-
https://clinton.presidentiallibraries.us/files/original/a8347fe1a1c4e8919bc304c18f6d318c.pdf
4954349619f2e0f059f5944b4547b342
PDF Text
Text
I
EXECUTIVE OFFICE OF T_HE PRESIDEN'T
COUNCIL
O~
ECONOMIC ADVISERS
WASHINGTON, D.C.
FOR IMMEIHATE RELEASE
Thursday> ,June 15,2000
2.050:?,
CONTACT: CHARLES STONE
(202) 395-5086
EDUCATIONAL ATTAINM£NT AND SUCCESS IN TIlE NEW ECONOMY:
AN ANALYSIS OF CHALLEI'<GES FOR IMPROVING HISPANIC STUDENTS'
ACHIEVE:\lENT
A Report by the Council of Econofl,lic Advisers
The President will announce today a new Council of Economic Advisers report on Hispanic
cdt:catio:1. This report focllses on education and the rewards to education arr.ollg L' 5. Hlspamcs.
At1er docJmcnting tbe gaps in educat,ional outcomes for Hispanics rela..ive to non-Hispanic
whiles, the report provides evidence about tbe increasing importance of education to economic
~
success by focusing on Hispanic:s working in the information technology ([T) sector of the new
economy, The report finds that those Hispanics who work In the highly paid, dynamk, arid
rapidly growing IT sector-where job:gro\\1h is much faster than in the etonomy at large-are
typically successful and cam far more than Hispanics who work in olher occupations, However,
Hispanics are' significantly underrcprcscl'.tcd in IT, primarily because they are less likely than
:hcir non~Hisranic peers to have the relatively high levels of education that IT jobs typically
ilurrovc the
;·cqGirc. Policic$ that close the ethnic education gap at alllcvt:!$ can be expected
future prosperity of Hispanic students and insure a greater flow into the labor force of worke:'S
prepared to contribute to the "new economy."
to
Among the Significant findings in the report are:
• The rnlJwnic populafion is a rapld!y growing, increasingly imporlant :·;egment of 11ie us.
POpuiOfion. In 20 years about I iii 6 C,S, residents will be of Hispanic origin, and by the
middl,; of this ccntury-wh~n today's yOU:lg cl:.ilcrcl'. ure middle aged-:his ralia win
incrca!;e to about 1 in 4. The future productivity of the U.S. labor force hinges to 'a
considerable degree on our nation's ability to provide high· quality education for Hispanic
young people, who will playa vital role in the labor market in future decades.
• Dcspil<'; fangible evidence of improvements for some groups, at present there are troubling
gaps In !he educational artainmcnt vf Hispanics. Over recent decades the average education
of U.S.-born Hispanics has i:1crca~d substantially. !.llld the gap between -them Jnd nOI1
Hispanic \.vhitcs i13S declined. r-.;oncthctcss, the high scbool completion nHe umung'young
Hispanic adults is only 63 perccrn-comparcd with about 8S percent for whites and African
Americans. And the fraction of Hispanics who graduate from 4-yc-ar colleges is less than
. balf that of whites, While these differences are partially attributable to the low education
�..
levels (If immigrant Hispanics, U,S,~born Hispanics also have relatively low edu<:utional
aHainment
'
importance of lroproving , " ,
edudtional outcomes for Hispanics is underscored by the increasing value of education' in .
the labor market. For example, two decades' ago. a male Hispanic' college graduate earned 67
percent more than a male Hispanic without a high school diploma, whereas today a male
Hispanic college graduate ea:":1S 146 percent more. These chrulges, in the rewards ·10 " ;,
cdu~ltion are similar to thQse observed for otber men. in the labor market.
• The economic advantages of education are growing, The
-',
",:
,'.
.
,',
'
• Currently, the rdalive(v low jcvcL~ of Hispanic earnings are explained in large measure by.
lower ievels ofeducational attainment. Earnings pr(!.miums associated with higher education
arc much the same for Hispanics as for non-HL\panic whites. Hispanics have much lower
c.l.\rnings tlum non-Hispanic whites; median earnings are 21 percent less for native·born
Hispanics. After accounting for differences in age and gender and in education) the eamings
gap declines to 6 percent for nati"e~born Hispanics (with the remaining "unexplained" gap
due to other factors not di~ectly examined in the study, such as quality of education,
geographic variation, and discrimimitory employment practices).
'. ..,"
• Hispanics are greatly underrepresented in the
high~paying
rr
seclOr, bUI those in iT
occupations are generally successful. While HtsjJanics are 11 percent of U,K workers, they
are only 4 percent of workers in five IT occupations, The Hispanic "digital divide" exists
because the rdatively low educational level of many Hispanics hinders entry into the IT labor
. ma~ket. This under-representation tn IT contributes to the economy-wIde Hispanic pay gap
bc-c<luso;: IT jobs pay considerably ntore than other jobs: non-Hispanic whites earn 62 perce:]1
more b IT t':lan non-Hispanic whlles in oth~r occupations, Dnd Hispanics cam twice as mueh
in 1T as in non-IT occupations. Hispanics who are in IT occupations cam O:1IY margina:ly
'less (3bout 6 to 8 percent) than non-Hispanic whites, after adjusting for differences in geeder,
age, and cducation.
.
2
'
�Educational Attainment and Success in the New Economy: An Analysis
of Challenges for Improving Hispanic Students' Achievement
June 2000
A Report by
The Council of Economic Advisers
�EXECUTIVE SUMMARY
Tbis repon focuses on education and Ihe rewards 10 education .among Hispanics in fhe United
Stale:!;' It documents the gaps in eduCational outcomes for Hispanics relative to non~Hispanjc
whiles The study also provides e\'idence about the increasing importance of education to the
economic su>ccess of Hispanics in the new economy, focusing particularly on a high~paying,
mpidly expanding sector. information technology (IT), Among the significant findings in the
report ;ue:
•
nU! I1ispa1lic population is a rapidly growing, increasingly important segment of the U.S,
population. In 20 yc<trs about 1 in 6 U$ residents will be of Hispanic origin. and by the
middle of this cemury-when 1(03)"S young children are middle aged~his ratio wtn
increase 10 about 1 in 4. The' future productivity of the U.S. labor foree hinges !O il
considerable degree on our nation's ability 10 provide high quality education for Hisp:inic
young people who will play ~ vita,l role in Ihe labor market of the future,
• Despile tangible evidence of improvements for ,,'ome groups, there are troubling lags ill tlte
eductltiulWI auaintr1CIll <if Hupanics. Over recent decades the average education of Hispanics
born in the United States has increased subslantiaUy. and the educational gap between U,S,~
born Hispanics and non-Hispanic whiles has narrowed. Nonetheless, the high school
Gompletion rate among all young Hispanic adults is only 63 percem--comparoo with 88
percent for whiles and African Americans, And the proponion of Hispanics who grJduate
from 4·year colleges is less than half that o(wrotes" While these differences firc partially
attributuble to the !ow education levels of immigrant Hispanics, U,S,·bom Hispanics also
have relatively low educa:ional attainment
• The ecmtomic reward. . oledlicatipn are on the r($(:. The importance of improving educational
outcomes for Hispunics is underscored by the increasing value of education in the tuhar
market. Two dec;)des ago, a male Hispanic college gruduate earned 67 percent more than u
Hispanic maJe with no high school education. an earnings premium that has increased to 146
" percem ,today" Similar increase:s in the earnings premium are observed for al! employed
males.
• Currently, the relatively low levels of lii.spattic eamings ore explained in large measure by,
. lower levels of educatio11al auainmem, Eamillgs premiums associmed with h.igher education
. are much' the same for Hispanic., as for non-Hispanics, Hispanics have much !ower earnings
than non-Hispanic whiles; median hourly earnings are 21 percenf les~ for US,~bom
Hispanics. After accounting for differences in age and gender, U.S.-born Hispanics earned
15 percenl less. and after controlling also fOf education. the gap narrows to 6 percent (with
the rent:lining "unexplained" gap due to other faciors not directly eJlamined in the study, such
as quality of educalion. geogmphic variation, and rli'scriminatory employment practices).
Educational differences also explain mu{;h of the wage gap for foreign-born Hispanics.
• Hispanics ure grcar;y underrepresenled in the high-puying IT sector, but ill gencrallhvse itl
IT occu[Jf1Iirms are successful. While Hispanics are I I percent of employed workers. they are
only 4 percent of workers in 5 IT occupations. This Hispanic "digital divide'· exists hc:causc
the relatively low educational level of m:my Hispanics hinders entry into the IT labor market.
'rhls undcrwreprescnt'ation conlributes 10 the cconomy~wide Hispanic pay gap becam.e these
IT jobs pll)' considcraMy more than olher jobs. Non-Hispanic whiles eam 62 percent more in
IT than nO!1 liispantc whites in (tlher occupations, :lnd Hispanics eam twice liS much in IT as
M
�in non-IT occupations. Hispanics who are in IT occupations earn only marginally less (about
6 to 8 percent) than non-Hispanic whiles after adjusting for differences in age, gender. and
education.
•
The IT ct....;e study iflustratcs tilm the consequences of u.ndcrachievemclIf in education are
fwofold: The students' future prosperity is harmed, and the economy aI large wiU have fewer
'Individuals prepared 10 contribute in "new economy" occupations, Individuals' economic
.succeSS in today's economy incr-casingly depends' on being well educated . .In tum. the strong
performance of tbe American economy is propelled by the ingenuity and ski!ls of our labor
force. exemplified by new economy Seclors like IT. Given the rapid grow!h of the U.S,
Hispanic population. the gap in educational achievement between Hispanics and their peers is
a maUer of critical importance for Hispanic young people and society generally.
"
�I. INTRODUCTION
Hispanics are an extmonHnarily vibrant.· rapidly g.rowing segment of the American
ye3f$, approximately I in 6 U,S, ('esidents will
be of Hispanic origin, and by the middle of the century, noou! one quarter of the population will
be Hispanic Clearly. Hispanic Americans will play an incfCusingly important role m American
life. In panicular, the success of the American economy qver the coming decades depends 10 a
considewhle degree Oil the pn:.'<iuctivity of a labor force in which Hispanic:> will play a
~pulallon. The Census Bureau projects that in 20
progressively larger role.
In this light, enhancing the current state of Hispanic education in the United Slates must
While Hispanic studcm achievement and educational
attainment have shown some progress over {he paSt decades, troubling gaps remain. Hispanics
lag behind non-Hispanics on a variety of educalional measures, A much smaller proportion of the
Hispanic population than the non-Hispanic population complcles high school. Similarly. college
entrance and completion ratcs arc much lower among Hi&panics than among non-Hispanic
whites,
be viewed as a public policy priority.
These educational achievement gaps are especially trouhling in a labor market in which
the economic rewards of education are large and increasing, Evidence suggests that demand has
increased for workers who bring strong proolem-solving ability and technical skills lO the
workpl:lce. Statistics presented below verify that the economic rewards of education are much
the Slime for Hispanics as for non~Hispanics. Those wno fall behind in educational achievement
wi!! aJso lag inlerms of economic success in the new economy.
-
-
To highlight these issues, this report focuses on one rapidly expanding, highly paid sector
of the economy-information technology (IT). An examination of labor market data indicates
that the generally well-educated Hispanics who attain positions in IT occupations earn twice as
much as Hispanics in Other occupations. Further, Hispanics in rr eam only slightly less than non
Hispanic whites with similar demof;raphk: characteristics and educalion, However. there is a
signiftcun~ "digital divide" in IT employm::!nt stemming from a dramatic underrcpresentation of
Hispanics in IT occupations. This undcrreprescntation appcurs in large measure to be the result
of educational differences between Hispanics and non-Hispanics. While Hispanic studeins who
ancnd college are as likely as other students to major in science and e.ngineering, Hispanics are
much less likely than Others fO attend college.
