
1991
Black
11.0%
8.4
9.0
10.0
11.3
39.3%
31.4
31.0
30.7
32.7
Hispanic
NA
21.9%
21.8
26.2
28.7
TABLE 6.11
Percent o Children in Poverty, b Race, 1979-1991
f
y
Children under 18
Year
Total
1979
1989
1991
16.4%
19.6
21.8
White
11.8%
14.8
16.8
Black
Hispanic
41.2%
43.7
45.9
28.0%
36.2
40.4
2.5
8.2
Point Change
1979-89
3.2
3.0
Children under 6
Black
Hispanic
Year
Total
1979
1989
1991
18.1%
22.5
24.6
13.3%
17.1
19.2
43.6%
50.1
51.7
29.2%
39.6
44.6
4.4
3.8
6.5
10.4
White
Point Change
1979-89
2H6
The Poverty Status of Different
Demographic Groups
Certain demographic groups are more vulnerahle lo pov
erty than others. Members of minority groups, families
headed by an unmarried female, and children are at the
highest risk of poverty. The poverty problem of the elderly,
though historically significant, has been ameliorated to a
degree by government transfers.
Table 6.10 shows the official poverty rate for different
racial and ethnic categories. While both white and black
poverty rates fell from 1967 to 1973, they have generally
risen ever since. In all cases, black and Hispanic poverty is
more extensive than that of whites. The black poverty ratehas been about three times that of whites since 1979, while
the Hispanic rate has climbed from 21.9% in 1973 to
28.7% in 1991.
Table 6.11 has poverty rates for children, by race/ethnic
ity. Child poverty grew in all racial categories between
1979 and 1991. Young Hispanic children (under six) experienced the highest growth rate, gaining 10 4 percentage
points between 1979 and 1989. By 1991, more lhan one
out of every five children was poor For children under six
years old, the rate was even higher, reaching 24.6% in
1991. Perhaps the most alarming statistic in the lahle refers
to young black children: in 1991 more lhan half of black
children under six were poor.
As noted above, the poverty of the elderly has been
greatly reduced by government transfers. Table 6.12 and
Figure 6D examine the reasons for this decrease by contrasting the "before-transfer" poverty rate, which is the rale
that would prevail if there were no gov ernment cash assist
ance, with the actual or "after-transfer" poverty rate, which
includes such assistance I he reduction in poverty among
Ihe elderly has come largely from expanded government
transfer payments, particularly between 1967 and 1973.
During this period, ihe helore-lransfcr poverty rate clipped
very slightly, from 58.3 to 58.0%, but higher Social Security
287
Perhaps the most
alarming statistic
refers to young
black children: in
1991 more than
half of black
children under six
were poor.
�l
TABLE 6.12
Poverty Among the Elderly and All Persons
Before and After Transfers, 1967-1990
Poverty Rate Among.
All Persons
The Elderly
After
Transfers
Before
Transfers
Year
1967
1973
1979
1989
1990
19.4%
19.0
20.4
20.9
21.5
14.3%
11.1
11.6
12.8
13.5
Before
Transfers
58.3%
58.0
58.9
51.9
51.0
After
Transfers
29.7%
16.1
15.1
11.4
12.1
FIGURE 6D
Poverty Before and After Transfers: Elderly vs. All Persons,
1967-1990
70
60
Elderly Befoie Tianslers
50
«
40
>.
•c
a>
> 30
CL
' \
E l d e r l y Aller Transfers
All Persons Before Transfers
20
10
All Persons A l l e i Transfers
I
I
288
\
I
I
I
benefits virtually halved after-transfer poverty, Irom 2 >.7
to 16 1% In 1989, more lhan half of the elderly would have
been poor without government transfer programs. The
effect of government benefits in 1989 was to dramatically
reduce the poverty rate among the elderly from SI 9 lo
1 1.-1%.
Family Structure and Poverty
I'amily structure has historically been an important
determinant of poverty status, as certain family types are
more vulnerable to poverty than others. However, the
important questions regarding family structure and pov erty are as follows: to what extent are poverty rates and
family structure changes causally related? 'lb what degree
are the high and intractable poverty rates documented
above driven by individuals' choices to form vulnerable
family types?
Table 6.13 takes a first look at these questions by show
ing the percentage of persons in female headed families,
married couple and male-headed families, and not in families f r o m 1959 to 1991 The table also shows the poverty
rates for each group i n each time period, along w i l h per
centage point changes over lime.
Persons in female headed families and persons not in
families are more likely to he poor than persons in married
couple or male headed families In 1991, 39.7% o f persons
in families headed by a w o m a n were poor f or persons not
in families in 1991, the poverty rate (211%) was also substantially higher than the rate for all persons. Those in married couple and male headed families are ihe least likely to
be poor; since 1967 their poverty rates have stayed below
10%.
Over time, the distribution of family types has shifted
toward family structures more vulnerable lo poverty. The
percentage of persons in married-couple and male headed
families, which have the lowest poverty rales, has consistently fallen, f r o m 8 5 . 9 % in 1959 l o 7 1 5 % in 1991 Con
versely, there has been a consistent expansion of femaleheaded families and an even faster g r o w t h ol singles.
Turning lo the poverty trends of these different groups,
all family types saw their poverty rales fall between I 967
and 1979, w i t h single persons showing the largest drop
289
Over time, the
distribution oi
family types has
shitted toward
family structures
more vulnerable
to poverty.
�TABLE 6.13
Changing Family Structure and Poverty, 1959-1991
Percent of Persons in:
Year
1959
1967
1973
1979
1989
1991
Point Changes
1967-79
1979-89
FemaleHeaded
Families
8.0%
9.1
10.5
12.1
13.2
13.9
3.0
1.1
MarriedCouple and
MaleHeaded
Families*
85.9%
84.2
80.7
76.2
72.5
71.5
-8.0
-3.7
Not
Living in
Families
6.1%
6.7
8.8
11.7
14.3
14.7
( 16 2 points) This trend was reversed in 1979 1989, when
poverty grew slightly lor persons in inarried-eouple and
female-headed families (0.9 and 1.0 percentage points,
respectively), and fell nnich more slowly for individuals.
What does lahle 6 13 reveal about the relationship
between demographic shifts and changes in poverty ratcsf
The p r i m a facie evidence is mixed. On the one hand, it is
clear that there has been a compositional shift lo families
more vulnerahle to poverty. However, when the demo
graphic shifts were occurring most rapidly, 1 9 6 7 - I V ) , the
overall poverty rates declined from I 1.2 to I 1.7". ami pov
erty fell among persons in female headed Iamilies and sin
gles. Conversely, when demographic forces slowed over the
1980s, the poverty trend reversed (as noted above), and
persons in married couple families saw a similar increase in
their poverty rate as did those in temale headed families
(0.9 versus 1.0 percentage points) Thus, while demo
graphic shifts to family types w i t h elevated vulnerability to
poverty have played a role in the high poverty rales of the
1980s, the extent of that role is unclear It is lo this issue we
now turn.
1
Total
100.0%
100.0
100.0
100.0
100.0
100.0
5.0
2.6
Poverty Rate of Persons in:
All
Persons
Year
1959
1967
1973
1979
1989
1991
49.4%
38.8
37.5
34.9
35.9
39.7
18.2%
9.6
6.0
6.4
7.3
7.9
46.1%
38.1
25.6
21.9
19.2
21.1
22.4%
14.2
11.1
11.7
12.8
14.2
Point Changes
1967-79
1979-89
-3.9
1.0
"Including unrelated sublamilies since 1979.
290
-3.2
0.9
- 16.2
-2.7
-2.5
1.1
Decomposing the Role of Demographics
One way to investigate the effect of demographic changeon poverty rates is to assume that poverty rates are influenced by three factors: I ) changes in family structure, i.e.,
the proportion of persons living in the different family
types shown in lahle 6.13; 2 ) changes in economic factors,
like lower wages and benefits, as well as changes in family
size; 3 ) the combined effect of these t w o factors
Table 6.14 is an example of this exercise I he first col
umn shows the actual changes in poverty rates over peak
years in the economy. Column t w o shows how poverty
rates would have been expected to change over the period
in question, given the changes that look place in lamily
structure l o r example, family structure changes occurring
between 1959 and 1967 are predicted to have led to 0 5
percentage point g r o w t h in overall povcrt)' However, nondemographic factors had a powerful countervailing elfect
and poverty rates fell 8 2 points oxer the period There was
no interaction effect during this period (column four)
291
// is clear that
there has been a
compositional
shift to families
more vulnerable
to poverty.
However, when
the demographic
shifts were
occurring most
rapidly the
overall poverty
rates declined.
�The s e c o n d c o l u m n s h o w s that t h e d e m o g r a p h i c
TABLE 6.14
Changing Family Structure and Poverty
shilt
over time into families more susceptible to p o v e r t y — t h e
increase- i n t h e p r o p o r t i o n of p e r s o n s f i x i n g a l o n e o r
in
t e m a l e headed h o u s e h o l d s ( lable 6 I S), has i n fact c o n t r i b
Period
Predicted
Change Due
to Family
Structure*
Change
in Poverty
Rate
Interaction
ol Family
Structure and
Other Factors**'
Change Due
to Other
Factors*'
uted
to
higher
poverty
rates
However,
the
post-1907
d e m o g r a p h i c p r e s s u r e o n p o v e r t y rates w a s at its l o w e s t i n
the p e r i o d f r o m 1979-1989, w i t h the e x c e p t i o n o l the t w o
year p e r i o d 1989-1991
T h i s d i l i e r e n c e is a result o l t h e
relatively s l o w recent g r o w t h o f female-headed
(D
(3)
(4)
-8.2
-3.1
0.6
1.1
1.4
1959-67
1967-73
1973-79
1979-89
1989-91
(2)
0.5
1.0
1.1
0.7
0.2
-8.7
-4.0
-0.3
0.5
1.2
0.0
-0.2
-0.2
-0.1
0.0
•Eflecl on overall poverty rate due to the changes in proportion ot persons in lemale-headed families,
married-couple and male-headed families, and not in families.
"Effect on overall poverty rate due to the changes in poverty rates within these three main demographic groups.
•••Effect on overall poverty rate due to the interaction of family structure changes and poverty rate
changes.
Iamilies
a n d s i n g l e p e r s o n s , a n d t h e f a l l i n g p o v e r t y rates o f these
same d e m o g r a p h i c g r o u p s p r i o r t o 1 9 7 9
C o l u m n t h r e e i n l a h l e 6.1 3 s h o w s that i n p r e v i o u s p e r i
ods, n o n - d e m o g r a p h i c f a c t o r s w o r k e d t o r e d u c e p o v e r t y .
B e t w e e n 1 9 5 9 a n d 1 9 7 3 , these c o u n t e r v a i l i n g f a c t o r s w e r e
far m o r e p o w e r f u l t h a n t h e d e m o g r a p h i c e f f e c t s s h o w n i n
c o l u m n t w o , so that t h e p o v e r t y rate f e l l s u b s t a n l i a l l y H o w
ever, b e t w e e n 1 9 7 9 a n d 1 9 8 9 , factors s u c h as l o w e r w a g e s
a n d a less e f f e c t i v e
transfer system actually
added
0.5
p o i n t s t o t h e p o v e r t y rate. D u r i n g t h i s same p e r i o d , d e m o g r a p h i c c h a n g e s a d d e d 0.7 p o i n t s , less t h a n t h e
demo
graphic impact in the t w o previous periods
T h e p o i n t o f t h i s d e c o m p o s i t i o n e x e r c i s e is that w h i l e
t h e s h i f t t o m o r e p o v e r t y - p r o n e f a m i l y t y p e s has c o n t i n u e d
t o c r e a t e u p w a r d p r e s s u r e o n p o v e r t y rates, that pressure
has lessened t h r o u g h o u t t h e I98()s. T h e s h i f t o v e r t i m e l o
TABLE 6.15
Poverty Rates for Female-Headed Families
f e m a l e - h e a d e d f a m i l i e s a n d p e r s o n s l i v i n g a l o n e has a d d e d
t o p o v e r t y rates, y e t s u c h s h i f t s fail t o f u l l y e x p l a i n (he h i g h
p o v e r t y rates o f t h e 1980s. M o r e o v e r , t h e m o s t i m p o r t a n t
All
Year
White
Black
24.56%
22.3
25.4
28.4
52.76%
49.4
46.5
51.2
Hispanic
f a c t o r i n r i s i n g p o v e r t y a p p e a r s t o b e the f a i l u r e o f econ o m i c f o r c e s t o lessen p o v e r t y
1973
1979
1989
1991
32.2%
30.4
32.2
35.6
51.4%
49.2
47.5
49.7
The Poverty of Female-Headed
Families
Female-headed f a m i l i e s w e r e i d e n t i f ied above as p a r t i c u larly v u l n e r a b l e t o p o v e r t y . T h e s e f a m i l i e s are s t e r e o l y p i e a l l y t h o u g h t t o he m o s t l y b l a c k w o m e n w h o have had c h i l
dren out-of-wedlock.
The n e x t f e w tables e x a m i n e these
m y t h s a n d f i n d that m o t h e r - o n l y
family
poverty
is not
e x c l u s i v e l y a m i n o r i t y p r o b l e m , n o r are most s i n g l e m o t h ers n e v e r - m a r r i e d w o m e n .
Table
6.15
examines
the
trend
in
poverty
rales
of
female-headed f a m i l i e s by race a n d H i s p a n i c o r i g i n . O v e r a l l , t h e i r p o v e r t y rales have c h a n g e d l i t t l e s i n c e 1973 Furt h e r m o r e , w h i l e m i n o r i t y f e m a l e h e a d e d f a m i l i e s have the
highest p o v e r t y rales. I h e i r rales d e c l i n e d b e t w e e n
292
1979
293
The shift over
time to
female-headed
families and
persons living
alone has added
to poverty rates,
yet such shifts
fail to fully
explain the high
poverty rates of
the 1980s.
�a n d 1 9 8 9 , w h i l e t h o s e t o r w h i l e s rose. T h e p e r c e n t a g e o f
TABLE 6.16
Increase in Poverty in Female-Headed Families,
b Race, 1973-1991
y
black female-headed families i n p o v e r t y fell f r o m
-i6.')'.V,; H i s p a n i c , f r o m
19 2 l o i 7 . W ,
l
i >. i t o
T h e rale for w h i t e s
d u r i n g t h i s s a m e p e r i o d i n c r e a s e d f r o m 22 S l o 25. I'i,.
W h e n t h e p o p u l a t i o n o f f e m a l e headed f a m i l i e s is e x a m
White
Year
Black
Other-
Total
Number(OOO)
i n e d as a w h o l e , it is c l e a r that w h i t e f a m i l i e s c o m p r i s e t h e
largest
share
(Table
6.16).
In
Iamilies
f o r 4 3 . 5 % . The b o t t o m p a n e l o f 'lable (> l(» s h o w s t h e share
of t h e t o t a l i n c r e a s e i n t h e n u m b e r o f p o o r f e m a l e headed
f a m i l i e s by race
W h i l e b l a c k s a c c o u n t e d for t h e largest
share o f t h e i n c r e a s e b e t w e e n
1973
1979
1989
1991
1989, w h i t e
a c c o u n t e d f o r VV0'*> o f t h e t o t a l ; black f a m i l i e s a c c o u n t e d
Percent of Poor
Female-Headed
Families Who Are in
Each Racial Group
54.3%
51.0
53.0
52.7
44.4%
46.7
43.5
44.1
1.3%
2.3
3.5
3.2
100.0%
100.0
100.0
100.0
2,193
2,645
3,504
4,161
1973 and
19^9 (57.5"„).
w h i t e s w e r e r e s p o n s i b l e f o r t h e largest share b e t w e e n 1979
a n d 1 9 8 9 ( 5 9 1 % ).
T h e s t e r e o t y p i c a l f e m a l e h e a d o f a f a m i l y is t h o u g h t to
have n e v e r b e e n m a r r i e d a n d t o have had a n u m b e r o f outo f - w e d l o c k b i r t h s . I n fact, as T a b l e 6 . 1 7 s h o w s , never-mar-
Share of Total
r i e d f e m a l e heads c o m p r i s e t h e s e c o n d smallest c a t e g o r y i n
Increase
1989. I n t h a t year, 2 2 . 6 % o f f e m a l e f a m i l y heads w e r e
1973-79
1979-89
35.4%
59.1
57.5%
33.8
7.1%
7.1
100.0%
100.0
452
859
n e v e r m a r r i e d , w h i l e .36.4% w e r e d i v o r c e d , l i v e n w i d o w e d
f e m a l e f a m i l y heads a c c o u n t e d f o r a l a r g e r p r o p o r t i o n o f
the total g r o u p than never-married mothers in
t h e y have s i n c e
"Includes Asians, Pacific Islanders, American Indians. Aleuts, and Eskimos.
1973). However, never-married
1 9 8 9 (as
mothers
w e r e t h e fastest g r o w i n g g r o u p ; i n t h e 1 9 7 9 1 9 8 9 p e r i o d
t h e y g r e w by 8 4 4 %
TABLE 6.17
Marital Status of Female Family Heads, 1973-1989
A great d e a l o f c o n t r o v e r s y e x i s t s r e g a r d i n g t h e rise of
o u t - o f - w e d l o c k b i r t h s d o c u m e n t e d i n t h e p r e v i o u s table.
C o n v e n t i o n a l t h i n k i n g suggests that a p r e c i p i t o u s rise i n
Marital Status of Female Family Heads:*
Year
1973
1979
1989
Married,
Husband
Absent
23.9%
21.0
17.1
Widowed
Divorced
Never
Married
25.9%
33.0
36.4
12.6%
16.2
22.6
37.7%
29.8
23.9
Total
10.6%
8.1
-0.4%
6.0
t i c u l a r l y a m o n g b l a c k s , has b e c o m e a m a j o r p o v e r t y p r o b l e m . I h i s p e r c e p t i o n has g r o w n o u t o f the fact that t h e r a t i o
o f e x p e c t e d u n m a r r i e d t o m a r r i e d l i f e t i m e b i r t h s ( t h i s w ay
o f m e a s u r i n g b i r t h t r e n d s is e x p l a i n e d b e l o w ) has g r o w n
100.0%
100.0
100.0
6,535
8,220
10,890
significantly, f r o m
23.2%
t o 6 2 . 2 % tor blacks, b e t w e e n
I 9 6 0 a n d 1 9 8 7 ( T a b l e 6 . 1 8 ) H o w e v e r , i h e data show that
w h i l e out ol w e d l o c k b i r t h s have b e c o m e m o r e
common
f o r b o t h b l a c k s a n d w h i t e s , t h e y have n o t i n c r e a s e d d r a m a t ically. W h a t is d r i v i n g t h e rise- i n t h e r a t i o o f u n m a r r i e d l o
Percent Increase in Number of Families:
1973-79
1979-89
female-headed f a m i l i e s l e d by n e v e r - m a r r i e d m o t hers, par-
Total
Number
(000)
married
60.2%
46.4
62.5%
84.4
25.8%
32.5
births,
particularly
among
blacks,
is less
an
increase i n o u t - o f - w e d l o c k b i r t h s t h a n a decrease i n m a r
ried births
liible 6 18 a n i l F i g u r e 6 E p r e s e n t t h e relevant data. T h e
table a n d l i g u r e
"Women without dependents are nol regarded as lamily heads.
29-1
use
the concept
of expected
lifetime
b i r t h s , i.e. t h e n i m i b c r of i n - a n d o u t o f w e d l o c k c h i l d r e n a
295
The stereotypical
temale head of a
family is thought
to have never
been married and
to have had a
number of
out-ot-wedlock
births. In fact,
never-married
female heads
comprise the
second smallest
category in 1989.
�TABLE 6.18
Expected Lifetime Births, by Marital Status and Race, 1960-1987
Fertility
I960
1965
1970
1975
1980
1987
0.08
1 05
0.11
1.08
0.14
1.16
0.12
1.09
0.18
1.25
0.29
1.43
3.45
3.49
2.67
2.75
2.25
1.93
1.56
1.15
1.57
1.01
1.47
0.87
3.53
4.54
2.78
3.83
2.39
3.10
1.69
2.24
1.75
2.27
1.77
2.29
Unmarried Births'
White
Black
Married Births'
White
Black
Total
White
Black
Unmarried Births as
Percent of Lifetime Births
2.3% 4.0% 5.7% 7.3% 10.2% 16.7%
23.2
28.2
37.6
48.8
55.5
62.2
White
Black
'The numbers in the table refer to the expected number of lifetime births given prevailing birth rates.
FIGURE 6E
Expected Lifetime Births By Marital Status and Race,
1960-1987
( 3 ] While Births. Unmamed
[ J
|
Black Births. Unmarried
| While Births. Married
Black Births. Married
c
m
S 2
3
1960
1965
296
1970
1975
1980
1987
woman could expect lo have throughout her lile, derived
from the b i r t h rates in each time period l o r example,
based on prevailing b i r t h rales al the lime, a married white
woman in I 9 6 0 would have been expected lo have an aver
age 3.-15 children over her lifetime.
The first t w o bars for each year in l igure 6li show that
while blacks have a higher level of expected unmarried
births than whites, both groups show a slight upward trend
While a rise in the number of unmarried births is not a triv
ial development, it is hard to imagine that this development
alone—a rise- over the period in expected black unmarried
births from 1.0S to 1.43—would have caused much controversy. The significant trend is the falling expected births to
married w o m e n o f both races, as the average white woman
in I 9 6 0 went from having 3.45 children in her lifetime to
having 1.47 children in 1987. The expected child-bearing
behavior o f her black counterpart changed even more dramatically, w i t h her expected births falling from 3.49 to
0.87. The ensuing rise in the ratio of unmarried to all
binhs (as shown in the last panel of Table 6.18) has been
driven to a greater degree by the falling birth rates of married women than the rising rales o f unmarried women
When comparing child hearing behavior of black and
white women, lahle 6 18 shows that in I960, expected lifetime births were about equal for married women of b o l l i
races ( 3 45 for whites; 3 49 for blacks) However, married
women began a dramatic decline in their child bearing
behavior, w i t h lifetime births to married black women
declining faster than married white women My 1970, a typical married black woman was expected lo have fewer children than her white counterpart, and by 1987 (the latest
year of available data), the expected births between the
t w o races diverged significantly. This trend had t w o effects
It lowered overall expected births ( liible 6 18). and il led
to the appeanuue of vastly accelerated out o f w e d l o c k
birth rates, especially for black women.
In sum, family structure changes cannot be held accountable for the failure of economic grow th lo reduce poverty
rates in the I98()s Although a demographic shift to family
types more vulnerable to poverty has contrihutcd lo high
poverty rales over lime, most of this shift occurred in earlier periods As liible 6.14 (column t w o ) shows, lamily
slrucmre created about the same degree of upward pressure on povcrt) rates from 1979-1989 as such changes did
While a rise in
the number ot
unmarried births
is not a trivial
development, it
is hard to
imagine that this
development
alone—a rise
over the period in
expected black
unmarried births
from 1.05 to
1.43—would
have caused
much
controversy.
�TABLE 6.19
1979-1989
Poverty Rates* Using Different Income Definitions,
All Persons
1979
1989
Point change
1979-89
Plus
MeansTested
Cash
Transfers
(D
Family Relationship
Market
Income
Before
Transfers
Plus Social
Insurance
(Including
Social
Security)
(2)
(3)
(4)
(5)
19.1%
19.9
12.86%
13.8
11.6%
12.8
9.6%
11.2
9.9%
11.8
Plus
Food and
Less
Housing Federal
Benefits
Taxes
Poverty Rate
Reduction
Due to
Government
Taxes and
Benefits
(1-5)
9.2
8.1
08
1.0
1.2
1.6
1.9
-1.1
Persons in Single Parent
Family with Related
Children under 18
1979
1989
50.0%
48.1
45.1%
44.6
40.2%
42.3
30.0%
36.4
30.1%
36.4
19.9
11.7
Point change
1979-89
-19
-0.5
2.1
6.4
6.3
-8.2
Persons in Married Couple
Family with Related
Children under 18
1979
9.4%
1989
105
7.7%
9.1
7.1%
8.5
5.9%
7.4
6.3%
8.0
Point change
1979-89
1.4
1.4
1.5
1.7
1.1
•Rates are based on adding benefits to income cumulatively, from left to right.
298
between 1959-1967, w h e n poverty fell 8.2 points. The
poverty of female-headed families, w h i l e high relative to
overall poverty, has risen only slightly over the 1980s
(Table 6.16). Furthermore, Tables 6.17 and 6.18 (and Figure
6E) show that the controversy surrounding out of-wedlock
births is overstated. Never-married mothers still comprise a
minority of poor female family heads, and the rise i n the
ratio of out-of-wedlock births to married births for blacks is
being driven by a dramatic fall i n the latter rather than a
substantial rise in the former.
3.1
2.5
-0.6
The Changing Effects of Taxes
and Transfers
Chapter 2 made the point that the decreased progressivity of the tax system over the 1980s favored the wealthy
and hurt the poor. I n tandem w i t h this development, government transfers became less generous over the decade,
particularly i n the early Reagan years. Since the official poverty statistics are calculated prior to tax payments (or
receipts), and excluding the value of in-kind benefits, they
shed little light on the effects of policy changes i n these
areas. Therefore, Table 6.19 uses a more inclusive measure
of poverty rates than the official rate—one that includes
five separate income definitions.
Table 6.19 examines the effect o f taxes and benefits on
poverty rates, for all persons and for family types w i t h children, 1979 and 1989. Column one shows the poverty rates
generated by the market—these rates represent the degree
of poverty that w o u l d exist i n the absence of any government intervention. Moving left to right, the table introduces different transfers and taxes, and shows how poverty
would be effected, showing i n the first row, for example,
that the poverty rate for all persons fell from 19.1 to 12.8%
due to the receipt of social insurance benefits. The final
column shows the extent to w h i c h government tax and
transfer policies mitigated the poverty rates generated by
market outcomes (here, negative signs mean less effective
poverty reduction). For instance, taxes and transfers
reduced the poverty rate by 9.2 percentage points in 1979;
i n 1989 they led to an 8.1 point reduction. Thus, the
reduced effectiveness of taxes and transfers led to a 1.1
point increase i n the poverty rate.
299
The reduced
effectiveness of
taxes and
transfers led to a
1.1 point
increase in the
poverty rate.
�FIGURE 6F
Poverty Before and After Transfers: Families with Children,
1967-1990
70
Female-Headed
Before Transfers
60
50
Female-Headed
Alter Transfers
0)
£ 30
Married-Couple and
Male-Headed Before Transfers
20 -
J
I
I
I
£ $ $i£
I
I
I
I
I
I
I
I
I
Married-Couple and
Male-Headed After Transfers
I I I l
J
I
1
L
£ i£ £ £ £ £ £ i i i S i i i i i i i
300
Almost every category of tax or transfer in the table
shows either worse market outcomes or less effective poverty reduction i n 1989 compared to 1979. Overall, tax and
transfer policies reduced poverty rates by 9.2 percentage
points in 1979, from 1 9 1 % to 9.9%. I n 1989, the market
was 0.8 points less effective in reducing poverty before
government intervention, and 1.1 points less effective after
taxes and transfers. Single-parent families w i t h children
were the biggest losers i n terms of reduced effectiveness
against poverty. I n 1979, through the receipt of various
cash and in-kind benefits net of taxes, their poverty rates
were reduced by 19.9 points; in 1989, taxes and transfers
took only 11.7 points o f f of their market outcome poverty
rates. The importance of the reduced effect for single-parents w i t h children is particularly notable in light of the fact
that their market outcome was slightly better in 1989 than
in 1979 ( 4 8 . 1 % in poverty versus 50.0%). The implication
is that the primary distribution (market outcome) was
somewhat more favorable to these persons in 1989 than in
1979; their entire increase i n post-tax, post-transfer poverty is due to reduced government assistance.
Married couple families also had higher rates o f poverty
in 1989 than i n 1979 The market was 1.1 points less effective at poverty reduction; each category of tax and transfer
once again led to higher poverty rates in 1989 so that by
the end of the period, poverty rate reduction fell by 0.6
points. Since the rates are based o n adding the value of benefits cumulatively, we can take the ratio of the point change
in column one to that in column five as an estimation of the
proportion of reduced effectiveness due to market forces.
For married couples, about two-thirds (1.1/1.7) is attributable to poverty-inducing market forces.
F i g u r e 6F portrays the effect of transfers (taxes are
ignored) on the poverty of non-elderly persons in families
w i t h children from 1967-1990, by comparing pre- and posttransfer poverty rates. The high poverty rates of femaleheaded families relative to married couple and male-headed
families are evident i n the figure. A subtler point is made by
noting the distance between pre- and post transfer poverty
rates for both types of families. The t w o lines for male-headed and married couple families maintain a fairly consistent
gap over time, suggesting that the effectiveness of transfers
has not varied a great deal for these family types. However,
the gap between the t w o lines for persons in fcmalc-hcadcd
301
The gap between
the two lines for
persons in
female-headed
families begins
to visibly narrow
in the 1980s
underscoring the
reduced
effectiveness of
taxes and
transfers.
�FIGURE 6G
Demographic and Benefit-Sum Trends,
1960-1989
Percent
40
Non-White
Female
Headship Rate
30
Black Female
Headship Rate
20
10
White Female
Headship Rate
Divorce Rate
20
15
10
Non-White Out-of-Wedlock Births
500
400
300
200
100
. (per 1000 live binhs)
White Oul-ol-Wedlock Births
Dollars;
1100
1000
900
800
700
Real Benefit Sum
I ' ' '
I I I I I I I
302
families begins to visibly narrow in the 1980s, echoing the
findings in Table 6.13Government
Benefits and Family
Structure
The previous sections examined the roles of family structure and benefit changes on poverty rates. This section
looks at the interaction between these two issues and
examines a question which has often been raised in the
debate over government provision to the poor: to what
extent have welfare programs caused poverty by creating
incentives that heighten the probability that a family will
remain poor? Specifically, welfare programs have been
attacked as the primary cause of the increased proportion
of female-headed families.
The argument regarding family structure has been developed as follows. Since the most valuable benefit package is
available to single-headed families with children, there
exists an incentive to form such families, either by delaying
or avoiding marriage, dissolving existing unions, or having
children out-of-wedlock. In fact, the evidence shows that
the proportion of families with children headed by females
has grown over-time, as has the benefit sum, at least up to
the mid-1970s (the benefit sum includes the value of Aid to
Families with Dependent Children [AFDC), Food Stamps,
and Medicaid; see Figure 6G, bottom panel). Critics of the
system of social provision have suggested that there exists a
causal relationship between these two phenomena: family
structure has been altered by the provision of benefits to
female-headed families with children.
Yet the fact that both the benefit sum and the proportion
of single-parent families with children has grown over time
does not necessarily imply that the former causes the latter.
Figure 6G challenges this causal connection. The figure
shows that all the demographic indicators of growing
female-headed families follow an upward trend since I960.
Throughout the initial part of the period, the benefit sum
(bottom panel in the figure) also grew.
However, beginning around the mid-1970s the benefit
sum began to fall, due primarily to the declining real value
of AFDC benefits. Yet none of the demographic indicators
follow the plunging benefit sum. The only series in the figure that reverses direction is the divorce rate, the one rate
in the figure that is least relevant to the low-income population. While white female headship and non-white out-of-
303
The apparent
independence
ot these
demographic
indicators and
the benefit sum
challenges the
overly simplistic
story that welfare
causes the rising
proportion of
female-headed
families.
�TABLE 6.20
AFDC Participation Rates of Female-Headed Families
with Children, 1967-1987*
Year
1967
1973
1979
1983
1987
Participation
Rates
36%
63
52
45
42
'Participation rate is the ratio of mother-only families on AFDC to mother-only families not participating in
AFDC.
FIGURE 6H
AFDC Participation Rates,
1967-1987
wedlock birth rates do appear to slightly decelerate by the
end of the time period covered, the white out-of-wedlock
birth rate does not appear to attenuate at all. The apparent
independence of these demographic indicators and the
benefit sum challenges the overly simplistic story that welfare causes the rising proportion of female-headed families.
Further evidence against the "welfare is the problem"
argument comes from an analysis of AFDC participation
rates, i.e., the proportion of the demographically eligible
population (female-headed families with children) participating in the program. If single-mother families were forming specifically to take advantage of the eligibility for welfare, we would expect to see a greater proportion of
female-headed families participating in AFDC over time
(i.e., rising participation rates). In fact, as Table 6.20 and
Figure 6H show, participation rates peaked i n the
mid-1970s and fell thereafter, apparently stabilizing in the
1980s.
How are falling participation rates to be explained in the
context of the growing proportion of female-headed families with children? The explanation is quite simple: femaleheaded families are forming for reasons other than to
ensure AFDC eligibility. The trend in Figure 6H suggests
that it is a mistake to necessarily associate the rise in
female-headed families with AFDC participation. In fact,
common sense dictates the independence of these two
events (family structure and AFDC eligibility). The decision
to form a family of a particular structure is most likely to be
based on a wide variety of factors, reflecting not only economic factors, but incorporating social influences such as
society's changing mores regarding single-parent families
and divorce.
Finally, the question of family structure and welfare has
been extensively examined by economists who have modelled the family formation process, estimating the impact of
a wide variety of variables thought to play a role in family
structure decisions, including, of course, welfare benefits. A
detailed discussion of this econometric work is beyond the
scope of this text. However, the findings of this literature
can be summarized as follows:
— the effect of the value of the benefit package has a small
effect on family structure;
— this small effect is most apparent in lower rates of
remarriage, and the decision of female-headed families
304
305
Female-headed
families are
forming for
reasons other
than to ensure
AFDC eligibility.
�TABLE 6.21
Shares o Workers Earning Enough to Maintain a Family o Four al 0.75
f
f
of Poverty and u to the Poverty Line, b Gender and Race
p
y
Workers
Share of Workers
Earning:
Less Than 0.75 of
the Poverty Line
Whites,
Non-Hispanics
1979
2.4%
Blacks,
Non-Hispanics
Point
Change
1989 1979-89 1979
8.1%
5.7
Hispanics,
Any Race
Point
Change
1989 1979-89 1979
4.2% 15.1%
10.9
Point
Change
1989 1979-89
3.7% 15.9%
12.2
Between 0.75 of
Poverty and the
Poverty Line
12.3
10.7
-1.6
20.9
18.2
-2.7
22.9
22.4
-0.5
Less Than the
Poverty Line
14.8
18.8
4.1
25.2
33.3
8.1
26.6
38.3
11.7
F
e
Workers
Whites,
Non-Hispanics
Blacks,
Non-Hispanics
Hispanics,
Any Race
S : of Workers
Earning:
1979
Point
Change
1989 1979-89 1979
Less Than 0.75 of
the Poverty Line
6.1%
16.4%
29.7
16.4
-13.4
33.1
19.3
-13.8
41.2
23.2
-17.9
35.8
32.7
-3.1
38.3
40.6
2.3
48.9
45.2
-3.8
Bi
0.75 of
Poverty and the
Poverty Line
Less Than the
Poverty Line
10.3
Point
Change
1989 1979-89 1979
5.26% 21.3%
16.1
Point
Change
1989 1979-89
7.8% 21.9%
4.2
to live independently (i.e., to form their own households instead of living with a relative);
— a review of the findings suggest that, over the period
covered in Figure 6G, the increase in the benefit sum
led to a 9 to 14 percent increase in the prevalence of
female-headed families with children.
Having examined the effect of the changes in taxes and
transfers over the 1980s, and the impact of benefits on family structure, we now turn to an explanation of the high
poverty rates over the decade.
Explaining the High Poverty Rates of
the 1980s
As noted above (see also Chapter 2), the tax system grew
less progressive over the decade, and government benefits
grew less effective at reducing poverty (Table 6.19) However, when we consider these findings in the context of the
relatively long economic expansion of 1983-89, another
question is suggested: why were the poor unable to "cash
in" on the economic growth that prevailed throughout
most of the decade? That is, given the expansion, we would
expect those at the bottom of the income distribution (at
least those with some connection to the labor market) to
respond to the less generous transfers and more regressive
taxes documented above by turning to the labor market. In
the following section, we argue that such a reaction did in
fact occur, but that the wages of low-income workers
declined significantly over the decade (see Chapter 3), and
this development served to keep the poverty rates high
throughout the decade.
Falling Wages and Poverty in the 1980s
The problem of declining wages has already been presented in Chapter 3. The relevant part of that discussion in
the context of this chapter is the expansion of low-wage
employment and the declining wages of low-wage workers.
The following tables specifically examine earnings trends
as they affect the poor.
Table 6.21 answers the following question: given the
wage structure at a point in time, what percentage of workers, working full-time/full-year, fail to earn an hourly wage
that would lift a family of four out of poverty? The table
306
307
The data in Table
6.21 show that,
the wage
structure at the
bottom of the
earnings
distribution
shifted in such
a way as to
increase the
percentage of
workers earning
poverty level
wages.
�TABLE 6.22
Wage Trends Relevant to the Poor and Near-Poor
Male Workers
Percent Changes
by Tenths
Female Workers
1973-79
1973-79
1979-89
0.3%
-1.7
2.1
6.1
Lowest Tenth
2nd Tenth
3rd Tenth
4th Tenth
1979-89
-17.5%
-16.0
-11.2
-12.8
10.0%
0.5
2.9
-1.7
-17.8%
-6.6
-5.4
1.8
TABLE 6.23
Work Experience of the Poor, 1979-1989
1979
-1.2
-2.9
-0.3
2.0
66.1
41.1
25.0
1.2
2.4
-1.2
100.0
29.9
20.9
Total
20,474
33.9%
12.9
9.7
11.2
100.0
Percent Employable
Worked
Did Not Work
16,907
35.1%
15.8
10.0
9.2
64.9
38.7
26.2
Number of Poor Aged 15 + (thousands)
Percent Not Employable*
III or Disabled
Going to School
Retired
1989
Point
Change
1979-89
34.1
22.4
Percent of Poor Workers Who Worked
Year-Round**
Full-Time. Year-Round
4.2
1.6
answers this question for 1979 and 1989 In addition, the
table breaks out the share of workers earning up to 0.75 of
the poverty line.
The data i n Table 6.21 show that, excepting white and
Hispanic females, the wage structure at the bottom of the
earnings distribution shifted i n such a way as to increase
the percentage o f workers earning poverty level wages. For
black men i n 1979, 25.2% earned poverty level wages. By
1989, this percentage had increased to 33 3%, an increase
of 8.1 points. For Hispanic men, the increase was even
larger ( 11.7 points).
The deterioration i n the wage has meant not only a
g r o w t h i n low-wage earners but also a g r o w i n g proportion
of workers w i t h very l o w earnings. This can be seen i n
Table 6.21 i n the increased share of workers who, even if
they worked full-time and full-year w o u l d not be able to
earn 75% of the poverty line. The proportion of these very
low earners rose between 1979 and 1989 in every demographic category, w i t h black females posting the largest
increase (16.1 percentage points). However, also i n all
cases, the percentage o f workers earning between 0.75 of
poverty and the poverty line decreased, suggesting a
strong d o w n w a r d shift at the very bottom of the earnings
distribution for all men and women.
While Table 6.21 shows the proportion o f low-wage
workers, Table 6.22 examines percentage changes i n wage
levels of males and females i n the bottom four-tenths of the
hourly earnings distribution. For low-wage men between
1973 and 1979, there was little change i n the bottom
three-tenths, and a 6 . 1 % gain i n the fourth. For low-wage
women, this was a period o f g r o w t h i n the lowest tenth.
However, this pattern of gains or small losses reversed over
the period of 1979-1989. A l l groups, excepting woman at
the fourth tenth, experienced falling wages i n recent years.
In addition, the wage loss was greatest for the lowest earners, further evidence of a d o w n shifting of the earnings
distribution.
Work Effort
""Employable" poor are those who are neither retired, ill, disabled, nor in school. Those who are keeping
house are in the employable group.
"At least 50 weeks.
308
The increase i n the prevalence of l o w wages clearly
increases the likelihood of poverty. However, it is possible
that employment possibilities—getting a job w i t h sufficient hours of w o r k — c o u l d have offset the effect of lower
309
All groups,
excepting woman
at fourth tenth,
experienced
falling wages in
recent years. In
addition, the wage
loss was greatest
for the lowest
earners, further
evidence of a
down shifting
the earnings
distribution.
�TABLE 6.24
Percentage of Half-Time and Full-Time "Worker
Equivalents" in Poor Families with Children, 1979-1989
At Least 1
Half-Time
Worker
Equivalent*
Families Wilh Children
All Families with Children
1979
1989
At Least 1
Full-Time
Worker
Equivalent
42.3%
44.0
26.4%
28.1
1.7
1.7
24.7%
27.8
10.0%
12.7
3.1
2.8
69.8%
72.7
52.4%
55.2
2.9
2.8
Point Change
1979-89
Female-Headed Families with Children
1979
1989
Point Change
1979-89
Married Couple Families with Children
1979
1989
Point Change
1979-89
'A "full-time worker equivalent" is defined as family labor supply between 1,750 and 2,080 hours worked
during the year.
310
wages. In this regard, it could be argued that the poor did
not benefit from the expansion not because of low wages,
but simply because they chose to work less. The following
tables challenge that argument by showing that, with the
exception of average annual hours worked by single mothers with children, the amount that poor (and near-poor)
people worked grew over the decade. Poverty rose despite
this extra work effort.
Table 6.23 begins the analysis of work effort by examining the work experience of poor in 1979 and 1989. A
slightly larger percentage of the poor were employable in
1989 than in 1979(66.1% versus 64.9% ). But what is more
central to the argument in question is the following: a
larger percentage of the employable poor worked in 1989
than in 1979, and of those who worked, a larger percentage
worked year round and full time.
The table considers only those poor persons age 15 and
older, and examines the labor market participation of those
who were employable (neither retired, ill, disabled, nor in
school). In 1989, 41.1% of these poor persons worked, a
2.4 percentage point gain over the 38.7% of the employable poor who worked in 1979. There was also an increase
in year-round work effort by the poor, as the percentage of
poor workers working at least 50 weeks of the year
increased by 4.2 points. The percentage of the poor working full-time/full-year also grew slightly (1.6 points).