The JT case study iUuslrnles that the consequences of underacbievement in education are
Iwo-fold, Undcmchievcmen[ not only hurtslbe futuf"C prosperity of students themSelves, but also
leduces lhc number of individa,;tls ill the US, labor market prepared to contribute in new
,economy occupations. JndjyjdUllls' economic. success in the modem economy depends On their
hcing well educated. In turn Ihe performance of the American economy is strong in part because
of the ingenuity and skills of our tabor force, especially in new economy sectors like IT. In light
of the mpid growth of the U.S. Hispanic population. the gap ill educational achievement between
Hispanics and their peers is a rrmtler of critical importance for Hispanic young people themselves
and also to society more generally.
�.,,
.,
2. A GRIEF OVERVIEW Or' TRENDS IN HISI'ANle EUUCA nON
Over the past .5 decades there has been a marked increase in the educational attainment of
young Americans Recent data itldici.ue that high school completion rales for young adults (aged
25-29) are approximately 88 pereen! f~r both whites an~ African Americans, with the earlier
pronounced differences bclween the races
Chart 1. High School Compleuon Rates of 25-lo
Hispanics,
disappearing by 1998 (Chart, 1}.1
·29·YeQI.()kls by Race ar>u ElhO!c:1y
....
howcver, have not ,experienced the same gains.
'"
The proponion of those :Jged 25-29 completing
high &Choal remains relalively low--about 63'
percent in 1998-and, though dat<l,are unavailable
for t~is series on Hispanics prior (0 1974. there has
been little growth in high school graduation rates
since that time.
"
_ _ _ _ _ _ _- - '
Similarly. as demonstrated' in Chart 2, the
1(14(1
'&so
'91\0
lS1(1
1900
1st)C
colle!;!; completion rAte for Hispanics have la,gged
behind those of whites and African Americans.
For whites the college completion rJle-:.the
fraction
earning
bnchelor's
degrees-rose
significantly. from 6 percent in 1940 to 28 percent
in 1998. Despite some progress. mcial and ethnic
gaps in college graduation rates remain large,
Currently, only W percent of Hispanic -adults aged
-
..'.'
.
..
~~~~
j. '
L-_~~
10 :
,~~~~":w.=-'~:J
~
,
.
,.
25·29 have graduated from college.
One major reason for the lower levels of.
education for Hispanics relative to non~Hispanics
is that new immigrants are much less educated. 2. If we look only al Hispanics born in the United
Statts ("native~born"}, there has been clear growth in edUCa1i{)nal attainment Census data from
1970. 1980, and 1990 indicate that among working-age adults, nativCM and forcign~boro Hispanics
twi! nativcMbom whiles in average educluional levels (see Table 1 on the nexl page}, However,
the educmion gap belween whites·and native-born Hispanics has been narrowing. In (,:ontrast, the
gap in average c.ducation betwccn whites and immigrant Hispanics has bct:orne wider. Meas!nes
4)
HMO
19!1O
1116(1
HITa
liElO
1{100
of educational achievement for Hispanics such as those given in Charls I and 2 combine the
relatively !ess educ~ted immigrant Hispanic group with [hose born in the United Stales?
. .,
I Charts! and 2 ;irt; based on Census aata, which indude both Hispanic and non-Ilispanic whites among
"while,," and similarly has some Hispanics included in'the African American group, The gaps. between
Hi$p:>uk~ and r.on-Hispanic:-; nre Ihus even larger than those pictured. Prior 10 the mid-19M fUlltual data
are nol availabk (the dOls in lhe chartS todic:ue poinls for which data are available;.
1 As of 1997. 38 pcrcenl of the Hi~panic population were fnreign·born, compared ~ilh S pet~ent of whites
:lI1d 6 ptrctm of African Amt::rica!\~
.1 For addittOfl!J.] analy~j;; scc Julian R Betts and Mag,h)} wfslwrn, "The Educalinnal Anainmeot ,If
Immigrun!s: Trends and Implications," National Bureau of E(:uootntc Resc<lr...:h Work'ng Paper 6757,
October 1998.
�Tab!e l. Average Years of Education for Individuals Aged
16~64
1970
1990
J 1.6
9.5
8.8
Men
Native White
..\Ju;ivc Hispunic
Immlgr.mt Hispanic ,
1980
12.7
JO.9
9.1
12.9
11.4
8.9
Women
)Jative White
11-5
12.4
12.8
Native Hispanic
9,2
10.5
11.3
Immigrant Hispanic
8,4
9.0
9,)
Sour;;c: Betts and Lofstrom (1998). baSt."d on data from the U.S, CCllSUS.
While the educational attainment of U,S.-born Hispanics has been increasing over lime,
U.S.;.born Hispanics continue to havc lower school cOn:Ipietion rates than do non-Hispanic whiles,
The average high school completion rate for 25- to 29-ycar*olds stood at about 80 percent for the
1995-1999 period. compared 'with_ a rate of 93 percent for non-Hispanic whites (Cbart 3)." In
contrast. the' completion rate for foreign-born Hispanics averages below 50 percent. Data on
dropout mtes for those aged 16~24-the fraction of individuals who are neither enrolled in high
schoo! nor have completed high school-show similar pauems, The dropout rate (in Chart 4) is
espe{;ially high for foreign~born Hispanics ("firSt generation immigrants") and for native-born
Hispanic youth who had at least one parent born outside the United States ("'second generation
immignmfs"").s However, even. for Hispanics who were born in the United States aud whose
parents were alsO' born in the United States ("third generation" or higher). the dropoul rate was
approximately twice as high for Hispanics as for non~Hispanic whites-IS.S percent vs, 7.7
.percent, CIMrly the Hispanic' education g3.f! is not solely the consequence of relatively low
educational attainment among immigrant Hispanics. A centrnJ'challenge for improying Hispanic
educational outconlCS, then. lies in i.mproving the educational prospects of both immigrant and
nativewbom Hispanic youth.'
Chart 4. Dropout Rates
Ct,art3. High Schoo! Completion Rates 1995·
~ 999, Aged 25·29
for HispaniC IM/j'ij;"anlS
and Whites 1998, Aged 16·24
11)0
""
re;~.1
..
,:
..
·
·
:
•
.
Non,HI$?II1'\\C
while
·
·
'"
r::
-
."
l~t
\llIM'ao,Ot\
'-"""~IWI
·
·
:
I
i+~pw'\'c, l\all~t
1i~"",. lo.-".~n
\:i<)jr
born
"
4 This completion rate of 93 percent for non-Hispanic White.:> is higher than tr.e 5S percent completion rate
reponed in Chart I whieh is for \\'hlte~ generally (including Hispanic white.,,). This analysis uses Ihe
Omen! Populatit)o Survey"{CPS) fOf 1995 through 1999. Consistent with the definition used by Ihe
Census Bureau, this analysts (as well as all orher original analYSIS woouc!ed for this 'repon) define~
individuals as "native born" jf lhe.)' "'-ere born in the United States Of ;m (jullying area of the L'ni!t!d Swtes.
or were born in.J (meig.n country bu: had 1I1 lellS! onc p:lrem born in Ihe United Stale:'>
5 Phillip KaJfmaR cl aI., "DrnpoUl Rates in !hc Unit(;d Stales: 1998," U.S. Department of EduC.:Jtbn,
~1I1itlnal
Cemcr EJr EduCa!lOn StallSlks, Novemher 1999. Their <:tn;;ly;;\'.
:lta:es ~nd thc District of Columbia 10 those born elsewhere,
3
t:omr~res
thu.;;e born ill the 50
�3. THE PATH TO HJGHER EDUCATIONAL ATfAINMENT
Early education in Ihe,ho~ and at school appears 10' be critical to successfully following
.a path towards higher educational attainment. Evidence suggeSls that the ethnic education gap
can arise from learning differences at very young ages. One report using t999 daw indicates that
among 3- 10 5-year-olds not yet enrolled in kindergarten, Hispanic children were Jess likely than
non"Hlspanic children to regularly engage in such "home literacy" activities as being read to, told
a story, or taught lellers, words, or numbers. These home literacy activities in turn were found
generally to be ll$suciated with higher leveis of "children's emerging literacy," Thus, the
Hispanic children in the study were less likely to recognize all letters. count to 20 or higher, write
their names. Of read or pretend to read storybooks. 6 StatistiCs also indicate that Hispanic 3~ and
4-year..olds are less likely than their wbite cOlmterpans to be enrolled in early chHdhood
education programs. and are underrepresented in Hearl Start enrollment.
on
At older ages. Hispanics
aver.tgc trail non-Hispanic whites in reading and"
mathematics proficiency, (at ages 9. 13, and 11, as measured by the National Assessment of
Educational Progress),7 Not surprisingly then, Hispanics on average also score lower than noo
HJsp'anic whites on college entrance exams. This latter difference can be traced in part to family
background, Hispanic students who take the Scholastic Aptitude Test (SAT) are much less likely
than non~Hispanic whiles to have a pa.rt~nl with a college degree, who might be in a beuer
position to assist a child in the college~prepar:l!ion process.s Hispanic SAT takers are also less
likel;r than their non-Hispanic counterparts to have taken the Preliminary SAT (PSAT).'1,
Careful research shows that much of tOe disparity betweet\ the cducatiorutl attainments of
Hispanics and Jion~Hispanic whiteS stems from iurge difrerences in' family background and
income.)~ One'study found thaI by age 15, 44 percent of tlispanic children had fallen one or two
years behind the exPected grade level--apparently because these students started school at older
ages or were not advanced along with other children in their elementary school classes. Onty
about hulf'as many nOIl~Hispanic whil!! children (23 percent) had falle!!. behind their expected
grad~ levd Statislical <tflulysis indicates thu! 'much of this educational gap can be explained by
differences in family background characteristics, such as bousehold income and parents'
education. Furthennore. fUlure prospeCtS of completing higb scbool and going on to college are
greatly diminished for children who fall behind bya.ge 15. For students who were 2 years behind
the expected grade level, 67 percent of Hispanics and SO percent of non~Hjspanic whiles failed to
(, Sec Christine WinqUIst Nnrd.' et a!., '''Home L:tef:lcy ActivitJe~ and Signs of Children's Emerging
Lilcracy: 1993 .and 1999," U"S. IX:partment of Education, NUliona! Center for Education Slalistics, ::000.
7 From rhe early 19805 to present there has been improvement on these SCores for all a.ge groups of
Hispanic$, a.lthough only slightly for reading.
'
About ooe tbird'ofHispanic SAT takers have a parent with a college educalion, comp:ued with more than
half of non-Hispanic whites. See the National Science Foundation, Women, Miflorities, and PerMlIlS wilh
DmthiluicJ ill SciellCC am! £ngineerillg/ 1998. 1!}()9.
'
1 Statislics also indicate that for HispanJc group a~ wdi as for oiher racial and ethnic gwup~, performaflce
on the American College Test (ACT) is dearly c()rre;ated with family :nc()me - (National Science
K
Foundation. Womell, ,\-linoTific!> and Per,tims with Di.wbilities
If)
Science and Ellgiflcariflg: 1998, 1999.}
These findings are relevant for IlIspanic families 1n particular be{:ause family Incomes are far lower for
Hispanic households than non-Hispanic while huuseholds ..
Iii Tlw rese<lrch reponed in lh:s paragraph is from S!cphen V. Cameron and JlllfleS J. Heckman, ''The
Dynamics of Educ;nional Allainment for Blacks, Hispanics, and Whi~cs." Na1iooal Bureau of Economic
Research working paper 7249. July !9t;9 Tel!:: :J.ulhllfS emph;,l~jze the mit: thaI economic baekgmunJ vlays
on ;:hildrcn'~ educut!'mal achievement.
.