In Table 6.24, the percentage of poor families with the
equivalent of at least a half- or full-time worker is shown for
1979 and 1989. For example, when members of all poor
families with children pooled their working hours in 1979,
42.3% of those families had the equivalent of at least one
half-time worker (full-time labor supply is defined as
between 1,750 and 2,080 hours of work per year). Once
again, work effort is seen to have increased between 1979
and 1989 in each category.
Poor married couples with children have the highest percentage of half- or full-time worker equivalents in both
years, a finding to be expected since they typically have
the most potential workers per family. In 1989, close to
three- quarters of these families had at least one half-time
equivalent worker. Conversely, female-headed families with
children have fewer potential workers (in addition to child
311
The amount of
work performed
by members of
poor families
with children
increased slightly
over the decade.
�TABLE 6.25
Hours Worked b Family Type in the
y
B to T o Fifths*, 1979-1989
ot m w
Annual Hours Worked
Lowest
Fifth
All Families
1979
1989
820
850
Second
Fifth
1,800
1,840
Percent change
3.7%
2.2%
280
260
570
560
-7.1%
1979-89
-1.8%
Female-Headed Families
With Children
1979
1989
Percent change
1979-89
Married Couples
With Children
1979
1989
2,240
2,350
2,950
3,250
Percent change
1979-89
4.9%
10.2%
Unrelated Individuals
(Non-Elderly)
1979
1989
590
640
1,350
1,540
Percent change
1979-89
8.5%
'Fifths are defined separately for each family type.
312
14.1%
care constraints, see Chapter 8 ) ; they have the smallest percentage of half- and full-time worker equivalents i n both
years. However, w i t h respect to the argument noted above,
the table shows that the amount of w o r k performed by
members o f poor families w i t h children increased slightly
over the decade.
Table 6.25 looks at average annual hours of work among
low- income families w i t h i n different family types: femaleheaded families w i t h children, married couples, and unrelated individuals. Whereas the previous table included only
the poor, Table 6.25 examines w o r k effort by those i n the
b o t t o m two-fifths of the income distribution, w h i c h
includes both the poor and near-poor.
As w i t h the previous table, the average hours of work i n
Table 6.25 represent the pooled w o r k effort of the family.
The table shows that, w i t h the exception of female-headed
families w i t h children, average annual hours of work grew
for ail family types between 1979 and 1989. However, even
i n light of extra w o r k effort, the wage decline constrained
these families from realizing any economic gains. W i t h the
exception of female family heads, they were r u n n i n g faster
yet losing ground
The largest g r o w t h i n percentage terms was for unrelated individuals. Those at the bottom of the income distribution increased their average hours of work by 8.5% over
the decade; those individuals i n the second q u i n t i l e
increased their average hours by 14.1%. Married couples
w i t h children have by far the most hours of work; in each
year these families supplied an average of w e l l over 2,000
hours of labor By 1989, the average married-couple family
w i t h children in the second quintile worked 3,250 hours,
10.2% more hours then i n 1979. Mother-only families w i t h
children consistently show a relatively small number of
hours of work, and their supply fell over the time period.
However, their annual change i n hours was quite small ( 2 0
hours in the bottom f i f t h , 10 hours i n the second f i f t h ) , and
their wages can be assumed to be low. Therefore, the magnitude of their annual income loss due to reduced hours
was negligible.
313
Even in light of
extra work effort,
the wage decline
constrained these
families from
realizing any
economic gains
. . . they were
running faster yet
losing ground.
�Chapter 7-
Conclusion: What Does the Evidence
Show?
This chapter began w i t h the observation that, based on
the movement of the economic variables usually correlated
w i t h poverty rates, poverty should have fallen after 1983.
Yet a wide array of evidence shows that the expected fall i n
poverty did not maleriali/.e. Poverty rates stayed high
simultaneous throughout the decade, whether measured by the ofticial
effects occurredthresholds or by alternative, more encompassing measures.
I urthermore. demographic
which kept family types unresponsive to explanationsr o(the) ,increase of
economic g w t h while pospoverty rates high
sibly relevant in earlier decades, are insufficient to explain
' the decade: the stuhhornly high rates. Neither is lack of work effort a
(1) the wages of sufficient explanation; the work effort of the poor and nearthose at the poor increased over the 1980s.
bottom ot the The conclusion we are left w i t h is that t w o simultaneous
income effects occurred w h i c h kept poverty rates high over the
distribution felldecade: ( I ) the wages of those at the bottom of the income
distribution fell precipitously, and ( Z ) the system of tax and
precipitously, and
(2) the system of transfers designed to mitigate such a failure of the market
fax and transfers was less effective than i n the past. In other words, the pri
was less effectivemary distribution of market income was such that more
individuals) were unable to
than in the past. families (as well xs unrelated poverty level through the
achieve incomes above the
wages they earned over the l9H0s This development nee
essarily puts additional pressure on the secondary distribution, i.e., taxes and transfers, w h i c h must devote more
resources to close a wider poverty gap (see Table 6.S).
However, as shown i n lahle 6.19, the secondary distrihution was in fact less effective at reducing poverty than in
the past
31 i
Regional Analysis: The
Differences in Growth by
Region and Residence
In this chapter, we shift our focus from the national to
the regional perspective, and examine diftcrcnccs in economic well-being at various state antl urban/suburban/rural
levels. O u r analysis follows the Census Bureau's d c l i n i t i o m
of geographical areas, which divide the country into lour
regions, nine divisions, antl, o f course, fifty states antl
Washington, D.C. The last section of the chapter analyses
urban/rural differences
We find that economic growth was t|uite uneven o n a
regional basis over the 1980N. T he Northeastern region,
comprising New Hngland and the Mitl-Atlantic divisions,
did comparatively well in terms of greater income and
wage g r o w t h , and lower unemployment. However, the
Northeast was the region most adversely cflccicd by the
economic d o w n t u r n that began in 1989, as indicaictl by
rises i n b o t h u n e m p l o y m e n t ( 2 . 8 percentage p o i n t s .
1989-1991) and poverty rates (2 2 points, 1989 1991) that
were the highest among all regions. Conversely, the l9H()s'
economy in the Midwestern antl the Western regions was
characterized by relatively Hat or negative income growth
over the 1980s, rising poverty, anil signilicanl wage loss
The South experienced somewhat more mixed results over
the l9K()s, but ditl sustain high unemployment antl poverty rates, particularly for children.
3 IT
Economic growth
was quite uneven
on a regional
basis over the
1980s.
�Income Inequality
TABLE 7.1
Median Family Income by Region, 1973-1991
(1991 Dollars)
Median Family Income
Northeast
1973
1979
1989
1991
$37,080
37,956
43,369
40.265
Midwest
$37,025
37,862
38,018
36,759
South
West
$30,665
32,301
33,500
31,940
$35,983
38,079
39,210
37,171
Percent Change
2.4%
14.3
17.0
-7.2
1973-1979
1979-1989
1973-1989
1989-1991
2.3%
0.4
2.7
-3.3
5.3%
3.7
9.2
-4.7
5.8%
3.0
9.0
-5.2
(
Chapter 1 documents the dramatic rise in income inequality over the course of the 1980s. In that chapter we
note that the share of total family income going to the top
20% of families grew by 5 3 percentage points between
1977 and 1989 Meanwhile the bottom 90% of families lost
income shares. In this section, we present a state antl
regional analysis of growing income inequality. We lintl
that inequality grew both within regions, as the tlistance
between rich and poor expanded, and between regions, as
the Northeast generally prospered more than the other
regions.
We begin by examining changes in the levels of income
by region and state. Table 7.1 presents the income of the
median family, 1973-1990 (1990 data is the most recent
available), by region, and gives a first impression of the
uneven nature of regional growth. The Northeast region
had the highest levels of both income and economic
growth throughout the peritKl; the Mitlwest saw the least
growth. Median income in the Northeast grew 1-1.3%
between 1979 and 1989, from 137,956 to $-i3,369 (1991
dollars). In the Midwest, however, median income growth
was flat over the 1980s; total growth was only 0 4%. The
South and the West also experienced slow growth between
1979 and 1989, 3 7% and 3 0% respectively, compared to
their levels of growth in the 1973-1979 peritKl, when the
median family income grew the fastest in those regions.
Although the Northeast experienced the most growth
between 1979 and 1989, it was "hit" the hardest during the
recent recession, a theme which resurfaces throughout this
analysis. Between 1989 and 1991, median lamily income
fell slightly in the Midwest (3 3%), and fell more steeply
in the South and West (-1.7% and 5 2%, respectively) Yet
the median family in the Northeast saw a dramatic 7.2%
drop in its income, the biggest regional loss While the
Northeast still had the highest median family income and
the most growth over the full period, this two year period
(I9H9-199I) took S3.104 off the median income
In Table 7.2 we look more closely at the trentl in
median family income by state, for liour-person families
(These tlata are only available for four-person families for
the years 197-4-1989.) Once again, the dramatic rise of the
median family's income in the Northeast is evident.
316
317
Although the
Northeast
experienced the
most growth
between 1979
and 1989, it was
"hit " the hardest
during the recent
recession.
�TABLE 7.2
Median Income for Four-Person Families by State, 1974-1989
(1991 Dollars)
Median Income for
Four-Person Families
Region/
Division/State
1974
Northeast
New England
Maine
New Hampshire
Vermont
Massachusetts
Rhode Island
Connecticut
Mid-Atlantic
New York
New Jersey
Pennsylvania
1979
1989
TABLE 7.2 (continued)
Median Income tor
Four-Person Families
Annual
Growth Rates
1974-79
1979-89
$32,940
36,703
34,496
41,017
37,800
43,238
$42,108
52,704
44,372
56,895
47,536
58,558
0.2%
2.3
0.6
1.3
1.0
0.8
2.4%
2.5
2.2
2.6
1.8
2.6
$39,806
43,896
38,023
$38,802
45,351
41,070
$47,992
58,466
44,379
-0.5%
0.7
1.5
2.1%
2.5
0.8
$39,682
37,994
42,907
42,445
40,409
$41,464
41,622
44,661
44,950
43,286
$45,549
41,959
46,801
47,038
44,547
0.9%
1.8
0.8
1.1
1.4
0.9%
0.1
0.5
0.5
0.3
$41,443
37,713
36,136
39,377
33,654
35,071
37,776
Midwest
East North Central
Ohio
Indiana
Illinois
Michigan
Wisconsin
West North Central
Minnesota
Iowa
Missouri
North Dakota
South Dakota
Nebraska
Kansas
$33,266
41,108
35,548
43,779
39,822
44,928
$44,926
41,535
39,192
35,927
35,355
38,189
42,053
$46,533
40,350
42,264
38,230
36,059
41,631
41,671
1.6%
1.9
1.6
-1.8
1.0
1.7
2.1
0.4%
-0.3
0.8
0.6
0.2
0.9
-0.1
Region/
Division/State
South
South Atlantic
Delaware
Maryland
District of
Columbia
Virginia
West Virginia
North Carolina
South Carolina
Georgia
Florida
East South Central
Kentucky
Tennessee
Alabama
Mississippi
West South Central
Arkansas
Louisiana
Oklahoma
Texas
Annual
Growth Rates
1974-79
1979-89
-0.5%
0.8
1.9%
1.9
-0.2
1.3
1.0
0.9
1.6
2.0
-0.3
1.3
1.6
0.1
1.5
0.7
1.0
0.7
1974
1979
1989
$39,970
43,694
$38,990
45,436
$47,000
55,079
39,608
39,705
32,985
34,596
34,260
35,863
38,808
39,222
42,288
34,742
36,163
37,094
39,715
38,204
44,563
49,526
34,941
41,813
39,666
43,956
41,079
$32,840
33,559
33,604
30,342
$35,224
35,775
34,258
32,526
$37,774
38,314
38,367
35,478
1.4%
1.3
0.4
1.4
0.7%
0.7
1.1
0.9
$31,203
33,066
33,184
36,540
$34,037
37,116
38,379
43,098
$34,987
37,791
37,861
38,419
1.7%
2.3
2.9
3.3
0.3%
0.2
-0.1
-1.1
$35,916
36,937
38,926
41,015
31,867
39,968
36,748
40,301
$36,905
37,600
41,731
46,433
38,710
42,332
39,111
46,855
$37,216
36,942
39,125
44,227
34,221
42,120
40,159
43,647
2.8%
1.8
1.4
2.5
3.9
1.1
3.0
1.0
0.8%
-1.8
-0.6
-0.5
-1.2
-0.1
0.3
-0.7
$40,416
39,398
41,807
50,827
44,794
$44,928
44,230
46,214
57,125
45,244
$45,833
42,533
47,025
53,174
49,414
2.1%
2.3
2.0
2.3
0.2
2.0%
-0.4
0.2
-0.7
0.9
Total U.S.
318
West
Mountain
Montana
Idaho
Wyoming
Colorado
New Mexico
Arizona
Utah
Nevada
Pacific
Washington
Oregon
California
Alaska
Hawaii
$38,700
$41,219
$44,774
1.3%
0.8%
319
�TABLE 7.3
Percent o Families* with Low, Middle,
f
and High Relative Income,
b Region, 1969-1989
y
Point Change
Region/
Income Group**
1969
1979
1989
1969-1979
1979-1989
27.6%
60.0
12.4
100.0
31.8%
55.8
12.4
100.0
29.5%
50.9
19.6
100.0
4.2
-4.2
0.0
0.0
-23
-4.9
7.2
0.0
28.7%
60.3
11.0
100.0
30.7%
57.0
12.3
100.0
34.4%
52.6
13.0
100.0
2.0
-3.3
1.3
0.0
3.7
-4.4
0.7
0.0
42.7%
49.0
8.3
100.0
40.4%
49.7
9.9
100.0
41.7%
46.2
12.1
100.0
-2.3
0.7
1.6
0.0
1.3
-3.5
2.2
0.0
29.5%
57.1
13.4
100.0
32.6%
53.0
14.4
100.0
35.5%
48.7
15.8
100.0
3.1
-4.1
1.0
0.0
2.9
-4.3
1.4
0.0
Northeast
Low
Middle
High
Total
Midwest
Low
Middle
High
Total
South
Low
Middle
High
Total
West
Low
Middle
High
Total
•Income adjusted tor family size. Individuals are considered one-person families.
"Low-income group refers to 0-0.75 of the median; middle income group refers to 0.75-2.00 of the
median; high income group refers to 2.00 and above times the median.
320
Between 1979 and 1989, the median income for a fourperson family rose at an annual rate o f 2.6% in Connecticut and Massachusetts, and 2.5% in New Jersey. Compare
this annual level o f growth to that o f the same states for the
period 1974 to 1979: 0.8%, 1.3%, and 0.7%, respecdvely.
Conversely, the median four-person family income in
many Western states saw l o w o r even negative growth in
the 1980s, after sustaining relatively high rates of growth in
the earlier period. California, for example, saw median
income grow at an annual rate o f 2.0%, 1974-1979; however, over the 1980s, income grew at a much slower 0.2%.
Many other Western states (eight out of 13) saw negative
growth. New Mexico, which had the highest annual growth
rate between 1974 and 1979 (3 9%), had the second largest negative growth between 1979 and 1989 (-1.2%). Similarly, annual growth rates were relatively flat or negative in
the Midwest, 1979-1989.
Table 7.3 sorts families i n t o low-, middle-, o r highincome classes based o n their distance from the median
(adjusted for family size). The findings show that the middle class contracted in all regions between 1969 and 1989
(with the one exception o f the South, 1969-1979, where it
grew slighdy) and shrunk faster in the more recent decade
(Figure 7A). With the reduction o f the middle class we
must examine whether middle-class families rose out o f
the middle class o r fell into the lower-income class.
Families in the Northeast region appear to have made
impressive gains over the 1980s, although as we show
below, income inequality increased in all Northeastern
states (in other words b o t h low- and high-income families
gained income shares, but the wealthy gained more). The
middle class in the Northeast grew smaller in both periods
(Figure 7A), but between 1969 and 1979 the shift was
exclusively d o w n w a r d into the low-income class. Between
1979 and 1989, 7.2% o f Northeastern families went from
the low- or middle-income class to the high-income class,
which contained 19 6% o f Northeastern families in 1989.
The Midwest d i d relatively poorly over the 1980s, both
compared to the other regions and to its o w n performance
over the prior decade. O f the 4.4% o f Midwestern families
that left the middle class between 1979 and 1989, most
(3.7% out o f 4.4%) fell into the lower class.
321
The middle class
contracted in all
regions between
1969 and 1989
and shrunk faster
in the more recent
decade.
�FIGURE 7A
Percentage Point Change in Middle Class by Region,
1969-1989
Northeast
Midwest
Note: Middle class refers to income 0.75-2.00 of the median.
322
South
West
I n the South and West, b o t h the lower- and higherincome classes grew at the expense of the middle class
over the 1980s. I n all three years, the South had the largest
l o w - i n c o m e class and the smallest m i d d l e - a n d h i g h income classes o f the four regions. O f the 4.3% o f Western
families to leave the middle class, 1979-1989, most (2.9%
o f families) fell to the lower class.
Shifting o u r focus to the degree o f income inequality
w i t h i n individual states, we find that income became more
unequally distributed over the 1980s in 43 states. Table
7.4 presents the ratio o f the average income o f families in
the t o p fifth o f the income d i s t r i b u t i o n to the average
income o f families i n the bottom fifth, by state. The growth
o f this ratio over time is an indicator o f greater income inequality. The years o f comparison are 1979 and 1987-1989
(data for the e n d period are pooled over three years to
enhance their reliability; such pooled data are unavailable
for the late 1970s).
I n 1979, 34 o u t o f 50 states had inequality ratios o f
between 6.0 and 8.0; only one state, Alaska, had a ratio
greater than ten. However, the ratio grew over the 1980s in
all but seven states (Figure 7 B ) , as shown in the point
change c o l u m n o f Table 7.4. In the Nonheast region, New
York saw a relatively large g r o w t h i n inequality o f 2.6
points. Furthermore, the ratio o f average income o f the top
fifth to the b o t t o m in New York was among the highest, at
10.7 at the end o f the 1980s. Many Southern states also
saw high and growing levels o f inequality. O f the 16 states
i n the South, by the end o f the 1980s, eight had inequality
ratios o f at least 10.0. In fact, outside o f the South only
New York and New Mexico had ratios greater than ten.
323
Income became
more unequally
distributed over
the 1980s in 43
states. The ratio
of average
income of the top
fifth to the bottom
in New York was
among the
highest, at
10.7at the
end of the
1980s.
�TABLE 7.4 (continued)
TABLE 7.4
Ratio of Average Income ot the Top Fifth to
Average Income of Bottom Fifth, by State
1987-89
Point
Change
Northeast
New England
Maine
New Hampshire
Vermont
Massachusetts
Rhode Island
Connecticut
6.3
5.7
6.8
7.2
6.5
6.3
7.8
6.2
7.7
8.6
7.6
6.4
1.5
0.5
0.9
1.4
1.1
0.1
Mid-Atlantic
New York
New Jersey
Pennsylvania
8.1
7.2
6.7
10.7
8.2
7.7
2.6
1.0
1.0
Midwest
East North Central
Ohio
Indiana
Illinois
Michigan
Wisconsin
6.8
5.8
7.6
7.0
6.4
8.5
7.9
9.8
8.8
7.0
1.7
2.1
2.2
1.8
0.6
West North Central
Minnesota
Iowa
Missouri
North Dakota
South Dakota
Nebraska
Kansas
7.0
6.2
7.5
8.4
7.7
7.5
6.4
7.7
7.2
8.5
8.0
8.3
7.2
7.1
0.7
1.0
1.0
-0.4
0.6
-0.3
0.7
324
1987-89
Point
Change
6.9
7.2
7.2
6.7
7.2
9.4
8.0
7.6
6.8
7.5
10.4
10.8
8.8
9.2
10.0
9.2
-0.1
0.3
3.2
4.1
1.6
-0.2
2.0
1.6
7.1
8.2
9.4
9.5
9.2
9.8
10.7
11.1
2.1
1.6
1.3
1.6
9.7
8.7
8.5
9.1
9.9
14.5
10.1
10.6
0.2
5.8
1.6
1.5
West
Mountain
Montana
Idaho
Wyoming
Colorado
New Mexico
Arizona
Utah
Nevada
8.8
7.2
6.1
6.4
8.8
7.2
6.2
6.2
8.0
7.4
7.0
9.3
10.8
9.1
6.3
6.6
-0.8
0.2
0.9
2.9
2.0
1.9
0.1
0.4
Pacific
Washington
Oregon
California
Alaska
Hawaii
1979
Region/Division/State
1979
7.6
6.7
7.8
10.1
7.5
7.1
7.5
9.8
9.4
8.1
-0.5
0.8
2.0
-0.7
0.6
Region/Division/State
South
South Atlantic
Delaware
Maryland
Virginia
West Virginia
North Carolina
South Carolina
Georgia
Florida
East South Central
Kentucky
Tennessee
Alabama
Mississippi
West South Central
Arkansas
Louisiana
Oklahoma
Texas
325
�FIGURE 7B
Distribution of Changes in States' Income Inequality,
1979-1989
Number of States with
1 pt. or more Increase
in Income Inequality
326
Number of States with
less than 1 pt. Increase
in Income Inequality
Number of States with
a Decrease in
Income Inequality
As shown in our analysis of national income trends
(Chapter 1), the panem of income growth over the 1980s
led to income gains for some families but losses for the
majority. Table 7.5 shows what the changes over the 1980s
have meant in terms of dollars for the average family in the
bottom, middle, and top fifth by state. In 28 states, the
average income of families in the lowest filth fell over the
1980s. Low-income families in the Midwest, particularly
the East North Central division, saw significant income
losses, from $889 in Wisconsin to 12,519 in Indiana. Lowincome Southern families also posted substantial losses;
the average family in the bottom fifth in West Virginia had
$3,381 less in the late 1980s than the average low-income
family in that sute in 1979On the other end of the income spectrum, certain
regions experienced dramatic income gains over the
1980s. In fact, the average income of the top fifth fell in
only six states over the period, and four of those six were
in the West (Mountain division). In the Northeast, the average family in the top fifth ended the period with between
110,701 (Pennsylvania) and 127,323 (New Jersey) more
than the best-off families in 1979. Smaller but significant
gains were also made by upper-income Southern families,
particularly in Virginia, South Carolina, Florida and
Georgia.
327
In 28 states, the
average income
of families in the
lowest fifth fell
over the 1980s
while the average
income of the i
fifth fell in only
six states.
�TABLE 7.5
Income Changes for Average Family, Bottom, Middle,
and Top Fifth, 1979 to Late 1980s
TABLE 7.5 (continued)
Bottom
Fifth
Middle
Fifth
Top
Fifth
$1,495
134
-1,578
-3,381
-591
2,033
-372
291
$2,942
5,038
5,122
-2,789
2,213
5,431
2,261
3,943
$8,700
5,560
23,429
3,914
11,328
17,111
16,688
18,021
$-1,798
-217
116
-814
$-1,856
145
1,059
-2,839
$4,910
11,633
10,598
3,140
$247
-2,617
-661
-1,027
$-135
48
-392
-1,331
$3,972
13,008
7,592
3,742
West
Mountain
Montana
Idaho
Wyoming
Colorado
New Mexico
Arizona
Utah
Nevada
$666
286
- 2,086
-4,766
-885
-987
655
-1,071
$-3,331
-2,060
-1,498
-4,923
-3,581
2,437
628
-2,789
$-1,591
4,385
-666
-1,596
7,716
14,398
5,811
-1,130
Pacific
Washington
Oregon
California
Alaska
Hawaii
$1,490
-127
-755
1,100
1,490
$1,261
898
1,925
-3,153
4,830
$4,659
8,836
15,697
2,214
19,206
Region/Division/State
Bottom
Fifth
Region/Division/State
Northeast
New England
Maine
New Hampshire
Vermont
Massachusetts
Rhode Island
Connecticut
Mid-Atlantic
New York
New Jersey
Pennsylvania
Middle
Fifth
Top
Fifth
$5,460
9,485
6,528
8,907
4,511
9,539
$18,813
21,024
21,096
24,220
15,390
24,096
$-858
1,860
-107
$4,037
7,898
-546
$19,735
27,323
10,701
Midwest
East North Central
Ohio
Indiana
Illinois
Michigan
Wisconsin
$-1,413
-2,519
-1,679
-1,483
-889
$418
-3,161
-250
-1,111
870
$8,181
7,206
10,276
9,399
1,045
West North Central
Minnesota
Iowa
Missouri
North Dakota
South Dakota
Nebraska
Kansas
$-56
-1,999
-33
334
84
1,174
-140
$3,957
-2,348
336
1,427
1,202
-1,060
906
$8,917
-1,249
9,872
-883
6,478
6,269
6,175
$416
2,268
1,463
787
224
3,675
328
South
South Atlantic
Delaware
Maryland
Virginia
West Virginia
North Carolina
South Carolina
Georgia
Florida
East South Central
Kentucky
Tennessee
Alabama
Mississippi
West South Central
Arkansas
Louisiana
Oklahoma
Texas
329
�Poverty
TABLE 7.6
Poverty Rates b Region and Division, 1980-1991
y
Point Change
1980-89
1989-91
Region/Division
1980
1989
1991
Northeast
New England
Mid-Atlantic
10.9%
9.5
11.4
10.0%
7.2
11.0
12.2%
10.4
12.8
-0.9
-2.3
-0.4
2.2
3.2
1.8
Midwest
East North Central
West North Central
11.2%
11.2
11.1
11.9%
11.9
11.7
13.2%
13.5
12.6
0.7
0.7
0.6
1.3
1.6
0.9
South
South Atlantic
East South Central
West South Central
16.3%
14.4
20.3
16.8
15.4%
12.7
18.6
17.8
16.0%
14.2
18.6
17.6
-0.9
- 1.7
-1.7
1.0
0.6
1.5
0.0
-0.2
West
Mountain
Pacific
11.4%
12.0
11.2
12.5%
13.1
12.3
14.3%
13.9
14.5
1.1
1.1
1.1
1.8
0.8
2.2
Total U.S.
13.0%
12.8%
14.2%
-0.2
1.4
330
Although national poverty levels were about the same in
1980 and 1989 (the fact that poverty rates ditl nol fall in
response to the economic expansion over the 198()s is
explored in Chapter 6), there were some notable regional
differences, particularly involving child poverty. Table 7.6
shows that in the Northeast, poverty fell 0.9 percentage
points between 1980 and 1989 (data for o u r usual comparative peak year, 1979, are not available by division). However, given the dramatic rise in median family income in
the Nonheast (see Table 7.1), poverty rates might have
been expected to show an even greater decline. This lack
o f response to g r o w t h at the median is driven by the
h e i g h t e n e d i n c o m e i n e q u a l i t y in the Northeast d o c u mented in Tables 7.4 and 7.5. Particularly in the Mitl-Atlantic states, w h e r e poverty fell o n l y 0.4 p o i n t s
( 1 8 8 0 - 1 9 8 9 ) , those families i n the b o t t o m f i f t h fared
poorly, while those in the top fifth made significant gains.
In b o t h the Midwest and the West, poverty rates rosebetween 1980 and 1989, even while the national average
fell very slightly, further emphasising the uneven nature o f
the 1980s expansion.
Once again, the initial regional effects o f the recession
beginning in 1989 are notable, particularly in the Northeast. Table 7.6 shows that poverty grew at least twice as
much in the Nonheast (1.4 percentage points) as in the
other regions in 1990. At the divisional level, the New
England and Mid-Atlantic divisions show the most growth
in poverty rates, o f 2.0 and 1.2 points, respectively.
Child poverty rates (for persons under 18) grew over the
1980s (see Chapter 6), and, as presented in Table 7.7,
show dramatic regional differences. Child poverty rates arcgenerally higher for children in the South and are consistently higher for black children in all regions. However,
child poverty fell in the South over the 1970s, driven by a
10.9 percentage point d r o p for black children Over this
same period, child poverty grew in the other three regions.
Yet by 1987 (the latest year o f available data) this trend of
falling poverty rales tor children in the South had been
reversed, as poverty grew from 19 6% to 23.4% I x t w e e n
1979 and 1987. In fact, child poverty grew in all regit>ns
over this latter peritKl, w i l h the rates lor white antl black
children g r o w i n g by similar amounts, w i t h the notable
331
Poverty grew at
least twice as
much in the
Northeast
(1.4 percentage
points) as in
the other regions
in 1990.
�exception of black children in the Midwest, whose rales
grew steeply, by 15.3 percentage points.
TABLE 7.7
Child Poverty Rates by Race and Region, 1969-1987
Point Change
1979-1987
1969
1979
1987
10.9%
10.6
23.5
12.9
15.5%
13.0
19.6
14.2
16.9%
18.4
23.4
18.9
1.4
5.4
3.8
4.7
15.1%
16.0%
20.0%
4.0
8.4%
8.3
15.0
11.2
10.7%
96
12.7
10.8
13.5%
13.6
15.7
16.9
2.8
4.0
3.0
6.1
10.7%
11.0%
15.0%
4.0
31.4%
31.5
50.9
31.4
37.1%
36.2
40.0
29.4
39.4%
51.5
45.9
35.4
2.3
15.3
5.9
6.0
42.0%
Race/Region
37.8%
45.1%
7.3
All Races
Northeast
Midwest
South
West
Total U.S.
White
Northeast
Midwest
South
West
Total U.S.
Black
Northeast
Midwest
South
West
Total U.S.
332
Wages and Taxes
'I'he reasons for the recent growth in income inequality
are analyzed o n a national (and international, see Chapter
9) basis throughout this book While data do not exist on a
regional o r state basis for each area e x p l o r e d in the
national analysis, we do have data o n two imponant causal
factors of slow growth and growing inequality and poverty:
wages and taxes.
The wage analysis i n the next few tables shows the
highly uneven nature o f economic growth in the 1980s
from a regional perspective. Table 7.8 shows the percentage change, 1979-1991, in average hourly wages for wage
and salary workers in each region, by gender and education level. The first panel has data for 1979-1989 liven in
the prosperous Northeast, average hourly wages fell lor
most workers with u p to three years o f college Male workers in the other three regions saw the biggest losses, and
even college-educated men saw wage erosion in the Midwest (-4.4%) and the South (-0.5%), o r flat growth in the
West (1.6%). Wage loss was particularly dramatic for malehigh school dropouts in the Midwest and West, their wages
fell by 21.4% and 23.9%, respectively. Female workers saw
less wage e r o s i o n than males, 1979-1989, particularly
those with at least some college education. However, in
the Midwest, even female workers with some college saw
their wages fall 3 0%.
The second panel of Table 7.8 shows the effect o f the
recession o n regional hourly wages. (Recall from Chapter 3
that wage loss among the college educated predated the
recession.) As noted above, the recession was particularly
severe in the Nonheast, where it was felt not only hy lesseducated persons, but by highly educated workers xs well
A salient example is the case of college-educated men in
the Northeast, whose average hourly wage grew by 4.6%,
1979-1989, and fell by 3 6%, 1989-1991, leaving ii practically flat (0.9% growth) over the hill 1979-1991 period
(third panel, lable 7.8). In fact, the 4 4% growth in the
Northeast for college-educated w o r k e r s over the f u l l
period appears to have been driven primarily by femaleworkers. College-educated workers also saw significant
334
During 19791989, even in
Ihe prosperous
Northeast,
average hourly
wages fell for
most workers with
up to three years
of college and
even collegeeducated men
saw wage erosion
in the Midwest
and the South.
�TABLE 7.8 (continued)
TABLE 7.8
Percentage Change in Average Hourly Wages
by Gender, Education, and Region, 1979-1991
Year
Region/Gender
Less Than
High School
High School
College
1 3 Years
College
4 Years
Year
Region/Gender
College
Plus
2 Years
All
Less Than
High School
High School
College
1 -3 Years
College
4 Years
College
Plus
2 Years
All
1989-1991 (continued)
All
Male
Female
All
1979-1989
Northeast
Male
Female
All
Midwest
-9.9%
-4.7
-9.6
-5.0%
42
-2.1
Male
Female
All
South
-21.4%
-15.5
-21.7
Female
All
Female
All
All
Female
All
-3.8%
0.3
-2.1
1.3%
03
10
2.5%
08
15
0.4%
06
02
2.0°,
1 4
08
- 1.1%
11.2
1.1
4.6%
15.8
5.9
12.6%
186
11.1
2.9%
15.1
5.4
- 14 8%
-6 7
12 9
- 13.2%
-3.0
- 12.3
- 4 4%
9.4
- 1.6
82%
83
54
-9.1%
0.9
-7.8
Male
Female
All
Midwest
- 165%
-30
13.3
-7.1%
4.2
-3.5
- 2 9%
11 2
-0 1
0 9%
18 8
44
15 5%
184
128
1 6%
172
53
- 15.6%
-11.2
- 14.9
- 14 0%
-4.1
-11.0
-9.8%
4 5
6 8
- 0 5%
125
09
8.3%
13 1
71
- 5 5%
5.9
-3.0
Male
Female
All
South
- 2 7 1%
- 19.4
-26 6
- 17.7%
7.5
-15 1
- 14 6%
- 10
- 12 4
- 5 0%
89
30
8 6%
70
52
10 6%
16
87
- 23.9%
- 14.9
-21 1
-15.5%
-4.2
-11.9
- 6 6%
5.5
3.8
1.6%
13.4
2.3
9.2%
8.0
62
-7.4%
5.9
-4.2
Male
Female
All
West
- 19.5%
-7.5
-16 9
- 18 2%
2.7
- 12 9
- 10 5%
1.5
-8.4
- 3.0%
11.6
10
82%
14 4
6.7
7.9%
72
42
- 18 2%
-11.7
- 17.1
- 12 7%
-3.2
9 9
- 8 3%
38
-6.0
02%
126
1.8
9.7%
126
7.5
-5.2%
5.6
-2.8
Male
Female
All
All
- 28.5%
- 14.8
-24 9
- 19.1%
33
- 13.6
-7.7%
60
-39
- 1 6%
16.8
14
73%
108
5.5
9 5%
77
48
Male
Female
All
- 2 3 2%
-110
-20.5
- 16 1 %
29
118
- 9 5%
35
6 9
- 2 3%
136
03
10 2%
132
7 8
7 1%
70
36
1969-1991
Northeast
Male
Female
All
Midwest
- 6.2%
0.8
4.2
-7.4%
1.8
-4.0
-2.3%
0.1
-1.5
-18%
0.1
1.2
-3.6%
26
-1.5
2.6%
0.2
15
- 1.2%
1.8
-0.2
7.2%
-4.6
6.2
-3.4%
0.9
2 6
-1.6%
2.1
0.2
0 6%
0.5
1.4
0 4%
- 1.2
-0.1
-1.6%
08
-0.9
Male
Female
All
West
- 4 .6%
4.1
-2.3
- 4 9%
1.5
2.2
0.8%
2 9
18
-2.5%
08
19
0.0%
1.2
0.4
-2.6%
13
12
f
Female
All
-6.1%
01
47
-4.2%
1.0
-19
- 1.3%
0.4
0.1
3.1%
3.0
08
17%
2.6
06
1979-1991
Northeast
2.3%
1.8
0.7
Female
All
South
335
�FIGURE 7C
Percentage Change in Median Hourly Wages By State,
1979-1989
FIGURE 7D
Percentage Change in Median Hourly Wages By State,
1989-1991
336
wage loss in the other regions, denoting the end ol the
198()s "white-collar" boom and showing the white-collar
nature of this recession (see Chapter -i). In the South and
West, even workers with post-graduate educations saw a
slight decline overall between 1989 antl 1991
In Table 7.9 and Figures 7C antl 71). we shilt to state
level analysis, by gentler ( fable 7.9), antl examine the percentage change in the real median hourly wage (half ol .ill
workers earn more than this wage, half earn less). Figure
7C shows that real median hourly wages lell in most ( S^)
states. Four states saw median wage growth ol greater than
10%; 12 saw wage growth of between 0% antl 10%: 18
states saw their median hourly wage fall lx-tween 0% antl
10%; and 17 states saw wages fall by more than 10%. The
figure amply dramatizes the regional nature ot economic
growth over the 1980s, as only stales on the Fa.Mcrn seahoard (plus Nevada) saw growth in their median wageWages at the median fell in all other slates
The state data show that the large wage losses in the
Midwestern antl Western regions were mostly driven by
the steep decline in male wages ( liible 7.9). |or example,
men in Colorado antl (California saw their median wage fall
by 19.9% and I 1.6%, respectively, between 1979 antl 198'J:
for women in those states, wage growth was either essentially Hat (Colorado, -0 1%), or positive (California. 3 8'A,)
However, women in Wyoming antl Alaska tlitl experience
substantial wage loss. In the Northeast, wage growth was
quite robust for both men antl women between 1979 antl
1989, though women's wages grew faster than those ol
men's in most Northeastern slates (though the It-rc/ ol
women's wages remainetl well below that of men's)
There, metlian wages for women grew by as much as
21.9% (New Jersey), while the largest growth rate (or males
was 14.7% (New Hampshire).
Figure 7D shows that, with the rising unemployment
beginning in 1989, the metlian hourly wage fell in 3-i
states, with notable diHcrcnccs in the eastern states Wages
grew in 16 states and the District of Columbia, although
for 12 slates antl the District of Columbia, the growth was
less than 2.0%; 20 states experienced a mixlcst decline (0
to 2.0%); 14 states saw wages decline more lhan 2.0%.
The large wage
losses in Ihe
Midwestern and
Western regions
were mostly
driven by the
steep decline in
male wages.
�TABLE 7.9
Percent Change in the Median Hourly Wages of
Men and Women, by State, 1979-1991
TABLE 7.9 (continued)
Percent Change
Percent Change
1979-1989
Men
Women
Division/State
Division/State
1989-1991
Men
Women
New England
Maine
New Hampshire
Vermont
Massachusetts
Rhode Island
Connecticut
2.5%
14.7
5.4
9.8
-4.2
8.5
1.7%
-1.5
-2.6
-4.8
5.9
1.2
3.9%
2.8
-0.4
1.4
1.8
4.0
-0.8%
2.9
-14.4
12.1%
21.9
-0.6
-2.6%
-5.3
0.1
3.2%
1.0
3.3
-14.1%
-16.1
-12.5
-9.8
-14.7
-0.8%
-3.8
-3.1
-6.4
-5.0
-0.6%
0.6
-1.0
-3.6
-3.9
-0.2%
-0.6
0.3
-0.7
-0.7
-17.7%
-15.0
-18.0
-19.2
-16.5
-12.2
-9.3
5.3%
-5.5
1.6
-2.7
-2.8
-6.5
0.7
0.1%
-3.1
-3.1
-5.2
-3.7
-2.0
-5.3
2.6%
-2.0
-2.8
-2.5
3.7
3.8
0.9
-8.3%
-11.0
-15.9
2.1
Mid-Atlantic
New York
New Jersey
Pennsylvania
12.1%
16.4
18.6
21.8
11.1
15.8
8.2%
8.4
9.9
11.9
-6.5%
0.5
1.2
-7.4
2.0%
4.7
-6.3
1.0
East North Central
Ohio
Indiana
Illinois
Michigan
Wisconsin
West North Central
Minnesota
Iowa
Missouri
North Dakota
South Dakota
Nebraska
Kansas
South Atlantic
Delaware
Maryland
District of Columbia
Virginia
338
1989-1991
1979-1989
Women
Men
Men
Women
South Atlantic (continued)
West Virginia
North Carolina
South Carolina
Georgia
Florida
-24.8
-2.7
4.2
-2.2
1.0
-5.7
1.4
2.2
4.7
10.6
-2.6
1.8
1.7
-4.2
-8.0
-4.4
2.2
2.0
4.4
-1.7
East South Central
Kentucky
Tennessee
Alabama
Mississippi
-14.1%
-13.8
-13.0
-11.9
- 5.3%
-2.9
2.3
3.2
-3.7%
-3.9
-6.4
1.8
-3.6%
5.0
-3.6
-1.7
- 7.6%
-12.1
-15.3
-14.7
-1.5%
-4.5
-3.0
4.5
2.0%
-10.1
-1.8
-0.3
-5.0%
4.7
-0.4
-2.1
-21.2%
-21.5
-20.4
-19.9
-19.8
-13.6
-13.5
-8.2
- 6.6%
-3.0
-13.0
-0.1
-5.9
1.0
0.0
4.1
-3.9%
4.1
-0.7
4.7
4.1
-7.6
-8.7
-1.9
-1.8%
-2.6
1.6
-2.8
-1.4
7.3
-4.7
-2.9
-17.1%
-17.4
-11.6
-23.6
-5.5
- 2.8%
-0.8
3.8
-12.2
6.7
2.7%
-4.0
-2.1
-0.6
-1.0
2.7%
0.0
2.0
-3.5
2.4
-11.7%
5.4%
-1.4%
-2.8%
West South Central
Arkansas
Louisiana
Oklahoma
Texas
Mountain
Montana
Idaho
Wyoming
Colorado
New Mexico
Arizona
Utah
Nevada
Pacific
Washington
Oregon
California
Alaska
Hawaii
Total U.S.
339
�TABLE 7.10
State and Local Effective Tax Rates in the Ten Largest States as
Percent o Income for Families o Four, 1985-1991
f
f
Lowest
Fifth
1991
Middle
Fifth
1985
1991
Top
Top
15%
1985
1991
4%
1991
1985
Top
1%
1985
1991
Region/State
1985
New York (NE)
Pennsylvania (NE)
New Jersey (NE)
14.4% 14.1% 13.4% 13.9% 13.4% 13.8% 13.1% 13.0% 12.0% 11.3%
6.4
5.5
8.2
7.3
8.6
8.3
9.9
9.8
15.6
15.9
6.8
9.7
8.3
10.7
8.4
107
10.8
13.3
9.3
152
Illinois (MW)
Ohio (MW)
Michigan (MW)
13.8% 16.5% 9.4% 10.8% 8 . 1 % 9.0% 7.6%
9.3
9.6
9.6
134
95
10.0
12.3
11.1
106
10.5
120
11.4
15.0
143
North Carolina
(South)
Florida (South)
Texas (South)
10.2% 10.6%
10.0
138
12.4
17.1
86%
59
6.2
9.7%
7.6
8.4
8.5%
4.6
5.0
94%
5.5
6.4
8.4%
4.1
4.5
California (West)
11.8% 14.1%
84%
8.8%
9.1%
9 7%
9.9% 10.7% 10.5% 10.6%
7.7%
9.8
9.6
6.1%
9.7
7.7
6.0%
9.6
7.6
9.0%
4.4
5.1
8.1%
2.7
3.0
8.4%
27
3.1
Point Change
1985-1991
New York (NE)
Pennsylvania (NE)
New Jersey (NE)
-0.3
03
1.9
0.5
-0.1
1.5
04
-0.3
23
-0.1
-0.9
2.4
-0.7
-0.9
29
Illinois (MW)
Ohio(MW)
Michigan (MW)
2.7
1.1
-0.7
1.4
0.5
-0.6
0.9
0.3
-0.5
0.1
0.2
-0.9
-0.1
-0.1
-0.1
North Carolina
(South)
Florida (South)
Texas (South)
0.4
3.8
4.7
1.1
1.7
2.2
0.9
0.9
1.4
0.6
0.3
0.6
0.3
00
0.1
California (West)
23
0.4
0.6
0.8
0.1
340
Most o f the N e w England, Mid-Atlantic, and Southern
states that experienced wage growth prior to the d o w n t u r n
saw their wages fall, 1989-1991.