4
�complete high school or earn a GED by age 24, Virtually none of these students (l percent of
Hisranics und 2 percent of non-Hisp:inic whiles} had :iUcnded coJJege by age 24, Thus, a
disparity in educational outcomes appears aroong young children-long before they reach the
ages when they are making decisions about completing high school and continuing on to college,
This evidence indicates chat the ethnic disparilies in high school complelion and college
attendance stem in large measure from a lifetime of disadvantage. The existing disparilies must
be addressed among disadvantaged students well before they reach the ,ages at which they are
maSl likely to drop out of high schooL
.While evidence suggests that children rrom low-income families are less likely (0' be
(by failing to eam 11 high school degree or otherwise f<liling lO acquire skills or
prepare to attend college). researchers also <Irgue that low family income can be an important
direct determinant of collcge altendance. ! I The high cos! of college education can pose a serious'
detenent. As, indicated in Table 2, high-income families are much more likely lhan low-income
families to send their children to college, and they are p'articulatly likely [0 send'them to four-year
colleges,:l The vast majority (90 percent) of sludents whose parents were in the top quanile of
the income distribution were pursuing post~secondary education within 2(J months of high schoo!
graduation, compared with only 60 percent of students whose parents were in the bottom quartile.
And of those lower income s'tudents enrolling in post-secondary education. fewer than half
'enrolled in a 4-year college, compared with almos.t tnree-qu3I1erS of slUdenL<;: from the top incorne
group. Much of these differences in youths' college attendance may arise from the differences in
preparedl1ess for college just discussed, ra:hcr than from financial barriers. However. even after
considering such family backgrot;.nd influences., parent.'il income remains an important
determinant of college attendance.
coHege~read>'
in
Tublc 2. llercentage of Students from Families
Each lfK'OOle Quartile EnrQlling in Post~
Secondary Schools within 20 Months of High School Graduation
Purentallncome Quartile
Total
Vocational.
Technical
Top
90
5
Second
79
70
6
2-Year
College
19
25
25
22
Third
7
10
Bottom
60
Source: Kane (1999), based on data from the high school class of 1992.
4-Year
College
66.
48
38
28
Young people, their familles, and the broader community continue 10 face the ch:J:J1cnge
of filldlng ways to insure that mQre disadvuntugcd young people complete high schoolund have
college atceliS. This must indudt: imprOVing educational prospects fol' tlisadvant.'iged children ;l,t
ever)' leve.l. and insuring that financial barriers do no! prove to be an obstacle at the coUege l~veL
II A!> Dr 19:18 medlar. income fur llispa:lic:; w..:-; $28.330 comjXl.red with $42.439 for non-Hispunic Vr;'lles.
DJ!;} fmm the 1993 Survey of Incnme und Program Parlicip3tion suggest th:.! !he median net w{lrth of non
Hispanic white household!. wa.\ over 10 times that of Hispanic households, The 19% £COI.lOmtc Rt'jJorl (If
Ihe P'CSiO;f/!f provides a detailed overview of ruci.al and ethnic disparity in income and J.ssels.
11 Thumas J. Kane, "Relhin\';ill~ the Way Amefic~n$ Pay for College," The Ali/kell 1lls1!twe Review, Third
Quarter 1999"
�4. TIU: IMPORTANCE OF EmJCATIOi" FOR ECONOMICSUCCI?:SS
On average. higher levels of etiuc!ltion lead to better labor market outcomes-to higher
rates of employment, lower rates of unemployment, and higher wages, And the wage premium
associuted with education has risen over time. In 1999, Hispanic men with a college degree
earned J46 percent more than Hispanic men who bad not completed high schoot 10 contrast. in
1979111is same premium was a mud! smaller G7 percent for college completion. (Over thc same
period the premium for college education for all mer. in the work force rose similarly. from 57
percc:nt to l47 percent.) The increasing premium appears to stem from the increasing value that
the market places on technology-intensive skills. induding computer skills that are used in
servicc sector jobs. Thc wage premium for completing high school relative to dropping out has
also risen over time for Hispanic men. increasing from 33 percent in 1979 to 40 percent in 1999.
Recent research suggests that employers seeking to hire high-school educated individuals are
looking for those with strong cognitive skills (including mastery of basIc reading. malh. and
prOblem-solving skills). This preference for cognitive rather than manual skills might account for
tile rising pay premium for high school education.
The raw comparisons' in wages across
education level described above do no! take
DH"""NCI~~,,"br:m : . , account of any differences in age structure or
.
gender between workers in lhese groups Cbart 5
• ri"f>limc: """,,... 0<>
CNon-r"S~Ne_1o
demonstrates that after controlling for age and
gender, the premium for education is even higher
for U.S,-born Hispanics than for nou*Hispanic
wbiles.):! The earnings premiums. which show the
percent in{:rease in earnings for specific
~-"~'--'200' educationallcveJs relative to those who drop out of
high school after receiving 10 or more years of
education, ilre given separately for non~Hisp(lnic
whites. native-born Hispanics and foreign,bolll Hispanlcs.)4 The geneml relationship between
educational atlaimnen( and labor market success clearly holds for both Hispanics and non~
HispJnics whites.\~
Chart 5. EaminQ$ premium by EducaliOfi
Re!ative10 Comp)eliog Only Grade 10 or 11
,
"'"
13 Specifically, these resl.lhs aTe based on ~egres~ji)o models estimated for each ethnic/nativity group using a
pooled sample of the_ 1995 through :o.hrch :::000 monthly caia from the CPS (with respoodents in 1995·
1998 included only in their Ima survey momDs ttlid n:::!ipondeols io 2000 included only In their fiml"ln survey
rnumh). The depc:1dcm 'IuriOible l); the ;0& of inclIVlduub;' per hour enrnlng.'i, und c)tplllnillOry variables nre
gender. age (included at' indicator variable.\ for )-year age groupings), nnd edLlcaliona~ Calegury (less inJfl
grade .10, an omitted calegory of grade 10 or more bUI no high school diploffi.1. high schooL some college,
SA Of tiS, gruduate cUucation). The analysis focuses on full-time \\'Orkcrs aged 20 or older wbo are not
self-employed, Earnings are conveneil to December 1999 dollars using the monthly CPt-V. Sample si:r.es
are 262.843 non-Hispanic whites and 30.650 Hispanics (just over' half of whom are foreign-born), Median
regresiiion is used, which allows one safely to ignore earnings lop-coding of the CPS data. Coefficients"
reporh:d in Chart 5 are for educational !eveh:·ofhigh school and above. They are transformed to represcn:
percent change~ in hourly earnings.
I~ The "earnings prem,iums" reported in Chart 5 rdlect ltl part the l:uus.u: effect of cciucJtion 00 workcis'
earmngs (e,g., the increased earnings due 10 the higher pmduC11Vit)' 01 wOTker;; in the [abo, markets). In
principle. these numbers may also reflect lhat on :lvernge workers who :lllain higher education may also
have valued unobserved characteristiCS (such a.<;; mherent cognitive ability or perSOnal dri'ie) thai differ
from those with lower levels of education. evidence sugge1its lila! the premiums reported in ordinary
regress-ion analysis are reasorubly good measures of ihe cau..;;al effects of educalio;'l on earnings. {See
6
�'J "
..
,
',-'
,.
".,
Since Hispanics have returns (0 education thaI arc at least as great ilS those of non·
Hispanic whites, the generally lower wages earned by Hispanics arise in large pan from their
"
lower levels of education. Specifically. over [he last half of the 19905. median hourly earnings of :: ·P,'}. i;" ";'
Hispanics were one~third less than those of non-Hispanic whites. Native-born Hispanics earned
21 percent less than white.s, while foreign~bom Hispanics earned 41 percent less (Chart 6). Part
of these wage gaps ure due to differences in gender and age composition; after adjusting fOf these
demogmphic factors, the gap is t5 percenl for native-b<un Hispanics and 39 percent for foreign·
"
born Hispunics, After controlling for available
Churt 6. Oilferences in W8ges oi Hispanic
measures of educational llnainmenl, the gap
ec mes ( urther to 6 percent for natiye~bom
Groups
d I'
SIil~jic~03P
Hispanics and J8 percent for foreign-born
Hispanics. Part of the remaining "unexplained
gaps" may be the consequence of differences in the
C(mtrol~ng lor
quality and type (If education at measured levels
(for example. if non-Hispanic whites typically live
Cor>!t¢lj,tlg fOf
In communities with higher quality public higb
1I~IJCr"opn ,,1$0
schools than Hispanics. or if immigranls educated
40
. so abroad
received
relatively
lower quality
"
education} Additionally, these gaps may reflect
differences In language ability, variations in regional labor markets. and any wage differentials
arising because of discriminatory employment practices. (Among foreign-born Hispanics [he
,differential might also stem in part from the inclusion of illegal immigrants.) The central
(;onciusion. though, is thaI for native·oom and immigrant Hispanics alike eamings disparities are
due in substantial measure;o differences in educational attainment Hi
:.' i
. ":
1.' ' : : .
" \ ,
Re!a'~i!V<~'O~!~!~i~~:;-l
,
,.
5. EDUCATION ASP gARNlNCS: A CASE STUDY OF TilE
rT SJ<::.CTOR
By most accounts the US. economy is: experiencing a technOlogical transforrna.tion that
has changed the nalure of work and placed a premIum on a new sel of skills. While this
transformation has affected many jobs in the economy, there is .a core set ofoccupations at the
forefront of the revolution-occupations in information technology (IT). In the last 10 years,
firms' expenditure on IT surg~d to be-come one of the largesl components of investment And
employers appear increasingly 10 need workers wilh lhe problem-solving skills and technical
expertise.necessary to efficiently utilize these new IT investments,
J);lVid Card, ''The Causal Effect of Education on Earnings," in J/rmdbo()k of Lahor Economics, 'volume 3.
edited by Orley Ashc:nfeher and David Card. North~Holland. 1999,)
;$ pm al! of the analysis ul>ing the CPS il
useful to note thaI some diffelCm:cfi between nadve~bom and
fmclgn.b(,rn His;mnics may slem from tl!!! inclusion in the CPS data of illegal immigrants. many of whom
lire presumably in a poor position to compete for good jobs in the Coiled Slates. For a discussion about the
pre.-.ence of illegal immiStants in the CPS data, see Gi.!illermina Jasso, et al.. 'Ihe New Immigrant Survey
Pitol (NlS·Pj:.O.. . erview i.ltld New Findings Aboul U.S. Legal1mmigmnts ai Admission," Demography.
is
FebnlalY 2000.
.
1(. Tht~ results aboul Ihe importance of education ior explaining the ethnic \Vilge gap arc cnr.sislcnt with
recent research indicating thaI thrce·qv<!.rtcn; or the wage gap bclW0ell McxicJn Americans and nO:"l
Hbpanic while.;. ii' llt\ri~utllbJe to Mcxic;m American5' relative youth, English languilge del1ciencics, and
especially 1heir lower educational at1>1mmem (Sleven J, Trejo, "Why Do Mexican Americans Earn Low
Wage~T
,'.
Joumal of Political Economy, 1997).
7
"
"
,
�This section exami~cs the role of Hispanic Americans in IT. The analysis provJdes a
vivid case study of the genera: problet:1 of l{pi" educational attainment for Hispanic Amcricuns,
and the importance dosing
cducatjm~a! gap.
or
!nc
Although there is no exact definition of an IT worker, there are a number of occupations
that quite clearly fnll into the general domain of IT:' The analysis in this repon. considers a
number of core IT occupations for which data lire available from the Currem Population Survey
(CPS), a large nationally representative sample with information -on workers' weekly earnings,
demographic characteristics, and occupation. These core IT occupations are:
•
•
•
•
•
electricalllnd electronic engineers;
computer systems analysts and scientists:
operations and sys!erns researchers and analysts:
computer programmers; and
computer operators"
Definitions of these occupations arc provided in the Appendix~
IT Occupations: Rapid Growth and
Hi~oh
Wages
The 'combined employment level in these five occupations has grown by almas! &1
percenl since 1983 (Chart 7), with 'particularly strong growth in the: last 5 years. [n contrast, lotal
employment in the overall economy gre\,i hy JUSt 32 percent since 1983. Today ihese IT
occupations comprise approximately 3.4 million
workers (about 2.6 percent of all employed
Chari 7. Workers EmplOyM J'lIT Occupations
workers). Employment projections by the Bureau
',000 ~;;;;;;;';;;-:---i
of Labor Statistics suggest that rapid growth for
00-_ _ __
PC--t>rt>9""""";'
computer-related occupalion.s is expected to
___
1'000
continue well into this century.
D~",
;
:;,'$l!l
Within specific occupations, the most
IT
"
~
notable feature is the strong and steady growth"of
computer systems analysts and scientists. In 1983.
1,000
o
""
,~,
ChartS. Medlan We'i!',dy Earrings ot F"m·T:me
WCllkars
~,~-!
Icch~ology.
~P'«l'-"I
10,. .....uu...
o~'
CQ-'"_",,,,,,
EI('~I''''.' en~,nOfl"
I
,
this occupation had just O\'Cf a quanef of a million
workers, or 14 percem of the total IT workforce.
By 1999. there were 1.5 million workers ill this
occupation, or 45 percent of the lotal. Also notable
is the decline in the number of computer operators,
p~rlJaps stemming from changes in computing
2~
~oo
&00
a;;::
1m""" ..,
'000
11:lO
In addition to experiencing high
cmploymem growth, tbese .occupations arc also
characterized by high wages (Chan. 8). Median
wcekly earnings for four of the fivl1 IT
occupatIons-ail bUI computer opcrJ.ton;......-...easil y
11 POI a fUrlher discussion of these and felO,ted issues set: Carol Ann Meo're~ c1 aI., "The Digital Workforce:
Building infOlech Skills at the Speed of lnoo\'~ti{!n:' U.S. Dep:utmem of Cornrncn:c. Office of Tcchnolog.y
\loliey, June 1999_
�exceeded median weekI)' earnings for all workers in 1999 as well as: in 1989. The median
earnings for the highest-paid IT occupatiOll---electricaJ and elcclronic cllg)r.ee~-was almost
!wice that of all workers ($1.073 vs. $549 iYl 1999).
Hispanics in IT Oceupalions
As in many other higher-paid occup,Hions, Hispanic Americans are severely
underrepresented in IT. There is some indication lhat this llnderrepresentation htls improved
modest;}' over t~e last decade. It i\onethclcss, an examination of dala from the CPS snows that
over the late 1990s (1995 to the most recent available data, Murcb 2000}. Hispanics represented
11 percent of all employed workers. bUI only 4.1 percent of the workforce i~ these five IT
occupatkms .
. The underrepresentation of Hispanics in IT contributes to Ihe ethnic economic gap
because Hispanics and non-Hispanics alike earned far more in IT than in other occupations.
Median hourly earnings for non-Hispanic- whiteS in IT were 62 percent higher than for nOn
Hispanic whites in non-IT occupations. and Hispanics earned twice as much in IT as in other
occupations. Moreover, evidence sugges1s th:'H Hispanics in IT earned only modestly less than
similar non~Hispamc whites: In an analysis of earnings that accounts for differences in education.
age and gender. native-born Hispanics earn about 6 percent less than non~Hlspani.c whites. And
foreign-born Hispanics earn an additional 2 percentage points less than native·born Hispanics (a
difference that is not statistically significant}.19 The "unexplained" pay gap of 6 percent is
comparable to the 6 percent gap that emerges in tne general labor market for native-born
Hispanics when controlling for demographics and education.
, The general conclusions about Hispanics in IT-tbat Hispanics carn only 'slightly less
than non-Hispanics but ~re greatly underrepresemed in IT-are reinforced when a somewhat
broader set of science ~d technology occuputions is examined. w In this expanded sample an
analysis that controls fOT age, gender. and education indicates thilt nalive-bom Hispanics eam
IH Thi.~ conclusion corne" from comparing average representation of Hillplmks in IT (,ccupations in ! 987-89
wilh 1997-99 (using various issues of Empioymel1l wuJ Earnings from Ihe Bureau (l.f Labor Statistics).
There were IOcreases in Hispanic represemation in four of the occupations--computer operators {up J.J
percentage points to 7.1 pe:r~ent}, compmer programmers {up l.2 percentage pvints to 4.4 perccm}.
computer scientists {up 1.0 percenuige points 10 3.6 percent). an<! electrical engineers (up 1.5 pcrccnt!1ge
points to 3.9 percent}. There was .(i decline in Hispanic representation fOf operations researchers (a 0.8
r*er~e~i~ge poini drop to 3A ~cent)..
.
"
, ThIs IS based on a regressIOn model estimated vSln!1, a pooJed sample of. Ihe 1995 through Mar~h 2000
monthly CPS dma. wjlh a dependenc variable. lug: of individu;;ls' per hour earnings, and explanatory
variables. gen<!er. age c;negory. Hispanic and foreign-horn Hispanic indicators. and educaliun;;l category
(less than high schooL high school, some college, :lsweialt!: degree, I3A nr BS, and graduate education).
The analysis focuses on ful!-lime workers aged 20 ()f older who afe not self-employed, Earnings are
convt:nct1 to December )999 dollars using (he monthly CPt-U. The "ample includes HisP3nics ,md oon·
Hispanic whiles, Tbe sample has 8.469 individuals. including 355 Hispanics. Median regreSSlOrl "''as u.sed.
The tOefficien! for the "Hispanic" indicator was .significanOy .differen! from 1.CTO (H,talistk of -2.0), and
"foreign-born Hispanic" wu~ nOt (t-sl(uislic of -0.5).
.
10 This broader set include:> the. ;5 IT occupations and 111~o engineers of all types (uerospllce, metallurgical
and nl1l1erlllls•. mining. pClmlowm, chemical, nudear, ciVIL agricllltuf:ll, industrinl. mechanical. marine and
naval architcct~). mathemntic:d sciemists (inc!\Jdinl,\ l!Ctuaries and SI:lliSlicillnS), niuural scientists
(physici~ts and astronomers, chemists. 'atmospheric and sp:lce sciel1lists, gcolngist£ Imd geodesisls, physku!
sdemtSlS. agricuf1Ural and f(lod scientists. biological and life scientists, (orestry and conservation scientbts.
and fI'ledic;,[ scien.tisls), aoo technicians of all sorts (c!eclficul and electronic:, industrial engineering.
mcch;mical engineering. er.ginecring. biological, chemical. and seier.!.'\! te<:hnicians).
9
�about 4 percent less thtln flon-Hispanic whiles, while forcign·bom Hispanics earn an additional 2
percentage ~oints less than r.a.tive~born Hispanics ((!;umings differences that are not statistically
significant).-! However, a large gap exists in Hispanic employment: Hispanics are 11 percent of
all {;mployed workers but only 4.3 percent of workers in· these science. and technQlogy
occupations.
As detailed in a 1999 Office of Technology Policy report, the Jack of Hispanic workers in
thes,; high-paid and rapidly-growing o;:;cupations stems from disparities in education that exist
umong young people prior to entering. the labor force. z2 In particular, the report indica~s that as
of 1996 Hispanic college students earned bachelor' $ degrees in science and engineering at the
same rate a~ whites (33 percent of students major in science, or engineering). And rates .are
comp~rahle also in engineering specifically (5.3 percent for Hispanics .and 4.9 percent for whites)
and computer science (1.8 percent for Hispanics and 1.7 percent for whiles), The shmtage of
Hispanic!; in new economy jobs is m;>i thc con'{equcncc of Hispanic college students shying away
from technical 1lelds. instead, the key to increasing Hispanic represerHation in science und
engineering appears 10 be identifying: and implementing strategies to increase the overall pool of
Hispanic undergraduates,
6. CONCLL'DING REMARKS
In light of the rapid growth of the U,S. Hispanic population. the gap in educ3tionul
uchievemcnt between Hispanics and their peers is a matter of critic;al policy importance" This
report emphasj~s a number of t;alient facts on Ihis issue. First, there is a large gap between the
education of Hispanics and non-Hispanic. The ethnic education gap stems in pan from the
comparatively low levels of education among immignmt Hispanics. However, while there has
been improvement in lhe educational ticlilevement of native-born Hispanic:;, much ot the g~p I.';
the I~onsequencc of poor educationul outcomes among native-born Hispar::ics, Closing: the
education gap will require improved educational outcomes for immigrolll und non-irmTIigranl
Hispariics alike. Second, this ethnic gap in educmion is a strong contribuling [actor to a
corresponding gap in economic outcomes, Hispanics e<lm substantially less than non-Hispanic
whites. in large measure because of the education gap. As a key example. the education gap
{';QnU ibutes to a serious "digilal divide" in employment in IT occupations and other science and
technology jobs. Hispanics '.vho work in these occupatio!1s generally have high earnings-only
modemteJy less (4 to 8 perCent) than similar J)Qn-Hisp>lnic whites. However, Hispanics are
sevclcly underrepresented in these new economy occupations in part because relatively few
Hispanics achieve the necessary educalional levels. Undcrachievement in education hurts the
future prosperity of the students themselves and also redu\:es the numtv~r of workers in thy labor
force. prepared \0 con~rjbute in new economy jobs.
Resc<tfch described in this reporl suggests Ibm the rcl<ttively poor educ.. tional outcomes
of Hispanic youth often stem from a lifetime ot disadvantage. The solution to the education gap
lies in finding and implementing initiatives that not only target students ut the ages when they afe
making decisions' about completing high school and continUing on IQ college, but that also focus
'I'll.: sample is 7lB Hi;.;panic~ and 16,495 non-Hisp.lOlC while;.;. The coeffiClcnl for "Hi:.panic" if; nO!
significant (t-statistic of - J £) nor :" the coefficient for "foreign-born Hispanic" {t-statistic of -(LO).
it ',he D:gital Work Fore;: Building Inf01ech Skills at the Speed of Innovation," U.S. Department of
Commerce, Office of Technology Policy, June 1999. This report also highlights that women generally arc
underrepresemed in IT OCCupatiOns. In contrast to racial and ethnic milloritie&, worne:-J are under
repre;.cnted becllUsc they are 1c~s likely to choo~ ;,cience and cnginC\!fing field5 when in college.
)j
10
�on children earlier in the educational process. In short, the education gap must be addressed at all
uSc levels. The accompuuying box list): some of (he ways in which the federal government is
seeking to improve educational outcomes for US. youth-programs. that indeed focus OIl all
educutionallevels. J( the ethnic education gap is to be narrowed substantially and r..tpidly, major
efforts will be req\lircd from families and communl;ies. and from the pri\'2te and public seCtors at
all kvels.
r Examplcs of Federal Gu\"crnment Efforts to Improve Educational Opportunity.
Research indicates that the carly preschool years, when human ability and motivation are being
shaped, are critical. to sklll formation. Developmental programs that interve!l(! early in life have
been show~ to be more cosl-.effective than later attempts a1 remediation. One such program is the
federally funded Head StaJ1 program. which. since 1965. has provided comprehensive
developmental services for low~income preschool children as well as social services for their
families, This program has been shown to have large positive effects on test scores and schooling
attainment for Hispanic children s:pecifically, (See Janet Currie and Duncan Thomas. "Does Head
Stitt" Help Hispanic Children?" National Bureau of Economic Research, working- paper 5805,
1996.) The success of Head Start has prompted the Administration to nearly double funding for
the program since 1993 and to seek a $1 billion (19 percent) increase in funding for the program
as part of.he fiscal 2001 budget
As part of their agenda to improve publk education. President Clinton and Vice President Gore
have insisted on high standards for aU students; demanded accountability for results; and
expanded investmem in strategies aimed at raising student achievement. The Clinton-Gore
: education agenda has focused 00 redudng class size in the early grades, expanding after·school
: and summer·school opponuflities. ensuring access to educational technology, improving teacher
quality. and expanding public school choice. (The 201X) Economic Report of the President details
federal initiatives targeting each of thesc agenda items.) As part of the Hispanic Education
Action Plan, the Administ~tion has rcquested funding in the fiscal 2001 budget for programs that
will improve the education of Hispanic students, ineluding Title [ grants to local educational
agencies. bilingual education, migrant education. an adult English literacy initiative. and
programs to help students prepare for and complete college.