The wage turnaround in the Northeast over the recent
recession is also apparent in the state-level analysis of male
median wages, 1989-1991 (Table 7.9). Males in five o f the
nine Northeastern states saw falling wages over the period,
w i t h men in New Jersey and Massachusetts posting large
wage losses o f 5.3 and 4.8%, respectively. However, except
for Vermont, w o m e n in the Northeastern states made wage
gains o f between 1.0% and 4.0%. The wages of Midwestern
male workers continued to fall over the recession, and by
the end o f the full period (full period data, 1979-1991, are
not in the table), the median wage for males in every Midwestern state had fallen by between 13-1% (Michigan) and
23 4% ( N o r t h Dakota). Males in Western states experienced similar wage declines over the f u l l p e r i o d , w i t h
m e d i a n wages falling f r o m between 5.6% ( H a w a i i ) to
24.3% (Montana).
As noted in Chapter 2, state and local taxes tend to be
regressive, meaning that the poor pay a larger proportion
of their income in taxes than the wealthy. Table 7.10 gives
average effective tax rates in the ten most populated states
for families o f four in the bottom fifth; middle fifth, and,
top 15%, 4%, and 1%. F i g u r e 7E portrays the effective
rates for the five most populated states. The data show that
state and local taxes became even more regressive over the
latter part of the 1980s (distributional data o n the state and
local tax burden is only available for these years). In California, the largest state (and one o f the least regressive),
the average family in the lowest fifth paid 14.1% o f their
income in state and local taxes in 1991, while the most
wealthy families paid 10.6%. Furthermore, the poorest Califomian fiamilies saw their tax burden rise by 2.3% o f their
income from 1985 to 1991, while those in the top 1 % had
only an extra 0 . 1 % added to their burden.
Texas, Florida and Illinois also saw their already highly
regressive state and local effective rates become more
regressive (Table 7.10). In Texas in 1991, the poorest fourperson families paid 17.1% of their income in taxes, while
the most wealthy paid 3 . 1 % . Similarly, in Florida, lowincome families paid over five times the p r o p o r t i o n o f
income paid by families in the top 1%.
341
Stale and local
taxes became
even more
regressive over
the latter part of
the 1980s. In
California the
average family in
the lowest fifth
paid 14.1% of
their income in
state and local
taxes in 1991,
while the most
wealthy families
paid 10.8%.
�Employment
FIGURE 7E
State and Local Effective Tax Rates in the
Five Most Populated States,
1991
0)
E
o
o
c
a
in
(0
c
I
m
Lowest 20%
Second 20%
342
Middle 20%
Fourth 20%
Next 15%
Next 4%
Top 1%
Table 7.11 shows annual rates of job growth by region
and division, 1973-1990. With the exception of the Northeast, employment grew faster over the 1973-1979 business
cycle than over the 1980s. In the West, employment grew
one and one-half times faster, on an annual basis, between
1973 and 1979, than over the 1980s. However, even in
light of the slowdown in employment growth, the Western
region led the pack in the annual rate of job growth over
the 1980s (2.7%). The Southern region had the second
fastest growth, most due to the Atlantic division, where
employment grew at an annual rate of 3 0%. In the Northeast, the job rate grew 2.3% annually in New England, and
1.5% in the Middle Atlantic states from 1979-1989.
Rates of job growth can be decomposed into two factors: growth due to a rising population, and growth due to
an increase in the ratio of workers to the adult population.
By these measures, the job growth in the Nonheast and
the West over the 1979-1989 period was driven by quite
different factors. In Table 7.11, we find that of the 2.7%
rate of Western job growth, 2 .2 points, or about four-fifths
of the total growth, came from that region's population
growth, which was quite strong over the decade. Conversely, 70% of the Nonheastem rate of 1.7% was generated by sending a larger proponion of the adult population into the workforce. This is most evident at the
divisional level in the Mid-Atlantic, where of the 1.5%
annual rate of job growth, 1.1 points (73%) are attributable
to growth in the employment to population ratio. This
finding suggests that much of the recent strong growth in
family income in the Northeast (Table 7.1) is due to families sending more workers into the workforce, a theme
explored in Chapter 1.
Interestingly, the recession beginning in 1989 appears to
have primarily affected job growth in the Northeast, where
growth turned negative, and did not affect other regions.
While employment grew in New England at an annual rate
of 2.3% between 1979 and 1989, it shrank at that same
rate between 1989 and 1990 (the latest data available).
Employment also fell slightly in the Mid-Atlantic states.
Meanwhile, job growth actually accelerated in the Midwest,
growing at an annual rate of 1.5%, 0.4 percentage points
faster than the rate over the prior cycle (Table 7.11). In the
343
Of the 2.7% rate
of Western job
growth about
four-fifths came
from that region's
population
growth.
Conversely, 70%
of the Northeastern rate of
1.7% was
generated by
sending a larger
proportion of the
adult population
into the
workforce.
�TABLE 7.11
Job Growth by Region and Division, 1973-1990
Employment Growth Due to:
1979-1989
Annual Rate of Growth
of Employment
Region/Division
73-79
79-89
89-90
Growth of
Adult
Population
Growth of
Employment/
Adult
Population
Northeast
New England
Middle Atlantic
1.0%
1.9
0.6
1.7%
2.3
1.5
- 0.7%
-2.3
-0.1
0.5%
0.7
0.4
1.2%
1.6
1.1
Midwest
East North Central
West North Central
1.8%
1.6
2.4
1.1%
1.0
1.2
1.5%
1.4
1.5
0.4%
0.4
0.5
0.7%
0.6
0.7
South
South Atlantic
East South Central
West South Central
2.7%
2.4
2.0
3.8
2.3%
3.0
1.5
1.5
1.8%
1.4
1.4
2.7
1.9%
2.2
1.1
1.8
0.4%
0.8
0.4
-0.3
West
Mountain
Pacific
4.2%
4.6
4.1
2.7%
2.6
2.7
2.4%
2.8
2.3
2.2%
2.4
2.2
0.5%
0.2
0.5
Total U.S.
2.4%
1.9%
1.3%
1.2%
0.7%
West, job growth slowed in the Pacific division, but grew in
the Mountain division.
Using the official measure of unemployment (the ratio of
those seeking work to the total workforce), the U.S. rate
fell slightly (0.5 percentage points) from 1979 to 1989
(Table 7.12 and Figure I F ) . From a regional perspective,
most of this decline was driven by a 2.1 point fall in the
Northeast, which had the lowest 1989 unemployment rate,
4.5%. Unemployment fell significandy in all but one Northeastern state (New Hampshire). Conversely, unemployment grew slighdy in the South (0 3 points), mostly due to
rises in the West South Central division (plus Mississippi).
The two other regions of the country saw unemployment
rates fall slightly between 1979 and 1989, 0.1 points in the
Midwest and 0.7 points in the West.
However, the rise in unemployment beginning in 1989
more than wiped out the progress made in lowering
unemployment over the 1980s. Between 1989 and 1991,
unemployment rose dramatically in certain Northeastern
states. For example, in Massachusetts in 1989, unemployment, at 4.0%, was well below the national average of
5.3%, and even below the regional average of 4.5%. Yet in
two years, unemployment in that state climbed 5.0 percentage points, to 9 0%, well above the national average of
6.7%. Of course, unemployment grew elsewhere as well,
particularly in the South. Yet the Northeast, which started
the recession with the lowest unemployment rates, had the
country's highest rates by 1991. This is most evident in the
Northeastern division of New England, which went from a
1989 rate of 3 9% to a 1991 rate of 8.0% (Figure 7F).
Urban versus Rural
In this section, we examine geographic differences in
certain economic indicators between urban, suburban and
rural settings. We use the Census Bureau's definitions of
these areas. Cities are areas with at least 50,000 residents
that form an economic and population nucleus to the surrounding communities. A suburban area is an area close to
a city, but with a smaller population than the city. The suburban economy is highly integrated with that of the nearby
city. For some comparisons, it is useful to pool cities and
344
345
The Northeast,
which started the
recession with
the lowest
unemployment
rates, had the
country's highes
rates by 1991.
�TABLE 7.12 (continued)
TABLE 7.12
Unemployment Rates by Region, Division, and State, 1979-1991
Point Change
State/Division/Region
1979
1989
1991
1979-89
6.6%
5.5
7.2
3.1
5.3
5.5
6.7
5.1
4.5%
3.9
4.2
3.6
3.6
4.0
4.0
3.6
7.2%
8.0
7.6
7.1
6.4
9.0
8.6
6.8
-2.1
-1.6
-3.0
0.5
-1.7
-1.5
-2.7
-1.5
2.8
4.1
3.4
3.5
2.8
5.0
4.6
3.1
Middle Atlantic
New York
New Jersey
Pennsylvania
7.0%
7.1
6.9
6.9
4.7%
5.1
4.1
4.5
7.0%
7.2
6.6
6.9
-2.3
-2.0
-2.8
-2.4
2.3
2.1
2.5
2.4
Midwest
East North Central
Ohio
Indiana
Illinois
Michigan
Wisconsin
5.5%
6.1
5.9
6.4
5.5
7.8
4.5
5.4%
5.7
5.6
4.8
6.0
7.1
4.4
6.4%
7.0
6.4
5.9
7.1
9.2
5.4
-0.1
-0.4
-0.4
-1.7
0.5
-0.7
-0.1
1.0
1.3
0.8
1.2
1.2
2.1
1.0
West North Central
Minnesota
Iowa
Missouri
North Dakota
South Dakota
Nebraska
Kansas
4.0%
4.2
4.1
4.5
3.7
3.6
3.1
3.3
4.5%
4.4
4.3
5.5
4.2
4.2
3.1
4.0
5.0%
5.1
4.6
6.6
4.1
3.3
2.7
4.4
0.5
0.2
0.2
0.9
0.6
0.6
-0.1
0.7
0.5
0.7
0.3
1.1
-0.1
-0.8
-0.4
0.4
5.4%
5.7%
6.6%
5.5
8.2
5.9
7.3
4.7
4.8
3.6
3.7
5.1
3.9
6.4
6.0
5.9
7.8
5.8
South Atlantic
Delaware
Maryland
District of Columbia
Virginia
346
1979
1989
1991
1979-89
6.7
4.8
5.1
5.1
6.0
8.6
3.5
4.7
5.5
5.6
10.5
5.8
6.3
5.0
7.3
1.9
-1.3
-0.3
0.4
-0.4
1.9
2.3
1.5
-0.5
1.7
East South Central
Kentucky
Tennessee
Alabama
Mississipppi
6.1%
5.6
5.8
7.1
5.8
6.3%
6.2
5.1
7.0
7.7
7.3%
7.4
6.6
7.2
8.6
0.2
0.6
-0.7
-0.1
2.0
1.0
1.2
1.5
0.2
0.9
West South Central
Arkansas
Louisiana
Oklahoma
Texas
4.7%
6.2
6.7
3.4
4.2
6.8%
7.2
7.9
5.6
6.7
6.7%
7.3
7.1
6.7
6.6
2.1
1.0
1.2
2.1
2.5
-0.1
0.1
-0.9
1.1
-0.1
6.0%
5.3%
6.7%
-0.7
1.4
Mountain
Montana
Idaho
Wyoming
Colorado
New Mexico
Arizona
Utah
Nevada
5.0
5.1
5.6
3.1
4.8
6.6
5.0
4.3
5.0
5.5
5.9
5.1
6.3
5.8
6.7
5.3
4.7
5.0
5.6
6.9
6.2
5.0
5.0
7.0
5.6
5.0
5.4
0.5
0.8
-0.5
3.2
1.0
0.1
0.2
0.4
0.0
0.1
1.0
1.0
-1.3
-0.8
0.3
0.4
0.3
0.4
Pacific
Washington
Oregon
California
Alaska
Hawaii
6.4%
6.8
6.8
6.2
9.3
6.4
5.2%
6.2
5.7
5.1
6.7
2.7
7.1%
6.3
6.0
7.5
8.5
2.7
-1.2
-0.6
-1.1
-1.1
-2.6
-3.7
1.9
0.2
0.3
2.5
1.8
0.0
5.8%
5.3%
6.7%
-0.5
1.5
State/Division/Region
1 989-91
1989-91
Northeast
New England
Maine
New Hampshire
Vermont
Massachusetts
Rhode Island
Connecticut
South
Point Change
0.3
1.0
-0.7
-4.6
-2.3
-2.2
-0.8
1.6
2.5
2.2
2.7
2.0
South Atlantic (continued)
West Virginia
North Carolina
South Carolina
Georgia
Florida
West
Total U.S.
347
�FIGURE 7F
Unemployment Rates by Region,
1979-1991
their suburbs into one area, w h i c h we w i l l call "urban
areas." Rural areas are, by definition, the areas outside o f
the urban areas; their economies are less closely integrated
w i t h cities.
Table 7.13 presents data o n the proportion o f families
w i t h low, middle and high incomes for city, suburban, and
rural areas (similar to the analysis in Table 7 3). The rural
areas had the largest lower-income group and the smallest
upper-income g r o u p in each year. However, the lower-income class in cities has grown consistently since 1969; i n
both 1979 and 1989, it was only a few percentage points
smaller than its rural counterpart. The middle-class shrinkage in the nation as a whole (see Chapter 1) appears to
have taken place i n each residential category between
The middle-class
shrinkage in the
nation as a whole
has taken place in
each residential
category...
with cities
experiencing the
largest decline.
TABLE 7.13
City/Suburb/Rural Comparisons o
f
Income Distribution, 1969-1989
y
^
•
*P y
y
0
+*r
y
y
Point
Change
0
/
Residence/Income Group*
1979-89
1969
1979
1989
33.8%
54.9
11.3
100.0
39.6%
49.8
10.6
100.0
42.7%
44.5
12.8
100.0
3.1
-5.3
2.2
0.0
22.8%
62.7
14.5
100.0
25.1%
58.7
16.2
100.0
26.8%
53.7
19.5
100.0
1.7
-5.0
3.3
0.0
44.1%
49.8
6.1
100.0
41.8%
50.5
7.7
100.0
46.8%
46.3
6.9
100.0
5.0
-4.2
-0.8
0.0
City
Low
Middle
High
Total
Suburb
Low
Middle
High
Total
Rural
Low
Middle
High
Total
•Low-income group refers to 0-0.75 of the median; middle-income group refers to 0.75-2 00 of the
median; high-income group refers to 2.00 and above times the median.
348
349
�1979 and 1989, w i t h cities c x p c r i c i u i n g the largest
decline. However, in rural areas there occurred a major
shift downward in the distribution of income Hy 1989,
4.2% antl 0.8% of ihe rural population had shifted from the
middle- antl upper-income classes lo the lower
TABLE 7.14
Poverty Rates by City, Suburban, and
Rural Residence, 1973-1991
Year
City
1973
1979
1989
1991
14.0%
15.7
18.1
20.2
Suburban
6.4%
7.2
8.3
9.6
Rural
14.0%
13.8
15.7
16.1
City/Rural
Difference
0.0%
1.9
2.4
4.1
Point Change
1979-1989
1989-1991
2.4
2.1
1.1
1.3
1.9
0.4
FIGURE 7G
City, Suburban, and Rural Poverty Rates,
1973-1991
tr
I
Q_
0.5
1.7
A slightly different picture emerges when we compare
recent poverty trentls a m o n g city, suburban antl rural
areas. In 1973, poverty rates for persons in rural areas antl
in cities were equal, at 14% (Table 7.14), while suhuriian
rates were less than half as large, at 6.4% As shown in Figu r e 7G, between 1973 and 1979, poverty declined in rural
areas, and grew in cities antl suburbs. From 1979 o n , however, poverty grew in both rural antl urban areas, antl by
the end o f the recovery, in 1989, poverty hatl grown (since
1973) by 4.1 percentage points in cities antl by 1 7 points
in rural areas, ( lable 7.1-r). As a result, by 1989, 2.-1% more
of the city population was ptK)r relative lo the rural popul a t i o n . Suburban poverty grew the least over the
1979-1989 pcritxl (up I I percentage points), so that in
1989 the suburban poverty rale was well below the rates in
both cities (9 8 points lower) antl rural areas (7 i points
lower).
The economic d o w n t u r n beginning in 1989 brought
higher poverty rates to all types of residences. I'ov crty grew
the most in the cities (2 1 points), and grew slightly (O.-t
points) in rural areas. Suburban areas experienced dramatic
growth o f poverty over the recession, as poverty rates grew
hy more ( 1 3 points) between 1989 and 1991 lhan between
the ten year peritKl, 1979-1989. This growth pattern further
increased the distance between suburban rates and those o f
the other types of areas
For recent years, 1988-1991, we can examine poverty
rates by residence w i t h i n the lour regions, liible 7.15
shows that in the Northeast antl Mitlwest, poverty is highest in urban areas, while in the Souih, poverty is most common (i.e., proponionately) in rural arexs In 1989, over
onc-tifth of the persons living in Midwestern cities were
poor. A similar proportion o f ' r u r a l Soulherners (19.5%)
were pt>t>r in that same year.
c
<u
o
2.
Uy 1991, urban poverty in Northeastern cities grew
steeply, hy 3 3 percentage points, driven hy the ialling
incomes, d e c l i n i n g w ages, antl rising u n c m p l o ) nieni
experienced in lhat region as the recession began Although
other areas also saw rising poverty rates, 1989 1990, the
350
35 1
In 1973, poverty
rates for persons
in rural areas and
in cities were
equal at 14%...
By 1989, 2.4%
more ot the city
population was
poor relative to
the rural
population.
�TABLE 7.15
Poverty by Region and Residence, 1988-1991
Point
Change
1989-1991
Region/Residence
1988
1989
1991
Northeast
City
Suburban
Rural
10.1%
18.0
5.6
10.2
10.0%
18.0
53
10.9
12.2%
21.3
7.0
12.3
2.2
3.3
1.7
1.4
Midwest
City
Suburban
Rural
11.4%
19.8
6.0
11.3
11.9%
20.8
5.7
12.4
13.2%
23.1
7.4
12.4
1.3
2.3
1.7
0.0
South
City
Suburban
Rural
16.1%
19.7
10.8
20.6
15.4%
18.6
10.6
19.5
16.0%
20.0
11.2
19.8
0.6
1.4
0.6
0.3
West
City
Suburban
Rural
12.7%
14.6
10.0
16.5
12.5%
15.1
9.8
14.8
14.3%
16.9
12.2
15.6
1.8
1.8
2.4
0.8
TABLE 7.16
Urban and Rural Employment Growth, 1979-1990
Employment Growth Due to:
1979-1989
Net Job Growth
(000)
1979-89
1989-90
Urban*
25.595
524
Rural
-5.198
48
Annual Rate of
Growth
Growth of
Employment/
Population
1979-89
1989-90
3.2%
0.6%
2.6%
0.6%
0.2
2.3
0.4
-1.9
•The urban category pools data from cities and suburbs
^2
Growth of
Population
rise o f poverty in cities o f the Northeast was hy tar ihe
largest, 0.9 points greater than the next biggest increase
the 2.-t point rise in the suhurhan V('est
'liible 7.16 examines job growth by urban (including
cities and suburbs) anil rural selling, and shows (hat job
creation in the 1980s was exclusively an uriian phenomenon. Uetween 1979 and 1989. rural sellings experienced a
net loss o f 5,198,000 jobs, 2 1 % o f the 1989 rural work
force. Over the same period, 25,595,00(1 new jobs were
added to the urban workforce (28% ot the 1989 udian
workforce), at an annual growth rate of 3.2%. As the recession began in 1989, annual job growth in cities slowed lo
0.6%, and turned positive (though practically Hat) at 0 2 " .
in non-uriian areas
A p p l y i n g ihe same d e c o m p o s i t i o n teehni(|ue usetl
above, it appears that changes in the annual rates o f job
growth were tlriven primarily by population changes in the
adult population over the 1980s In urban areas. 2.6 of the
3 2% growth rate can be attributed to population growth;
the employment to population growth rate only accounted
for the other 0 6 percentage points o f job growth In fact,
the growth rate o f employment it) population was similar
in both areas, 1979-1989 (the rural rale was 0 2 points
lower lhan the urban), but the non-urixm atluli population
fell at an annual rate o f 2 3%, tlriving the negative job
growth rate o f 1.9%. Over the same peritKl, urban population grew 2.6%, annually.
Job creation in
the 1980s was
exclusively an
urban
phenomenon.
Between 1979
and 1989, rural
settings
experienced a
net loss ol
5,198,000 jobs.
�TABLE 7.17
Unemployment Rates, Regular and Adjusted,
Urban/Rural, 1979-1990
Urban
Year
Unemployment
Rate
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
Rural
Adjusted
Unemployment
Rate*
6.0%
9.5
10.3
13.1
13.1
10.4
9.9
9.5
8.7
7.9
7.5
7.9
5.8%
7.0
7.5
9.5
9.4
7.3
6.9
6.6
5.9
5.3
5.2
5.4
Unemployment
Rate
Adjusted
Unemploym
Rate*
5.7%
7.3
7.9
10.1
10.1
8.1
8.4
8.3
7.2
6.2
5.7
5.9
8.5%
10.7
11.5
14.9
14.9
12.2
13.0
12.8
11.3
10.1
9.1
9.4
0.0
0.2
The rural job contraction noted above led to higher
unemployment outside o f cities over the 1980s (Table
7.17 and Figure 7 H ) . Unemployment rates, both as usually measured and adjusted to include discouraged workers (discouraged from seeking w o r k by lack of previous
job-seeking success) and half o f the involuntary pan-time
workers (those w h o w o u l d prefer to w o r k full-time, but are
unable to find full-time w o r k ) , were at similar levels i n
urban and rural settings in 1979. However, by 1990, the
more comprehensive measure (the adjusted rate) was
9.4% i n rural areas while the urban rate was 1.5 points
lower at 7.9%. In fact, adjusted unemployment in rural
areas was consistently higher than urban unemployment
(Figure 7 H ) , diverging by a high o f 3 3 percentage points
in 1986.
0.6
0.3
FIGURE 7H
Urban and Rural Unemployment Rates,
1979-1990
Point Change
1979-89
1989-90
-0.6
0.2
-0.5
0.4
The rural job
contraction led to
higher
unemployment
outside of cities
over the 1980s.
Adjusted Rural'
•Unemployment rate adjusted to include discouraged workers and half of the involuntary part-time
workers.
S
^ ^ ^ ^
^
Adjusted Urban"
^ ^ ^ ^
C
D
Q.
2
-
1979
1980
1981
_l_
1982
1983
1984
1985
1986
' Unemptoymenl rate adjusted lo include discouraged workers and
half oi the involuntaiy parMimers.
354
355
-L
1987
1988
-L
1989 1990
�Chapter 8.
Conclusion
When we examine the economic problems that confronted American families over the 1980s, we find significant regional differences. While wage losses were common
to some states in all regions, 1979-1989, the Midwest and
the West saw big losses, while Northeastern states and
Southern states on the Eastern seaboard saw wage growth
Early indicatorsat the median. Income inequality also grew over the 1980s,
as
suggest that the the ratio of average family income in the top fifth to averrecession was age income in the bottom fifth grew in 43 states.
The onset of the recession in 1989 saw unemployment
much more
rates rise in all but nine states. However, as stressed
acutely felt in the
throughout the chapter, the early indicators included in
Northeast than in analysis suggest that the recession was much more
this
any other region.
acutely felt in the Northeast than in any other region.
Finally, our urban/rural analysis reveals that a downward
shift in the rural income distribution occurred over the
1980s. Over the same period, poverty grew in all types of
residences, but the largest increase was in cities. Employment growth was negative in rural settings, 1979-1989,
leading to higher rural (versus urban) unemployment
rates.
Losing Access to Basic
Necessities: Education,
Housing, Health Care,
and Child Care
So far, most of our research has defined well-being in
terms of economic resources—primarily wealth, earnings,
and income. This chapter adds another dimension,
expanding our scope to include Americans' access to those
basic necessities that have a vital impact on our living standard: education, housing, health-care, and child care. We
find that the same economic problems documented in earlier chapters—unequal growth, stagnant incomes for most
families, and increased poverty—have had a negative
impact on our educational and health-care systems, housing markets, and access to child care.
Education
As a society, Americans place a high value on equal
access to quality education for our children Yet, as the following data show, we are far from providing it. An examination of test results from public elementary and high schools
shows that minority and economically disadvantaged students do comparatively poorly in mathematics, science, and
reading. The same students tend to have limited access to
higher education. Finally, the U.S. devotes fewer public
resources to public education than other countries; we also
pay our public school teachers less than other equally welleducated professionals.
356
357
Unequal growth,
stagnant incomes
and increased
poverty—have
had a negative
impact on our
educational and
health-care
systems, housing
markets, and
access to child
care.
�TABLE 8.1
Proficiency in Mathematics and Science, Age 9,13, and 17 b
y
Race/Ethnicity, 1970-1990: Average Scores
Mathematics
Aqe and Race/Ethnicity
1973
1981
Science
1990
1970
1981
1990
Age 9
Race/Ethnicity
White
Hispanic
Black
As a Percentage of White Score
Hispanic
Black
225
202
190
224
195
204
235
214
208
236
n.a.
179
229
189
187
238
206
196
89.8%
84.4
87.0%
91.1
91.1%
88.5
n.a.
75.8%
82.5%
81.7
86.6%
82.4
274
239
228
274
240
252
276
255
249
263
n.a.
215
257
217
226
264
232
226
87.2%
83.2
87.6%
92.0
92.4%
90.2
n.a.
81.7%
84.4%
87.6
87.9%
85.6
310
277
270
304
272
277
310
284
289
312
n.a.
258
293
249
235
301
262
253
89.4%
87.1
89.5%
91.1
91.6%
93.2
n.a.
82.7%
84.8%
80.1
87.0%
84.1
Aqe 13
Race/Ethnicity
White
Hispanic
Black
As a Percentage of White Score
Hispanic
Black
Age 17
Race/Ethnicity
White
Hispanic
Black
As a Percentage of White Score
Hispanic
Black
358
Educational Proficiency: Low Scores for
Minorities and the Poor
Scholastic achievement is a reasonable indicator of subsequent opportunity, regarding both college and future
employment. Historically, achievement is highly influenced
by family background: parental education, family income,
community rank, and minority status. To the degree that
schools can offset these factors, opportunity is somewhat
equalized.
The next set of tables examines the performance of elementary and high-school students in mathematics and science, by race/ethnicity, community, family background, and
school quality. The findings suggest that, while minority
students have shown some improvement in mathematics
and science, inequalities exist in public schooling, and children from disadvantaged communities do worse relative to
their affluent, white counterparts.
Tables 8.1 and 8.2 (as well as Table 8.4) present the
National Assessment of Educational Progress (NAEP) findings in science, mathematics, and reading, using a proficiency scale from 0 to 500. Level 200 represents an understanding of simple principles; level 250, the application of
general information; level 300, the analysis of procedures
and data; and level 350, the integration of specialized
information.
In Table 8.1, the mathematics and science scores for 9-,
13-, and 17-year-olds are given by race and Hispanic ethnicity, along with black and Hispanic scores relative to whites'
scores, for the years 1970-1990. The levels show that while
all races have improved marginally over the period in both
subjects, only 17-year-old whites reach the 300 level (in
mathematics). In addition, Hispanics and blacks show consistently poorer outcomes than whites.
However, black students showed marked improvement
between the early 1970s and 1981, in their absolute
scores. In addition, they also narrowed the gap relative to
whites. For example, between 1973 and 1981, the ratio of
black to white mathematics scores for 13-year-olds rose
from 83-2 to 92.0%. Similar gains at a lower level were
made by blacks in science. Yet between 1981 and 1990,
359
The findings
suggest that, while
minority students
have shown some
improvement in
mathematics
and science,
inequalities exist in
public schooling,
and children from
disadvantaged
communities
worse relative to
their affluent, white
counterparts.
�TABLE 8.2
Scholastic Aptitude Test Scores
B Race/Ethnic Group, 1976-1990
y
Verbal
1976
1980
1985
1990
Change
1976-90
American Indian
Asian American
Black
Mexican American
Puerto Rican
White
388
414
332
371
364
451
390
396
330
372
350
442
392
404
346
382
368
449
388
410
352
380
359
442
0
-4
20
9
-5
-9
All Students
431
424
431
424
-7
American Indian
Asian American
Black
Mexican American
Puerto Rican
White
420
518
354
410
401
493
426
509
360
413
394
482
428
518
376
426
409
490
437
528
385
429
405
491
17
10
31
19
4
-2
All Students
472
466
475
476
4
Math
360
blacks made either slower gains or lost ground, relative to
white students. Nine-year-old black students, having narrowed the gap with whites by 6.7 percentage points in
mathematics between 1973 and 1981 (91.1% minus
84.4%), widened the gap by 2.6 points by 1990. The same
students gained 5.9 points in science over the earlier
period and made a smaller gain of 0.7 points over the
1980s.
Hispanic students also showed improvement over the
period covered in Table 8.1. Except for nine-year-olds in
mathematics, 1973-1981, their scores consistently
improved in both subjects. Similarly, they also made progress toward closing the gap with whites between 1981
and 1990. For 13-year-old Hispanic mathematics students,
the ratio to the score of whites was 92.4% in 1990, a 4.8
percentage point gain over their 1981 ratio of 87.6%.
Further evidence of black high-school students' improvement is seen in Table 8.2, Scholastic Aptitude Test (SAT)
scores in selected years, 1976-1990 (unlike the other tables
in this section, SAT scores range from 200 to 800). Over
the period covered in the table, black students posted
gains that are substantially larger than those of any other
group. The average scores of blacks grew by 20 points
between 1976 and 1990 in the verbal category and 31
points in mathematics. Mexican-Americans saw smaller
gains in verba] skills, and all minority groups saw smaller
gains (relative to blacks) in mathematics. However, even
though white students scored lower in both areas in 1990
than in 1976, there still exists a substantial gap between
the scores of whites and those of minorities, with the
exception of Asian-Americans' scores in mathematics.
361
Even though white
students scored
lower in 1990
than in 1976,
there still exists a
substantial gap
between the
scores of whites
and those of
minorities.
�TABLE 8.3
Proficiency in Mathematics and Science, Grades 4, 8, and 12
b Community, 1990: Average Scores
y
Test Subject
Mathematics
Science
Grade 4
Advantaged Urban
Disadvantaged Urban
Rural
As a Percentage of Advantaged Urban
Disadvantaged Urban
Rural
231.0
200.0
217.9
86.6%
94.3
251.6
208.6
235.0
82.9%
93.4
Grade 8
Advantaged Urban
Disadvantaged Urban
Rural
As a Percentage of Advantaged Urban
Disadvantaged Urban
Rural
283.2
251.7
261.1
88.9%
92.2
283.4
242.2
257.4
Table 8.3 presents a similar analysis as in Table 8.1 for
4th, 8th, and 12th graders by type of community in 1990.
Advantaged urban areas are metropolitan areas where a
high proportion of the students' parents are employed in
professional or managerial positions. Disadvantaged urban
areas are metropolitan areas where a high proportion of
students' parents are receiving welfare or are frequently
unemployed. Rural areas are those with populations below
10,000 and where many students' parents arc farmers or
farm workers.
The table shows that students from advantaged communities consistently score higher than those from disadvantaged or rural settings. The difference in achievement is
most pronounced for 4th graders, with disadvantaged science students scoring 82.9% relative to those from advantaged communities. While the differences narrow with age,
they still exist for high school seniors: disadvantaged mathematics and science students score 91.6% and 89.7%,
respectively, relative to advantaged urban students.
85.5%
90.8
Grade 12
Advantaged Urban
Disadvantaged Urban
Rural
As a Percentage of Advantaged Urban
Disadvantaged Urban
Rural
362
308.4
282.6
290.2
91.6%
94.1
304.1
272.9
290.7
89.7%
95.6
363
Students from
advantaged
communities
consistently
score higher than
those from disadvantaged or
rural settings.
�TABLE 8.4
Effects of Family Background on Scores
of 8th Graders, 1988
Family Background
Test Subject and
Score Group
Disadvantaged
Average
Advantaged
44.3%
30.8
17.0
7.9
100.0
23.8%
26.9
27.0
22.3
100.0
9.3%
16.1
26.2
48.5
100.0
44.0
29.1
18.5
8.4
100.0
24.2
27.0
25.8
23.0
100.0
11.1
16.9
29.4
42.5
100.0
42.3
29.2
19.3
9.1
100.0
24.7
25.5
27.3
22.6
100.0
11.1
16.9
29.4
42.5
100.0
Mathematics
Lower Fourth
Second Fourth
Third Fourth
Upper Fourth
Total
Reading
Lower Fourth
Second Fourth
Third Fourth
Upper Fourth
Total
Table 8.4 introduces family background, measured on a
composite scale w h i c h factors in parental education, occupation, and income, for 8 t h graders in 1988 (the latest available data of this sort). The pattern i n the table is strikingly
similar for each subject: the higher the student's socioeconomic status, the more likely he or she is to have a high
score. Among those children w i t h the most advantaged
upbringing, between 42.5% and 48.5% were i n the top
fourth of all students i n science, reading, and mathematics.
Conversely, a m o n g the most disadvantaged students,
between 42 3% and 44.3% of all students were i n the lowest fourth of all students. Furthermore, less than 10% of
students from disadvantaged families ever made it into the
top fourth.
These racial, community, and family differences reflect a
two-tiered system of public education provision. Table 8.5
shows that w h e n schools are sorted into thirds based on
student performance, the gap between the top and bottom
students is disturbingly wide. The average proficiency
TABLE 8.5
Average Proficiency of 4th, 8th, and 12th Graders b
y
School Rank, 1990
Science
Lower Fourth
Second Fourth
Third Fourth
Upper Fourth
Total
The higher
the student's
socio-economic
status, the more
likely he or she is
to have a high
score.
Pecentage of Students
at or Above
Test Subject
Average
Proficiency
Level
200
Level
250
Level
300
Level
350
232
198
34
90%
46
22%
3
0%
0
0%
0
Grade 4
Top One-Third Schools
Bottom One-Third Schools
Point Difference
Grade 8
Top One-Third Schools
Bottom One-Third Schools
Point Difference
284
246
38
100
94
88
44
29
4
1
0
312
273
39
100
100
97
77
66
18
10
1
Grade 12
Top One-Third Schools
Bottom One-Third Schools
Point Difference
364
365
�FIGURE 8A
Total College Costs by Type of Four-Year Institution,
1976-1989
score difference between students in the top and bottom
schools is 34 points for fourth graders and 39 points for
high-school seniors. In addition, fewer than half of the 4th
graders from the bottom schools (46%) demonstrated a
systematic grasp of basic grade-level principles (in this case,
addition with whole numbers), while 90% of 4th graders
from the top schools showed an understanding of these
principles. A similar pattern exists at the 300 level for seniors, with 66% of those at the top schools preforming
"on-grade," as opposed to 18% of those from the lowerperforming schools.
16,000
14,000
-
Private
•
12,000
j
10.000
All Institutions
CD
1991
Doll
<
P
-
6,000
4.000
i
2,000
i
i
i
i
i
i
i
i
i
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
TABLE 8.6
The Shrinking Federal Student Aid Grant,
School Years Beginning in 1977-1989
(1991 Dollars)
Year
1977
1979
1981
1983
1985
1987
1988
1989
Maximum
Grant
$3,017
3,313
2,524
2,461
2,658
2,518
2,533
2,526
366
Student Cost
(Tuition, Room,
and Board)
$5,196
5,170
5,274
5,698
6,183
6,606
6,757
6,886
Maximum Grant
as Percentage
of Student Cost
58.1%
64.1
47.9
43.2
43.0
38.1
37.5
36.7
Access to College: Higher Costs, Less A i d
As Figure 8A shows, the real cost of attending college
has increased since 1976. The figure shows an increase in
the total cost, inclusive of room and board, of both public
and private four-year institutions, from 1976 to 1989- The
rate of increase appears especially precipitous for private
colleges and universities; their average total cost of attendance had reached »13,563(1991 dollars)by 1989- The figure also shows that, throughout the 1980s, the "affordability gap" between private and public institutions has grown,
a result of the faster growth of private, relative to public,
tuition.
However, it is frequently noted that the majority of
middle- and low-income college students receive some
form of government student aid, either from federal grants
or loans, or from work-study. The next few tables include
this factor in the cost burden of seeking higher education.
The evidence shows that aid as a percentage of tuition has
fallen over time, and that families with middle- and lowincomes are particularly hard-pressed to meet their
expected contribution to their children's college
education.
Table 8.6 presents the maximum grant, average student
cost (an average of costs at all types of higher institutions),
and the grant as a percent of student cost, from 1977-1989.
The final column shows the dramatic drop in aid as a percentage of the total tuition. In 1977, the maximum grant
accounted for 58.1% of student costs; by 1989, that percentage had dropped 21.4 points, to 36.7%.
367
Fewer than half of
the 4th gratters
from the bottom
schools (46%)
demonstrated a
systematic grasp
of basic
grade-level
principles while
90% of 4th
graders from the
top schools
showed an
understanding of
these principles.
�TABLE 8.7
Costs of College: The Percent of Students with College Costs Covered b
y
Expected Family Contribution and the Average Percent of
Total Cost Met b Aid, b Family Income, 1987
y
y
Percent of Students wilh Costs
Covered by Expected
Average
Family Contribution
Percent Met by Aid
Public College Private College Public College Private College
Family Income
Less than $11,000
$11,000-17,000
$17,000-30,000
$30,000-50,000
More than $50,000
44.9%
47.8
56.2
80.5
95.7
37.6%
35.4
40.8
52.7
83.3
62.9%
64.5
44.9
23.0
7.8
64.1%
69.9
61.2
40.3
17.0
TABLE 8.8
Net Cost of College as a Ratio of Expected Family
Contribution, by Family Income, 1987*
Public College
Average
Less than $11,000
$11,000-17,000
$17,000-30,000
$30,000-50,000
More than $50,000
Private College
Average
1.99
1.59
1.34
0.76
Family Income
3.58
2.88
1.95
1.35
0.68
0.35
*A ratio of one means that the cost of college is equal to Ihe expected family contribution.
368
Table 8.7, introduces the concept of expected family
contribution (EFC). Established by Congress, this formula
takes into account a family's assets and expenses and then
determines what the family is expected to contribute to
their child's education.
Table 8.7 shows the average percent of total costs met by
EFC and by aid (including grants, loans, and work-study) for
students i n different family income brackets in 1987. Less
than half of low income families (44.9% ) were able to meet
their expected contribution to public colleges; conversely,
for wealthy families, the vast majority ( 9 5 . 7 % ) met their
EFC. Moreover, the table shows that as family income rises,
the percentage of families that can meet their children's
college expenses rises. I n the case of private college, even
among families earning J30,000 to $50,000 in 1987, only
about half ( 5 2 . 7 % ) were able to meet their EFC. The last
t w o columns make the obvious point that aid is substantially more important to poor students relative to wealthy
students. For students from families w i t h incomes below
>30,000 in 1987, the aid package covered an average of
over 60% of total costs for private colleges. For the most
wealthy students, this percentage was 17.0% for private
college, 7.8% for public.
Table 8.8 demonstrates the difficulty less wealthy families have funding their children's college education. The
table presents the ratio of the net cost of college (subtracting out grants, loans, and work-study) to the EFC. When this
ratio exceeds one, it means that net costs of college attendance are higher than the EFC. In fact, this was the case for
most income groups in the table. For public colleges, net
costs exceeded the EFC (on average) for those families
earning less than *30,000 i n 1987. For those families w i t h
incomes below SI 1,000, net costs were about twice the
EFC for public colleges; for private schools, the ratio was a
substantially higher 3 58. In fact, for private colleges, only
the most wealthy families had a ratio below one.
369
In 1977, the
maximum grant
accounted for
58.1% of student
costs; by 1989,
that percentage
had dropped 21.4
points, to 36.7%.
�FIGURE 8B
Enrollment Rates for 3- & 4-year-olds in Preprimary
Education by Race/Ethnicity, 1974-1989
Minority Education: Uneven Results
9>
a.
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
FIGURE 8C
Percent of 18- to 24-year-olds Completing High School,
1970-1989
90
85
80
-
While
" — ^
Table 8.1 gave evidence o f the disadvantage minority students face regarding educational proficiency. The following
data suggest that m i n o r i t i e s also suffer disadvantages
regarding educational enrollment and attainment as well.
However, at least i n the case of high school completion
rates, the position of blacks has improved.
Pre-kindergarten education for three- and four-year-olds
(including Head Stan), has been highly touted as the most
important component o f a national education goal: readiness to learn. However, as Figure 8 B shows, preprimary
enrollment has diverged for black and white children over
the 1980s. From the beginning o f the figure until about
1980, an approximately equal p r o p o n i o n o f black and
w h i t e c h i l d r e n , b e t w e e n 2 1 % and 30%, received
preprimary instruction. Throughout the 1980s, the trend
for white enrollment has continued to rise while black
e n r o l l m e n t has flattened o u t . Hispanic e n r o l l m e n t has
been relatively flat throughout the period, and has consistendy been at the lowest level.
F i g u r e 8C, covering the years 1970-1989, shows the
percenuge of 18- to 24-year-olds of a given race w h o have
completed high school. The rate for whites has remained
relatively constant, between 80 and 85 percent. However,
the completion rate for blacks rose throughout the 1980s,
and they can be seen to have gone a long way toward closing the gap w i t h whites. Hispanics, however, show the lowest rates, and they began a d o w n w a r d trend in the late
1980s.
All
75
70
-
2.
65
60
Hispanic
55
50 -I
1
1
1
1
1
1
1
1
1
L
1
1
£ £ &£ * £ £ £ £ £ $ i M
370
1
1
J
1
1
1
1
i i i * a# s
371
Pre-kindergarten
education for
three- and fouryear-olds has
been highly touted
as a national
education goal.
However, preprimary
enrollment has
diverged for
black and white
children over
the 1980s.
�FIGURE 8D
Percent of 18- to 24-year-olds in College,
1970-1989
The percentage o f 18- to 24-year-olds enrolled in college
is shown in Figure 8 D , also for the years 1970-1989 Once
again, blacks and Hispanics are at a distinct disadvantage
relative to whites. And in this case, although blacks began
to show an upward trend i n the 1980s, the gap between
them and whites remains. Hispanics have particularly l o w
rates o f college enrollment, and, as w i t h high school completion rates, show a d o w n w a r d trend in the late 1980s.
Minorities and Advanced Degrees
TABLE 8.9
Postsecondary Enrollment and Degree Attainment by Race/Ethnic Group,
1988-89
(U.S. Residents Only)
Undergraduate
Enrollment
Group
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other"
Total
(U.S. Residents)
Bachelor's
Degrees
Fall 1988
80.2%
9.4
5.7
4.7
100.0
Doctorates*
1988-89
Master's
Degrees
1988-89
86.8%
5.9
3.0
4.3
87.9%
5.1
2.6
4.3
88.9%
3.8
2.2
5.1
100.0
100.0
1988-89
In addition to their low rates of high school completion
and college enrollment relative to whites, few m i n o r i t y college graduates go on to pursue advanced degrees. Table 8.9
shows that in 1988, whites made up 80.2% of undergraduate enrollment, w h i l e blacks and Hispanics made up 9.4%
and 5.7%, respectively ( i n this table, there are no Hispanics
in the white and black categories). However, given this distribution of enrollment, whites are disproportionately represented among degree recipients, and this result becomes
larger as the level o f advancement increases. In 1988,
blacks received 5.9% of bachelor's degrees, 5 . 1 % of master's, and 3 8% of doctorates. Hispanics post a similar
decline, at a lower level. Conversely, the percentages for
whites were bachelor's: 86.8%, master's: 87 9%, and doctorates: 88.9%. (The high percentage of doctorates in the
"Other" category, relative to blacks and Hispanics, was
driven exclusively by Asian-Americans.)
Educational Expenditures: Less Investment
than Other Countries and Greater Inequality
Within the U.S.
One indicator of a government's commitment to public
education is the percent of its gross domestic product
(GDP) spent on education. Another indicator of society's
commitment to education is seen in the degree to w h i c h
teachers are remunerated relative to other professionals.
On both of these indicators, the U.S. falls short.
100.0
'Ph.D., Ed.D., and comparable degrees.
"Other includes Asian/Pacific Islander and American-lndian/Alaskan Native.
372
373
Few minority
college graduates
go on to pursue
advanced
degrees.
�TABLE 8.10
Public Expenditures on Education as
a Percent of GDP, 1986
Country
1986
Australia
Austria
Belgium
Canada
Denmark
Finland
France
Greece
West Germany
Ireland
Italy*
Japan
Luxembourg
Netherlands
New Zealand
Norway
Portugal
Sweden
Switzerland
Turkey
United Kingdom
Yugoslavia
United States
5.3%
6.1
5.4
6.7
7.1
5.3
5.6
2.7
4.3
5.8
5.1
5.0
7.0
6.6
4.9
6.6
4.1
7.3
5.1
2.0
5.0
3.5
4.8
'1986 data unavailable, figure is for 1984.
Table 8.10 gives public educational expenditures as a
percent of GDP for the industrialized countries, for the year
1986, the latest available data. Public expenditures in the
U.S. accounted for 4.8% of GDP, a figure that places the U.S.
18th out of 23 countries.
However, when private sector spending is factored into
educational expenditures, as in Table 8.11, the U.S. rises to
a middle position of those countries with valid data on private expenditures. This reflects the fact that the U.S., at
1.7% of GDP has the highest proportion of private educational expenditures of the countries shown. Thus, those
U.S. families with fewer resources of their own to devote to
their children's education, i.e., the families that depend on
public provision, are at a distinct disadvantage, relative to
most other industrialized countries.
Public
expenditures on
education in the
U.S. accounted
tor 4.8% ol GDP, a
ligure that places
ihe U.S. 18th out
ot 23 countries.
TABLE 8.11
Private and Total Educational Expenditures,
as a Percent o GDP, 1986*
f
Country
Australia
Canada
Denmark
Finland
France
Greece
West Germany
Ireland
Japan
Netherlands
Norway
Switzerland
United States
Private
Total,
Public and
Private
0.4%
0.6
0.1
0.5
1.0
0.2
0.2
0.3
1.4
0.2
0.2
0.1
1.7
5.7%
7.3
7.2
5.8
6.6
2.9
4.5
6.1
6.4
6.8
6.8
5.2
6.5
'Table includes only those countries with valid data on public and private expenditures.
374
375
�TABLE 8.12
Public School Expenditures Per Pupil, 1989-1990
Adjusted for Cost of Living Differences Between States
Adjusted
Spending
Per
Pupil
1989-90
State
Highest Ten States:
New York*
New Jersey*
Connecticut*
Pennsylvania*
Maine*
Rhode Island*
Wisconsin
Alaska
Vermont*
Wyoming
$6,994
6,708
6,261
6,033
6,028
6,016
5,978
5,896
5,759
5,756
Percent
ol U.S.
Average
141.2%
135.5
126.4
121.8
121.7
121.5
120.7
119.1
116.3
116.2
Lowest Ten States:
South Carolina**
North Dakota
Tennessee"
Oklahoma
Hawaii
Alabama**
Arkansas**
Idaho
Mississippi"
Utah
4,029
3,869
3,779
3,718
3.640
3,606
3,600
3,499
3,454
2,928
81.4
78.1
76.3
75.1
73.5
72.8
72.7
70.7
69.8
59.1
U S Average
..
4,952
Resources are also distributed i n a highly unequal manner w i t h i n the U.S., i.e., between the states. Table 8.12
shows that, even after adjusting for different costs of living,
some states devote substantially more resources to education than others. The adjusted expenditures for the highest
and the lowest ten states are given i n the table, along w i t h
the percent of the U.S. average, for the school year
1989-1990. The wide disparity is seen in the range of percentages, from a low of 59.1 % in Utah to a high of 141.2%
in New York. Furthermore, the table highlights regional differences, as northeastern states are over-represented among
the highest ten states, w h i l e southern states are over-represented among the lowest.
Finally, although much public rhetoric emphasizes the
importance of well-trained teachers, society's economic valuation of teachers belies this emphasis. F i g u r e 8E shows
the average salaries of beginning public school teachers ( i n
1991 dollars), compared to the beginning salaries of college graduates in five other fields, from 1972 to 1991.
Resources are
distributed in a
highly unequal
manner between
the states.
100.0
FIGURE 8E
Beginning Teacher Salaries vs. Other College Graduates,
1972-1991
45,000
reaching
^
Accounfing
SatosMaKctlng
Busness Admin
40,000
Enpnmnng
Liberal Arts
Main oi Slat,sbcs
Economia
ui
s
o
Q
35,000
8
?
30.000
-
—*
'A
w
cn 25.000
•Northeastern Slates
"Southern States
20,000
- Teaching
-
15 000 '—'
1
1
1
"
1
1
1
'
1
1
1
'-
J
I
1
1
L
££££ £ £ £ £ i i i i i i i i i i i
376
377
�TABLE 8.13
Housing Costs, 1967-1990
(1989 Dollars)
Owner Cost
Annual
After-tax
Cash Cost
Year
1967
1973
1977
1979
1987
1989
1990
$5,801
6,825
8,106
9,171
8,202
8,599
8,341
Renter Cost
Monthly
Gross Rent
$368
368
361
360
411
409
406
FIGURE 8F
After-Tax Home Owning and Rental Costs,
1967-1990
Annual Home Owning Cost (1989 Dollars)
11,000
££i£Sg£i&£$
378
Monthly Rental Cost (1989 Dollars)
420
W h i l e the salaries of most professionals cluster together in
the middle range ( w i t h engineers notably higher), teachers'
salaries are consistently the lowest. Even though their real
earnings grew i n the mid-1980s (after falling i n the late
1970s), there still remains a substantial pay gap between
teachers and other college graduates.
Housing
Housing Costs: Rising Faster than
Incomes
The affordability of housing, both for renters and owners,
depends on trends i n housing costs and trends in family
income. The f o l l o w i n g tables and figures show that the cost
of o w n i n g a home rose quickly i n the late 1970s and early
1980s, both i n real terms and as a percentage of income.
Rental costs also grew throughout the 1980s. Oversupply
of housing stock and falling mortgage rates i n the latter
part of the 1980s led to a fall i n home ownership costs.
However, stagnant income g r o w t h has kept the cost of
housing as a percentage of income high, relative to earlier
periods. Furthermore, large, low-income, and female-headed families continue to face inordinate difficulty i n the
housing market.
Table 8.13 and F i g u r e 8F show the after-tax cost (mortgage payment less tax savings) o f home ownership from
1967 to 1990, i n 1989 dollars. Costs began to accelerate in
a b o u t 1977, w h e n they began a five-year c l i m b , f r o m
18,106 to $10,617 (Figure 8F). By the latter p a n o f the
1980s, the trend had declined and flattened, but costs
remained high relative to the beginning p e r i o d i n the
figure.
Gross rental costs (inclusive of fuel and utility costs,
taxes, and insurance), also plotted i n Figure 8F, show a steep
incline lasting u n t i l the mid-1980s. At $360 in 1979 (Table
8.13), gross rent peaked at $411 i n 1987 and has fallen
only slightly since then. As we show below, this rise in the
cost of rental housing, in tandem w i t h the particularly stagnant g r o w t h of renters' incomes, led to a steady incline in
the housing burden for renters.
*********
379
Stagnant income
growth has kept
the cost of
housing as a
percentage of
income high,
relative to earlier
periods.
�FIGURE 8G
Home Owning and Rental Costs a s a Percent of Income,
1967-1990
50
45
-
Home Owning Costs
/
\
o
u
_c
o
c
35
o
30
-
25
-
^^^^
20
- 1 — 1 - 1 1 1 .1
J
y
l_l
—"
Rental Costs
II 1
1
1
I
1
1
1
1
1
1
i
1
1
TABLE 8.14
Burdens of Buying Homes and Renting, 1967-1989
Housing Expenses as Percent
of Income for:
Prime
First-Time
Home Buyers*
Renters
At Business Cycle Peaks:
1967
1973
1979
1989
21.2%
24.3
36.0
32.3
24.4%
23.3
25.8
28.4
Period Averages:
1967-73
1974-79
1980-89
22.5
29.9
37.2
23.7
24.9
29.0
1
Figure 8G and Table 8.14 introduce income growth
into the analysis, by showing rental costs and first-time
home purchasing costs as a percentage of income between
1967 and 1990. The housing cost burden for first-time
home buyers is the expense of the median-priced home,
expressed as a percentage of the median income of married-couple renters aged 25-29. These families are selected
because they are especially likely to be considering the purchase of a first home. The cost burden for renters is the
median gross rent as a percentage of the median renter's
income.
Let us first consider first-time home buyers. Driven by
the steep cost increases portrayed in Figure 8F, the housing
cost burden rose from 21.2% of income in 1967 to 36.0%
in 1979 (Table 8.14). As the housing market "softened," the
burden fell, as seen in the decline toward the end of Figure
8G. However, as noted above, since family income at the
median grew at a slower rate in the 1980s than in earlier
periods, first-time home buyers were unable to reap the full
benefits of the softer housing market in the late 1980s.
While the cost burden fell from 36.0% in 1979 to 32.3% in
1989 (Table 8.14), the average over the 1980s, at 37.2%,
was significantly higher than that of earlier periods, when it
averaged 29 9% from 1974 to 1979, and 22.5% from 1967
to 1973.
With the exception of a slight decline at the end of the
1980s, renters saw a constant rise in their housing cost burden. Once again, the steep incline in gross rents over the
1980s (see Figure 8F) coupled with the previously mentioned sluggish income growth, led to the increased burden
shown in the figure. Unlike first-time buyers, the cost burden for renters was higher at each peak year, as shown in
Table 8.14, climbing from 24.4% in 1967 to 28.4% in
1989. Similarly, the average burden grew consistently over
the period; for 1980-1989, the average cost burden for renters was 29 0%.
As earlier chapters have noted, the income growth over
the 1980s has generated highly unequal outcomes, and certain types of families—larger families and female-headed
families—were more vulnerable than others. These outcomes are, of course, evident in the housing market. Below
we examine the housing experience of low income fami-
•Married-couple renters aged 25 to 29.
380
381
The average
housing cost
burden over the
1980s, at 37.2%,
was signif icantly
higher than that of
earlier periods.
�TABLE 8.15
Percent of O n r and Renter Families
we
W o Could not Afford Median Priced Home, 1988
h
Family Type
Owners
Renters
All Families
30.9%
90.4%
Married Couple Familes
With Children
25.5
32.2
86.5
90.7
Female-headed Familes
With Children
61.0
71.5
97.2
97.2
TABLE 8.16
Shelter Poverty, by Household Size
1970-1989
A. By Household Size
Household Size
Year
1
2
3
4
5
1970
1979
1989
1991
39.0%
28.9
24.4
26.7
27.1%
24.4
23.5
25.3
20.5%
24.6
27.1
29.5
26.6%
28.8
30.4
32.2
31.7%
37.7
40.0
43.3
B. Small vs. Large Households
Year
1970
1979
1989
1991
All
Households
29.4%
34.2
36.6
38.9
Small (1-2
Persons)
All Households
Large (3+
Persons)
29.7%
31.5
32.7
34.9
30.3%
25.6
23.8
25.7
382
6+
43.5%
51.8
55.6
57.6
lies, but Table 8.15 shows the percentage of a l l owner and
renter families in 1988 who would not have been able to
purchase a median-priced home in that year, by family type.
The table shows that the median home was out of reach for
more than 90% of all families of renters in 1988. Among
married couples with children who owned homes, the
housing market and their family incomes were such that in
1988, 32.2% would have been unable to buy the medianpriced home. The housing market was particularly inaccessible to female-headed families with children who
rented in 1988, as 97.2% of these families could not afford
the median home.
The high costs of housing in the 1980s has led analysts of
the housing market to consider the amount of income families have left over for other necessities, after their housing
expenditures. This work has led to the notion of "shelter
poverty." A household is considered shelter poor if, after
housing expenses, it does not have enough to pay for a minimum amount of non-housing necessities, as defined by the
U.S. Bureau of Labor Statistics.
As shown in Table 8.16, 32.7% of all households
(including both renters and homeowners) were "shelter
poor" in 1989, higher than both the 1970 rate of 29 7%
and the 1979 figure of 315%. In other words, in 1989,
close to one-third of households were unable to buy
enough food, clothing, and other necessities after paying
for housing. Table 8.16 also shows that shelter poverty
rates have risen over time for larger households and fallen
for smaller households. For example, 21.7% of two-person
households were shelter poor in 1970; by 1989, this rate
had fallen to 23.5%. However, for households with four or
more persons, shelter poverty rates have increased since
1970, and the larger the household size, the greater the
increase. Well over half (55.6%) of the largest households—those with six or more persons—were shelter poor
in 1989. As the figures for 1991 show, shelter poverty has
risen for all households over the recent recession.
H o m e O w n e r s h i p Rates: F a l l i n g f o r Young
F a m i l i e s a s a R e s u l t o f R i s i n g Costs
As would be expected, the rising costs of housing and
sluggish income growth have depressed home ownership
rates in the 1980s for those families entering the housing
383
In 1989, close to
one-third of
households were
unable to buy
enough food,
clothing, and
other necessities
after paying for
housing.
�TABLE 8.17
H m Ownership Rates b Age of Household Head,
o e
y
1973-1990
Age of
Household
Head
1973
1980
1989
1990
Under 25
25-29
30-34
35-39
40-44
45-54
55-64
65-74
75 +
23.4%
43.6
60.2
68.5
72.9
76.1
75.7
71.3
67.1
21.3%
43.3
61.1
70.8
74.2
77.7
79.3
75.2
67.8
17.6%
35.4
53.6
63.9
70.8
75.3
80.2
78.2
70.3
15.3%
35.9
51.5
63.1
70.4
76.1
80.4
78.2
71.0
Total
64.4
65.6
63.9
64.1
Home Ownership Rate in:
Point Change
1973-80
1980-89
-2.1
-0.3
0.9
2.3
1.3
1.6
3.6
3.9
0.7
-3.7
-7.9
-7.5
-6.9
-3.4
-2.4
0.9
3.0
2.5
1.2
-1.7
market. Table 8.17 shows home ownership rates, by the
age of the household head, i n selected years from 1973 to
1990. Overall, the percentage of households o w n i n g homes
grew slightly (1.2 percentage points) between 1973 and
1980, and fell slightly (-1.7 points) over the 1980s
However, more dramatic changes in ownership rates are
evident when we examine changes by age group Here we
see that the percentage of home owners among all households headed by someone under 55 fell by at least 2.4 percentage points between 1980 and 1989- The biggest losses
went to young households. In 1980, 43-3% of those household heads between the ages of 25 and 29 owned homes.
By 1989, that percentage had fallen 7 9 points, to 35.4%.
Households headed by a person aged 30-34 also did substantially worse over the period: their ownership rates fell
by 7.5 points.
O w n e r s h i p rates for older persons ( 5 5 and o v e r )
increased over both periods, though less so between 1980
and 1989. This increase is likely the result of older persons'
incomes g r o w i n g faster than those of younger persons over
the period. Also, many older heads of households owned
homes prior to the steep rise in housing prices shown i n
Figure 8F. The added wealth from the increased home
equity realized by those households headed by an older
person helped to insure that home ownership rates have
remained high—above 75%.
The years 1983-1989 saw one of the longest expansions
of the United States economy since World War I I , yet housing costs as a percentage of income and shelter poverty
were excessively high throughout the 1980s, and home
ownership rates fell dramatically for most families. Once
again, as we saw i n Chapter 6, the uneven and unequal
nature of the expansion is evident. There were some economic winners over the 1980s; many more lost.
Low-Income Housing: Reduction in Housing
Stock and Deteriorating Conditions
Since 1970, the gap between the supply of, and the
demand for, low-income housing has expanded (this is
one factor leading to the high rates o f shelter poverty
noted above). In this section, we examine the housing
problems o f the poorest members o f society. As Chapter 6
documented, poverty rates were high in the 1980s, partic-
384
385
Households
headed by a
person aged
30-34 also did
substantially
worse from 1980
to 1989: their
ownership rates
lell by 7.5 points.
�u l a r l y w h e n c o m p a r e d t o o t h e r periods o f e c o n o m i c
g r o w t h . The f o l l o w i n g tables and figures show that the
number o f low-income units has fallen way behind the
number o f low-income households. We show that p a n o f
this development is a result o f a diminished federal commitment to low-income housing over the 1980s. In addition, the housing conditions experienced by the p o o r are
substantially worse than those o f the non-poor.
F i g u r e 8 H shows that the number of low-rent units has
fallen since 1970, w h i l e the number of low-income renters
has increased. Low-income households in this figure are
those whose annual income is i 10,000 or less ( 1 9 8 9 dollarsX approximately the poverty line for a family of three in
1989- Low-rent units are those units renting for J250 a
month or less, an expenditure of 30% of monthly income.
There were 6.8 m i l l i o n units meeting this criterion in
1970. Since then, there has been a steady decline, until
there were only 5.5 m i l l i o n units i n 1989. However, on the
demand side, the number of low-income renters has steadily grown, from 6.4 m i l l i o n i n 1970 to 9.6 m i l l i o n i n 1989.
The cause of this w i d e n i n g gap stems from the g r o w t h i n
poverty, as w e l l as a reduced government commitment to
house the poor. F i g u r e 81 shows the net additional units of
federally funded low-income housing from 1977 to 1991.
In 1977, 498,000 units were added to the stock of lowincome units. W h i l e the numbers of new units began to fall
in the late 1970s, by 1982 there had been a radical drop i n
net additional units, reflecting the Reagan administration's
reliance on existing housing stock. By 1990, net new units
hit a low of 102,000.
Besides the shrinking supply o f affordable housing, the
p o o r also face significandy worse housing conditions than
the non-poor. Table 8.18 shows t w o indicators of housing
conditions, by poverty status: deficient housing, defined as
housing w i t h moderate to severe physical or structural
deficiencies, and overcrowded housing, defined as housing more than one person per r o o m . In 1989, 17 9% o f all
poor households were housed in deficient structures, compared w i t h 6.6% o f n o n - p o o r households. Poor black
households were in the worst condition: 29.2% o f their
structures were deficient in 1989-
FIGURE 8H
Low-Rent Units and Low-Income Renters,
1970-1989
1970
1978
1983
1989
FIGURE 81
New Federally Funded Low-Income Housing Units,
1977-1991
600
-8
1977
1979
1981
386
1983
1985
1987
1989
1991
387
In 1989, 17.9%
of all poor
households were
housed in
deficient
structures,
compared with
6.6% of non-poor
households.
�TABLE 8.18
Percent of Households Living in Deficient and
Overcrowded Housing, 1989
Deficient
Housing
Overcrowded
Housing
17.9%
19.4
15.8
Households
7.9%
10.0
4.7
Poor Households
All Poor Households
Renters
Owners
Black
Hispanic
White
29.2
23.1
12.5
Black
Hispanic
White
6.6
9.8
5.1
2.1
3.7
1.3
13.0
12.0
5.6
Non-Poor Households
All Non-Poor Households
Renters
Owners
9.1
26.0
4.0
3.3
11.3
0.2
TABLE 8.19
Estimates of Homelessness, 1987
Subject
Range
Estimated Range of Homeless
Persons in a Given Month
496,000-600,000
Estimated Range of Homeless
Children in a Given Month
The tight, low-income housing market documented
above has led to overcrowded conditions for many poor
households. In 1989, 7.9% of all poor households lived in
overcrowded conditions, as opposed to 2.1% of non-poor
households. Overcrowding is particularly severe for Hispanics. Poor Hispanics experienced the highest percentage
of overcrowded households, with more than one-quarter
(26.0%) overcrowded. In fact, non-poor Hispanic households had a higher rate of overcrowding ( 11.3% ) than all
non-Hispanic poor households.
Homelessness
The housing problems of the low-income population—
high costs, low supply, deficient and overcrowded housing—have led to the highly visible increase in homelessness over the 1980s. The problem of homelessness has
proven to be a difficult one to analyze. Due to the difficulty
in both defining and counting the homeless, it is not easy to
determine the number of homeless persons This brief section presents the most recent data on homelessness, emphasizing the numbers and characteristics of the homeless.
Unfortunately, the most recent data arc from the year 1987.
Most analysts agree that the magnitude of the problem has
grown since that time. In fact, in a recent survey of major
cities, 25 of the 28 cities in the survey report an average
13% rise in requests for emergency shelter over the course
of 1991.
Table 8.19 gives a range of the number of homeless persons between about 500,000 and 600,000 in a given
month, including 35,000 to 90,000 children. This estimate
is an extrapolation from the number of homeless found to
be using homeless services (soup kitchens and/or shelters)
during the month of March 1987. There are many assumptions embedded in this estimate, including an estimate of
the number of homeless not using services, given the number of known service-users. However, this estimate is well
within the range of most studies of homelessness.
35,000-90,000
388
389
In a recent survey
ot major cities, 25
of the 28 cities in
the survey report
an average 13%
rise in requests
for emergency
shelter over the
course of 1991.
�TABLE 8.20
Characteristics of Service-Using Homeless
Adults, in Percents, 1987
Characteristic
Percent
Gender
Male
Female
81%
19
Race/Ethnicity
White (Non-Hispanic)
Black
Hispanic
Other
46
41
10
3
Institutionalization*
Mental Hospitalization
Detox or Alcohol/Drug
Treatment Center
Jailed for 5 or More Days
State or Federal Prison
19
33
52
24
The characteristics of those adults w h o use homeless services arc given i n Table 8.20. When examining the race o f
those homeless persons w h o use services, the largest percentage are white ( 4 6 % ) , w i t h blacks accounting for 4 1 %
and Hispanics, for 10%. A n o v e r w h e l m i n g majority, 8 1 % ,
of the homeless are male. I n addition, about one f i f t h o f
homeless service users ( 1 9 % ) have been hospitalized for a
psychiatric condition; over half have experienced some
degree of imprisonment (possibly for being homeless).
A further important point is made regarding the length
of time spent homeless. Recall that the estimates i n Table
8.19 refer to the number o f homeless i n a particular month.
Since, over the course of a year, there is a good deal of turnover i n the homeless population, the number of homeless
i n a month w i l l be substantially less than the number
homeless during the course o f a year. Table 8.20 shows that
among the homeless adults using services, over half ( 5 4 % )
had been homeless one year or less, w h i l e a smaller group,
comprising about one-fifth ( 19% ) were homeless for over
four years. Homeless analysts estimate that the number o f
homeless i n a given year is about twice the number o f
homeless i n a given month. This suggests, based on the data
i n Table 8.19, that there w e r e approximately 1 m i l l i o n
homeless persons i n 1987.
Length of Time Spent Homeless
Less than 1 Month
2-3 Months
4-6 Months
7-12 Months
13-24 Months
25-48 Months
More than 48 Months
8
13
19
14
16
12
19
'These categories are neither exhaustive nor mutually exclusive.
390
391
There were
approximately
1 million homeless
persons in 1987.
�Health Care: Higher Costs, Poor Health
Outcomes, and Many Uninsured
TABLE 8.21
OECD Spending on Health, 1960-1989
Total Health Expenditure as Percent
of Gross Domestic Product in:
Country
1960
1970
1980
1989
Australia
Austria
Belgium
Canada
Denmark
Finland
France
West Germany
Greece
Iceland
Ireland
Italy
Japan
Luxembourg
Netherlands
New Zealand
Norway
Portugal
Spain
Sweden
Switzerland
Turkey
United Kingdom
4.6%
4.6
3.4
5.5
3.6
3.9
4.2
4.7
3.2
1.2
4.0
3.3
2.9
n.a.
3.9
4.4
3.3
n.a.
2.3
4.7
3.3
n.a.
3.9
5.0%
5.4
4.0
7.2
6.1
5.7
5.8
5.5
4.0
4.3
5.6
4.8
4.4
4.1
6.0
5.1
5.0
n.a.
4.1
7.2
5.2
n.a.
4.5
6.5%
7.9
6.6
7.4
6.8
6.5
7.6
7.9
4.3
6.4
8.5
6.8
6.4
6.8
8.2
7.2
6.6
5.9
5.9
9.5
7.3
n.a.
5.8
United States
5.2
7.4
9.2
11.8
2
1
2
5.3
7.1
of 24)
Americans spend more o n health care than any other
nation i n the w o r l d , over $650 billion i n 1990. This fact by
itself is not necessarily cause for alarm. The question is:
what are we getting for our money? Unfortunately, the picture that develops upon analysis o f our health-care system
is one of inflated costs, unsatisfactory results, and uneven
insurance coverage.
1
3.8
Rank in
1989 (out
7.6
U.S. Rank
OECD Average (excluding
7.0%
8.2
7.2
8.7
6.3
7.1
8.7
8.2
5.1
8.6
7.3
7.6
6.7
7.4
8.3
7.1
7.6
6.3
6.3
8.8
7.8
3.5*
5.8
17
7
14
4
21
15
3
8
23
5
13
11
18
12
6
16
10
20
19
2
9
24
22
U.S. H e a l t h - C a r e C o s t s :
Highest Among the OECD Countries
Total health-care expenditures as a percentage o f Gross
Domestic Product (GDP) are s h o w n i n Table 8 . 2 1 f o r
countries i n the Organisation for Economic Co-operation
and Development (OECD), an association that includes
most o f the advanced industrialized countries. Available
data are for selected years from I 9 6 0 to 1989. The U.S.
ranks either first o r second in each year o n health-care
expenditures as a percent. I n 1989, we spent 11.8% o f o u r
GDP o n health care, the only nation to reach double-digits,
and well above the OECD average o f 7.6%. Sweden, the
nation ranked second i n the table, spent 8.8% percent o f
its domestic product o n health care, a fiill 3% o f GDP less
than the U.S.
1
Turkey, Portugal, and
Luxembourg)
•This percenlage refers to 1987.
392
393
In 1989, we spent
11.8% of our GDP
on health care,
the only nation lo
reach doubledigits, and well
above the OECD
average of 7.6%.
�TABLE 8.22
The Main Components of Health Care Expenditures
in the U.S., Per Capita, 1960-1990
(1991 Dollars)
Year
Hospital
Care
Physicians'
Services
1960
1970
1980
1989
1990
$225
429
720
994
1,027
Drugs and
Other Medical
Nondurables
$129
208
295
485
504
Percentage Change
355.7%
1960-1990
Annual Average
Growth Rates
1960-1990
Nursing
Home Care
$101
135
152
216
219
$23
76
141
204
214
291.5%
828.5%
4.5%
5.1%
116.2%
2.6%
7.4%
FIGURE 8J
Per Capita National Health Care Expenditures,
1960-1990
1,200
1,000
Hospital Care
800
1
2
M
O
Q
What has made medical spending grow so quickly i n the
U.S.? Table 8.22 and F i g u r e 8J show the per capita g r o w t h
of the four largest components of our health care bill: hospital care, physician services, drugs and other medical nondurables (e.g, bandages), and nursing-home care. As the figure shows, the costs of these services have risen sharply
over time, w i t h the fastest g r o w t h period o c c u r r i n g
between 1970 and the mid-1980s. Hospital care, w h i c h
cost »225 per capita i n real terms i n I 9 6 0 , rose 355.7% by
1990, to * 1,027. Physician services account for the second
largest component of the g r o w t h of health-care expenditures, g r o w i n g by 291.5% between I 9 6 0 and 1989. Nursing-home care, w h i l e representing a relatively small share
of per capita expenditures, has had the most growth over
the period, at an annual average rate of 7.4%. Demographic
pressures—the aging of the p o p u l a t i o n — w i l l continue to
create upward pressure on this component.
An important reason w h y health care is so costly i n our
country has to do w i t h medical inflation. As is w e l l k n o w n ,
w i t h the passage of time, prices rise. However, prices for
certain items rise faster than others; such is the case w i t h
medical services throughout the 1980s. This means that we
are spending more on medical care and getting less for our
money, relative to other goods and services. F i g u r e 8 K
plots the general price index we use to adjust prices
throughout this text, the CPI-U-X 1, against the price index
for medical care. By about 1980, medical prices began to
rise at a faster rate than the prices of the other items (one of
w h i c h is medical care) used to compute the more general
price index. Medical prices continued to accelerate, and as
the figure shows, the price index for medical care has surpassed the general price index since 1983.
As health-care costs have risen, families have had to meet
the rising costs from their budgets (businesses have also
had to contend w i t h rising health-care costs, as discussed
600
Physicians' Services
^
400
Drugs, Non-Durables
200
,
l
t
M ^ * ^ - M U i t i U i t t i
J*UJ**«-
Nursing Home Care
_L
1960
1965
1970
1975
394
-L
1980
1985
1986
1987
J.
1988
1989 1990
395
Hospital care rose
355.7% in real
terms between
1960 and 1990.
�FIGURE 8K
Medical Care Inflation vs. Inflation on All Items,
1969-1991
200
S
150
All Items
0 L-l
1969
below). However, health-care expenditures derive from a
variety of sources. As Table 8.23 shows, families have had
to devote a greater proportion of their income (2.7% more)
to health-care payments between 1980 and 1991. While
fewer of these costs reflect out-of-pocket expenses, a
greater proportion have been shifted to insurance and
health-related taxes. As noted, the overall burden to the
average family grew between 1980 and 1991, from 9 0% to
11.7% of family income (Table 8.23) I n addition, costs in
real terms grew i n each category, w i t h the average total
family expenditure rising by 49.0%, or SI,412. The cost of
Medicare premiums doubled, from $56 to $112 (Medicare
is government sponsored health insurance, primarily for
the elderly); insurance costs grew by 72.4%
1
)
1
1
1
I
I
I
I
I
i
1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991
The overall
burden to the
average family
grew between
1980 and 1991,
from 9.0% to
11.7% of family
income.
TABLE 8.23
The Family Health Care Burden, 1980 and 1991
(1991 Dollars)
Expenditure and
Percentage
Point Changes,
1980 and 1991
Expenditure
Source of Payment
1980
Average Family
$2,885
Expenditure
Out-of-Pocket
1,082
Insurance
429
Medicare Payroll
218
Medicare Premiums
56
General Taxes'
1,099
Average Family Income 33,201
1991
Percent
of Total
100.0% $4,297
100.0%
1,412
0.0
31.7
17.2
8.6
2.6
39.9
11.7"
280
310
151
56
616
4,445
-5.8
2.3
1.0
0.7
1.8
2.7
Percent
of Total
37.5
14.9
7.6
2.0
38.1
9.0"
1,362
739
369
112
1,715
33,646
Point
Expenditure
Changes
Change
Percent
Growth
49.0%
25.8
72.4
68.9
100.0
56.1
13.8
•General taxes include federal contributions to Medicaid, government contributions to employer-sponsored insurance, and funding for other public health programs. Health payments cover the 1991 cost of
all health services and supplies.
"Percent of lamily income spent of health care.
396
397
�TABLE 8.24
Health Indicators in U.S. and OECD Average*
Indicator
Year
U.S.
Rank
(Out
ol 23)
Inlant Mortality Rate*"
1970
1980
1989
1970
1980
1986
15
17
19
12
14
15
1960
1980
1989
1960
1980
1989
12
15
17
16
18
19
Perinatal Mortality Rate**
Female Life Exp. at Birth
Male Life Exp. at Birth
U.S.
Level
OECD
Average
2.00
1.26
0.97
2.30
1.32
1.03
2.05
1.14
080
2.29
1.28
0.99
73.10
76.70
78.50
66.60
69.60
71.50
U.S. Level
as Percent
of OECD
Average
72.90
77.50
79.20
67.80
70.90
72.80
97.5%
110.1
121.3
100.4
103.0
103.9
100.2
99.0
99.1
98.2
98.2
98.2
• Excluding Turkey.
"Infant mortality is deaths of infants less than one year old, as a percent of all live births. Perinatal mortality is fetal deaths occurring after 28 weeks of pregnancy, plus infant deaths occurring prior to 7 days
after birth, as a percent of all live births and fetal deaths.
TABLE 8.25
Infant* and Neonatal** Mortality Rates, by Race, 1960-1988
Blacks
Whites
Years
Infant
Mortality
Neonatal
Mortality
1960
1970
1980
1988
2.29%
1.78
1.10
0.85
1.72%
1.38
0.75
0.54
Life
Expectancy
70.60
71.70
74.40
75.60
Inlant
Mortality
Neonatal
Mortality
4.43%
326
2.14
1.76
2.78%
2.28
1.41
1.15
-3.0%
-4.1
-2.4
- 2.0%
-4.7
-2.5
Life
Expectancy
63.60
64.10
68.10
69.20
Average Annual Rate of Change
1960-1970
1970-1980
1980-1988
- 2.5%
-4.7
- 3.2
-2.2%
-5.9
-4.0
0.2%
0.4
0.2
0.1%
0.6
02
Are We Getting Our Money 's Worth:'.
Lagging Health
Indicators
Clearly the U.S. is devoting a great deal more ol our
national resources to health care than the other Ol t I)
countries. But are we getting our money's worth? Thai is.
given our extensive expenditures, how do our health indicators compare to those of other countries?
Table 8.24 shows that despite our high expenditure
level, the U.S. ranks quite p<x>rly on lour basic health indicaiors: infant mortality rates, perinatal mortality rates, and
male and female life expectancy at birth. I unhennore, our
rankings have fallen over time. Column two shows the
rank of the U.S. out of 23 OliCI) countries, all of whom
spentl a smaller proponion of their national product on
health care. Kcgarding infant mortality rates (the deaths ol
all infants under one year old, as a percentage of all live
births), the U S. ranked 19 out of 23 in 1989, our worst
ranking in the table. The U.S. perinatal mortality ranking
was also high in 1986 (the latest year ol data availability),
and had risen over time. Regarding lile expectancy, the U.S.
rank was 17 for females and 19 for males in 1989 (again,
out of 23). In both eases, the U.S. lile expeciancy—5
years for females, 71.5 for males—wxs below the OliCI)
averages, 79.2 and 72.8, respectively
There also exists a high degree of inequality in health
indicators between whites and blacks within the I i.S Table
8.25 has infant and neonatal (death within the first 28 d.iys
of life) mortality rates, and life expectancy from hirlh for
selected years between 196(1 and I9H8, by race. All of the
health indicators show blacks to have a poorer health status
than whites The black infant mortality rale in I98H al
1.76%, was higher than that of whites, O.H5%. The black
neonatal mortality rate, at 1.15% in 1988 is almost twice
the while rate of 0 54%. Similar results are found tor lile
expectancy; whites live longer, on average, than blacks
Turning to average annual rates of change, both races
show improvement over time, although the rales of
improvement have slowed lx-tween 1980 and 1988. l or
both races, infant and neonatal monality rates improved
lx-tween I960 and 1970, falling at an average annual rate
between 2% and 3% lor both races. The monality rates fell
significantly faster Ixtween 1970 and 1980, lx>th lor white
and black infants; neonatal monality rales lell at an annual
•The number of deaths in the first year ot lite as a percent ol live births.
"The number of deaths in the first 28 days of life as a percent of live births.
399
Regarding inlant
mortality rates
(the deaths of all
infants under one
year old, as a
percentage ol all
live births), the
U.S. ranked 19
out ol23 in 1989,
our worst ranking
in the table.
�rale o f 5 . 9 % f o r w h i t e s , 4 . 7 % l o r b l j c k s . I I O W C M T , the rale
TABLE 8 .26
Number and Percent of the Non-aged* Population
Without Health Insurance, 1987-1990
o f i m p r o v e m e n t .slowed d o w n o v e r t h e l9H<)s
O v e r the
1980.S, the d e c l i n e i n t h e w h i l e i n f a n t m o n a l i t y rale s l o w e d
f r o m its I97()s average g r o w t h rate o l - H t o -.V2% j x r
Non-aged Population
1987
1988
1989
1990
Number Uninsured
(Million)
30.7
32.4
33.0
34.4
Percent Uninsured
14.4%
15.1%
15.3%
15.7%
year; f o r blacks, t h e rate s d e c l i n e s l o w e d f r o m - i I "„ l o
-2.4%.
Health Insurance
Coverage
Incomplete
ami
Uneven
Despite t h e fact that w e i n t h e ILS s p e n d .1 gre.il i l e a l o n
h e a l t h care, m a n y A m e r i c a n s are w i t h o u t h e a l t h i i i M i r a n c c
•Persons under 65.
a n d are t h u s e x p o s e d t o great m e d i c a l a n i l t i n a n i ial risks
TABLE 8.27
Breaks in Health Insurance During a 28-Month Period, 1985-1987
Group
All
Ethnicity
White
Black
Hispanic*
Age
Under 18
18-24
25-44
45-64
65 and over
Ratio of Income to Poverty Line
Under 1.0
1.0 to 1.99
2.0 to 2.99
3.0 to 3.99
4.0 to4.99
5.0 to 5.99
6.0 and over
Work Experience
Worked Full-Period, Full-time
Worked Full-Period, Part-Time
Worked with Interruptions
Percent
Covered
Less Than
28 Months
28.1%
Percent
With No
Coverage Over
28 Months
4.3%
A p p r o x i m a t e l y 3S.4 m i l l i o n A m e r i c a n s , I 1 V A , o l t h e p o p u
lation, were w i t h o u t
health insurance
adults, c h i l d r e n , m i n o r i t i e s ,
and
in
workers
1991
wilh
Young
gaps
in
e m p l o y m e n t are t h e g r o u p s m o s t l i k e l y l o lack c o v e r a g e .
Also, w o r k e r s i n the s e r v i c e s e c t o r a n d i n s m a l l f i r m s are at
greater risk o f b e i n g u n i n s u r e d
As s h o w n i n T a b l e 8 . 2 6 , b o t h t h e n u m h e r a n i l p e r c e n t
age o f t h o s e l a c k i n g h e a l t h i n s u r a n c e has g r o w n
in the
recent past ( e a r l i e r data are not c o m p a r a b l e t o these most
26.4
37.7
52.0
4.0
5.9
11.3
34.5
51.9
27.7
19.9
0.7
5.1
6.0
45
43
0.1
55.3
45.7
29.9
19.2
13.3
11.9
8.9
14.8
8.5
30
1.7
0.8
0.6
0.5
140
25.4
44.7
n.a.
n.a.
n.a.
recent f i n d i n g s ) . The f i n d i n g s i n I h e l a h l e apply to t h e n o n
e l d e r l y ( u n d e r 6 5 ) p o p u l a t i o n , s i n c e these p e r s o n s are not
e l i g i b l e for M e d i c a r e
In
I9H7,
l i t " , of the
non-elderly
p o p u l a t i o n w e r e u n i n s u r e d . T h i s p e r c e n t a g e has
steadily, a n d r e a c h e d
grown
15. VA., o r .VVO m i l l i o n p e r s o n s by
1 9 8 9 , t h e last n o n - r e c e s s i o n a r y year o f data a v a i l a b i l i t y
T h e data i n Table 8 . 2 6 are f r o m a ' p o i n t i n t i m e " analysis,
i.e., s u c h data i d e n t i f y p e r s o n s w h o l a c k e d h e a l t h i n s u r a n c e
w h e n that p a r t i c u l a r s u r v e y w a s a d n i i n i s t e r e d
However,
m a n y p e r s o n s e x p e r i e n c e gaps i n t h e i r i n s u r a n c e o v e r a
g i v e n t i m e p e r i o d , p o s s i b l y d u e t o t h e loss o f a j o b that |>ro
v i d e d i n s u r a n c e o r a t e m p o r a r y loss o f i n c o m e T a b l e 8 . 2 7
s h o w s t h e result o f a s t u d y that t r a c k e d t h e h e a l t h insurance c o v e r a g e o f a r e p r e s e n t a t i v e g r o u p o f p e r s o n s o v e r a
2H m o n t h p e r i o d , b e t w e e n
I98S
and
IW7
Over
JH "„
e x p e r i e n c e d a g a p o f at least o n e m o n t h i n t h e i r cov erage at
s o m e p o i n t o v e r I h e 2 8 m o n t h p e r i o d , s h o w i n g that
the
p o i n t i n t i m e analysis u n d e r s t a t e s t h e e x t e n t o f t h e p r o b
l e n i O n t h e o t h e r h a n d , o n l y -t-.V*'. o f t h e o v e r a l l p o p u l a t i o n w a s u n i n s u r e d l o r t h e e n t i r e p e r i o d , s u g g e s t i n g that
l o n g t e r m lack o f c o v e r a g e is not a p r o b l e m l o r most p e o p l e
•Persons ol Hispanic origin may be ol any race
mi
Approximately
35.4 million
Americans,
14.1% ol the
population, were
without health
insurance in
1991.