Finally, ·the federal government has a number' of programs to aid students in preparing for post·
secondary education <lnd to, hclp make college affordable. GEAR UP partnerships of middle
schOOls. colleges, and community organizations provide low-income students with n)Cntoring,
tutoring. and information on financia: aid, starting no laler (han 7th grade. Another example is- lhe
TRIO programs-educiltional outreach progmms designed to motivale and support students from
low-income families, Other examples include programs that provide financially needy students.
with assistance. most prominently the $4.9 billion Hope Scholarship. $2.4 billion Lifelong
Learning tax credits, and $7,6 billion rovided in the 2000 budget for Pell grants.
II
�,,
:
'.,
....;:.,'
.
ApPENDIX. DESCRIPTIONS OF IT OCCUI'A TIONS STUDIED IN THIS REPORT
Electrical and Electronic Engineers design. develop, test. and supervise the manufacturing of
electrical and electronic equipment. These engineers specialize in different areas such as power
generation, transmission. and distribution; communications; computer electronics; and electrical
equipment manufacturing ~ or a subdivision of these areas. They design new products, write
performance requirements, and develop maintenance schedules. They also test equipment:solve
operating problems, and estimate the time and cost of engineering projects.
Computer Systems Analysts. Engineers. and Scientists is a category which includes a wide range
of computer-related occupations. Systems analysts solve computer problems and enable
computer technology to meet the individual needs of an organization. Computer engineers work
with hardware and software aspects of systems design and development. Computer scientists
include a wide range of computer professionals who design computers and the software that runs
them, develop information technologies, and develop and adapt principles for applying computers
to new uses.
Operatiolls Researchers and AnalysIs conduct research and perform analyses to support
management in increasing the performance of an organization. Managers begin the process by
presenting the symptoms of an operations-related problem to the analyst. who then formany
defines the problem and selects the most appropriate analytical technique to examine it. Upon
completion of the analysis, the analy"st presents management with recommendations based on the
results of the analysis.
'
,
:."
, , ' .: : . i .
.. , '
"", .
'.,.:
",'
,
.
.::l. ::;':,
,
"
......
,,'
,
';'
'
•• !..
,'. '
:: .
"
Computer Programmers write, test, and maintain the detailed instructions, called programs or
software, that computers must follow to perfonn their functions. in many larger organizations,
programmers follow descriptions' that have been prepared by software engineers or systems
analysts. The transition from' mainframe to personal computers has blurred the once rigid
distinction between the programmer and Ihe user. increasingly, adcpt users are taking over many
of the lasks previously. performed by programmers. such as writing simple programs to assess
data or perform calculations.
Computer Opera/or.l· oversee the operation of computer hardware systems to ensure that they are
being used most efficiently. These systems include mainframes, minicomputers, or networks of
personal computers. Computer operators must amici pate problems and· take preventative action;
as well as solve problems that occui during operations. Increased automation and other
technological advances are 'shifting the responsibilities of many computer operators to areas such
as network operations, user SUppOI1, and dalubasc maintenance.
Source: 'Bureau of Labor Statistics, U.S. Department of Labor. Occupatio/wi Outlook Handbook, 2000-01
Edition. 2000.
,
,
12
'
�,la'"
, .. iail::tEl::Ui=e '4,,''''JO%
'9
'9'*""
iff
""d!!! till....'
liii'''.'
t
.IN
Clinton Presidcntial Rccords
Digital Records iVlarlwr
_JAW·'
Tili:; i,; not n presidential record. This is lIsed a, an administrativ\)
lIlarker by the Willial11 J. Clinton Presidential Library Stare
This l11arker ideilli lies the place or a publ icatioll.
Publications have nOI beell scanlled inlhcir entirely for the purpose
or digitizalion. To see the full publication plcase semch online or'
visit the Clinton Presidential Library's Research Room .
•
•
�,:.r.
"
.
'
. l~
.'
.,'
,
I
I
I
1
, I
I
I
REACHING THE UNINSURED:
ALTERNATIVE APPROACHES TO EXPANDING
,
HEALTH INSURANCE ACCESS
I
I
September 2000
A Report by
The Council of Economic Advisers
1
,
':
,
�THE·EcONOMIC 1MI'ACT OF THIRD-GESERATION WIRELESS TECHNOLOGY
. October 2000
. A Report by
The CounciJ of Economic Advisers
�THE ECO;-"'OMIC IMPACT OF TlHko-GeN£RATlON \VIlU~I.ESS TECU;-';OLOC\'
EXECIJTlVE SUMMA1H'
"rhird~generatiol1l1
(3G) wireless tec-hnology provides high~speed mobile access to the Intemet
and.other communications net~orks. This tediltOlogy offers significant benefits to consumers
and telecommunications providers and complementary benefits to the U,S. economy. It is urgent
that the United States follow other advanced countries in making adequate spectrum available for
3G applications. This report documents the likely beru!fitS of 30 technology and explains why
,,: adequate spectrum is needed to provide these services efficiently, The key points are these:
• Telccommunications and the Internet !lre among tbe most import.mt sectors of the New
Economy. Telecommunications represented.3 percent of GDP in J 998, having grown at a i
percent annual rate over the preyious 10 years, Wireless .carriers employ over 150,000 peo
ple In the United States and generate $44 billion in annual revenue, At the end of 1999, the
United States had 86 million wireless subscribers; today that number exceeds 100 mimon,
By year·end 2000 there will be over 600 million'wireless subscribers worldwide. The Inter
net has spawned thousands of companies, as entrepreneurs have raced to provide content,
commerce) and ncw services to consumers and firms, Consumers purchased $5.5 billion of
goods and services over the Internet during tht: second quarter or 2000 alone. Sales over the
Internet between businesses are estimated 10 hit $251 billion in 20001 up from only $43 bil
lion in 1998. The most successful Internet startup companies have created hundreds of bil
lions of dollars of market value.
• Third-generation wireless technology combines two powerfuJ innovations: wireless
communications and the Internet. Today's wireless devices are designed to transmit voice
and brief text messages and cannot handle digital multimedia and other high-bandwidtb
lntemet content 30 devices, by contrast; provide high-speed mohile cOlmections to the
lntemet and other communications networks, giving users full access to the rich content and
"commercial possibilities of the "information superhighway,"
• This new technolugy promises substantial benefits to eonsumcrs~ producers, and the
ecunumy as a "'hole. The annual.consumcr.benefit from today's wireless telephone services
is estimated at $53-$1] 1 billion. The consumer benefits from 30 services will likely be of
this order of magnitUde, , Providers also stand to reap substantial guins, Recently completed
3G spectrum" Iluctions in Europe have raised $150-$600 per capl1a, These at.:.ction revenues
indicHte the expected producer benefits from operating 30 licenses.
• To provide 3G applications most efficiently, adequatC' spectrum must be made available
for commercial use. ln telecommunications, the most important scarce resource is spec
trum. While current U.S. carriers can develop 30 applications uSlng currently allocated
5pectrum~ the allocation of additional spectrum could lower the cost of bringing 30 to U.S.
consumers, Ho\vevcr, parts of the spectrum being considered for 3G applications are already
In use.
�• Delays in introducing 3G produets and s{'rviccs can bt' C()stJ~'. Besides the foregone benc
HIS to 3G consumers and providers, delay may be hnnnfullO U.S, firms seeking to provide
complementary products and services. Early investments are necessary to deyelop a vibrant
U.S. industry for 3G applications. Knowledge spilloverS, which are important in hlgh~tech
industries) tend to be geographically lo,:alized .. Finland. which aHocated its 3G spectrur!1 Ii~
ccnses in March 1999, has already taken the lead in developing an industry 10 provide mobile
applications.
• Government poJicy in allocating spectrum must weigh carefully all benefits and costs.
Consumer benefits, provider profits, and the potential benefits of iadt:stry leadership sbould
be weighed against the possible costs of moving incumbent users to ensure that adequate
. spectrum is made-available for 3G applicatior;s.
2
�TilE ECOi\'Ol\'HC IMPAC1'OfTlIIRD~GEN£RATIOi\' \VIRELESS TECHNOLOG\'
1.
I~TnootJcTIO,.'I:
. The U.S. economy has perfonned rcmark?-bly over the last several years. Productivity
growth hus .')ccelerated from about 1Yl percent per year from 1973 10 1995 H? about 3 percent per
year from 1995 to 1999. This acceleration is heavily related 10 technology, both the investment
in IT hardware and software and the extraordinary productivity of the industries producing the
technology. Between January 1993 and September 2000 the total market value of finns on the
NYSE and ~ASD;\Q increased by 400 percent Knowledge and intangible capital arc incrctl.$~
iogly important: R&D spending has soared. along with the numOc!"s of patents.
.
At the heart of this "New Economy" lie a se:ics of dramatic technological innovations.
Advances in computing, information storage, and data transmission have reduced costs, created
new market!;, and expanded ~xisting markets. These innovations came from a remarkable flour
ishing of entrepreneurship, often concentrated in Silicon Valley and other high-technology corri~
dors in the United States. Firms and other organizations have moved quickly to exploit the OPA
portunities pruvided by these new technologies. Firms are spending billions on enterprise sys
tems, sophistit8ted software and hardware, packages that integrate ordering, procurement, in
vento!}', finance, and human resources, Consumers are offered an increasing array of goods and
services for communication, entertabunent, shopping, education, and other activities. In some
industries, firms arc taking advantage of technological improvements by expanding and consoli
dating their operatio:1s to reduce COStS; in other industries, startup companies are using technol~
ogy to create new products and markets, These changes explain a large portion of rccen1.LS,
productivity gains.
The tek'Communlcations sector has been a primary beneficiary of these technological ad~
vances. Radical improvements in computing power, along witb healthy competition in· the
comm;.tnications sector, have reduced the costs of communications dramatically. As costs have
fallcn, and capabilities have expanded, the wireless telephone and pager markets have expanded
rapidly. Wireless carriers employ mOre than 150,000 people in' the United States and generate
over $44 billion in annual revenue (see Figures 1 and 2). Mobile-phone penetration in the
Unitcd States now exceeds 35 percent. Today, the number of U.s. \\<1.reless subscribers exceeds
100 million. Experts estimate that by ye3r~end 2000, there will be over 600 million wireless sulr
scribers worldwide.! .
.
'The Internet is also transforming the \vays individuals and organizations communicate
and managc information. Nearly 54 percent ofU.8. households have access 10 the Internet and
sur\'c~s indicate that over 50 percent of U.S: businesses will scI] pr9ducts online in tbe year
2000."" Truditional firms and new firms alike are competing to deliver consumers higher~spccd
aCcess to the Jnternet and more sophisticated services for this new medium. Internet sales to
I Cellular Telephone industry Association (W\\::\~:,wo~\'.COt1l.C(lm); Electronic Trend Publications, "The \\iorldwide
WireJess Network." July 2000,
NUJ\ Internet. "How Many Online," Seplember 2000 (www.tl\l::l.ic:survc\.s/hnw
Indicators, Cktober 6, 2000 (www.inlcrnetimlicUlopj colll/facts,HPJ!).
2
3
InP.O\'
odhw); l;'ltemct Economy
�Flgur(l 1, AfHllud RIIvilfluo of U.S. Mobile.
Tllklphone Carriers (1985.1999)
Figure 2. Direct Wireless Employees of U.S.