�TABLE 8.28
Percent of U.S. Population Obtaining
Health Insurance From Specified Sources, 1990
Family
Member's
Own Job
Job
Medicare
29.4%
29.7%
Other
6.8%
11.2%
Medicaid
Uninsured Total
9.0%
13.9%
100.0%
TABLE 8.29
Persons without Health Insurance: Their Attachment
to the Workforce, 1988-1990
1988
Total*
(000)
Non-worker
Full-Year/Full-time Worker
Part-time Workers'*
6,017
12,973
17,840
Total
36,830
1989
Percent of
Uninsured
Total
(000)
16.3% 4,975
35.2
11,573
16,837
48.5
100.0
33,385
1990
Percent of
Uninsured
14.9%
34.7
50.4
100.0
'Other than non-worker category, totals include workers and their dependents.
"Includes full-year/part-time, part-year/full-time, and part-year/part-time workers.
402
Total
(000)
4,958
17,059
12,612
34,629
Percent of
Uninsured
14.3%
49.3
36.4
100.0
Those most likely to experience a gap in health insurance coverage tended to be minority, young, low-income,
and with a tenuous connection to the workforce. More than
half of Hispanics (52.0%) and 37.7% of blacks were covered for less than the full period; 11.3% of Hispanics had no
coverage at all for the full 28 months. For young adults
between the ages of 18 and 24, 51.9% experienced a gap in
coverage, while 6.0% were without insurance over the full
period. Poor persons (those whose ratio of family income
to the poverty line was less than one) were the most likely
to be without coverage, both for at least one month
(55.3%) and for the entire 28 months (14.8%). Those
whose income to poverty ratio was above 4.00, were less
likely to be without coverage; less than one percent were
without coverage for the whole period.
The last panel of Table 8.27 shows that the likelihood of
experiencing a gap in coverage rises for those who either
worked part-time or with employment interruptions. Table
8.28 confirms that most persons ( 5 9 1 % ) receive their
health insurance either through their own or a family
member's job. The rest of the insured population (27.0%)
depend on government sponsored or privately purchased
insurance.
However, having a job, even a full-time job, does not
guarantee the provision of health insurance. In fact, as
shown in Table 8.29, in 1990 almost one-half of the uninsured had some connection to the full-time/full-year workforce, and a minority of uninsured persons in each year
were non-workers. The table shows the numbers and percentages of workers and their dependents without health
insurance by their attachment to the labor force,
1988-1990. In 1988, 35.2% of the uninsured were lull-year,
full-time workers or their dependents. By 1990, this percentage had risen by 14.1 points, to 49.3%. Of the population of persons lacking health insurance in 1990, approximately 29.7 million persons, 85.7% of the uninsured, had
some connection to the workforce.
403
Having a job Q
not guarantee the
provision of
health insurance
...in 1990
almost one-half
the uninsured,
some connection
to the full-time/
full-year
workforce.
�TABLE 8.30
Health Insurance Coverage of Workers, b Industry, 1989
y
Percent Receiving Insurance from:
Industry
Own Job
Other
Source
Not
Insured
Total
Agriculture, Forestry, and
17.8%
54.0%
28.2%
100.0%
22.7
52.8
24.5
100.0
29.0
31.8
41.0
43.9
49.4
47.9
37.5
28.8
21.6
20.3
21.5
27.3
100.0
100.0
100.0
100.0
55.6
60.3
35.7
26.5
8.7
13.2
100.0
100.0
61.8
29.7
8.5
100.0
68.8
18.8
12.4
100.0
73.5
73.9
76.0
77.1
Fisheries
Personal Services, Including
Household
Entertainment and Recreation
Services
Retail Trade
Business and Repair Services
Construction
Professional and Related
Services
Wholesale Trade
Finance, Insurance, and
fleal Estate
Manufacturing, Nondurable
Goods
Transportation, Communications,
and Public Utilities
Public Administration
Mining
Manufacturing, Durable Goods
16.6
21.1
14.1
14.2
9.9
5.0
9.9
8.7
100.0
100.0
100.0
100.0
TABLE 8.31
for
Employer Health Care Burden the Average Business, 1980 and
Percent
of Total
Healthcare
1980
Insurance
Medicare Payroll
Tax
General Taxes*
Other"
Total Burden
71,605
19,437
21,630
9,930
$122,602
58.4
15.9
17.6
8.1
100.0%
1991
131.360
39.204
47,965
19,112
$237,641
Percent
of Total
55.3
16.5
20.2
8.0
100.0%
Expenditure
Change
1991
Point
Change
Percent
Growth
59,755
-3.1
83.5
19.767
26,335
9,182
$115,039
0.6
2.6
-0.1
0.0
101.7
121.8
92.5
93.8%
'General taxes include federal contributions to Medicare and Medicaid, government contributions to employer-sponsored insurance, and other public health programs.
One factor contributing to the high rates o f those w h o
are employed and uninsured is the job growth in industries that are less likely to provide their employees w i t h
health insurance. Table 8.30 has the percentages o f covered workers by industry, for the year 1989. Service sector
firms, such as those in personal services, entertainment
and recreation services, and retail trade have low rates of
coverage (22.7%, 29 0%, and 3 1 8 % , respectively) and correspondingly high rates o f uninsured workers (24.5%,
21.6%, and 20.3%). As noted in Chapter 3, the employment share of these industries rapidly expanded over the
1980s. M a n u f a c t u r i n g , an i n d u s t r y w i t h a s h r i n k i n g
e m p l o y m e n t share, has t r a d i t i o n a l l y h i g h rates o f
c o v e r a g e — 7 7 . 1 % f o r d u r a b l e goods m a n u f a c t u r i n g i n
1989, the highest rate of coverage in the table.
Another factor influencing the lack of coverage through
employment is the rising health-care burden faced by businesses. Whereas Table 8.23 examines the family health-care
burden, Table 8.31 shows a similar analysis for businesses.
Between 1980 and 1991, the total real cost of health-care
provision rose by 93 8% for the average business, from
$122,602 to 1237,641. The largest g r o w t h came from the
g r o w t h of health-care-related taxes, w h i c h grew 121.8%,
and from increased contributions to the Medicare program,
w h i c h grew 101.7%. Insurance costs, while representing a
smaller share of the average 1991 business health-care burden (compared to 1980), grew by 83 5%. F i g u r e 8L plots
the g r o w t h of the total cost to employers of health insurance provision, 1965 to 1990. I n real 1991 dollars, these
costs grew from $23 7 b i l l i o n i n 1965 to S144.8 billion in
1990. The rising health-insurance expenditures reflect
both the expanding size of the insured workforce and the
rising costs of health insurance.
The health-care burden on families and businesses is
greater in the United States than in the other OECD countries primarily due to our lack of public commitment to
health-care provision. Table 8.32 gives the percentage of
public health expenditures i n the 24 OECD countries, in
1989. At 4 2 % , the U.S. has the second lowest level of public
expenditures as a percent of all health care spending. In
Germany, public spending accounts for 72% of its health
care budget; i n France, 75%. Countries w i t h well-developed national health plans have an even larger public commitment: the Swedish public sector covers 90% of that coun-
"Worker's compensation, temporary disability, and industrial health benefits.
405
404
One factor
contributing to the
high rates of
those who are
employed and
uninsured is the
job growth in
industries that are
less likely to
provide their
employees with
health insurance.
�FIGURE 8L
Employer Spending on Private Health Insurance,
1965-1990
TABLE 8.32
Public Health Expenditures as a Percentage of
Total Health Expenditures, OECD Countries, 1989
Country
1965
Public
Spending
1989
406
1960
1985
1989
1990
70%
67
89
75
84
79
75
72
89
88
84
79
73
92
73
85
95
62
78
90
68
37
87
42
Average
1975
1970
Australia
Austria
Belgium
Canada
Denmark
Finland
France
West Germany
Greece
Iceland
Ireland
Italy
Japan
Luxembourg
Netherlands
New Zealand
Norway
Portugal
Spain
Sweden
Switzerland
Turkey
United Kingdom
United States
76
407
�FIGURE 8M
Women in the Labor Force by Presence of Children,
1947-1991
70
Women with Children
60
Child Care: A Constraint on Working
Parents
-
/
so -
Women without Children
0)
40
30
-
20
10
J
tryTs health costs; the United Kingdom, 87%. Furthermore,
it cannot be argued that the smaller proportion of private
expenditures compromise the quality of health care in
these countries with a greater public commitment. Each
one of the five countries cited above surpasses the U.S. on
the health indicators listed in Table 8.24.
L
J
I
I
I
I
I
I
L
J
L
J
TABLE 8.33
Percent of W m n with Children in the Labor Force,
o e
by Age of Child, 1950-1991
Age of Child
Year
Total
1950
1960
1970
1980
1990
1991
21.6%
30.4
42.4
56.6
66.7
66.6
408
Under Age
6
13.6%
20.2
32.2
46.8
58.2
58.4
Age 6 to
17
32.8%
42.5
51.6
64.3
74.7
74.4
L
Earlier in this text, we documented the increasingly
important role played by working spouses in maintaining
family incomes. As more mothers enter the workforce,
there has been a growing need for an essential corollary to
mothers' workforce participation: child care. Along with
this increased demand are a variety of concerns regarding
the availability, the quality, and the costs of child care. Lack
of affordable, high-quality child care can be a constraint,
keeping mothers either out of the workforce or working for
fewer hours than they would prefer (i.e., underemployed),
or simply cutting into a family's resources and adding stress
to family life. This constraint can be particularly severe for
low-income families, who both need the extra earnings,
and tend to earn less than better-educated, more experienced women. In fact, we w i l l show that these low-income
families tend to spend a higher proportion of their incomes
on child care than wealthier families, and are therefore left
with fewer resources to pay for other necessities.
Figure 8M demonstrates the important social and economic trend alluded to above: the dramatic increase of
mothers in the labor force. Although a larger percentage of
women without children than mothers worked (or sought
work) in 1947, by 1970 the situation began to reverse.
Over the course of the 1980s, the percentage of working
mothers continued to grow, reaching two-thirds by 1991
(Table 8.33). Those mothers with children over age 6 have
had the highest participation rates, as close to three-quarters of them were in the labor force in 1991. Clearly, a
shrinking minority of mothers has come to provide lulltime care for their children.
Since the primary responsibility for child care typically
falls on women, research has focused on the impact of
child-care constraints on mothers. This research has found
409
Mothers out-ofwork due to child
care were
disproportionately
poorly educated,
lowering their
potential to earn a
wage that might
have allowed
them to purchase
the very child
care they needed
to go out and
work. Thus, they
were stuck in a
vicious cycle.
�TABLE 8.34
Child-care Problems as a Constraint on Work,
W m n 21-29 Years Old, 1986
o e
Main Reason for Work Status
Child-care Problems
Number
(000)
Work Status
Never Employed
Other Problems
Number
(000)
Percent
Percent
65.5%
1,519
383
34.5
2,098
58.0
1,109
100.0
3,617
100.0
726
Under-Employed*
Total
42.0%
"Employed less lhan (ull-time or full-year even though preferring more hours of work.
TABLE 8.35
Mothers, 21 to 29 Years Old, Not in the Labor Force,
By Selected Characteristics, 1986
Total,
All Mothers
Not in the
Labor Force
(000)
Characteristics
Total
Mothers Not in the Labor Force
Because of Child-care Problems
Number
(000)
As a Percent of All
Mothers Not in the
Labor Force
4,726
1,109
23.5%
433
860
3,433
176
250
683
40.6
29.0
19.9
1,854
1,912
961
376
428
305
20.3
22.4
31.8
1,217
2,763
414
509
34.0
18.4
1,157
3,547
372
727
32.2
20.5
Race or Hispanic Origin*
Hispanic Origin
Black
White
Total Number of Children
One
Two
Three or More
Poverty Status
At or below Poverty Line
Above Poverty Line
Educational Attainment
No High-School Diploma
High-School Graduate
that for over one m i l l i o n mothers, difficulties attributable
to child care kept them either under- or unemployed
("under-employed" refers to less than full-time/full-year
workers w h o w o u l d have preferred to w o r k more hours).
Table 8.34 shows the result of a national survey that asked
mothers for the main reason w h y they were not w o r k i n g or
were underemployed at any period i n 1986, the year the
survey was conducted. For mothers between the ages of 21
and 29, 1,109,000 answered that child-care problems
restricted their labor force participation. The majority of
these women (65.5% ) never worked at all i n the survey
year and 34.5% were underemployed.
As Table 8.35 shows, these w o m e n w h o were struggling
to balance employment w i t h day-care constraints, represent almost one-quarter o f all mothers of this age range not
i n the 1986 labor force. Mothers w h o were kept out of the
workforce by child-care problems i n 1986 tended to be
poor, minority, and w i t h large families ( 3 or more children> O f those mothers not i n the workforce because of
child care, 34.0% were poor (about twice as many as those
w h o were non-poor). Among non-working m i n o r i t y mothers, 40.6% of Hispanics and 29 0% of blacks reported childcare problems as keeping them out o f the workforce. The
percentage for whites, 19-9%, was substantially lower. Furthermore, mothers out-of-work due to child care were disproportionately poorly educated, lowering their potential
to earn a wage that might have allowed them to purchase
the very child care they needed to go out and work. Thus,
they were stuck i n a vicious c y c l e
Child care can also present problems for w o m e n w h o are
working, from lateness and absenteeism to j o b loss. Table
8.36 shows that i n 1988 a total of 3 7% of mothers missed
work i n a given month due to child-care problems. As might
be expected, that proportion is higher for families w i t h
infants i n child care ( 5 3%), since infants are likely to
require more parental care. Wealthy parents (those earning
an annual income of 550,000 or more i n 1988), also lost
more work than average, suggesting that child care is not
exclusively a low-income problem. The final column of
Table 8.36 shows that 2.3% of w o r k i n g mothers i n 1988
actually left the labor force (voluntarily or otherwise) due
to child-care problems. Again, loss of w o r k is more common
for women w i t h infants ( 3 0% ), as w e l l as for w o m e n from
low-income families.
'In this table, race and Hispanic origin are mutually exclusive categories.
410
411
Wealthy parents
(those earning an
annual income of
$50,000 or more
in 1988), also lost
more work than
average,
suggesting that
child care is not
exclusively a
low-income
problem.
�TABLE 8.36
Child-care Related Constraints on Women's Employment, 1988
Mother (23-31)
Missed Work w/in
Last 4 Weeks Due to
Child-care Problems
Mother (21-31)
Dropped Out of
Labor Force Because of
Child-care Problems
Total
3.7%
2.3%
Age of Youngest Child
Birth to 1 Year
2-4 Years
5 Years
5.3
2.8
3.4
3.0
2.0
2.0
Net Family Income
(in 1988 Dollars)
$0-14,999
$15,000-24,999
$25,000-49,999
$50,000 and more
3.7
3.4
3.1
4.7
4.2
2.6
1.6
0.3
FIGURE 8N
Distribution of Child-Care Choices for Preschool Children,
1965-1990
Child-Care Arrangements
Even in the context o f the child-care induced constraint
o n women's labor force participation, Figure 8M shows
that mothers have continued to enter the workforce. Families continue to find ways to see that their children are
cared for when n o family caretakers are available. The following section examines the different types of child-care
arrangements, h o w they are utilized, and what they cost.
Child care is provided i n a number of formats: child-care
centers, family day-care homes (care for a small number of
children in a care-giver's home), in home care (care pro
vided by a non-relative i n the child's home), and care provided by a relative. F i g u r e 8 N shows the trend in the use of
these different venues i n selected years since 1965, for preschoolers (under age 5 ) in families w i t h an employed
/ mother. Each bar represents the proportion of families
using a specific arrangement i n the given year.
One of the most pronounced trends i n Figure 8 N is the
increased proportion of families reliant on child-care centers, from 6% in 1965 to 28% i n 1990, resulting from the
increased proportion of w o r k i n g mothers w i t h young children (see Table 8.33, column two). An associated trend is
the fall in in-home care, from 15% to 3% over the period,
and the fall i n care provided by a relative, from 33% to
19%. A surprising trend is the recent increase in parent
care. Given the k n o w n increase i n labor force participation
of mothers, this trend probably reflects the increased
involvement of fathers in child care, as w e l l as more flexible
w o r k scheduling by both parents.
Regarding availability, the demand for child care appears
to have led to a g r o w t h in providers, although there is little
reliable data on nation-wide availability. Nevertheless, the
anecdotal evidence suggests that many parents have difficulty f i n d i n g affordable, high-quality child care (costs are
discussed below). A 1990 p o l l of mothers whose children
were i n child-care centers reported that more than half said
they w o u l d have chosen another program had one been
available. Infant care, w h i c h is both more costly and more
labor intensive than average, appears lo be in particularly
short supply. From the late 1970s to 1990, the proportion
of children under one year old in child-care centers went
from 1% to 4 % ; for one- to two-year-olds, the proportion
grew from 3% to 5%. In New York City in 1987, licensed
providers could accommodate only 6% of the children
1977
1965
412
1982
1985
1990
413
In New York City
in 1987, licensed
providers could
accommodate
only 6% of the
children aged two
or younger who
needed care.
�TABLE 8.37
Average Percentage of Family Income Spent on Child Care,
By Family Income, 1990*
(1990 Dollars)
Percent
Paying
For Care
Percent
of Income
Spent on Care
57.0%
$63.2
10.4%
n.a.
n.a.
Total
Average
Weekly Cost
37.3
65.5
23.2
8.7
37.9
50.7
50.7
64.5
85.1
24.8
13.2
8.8
10.2
6.2
Poverty Status
Below Poverty
Above Poverty
Annual Family Income
Less than $15,000
$15-24,999
$25-34,999
$35-49,999
$50,000 or more
42.0
48.0
52.0
54.0
68.0
'Data is for employed mothers with youngest child under five.
aged t w o or younger w h o needed care In Connecticut i n
1990, fewer than one-third of the licensed centers or family
day care homes w o u l d accept infants.
There is also a lack of hard data on the quality of the
child care provided i n the various settings noted above.
Although those centers and family day-care homes that are
licensed are required to meet state regulations, the Children's Defense Fund estimates that as many as 43% of the
children i n child-care situations outside their homes are in
unregulated settings. Such settings are likely to be understaffed, w i t h poor health standards and poorly trained
staffs.
Moreover, simply the fact that a child-care setting is regulated does not insure quality care. Research on individual
sites has found that i n certain locales, regulations are so lax
that even licensed providers deliver below-standard care. A
study on licensed centers i n Atlanta, Georgia found preschoolers to be ignored 79% o f the time; infants, 6 1 % of
the time. Other states' regulations do not require children
to be immunized against contagious diseases, or allow for
excessively l o w staff to child ratios. A national study exami n i n g trends between the mid-1970s and 1990 found that
the average n u m b e r o f c h i l d r e n e n r o l l e d i n centers
increased by 39% over that period, while staffs increased
by an average 25%. This trend has lowered the staff to child
ratio in licensed centers. In addition, thirty-five states
require no training for child-care staff members.
The Costs of Child Care
The costs of child care for employed mothers w i t h preschoolers are examined in Table 8.37. Most families w i t h
children under five ( 5 7 % ) were paying for child care i n
1990, and wealthy families were paying a smaller percentage of their income than middle- and low-income families.
In fact, low-income families spent as much as one-quarter of
their income for child care, significantly constraining their
family budgets.
The l i k e l i h o o d of purchasing c h i l d care rises w i t h
i n c o m e b o t h because w e a l t h i e r f a m i l i e s have more
resources to purchase child care, and because they w o r k
more hours. Less than half ( 4 2 % ) of the families w i t h lowincomes purchased child care w h i l e over two-thirds ( 6 8 % )
of wealthy families did so i n 1990. Although families w i t h
414
415
For a single
parent working
full-time at the
minimum wage,
with a one-yearold, child care
uses between
44.4% and 69.6%
ot their income.
�TABLE 8.38
Child-Care Cost Burden, as Percentage of Income for Low- and Middle-Income
Families: Licensed Centers in Four Cities, 1990
One-year-old Child
Low Income'
Four-year-old Child
Middle Income"
Low Income'
Middle Income"
Single
Cities
Two
Single
Two
Single
Two
Single
Two
Parent
Parents
Parent
Parents
Parent
Parents
Parent
Parents
Oakland, CA
60.8%
32.8%
27.5%
13.7%
50.9%
27.5%
23.0%
11.5%
Boulder, CO
69.6
37.5
33.7
17.0
47.1
25.4
22.8
11.5
Dallas, TX
41.1
22.2
22.4
11.4
35.6
19.2
19.4
9.9
Orlando, FL
44.4
23.9
24.4
12.5
32.9
18.1
18.1
9.3
'One (in the case ol single parents) or both parents working full-time/full-year at the minimum wage.
"One (in the case of single parents) or both parents working full-time/Jull-year at Ihe state-specific
median wage.
incomes over J50,000 spent more i n absolute terms on
child care than other families, their expenditures, on average, amounted to only 6.2% of their income. Conversely,
the least w e l l - o f f families, those w i t h incomes under
$15,000, spent 24.8% of their annual income on child care.
I n addition, costs vary by regional location, by the age o f
the child under care, and by child-care setting. Table 8.38
has the percentage o f family income required to send a
one- o r four-year-old child to a licensed center i n four different cities in 1990, for one- and two-parent families. For
a single parent w o r k i n g full-time at the m i n i m u m wage,
w i t h a one-year-old, child care is clearly prohibitive, using
u p between 44.4% and 69 6% o f their income. The situation is not much better if the child is four (thereby requiring less expensive care); the expenditure still calls for
between 32.9% and 50.9% o f income. Even for middle-income single parents (working full-time at the median wage
in that state) w i t h a one-year-old, child-care costs between
22.4% and 33-7% o f income. For two low-income parents,
child-care costs require approximately one-fifth to onethird o f family income. Only for two middle-income parents, both w o r k i n g full-time, d o child-care expenditures
potentially leave enough income left over for other necessities. For such a family in Dallas o r Orlando, child care for a
four-year-old requires slightly less than 10% o f family
income. However, if this family had a one-year-old in Boulder, the e x p e n d i t u r e w o u l d call for 17.0% o f family
income.
In sum, as the proportion of mothers entering the workforce has risen, so has the use of child care become a necessity for the majority of families w i t h young children. While
the supply of child-care services has g r o w n i n response to
the demand, anecdotal data suggest that there are problems
w i t h the availability and quality of child care. Furthermore,
expenditures on child care are regressive: though lowincome families pay less in absolute terms than wealthy
families, in relative terms (proportion of income) they pay
substantially more. Finally, child care can serve as a constraint on the workforce participation of mothers Both low-
416
417
For two
low-income
parents, childcare costs require
approximately
one-fifth to
one-third of family
income.
�Chapter 9-
Conclusion
Our examination of those services vital to a rising standard of living reveal a trend of decreased access for most
families. Our education system is found to be two-tiered:
disadvantaged students score poorly on tests of basic skills
compared to their more wealthy counterparts. Further,
blacks and Hispanics account for well under 10% of the
Our examinationadvanced degrees in 1989 (Table 8.9). The housing market
over the 1980s has been
as a
of those servicesproportion of income forcharacterized by rising costsmarpersons trying to enter the
vital to a rising ket, leading to home ownership rates that fell as much as 8
standard of livingpercentage points for young families (Table 8.17). Regardreveal a trend of ing health care, the American system is the most cosdy in
decreased access world, yet our health indicators show us behind most
the
for most famUies.
industrialized countries. Our health-care system also leaves
a substantial number of persons uninsured or partially
insured. Finally, the sharp rise of mothers in the workforce
has led to a growing need for child care. However, the
costs of child care are prohibitive for many mothers; even
middle-class families report that child care is a constraint
on their workforce participation
International
Comparisons:
The United States
is Falling Behind
Introduction
In this chapter we examine how American workers and
their families have fared in comparison to their foreign
counterparts. While Americans still have the highest per
capita income as measured by relative purchasing power,
many indicators of economic well-being still show the U.S.
to be falling behind. Compared to other countries, U.S.
productivity growth has been relatively stagnant. Likewise,
wage growth (which is related to productivity) has been
sluggish, while other countries have seen strong growth.
Finally, though the U.S. economy seemed to have a faster
rate of job growth than other countries over the 1980s,
taking account of population growth, other countries have
had higher rates of job creation.
The growth of U.S. income inequality documented
throughout this text is also notable in an international context. Poverty in the U.S. is both higher and more persistent
than in other countries. Furthermore, the U.S. tax and transfer system has become less effective at poverty reduction
over the 1980s, while the systems of other countries have
grown more effective.
Incomes and Productivity: Sluggish
Growth i n the U.S.
One widely accepted measure of living standards is per
capita Gross Domestic Product (GDP): the total value of
418
419
While Americans
still have the
highest per capita
income as
measured by
relative
purchasing power,
many indicators
of economic
well-being still
show the U.S. to
be falling behind.
�TABLE 9.1
Per Capita G P Growth in Ten Countries, 1979-1989,
D
Broken D w into Growth of Employment and Productivity
on
Growth Rate
of Real
Per Capita
GDP,
1979-1989
Country
Australia
Canada
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
1.8%
2.0
1.6
1.6
2.2
3.3
1.0
1.8
2.1
1.8
Weighted Average*
2.1
Percent of Growth
Due to
More
Workers
Higher
Productivity
48%
42
-24
34
9
16
74
34
19
40
52%
58
124
66
91
84
26
66
81
60
24
76
Total
100%
100
100
100
100
100
100
100
100
100
100
'Weighted by each country's population, which is taken to be the average of the populations of 1979
and 1989.
420
goods and services produced in the domestic economy
divided by the population. Growth o f per capita GDP is
considered an essential component o f rising living standards. Such growth comes from t w o sources: greater output per w o r k e r (i.e., higher productivity), and sending
more workers into the workforce (i.e., expanding the ratio
o f employment to the population). Since sending more
workers into the workforce (and/or w o r k i n g longer hours)
can lower living standards, it is important to decompose
GDP growth into these t w o factors. Table 9.1 does so by
examining the growth o f GDP over the 1980s for the seven
major industrialized countries, as well as for three others
for w h o m there are comparable data.
Column one of Table 9-1 shows the annual growth rates
of real per capita GDP for the years 1979-1989. Although
GDP per person i n the United States rose over the decade, it
grew at a below average pace ( 1.8% versus 2.1%), meaning
other countries' per capita incomes were rising faster than
ours. At 1.8%, the g r o w t h rate for the U.S. was below the
growth rates sustained by Canada (2.0% ), Italy (2.2% ), and
the United K i n g d o m ( 2 . 1 % ) , and was almost half of Japan's
rate ( 3 . 3 % ) .
Of equal importance is whether per capita growth of GDP
is generated through higher productivity or through an
increase in the proportion of the population in the workforce In general, if productivity growth is strong and wages
rise accordingly, families can live better without sending
more workers into the workforce; if it is weak, living standards can rise only if more people work (or those working
put in longer hours). Columns two and three of Table 9.1
show that a relatively large part of our per capita income
growth ( 4 0 % ) comes from more workers—increasing the
proportion of the employed population. Only three countries
have a higher percenuge of their growth rate coming from
this source. The U.S. rate of growth attributable to more workers is also w e l l above the ten country average of 24% Conversely, the U.S. percentage of growth attributable to higher
productivity ( 6 0 % ) is well below the average of 76%.
421
Although GDP per
person in the
United States rose
over the decade,
it grew at a below
average pace,
meaning other
countries' per
capita incomes
were rising faster
than ours.
�TABLE 9.2
Living Standards vs. Productivity Growth, 1979-1989
Indices of 1979-89 Growth Rates
(Weighted Average = 100)
Real Per
Capita GDP
Productivity
(GDP/Worker)
Australia
Canada
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
88
98
77
77
105
159
46
86
99
86
61
75
127
67
126
176
16
76
106
67
Weighted Average*
100
100
Country
•Weighted by each country's population, which is taken to be the average ol the populations of 1979
and 1989.
TABLE 9.3
Productivity Growth Rates, 1960-1989
Country
Australia
Canada
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
1960-1967 1967-1973 1973-1979 1979-1989
n.a.
2.7%
5.1
4.6
7.1
9.1
n.a.
4.7
2.7
3.1
2.8%
2.6
4.2
4.3
5.0
7.7
n.a.
2.8
3.3
1.0
1.7%
1.2
2.4
2.9
2.8
3.4
0.0
0.6
1.3
-0.1
1.0%
1.2
2.0
1.1
2.0
2.8
0.2
1.2
1.7
1.1
Cumulative
1960-1989*
Table 9.2 provides a direct comparison of per capita
GDP and productivity g r o w t h rates. Column one shows that
the U.S. per capita income grew at 86% of the average rate,
as suggested by the data i n Table 9.1. However, our productivity rate grew at an even slower relative pace, at 67% of
the average. Japan, o n the other hand, saw its per capita
income grow at 159% of the average rate, fueled by productivity g r o w t h that was 176% of the average. I n fact, only
t w o countries—Australia and the Netherlands—posted
lower productivity g r o w t h rates than the U.S.
The above tables have emphasized the importance of
worker productivity (GDP per worker) to our standard of
living. Table 9.3 presents the average annual g r o w t h rates
of this important statistic for various periods, 1960-1989.
The g r o w t h rates for the U.S. are relatively low; for the
cumulative period the U.S. rate is the second lowest i n the
table. U.S. productivity was at its highest between I 9 6 0 and
1967, when it grew at an annual rate o f 3 . 1 % . Since that
time, it has fallen precipitously and was even negative
between 1973 and 1979The rates for other countries have also slowed d o w n over
the period, though all but the g r o w t h rate o f the Netherlands remain higher than that o f the U.S. ( F i g u r e 9A).
Japan's productivity fell from an average annual rate of
9 . 1 % between I 9 6 0 and 1967, to 2.8% over the most
recent period. Similarly, West Germany's rate fell from 4.6%
to 1.1%. However, even i n the context o f this general slowd o w n , the majority o f the other countries still had faster
productivity growth rates than those o f the U.S. since 1973-
1.7%
1.8
3.1
2.8
3.8
5.1
0.2
2.1
2.1
1.2
•Netherlands: 1973-1989; Australia: 1967-1989.
422
423
Only two
countriesAustralia
and the
Netherlandsposted lower
productivity
growth rates
than the U.S.
�"Wages: Falling for Production Workers
F I G U R E 9A
Productivity Growth Rates,
Given that wage growth is in part dependent on productivity growth, we would expect to see US wages reflect
the relatively stagnant growth of productivity presented
above. In fact, Table 9.4 and Figure 9B reveal that the U.S.
annual growth rate for hourly manufacturing compensation
(wages plus fringe benefits) is well behind that of most
comparable nations. (The only comparative data are from
the manufacturing sector )
Between 1979 and 1989, real hourly manufacturing
compensation in the U.S. grew at an average annual rate of
0.2%. Only Denmark had a lower growth rate, at 0.1%.
Meanwhile, compensation in West Germany grew at an
average rate of 2.4% a year, ten times the U.S. growth rate.
Furthermore, what little growth U.S. manufacturing wages
experienced benefitted only supervisors and managers.
1960-1989*
2
3
What tittle
growth U.S.
manufacturing
wages
experienced
benefitted only
supervisors
managers.
Production
workers saw thei
compensation
at a rate of 0.6%
a year.
4
Annual Growth Rate (Percent)
FIGURE 9B
Hourly Manufacturing Compensation Growth,
1979-1989
•Nelhertands. 1973-1989,Auslmlia: 1967-1989.
TABLE 9.4
Hourly Manufacturing Compensation Growth,
1979-1989
West Germany
United Kingdom
Japan
Growth Rate of Compensation
in Manufacturing, 1979-1989
France
-
All Employees
0.4%
-0.1
1.7
2.4
0.9
1.8
0.8
0.5
2.0
0.2
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
424
Production Workers
0.5%
0.3
2.1
1.8
0.6
1.3
0.9
0.9
1.7
-0.6
-
Sweden
Country
Italy
Netherlands
-
Canada
-
United States
-
Denmark
0
t
2
Annual Growth Rate (Percent)
All Employees
Q
Production Workers
425
�TABLE 9.5
Manufacturing Hourly Compensation Compared, 1979,1989, and 1990
(Using Purchasing Power Parity Exchange Rates)
Country
1979
1989
1990
81
68
86
82
80
48
96
83
62
100
84
64
94
94
77
55
98
79
69
100
86
65
93
96
79
57
97
78
71
100
68
73
74
82
76
69
94
92
51
95
82
64
100
92
77
85
111
93
61
106
89
75
100
97
80
85
113
94
63
104
90
77
100
71
80
83
All Workers
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
Average*
Production Workers
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
Average*
"Averages are weighted by hours worked; they exclude the United States.
426
Production workers saw their compensation fall at a rate of
0.6% a year. The U.S. was the only country whose production workers experienced wage loss during this period.
Whereas Table 9 4 examined the growth rates of manufacturing compensation, the next two tables compare the
actual levels of compensation, indexed to the U.S. In order
to make such comparisons, we first converted wages from
their national currency into dollars, using two different
exchange rate series: Purchasing Power Parity exchange
rates (PPP's) and market exchange rates. PPP s are currency
values intended to measure the cost of buying the same
"basket" of goods and services in all countries. Market
exchange rates represent the value of the dollar relative to
other currencies in world financial markets. While both
exchange-rate-measures have their drawbacks, the PPP's are
commonly considered an indicator of the relative price of
consumption. Market exchange rates reflect the relative
value of American goods, services (including labor), and
assets in international markets.
When we use PPP's to adjust for currency differences, as
in Table 9.5, U.S. compensation ranks the highest. However, by 1989, France, West Germany, and the Netherlands
had compensation levels that were 94%, 94%, and 98%,
respectively, of the U.S. levels for all workers. Moreover,
other countries gained on the U.S. throughout the 1980s, as
seen by the rise in the "All Workers" averages (averages in
this and the next table exclude the U.S.), from 68% of U.S.
compensation to 74%. Given the findings in Table 9.4
regarding the falling compensation of U.S. production workers, it is not surprising that production workers in two
countries have surpassed the U.S. in this category (Figure
9C shows the PPP adjusted wages). By 1989, production
workers in West Germany and the Netherlands had higher
levels of compensation than their American counterparts,
by 11% and 6%, respectively.
When we use market exchange rates (Table 9-6), the
story changes significantly: American workers no longer
lead the pack. Since 1979, workers in many countries (and
production workers in most countries) have received
higher hourly compensation than U.S. workers. The average
hourly compensation for non-managerial (i.e., production)
workers was only 6% below that of the U.S. in 1989. By
427
By 1989,
production
workers in West
Germany and the
Netherlands had
higher levels ol
compensation
than their
American
counterparts, by
11% and 6%,
respectively.
�FIGURE 9C
Comparative Manufacturing Compensation, 1989
(Using Purchasing Power Parity Exchange Rates)
United States
TABLE 9.6
Manufacturing Hourly Compensation Compared, 1979,1989, and 1990
(Using Market Exchange Rates)
Country
1979
1989
1990
83
105
105
110
69
57
127
127
60
100
93
84
98
104
77
80
100
110
62
100
96
99
113
123
93
77
115
123
77
100
78
85
93
84
117
85
125
79
61
126
126
62
100
103
101
89
123
92
88
109
123
67
100
108
121
103
146
111
86
123
142
84
100
82
94
104
Netherlands
West Germany
All Workers
France
Canada
Sweden
Italy
Weighted Average
United Kingdom
Denmark
Japan
10
1991 Dollars
All Employees
Q
Production Workers
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
Average*
Production Workers
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
Average*
'Averages are weighted by hours worked; they exclude the United States.
428
429
�FIGURE 9D
Comparative Manufacturing Compensation, 1989
(Using Market Exchange Rates)
Sweden
West Germany
Netherlands
United States
France
Canada
Weighted Average
Denmark
Japan
Italy
1990, the U.S. was 4% below the average; production workers in West Germany and Sweden were earning 146% and
142% of the U.S. production worker's earnings, respectively
(Figure 9D shows the wages adjusted by market exchange
rates). As market exchange rates reflect the relative attractiveness of U.S. products in international markets, the comparisons in this table further underscore the charge that
the U.S. is becoming a low-wage country (see Chapter 3).
These wage data make two important points regarding
the economic position of the U.S. in international markets.
Firstly, we have had to let the market value of the dollar
decline in order to offset our competitive decline. This has
caused the earnings of U.S. workers to fall relative to workers of other countries. Secondly, these data suggest that the
trade imbalances we have with many of these countries
cannot be blamed on an overpaid U.S. workforce.
United Kingdom
10
15
Poverty: Highest and Most Persistent
in the U.S.
1991 Dollars
All Employees
430
Q
Production Workers
In discussions of America's relative prosperity, much is
often made of our high per capita income. Table 9-7 shows
per capita income adjusted by PPP's and indexed to the U.S.
Indeed, our per capita income was the highest in 1979 and
1989, although the rising averages (exclusive of the U.S.)
show the other countries to be gaining on us. However, a
per capita income comparison contrasts the "average" person in each country. Given different degrees of income inequality, this average measure reveals little about how persons in the different countries are faring.
To make such comparisons, we need to examine evidence of how income is distributed throughout each economy. In Table 9-8, the after-tax income distributions of nine
industrialized countries are presented, with the percentage
of the poor and near-poor, the middle class, and the wellto-do given for each country. These data, from the
mid-1980s, show the U.S. to have the most unequal distribution, after taxes and transfers. At one end of the scale, we
find that the middle class in the U.S. is the smallest
(53 7%); the low-income class is the largest (24.2%). West
Germany and the Scandinavian countries have much
greater income equality, with relatively large middle classes
and smaller low-income and wealthy classes. At the other
end of the scale, the U.S. has the third largest wealthy class.
431
These data
suggest that the
trade imbalances
we have with
many ol these
countries cannot
be blamed on an
overpaid U.S.
workforce.
�TABLE 9.7
Per Capita Income Compared to U.S., 1973-1989
Per Capita Incomes Adjusted
for Purchasing Power (U.S. = 100)
Country
1973
1979
1989
Australia
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Norway
Sweden
Switzerland
United Kingdom
United States
74
80
76
77
79
63
63
75
58
78
107
70
100
75
88
77
81
85
70
67
77
70
79
97
71
100
74
90
77
79
83
73
78
71
73
79
97
73
100
71
75
78
Weighted Average*
'Weighted by population, excluding U.S.
TABLE 9.8
Income Distribution: International Comparisons,
Mid-1980s
All Persons
Poor and
Near-Poor
Middle
Class
Well-to-Do
Total
Australia
Canada
West Germany
Netherlands
Norway
Sweden
Switzerland
United Kingdom
United States
21.4%
21.0
12.6
14.2
13.2
10.5
15.9
21.4
24.2
56.0%
58.5
70.1
62.5
73.4
79.0
67.2
58.5
53.7
22.6%
20.6
17.3
23.3
13.4
10.5
16.9
20.2
22.1
100.0%
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
Average
17.2
64.3
18.5
100.0
'Income is post-tax and transfer, and is adjusted for family size. Poor and near-poor have income up to
62.5% of the median income for that country. Middle class is between 62.5% and 150%; well-to-do is
above 150% of median income.
432
This means that, relative to these other countries, American
incomes are d i s t r i b u t e d much more unevenly, w i t h a
greater proportion of Americans having relatively low
incomes, proportionately fewer w i t h middle-class incomes,
and w i t h a larger group of well-off families relative to most
other countries.
Another important measure of income inequality is the
extent of poverty. Chapter 6 presented an analysis of the
problem of poverty i n the U.S. In this section, we introduce
an international perspective and examine both the extent
of poverty i n other countries, and how those countries
have responded. We f i n d that American poverty rates are
the highest, and that our system of social protection is the
weakest.
The next three tables examine poverty rates in the late
1970s and the mid-1980s, the years of available data. The
poverty lines used are 40% of median income in each
country, a relative definition of poverty (Chapter 6 discusses this approach to measuring poverty), producing a
poverty threshold close to the official U.S. poverty lines in
the mid-1980s. Since the levels of income and wealth vary
significantly between countries, a relative threshold is
essential to international comparisons. Like the official U.S.
definition, these rates take cash transfers into account and
are adjusted for family size, but unlike the U.S. definition,
these data also account for taxes and the value of food
sumps.
Table 9.9 has the poverty rates from eight industrialized
countries, for all persons, all adults (age 18-64), and children (17 and under). The U.S. poverty rates are by far the
highest for each category. I n the broadest category, 13 3%
o f all Americans were poor in 1986, nearly twice the percentage as in the next highest country (Canada), and 2.3
times the average level o f 5.9%. As noted in Chapter 6,
American child poverty is particularly high. At 20.4%, the
U.S. child poverty rate is almost three times the average
rate o f 7.4%.
The next u b l e , Table 9-10, examines the impact of the
tax and transfer systems of the various countries, for the
same categories as i n the previous table. The "Pre" columns
give the poverty rates generated by the market, prior to the
introduction of taxes and transfers. Thus, these percentages
represent the degree of poverty that w o u l d exist w i t h no
government intervention.
433
The U.S. poverty
rates are by far
the highest for
each category.
�TABLE 9.9
International Poverty Statistics
Poverty Rates in the Mid-1980s
Country
All Adults
(18-64)
All Persons
Australia
Canada
France
West Germany
Netherlands
Sweden
United Kingdom
United States
All Children
(17 or under)
6.7%
7.0
4.5
2.8
3.4
4.3
5.2
13.3
6.16%
7.0
5.2
2.6
3.9
6.6
5.3
10.5
9.0%
9.3
4.6
2.8
3.8
1.6
7.4
20.4
Average
5.9%
5.9%
7.4%
Ratio of U.S.
to Average
2.3
1.8
The extent of market-generated poverty i n these eight
industrialized countries is quite alarming ( F i g u r e 9E). O n
average, over one-fifth of all persons were poor prior to
government intervention. In France, the United Kingdom,
and Sweden, that proportion rises to over one-fourth. To
some degree, these high overall rates are driven by exceptionally high poverty rates of elderly persons (not shown),
but the rates for adults and particularly children are also
disturbingly high. Interestingly, by this measure, the U.S.
rates are relatively l o w ( w i t h the notable exception of child
poverty). The U.S. pre-tax and transfer poverty rate for all
persons, 19 9%, is higher only than Canada and Australia,
and below the average rate of 22.4%.
However, the post-tax and transfer column tells quite a
different story. Here, the U.S. rates are the highest i n a l l
cases. Clearly, our market outcomes, w h i l e leaving one-fifth
of all persons impoverished, generate relatively less poverty than most other countries. Nevertheless, it is equally
clear that our system of taxes and transfers is much less
Our system ot
taxes and
transfers is much
less effective in
reducing poverty
than that of any
other country.
2.8
FIGURE 9E
Impact of Taxes and Transfers on Poverty Rates,
Mid-1980s
TABLE 9.10
The Impact of Taxes and Transfers on
International Poverty Rates, Mid-1980s
All Persons
Adults (18-64)
35
Children (17 or under)
Reduction ol Poverty Due
to Taxes and Transfers
30
Post-tax and transfer
Poverty Rales
27.7
25 9
Country
Pre*
Post'
Point
Change
Point
Pre'
Post'
Change
Point
Pre*
Post'
25
Change
21.5
21.6
19.9
Australia
Canada
France
West Germany
Netherlands
Sweden
United Kingdom
United States
Average
19.1%
6.7%
17.1
7.0
26.4
4.5
21.6
2.8
21.5
3.4
25.9
4.3
27.7
5.2
19.9
13.3
22.4%
5.9%
-12.4
-10.1
-21.9
-18.8
-18.1
-21.6
-22 5
-6.6
-16.5
12.9%
11.5
17.6
9.8
17.4
13.4
18.1
12.8
14.2%
6.1% -6.8
7.0
-4.5
-12.4
5.2
2.6
-7.2
3.9
-13.5
6.6
-6.8
5.3
-12.8
10.5
-2.3
5.9% - 8 . 3
16.4% 9.0% - 7 . 4
15.7
9.3
-6.4
21.1
4.6
-16.5
8.4
2.8
-5.6
14.1
3.8
-10.3
7.9
1.6
-6.3
27.9
7.4
-20.5
22.3
20.4
-1.9
16.7%
7.4% - 9 . 3
'"Pre" refers lo pre-tax, pre-transfer income; "posf refers to post-lax, post-transfer income.
434
435
264
�TABLE 9.11
Tax and Transfer System Effectiveness, Early lo Mid-1980s
United States
1979
1986
Pre**
Post**
Point Change
18.5%
10.8
-7.7
19.9%
13.3
-6.6
Percentage of
Poverty
Reduced
41.6%
11.2%
8.3
-2.9
12.8%
10.5
-2.3
Percentage of
Poverty
Reduced
25.9%
Period 1
Period 2
Point
Change
1.4
2.5
1.1
20.6%
5.7
-14.9
21.6%
6.0
-15.6
1.0
0.3
-0.7
72.3%
72.2%
12.6%
5.6
-7.0
13.6%
6.0
-7.6
55.6%
55.9%
13.7%
6.5
-7.2
15.1%
7.4
-7.7
52.6%
51.0%
33.2%
Pre**
Post"
Point Change
Average of Other Nations*
Point
Change
18.0%
All Persons
Adults (18-64)
1.6
2.2
0.6
1.0
0.4
-0.6
Children (17 or Under)
Pre**
Post**
Point Change
19.0%
14.7
-4.3
22.3%
20.4
-1.9
Percentage of
Poverty
Reduced
22.6%
8.5%
3.3
5.7
2.4
1.4
0.9
-0.5
'Average ol Canada, Australia, Sweden, Germany, Netherlands, and France. Period 1 is either 1979 or
1981, except for Netherlands (1983). Period 2 is either 1986 or 1987, except for France and Germany
(1984).
""Pre" refers to pre-tax, pre-transfer income; "post" refers to post-tax, post-transfer income.
436
effective in reducing poverty than that of any other
country.
The effect of taxes and transfers is shown in the point
change columns of Table 9-10, and in the lighter area in the
bars of Figure 9E. For example, in the United Kingdom, the
market economy of the mid-1980s left 27.7% of all persons
poor. However, after taxes and benefits, the U.K. poverty
rate fell to 5.2%, a change of 22.5 percentage points. By this
measure of poverty reduction, the U.S. is an extreme laggard. For all persons, our redistribution system lowered
poverty by only 6.6 points, well below the average reduction of 16.5 points. For adults, poverty was reduced by a
lesser amount, 2.3 points (note, however, that all countries
show less poverty reduction for adults). For U.S. children,
poverty barely fell at all (1.9 points). This is in stark contrast to the child poverty reduction of the other countries,
where poverty was reduced between 5.6 and 20.5 points.
Table 9-11 introduces changes over time into the analysis, examining changes between two periods (see table for
the description of the time periods). The table compares
the U.S. to the average of the other countries in the study,
omitting the U.K. due to lack of data from the earlier year.
The first point that may be seen in the table is that poverty
worsened over the 1980s in all countries for each category,
both in pre- and post-transfer terms. However, turning to a
comparison of the U.S. and the average of the other countries, we find that poverty reduction is not only less effective in the U.S., as shown in the previous table, it has
become even worse over time. In contrast, poverty reduction in other countries has grown more effective.
This point is made in the point change column of Table
911. For example, in the U.S. in 1986, the market left 1.4%
more persons in poverty than in 1979- After taxes and
transfers, 2.5% more persons were poor in 1986 compared
to 1979. Thus, poverty reduction was 1.1 (2.5 minus 1.4)
percentage points less effective in 1986 as in 1979. Similarly, U.S. poverty reduction was slightly less effective for
adults (0.6 points) and significantly less effective for chil
dren (2.4 points). Turning to the average of the other countries, we observe an opposite trend. For all persons, poverty was reduced by an extra 0.7 percentage points for all
persons in Period 2 as compared to Period 1. For adults and
children, poverty reduction was 0.6 and 0.5 points, respectively, more effective in the latter period. Thus, although
437
Poverty reduction
is not only less
effective in the
U.S., it has
become even
worse over time.
In contrast,
poverty reduction
in other countries
has grown more
effective.
�TABLE 9.12
Poverty Rates and Transitions O t of Poverty
u
for Families wilh Children, Mid-1980s
Country
Poverty Rate*
Transition
Rate**
Percent of
Families Poor
in All 3 Years
of a 3-year
Period
17.0%
4.0
12.0%
27.5
11.9%
1.6
7.8
6.7
18.0
11.0
4.4
2.7
2.7
25.6
26.9
20.0
25.2
26.0***
44.4
36.8
1.5
1.4
4.0
n.a.
0.4
0.4
n.a.
20.3
15.3
49.3
Canada
France-Lorraine
West Germany
All
German
Foreign
Ireland
Luxembourg
Netherlands
Sweden
United States
All
White
Black
13.8
17.0
7.7
14.4
9.5
41.5
'Percent of families with income bebw 50% of that countr/s median income in year one.
"Percent of families who were poor in year one who had more than 60% ol median income in year two.
"Based on 10-30 cases.
438
markets were generating more poverty abroad, the tax and
transfer systems lessened the ultimate outcomes so that
poverty rose only slightly.
A more encompassing measure is shown in Table 9 1 1 as
the percentage of poverty reduced. If we consider pre-tax
and transfer poverty to be the total amount of poverty that
the redistributive system can effect, then the percentage of
poverty reduced is the ratio of the point change to the poverty induced by the market, expressed as a percent. The
table shows a smaller percentage reduction of pre-tax and
transfer poverty in the U.S. than in other countries. In addition, this percentage fell over time in the U.S., while holding constant elsewhere. For all Americans in 1979, 41.6%
of the pre-transfer poor were lifted out of poverty by taxes
and transfers, but in 1986, this percentage fell to 33 2%. In
the other nations, the percentage of poverty reduced was
both much higher than the U.S. and essentially the same in
both periods. For example, in the average foreign nation,
child poverty was reduced 52.6% in Period 1 and 510% in
Period 2. In the U.S., child poverty was reduced a much
smaller 22.6% in 1979, and an even smaller 8.5% in 1986.
Comparing Economic Mobility
Between Countries
In Chapter 6, we note that poverty is a dynamic, rather
than a static, phenomenon. Research on the length of time
spent poor has found a great deal of transition into and out
of poverty. In Table 9-12 and Figure 9F, we present the
results of an international study of the economic mobility
of poor families with children in the mid-1980s, enabling
us to compare U.S. economic mobility to that of other countries. As above, a relative definition of poverty is applied;
however, in this analysis, the poverty lines are set at 50% of
median income (the previous tables used 40% of the
median income as the standard; again, income is post-tax
and transfer and adjustments are made for family size).
The first column of Table 9.12 gives the poverty rates for
families with children in the countries in this study. Once
again, the U.S. rates are the highest; this is particularly the
case for black families, whose poverty rate was 49.3%.
439
In the average
foreign nation,
child poverty was
reduced 52.6% in
Period 1 and
51.0% in Period
2. In the U.S.,
child poverty was
reduced a much
smaller 22.6% in
1979, and an
even smaller
8.5% in 1986.
�FIGURE 9F
Poverty Rates and Transitions out of Poverty of Families with Children,
Mid-1980s
50
Nattwriamls
•
40
-
Sweden
•
ID
(L
30
<
D
08
IT
c
o
—
-,
PianM
w Germany-Gemian
%
•
W. Germany-All
20
Ireland
^ W . Germany-Fofetgn
United Slales-While ^
55
Canada^
10
United Stale«-AII
•
-
Ureled Slales-Black
•
1
10
1.
20
._
1_
30
1
1
40
60
so
Poverty Rate in Year One (Percent)
TABLE 9.13
Absolute Poverty A o g
mn
Children, Adults, and the Elderly, 1979-1982
Poverty Rates in Poor Families
Absolute Poverty
Children
Adults
Australia
Canada
West Germany
Norway
Sweden
Switzerland
United Kingdom
United States
16.9%
9.6
8.2
7.6
5.1
5.1
10.7
17.1
10.5%
7.5
6.5
7.1
6.7
6.2
6.9
10.1
440
Elderly
19.2%
4.8
15.4
18.7
2.1
6.0
37.0
16.1
Overall
13.2%
7.4
8.3
8.6
5.6
5.8
11.8
12.7
However, it is the second and third columns in which we
are most interested, as here (column two) wc find the percent of poor families in year t, who escaped poverty by
year t + 1 ("t" refers to year one or two of the study; t + 1
refers to years two or three), and the percent of families
who were poor for three consecutive years (column three).
The horizontal axis of Figure 9F is the percent of families
who were poor in year t of the study (column one, Table
912). The vertical axis is the percentage of poor families
who escaped poverty by year t + 1 . Therefore, as the data
points in the figure move out to the lower right (higher
poverty in year t, a lower probability of escape by year
t+1), the economic situation worsens.
Most of the countries in Figure 9F show relatively low
poverty rates and relatively high "escape" rates. France, for
example, has a family poverty rate of 4% in year t, and
27.5% of the poor in that year had escaped poverty by the
next year (Table 9 12). However, the U.S. position in the
figure reveals high poverty and low mobility. The U.S. overall poverty rate (20.3%) is the highest in the table, and the
escape rate of 13 8% is the second lowest. The situation for
U.S. blacks is particularly severe. As noted above, the poverty rate for American black families was 49.3% in the
mid-1980s, the highest in the table. Blacks also had the
lowest probability of escaping poverty; only 7.7% of the
poor in year t were non-poor in year t + 1 . Only the Canadian poor are less likely than the U.S. poor to leave poverty. In fact, even foreigners within Germany, who have
higher poverty rates than U.S. whites, have higher escape
rates than the U.S. poor.
Column three of Table 9.12 shows further evidence of
low economic mobility in the U.S. compared to the other
countries in the table. The probability of a poor family with
children remaining poor for the full length of the threeyear period of the study was highest in the U.S., at 14.4%.
While the rate for Canada was slightly lower (11.9%), the
other countries in the study had much lower persistent
poverty rates than the U.S. Black families were most likely
to experience persistent poverty, as 41.5% were poor for
the full period.
As noted, the previous tables use a relative definition of
poverty, based on the median income in each country. By
this measure, the U.S. has been found to have the highest
rates of poverty. However, some might argue that absolute
441
The poverty rate
for American
black families
was 49.3% in the
mid-1980s, the
highest in the
table. Blacks als
had the lowest
probability of
escaping povert
�TABLE 9.14
Jobs Created in Ten Countries, 1979-1989
Country
Total
Created
1979-89
(000)
Australia
Canada
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
1,617
2,091
160
1,690
840
6,460
720
364
1,460
18,518
New Jobs
as Percent
of 1979
Employment
26.5%
20.1
0.8
6.6
4.2
12.0
13.5
8.7
5.9
18.7
TABLE 9.15
Contributions to Employment Growth in Ten Countries
1979-1989
Breakdown of Employment
Growth into Sum of
Country
Employment
Growth
Rate,
1979-1989
Australia
Canada
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
2.4%
1.8
0.1
0.6
0.4
1.1
1.3
0.8
0.6
1.7
442
Population
Growth
Rate,
1979-1989
1.5%
1.0
0.5
0.1
0.2
0.6
0.6
0.2
0.2
1.0
Employment/
Population
Ratio
Growth
Rate,
1979-1989
0.9%
0.8
-0.4
0.5
0.2
0.5
0.7
0.6
0.4
0.7
poverty rates, i.e., a fixed income threshold, adjusted for
family size, should show American poverty to be relatively
low. That is, despite our relative inequalities, our high per
capita income should leave fewer Americans than foreigners w i t h purchasing power below an absolute poverty line.
In fact, the last analysis of this question, w h i c h took place i n
1979-1982, found the opposite to be true.
Table 9.13 shows that the p r o p o r t i o n of Americans i n
absolute poverty, 12.7%, is only surpassed by Australia's
poverty rate, 13 2%. The situation is similar for children
and adults; i n fact, American children have the highest
absolute poverty rates i n the table, 17.1 %. Only elderly persons i n the U.S. have less absolute poverty than their foreign counterparts. These data mean that even w h e n using a
poverty standard set by the U.S., w i t h its relatively high purchasing power, the percentage o f persons w h o cannot
afford the basic necessities is higher i n this country than i n
most industrialized nations. Furthermore, a number of factors, including the rising average o f per capita GDP exclusive of the U.S. (Table 9 7), high absolute poverty rates over
the 1980s (Chapter 6 ) , and o u r high relative rates shown
above, suggest that high U.S. absolute poverty rates relative
to other countries have most certainly risen over the
1980s.
Job Growth and the Public
Commitment to Job Training and
Unemployment Compensation
We have already shown (Chapter 4 ) that most of the net
job g r o w t h over the 1980s occurred i n low-paying service
industries. Here w e examine j o b g r o w t h from an international perspective, and f i n d that other countries have had
faster job g r o w t h and lower unemployment than the United
States.
Table 9.14 shows that f r o m 1979 to 1989 thousands
more jobs were created i n the U.S. than i n other countries
(column one). However, the U.S. should be expected to
have the largest absolute number o f jobs, as it is by far the
largest country. A better indicator is the rate of job creation. Column t w o of Table 9-13 shows that U.S. employment
i n 1989 was 18.7% greater than i n 1979, compared to
increases of 26.5% i n Australia and 2 0 . 1 % i n Canada over
the same period.
443
The U.S. annual
employmenl
growth rate
(1.7%) turns out
to be due more to
population growth
(1.0%) than to
a rising
employment/
population
ratio (0.7%).
�TABLE 9.16
Hours Worked Per Employee in Manufacturing,
Indexed to the United States, 1970-1989
Country
1970
1979
1989
Canada
Denmark
France
West Germany
Italy
Japan
Netherlands
Sweden
United Kingdom
United States
100
96
98
99
100
119
98
91
104
100
98
86
90
91
92
114
88
80
100
100
95
82
83
82
95
111
83
79
95
100
TABLE 9.17
Paid Vacation in European Countries
Country
Weeks
Mandated
by Law
Austria
Belgium
Finland
France
Greece
West Germany
Iceland
Ireland
Luxembourg
Malta
Netherlands
Norway
Portugal
Spain
Sweden
Switzerland
United States**
5
4
5
5
4
3
5*
3
5
4
4
4*
30 Days
30 Days
5
4
16.1 Days
"Iceland has 4 weeks, 4 days; Norway, 4 weeks, 1 day.
'Estimated average, not mandated.
444
Another reason that simple job counts are misleading is
that they do not take into account population growth, a factor that limits both the number of jobs created and the rate
of job growth. Some countries (e.g., West Germany) have
had very little population growth while most others have
grown at a much slower rate than the U.S. Without population growth the only manner in which additional job
growth can be achieved is by raising the percentage of the
population with employment, which is always a slow process. On the other hand, countries with fast growing populations have the possibility of creating substantial growth in
employment.
To address this issue, Table 9-15 breaks down the
growth in employment into growth of population, on the
one hand, and of the employment/population ratio on the
other. The US. annual employment growth rate (1.7%)
turns out to be due more to population growth ( 1.0% ) than
to a rising employment/population ratio (0.7% ). By a more
appropriate standard, the employment/population ratio, the
U.S. has still done well in creating employment, but not
orders of magnitude better than the other countries. And as
shown above, we have accomplished this at the expense of
wage growth.
Workers in the U.S. have also been putting in more hours
than workers in most other countries, in part, to compensate for falling wages. Table 9.16 shows the hours worked
per employee in manufacturing (the only data of this type
available), indexed to the U.S. While two countries had
average hours of work in 1970 that exceeded ours, by 1989
only Japan had more hours of work per average manufacturing employee. Furthermore, American workers had less
generous vacation benefits than European workers (Table
9.17). The majority of European workers receive four or
five weeks of paid vacation, mandated by law. American
workers typically get slightly over two weeks of paid vacation. Furthermore, American workers' lesser vacation time
only partially explains our extra hours of work.
Another way in which European countries arc more generous to their workers is in the provision of publicly provided job training, placement and job creation. Such programs are vital to an economy like the United States, both as
we shift from manufacturing to services, and as the armed
forces reduce their workforce. When compared to the U.S.,
only Japan and Australia devote less of their national
445
Workers in the
U.S. have been
putting in more
hours than
workers in
most other
countries...
American workers
have also had
less generous
vacation benefits
than European
workers.
�TABLE 9.18
Public Spending on Training and Placement, Direct Job
Creation and Subsidies, as Percent o GDP, 1989
f
Training and
Placement
Country
Australia
Belgium
Canada
Denmark
Finland
France
West Germany
Italy*
Japan
Luxembourg
Netherlands
Norway
Spain
Sweden
United Kingdom
United States
0.20%
0.49
0.49
1.32
0.53
0.69
0.83
0.80
0.07
0.33
1.03
0.72
0.30
1.43
0.63
0.24
Direct Job
Creation
and
Subsidies
0.04%
0.63
0.02
0.03
0.45
0.04
0.19
0.00
0.10
0.02
0.05
0.16
0.50
0.13
0.03
0.01
Total
0.24%
1.12
0.51
1.35
0.98
0.73
1.02
0.80
0.17
0.35
1.08
0.88
0.80
1.56
0.66
0.25
resources to promote these programs. As shown in Table
9.18, in 1989 the U.S. government spent only 0.24% of
GDP on job training and placement, and much less (0.01 % )
on job creation and subsidies.
As Chapter 4 points out, unemployment rates provide
only a partial measure of labor force supply and demand, as
they fail to capture the under-employed and discouraged
workers. However, a general sense of the performance of
the U.S. labor market from an international perspective is
given in Table 9.19- In 1979, U.S. unemployment, at 5.8%,
was well above the average of 4.9% for the major industrialized countries. By 1989, unemployment had fallen in the
U.S. to 5.2%; among the seven major countries, only Japan's
rate was lower. Among the smaller countries, the Scandinavian countries had less unemployment than the U.S.
When compared
to Ihe U.S. only
Japan and
Australia devote
less of their
national
resources to
promote job
training and
placement
programs.
TABLE 9.19
Unemployment in O C Countries, 1973-1989
ED
Unemployment Rate
1973
1979
1989
Canada
France
West Germany
Italy
Japan
United Kingdom
United States
5.5
2.7
0.8
6.2
1.3
3.0
4.8
7.4
5.9
3.2
7.6
2.1
5.0
5.8
7.5
9.4
5.6
10.9
2.3
7.1
5.2
Seven Major Countries
3.4
4.9
5.7
Australia
Belgium
Finland
Netherlands
Norway
Spain
Sweden
2.3
2.7
2.3
2.2
1.5
2.5
2.5
6.2
8.2
5.9
5.4
2.0
8.5
2.1
6.1
8.0
3.4
8.3
4.9
16.9
1.4
Total OECD
•1988
Country
3.3
5.1
6.2
Smaller Countries
446
447
�TABLE 9.20
Unemployment and Unemployment Compensation, 1989
Country
Percent of
GDP Spent on
Unemployment
Compensation
Australia
Canada
France
West Germany
Italy*
Japan
Netherlands
Sweden
United Kingdom
United States
0.87%
1.57
1.27
1.20
0.40
0.34
2.42
0.55
0.84
0.47
1989
Unemployment
Rate
6.1%
7.5
9.4
5.6
10.9
2.3
8.3
1.4
7.1
5.2
Percent of GDP
per Percent
of
Unemployment
0.14%
0.21
0.14
0.21
0.04
0.15
0.29
0.39
0.12
0.09
Most other countries are also more generous than the
U.S. in the provision of unemployment compensation.
Table 9.20 shows that the U.S. spends only 0.47% of GDP
on compensating those w h o are out of work; only Italy and
Japan spend less. However, this figure does not take the
level of unemployment into account. That is, a country
might devote very few resources to unemployment compensation because of low unemployment. Therefore, colu m n t w o gives the unemployment rates for the various
countries in the table, and the t h i r d column is the ratio of
compensation to unemployment, a measure of relative generosity given the country's unemployment rate. The numbers in this column tell us the percent of GDP that would
be spent on compensation if unemployment rose by 1 % .
The U.S. spent 0.09% of GDP per 1% of unemployment i n
1989; every country except Italy spent more. West Germany had unemployment rates similar to the U.S., yet their
expenditure o n unemployment compensation was over 2.5
times that of the U.S.
Conclusion
The figure in column one for Italy is from 1988.
448
Compared to other industrialized countries, the U.S. is
falling behind o n many important economic indicators.
O u r comparatively sluggish p r o d u c t i v i t y g r o w t h and
declining competitiveness are l i m i t i n g our wage g r o w t h ,
while most other countries are experiencing rising wages
(for an explanation of additional factors causing wage
dechne, see Chapter 3 ) This development, in tandem w i t h
the diminished effectiveness of our tax and transfer system,
has left the U.S. w i t h the highest and most persistent poverty rates among the industrialized countries examined i n
this chapter. The stagnant wages and productivity i n the
U.S., along w i t h the diminished effectiveness of our system
of taxes and transfers, have begun to seriously undermine
the U.S. standard of living, once considered the highest in
the world.
449
The stagnant
wages and
productivity in th
U.S., along with
the diminished
effectiveness of
our system of
taxes and
transfers, have
begun to seriousl
undermine the
U.S. standard of
living, once
considered the
highest in the
world.
�Table Notes:
Frequently Cited Sources
The following abbreviations are used throughout the Table Notes
HHP.
President of the United Slates, iitonoinic h'c/mrl i if I be
I'rcsnicnl.
1992
Cireeti
Book
U.S. Mouse of Representatives, diccn Hook' Various years
F-60
Series.
U.S. Department of Comnicrcc. Ihne.iu ol the Census. Series l'-(>(>.
Various dales.
SCB.
U.S. Departmeni of Commerce. Simvy o f ( i/rrcnt Business. Moiuhlv
Employmerit ami
Earnings-.
U S. Departmeni of Uilx>r. Employmenl tnul Etirniii^s Momhly and
historical supplemcnis.
NIEA:
U.S. Department of Commerce \alional
Accounts. Revisions as ol Spring I W J
Income and I'nulm l
Table Notes:
Introduction
A. Incidence of labor Market Distress Based on tabulations of the March I WO
Current Population Survey (CPS) provided by lidie Rasell The sample is all i i\il
ians ages 25-64, excluding the unineorporaied self-employed Poverty level
wage earners are those who averaged S(>.(>9 per hour or less over the year
(based on annual wages and usual weekly hours and weeks worked). This is ihe
four-person poverty level ($12,075) in I9K9 divided by 20H0 hours We
excluded zero or negative earners from our count Involuntary pan-iimers are
those who could not find lull-time work and worked part-time involununly or
those who because of slack work were working pan-time involuntarily. Discouraged workers are those who did not work or look lor work at some lime during
the year because they believed no work was available Most discouraged work
ers fell into an additional category of distress. We did not count 16-2-1 year olds,
or those involuntary pan-lime or unemployed lor less than lour weeks so om
measure focused on unmistakably serious distress among the prune-age
worklorce.
nl
�income thresholds depend on family size, and we use the lour-person threshold for illustrative purposes.
Chapter 1
1.1
Median Family-Income. P - M Series. No 97, p .30; P-60 Series, No. 166, p
I I ; P - M Series, No 174, p. 201; P - M Series, No 180, liihlc I V
1.2
Annual Growth o f Median Family Income. Yearly dollar change is annual
average of total dollar change in period Size-adjusted median lamily income
is weighted by families and adjustetl for family size using the poverty' ecjuivalence scale. This is actually the average income in the middle fifth, not the
median. Based on data in Table I I and from the Green Hook (1992), p. 14 16
1.3
14
1.5
Growth of Median Family Income by Age of Household Head 1991 innmP-Otl
Series, No. 180, Table 13; 1989 from P - M Series, No. 174, p. 52; 1979 from
P-OO Series, No. 129, p. 32; 1973 from P-60 Series, No. 97, pp. 49 51; 1967
from P-6<) Series, No. 59, pp 32- 34 (through telephone conversation with Ed
Welniak on June 15, 1990).
Growth o f Median Family Income Hy Race/Ethnic Group. 1991 from P-60
Series, No. 180, Table 13; 1989 from P 60 Series, No. 174, pp. 53-55; 1979
from Fisher (1986): Table IB; 1947-1973 from P-MSeries, No 129, pp. ^ - 4 3
Income Growth By type of Family. For 1991 from P-OO Series, No. 180, lable
13 Other data from Departmeni of Commerce unpublished historical tables
(1991c): Table 13.
1.6
Real Family Income Growth by Fifth P-6<) Series, No. 174, p. 216 and, for
1947, Department of Commerce unpublished historical tables (1991c): Table'
10, and P-60 Series, No. 180, Table B-7.
Change in Family Income by Income I erel I sing HLV Return Data I rom I S
Treasury (1992): liible A2. These are consianl law adjusted gross incomes
1.12
Shares of Consumption. From Cutler and Katz (l ) )2) 'liible I
1.13
Distribution of Aggregate Income Gains See note to liible I 10 Shares lor
groups add up to slightly more than 100% because families with negaiive
incomcs are excluded Irom the lowest lifth but included in "All" .md. iherefore, aftect the aggregate growth
1.14
Source of Family Income f o r F.acb Fifth of Families In >in impuMishc il ( < >ngressional Budget Office (CBO) tahulalions. Sh.ires .md Sources of I .imily
Income: All Iamilies, 1977. I9K0. 1985. I'>HK l'W>. proxided hx l-'rank
Sammanino. Markcl-hased income is total income less iLiiislcr and other
(e.g., alimony, pension) income.
1.15
Real Income Growth By Type of l \ rsoiial Income XII'A. 'liible 2 1 The earliest data available are lor 1959
1.16
Shares of Markel-Based Income By Type From SlI'A. 'liible 1 I The e.uliesi
data available are for 1959
1.17
Shares of Income by lype. By Sector Based on A7/M. 'liible I 15 'I'he < orporate and Business" sector includes "corporate,'' "oilier private business and
"rest of world." The "govemment/nonprolii" sector includes die household,
government enterprise, and government sectors, all ol w hich generate no
capital income
1.18
Distribution of labor and Capital Incomes See note to 'liible I I I
1.19
Sources of Income Growth of Top Fifth. Data derived Irom shares of income
by type of income from unpublished Congressional Mudget Office (CIU>)
data (see note lo liible 1.14) and from income levels from 'liible 1.9.
1 20
Shares of Income Growth by iy/>e of Income f o r lop Fifth Based on d.u.i in
Table 1.19.
1.21
Distribution of Persons. Households and Iamilies by Income l e i v l Based i >n
a Census Bureau analysis presented in Green Book (1992) lahle 90. p. I i55
I 22
Distribution of Prime . \ge Adults by Relative Income l evel Based on a ( i nsus Bureau analysis presenied in Green Book (1992): liible 93. p. I |5S
I 23
Changes in Incomes of Married Couple Families with Children hy Source
Joint Economic Committee (1992): liible I
1.24 Husbands' and Wives' Hours of Work Joint Economic Commitiee
(1992): liihles 3 and 4 The hours lor employed wives and husbands are cm
reeled numlxrs from tabulations provided hy ihc .ll C
Shares of Family Income Going lo Various Fifths, and tit Top 5%. P-6() Series,
No. 174, p. 216, and P - M Series, No 180, lable B-7.
1.7
1.11
1.8
Income Growth By Fifth a n d Family Type. Green Book (1992): lable 45, pp.
1371-72.
1.9
Income Growth Among Top Fifth and by Fifth. From House of Representatives, Ways and Means Committee (1991)
Fersons are placed into fifths and percentiles based on adjusted family
income; the numbers in the table, however, are average unadjusted family
incomes for the families in each group. The numbers of persons in each fifth
are the same, though the numbers of families differ. Averages are weighted by
families, not by persons. "All" includes families with zero or negative incomes,
but the bottom fifth excludes these families.
1.10
Changes in Family Income Shares. House of Represeniatives, Ways and
Means Committee (1991). See note to Table 1.9. Shares for groups add up to
slightly more than 100% because families with negative incomes are excluded
from the lowest fifth but included in "All." Maximum income thresholds are
for a family of four as presented in Green Book (1992), p. 1515. I he actual
^2
i
t
153
�1.26
Role of Higher Wives' Earnings and Hours on Family Income Growth. Based
on the hours, wages and income data presented in Tables 1.23,1-24 and 1.25
and Joint Economic Committee (1992): Tables 1 and 7. The hypothetical calculations use the hours, wages and incomes in 1979 as a base and ask what
if the hours or earnings of wives remained at the 1979 levels but other
dimensions of income changed as they did over the 1979-1989 period. For
the hypothetical calculations of income changes without higher wives' hours,
the 1989 hourly wage levels of wives are used.
2.4.
Shares of After-tax Income for All Families. U.S. House of Representatives,
(1991), p. 71.
2.5.
Effective Federal Tax Rates. 1977 and 1989 actual effective rates were applied
to 1989 average before-tax incomes—from Green Book (1991), p. 67. The
resulting average after-tax incomes were multiplied by the number of famUies
in each group (provided by Frank Sammanino, Congressional Budget Office
(CBO)) to get total after-tax incomes per group. Then these were converted
to shares by dividing each group's total by the sum over the groups. Shares in
the second and fourth column differ somewhat from corresponding shares in
Table 2.4 because of rounding error in Green Book (1991) tables— particularly for the number of families in the top 1%— and because total income for
all families in Table 2.5 necessarily excludes families with zero or negative
incomes, while in Table 2.4 such families are included.
2.6
Effective Average Tax Rates. Green Book (1991), p. 73.
2.7
Effect of Federal Tax Changes on Shares ofAfter-Tax Income. Average pre-tax
family income is from Green Book (1991), p. 67. See note to Table 2.6 for
effective rates.
1.30 Effect of Second Earner on Household Expenditures. Democratic Study
Group (1990): Table 1, pp. 8-9-
2.8
Effective Tax Rates for Selected Federal Taxes. Data provided by Frank Sammanino, Congressional Budget Office (CBO).
1.31 Child Care Expenditures and Family Income for Employed Women With
Children. Democratic Study Group (1990): Table 3.
2.9
Changes in Effective Federal Taxes. See note to Table 2.8.
2.10
Taxed and Untaxed Corporate Profits. Figures are from NIPA: Tables 1.1 and
1.16. Taxes include federal, state and local combined. Actual profits are taxed
profits plus net interest (interest paid minus interest received) and the difference between the allowance for inventory investment and capital depreciation allowed in the tax code, on the one hand, and actual inventory investment and capital depreciation, on the other. The idea for this analysis is
based on Thomas Karier (1990). Only nonfinancial corporations are included
because banks do not pay net interest—they receive net interest.
1.27
Effect of Wives' Earnings on Income Shares Among Married Couples with
Children. Derived from data presented in Joint Economic Committee (1992):
Table 8.
1.28
Changes in Hours Worked in All Families and in Married Couple Families.
Based on Congressional Research Service computations of March Current
Population Survey (CPS) data, presented in Green Book (1992): Table 70, pp.
1424-1425.
1.29 Effect of Wives' Earnings on Income Inequality Among All Families and Married Couples. From Cancian, Danziger, and Gottschalk (1992): Table 7.
1.32
Hours Worked by Employed Parents by Gender. Leete-Guy and Schor (1992):
Table 9.
1.33 Median Family Income by Age of Householder. From unpublished Census
Bureau historical tables, Table 15.
1 34 Distribution of Individuals in Final Year by Family Income Fifth in Starting
Year. SawhiU and Condon (1992): Table 1.
1.35
Transitions Into or Out of Middle Income for Families with Children. Based
on research by Duncan et al. (1991) as presented in Green Book (1992):
Table 99, p. 1467.
2.11 Corporate Profits Tax Rates. NIPA. Tables 1.1 and 1.16.
2.12
Chapter 2
Total State and Local Taxes in 1991 (Effective Rates) as Shares of Income for
Families of Four. Mclntyre, et al. (1991), p. 18. Any tax analysis of this nature
must make a number of assumptions regarding tax incidence. For example,
in this study half of the property tax paid on rental units is allocated to the
owners of the property, and half is assigned to the renters in the form of
higher rents. We consider the assumptions in this study to be reasonable and
unbiased. The methodology of the study is discussed on pp. 70-72.
2.1
Total U.S. TaxBurden. VromNIPA Tables 1.1, 3.2, 33. 1991 data from Survey
of Current Business (SCB) (February 1992).
2.2
Tax Revenues in OECD Countries. From Organization of Economic Co-operation and Development (OECD) (1990a), p. 71.
2.13 Federal vs. State and Local Tax Burdens. NIPA: Tables 3.2, 3.3
Average After-Tax Family Income. Green Book (1991), pp. 67-68.
2.14
2.3
454
The Composition of Taxes as Percent of GDP. NIPA: Tables 1.1,32,33
455
�Chapter 3
3.1
3.2
3-9
Trends in Average Wages and Average Hours. Unpublished tabulations from
Kevin Murphy from an update of Murphy and Welch (1989), based on March
Current Population Survey (CPS) files, 1963-1990. Hours of work derived
from differences between annual, weekly and hourly wage trends. Unfortunately, the data include self-employed as well as wage and salary workers. We
suspect that the wage growth from 1982 to 1989 is overstated in their series
(particularly the hourly wage). Productivity data for 1959 to 1989 are from
ERP (1992): Table B-44, p. 348 for the non-farm business sector. 1990 productivity data from Joint Economic Committee Economic Indicators (June
1992a), p. 16.
Change in Hourly Compensation, Wages and Benefits. Based on employment cost levels from the Bureau of Labor Statistics (BLS) Employment Cost
Index series for March 1987 to March 1992 for private sector workers. We categorize pay differendy than the BLS, putting all wage related items (including
paid leave) into the hourly wage. Benefits, in our definition, only include payroll taxes, pensions, insurance and "other" benefits. See Mishel and Bernstein
(1992): Appendix A for further discussion.
It is imponant to use the current-weighted series rather than the fixedweighted series because composition shifts have a large effect, as we show in
Tkble 3 26. Data for 1966, 1972 and 1977 are based on BLS surveys of
employer expenditures for employee compensation (EEEC) in the private
non-farm economy, provided by BLS economist Albert Schwenk.
3.3
Hourly Wages a n d Benefits By Occupational Groups. See note to Table 3.2.
Our analysis in which we tried to match occupational data from the earlier
employer expenditures for employee compensation (EEEC) data to the
Employment Cost Index (ECI) data yielded nonsensical results, perhaps
based on changes in definitions.
310
Distribution o f White Employment by Wage Level. See note to Table 3 9.
These are non-Hispanic whites.
311
Distribution o f Black Employment by Wage Level. See note to Table 3.9.
These are non-Hispanic blacks.
312
Distribution of Hispanic Employment by Wage Level. See note to Table 3.9.
Hispanics may be of any race.
3.13
Employer Hourly Benefit Costs, By Type. Based on employer expenditures for
employee compensation (EEEC) and Employment Cost Index (ECI) as
described in note to Table 3-2. Inflation in medical care services from ERP
(1992): Table B-58, p. 364, and from Consumer Price Index (CPI) news
release with seasonally adjusted data for the first three months of 1992.
314
Changes in Private Sector Benefit Coverage. Based on tabulations of the
March 1980 and March 1990 Current Population Survey (CPS) data files, provided by Edie Rasell. The sample includes wage and salary workers ages
18-64 in the private sector who worked at least 20 hours weekly and for at
least 26 weeks. Health insurance coverage is where the employer offers coverage and pays for at least some of the costs. Pension coverage is where worker's employer has a plan in which the employee participates. Unfortunately,
the 1979 data are based on 1970 Census weights.
3.15
Changes in Private Sector Benefit Coverage by Wage Fifth. See note to Table
3.14.
3.16
Ti-end in Days Off With Pay. Based on unpublished Bureau of Labor Sutistics
data, provided by Bill Gullickson, on the ratio of hours at work to hours paid,
for production and nonsupervisory workers in nonagricultural business (limited to 1981-1989) and in manufacturing (1947-1989). The days off of paid
leave is computed from one less the ratio of hours at work to hours paid
times 260 days (assumes a five-day week for 52 weeks). Data for non-form
business for 1966, 1972 and 1977 derived from employer expenditures for
employee compensation (EEEC) data (see note to Table 3-2) from the ratio of
hourly compensation per hour paid to hourly compensation per hour
worked.
317
TheJoint Effect of Education and Experience on Wages. Based on tabulations
of Current Population Survey (CPS) wage data described in the Appendix.
The basic data are the hourly wages and employment shares of workers by
Hourly and Weekly Earnings o f Production and Nonsupervisory Workers.
Employment and Earnings (March 1992): Table C-l, p. 133; and for 1947,
Supplement to Employment and Earnings (March 1985), p. 5.
3.4
Distribution o f Total Employment by Wage Level. Based on analysis of Current Population Survey (CPS) wage data described in the Appendix. The poverty level wage was defined as the four-person poverty threshold in 1979
divided by 2080 hours and deflated by CPI-U-X 1 to obtain levels for other
years. We calculated more intervals than we show but aggregated for simplicity of presentation (no trends were lost).
3.5
Wages f o r All Workers by Wage Percentile. Based on analysis of Current Population Survey (CPS) wage data described in the Appendix. These data differ
slighdy from those in Mishel and Bernstein (1992) because of some corrections to the 1989 data.
3.6
Wages f o r Male Workers by Wage Percentile. See note to Table 3.5.
3.7
Wages f o r Female Workers by Wage Percentile. See note to Table 3 5 .
3.8
Changes in the Gender Wage Differential. Based on data from Tables 3-6 and
3.7.
456
457
�experience (five-year categories starting with 1-5 years, antl going to 31-35
years, and a 35 plus years category) anil education (less lhan high school,
high school graduate, some college, college graduate, more than college) for
all workers, men and women The dccom|x>sition is based on three terms:
(1) the initial wage times the difference in employmenl shares; (2) the iniiial
employment share times the difference in wages; and (3) the change in
shares limes the change in wages. I he lirsi column represents the first term,
the second column the second term and the third term is not presented, but
can Ix- derived from the difference Ixtween the last column and the first two
columns.
3.18
Ibe Separate Effects of Education and Experience on Wages. Same data and
analysis as in Table 3 17, but we separate the education anil experience
effects.
also incorporates differences in how households anil employers classily the
workforce by occupation.
Household data yields a larger share of executives, managers anil sales
workers and a lesser share of administrative anil service workers. We corn el
for this by calculating the effect of the dilierence Ixtween household (CPS)
data and employer (Occupational Fmployment Statistics) survey data on pay
levels in 1990 using a shih-share analysis. Pay levels are w ages or compensation from the l-mployment Cost Index (FCI) for 19^0 The Curirni I'opulalion Survey (CPS) household employmenl data are from Employment ami
Earnings (January 1991): lable 21, p. IHH The OTS employer data are from
Silvestri and l.ukasiewie (1991): lable I , p 05
3.27
Pay in Expanding and Shrinking Industries Costrell ( IWH): Tables A l and
A5.
3 28
Effect of Industry Employment Shifts on Hiy. Baseil on Costrell (IMKH) as
shown in Mishel (1989): Table 3
3 29
Industry Employmenl Shifts and Production Worker Weekly Wages. Based on
Costrell (1988) as shown in Mishel (1989): liible i
3.30
Trade Induced Changes in labor Supply I rom Borjas et al (199 I): Table I
3.31
Immigration and Labor Supply. From Borjas et al. (1991): 'liihles 2 and V
3 .32
Effect of Trade on Labor Supply of High School and College Workers From
Borjas et al. (1991): Table 5
3.33
Effect of Trade and Immigration on Education Wage Differential From Borjas et al. (1991): lable 6
3-34
Employment Crowtb By Sector. Employmenl and Earnings, (March 1990):
Table B-l with 1989 data as revised in Employment and Earnings (June
1991).