Providers at Yeal·End
roTI----------------------~1
"
40
lS
30
CI Hoarrnr Revenue
iii Service Ravenue
"
'"
•
n
,
o
Source; CTIA Semi-Annual Wireless ~urvey, 1999.
consurners--s{)~ctilled B2C ecommerce-were $5.5 billion for the second quaner of 2000 alone. 3
Sales over the Internet between businesses (B2B) have increased even more dramatically. B2B
sales are estimated to hit $25 J billion in 2000, up from only $43 billion in 1998,'
The latest advance in mobile communications technology, "third-generation" (3G) wire
less, will be capable of combining the powerful technologies of wireless communications' and the
Jntcrnet. s Today's wireless service, used for·ar.alog and digital cellular phones and page:'s, was
designed 10 transmit voice l)l1d brief text messages. These devices transfer data at relatively slow
speeds) around 9.6 kilobits per second (kbps}6_signifi~a;!tly slower than conventional 56 kbps
dial-up modems. 30 devices, by cont:'ast, w)Jl transmit da1a at speeds between 144 kbps and 2
megabits per second, about as fast as a cable modem Or digital subscriber hne. Increasing the
data~transrer ralc allows mobile phones, hand~hcld computers, and other products to become
multimedia access devices. Further, the international standards that have been developed for 30
allow global roaming ....:1th a single device.?
'
The market for high-speed, or "broadband;" wireless access has tremendous potential.
Broadband appl~cations such as streaming audio and video are -already becoming increasingly
popular on the Internet, as evidenced by the rapid growth of high-speed cable and DSL modems,
U.S, Department of Commerce, Press Release, August 31,2000.
Forrester Research, tnc_, "Resizing Online Business Trade:' November 1998.
~ First-generation (lG) wireless phones, introduced in the United States in 1983, U5C analog techno!o-gy to transmit
voice caUs, &:cond~!;etletatl(l:') (2G) wireless phones use digital tedmoJogy and were introduced into widespread
commercial service in 1996 f(lllowing the FCC's auction of pes spectrum licenses in 1994 and 1995. While both
technologies are clUre-nlly lIsed in the United Stales, since 1999 the number of 2G subscribers has exceeded the
number uf IG subscriber.;. Judy Bcrck, "A Brief History of pes. (Oigital Cellular) Technology Development in the
United States," April 19-98 (W\\w,pcsdat,tcQm/histon'_hlm); Federal Communications Commission, Fifl-h CQmpeti.
lion Reporl, AllgUSt 18,2000..
6 Competitive Intclligence Public:ltioJ's, "3G Mobile: F'ulure Markets," Research Report #103, Chapter 2, May :WOO
(w\\'w. eI<.':ctrol'l ics.cnlrepGrtzh:)(}buJ/c i: ! 03 ,hIm I).
7 Throughout this doc-umcnl we g.enerally use "3G" to refer to the entire class of high~speed wireless communica
tions technologies, Other writers distinguish between 3G and an intermedillJ)' Set of technologies, "2.50," which
offer mobile data services al rates between 56 kbps and 144 kbps, the speeds of conventional modems ane' ISDN
lines, respectively. BOllI 3G and 2,5G wiE nffer subsUlmia) upgrades to the existing rnobi]e aa:a lransmlss;nn capa
bilities, and development of both technologies benefit from allocation of additional spectrum.
l
!
4
�As these and other applications multiply, wireless devices will require 3G capabilities to access
existing Interne1 materials, along with new lnterncr sites optimized for mobile aecess, The
bandwidth provi~ed by 30 facilitates secure mobIle commerce, real-time vidcotonferencing) onw
line gaming, and other, nOl-ret-imagined applications. The 30 technology also gives the user an
i'always-on;' mobile Internet connection.
More iI1.1portantly, tne development of 3G tcchnologies will encourage investment and
innovation in complementary services such as specialized content and billing and payment sys
tems. The Internet has spawned thousands of companies as entrepreneurs hi:wc raced 10 provide
eon tent, products and new services to consumers and to firms. The most successful of these
startup companies have created huncreds of hi:lions of dolhlrs of market \'alue .and havc im
. pactcd the economy dramatically. The combined market capitalization of 15 leading internet ap
plications companies-Yahoo, Verisign, eBay,lnktomi, Commerce One, Amazon, erv1G!,
Infospace, Vignette) Lycos. Inlernet Capital Group, Akamai, Real Networks, Hcal
theonfWebMD, and Cache flow-was '$193 billion on October 2, 2000, An appropriate .lIoe.
lion of commercial spectrum licenses and other policies that favor investment have the potential
to unleash a wave of innovation in 3G applications. The impact of these 'yet-to-be-developed
applications is impossible to predict precisely, but history suggests that they may be profound.
Severa! other counlries 1 including Finiand, Japan, Spain, the·l.l.K., the Netherlands, and
Gemluny, have already allocated nev·,' spectrum specifical1y for high~speed wireless devices and
applications. s It is urgent that the United States follow other ud":'anced countries in mak:in~ ade~
quate spectrum available for 3G applications. As explained below, delay is costly.
,
This report documents the likely benefits or 3G technology and expJains why an adequate
supply of commercial spccU'um licenses is needed to provide these services efficient;y. In gcn~
era.!, benefits oftcchnological innovation accrue to the consumers who use,tne nev.' technology,
the producers who provide it, and other firms that supply complementary goods and services,
Introducing new technologies is also costiy: research and development must be funded;. existing
technolo-gies must be modified or abandoned, and ne.. v capital must be provided. 1n telecommu
nications, the most important scarce resource is spectrum. Commercial spectrum licenses allow
fiW1S to transmit data Over a particular frequency in a particular orca. To provide high-speed and
other wireless applications efficiently, spectrum must be allocated to its highest valued use. This
may require a reallocation of spectrum.
2, Ill:NEFITS FROM NEW TECHNOLOGIES
Tech:lOlogical innovation does nOi'occur in a vacuum; it requires a particular Structure of
incentives and institutions. Finns' demands for new technologies are derived from consumers~
demands for new products and services. Those firms that quickly learn 10 satisfy consumer
net:ds stand to reap substantial gains, particularly in markets where network effects and firsl
mover advantages arc important. There can also be significant spillover benef.ts 10 firms that
provide complementary goods and se,viees.
9 European regulalors have manclHed that newly allocated spectrum be used only (or 30 Icch:!.ology_ C.S" law gen
erally permits carriers to use their allocated spectrum for a variety of technologies,
5
�A. Benefits to Consumers
The potcntial 'Consumer benefits from introducing 3G technology are substantial. While
it 1S impossible 10 predict the precise demand for any future product, one can see the order of
magnitude by studying the introduction of related technologies. For instanc-e, a well-knov.n
study attempts to measure the "consumer surplus H created by the introduction of analog cellular
service (IG),9 Eeonomists define consumer surplus as the difference'between the prkes con
sumers actually pay and the maximum amOunts they would be willing to pay for a particular
good or service. Consumer surplus is thus a measure of the net benefits to consumers created by
a particuhlT market. Using data on price and number of subscr1bcrs in the, top 30 cellular phone
markets between 1989 and 1993, the study estimates that consumer surplus generated -by the in
troduction of the cellular telepho.nc was in the range of $31 billion to S5Q billion per year in COfi¥
stant 1994 dollars,lo In light of sucn potential benefits. delays in the introduction of these serv
ices can be extremely cos-tiy to consumers.
How have the benefits- from the introduction of digital wireless (2G)cornpared with the
benefits of (l0)? Updated cakL:Jations estimate that the combined consumer surplus from IG
and 2G was between $53 and $111 billlon in 1999. t! This new consumer surplus is the product
of severn! factors. First, to the extent that consumers value the quality improvements such as
lmproved clarity provided by digital ~irdcss, their willingness to pay rises and overall demand
increases" Second, because digital wireless uses: spectrum more efficiently, providers can offer
the same service at lower cost. Consumers benefit to the extent that providers pass along these
gains through price reductions. 1l.ird, allocating new spectrum for digital ",;irclcss introduced
new competitors into the market Tbe avcrage number of <;:ompetitors in major metropolitan ar·
eas has incn.:ascd from t'\\'O to more than f{ju~, Increased competition pressures lirl11S to lower
costs; ~nsuring that the cost savings from technological improvement are passed on 10 consum~
crs,
The combined results have been dramatic, as shown in the figures below, Following the
allocation of new spectrum for digital ser\'ices starting in 1994~ total wireless use has risen
sharply, prices have fallen rapidly, and 5ubscribership has increased subS1antially, As 5ho\\1) in
Figure 3. Wta] minutes of use by U.S. \\~reless cusWmers more than tripled,Trom 199510 1999,
During the same period, consumers' fully weighted cos.t pCI' minute dropped by nearly 50 percent
(Figure 4); and average local monthly prices fell from SSI in 1995 $41 in 1999 (Figure 5). In
1999, m~re than half of all mobHe subscribers were using digital technology (Figure 6).
'0
Jerry A- Hausman, "Valuing the Effect of Regulation on New Services in Telecommunications," Brookings Papers
on Economic Activity, MiCTfJPconQmics (1997), pp. 1-37.
10 An earlier study concluded that the total consumer welfare loss from the !(J-year delay in licens:ng tbe cellular
(lO) spectrum a! $86 billion in 1991, or 2 percent ofGDP in 1983 when the Ikensing finally occurred. J. Rohlfs, C.
L. Jacks-on. and T. E, Kelley, "Estimate of the Loss 10 the Uniied States Caused by the FCC's Delay in Licensing
Cellular Telecommunications," Natlonnl Economic Res::arch Assodales Report (J 99!).
H Jerry A- Hausmnn, "Mnblle Telephone," H(lndboDk ofTei~comm/lnica{i(}/J5 Economics, forthcoming.
• 9-
6
�N9IJru 3. Total Minutu O'f Use fO'r u.s.
'"
Wir0leU $uba.criofttS (1991·1999)
>l()
Ftgure 4, Estimated Total &p*nse Per
Minut;) fOf All Calls Mad* by U.S. Wireless
Subscribers (1li91-1'999j
...
,,~------
MB
7.S
61.0
,
51.J
...
,
'"
Figure S. AVerage Lecal Monthly Bill for U.S.
Wireless Subscribers (1988.-1999)
-
-
~
·
~
.
O
15 ~
__________- -____-'
Agora 6. US. Analog and Digital Mobile
Telephone Subscribers .
a Dg~~1
OAnalog
e
eo
'11.1 69.1
~w
~7,1
g
:
PH'~ P~ ~,
Source: CTIA
P" PH P"
Semj~Anol.lal
m, m,
42,8 J\I_~ ~t2
m~!!-
I
'''''
1998
1if99
Wireless Survey, December 1999; fCC, Fi/th Report on Commercial Mobjle
Services, AU~!lSl IS. 2000,
'
.
Moreover'),digital wireless has allowed. new services) sllch as' voice messaging l text mes~
saging, and caller ID, to be integrated into mobile phones, The introduction of voice messaging
sen,kes for basic telephony created an estimated $1.3 hillion in COlis-umer surplus in constant
1994 dolbrs. 12 This t~chnology, which is included in the service provided to many digital wire
less subscribers) may be even more valuable to consumers when combined with the freedom that
mobility provides. J3
Consumers in other countries are already enjoying wireless Internet applications using
2G technology. In Japan, for example, Nippon Telegraph and Telephone's DoCoMo subsidiary
has launched a service called i-mode. Over 10. million lapanese customers have subscribed.
Subscribers use an i~mode phone that can send and receive e~mail as well as access websites op~
timlzed for tiny screens. With a thumb-controlled joystick, subscribers can tap into online newS t
browse through restaUfa:lt guides, buy airline tickets, and trade stocks, Using another technology
called wireless application protocol (WAP), several European finns have tumed phones into
12
Hausman. "Valuing the Effect of Regulation."
.
l~ Hausman, "Telecommunications: Building tbe lnfrastrJclure for Value Creation," in R. Nolan and S. Bradley,
eds., Sense and Respond (Cambridge, Mass.: Harva~d Business School Press, 1998), provides a method to estimate
fln upper and lower bound for consumer surplus for olher goods using limited earn, and he applies this method
internet access,
7
lO
�>
electronic waHets, allowing customers 10 pay for goods nnd services via their mobile phone bill
rather than via credit cards or cash, According 10 recent news reports, Finnish consumers can
make vending machine purchases~ pay rent, phone, or electricity bills, and pay for parking spaces
with their mobile phones.