Comparison of Union and Nonunion Hourly Wages and Benefits, hmployment Cost Index pay level data in U.S. Departmeni of hilxir (1989a): lahle
10.
3.35
Iffect of Deunionization on Male Occupation and Education Difjereiiluils.
From Freeman (1991): liible 2.
3.25
Changes in Employment Share By Sector. Based on data in Table 3 24.
3.36
3 26
Effect o f Structural Employment Shifts on Pay Levels. This comparison is
based on two computations of pay levels in March 1989 One is the current
weighted pay levels from the Hmployment Cost Index See note to Table 3 2
The other is from the underlying, unpublished data used to compute quarterly changes in hourly wages and compensation, which is based on 1980
Census weights. The major difference between the pay levels in these computations is the weighting scheme. Because the data with current weights are
baseil on an employer survey, and the data with 1980 weights are based on a
household survey, the difference in the weighted and unweighted pay levels
Impact of Deunionization on Changes in Average Hourly Earnings From
Blackburn el al. (1991): Appendix Table ^
3.37
Effect of Unions on Wages, By Wage f ifth From Card (1991): 'liible H 'The
effect of deunionization is the change in union coverage limes the union
wage premium.
3 38
Effect of Unions on Male Wage Inequality From Card (1991): 'liible 9 and
from Freeman (1991): Table 6.
3 39
Value oj Minimum Wage. Historical values of minimum wage Irom Shapiro
(1987), p. 19. Wages for 1990 and 1991 are based on Icgislaicd increases lo
3.19
Change in Real Hourly Wage by Education. Based on tabulations of Current
Population Survey (CPS) wage data described in the Appendix. The group
with 17 years of schixding (college plus one) is omitted for simplicity of
presentation.
3 .20
Change in Real Hourly Wage of Men by Education. See note to Table 3.19.
3.21
Change in Real Hourly Wage f o r Women by Education. See note to Table
319.
3 22
Change in Real Hourly Wage by Work Experience. Based on tabulations of
Current Population Survey (CPS) wage data described in the Appendix.
3 23
Entry Level Wages and Employment Shares. Based on tabulations of Current
Population Survey (CPS) wage data described in the Appendix. Following
Katz and Murphy (1990) we examine the wages of workers with 1-5 years of
work experience (traditionally defined).
3.24
|5K
159
�$3 80 on April 1, 1990 and to $4.25 on April 1, 1991. Inflation projection for
1992 from a Merrill Lynch (May 4, 1992) forecast.
3 40
3 41
Amount by Which Earnings of a Eull-Ttme, Iwll-Year Minimum Wage Worker
are Above (Below) the Poverty Line. See note to Table 3 39 for source of minimum wage levels and 1992 inflation projection Annual earnings are based
on 2,080 paid hours of work, and assuming no additional income. Poverty
lines are from P-60 Series, No. 166, p. 88. This source gives poverty lines
adjusted by the CPI-U, which is the official method. We convert these poverty
lines back to actual (nominal) poverty lines for each year using the CPI-U,
take the difference between these lines and the actual salary of a minimum
wage worker, and adjust this "gap" for inflation using the CPI-U-X 1
Minimum Wage Workforce Demographic Composition. Based on tabulations
of Current Population Survey (CPS) wage data described in the Appendix.
Because of the change in the legislated minimum wage in April 1991 the definition of a minimum wage worker differs between the first three and the last
nine months.
3 .42
Distribution o f Wage Earners by Minimum Wage Status and Family Income.
Based on tabulations of March 1990 Current Population Survey (CPS) data
file prepared by Bill Spriggs.
3.43
Distribution o f Workers Before and After Minimum Wage Increases. Based
on tabulations of 1990 and 1991 CPS Outgoing Rout ion Group files prepared
by Bill Spriggs.
3.44
Effects of Computer Usage on Wage Structure. Derived from Krueger (1991):
Tables 1 and 8. The computer-use wage premium is based on the coefficients
of the computer-use variable (a) and its interaction (b) with schooling (S) as
exp (a+bS)-l.
3.45
Executive Pay Levels. The Economist (December 23, 1989), using data on
gross and net income in dollars.
3.46
Real Growth in Executive Pay. Based on data in Table 3.40, converted from
dollars to national currency using exchange rates from ERP (1990): Table
C-109, p. 418 and then convening to constant currency values using an index
for domestic consumer prices from International Monetary Fund, International Financial Statistics, except the Personal Consumption Expenditures
(PCE) index was used for the U.S.
3.47
Comparative Pay Levels f o r Workers and CEOs. Towers, Perrin and Company
(October 1988): Exhibits 8 and 9, pp. 18-19.
3.48
The Gap Between Executives and Production Workers. I lourly compensation
of production workers from U.S. Department of l-abor (1989c). CEO pay
from lowers, Perrin and Company (October 1988).
460
3.49
Black-White Wage Differentials Among Young Workers From Bound and
Freeman (1991): Table 3 The data were transformed from log points to |x-rceniage changes.
3.50
Employment Opportunity and Wage Growth From Blackbum et al (1991):
lables 1 and 3. The wage growth in the last column is for 1979-19H8.
3.51
The Effect of Occupation and hulustry Employment Shifts on Skill and Education Requirements. From leixeira and Mishel (1992): lahle 1
3 52
The Effect of Occupation and Employment Shifts on fay From leixeira and
Mishel (1992): Table 2.
Chapter 4
4.1
Unemployment Rates. Employment and Earnings (January 1990) 'liible 3,
pp. 162-64; Table 39, p. 206; and, Employment and Earnings (January
1992): Table 40, p. 209; Table 3, pp 164-66.
4.2
Rates of Underemployment. For 1973 and 1979, US Departmeni ol ljlx>r
(1985): Table 1, p. 6; Table 4, p. 14; Table 21, p 58 For 1989, Employment
and Earnings (January 1990): lable 1, p 160; Table 31, p 199; lahle 35. p
202.
4.3
Changes in Unemployment in Postwar Recessions Unemployment rates lor
all but the 1990s recession from Meisenheimer et al (1992): 'liible 4 1990
data from Employment and Earnings (January 1992): lable A-43. p 54 1992
data from July 2, 1992 Bureau of l-abor Statistics (BLS) release on unemployment data. Unemployment rate due to "permanent job loss" is based on
"other job loser" category with "other reasons" including layoffs, job leaves,
re-entrants and new entrants."
The permanent job loser unemployment rate was derived from the aggregate unemployment rate and from data on the percentage of the unemployed that were "other job losers" from U S Department of I j l x i r (I9H2):
Table E-23, p. 417 for 1967-1981; from Bureau of I j b o r Statistics (BLS) economist Gloria Green for 1982; from Employment and Earnings (January
1991): Table A-52, p 60 for 1990; and, from the July 2, 1992 Bureau of Lihor
Statistics (Bli>) release on unemployment for 1992 Ungth of recessions arcbased on National Bureau of Economic Research (NBER) dating, as presented in Henwood (1991), p 9.
We assume the trough of the early 1990s' recession was in May 1992 /Vs of
this writing there has been no "officially declared" end ol the recession
Although output increased in most of 1991 and in 1992, there was a continued growth of unemployment, especially in the first six months ol 1992
l.ast, in contrast to the Bureau of Ubor Statistics (BLS) analysis, we identified 1990:2 and not 1990:3 as the peak quarter because of the 0.3% rise in
i(»l
�unemployment in the 1990:2 period. The other peak quarters identified by
Bureau of Labor Statistics (BLS) were output peaks, not unemployment
peaks. However, choosing the unemployment peak, as we did for 1990,
would only affect the 1980:1 peak and not other peaks.
4.12
Changes in Labor Force Levels and Participation in Recessions. Labor force
size and participation data for 1969 to 1981 from U.S. Department of Labor
(1982): Table E-l, p. 300; Table E-2, p. 316; for 1982 from Bureau of Labor
Statistics (BLS) economist Gloria Green; and, for 1990 from Employment
and Earnings (January 1991): Table A-43, p 53. Rates of change are annual
rates (but not log annual rates).
4 .14 Wage Differences Between Part-Time and Full-Time Workers by Gender and
Selected Occupation. From analysis of Current Population Survey (CPS) wage
data as described in the Appendix.
4.15
Difference in Fringe Benefits, for Full-Time and Part-Time Workers, By Sex or
Marital Status and Occupation. Blank (1990): Table 4, p. 30.
4.5
Change in Unemployment Rate by Occupation, Gender, Race, and Industry.
For 1990 bom Employment and Earnings (January 1991): Table A-50. For
1992 from Bureau of Labor Sutistics (BLS) economist Leo Rydzewski.
4.16
4.6
Changes in Employment and Unemployment in Recessions by Occupation.
Data through 1982 from unpublished tabulation, provided by Bureau of
Labor Statistics (BLS) economist Joseph Meisenheimer, of data used in
Meisenheimer et al. (1992): Table 5. For 1990 from Employment and Earnings (April 1992): Table A-46 and for 1992 from (July 2, 1992) Bureau of
Labor Statistics (BLS) release on unemployment.
Growth of Multiple Jobholding, All Workers. Allfiguresare from May of the
given year. 1973 and 1989 are from U.S. Department of Ubor (1989d). 1985
is from Stinson (1986). 1979 is from Sekscenski (1980). 1991 is from U.S.
Department of Ubor (1991b).
4.17
Growth of Multiple Jobholding By Gender. See note to Table 4.16.
4.18 Distribution of Multiple Jobholders Experiencing Economic Hardship. See
note to Table 4.16.
4 .7
Proportion of the Unemployed Who Receive Some Unemployment Insurance
Payment. Center on Budget and Policy Priorities (1992): Table 1.
4 .20
4.8
Employment Growth in Expansions. Length of expansion from National
Bureau of Economic Research (NBER) dating reported in Henwood (1991),
p. 9- Employment data are from the establishment survey as presented in U.S.
Department of Labor (1991a), pp. 2, 6, and 689.
The Use of Various "types of Contingent Labor. The two surveys are the Conference Board survey (1989) and the Bureau of National Affairs survey
(1986). R)r detaUs see Carre (1992).
4.21
Employment Growth. For lull-time equivalent, NIPA: Table 6.5C. For hours,
NIPA: Table 6.9C (earliest year is 1948 so growth rate for earliest period
adjusted accordingly). For Labor Force Participation, Working Age Population
and Civilian Employment, Employment and Earnings (January 1992): Table
1, p. 162. Note the slowdown in labor force participation is even greater
when the measure is decline in non-participation.
The Growth of Personnel Services Industry Employment. Standard Industrial
Classification (SIC) 736 from Employment and Earnings (Supplements for
1982 and earlier years; for 1989, March 1990: Table B-2; and for 1991, March
1992: Table B-2). 1989 and 1991 data for women were obtained from
Bureau of Ubor Sutistics (BLS) by telephone.
4.22
Growth in Temporary Help Industry Employment. Employment in Standard
Industrial Classification (SIC) 7362 from same sources as Table 4.21.
4.23
Elements of the Marginal Workforce. Carre (1992).
Composition of Non-Agricultural Employment. For 1973 and 1979 from U.S.
Department of Ubor (1989b): Table 23, p. 121. For 1989 from Employment
and Earnings 0anuary 1990): Table 32, p. 199- For 1991 bom Employment
and Earnings (January 1992): Table 32, p. 201.
4.24
The Growth of Self-Employment. U.S. Department of Labor (1989b): Table
21, pp. 112-13; Employment and Earnings, (January 1990): Table 23, p. 189;
and Employment and Earnings, (January 1992): Table 23, p. 191; Table 19,
p. 18.
4.11 Age and Gender Composition of the Labor Force and Rate of Part-time
Employment. From Tilly (1991): Table 1.
4.25
Self-Employment and Paid Employment Earnings. Haber et al. (1987): Table
4, p. 20.
4.4
4.9
4.10
462
Industry Composition of the Labor Force and Rate of Part-time Employment.
From Tilly (1991): Table 3.
4.13 Rate of Voluntary and Involuntary Part-time Work By Industry. From Tilly
(1991): Table 4.
4.19 Hours Worked By Multiple Jobholders, By Gender. Unpublished data provided by Bureau of Ubor Statistics (BLS) economist John Stinson.
463
�Chapter 5
5.1
5.12
Growth o f Household Wealth. Assets and debts are year-end outstanding
values from balance sheet data for households, personal trusts and nonprofit
organizations from Board of Governors (1992), pp. 14-17 for 1967-1991;
and (1990), pp. 19-24 for 1949 data. Nonprofit organizations, a small component judging from the breakout on tangible assets, were included because
the Federal Reserve System does not give the breakout for financial assets.
Data converted to real dollars using the Personal Consumption Expenditures
fixed-weight price index from NIPA. The growth rates in the table are assets
net of debt per adult. The number of adults from ERP (1992): Table B-29,
and for 1991 from telephone conversation with the Census Bureau's Population Estimates Department. Adult population includes the Army.
5.2
Distribution o f Wealth. From 1989 Survey of Consumer Finance (SCF) tabulations in Wolff (1992b). Table 2. Net financial assets breakdown from Wolff
over the telephone.
5.3
Selected Holdings of Assets by Family Wealth Level. 1989 Survey of Consumer Finance (SCF) data from Wolff (1992b): Table 6.
5.4
Share of Total Household Wealth Held by Richest One Percent of Individuals.
From Wolff (1992a): Table 1.
5.5
Composition of Aggregate Household Wealth. 1989 Survey of Consumer
Finance (SCF) data from Vffolff (1992b): Table 5.
5.6
Changes in Distribution of Net Worth and Family Income. 1983 and 1989
Survey of Consumer Finance (SCF) data from Wolff (1992b). Table 2.
5.7
Sources o f Wealth Growth. Reported in Wolff (1992b), p. 21.
5.8
Change in Net Worth by Wealth Class. Computed from data in Wolff (1992b):
Table 2.
5.9
Change i n Wealth by Income Class. From Census Bureau study of Survey of
Income and Program Participation (SIPP) data presented in Eargle (1990):
Table H, p. 8.
5.10
5.11
Household Debt Burden. Asset and debt data from Federal Reserve Board balance sheet data (see note to Table 5.1). Personal income data for 1959 to
1989 from ERP (1992): Table B-23; for 1991 from Joint Economic Committee, Economic Indicators (May 1992), p. 5.
Because current NIPA data start in 1959, personal income data for 1949 are
.taken from earlier NIPA data, Table 2.1.
Ratio of Household Income and Wealth Between Non-White and White Families. From Wolff (1992b). Table 8.
464
Net Worth by Race and Household Characteristics. From Wolff (1992b):
Table 9.
Chapter 6
6.1
Percent and Number of Persons in Poverty with Averages Over Peak Years.
Actual poverty rates are from P-60 Series, No. 176, p. 16. Predicted rates are
from unpublished data provided by Rebecca Blank. The Blank model is
described in Blank (1991).
6.2
Consumption-Based Poverty Measures. Ruggles (1992), p. 7.
6.3
Poverty Rates when (Nonmedical) Non-cash Benefits are Included. Green
Book (1992), p. 1305.
64
Percent of Persons with Low Relative Income Adjusted f o r Family Size. P-60
Series, No. 177, p. 8. Equivalence factors (to adjust for family size) are from
Ruggles (1990). Unrelated individuals treated as one-person families.
6.5
Poverty Gap. Aggregates and Means. Center for Budget and Policy Priorities
(1992), p. 33.
6.6
Persons Below 50% o f the Poverty Level. Center for Budget and Policy Priorities (1992), p. 37.
6.7
Distribution of Poverty Spells for Nonelderly Persons
Entering Poverty. Kane and EUwood (1986), p. 12.
6.8
Events Leading to Poverty Spells, and Average Spell Length. Bane and
Ellwood (1986), p. 18.
6.9
Length of Poverty Spells for Children Age 1-10, by Race, in Percent. Green
Book (1992), p. 1177.
6.10
Poverty by Race/Ethnicity. Center for Budget and Policy Priorities (1992), p.
3.
6.11
Percent of Children in Poverty, by Race. P-60 Series, No. 175, p. 24.
6.12
Poverty Among the Elderly and All Persons Before and After Transfers. Danziger and Weinberg (1992): Tables 2 and 7.
6.13
Changing Family Structure and Poverty. Current Population Survey (CPS)
P-60 Series, No. 174, p. 58. Number of persons in female-headed famUies in
1959 is an estimate based on the number poor and in such famUies, together
with the poverty rate among such famUies.
6.14
Changing Family Structure and Poverty. See note to Table 6.13. This type of
demographic decomposition assumes the changes in population shares
occur independently of changes in the poverty rates of each specific group.
465
�The interaction term in column three reflects the extent to which changing
demographics are causally linked to changes in poverty rates.
7.3
Percent of Families with Low, Middle, and High Relative Income, by Region.
P-60 Series, No. 177, p. 18-19.
6.15
Poverty Rates f o r Female-Headed Families. Center for Budget and Policy Priorities (1992), p. 5.
7.4
Ratio of Average Income of the Top Fifth to Average Income of Bottom Fifth.
Barancik and Shapiro (1992), p. 22.
6.16
Increase in Poverty i n Female-Headed Htmilies, by Race. Center for Budget
and Policy Priorities (1992), p. 6.
7.5
Income Changes f o r Average Family, Bottom, Middle, a n d Top Fifth. Barancik and Shapiro (1992), pp. 50-51.
6.17
Marital Status of Female Family Heads. Women without dependents are not
included. 1973, 1979 are for women aged 14 and over, while 1989 is
restricted to women aged 15 and over. 1973 numbers are from U.S. Depanment of Commerce Series P-20, No. 255, pp. 25-26; 1979 data come from
U.S. Department of Commerce, Series P-20, No. 349, pp. 31-32; 1989 data
are from U.S. Department of Commerce, Series P-20 (1989), p. 55.
7.6
Poverty Rates by Region and Division. 1990 from P-60 Series, No. 175, p.
155; \9S9 from P-60 Series, No. 171, p 149; I960 from P-60 Series, No. 175,
pp. 219-21; and Bureau of Economic Analysis diskettes, released April 22,
1992 and September 6, 1991 (for population).
7.7
Child Poverty Rates, by Race and Region. ChUdren's Defense Fund (1991), p.
147.
6.18
Expected Lifetime Births, by Marital Status and Race. Jencks (1991), p. 86.
7.8
6.19
Poverty Rates Using Different Income Definitions. Green Book (1991). All
persons: p. 1164. Single parent families: p. 1168. Married-couple families: p.
1170.
Percentage Change, Average Hourly Wages by Gender, Education, and
Region. Authors' tabulation of Current Population Survey (CPS) wage data,
see Appendix.
7.9
6.20
AFIX; Participation Rates of Female-Headed Families with Children. Moffitt
(1992), p. 9.
Percent Change in the Median Hourly Wages of Men a n d Women, by State.
Authors' tabulation of Current Population Survey (CPS) wage data, see
Appendix.
6.21
Shares o f Workers Earning Enough to Maintain a Family of Four at 0.75 o f
Poverty and up to the Poverty Line, by Gender and Race. Authors' tabulations of Current Population Survey (CPS) wage data; see the Appendix for
description of analytic methods.
7.10
6.22
Wage Trends Relevant to the Poor and Near-Poor. Same source as Table 6.21.
6.23
Work Experience of the Poor. 1979 from P-60 Series, No. 130, p. 58; 1989
from P-60 Series, No. 168, p. 65.
State and Local Effective Tax Rates in the Ten Largest States as Percent of
Income f o r Families of Four. Mclntyre et al. (1991), pp. 23, 28, 32, 41, 49,
51, 52, 54, 57, 62. Any tax analysis of this nature must make a number of
assumtions regarding tax incidence. For example, in this study half of the
property tax paid on rental units is allocated to the owners of the property,
and half is assigned to the renters in the form of higher rents. We consider
the assumptions in this study to be reasonable and unbiased. The methodology of the study is discussed on pp. 70-72.
6.24
Percentage of Half-lime and Full-Time Worker Equivalents' in Poor Families
with Children. Green Book (1992), pp. 1282-83. FamUies include unrelated
subfamilies.
7.11 Job Growth by Region and Division. Calculated from employment and population figures in Bureau of Economic Analysis diskettes, released April 22,
1992 and September 6, 1991.
6.25
Hours Worked by Family Type in the Bottom Fifths. Green Book (1992), pp.
1424-25.
7.12
Unemployment Rates by Region, Division, and State. Bureau of Labor Statistics diskette (1992).
7.13
City/Suburb/Rural Comparisons of Income Distribution. P-60 Series, No.
177, p. 18-19. "Cities" refer to the Census Bureau terminology "central
cities," "Suburbs" to "not central cities," (but within metropolitan), and
"Rural" to "not metropolitan."
7.14
Poverty Rates by City, Suburban, and Rural Residence. Center on Budget
and Policy (1992), p. 20.
Chapter 7
7.1
Median Family Income, by Region. U.S. Department of Commerce, unpublished historical data (1991c): Table 12.
7.2
Median Income f o r Four-Person Families, by State. U.S. Department of Commerce (1992). Table 1.
466
467
�7.15
Poverty by Region and Residence. P-60 Series, No. 175, p. 154; P-60 Series,
No. 171, pp. 149, 315.
8.11
Private and Total Educational Expenditures, as a Percent of GDP. See note
to Table 8.10.
7.16
Urban a n d Rural Employment Growth. Department of Agriculture (Winter
1991/92), pp. 14-15.
8.12
Public School Expenditures per Pupil. Adjustedf o r Cost of Living Differences
Between States. U.S. Department of Education (1991), p. 28.
7.17
Unemployment Rates, Regular and Adjusted, Urban/Rural. See note to Table
7.16.
8.13
Housing Costs. Apgar et al. (1991), p. 24. See note from Table 8.14 for
description of the methodology used to calculate housing costs.
8.14
Burdens of Buying Homes and Renting. Apgar et al. (1991), p 24. Incomes:
1970 is from the 1970 Census of Populations; 1967-1969 are from the Panel
Survey of Income Dynamics (PSID); 1971-1972 are interpolated from the
PSID and 1970 Census of Population; 1973-1983 are from the American
Housing Survey (AHS); 1983-1989 are from the AHS, adjusted by the Current
Population Survey.
Prime first-time home buyers are married-couple renters ages 25 to 29.
House prices are the AHS median values of houses purchased by first-time
home buyers ages 25 to 29 in 1977, indexed by the Census Department's
Construction Reports C-27 Constant Quality Home Price Index. Mortgage
rates equal Federal Home Loan Bank Board contract mortgage rate. Yearly
payments are calculated based on a 30-year mortgage with a 20% down
payment.
Also included in yearly home costs are property taxes, insurance, fuel and
utilities, and maintenance.
The tax benefits are deducted. Rent is median 1977 contract rent from
AHS, indexed by the Consumer Price Index (CPI) residential rent index, with
depreciation adjustments. Also included in rental costs are fuel and utilities,
property taxes, and insurance.
8.15
Percent of Owner and Renter Families Who Could Not Afford Median Priced
Homes. U.S. Department of Commerce, Series P-20 (1991), p. 11. In this analysis, "family" refers to a group of two or more persons related by birth, marriage or adoption who reside together. Affordability calculations were made
using a conventional fixed-rate 30-year mortgage. The interest rate used is the
average contract interest rate on loans made in the months of February
through May 1988. The cost of the median home is specific to each of the
four Census regions. Income includes wages, salaries, wealth and other types
of "permanent" income.
8.16
Shelter Poverty, by Household Size. Stone (1993), pp. 419, 420, 422, 423.
Averages for panel B use 1979 weights for the number of households in each
category, so that changes in shelter poverty are not a function of the shift to
smaller households over the 1980s. The data for 1991 are estimates based on
predicted family incomes, housing costs, unemployment, and changes in the
number of households.
8.17
Home Ownership Rates by Age of Household Head. Apgar et al. (1991), p. 9-
Chapter 8
8.1
Proficiency in Mathematics and Science, Age 9, 13, and 17 by Race/Ethnicity.
U.S. Department of Education (1991a), p. 64.
8.2
Scholastic Aptitude Test Scores by Race/Ethnic Group. U.S. Department of
Education (1991a), p. 123.
8.3
Proficiency in Mathematics and Science, Grades 4, 8, and 12, by Community. Mathematics: U.S. Department of Education (1991b), pp. 485, 497, 511.
Science: U.S. Department of Education (1992), pp. 145-47.
8.4
Effects of Family Background on Scores of 8tb Graders. U.S. Department of
Education (1989), p. 118. Family background, or "socioeconomic status," is
based on a composite score of parental education and occupations, lamily
income, and household characteristics. The "disadvantaged" group is the
lowest fourth; the "average" is the middle one-half; the "advantaged" group is
the top fourth.
8.5
Average Proficiency of 4tb, 8tb, and 12th Graders by School Rank. U.S.
Department of Education (1991b), p. 13-
8.6
The Shrinking Federal Student Aid Grant. Pell grant data are from The College Board (1991), p. 10. Cost data are for all institutions and are from U.S.
Department of Education (1991a), p. 296. Figures for 1985 and 1987 are
estimates.
8.7
Costs o f College: The Average Percent o f Total Cost Met by Aid and the Percent o f Students with Cost Covered by Expected Family Contribution, by
Family Income. U.S. Department of Education, The Condition o f Education
(1991), pp. 113-14.
8.8
Net Cost o f College as a Ratio of Expected Family Contribution, by Family
Income. U.S. Department of Education, The Condition of Education (1991),
p. 114.
89
Postsecondary Enrollment and Degree Attainment by Race/Ethnic Group.
U.S. Department of Education (1991a), pp. 200, 264, 267, 270.
8 .10
Public Expenditures on Education as a Percent of GDP. Organisation of Economic Co-operation and Development (OECD) (1990b), p. 87.
468
469
�8.18 Percent of Households Living in Deficient and Overcrowded Housing. Lazere
et al. (1991), pp. 24-25. The specific physical and structural defficiences, as
defined by the Bureau of the Census and the Department of Housing and
Urban Development, are given on p. 22 of the report.
8.28
Percent of U.S. Population Obtaining Health Insurance from Specified
Sources. Green Book (1992), p. 312.
8.29
Persons without Health Insurance: Their Attachment to the Workforce.
Green Book (1992, 1991, 1990), pp. 315, 311, 291, respectively.
Health Insurance Coverage of Workers, by Industry. Green Book (1991), p.
312.
8.19
Estimates of Homelessness. Burt and Cohen (1989), p. 2. Estimate of children, p. 28. Upper-bound for children is from Green Book (1992), p. 1183.
8.30
8.20
Characteristics of Service-Using Homeless Adults, in Percent. Green Book
(1992), p. 1187.
8.31 Business Health Care Burden. See note for Table 8.23; Tables 5A and 5B.
8.21 OECD Spending on Health. Schieber and POullier (1991), p. 109.
8.22
The Main Components of Health Care Expenditures in the U.S., per Capita.
Green Book (1992), p. 289.
8.23
The Family Health Care Burden. Families USA Foundation (1991): Tables 4A
and 4B. The costs of health services and supplies come from the Office of
National Cost Estimates. Direct sources of financing health expenditures
include out-of-pocket expenditures, health insurance, and public programs.
Thesefinancingsources were split into public and private sources; private
sources were then divided into household and business contributions. See
methodology section in the Families USA Foundation report for further
description of data sources involved in these calculations.
8.24
8.25
Health Indicators in U.S. and OECD Average. Health Care Financing Administration (1989), pp. 177-78, 185-86 for data from 1970-1986. Data for 1989
are from OECD (1992b), pp. 46-47. TXirkey is excluded from OECD averages
as it has the health statistics and living standards of a Third World country
and would skew the averages.
Infant monality is deaths of infants less than one-year-old, as a percent of
all live births. Perinatal mortality is fetal deaths occurring after 28 weeks of
pregnancy, plus infant deaths occurring prior to seven days after birth. 1986
perinatal mortality for Belgium, France, Italy, and Spain are actually from
1984, and for Ireland, from 1985. Canadian life expectancies in I960 and
1980 are actually from 1961 and 1981, respectively. Portuguese life expectancies in 1980 are from 1981.
Infant and Neonatal Mortality Rates, by Race. National Center for Health
Statistics. United States (1990), p. 107.
8.26
Number and Percent of the Non-aged Population without Health Insurance.
Green Book (1992), p. 318.
8.27
Breaks in Health Insurance During a 28-Month Period. Nelson and Short
(1990). Sample is those numbers of the Survey of Income and Program Participation 1985 panel (interviewed February 1985-August 1987) for whom 28
months of continuous information was available.
470
8.32
Public Health Expenditures as a Percentage of Total Health Expenditures,
OECD Countries. Green Book (1992), p. 330.
8.33
Percent of Women with Children in the Labor Force, by Age of Child. Green
Book (1992), p. 935.
8.34
Child-care Problems as a Constraint on Work, Women 21-29 Years Old. Cattan (1991), p. 5.
8.35
Mothers, 21 to 29 Years Old, Not in the Labor Force, by Selected Characteristics. Cattan (1991), p. 6.
8.36
Child-care Related Constraints on Women's Employment. Veum and Gleason
(1991), p. 16.
8.37
Average Percentage of Family Income Spent on Child Care by Family
Income. Green Book (1992), p. 944.
8.38
Child-care Cost Burden, as Percentage of Income for Low- and Middle-Income Families: Licensed Centers in Four Cities. Cost data are from Children's
Defense Fund (1991), p. 43. State specific median wages are from Current
Population Survey (CPS) wage data; see Appendix. The value of the Earned
Income Tax Credit was added to the income of both family types. The value
of the Dependent Care Tax Credit was added to the income of two-parent
families only, since the income of single-parent families was too low to qualify
for this non-refundable credit. Since these are the only effects of the tax system factored into these estimates (i.e., no tax liabilities are accounted for),
the percentages are underestimates of the actual child-care burden.
Chapter 9
91
Per Capita GDP Growth in Ten Countries. GDP stands for "Gross Domestic
Product," and is essentially the total of all money received for goods and services produced in the economy in a single year. It is not exactly the same
thing as Gross National Product, or GNI> which is total money received from
all productive sources, because GNP also includes net investment income
from abroad. However, the differences between GDP and GNP are usually
471
�very small, which is why GDP per capita is commonly used as an indicator of
national living standards.
Real GDP comes from Organisation of Economic Co-operation and Development (OECD) (1992c); civilian employment comes from U.S. Department
of Labor, Monthly Labor Review (July 1986), p. 96 and (June 1992), p. 113;
and total population comes from OECD (1992c), pp. 156-57. Ideally, this
type of decomposition would only use the adult population since this is the
relevant population when examining labor market participation. However,
international data on adult populations were not readily available. At any
rate, the findings would be similar.
9.2
Living Standards versus Productivity Growth. Growth rates for real per
capita GDP and for productivity (GDP per civilian worker) were computed,
and a straightforward weighted average of these rates was computed. Then
the growth rates were converted to indices, with the weighted average
growth rates equal to 100. The weighted averages were weighted by the populations of the countries; a country's population was taken to be the arithmetic average of the populations in 1979 and 1988. These countries were
selected because the Bureau of Labor Statistics regularly compiles information about their employment levels. Civilian employment levels come from
U.S. Department of Labor, Monthly Labor Review (June 1992), p. 112. Real
GDP comes from Organisation of Economic Co-operation and Development
(OECD) (1992c).
9.3
Hourly Manufacturing Compensation Growth. The consumer price index
listed in the International Monetary Fund (1991) was used to deflate hourly
compensation in all countries except the U.S. We deflated U.S. compensation
growth using the CPI-U-X1. Data for all employees is from U.S. Department of
Labor (1992), p. 3; data for production workers in manufacturing is from
U.S. Department of Labor (1990b).
Per Capita Income Compared to U.S. Organisation of Economic Co-operation
and Development (OECD) (1992c), pp. 146-47
98
Income Distribution: International Comparisons. Green Book (1992), p.
1293.
99
International Poverty Statistics. Smeeding (1992), p. 31. The poverty thresholds used are 40% of the median income in the relevant country, adjusted for
family size. Income is post-tax and post-transfer, and includes the value of
food stamps.
910
The Impact of Taxes and Transfers on International Poverty Rates. Smeeding
(1992), p. 33.
9.11
Tax and H-ansfer System Effectiveness. Smeeding (1992), p. 34.
9 12
Poverty Rates and Transitions Out of Poverty f o r fiamilies with Children.
Duncan et al. (1991): Table 1.
913
Absolute Poverty Among Children, Adults, and the Elderly. U.S. Department
of Labor, Monthly Labor Review (July 1986), p. 96, and (June 1992), p. 112.
914
Jobs Created in Ten Countries. Civilian employment from U.S. Department of
Labor, Monthly Labor Review (July 1986), p. 96, and (June 1992), p. 112.
915
Contributions to Employment Growth in Ten Countries. See note Table 9.14.
Population from Organisation of Economic Co-operation and Development
(OECD) (1992), pp. 156-57. Ideally, this tyjje of decomposition would only
use the adult population since this is the relevant population when examining labor market participation. However, international data on adult populations were not readily available. At any rate, the findings would be similar.
916
Hours Worked per Employee in Manufacturing, Indexed to the United States.
Leete-Guy and Schor (1992), p. 16.
917
Paid Vacation in European Countries. Leete-Guy and Schor (1992), p. 19.
9-18
Public Spending on TYaining and Placement, Direct Job Creation and Subsidies, as Percent of GDP. Organisation of Economic Co-operation and Development (OECD) (1991), pp. 238-49.
9.19
Unemployment in OECD Countries. Organisation of Economic Co-operation
and Development (OECD) (1992a), p. 192.
9.20
Unemployment and Unemployment Compensation. See notes to Tables 918
and 9.19.
Productivity Growth Rates. Civilian employment levels are from U.S. Department of Labor, Monthly Labor Review (June 1992), p. 112. Real GDP comes
from Organisation of Economic Co-operation and Development (OECD)
(1992c).
9.4
97
9.5
9.6
Manufacturing Hourly Compensation Compared, (Using Purchasing Power
Rzrity Exchange Rates). Exchange rates are from Organisation of Economic
Co-operation and Development (OECD) (1992), p. 156-57. Data for all workers are from U.S. Department of Labor (1992), p. 3. Data for production
workers are from U.S. Department of Labor (1990b), and from U.S. Department of Labor (1992), p. 8.
Manufacturing Hourly Compensation Compared, (Using Market Exchange
Rates). See note to Table 9 5.
472
473
�Figure Notes
Figures:
1A
Median Family Income, 1967-1990. See note to Table 1.1.
IB
Annual Growth of Median Family Income, 1947-1990. See note to Table
1.2.
IC
Ratio of Family Income of Top 5% to Lowest 20%, 1947-1989- See note to
Table 1.7
ID
Income Growth Among Top Fifth and By Fifth, 1980-1S>89. See note to
Table 1.9.
IE
Shares of Aggregate Income Growth Among Top Fifth, 1980-1989. See note
to Table 1.13.
IF
Distribution of Labor and Capital Incomes, Top 20%, 1980-1989- See note
to Table 1.18.
IG
Growth in Annual Hours Worked by All Wives, 1979-1989- See note to
Table 1.24.
1H
Growth in Family Income, By Wives' Higher Earnings, 1979-1989- See note
to Table 1.26.
II
Family Income By Age o f Householder, 1989- P-60 Series, No. 172, pp.
58-59.
1J
Median lamily Income By Cohort and Age, 1949-1989- See note to Table
133.
2A
Federal Tax Burden, 1977 and 1989- See note to Table 2.6.
2B
Federal Income Tax: Cumulative Returns by Marginal Rates, 1979 and
1988. Bakija and Steuerle (1991): Table A.5.
2C
Excise Tax Burden, 1977 and 1989- See note to Table 2.8.
2D
Personal Income Tax Burden, 1977 and 1989. See note to Table 2.8.
2E
Taxed and Untaxed Corporate Profits as a Percent of GDP, 1946-1991. See
note to Table 2.10.
2F
Corporate Profits Taxes, 1946-1991- See note to Table 2.11.
2G
Total Tax Receipts As a Percent o f GNP, 1930-1990. Bakija and Steuerle
(1991): Table A l.
2H
Federal Revenue Sources, 1991- See note to Table 2.13.
21
State and Local Revenue Sources, 1991. See note to Table 2.13.
3A
Real Hourly Earnings of Production and Nonsupervisory
Workers,
1967-1991- See note to Table 3 3.
3B
Real Weekly Earnings of Production
and Nonsupervisory
Workers,
19671991- See note to Table 3 3.
3C
Real Wage Growth By Wage Percentile, 1979-1991- See note to Table 3 5.
475
�3D
3E
3F
3G
3H
31
3J
3K
3L
3M
3N
4A
4B
4C
4D
4E
4F
4G
5A
5B
5C
5D
5E
6A
6B
6C
6D
Employment at or Below Poverty Level Wages, By Race, 1973-1991. See
notes to Table 3.10, Table 3.11, and Table 312.
Change in Real Hourly Wage By Education, 1973-1991. See note to Table
319.
Real Entry Level Wages f o r High School Graduates, 1973-1991. See note to
Table 3 23.
Real Entry Level Wages f o r College Graduates, 1973-1991. See note to Table
3 23.
Bay in Expanding and Shrinking Industries, 1948-1987. See note to Table
3.27.
ftade Induced Changes in Labor Supply, 1967-1985. See note to Table
330.
Union Play Premiums, 1989 Dollars. See note to Table 3.34.
Value of Minimum Wage, 1960-1992. See note to Table 3.39.
Real Growth in Executive After-Tax Pay, 1979-1989- See note to Table 3.46.
Black-White Wage Differentials Among Young Workers, 1973-1989- See note
to Table 3.49.
Growth in Skill and Education Requirements, 1970-2000. See note to Table
351.
Real Rates of Underemployment, 1973-1991- See note to Table 4.2.
Duration of Recessions, 1948-1992. See note to Table 4.3Civilian Employment Growth, 1947-1991- See note to Table 4.9.
Part-Time Employment Rates, 1973-1991. See note to Table 4.10.
Partially or Fully Paid Medical Benefits Offered to Employees, by Full- or
Bart-time Status, 1991. Hewitt Associates survey as in Callaghan and Hartmann (1991): Figure 13, p. 16.
Multiple Jobholding Rates, 1979-1991- See note to Table 4.16.
Temporary and Personnel Services Industry Employment, 1973-1991. See
note to Table 4.21.
Growth of Household Wealth Per Adult, 1949-1989- See note to Table 5.1.
Distribution of Income and Wealth, 1989. See note to Table 5 2.
Distribution o f Wealth, 1983-1989- See note to Table 5.6.
Change in Net Worth By Wealth Class, 1983-1989- See note to Table 5 8.
Distribution o f Net Worth By Wealth Class, 1983-1989- Based on data in
liible 5.8, last column, adjusted to sum to 100%, as table data does not sum
to 100 due to negative growth for the middle and second quartiles.
Predicted versus Actual Poverty Rates, 1959-1990. See note to Table 6.1.
Poverty Rates by Price Index, 1974-1990. CPI-U-X1 rates to 1989 from P-60
Series, No. 171; 1990 figure from authors' communication with the Bureau
of the Census; for CPI-U rates see note to Table 6.1.
Persistent Poverty i n Industrialized Countries.- Households With Children,
Mid-1980s. McFate (1991): Figure 2.
Poverty Before and After Transfers: Elderly versus All Persons, 1967-1990.
See note to Table 6.12.
476
6E
6F
6G
6H
7A
7B
7C
7D
7E
7F
7G
7H
8A
8B
8C
8D
8E
8F
8G
8H
Expected Lifetime Births by Marital Status and Race, 1SH50-1987. See note
to Table 6.18.
Poverty Before and After Tt-ansfers. Families with Children, 1967-1990. See
note to Table 6.12.
Demographic and Benefit-Sum Trends, 1960-1989- Mofifit (1992): Figure 4.
Data brought forward to 1989 from P-60 Series, 1981-1989 (headship); U.S.
National Center for Health Statistics, Monthly Vital Reports, 1981-1989
(divorce); U.S. National Center for Health Statistics, Advance Report o f
Final Natality Statistics, 1981-1989 (out-of-wedlock births). Added data for
benefit sum are from Green Book, (1992), pp. 1210-11.
AFDC Participation Rates, 1967-1987. See note to Table 6.20.
Percentage Point Change in Middle Class by Region, 1969-1989 See note
to Table 7.3Distribution of Changes in States' Income Inequality, 1979-1989. See note
to Table 7.4.
Percentage Change in Median Hourly Wages By State, 19791989. Authors'
tabulations of Current Population Survey (CPS) wage data, see Appendix.
Percentage Change in Median Hourly Wages By State, 1989-1991. Authors'
tabulations of Current Population Survey (CPS) wage data, see Appendix.
State and Local Effective Tax Rates in the Five Most Populated States, 1991See note to Table 7.10.
Unemployment Rates by Regions, 1979-1991- See note to Table 7.12.
City, Suburban and Rural Poverty Rates, 1973-1990. See note to Table
7.14.
Urban and Rural Unemployment Rates, 1979-1990. See note to Table 7.17.
Total College Costs by Type of Four-Year Institution, 1976-1989- U.S. Department of Education (1991), pp. 296-97.
Enrollment Rates for 3- & 4-year-olds in Primary Education by Race/
Ethnicity, 1974-1989- U.S. Department of Education (1992), p. 18.
Percent of 18- to 24-year-olds Completing High School, 1970-1989- Carter
and Wilson (1992): Table 1. Series for blacks and Hispanics smoothed with
three-year moving average to reveal trend.
Percent of 18- to 24-year-olds in College, 19701989- Carter and Wilson
(1992): Table 1. Series for blacks and Hispanics smoothed with three-year
moving average to reveal trend.
Beginning Teacher Salaries versus Other College Graduates, 1972-1991Educational Research Service, (editions since 1973-1974).
After-Tax Home Owning and Rental Costs, 1967-1990. See note to Table
8.13.
Home Owning and Rental Costs as a Percent of Income, 1967-1990. See
note to Table 8.14.
Low-Rent Units and Low-Income Renters, 1970-1989. Lazere, et al. (1991),
p. 5.
477
�81
8J
8K
8L
8M
8N
9A
9B
9C
9D
9E
9F
New Federal Funded Low-Income Housing Units, 1977-1991- Lazere, et al.
(1991) , p. 33.
Per Capita National Health Care Expenditures, 1960-1990. See note to
Table 8.22.
Medical Care Inflation versus Inflation on All Items, 1969-1991. Economic
Report of the President (ERP) (1992), p. 361.
Employer Spending on Private Health Insurance, 1965-1990. Levit and
Cowan (1991), p. 86.
Women in the Labor Force by Presence of Children, 1947-1991. Green Book
(1992) , p. 935.
Distribution of Child-Care Choices for Preschool Children, 1965-1990.
Wilier et al. (1991), p. 44.
Productivity Growth Kates, 1960-1989. See note to Table 9.3.
Hourly Manufacturing Compensation Growth, 1979-1S>89. See note to
Table 9.4.
Comparative Manufacturing Compensation, 1989 (Using Purchasing
Power Parity Exchange Rates). See note to Table 9.5.
Comparative Manufacturing Compensation, 1989 (Using Market Exchange
Rates). See note to Table 9.6.
Impact of Taxes and Transfers on Poverty Rates, Mid-1980s. See note to
Table 910.
Poverty Rates and Transitions out of Poverty of Families with Children,
Mid-1980s. See note to Table 9.12.
478
Appendix: Wage Analysis Computations
This appendix provides background information on our analysis of wage data
from the Current Population Survey (CPS), which is prepared by the Bureau of the
Census for the Bureau of Labor Statistics (BLS). Specifically, we analyze computer
tapes provided by the BLS which have a full year's data on the outgoing rotation
groups (ORG) in the CPS. We believe that the CPS ORG files allow for a more timely,
up-to-date, and more accurate analysis of wage trends then the traditionally used
March CPS files (which, for 1991, will not be available until the end of 1992) while
keeping within the familiar labor force definitions and concepts employed by the
CPS.
The ORG files provide data on those CPS respondents in either the fourth or
eighth month of the CPS (i.e., in groups four or eight, out of a total of eight groups).
Therefore, in any given month the ORG file represents a quarter of the CPS sample.
For a given year, the ORG file is equivalent to three months of the entire CPS (one
quarter of the 12 monthly surveys). For our analysis, we use the full year ORG samples, with sample sizes ranging from 155,265 in 1979 to 171,296 in 1991.
Changes in annual or weekly earnings can result from changes in hourly earnings
or from more working time (either more hours per week or more weeks per year).
Our analysis is centered around the hourly wage, which represents the pure price of
labor (exclusive of benefits), because we are interested in changing pay levelsforthe
workforce and its subgroups. We do this to be able to clearly distinguish changes in
earningsresultingfrommore (or less) work rather than more (or less) pay. Our analysis, therefore, does not take into account that weekly or annual earnings may have
changed because of longer working hours or lesser or greater opportunities for
employment.
In our view, the ORG files provide a better source of data for wage analysis than
the traditionally used March CPS files. In order to calculate hourly wages from the
March CPS, analysts must make calculations using at least two of three retrospective
variables: the annual earnings, weeks worked, and usual weekly hours worked in
the year prior to the survey. Limiting the March sample to full-time, full-year workers
is increasingly unsatisfactory since this is only two-thirds of wage earners and there
can be a significant variation of hours worked in this group over time and year by
year. In contrast, respondents in the ORG are asked a set of questions about hours
worked, weekly wages, and (for workers paid by the hour) hourly wages in the week
prior to the survey. In this regard, the data from the ORG are likely to be more reliable than data from the March CPS. The ORG files are also more current and have a
much larger sample size.
Our sub-sample includes all wage and salary workers with valid wage and hours
data, whether paid weekly or by the hour. Specifically, in order to be included in our
sub-sample, respondents had to meet the following criteria:
—Aged 18 to 64;
—Employed in the public or private sector (self-employed were excluded);
479
�-Hours worked within the valid range in the survey (0-99 hours per week);
and,
—Either hourly or weekly wages within the valid survey range (top-coding
problems discussed below).
For those who met these criteria, an hourly wage was calculated in the following
manner. If a valid hourly wage was reported, that wage was used throughout our
analysis. For salaried workers (those who only report a weekly wage), the hourly
wage was their weekly wage divided by their hours worked. CPS weights were
applied to make the sample nationally representative.
For the survey years 1979, 1987, and 1988, the weekly wage is top-coded at
$999.00. Particularly for the later years, this truncation of the wage distribution creates a downward bias in the mean wage. This is especially problematic for comparisons between 1987 and 1988 to later years, when the top code was raised to
$1,923.00. Fortunately, the 1987 and 1988 ORG files have an unedited field that
allowed us to impose the new higher top-code present in the ORGfilesfor1989 and
beyond. This unedited field in the 1987 and 1988 ORG files includes reported
weekly wages up to the new top-code of $1,923. There are no imputations in that
data field.
If a weekly wage from the edited field (with the lower top-code) was top-coded,
and a valid wage existed in the unedited field, we used the wage with the higher topcode. At the top of the wage distribution, this procedure yielded substantially higher
mean houriy wages. For example, in 1988, for males with more than six years of
education beyond high school the mean hourly wage increased by $1.65, from
$16.36 to $18.01, when using the field with the higher top-code.
We extended our analysis back to 1973 by pooling the 1973 and 1974 May CPS.
The 1974 data were deflated to 1973 dollars using the Personal Consumption
Expenditures (PCE) (1972 base year).
Demographic variables are also used in the analysis. Education refers to years of
school completed. Our race variable is comprised of four mutually exclusive
categories:
—White, non-Hispanic;
—Black, non-Hispanic;
—Hispanic, any race;
—All others.
Time Period
Our analysis tracks wages from 1979 onward by examining trends from one cyclical peak, 1979, to another, 1989, and then from 1989 to the most recent year, 1991.
The post 1989 wage trends are influenced by the cyclical downturn and may partiallyreflectshort-term,recession-induceddevelopments. Vtfe believe, however, that
the 1989-1991 wage trends primarilyreflectlonger- term trends. For one, the use of
hourly wages (especially person-weighted) as our measure moderates the impact of
the cycle. Two, the trends we observe from 1989 to 1991 prevailed before the recession. This can be seen in the year-by-year wage trends from 1987 to 1991.
We do analyze a number of trends from 1987 onwards. There arc several reasons
why this particular time period is analyzed. In order to show that the wage trends
we identified preceded the onset of the recession it was necessary to examine some
pre-recession years. However, for technical reasons—the ORG data has a different
top-code (and no unedited field with the new top code) in 1986—we could not
examine trends prior to 1987. Therefore, the wages of high-wage workers, such as
college-educated workers, are not comparable between 1986 and 1987 (although
there is a comparability between 1979 and 1987 because the old top-code in 1979
was hardly binding).
Inflation Adjustment
To adjust for inflation we used the CPI-U-X1, which corrects the more commonly
used Consumer Price Index (CPI) for allegedly overstating inflation in the 1970s
(starting in 1983 the two indices are the same). This is the deflator we use throughout the book.
480
481
�Bibliography
American Federation of Teachers. Survey & Analysis of Salary D-ends, 1989- Washington, DC: AFT, 1989.
Apgar, William C, Jr., Denise DiPasquale, Jean Cummings and Nancy McArdle. "The
Sute of the Nation's Housing: 1991" Cambridge, MA: Joint Center for Housing
Studies of Harvard University, 1991.
Bakija, Jon, and C. Eugene Steuerle. "Individual Income Taxation Since 1948."
National Tax Journal, Vol XLLiy No. 4, Part 2, December 1991, pp. 451-475.
Bane, Mary Jo and David T. Ellwood. "Slipping Into and Out of Poverty: The Dynamics of Spells." The Journal of Human Resources, Vol. XXI, No. 1, Winter 1986, pp.
1-23.
Barancik, Scott, and Isaac Shapiro. Where Have all the Dollars Gone: A State by
State Analysis of Income Disparities in the 1980s. Washington, DC: Center on
Budget and Policy Priorities, 1992.
Blackbum, McKinley L., David E. Bloom, and Richard B. Freeman. "Changes in Earnings Differentials in the 1980s: Concordance, Convergence, Causes, and Consequences." Working Paper No. 3901. Cambridge, MA: National Bureau of Economic Research, 1991.
Blank, Rebecca M. 'Are Part-Time Jobs Bad Jobs?" In Gary Burtless, ed., A Future of
Lousy Jobs? Washington, DC: Brookings Institution, 1990.
Blank, Rebecca M. "Why Were Poverty Rates So High in the 1980s?" Working Paper
No. 3878. Cambridge, MA: National Bureau of Economic Research, 1991.
Board of Governors, Federal Reserve System. Balance Sheets for the U.S. Economy:
1945-1989. Washington, DC: FRS, April 1990.
Board of Governors, Federal Reserve System. Balance Sheets for the U.S. Economy.
1960-91- Washington, DC: FRS, March 1992.
Borjas, George J., Richard B. Freeman, and Lawrence F. Katz. "On the Labor Market
Effects of Immigration and Trade." Working Paper No. 3761. Cambridge, MA:
National Bureau of Economic Research, 1991.
Bound, John, and Richard B. Freeman. "What Went Wrong? The Erosion of Relative
Earnings and Employment Among Young Black Men in the 1980s." Unpublished,
1991.
Burt, Martha R., and Barbara E. Cohen. 'America's Homeless: Numbers, Characteristics, and Programs that Serve Them." Urban Institute Report 89-3 Washington,
DC: Urban Institute Press, 1989.
Callaghan, Polly, and Heidi Hartmann. Contingent Work: A Chart Book on Fart-Time
and Temporary Employment. Washington, D.C: Economic Policy Institute, 1991.
483
�Cancian, Maria, Sheldon Danziger, and Peter Gottschalk. "The Changing Contributions of Men and Women to the Level and Distribution of Family Income,
1968-1988." Prepared for the Jerome Levy Economics Institute Conference: Economic Inequality at the Close of the Tiventieth Century, Bard College, June 1991.
Democratic Study Group. "They Didn't Come to the Party: A Tough Decade for Families in the Middle." Special Report, No. 101-32. Washington, DC: U.S. House of
Representatives, 1990.
Card, David. "The Effect of Unions on the Distribution of Wages: Redistribution or
Relabelling?" Working Paper No. 287. Depanment of Economics, Princeton:
Princeton University, 1991.
Duncan, Greg, et al. "Poverty and Social Assisunce Dynamics in the United States,
Canada and Europe." Paper presented at the conference Poverty and Public Policy, Washington, DC: Joint Center for Political and Economic Studies, September,
1991.
Card, David. "The Effect of Unions on the Distribution of Wages: Redistribution or
Relabeling?" Working Paper No. 287. Department of Economics, Princeton:
Princeton University, 1991.
Duncan, Greg J., Timothy Smeeding and Willard Rodgers. W(h)ither the Middle
Class: A Dynamic View? Survey Research Center. Ann Arbor: University of Michigan, 1991.
Carre, Francoise J. "Temporary Employment in the Eighties." In Virginia L. duRivage,
ed., New Policies for the Part-Time and Contingent Workforce. Economic Policy
Institute Series. Armonk: M.E. Sharpe, Inc., 1992.
Eargle, Judith. Household Wealth and Asset Ownership: 1988. Current Population
Reports, Series P-70, No. 22, U.S. Bureau of the Census, Washington, DC: U.S.
Government Printing Office, 1990.
Carter, Debroah J., and Reginald Wilson. Minorities in Higher Education. Washington, DC: American Council on Education, 1992.
The Economist. "Bosses Get Rich." December 23, 1989, pp. 92-93.
Cattan, Peter. "Child-care Problems: An Obstacle to Work." Monthly Labor Review,
October 1991, pp. 3-9.
Center on Budget and Policy Priorities. 19SX) Poverty Tables. Washington, DC: CBPI^
1992.
Children's Defense Fund. The State of America's Children: 1991. Washington, DC:
CDF, 1991.
Children's Defense Fund. The Health of America's Children-. Maternal and Child
Health Data Book. Washington, DC: CDF, 1989.
College Board. Trends in Student Aid. 1981 to 1991- New York: College Board Publications, 1991.
Congressional Budget Office. "Shares and Sources of Family Income: All Families,
1977, 1980, 1985, 1988, 1989 " Unpublished, 1992.
Costrell, Robert M. The Effects of Industry Employment Shifts on Wage Growth.
1948-87. A study prepared for the Joint Economic Committee, Washington, DC:
JEC, August 1988.
Cuder, David M., and Lawrence F. Katz. "Rising Inequality? Changes in the Distribution of Income and Consumption in the 1980s." Presented at the American Economic Association Meetings, January 1992.
Danziger, Sheldon H., and Daniel H. Weinberg. "Market Income, Income Transfers
and the Trend in Poverty." Paper presented at the conference Poverty and Public
Policy. What Do We Know? What Should We Do? Institute for Research on Poverty.
Madison: University of Wisconsin, May 28-30, 1992.
484
Educational Research Service. Salaries Paid Professional Personnel in Public
Schools. Reston.- ERS, various dates.
Educational Research Service. Salaries Paid Support Personnel in Public Schools.
Reston: ERS, various dates.
FamUies USA Foundation. Health Spending: The Growing Threat to the Family Budget. Washington, DC: FamUies USA Foundation, December 1991.
Faux, Jeff. Family Incomes in Trouble. Briefing Paper. Washington, DC. Economic
Policy Institute, 1986.
Fisher, Gordon M. Thends in Money Income and Poverty. Unpublished. U.S. Census
Bureau, November 1986.
Freeman, Richard B. "How Much Has De-unionization Contributed to the Rise in
Male Earnings Inequality?" Working Paper No. 3826. Cambridge, MA: National
Bureau of Economic Research, 1991.
Haber, Sheldon E., Enrique L. Lamas, and Jules H. Lichtenstein. "On Their Own: The
Self-employed and Others in Private Business." Monthly Labor Review, May 1987.
Harrison, Bennett, and Barry Bluestone. The Great U-Tum. New York: Basic Books,
1988.
Health Care Financing Administration. Health Care Financing Review. 1989 Annual
Supplement. Washington, DC: U.S. Government Printing Office, December 1989Henwood, Doug. Left Business Observer. Supplement No. 1. New York: Henwood,
September 1991.
International Monetary Fund. International Financial Statistics. Washington, DC:
IMF, monthly.
485
�Jencks, Christopher. "Is the American Underclass Growing?" In Christopher Jencks
and Paul E. Peterson, eds., The Urban Underclass. Washington, DC: Brookings
Institution, 1991.
Jerome Levy Economics Institute Conference on Economic Inequality at the Close
of the Tiventieth Century, Bard College, June 1991. Revised January, 1992.
Mishel, Lawrence, and Jacqueline Simon. The State of Working America. Washington, DC: Economic Policy Institute, 1988.
Mishel, Lawrence R. "The Late Great Debate on Deindustrialization." Challenge, January/February 1989.
Joint Economic Comminee. Economic Indicators. Washington, DC: JEC, Monthly.
Mishel, Lawrence, and Jared Bernstein. "Declining Wages for High School and College Graduates." Washington, DC: Economic Policy Institute, 1992.
Joint Economic Committee. Families on a Theadmill: Work and Income in the
1980s. Washington, DC: JEC, January 1992.
Mishel, Lawrence, and David Frankel. The State of Working America, 1990-91 edition. Economic Policy Institute Series. Armonk: M.E. Sharpe, Inc., 1991-
Karier, Thomas. "What Happened to the Corporate Profit Tax?" Working Paper No.
37. Annandale-on-Hudson: Jerome Levy Economics Institute, May 1990.
Mishel, Lawrence, and Ruy Teixeira. The Myth of the Coming Labor Shortage. Jobs,
Skills, and Incomes ofAmerica's Workforce 2000. Washington, DC: Economic Policy Institute, 1991.
Katz, Lawrence F., and Kevin M. Murphy. "Changes in Relative Wtges, 1963-1987:
Supply and Demand Factors." Unpublished. Cambridge: Harvard, 1990.
Krueger, Alan B. "How Computers Have Changed the Wage Structure: Evidence
from Microdata, 1984-89 " Presented at National Bureau of Economic Research
Labor Studies Workshop, July 1991.
Lazere, Edward B., et al. A Place to Call Home: The Low Income Housing Crisis Continues. Washington, DC: Center on Budget and Policy Priorides, 1991.
Moffit, Robert. "Incentive Effects of the U.S. Welfare System: A Review." Journal of
Economic Literature, Vol. XXX, No. 1, March 1992, pp. 1-61.
Murphy, Kevin, and Finis Welch. "Recent Trends in Real Wages: Evidence From
Household Data." Paper prepared for the Health Care Financing Administration of
the U.S. Department of Health and Human Services. Chicago: University of Chicago, January 1989.
Leete-Guy, Laura, and Juliet B. Schor. The Great American Time Squeeze: Trends in
Work and Leisure, 1969-1989- Washington, DC: Economic Policy Institute, 1992.
Nelson, Charles, and Kathleen Short. Health Insurance Coverage: 1986-88. Current
Population Reports, Household Economic Studies, Series P-70, No. 17. Washington, DC: U.S. Government Printing Office, 1990.
Levit, Katharine R., and Cathy A. Cowan. "Business, Households, and Governments:
Health Care Costs, 1990." Health Care Financing Review, Vol. 13, No. 2, Winter
1991, pp.83-91.
Nelson, F. Howard. An Interstate Cost-of-Living Index. Washington, DC: American
Federation of Teachers, November 1989-
Levit, Katharine R., Mark S. Freeland, and Daniel R. Waldo. "Health Spending and
Ability to Pay: Business, Individuals, and Government." Health Care Financing
Review.Vol 10, No. 3, Spring 1989.
McFate, Katherine. Poverty, Inequality and the Crisis of Social Policy. Washington,
DC: Joint Center for Political and Economic Studies, 1991.
Mclntyre, Robert S., et al. A Far Cry From Fair. Washington, DC: CitizensforTax Justice, April 1991.
Meisenheimer II, Joseph R., Earl F. Mellor, and Leo G. Rydzewski. "The Labor Market
in 1991" Monthly Labor Review, Vol. 115, No. 2, February 1992, pp. 3-17.
Merrill Lynch. Weekly Economic and Financial Commentary. May 18, 1992.
Mishel, Lawrence. The Polarization of America: The Loss of Good Jobs, Falling
Incomes and Rising Inequality. Washington, DC: Industrial Union Department,
AFL-CIO, October 1986.
486
OECD. Revenue Statistics of OECD Member Countries: 1965-1989. Paris:
Organisation for Economic Co-operation and Development, 1990a.
OECD. Education in OECD Countries: 1987-88. Paris: Organisation for Economic
Co-operation and Development, 1990b.
OECD. Employment Outlook. Paris: Organisation for Economic Co- operation and
Development, 1991.
OECD. OECD Economic Outlook. No. 51. Paris: Organisation for Economic Co-operation and Development, 1992a.
OECD. OECD in Figures. 1992 Edition. Paris: Organisation for Economic Co-operation and Development, 1992b.
OECD. National Accounts: Main Aggregates. Paris: Organisation for Economic
Co-operation and Development, 1992c.
President of the United States. Economic Report of the President. Washington, DC:
U.S. Government Printing Office, annual.
487
�Rasell, M. Edith, and Lawrence Mishel. Shortchanging Education: How U.S. Spending on Grades K-12 Lags Behind Other Industrial Nations. Washington, DC: Economic Policy Institute, 1990.
Ruggles, Patricia. Drawing the Line: Alternative Poverty Measures and Their Implications f o r Public Policy. Washington, DC: Urban Institute, 1990.
Ruggles, Patricia. "Measuring Poverty." Pocus. Vol. 14, No. 1, Spring 1992, pp. 1-9
Sawhill, Isabel V, and Mark Condon. "Is U.S. Income Inequality Really Growing?"
Policy Bites, No. 13. Washington, DC: Urban Institute, June 1992.
Tilly, Chris. "Continuing Growth of Part-Time Employment." Monthly Labor Review,
Vol. 114, No. 3, March 1991, pp. 10-18.
Towers, Perrin and Company. Worldwide Total Remuneration, 1988. October 1988.
U.S. Department of Agriculture, Economic Research Service. Rural Conditions and
Thends. Vol. 2, No. 4. Washington, DC: ERS, Winter 1991/92.
U.S. Department of Commerce, Bureau of Economic Analysis. Survey of Current
Business. Washington, DC: U.S. Government Printing Office, various dates.
Schieber, George J., and Jean-Pierre Poullier. "International Health Spending: Issues
and Trends." Health Affairs, Vol. 10, No. 1, Spring 1991, pp. 106-116.
U.S. Department of Commerce, Bureau of the Census. "Marital Status and Living
Arrangements." Current Population Reports: Series P-20. Washington DC: U.S.
Government Printing Office, various dates.
Sekscenski, Edward S. "Women's Share of Moonlighting Nearly Doubles During
1969-79" Monthly Labor Review,Voi. 103, No. 5, May 1980.
U.S. Department of Commerce, Bureau of the Census. Current Population Reports:
Series P-60. Washington, DC: U.S. Government Printing Office, various dates.
Shapiro, Isaac. "No Escape: The Minimum Wage and Poverty." Washington, DC: Center on Budget and Policy Priorities, June 1987.
U.S. Department of Commerce, Bureau of the Census. Current Population Reports:
Series P-70. Washington, DC: U.S. Government Printing Office, various dates.
Shapiro, Isaac, and Marion Nichols. Far From Fixed. An Analysis of the Unemployment Insurance System. Washington, DC: Center on Budget and Policy Priorities,
March 1992.
U.S. Department of Commerce, Bureau of the Census. Who Can Afford to Buy a
House? Current Housing Report, Series H121/91-1. Washington, DC: U.S. Government Printing Office, 1991a.
Sivestri, George, and John Lukasiewicz. "Occupational Employment Projects."
Monthly Labor Review. Vol. 114, No. 11, November 1991.
U.S. Department of Commerce, Bureau of the Census. Trends in Relative Income:
1964 to 1989. Current Population Reports, P-60 Series, No. 177. Washington, DC:
U.S. Government Printing Office, 1991b.
Smeeding, Timothy M. "Why the U.S. Antipoverty System Doesn't Work Very Well."
Challenge, January-February 1992, pp. 30-35.
Smeeding, Timothy, Barbara Boyle Torrey, and Martin Rein. "Patterns of Income and
Poverty: The Economic Status of Children and the Elderly in Eight Countries." In
Palmer, Smeeding, and Torrey, eds., The Vulnerable. Washington, DC: Urban Institute, 1988.
U.S. Department of Commerce, Bureau of the Census. "Historical Tables." UnpubUshed, 1991c.
U.S. Department of Commerce, Bureau of the Census. "Estimates of Median FourPerson Family Income, by State: 1974-89 " Current Population Reports: TP-61.
Washington, DC: U.S. Government Printing Office, 1992.
Stinson, John F, Jr. "Moonlighting by Women Jumped to Record Highs." Monthly
Labor Review, Vol. 109, No. 11, November 1986.
U.S. Department of Commerce, Bureau of Economic Analysis. National Income and
Product Accounts. Washington, DC: U.S. Government Printing Office, 1992.
Stone, Michael E. Shelter Poverty: New Ideas on Housing Affordability. Philadelphia:
Temple University Press, forthcoming 1993.
U.S. Department of Education, National Center for Education Statistics. The Condition o f Education. Washington, DC: U.S. Government Printing Office, various
dates.
Swartz, Katherine, and Timothy D. McBride. "Spells Without Health Insurance: Distributions of Durations and Their Link to Point-in-Time Estimates of the Uninsured." Inquiry. Vol 27, Fall 1990, pp. 281-88.
Tax Foundation. Tax Burden by Income Class: 1986-1987. Washington, DC: Tax
Foundation, 1989.
Teixeira, Ruy, and Lawrence Mishel. Myth of Rural Labor Shortage. Washington, DC:
Economic Policy Institute, 1992.
488
U.S. Department of Education, National Center for Education Statistics. Digest o f
Education Statistics, 1989. Wtshington, DC: U.S. Government Printing Office,
1989U.S. Department of Education, National Center for Education Statistics. Digest of
Education Statistics, 1991. Wishington, DC: U.S. Government Printing Office,
1991a.
489
�U.S. Depanment of Education, Office of Educational Research and Improvement.
The State of Mathematics Achievement: NAEP's 1990 Assessment of the Nation
and the THal Assessment of the States. Prepared by Educational Testing Service.
Washington, DC: U.S. Government Printing Office, 1991b.
U.S. Department of Education, Office of Educational Research and Improvement.
Trends in Academic Progress. Prepared by Educational Testing Service. Washington, DC: U.S. Government Printing Office, 1991c.
U.S. Department of Labor, Bureau of Labor Statistics. Multiple Jobholding Reached
Record High in May 1989- USDL: 89-529, Washington, DC, November 6, 1989d.
U.S. Department of Labor, Bureau of Labor Statistics, Comparative Real Gross
Domestic Product, Real GDP Per Capita, and Real GDP Per Employed Person.Fourteen Countries.-1950-1989. Washington, DC: Office of Productivity and Technology, June 28, 1990a.
U.S. Department of Education, Office of Educational Research and Improvement.
The 1990 Science Report Card. Prepared by Educational Testing Service. Washington, DC: U.S. Government Printing Office, 1992.
U.S. Department of Labor, Bureau of Labor Statistics, Office of Productivity and Technology. Hourly Compensation Costs for Production Workers, All Manufacturing,
34 Countries, 1975 and 1979-89- Washington, DC: U.S. Government Printing
Office, September 1990b.
U.S. Department of Health and Human Services. National Center for Health Statistics. Health United States: 1989- Washington, DC: U.S. Government Printing
Office, 1990.
U.S. Department of Labor, Bureau of Labor Statistics. Employment, Hours, and
Earnings, United States. Volume I , 1909-90. BuUeting 2370. Washington, DC:
U.S. Government Printing Office, 1991a.
U.S. Department of Health and Human Services. National Center for Health Statistics. Advanced Report of Final Natality Statistics . Washington, DC: U.S. Government Printing Office, various years.
U.S. Department of Labor, Bureau of Labor Statistics. Multiple Jobholding
Unchanged in May 1991- USDL No. 91-547, News, October 28, 1991b.
U.S. Department of Health and Human Services. National Center for Health Statistics. Monthly Vital Reports . Washington, DC: U.S. Government Printing Office, various years.
U.S. Department of Labor, Bureau of Labor Sutistics. Employee Compensation in
the Private Nonfarm Economy, 1974. Bulletin 1963. Washington, DC: U.S. Government Printing Office, 1977.
U.S. Department of Labor, Bureau of Labor Sutistics. Labor Force Statistics Derived
From the Current Population Survey: A Databook, Bulletin 2096, Washington,
DC: U.S. Government Printing Office, September 1982.
U.S. Department of Labor, Bureau of Labor Sutistics. Supplement to Labor Force
Statistics Derived From the Current Population Survey: A Databook, Bulletin
2096-1. Wtshington, DC: U.S. Government Printing Office, May 1984.
U.S. Department of Labor, Bureau of Labor Statistics, Office of Productivity and Technology. Underlying Data For Indexes of Output Per Hour, Hourly Compensation,
and Unit Labor Costs in Manufacturing, Twelve Industrial Countries, 1950-1990
and Unit and Labor Costs in Korea and Taiwan, 1970-1S>90. Washington, DC:
U.S. Government Printing Office, February 5, 1992.
U.S. Department of Labor, Bureau of Labor Statistics. Employment and Earnings.
Washington, DC: U.S. Government Printing Office, monthly.
U.S. Department of Labor. Monthly Labor Review. Washington, DC: Bureau of Labor
Sutistics, monthly.
U.S. House of Represenutives, Committee on Ways and Means, Subcommittee on
Human Resources. Background Material on Family Income and Benefit Changes.
Washington, DC: U.S. Government Printing Office, 1991.
U.S. Department of Labor, Bureau of Labor Sutistics. Handbook of Labor Sutistics,
Bulletin 2217. Washington, DC: U.S. Government Printing Office, June 1985.
U.S. House of Represenutives. Green Book: Background Material and Data on Programs Within theJurisdiction of the Committee on Ways and Means. Washington,
DC: U.S. Government Printing Office, various years.
U.S. Department of Labor, Bureau of Labor Sutistics. "Employer Costs for Employee
Compensation, March 1989 " USDL:89-295, Washington, DC, 1989a.
U.S. Department of Treasury. "Treasury Report on Income Mobility." Tax Notes. Special Supplement. Vol. 55, No. 9, June 1, 1992.
U.S. Department of Labor, Bureau of Labor Statistics. Handbook of Labor Sutistics,
Bulletin 2340. Washington, DC: U.S. Government Printing Office, August 1989b.
Wilier, Barbara, et al. Ibe Demand and Supply of Child Care in 1990. Washington,
DC: National Association for the Education of Young Children, 1991.
U.S. Department of Labor, Bureau of Labor Sutistics, Office of Productivity and Technology. Hourly Compensation Costs for Production Workers; 40 Manufacturing
Industries; 34 Countries, 1957 and 1978-88. Washington, D.C: U.S. Government
Printing Office, August 1989c.
Wolff, Edward N. "Changing Inequality of Wealth." Presented at American Economic
Association Meetings, January 1992a.
490
Wolff, Edward N. "Trends in Household Wealth During the 1980s." Economic Policy
Institute, preliminary draft, June 1992b.
491
�Worid Bank. World Development Report 1992. Oxford: Oxford University Press,
1992.
EPI Book Series
Veum, Jonathan R. and Philip M. Gleason. "Child<are. Arrangements and Costs."
Monthly Labor Review, Vo\. 114, No. 10, October 1991, pp. 10-17.
Unions and Economic Competitiveness
Lawrence Mishel and Paula B. Voos, editors
0-87332-828-0 (paper)
$16 95
0-87332-812-4 (cloth)
$37.50
The State o f Working America, 1990-1991
Lawrence Mishel and David Frankel
0-87332-812-4 (cloth)
$29 95
0-87332-813-2 (paper)
$14.95
Beyond the TWin Deficits: A Trade Strategy f o r the 1990s
Robert Blecker
1- 56324-090-4 (cloth)
$42.50
1-56324-091-2 (paper)
$17 95
New Policies f o r the Part-Time and Contingent
Workforce
Virginia L. duRivage, editor
1-56324-164-1 (cloth)
$45 00
1-56324-165-x (paper)
$ 17 95
To order Economic Policy Institute books, contact
M.E. Sharpe, Inc.
80 Business Park Drive
Armonk, NY 10504
or call toll-free 1-800-541-6563-
EPI Study Series
America's Industrial Policy Successes: The Forgotten
History of Picking Winners and Losers
Robert Cohen
0-944826-47-4
(Vol. 36) $12.00
Does America Need Cities?: An Urban Investment Strategy for National Prosperity
Joseph Persky, Elliott Sclar,
and Wim Wiewel, with the assistance
of Walter Hook
0-944826-47-4
(Vol. 35) $12.00
492
493
�The Roots o f the Public Sector Fiscal Crisis
Max B. Sawicky
0-944826-43-1
(Vol. 34) J10.00
Short Hours, Short Shrift: Causes and Consequences o f
Rtrt-time Work
Chris Tilly
0-944826-29-6
(Vol. 25) $10.00
Contingent Work: A Chart Book on Part-Time and Temporary Employment
Polly Callaghan and Heidi Hartmann
0-944826-45-8
(Vol. 33) $10.00
Export Controls: Industrial Policy in Reverse
Robert Kuttner
0-944826-39-3
(Vol. 24) $12.00
Industrial Policy i n Developing Countries: Reconsidering the Real Sources of Export-Led Growth
Stephen C. Smith
0-944826-42-3
(Vol. 32) $12.00
One-Third o f A Nation: A New Look at Housing Affordability in America
Michael E. Stone
0-944826-31-8
(Vol. 23) $12.00
The Myth of the Coming Labor Shortage: Jobs, Skills,
and Incomes of America's Workforce 2000
Lawrence Mishel & Ruy A. Teixeira
0-944826-33-4
(Vol. 31) $12.00
Job Displacement and the Rural Worker
Michael Podgursky
0-944826-14-8
(Vol. 22) $10.00
The Low Wage-Challenge to Global Growth: The Labor
Cost-Productivity
Imbalance in Newly
Industrialized
Countries
Walter Russell Mead
0-944826-21-0
(Vol. 30) $12.00
Public Investment and Private Sector Growth: The Economic Benefits of Reducing America's "Third Deficit"
David Alan Aschauer
0-944826-38-5
(Vol. 29) $12.00
Trade Deficits and Labor Unions: Myths and Realities
Thomas Karier
0-944826-36-9
(Vol. 28) $10.00
The Japanese Trade Challenge and the U.S. Response:
Addressing the Structural Causes of the Bilateral Thade
Imbalance
Dominick Salvatore
0-944826-34-2
(Vol. 27) $12.00
Deeper i n Debt: The Changing Financial Conditions o f
U.S. Households
Robert Pollin
0-944826-37-7
(Vol. 26) $12.00
494
Capital Flight and the Latin American Debt Crisis
Manuel Pastor
0-944826-19-9
(Vol. 21) $12.00
Are Americans on a Consumption
Reconsidered
Robert A. Blecker
0-944826-22-9
Binge?: The Evidence
(Vol. 20) $12.00
Flying Blind: The Failure of Airline Deregulation
Paul Stephen Dempsey
0-944826-23-7
(Vol. 19) $12.00
Modernizing M a n u f a c t u r i n g : New Policies to Build
Industrial Extension Services
Philip Shapira
0-944826-24-5
(Vol. 18) $12.00
Beyond Free Trade and Protectionism: The Public Interest i n a U.S. Auto Policy
Daniel Luria
0-944826-08-3
(Vol. 17) $10.00
Shortchanging the Workforce: TheJob Training Partnership Act and the Overselling of Privatized Training
John Donahue
0-944826-18-0
(Vol. 16) $12.00
495
�Keeping Jobs i n Fashion: Alternatives to the Euthanasia
of the U.S. Apparel Industry
Richard Rothstein
0-944826-11-3
(Vol. 15) $12.00
The Consumer Electronics Industry and the Future o f
American Manufacturing: How the U.S. Lost the Lead
and Why We Must Get Back i n the Game
Susan Walsh Sanderson
0-944826-12-1
(Vol. 14) $12.00
Telecommunications Policy, High Definition Television, and U.S. Competitiveness
Robert B. Cohen and Kenneth Donow
0-944826-10-5
(Vol. 13) $12.00
A Progressive Answer to the Fiscal Deficit
Ame Anderson
0-944826-09-1
(Vol. 12) $10.00
Managed Trade and Economic Sovereignty
Robert Kuttner
0-944826-17-2
(Vol. 11) $12.00
The Emperor's New Clothes: Transit Privatization and
Public Policy
Elliott D. Sclar, K. H. Schaeflfer,
and Robert Brandwein
0-944826-15-6
(Vol. 10) $10.00
Strengthening the Progressive Income Tax: The Responsible Answer to America 's Budget Problem
Richard A. Musgrave
0-944826-07-5
(Vol. 9) $10.00
The State o f Working America, 1988-1989
Lawrence Mishel and Jacqueline Simon
0-944826-04-0
(Vol. 6) $12.00
Manufacturing Numbers: How Inaccurate Statistics Conceal U.S. Industrial Decline
Lawrence Mishel
0-944826-03-2
(Vol. 5) $12.00
Prisons for Profit: Public Justice, Private Interests
John Donahue
0-944826-02-4
(Vol. 4) $10.00
The l i m i t s of Privatization
Paul Starr
0-944826-01-6
(Vol. 3) $10.00
Economic Competitiveness: The States Take the Lead
David Osborne
0-944826-00-8
(Vol. 2) $12.00
No Longer Leading:
Performance
Lucy Gorham
0-944826-32-6
A Scorecard
on U.S.
Economic
(Vol. 1) $12.00
Special Offer
A complete set of EPI Studies (Volumes 1-36) is available at
a reduced price of $299 00 (a 25% savings on the full
price). When ordering by phone use the order code
EPIPS-FALL.
Public Interest Publications also offers EPFs Studies
through a standing order plan. Please call
1-800-537-9359 for more detailed information.
Toward a High-Wage, High-Productivity Service Sector
Lester Thurow, with background
paper by Louise Waldstein
0-944826-06-7
(Vol. 8) $12.00
Workforce Policies for the 1990s
Ray Marshall and Paul Osterman
0-944826-05-9
496
(Vol. 7) $12.00
497
�EPI Working Paper Series
No. 101
Better Jobs or Working Longer For Less
Lawrence Mishel
0-944826-25-3
No. 102
Trading Away Jobs: The Effects of the U.S.
Trade Deficit on Employment
Faye Duchin and Glenn-Marie Lange
0-944826-26-1
$10.00
M0.00
Merchandise
No. 103
Family Incomes in the 1980s: New Pressure on Wives,
Husbands, and Young Adults
Stephen Rose and David Fasenfest
0-944826-28-8
$10.00
No. 104
Trade Protectionism and Industrial Revitalization:
American Steel in the 1980s
Robert E. Scott, Thea M. Lee, and Robert
Blecker
$10.00
No. 105
Reconsidering the Benefits and Costs of Trade
tion: The Case of Textiles and Apparel
Robert E. Scott and Thea M. Lee
0-944826-41-5
$10.00
$12.00
Declining American Incomes and Living Standards
Frank Levy, Barry Bluestone,
Lester Thurow, Ralph Whitehead, Jr.,
Jeff Faux
xerox
$12.00
EPI Briefing Paper Series
The Effect of George Bush's NAFTA On American workers: Ladder Up or Ladder Down
Jeff Faux and Thea Lee
(BP31) $5.00
The Myth of the Coming Labor Shortage in Rural Areas
Ruy A. Teixeira and Lawrence Mishel
(BP30) $5.00
Unprepared for Recession: The Erosion of State Unemployment Insurance Coverage Fostered by Public Policy in
the 1980s
Marc Baldwin and Richard McHugh
(BP29) $5 00
Protec-
EPI Seminar Series
Employee Rights in a Changing Economy: The Issue of
Replacement Workers
Julius Getman, William Gould i y
Cynthia Gramm, Ray Marshall, Lawrence
Mishel, Brian Shell, William Spriggs,
Jeff Faux
0-944826-40-7
$12.00
498
Macroeconomic Policy
Barry Bosworth, Paul Davidson,
Robert Eisner, James Galbraith,
Hyman Minsky, Lawrence Summers,
Edward Yardeni, Jeff Faux
0-944826-20-2
The Great American Time Squeeze: Trends in Work and
Leisure, 1969-1989
Laura Leete-Guy and Juliet B. Schor
(BP28) $5.00
A Report Card on the Greenspan Fed
Gary Dymski, Gerald Epstein,
James Galbraith, Robert PoUin
(BP27) $5.00
Investment-Led
Stimulus: A Plan for Short-Term Recovery and Long-Term Economic Growth
Jeff Faux
(BP26) $5.00
New Policies for the Part-Time and Contingent
Workforce
Virginia L. duRivage
(BP25) $5.00
499
�
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Michael Waldman
Description
An account of the resource
<p>Michael Waldman was Assistant to the President and Director of Speechwriting from 1995-1999. His responsibilities were writing and editing nearly 2,000 speeches, which included four State of the Union speeches and two Inaugural Addresses. From 1993 -1995 he served as Special Assistant to the President for Policy Coordination.</p>
<p>The collection generally consists of copies of speeches and speech drafts, talking points, memoranda, background material, correspondence, reports, handwritten notes, articles, clippings, and presidential schedules. A large volume of this collection was for the State of the Union speeches. Many of the speech drafts are heavily annotated with additions or deletions. There are a lot of articles and clippings in this collection.</p>
<p>Due to the size of this collection it has been divided into two segments. Use links below for access to the individual segments:<br /><a href="http://clinton.presidentiallibraries.us/items/browse?advanced%5B0%5D%5Belement_id%5D=43&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=2006-0469-F+Segment+1">Segment One</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?advanced%5B0%5D%5Belement_id%5D=43&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=2006-0469-F+Segment+2">Segment Two</a></p>
Creator
An entity primarily responsible for making the resource
Michael Waldman
Office of Speechwriting
Date
A point or period of time associated with an event in the lifecycle of the resource
1993-1999
Identifier
An unambiguous reference to the resource within a given context
2006-0469-F
Extent
The size or duration of the resource.
Segment One contains 1071 folders in 72 boxes.
Segment Two contains 868 folders in 66 boxes.
Provenance
A statement of any changes in ownership and custody of the resource since its creation that are significant for its authenticity, integrity, and interpretation. The statement may include a description of any changes successive custodians made to the resource.
Clinton Presidential Records: White House Staff and Office Files
Publisher
An entity responsible for making the resource available
William J. Clinton Presidential Library & Museum
Format
The file format, physical medium, or dimensions of the resource
Adobe Acrobat Document
Text
A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text.
Original Format
The type of object, such as painting, sculpture, paper, photo, and additional data
paper
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
MN - 1993 Budget: The State of Working America 1992
Creator
An entity primarily responsible for making the resource
Office of Speechwriting
Michael Waldman
Is Part Of
A related resource in which the described resource is physically or logically included.
Box 28
<a href="http://clinton.presidentiallibraries.us/items/show/36404"> Collection Finding Aid</a>
<a href="https://catalog.archives.gov/id/7763296">National Archives Catalog Description</a>
Identifier
An unambiguous reference to the resource within a given context
2006-0469-F Segment 2
Provenance
A statement of any changes in ownership and custody of the resource since its creation that are significant for its authenticity, integrity, and interpretation. The statement may include a description of any changes successive custodians made to the resource.
White House Staff and Office Files
Publisher
An entity responsible for making the resource available
William J. Clinton Presidential Library & Museum
Format
The file format, physical medium, or dimensions of the resource
Adobe Acrobat Document
Medium
The material or physical carrier of the resource.
Preservation-Reproduction-Reference
Date Created
Date of creation of the resource.
6/3/2015
Source
A related resource from which the described resource is derived
7763296
42-t-7763296-20060469F-Seg2-028-003-2015