'
Possible 3G applications me even more in~pressivc, According to the International Tele
communication Union (lTU), 30 devices will be compact enough to fit into a pocket or handbag
and win integrate the functions oCa range of existing devices, The ITe suggests that the 3G de
vice
will function as a phone, a computer, a television, a pager, a videoconferencing center, a
newspaper, a diary and even a credit carel. [It will] support n01 only voice communica
tions, but also real~time video and full-scale multimedia via a screen that can be pulled
out and flexible. It will olso runction as a portabll.) address book .md agenda, contoining
all tlw ,nfonnation about meetings and :.ontacts and able to remind you automal1cally be
fore an i;l)pOrumr arp(liml~\enl or automatically eonnccll0 an audio or videoconference at
a spcdfted lir:ll.). It will automatically search the lntemet for rclevant news and ilif()nnli~
tion on pre-selected subjecls, book your ne).! holiday for you ~n-Jine, and download a
bedtime story for your child, complete with moving pictures. 11 will even be able to pay
(or goods when you shop via wireless electronic funds transfer. in short, the nev.' mobile
handset will become the single, indispensable "life tooL" carried everywhere by every
one, just like a wallet or purse is today,14
B. Benefits to Providers
In a dynamic, rivalrous market such' as the U.s, telecommunications nlarketl firms com
pete aggress:vely to provide neW goods and se:vlces to COnSU!11CTS, First-mover advant.:tges can
be importa,nt in many telecommunications markets so the profits from esta~lishing an carly lead
in these markets can be substantial. Of course. the precise value 10 U.S. operamrs of additional
spectrum for 3G technology is uncertain. A simple analysis of the existing wireless industry in
dicates that, in thc aggregate, U.S. wireless operators earned $23B million of revenue per MHz
under the existing spectrum allocation in 1999. At simifar rates, an additional 150 MHz of spec
trum could bring an additional $35.7 billion of service revenues per year, depending on. what
services are provided. Mobile datu technology may also facilitate new business models for pro
viders, as revenues from 'adVertising, licensing content and applications providers, transaction
processing, and billing may augment or replace tbe traditkmal fee-for-service (subscription)
j
model.
A second, morc precise measure of the order of magnitude' of provider benefits is given
.by the recently completed auctions for 3G spectrum in Europe. Auctions in Germany and the
U.K. raised $46 and $-35 billio!1~ respectively, representing 10lal payments in excess of $500 per
inhabitant in these two countnes. An auction in the Netherlands raised about $2.5 billion, or
$ j 50 per inhabitant Table 1 describes the results of these auctions,
14 lntcrmH:onal TeleJ;:or:'lr:lu~katjon Cnino, ''The Next Generation of Mobile Communications," October 10. 2QO()
{jilin :/.'w'\\'\ \', it\l., in:(~ Jr. t iw;\,lt i ~t3 I'd \!, eniindcx. h\In I).
�Table 1. Comparison of European Spectrum Auctions
412712000
End Date
8/17/2000
712412000
Net
Proceeds
$35,4 billion
(£22.5}
$46.2 billion
(98.6 OM)
S2.5 billion
Net
I?(OCHd~;
$599
$503
$158.4
5'
6"
:5"·
per Caolla
."J<lmbe~ o~
licenses
Fees Paid by Win
ners
Wr,niflg :::irms
{parent company
country 01 origin)
$7.1 hiUion
$7.7
bU,~n
$0.5 b\:!ion
,
•
Vodafone Airlouch
•
BT Cellnet (UK)
Orange (FrartCo Tele
rom)
One20ne (Deutsche
Tefekom)
Te;esystem Interne:
lional Wireless {Tele·
(UK)
• Deutsche Telekom
(Gfmnsny)
•
•
Libe:1el (Nefhe;ftJTlds)
KPN Mc~ile (Nether
• Viag Te!ekcm (British
lands)
Telecom}
• Dutchlone (Nef.her~
• Manflesmann (Voda
lands)
!ono)
.' Telfort (British Tele.
•
• Telefonu I Sonera
com;
(Spain/Finlend)
• 3G Bluoconsorlium
•
• E-Plus (Netherlands)
(TeJe Danmalk /
globe - Ganado}
• MobllCom (Genn8nyJ
Deutsche Telecom /
Sc/gacom)
I France Telecom
Source: UMTS Forum; populatiOn figures (rom SU;tiSfiCa/ Abstrocf of the United Stotes, HUHi. All figur.es con
•
verted 10 current U.S. dollars.
.
"National licenses
~'Each opera:ot purchased 2 sets ol2x5 MHz llCe:tiSes. The result is 6 national licenses.
"'HNationallie-enses
The most a company will be willing 10 pay 10r a spectrum licc!ise is the present value of
~
,
,'16
',
t1 f uturc prOtlts (Iter tax) It expects to rnak "HEllig t l' I'
lC
u
e tram '
l1S tcensc.' 1n a competttive aL.;C~
tion with ~ult~p~e bidders~ the f!ice paid by eac-h witini.ng firm \\:ill come close to,. but wiH no:..
exceed, ttlJS wIllmgness to pay.. USlflg the data from 1able 1, thIS suggests that wmners of the
Gennan uuctions l for example, expect to earn at least $7.7 billion in present value of profits from
operating 3G.licenses in Germany. Annuitizing this p-resent nllue at a f 5 percent ra~e suggests
U The present value of expe<!cd future profits is the Sum of all expected fU1ure profits discounted by the projec!'s
COSt of enpitaL future pl'ofits are all cash Oows from operating !he :service less operating costs and additional in
vcStments required 10 brin~ the service onhne.
.
14 A more refined view also eonsl{lcrs the value of profits fo~goile If the finn does nOi win the licen~e Since 3G is
partly a substitute for existing services, incumbem firms must consider their expected reduNion in profits from IG
. and 2G sen'ice., in the case in which they do operate a 3G license ana- in the case in which they do not operate a 30
license. Fo: example, incumbents without the new technology may lose customers t() emrums that provide the
:lev,.er services:. In theory, this can incrense a firm's willingness to pay for the license (and will depend on its exist
ing, mar;;et share wilh the Cl:rrent technology). By contrast, new entr.mlS consider only their expected fu;ufe proEtf>
from operating using the license.
(1 11 is possible for a firm to ovcrpav jfits expectation and (hat or other bidders is 100 oplimislic
' -
9
.
�that each of the six winning firms expects future
year,li
ancr~tax
profits in excess of $1 billion per
\Vill 3G be as profitable for U.S. companies? While these auction results suggest that
European firms have high expectations for 3G, European and the C.S "vireless markets differ in
important ways. First. three of the bands under consideration for 30 applications in Jhe United
Statcs-the 306-960 MHz, 1710-1850 \1Hz, and 2500-2690 MHz bands-are currently used by
analog cellular phone providers, the Department of Defense, fixed wireless providers, satellite
broadcasters I school systems! and private video teleconferences, The u.K., GermanYr and the
Netherlands. by contrast, did no! face significant incumbency problems v.:hen sp!!ctrum was aue~
lioned for 3G applications,
';\1oreoyer, "",':ireless Internet access may be less popular here than abroad because D5,
prices for wireline Internet access arc already lo\v. The average monthly U.S. price for 30 hours
of Internet access at off~pe:1k times is $22; the average monthly price for
OECD countries IS
$3SY' To th,; extent that \Yirclcss and \vireline Internei access are substitutes these price differ
ences could recuce the potential mnrket for 3G services in the United States. On the other pand,
wireless and \\':ireline Internet access may be complements, a.nd providers could choos~ to pro:
vide combined service. or course, finns in the United States and abroad may change their pric~
ing strategies for wireline Internet access once 3G services become available.
all
j
Finally, finns' expectations about the profitability of 3G may change. Carriers willicam
more about the technology an.d about consumer demand between now and a U.s. auction. If 3G
applications developed within the next 2 years lurn OUt to be highly successful, carriers may de
cide that U.S. licenses arc more vaiuable tban previously thought Firms that win. 3G licenses in
other countries may also vicw U.S. licenses as more valuable if bargaining power with equip
ment suppliers and learning-by~doing decreases anticipated costs. Additionally; as information
about 30 emerges, financiul markets' willing:i.ess to finance license purchases may change.
Early eviden~e suggests that financial markets are not as willing 10 finance European 3G licenses
as firms had anticipated. After bidding an average of $7.7 bi!lion for German UMTS licenses,
companies including Deutsche Telckom have Seen their credir rulings fall. France Telecom's
credit rating was lowered from AA- to A after it supported winning hidders in the U,K, and
Germany. (Of course, 6cse do~ngrades may reflect Other factors as welL) A ratings downgrade
ofthig son typically increases 3 fiml's cost of borrowing signifieantly, Macroeconomic changes,
too, may have an impact OIl finns' cost of borrowing. A significant increase in U.S. interesl
rutes, for example, would likely depress finns' aids.
13 Besides the cost of the license, firms will h;;ve additional capital expenditJres to operate their ne:works in Ger~
mallY, Cash fiow from ot"crntit1!1s mil'll cover tl:c expense :~o;, :l1is as well,
\9 GEeD, D:rec!orate of Scic:1ce T<:tbnology und Industry, H:ntcrnct Access Price Comparison," September:;; J,
2000 (~'r.':.':';:. <ll'cd. (If-gIJs: i/~ti;i:: cn! i).
10
�C. Benefits to U.S. Industry
Besides the direct benefits to consumers and 30 providers, the introduction of this tech
nology could Imleash a wave of secondary innovations in related goods and services. and to fos
ter the development of new "technology corridorst , sueh as Silicon Valley, The spHlover bene
fits to the U.S economy could be significant.
The emergence of the Internet economy, particularly in ;-he United States, shows hm\'
technological innovation can generate large social 'rei'Jnts, Communication.) protocols such as.
Tep/IP and HTML provide a standard platform :or exchanging information between computers.
Opening a new platform stimulates investment not only for the provision of the necessary hard- ,
warC' and sortware, but also for applications and content delivered over that platform, Wide
spread diffusion of these communications standards has given rise 10 entire industries devoted to
providing Internet content and commercial services to consumers and businesses, Startup com~
panics, along with established retailers and information services, have created hundreds of bil
lions of dollars of shareholder wealth through Internet-related activities. Employment in severa!
IT sectors more than doubled between 1993 and 1999. 20 These investments in IT and .comple
mentary $CrVH::es have been major contributors to productivIty Improvements over the latter half
OflhcI990s,21,
,
Importantly, the sectors producing these technological innovations often cluster geo~
graphically. One reason is that knowledge spillovers betwee:l firms, and spillovers between
firms ilnd academic institutions, arc particularly ,significant in high-technology sectors, A recent
study of kn0wledge flows used patent citations to show that these spillovers tend to be geo~
grapnically localized, even af1er controlllng for pre~existing research activhy. 22 In the technol
ogy sector much of the relevant, knowledge is '''tacit!'' rather than explicit! making close social
23
. ties (between entrepreneurs and venture capitalists! for example) all the more important.
In
vestigators have sbown that spatial concentration of innovations was significantly higher in in·
dusuics in which knowledge generation-as measured by industry R&D/sales, the use of skilled
labor, and the importance of academic research-was particularly important 24 In shori) location
R
matters.
u.s, Department or Labor, Bureau of-Labor St::ui!nks, ")/ationa! Ern:p!oyment, Hours, <I:'ld Earnings," series
EEUOOlOOODl and EEUSOmOOJ.
21 Dale Jorgenson and Kev:n Stiroh, "Raising the Speed Limit: US Economic Growth in the Information Age,"
Working PUp<!f, Department o;Economics, Harvard Univc.rsity (May 2000); Stephen Qiiner lltlG Dar,iel Sichel, ''Tl:f
Resurgence orGrowth in the Late 19905: Is Inrormation Techno!osy the Sto~y?" Working Paper, Federal Reserve
Board (Pebruary 2000).
l2 Adam B. Jaffe, Manue! Traj:enberg. and Rebtx:ca He:1aerson, "Geographic Localization of Knowledge Spillovers
as Evideneed by Pmen! Citalions," QuartCl'iy Journal c:/Economics, VeL 108 (l993). PI", 577-98,
.
:'l Gunnar Eliassen" "Business Competence, Orsani41lional Learning, and Economic GroV.1I:: Establishing ,he
Smith~Schumpctc(- Wid:sell COImection," in F. M, Scherer and M. Pcrman, cds.• ElllreprCllCl1l'Shlj), Teclt/w/ogieal
, lnnovalion, una £c()Ilomic Growlh." Studies in 'he Schumpeteriatl Tradition (Ann Arbor:' University of Michigan
Press. 1992); Jacqueline Senker, "Tacit Knowledge and !\1ode!s of Innovation," industrial alla Corporate Change.
Vol. 4 (1995), flP, 425-77.
1( David 8, Audrelsch and M. P. Feldman, "R&D Spillovers and thc Geography of lnn(}valion ane Production,"
AmeriCan Economic RtJView, VoL 86 (l996), pp. 630-40.
:w
II
�Besides this acadeniic work on spillovers,' strong anecdotal evidence suggests ~hat
location can be important in the early-stages of high technology industries. Silicon .Valley is the
most famous example. Moreover, in Fbland-which allocated its 30 spcel!"1l111 i:1 March
1999-a vib~ant cluster of startups devdop:ng commercia! applications for 3G and existing
digital wiz-eless technologies has emerged. Nearly 3,000 companies in Finland are involved in
telecommunicutions and oilier IT industries, including work on wireless technologies and
applications ranging from bil1~payment systems to wireless portals and entertainment. Recently,
major companies such as Hewlett-Packard have chosen to base their ·wireless applications
development programs there, where wireless' penetration is the highest among the OECD
economics. (Sec Appendix 2 for a description of the Finnish wireless cluster.)
Economic clusters such as these playa major role in advanced economies.:5 Firms
within economic clusters are often able to perceive new customer needs more clearly and more
rapidly. According to one important study on economic dusters, "cluster participation also of~
fees advantages in perceiving new lcc1mological; operating, or delivery possibiilties,,,26 Moreo~
vcr. new business fonnation occurs more readily in economic clusters, because the barriers to
entry are lower there than elsewhere. The required assets, skills, inputs, and starr are readily
available at the cluster location and are more'easily assembled there?'
Finally; it should be noted that first-mover advantages are particularly important in mar
kets with network externalities.1!! Many In1c~et markets display strong network eXlcrnalitics,1.9
and wireless Internet markets may be subject to the same effects, In short, to promote a domestic
cluster of internationally competiti\;e wireless finns, it is essential that adequate spectrum be
made a\'ailable for commercial use,
.
3. THE NIttI) fOH ADEQUATE SPECTRUM
, If the benefits to firn1.s from operating 3G are so large, why aren't· U,S. mobile operators
and o\vners of other spectrum alrea~y scrambling to offer this service? No law pre\'ents provi'd
ers from using their currently licensed spectrum for mobile data services such as 3G. In princi
ple, ;some (or aU) of the roughly 200 MHz currently in use for wireless telephone technologies
could be convened by its owners to provide 3G service. However, there ate several reasons why
converting currently used spectrum to this new technology may he costly.
Michael E. Porter, "Location, Compe:ition, and Economic Development: Loeul Clusters in a Global Economy,"
pp. 15-34. ,Porter defmes economic dusters as "geographic con·
centraW:IOS of lnlerconoecled companies, spedalized suppliers, service providers, firms in related industries, and
associated institutions (e.g.• universities, standards agencies, trade associations) in a particular field thal compete but
also coopera!e.'· See also Ponc" The Compr!filtw: Am'an/ug(! o/Nariol?! (New York: The Free Press. 1990).
2/. Porter, "Localicn, Competit:on, and Economic Devclopment"
2:
Economic Dltve/cpmclIf Journal, VoL 14 (2000).
17
Jbid.
Michael L, Kh,tz arid Carl Shzpiro, "SYSlems Competitlor. and Network Effects," Journal olEcanomic !'crspcc~
Vol. S (1994), pp. 93-115,
:'J For example, consider !his explanation from CEO Meg Whitman for eSay's dominance of the online-auction
busi!"less: "We han: the largest r:larketplace by far. Thai dOC5 matter because the sellers want to be where the buyers
are and the buyers W,WI to be where the sellers are." Wall Sweet Joumol, November 22, 1999.
li
lives,
. 12
�First, as ba!1dwidth becomes in~
creasingly scorce, the costs and prices for
current mobHe phone services such as voice
will increase". Second, much of the existing
capital stock would have to be replaced.
Through the cnd of 1999, wireless carriers
had invested over $70 !>illion in capital
cquipmcm (see Figure 7). A carrier that
tried to use its cxisti:1g spectrum fo!' 3G
would fHld some fraction of its current
capito! stock obsolete. Third, the allocation
of new spect~um "licenses could luwer the
cos: of entry into the wireless market, re~
dueing costs by increasing competition.
Figure 1, Cumulative Capita! Investment of
U,S, Wire len Carriers by Year·End f1985
~r---------~1~~"~----------'~1
'"
ro
ItO
~so
:5",
;:.
.!2 30
a
oro
..
,
U.,
JM
:
·
·
·
·
u
Moreover, physical capacity limitations may set in with wireless technology before the
consumer demand for additional bandwidth is exhausted, Although technological improvements
have increased the amount of data that can be transmitted per unit of spectrum, transmitting more
wireless data will. at some point, require allocation of more spectrum for these services. 30
Given these considerations, the provision of additional spectrum for hjg:h~speed applica
tions should be considered a c-Ost reduction for mobile data services, Depending on competitive
conditions) (his cost reduction could lead to substantially lower prices and higher quantities for
consumers.
Cncertuinty itself can also cause firnls to delay investments and hinder the diffusion of
new technologies. 3l In the current environment, U.S. firms face th~ee types of uncertainty:
. regulatory, tl!chnical, qnd business. Whether and when the FCC will allocate new. spectrum li
censes are the key elements of regulatory uncertainty. 1f firms are required to use existing spec
trum to introduce 30 services, technical cnccrtainty will be high, because equipment manufac- .
turers and service providers must Jearn to squeeze both existing and 3G applications into existing
bandwidth. Customer demand for new services is the major source of business uncertainty. Be
cause the demand for mobile data servIces will be dependent on the applications de\'cloped for it
(i.e., the software that will run on the 3G hard\\:are). the timing of customer demand must also be
considered. The decisions made by sof1\\'are developers will depend on their estimates of the
size nfthe uscr base, If developers believe that the user base will be small or slow 10 develop
because of high service' prices or because service providers themselves will delay inveSlments
they will choose 10 develo~ fewer applications. This may, in turn; stall the development and
diffusion of the teciu)ology. 2
Splitting cells requires vcry expensing uddit:onal ne~w{)ri< bfrastructufc, especially in congested ureas ,(Berek, "A
Bricf Ilistory of reS"), Gokimna Sachs (W:re!ess Data. 2000) points oul that in large metro areas, c:ardws arc al
ready hiHir;g, capacity constrabts. This allows 1.hem \0 sustain higher prices.
}; Michael E. Porter and A. M1c~ae! Spence, "The Capacity Expansion Process in a Growing Ongopoly: The Case
of Corn We: Millir.g," in J. McCall, ed., The Economics ofInformation and Uncertainry (Chicago: University of
::hicugo P:es~ 1(82).
~; Kat?; and Shapiro, "Systems Competition and Network Effects."
J(
w
13
�In short, while some mobile data services would probably be forth<::oming without the
provision of additional commercial spectrum licenses: one can assume lhat the amount would be
dramatically lower (al significantly higher prices) without adequate spectrum,
4. COSTS OF J)£LA Y
The process of allocating additional U,S, spectrum for 30 applications is complicated by
the presence of incumbent users. The costs borne by these incumbents must be figured into any
calculation of costs and benefits. ;.Jonet!1cicss, the potential benefits from the allocation of ad~i~
tional spectrum that have been docun:ented in this paper are substantial. Each year of delay in
introducing 30 wi:1 deprive consumers of the surplus that technology will generate. Producers,
of cocrsc, will also lose the ·potential profits rrom providing 30 devices and app~ications, Fi
nally, the U.S. Treasury will lose the interest on delayed auction revenues, which could be s~b
SllintmJ.
Perhaps the most important cost of delay is the forgone benefits from the creation of in~
tCfnatioMlly competitive industry dusters dedicated to 3G products and services. As discussed
above, these clusters are already emerging in Finland and elsewhere. The most important pro~
viders of wifeline Internet services-finns like AOL 1 Amazon.com, Yahoo!, and eBay-ure lOA
cated in the United States, For U.S. fim1s to develop similar leadership in wireless technologies,
it is essential that the supporting institutions be developed as quickly as possible.
5. COSCLUSIO~
3G applications promise subS121ntial benefits. In the United States., however~ parts of 'the
spectrum suitable for 30 applications are already in use, In judging the costs of delaying 30 de~
velopmcnt, it is hnportan: to take into account not on!: the expected revenues from auctioning
spectrum lic~:nscs, but also the expected consumer benefits, These, benefits are likely to be sub~
stantinl-on the orde~ of tens ofbi!lions of dollars per year. f.urther, greater delay in providing
additional spectrum licenses for high~speed applications rcdw.:es the iikelihood that U,S, industry
wilt take ,the lead in developing wireless technology and appiicati.ons..
·14 .
�AI'PE::o.:DlX 1. CHARTS AND T AliLES
Table A~1, Schedule of Allocations of Commercial LIcenses to 3G Spectrum
Country
Date Scheduled
Comoleted
Mon:h
Y~llIr
Comment
Type
Mar
1999
,
Beauty Conteat
•
Mar
Apr
2000
2000
,
Beau:y Conies!
Auclic'1
• 4: nalionallicenses
• 5 national licenses
Beauty Conte~d
•
•
?
Ff~;and
Spain
United Klngdol)1
v
4 national licenses
awa~ded
awarded
Jun
Japa.-"
2000
v
·Jul
2000
y
Auction
Germany
Jwl
Jul
2000
2000
,
Aucton
Auction
France
Sep
2000
Beauty Contest
Sweden
Italy
No,
Nov
2000
2000
Year
end
Yea(
end
Jan
2000
Beauty Con:est
H1brid Auction! Beauty
Contest
Beauty Conlest
The Netherlands
New Zealand
-
South Korea
Singapore
Australia
lama;"!
U,S
Early
Sep...
2000
2001
2001
2002
•
•
•
•
•
•
•
•
$35 bililOt'!
3 licenses
awarded; service to
commence 5101
5 nalional licenses
$2.5 billion
4 national licenses
6 national licenses
$45 billion
4 nationa!licenscs
fixed cost of FFr
32,5 billion ($4 bil- '
lion} per license
4 nationa~ licenses
Hybrid Auction I B!'l8uty
Conlest
Auction
undecided
auction
Source. UMTS Forum, August 18,2000 (www.iJM:s-forum.o:o}.·
Note: In a beauty contest, license winners aYe ge.ierally chosen by government regulators on the basis of tnns:
competil1g business plans. Firms' business plans include desCriptions ot s(:Nice offerings, pricing, geographIC (:O\I$r
age, and timing of new teChnology inlroduction,
15
�Table A-2. Wireless Subscribers as a Fraction of the Population in G7 Countries and Scandinavia
1995
Norway
Sweden
Italy
Japan
5itO"/;
486
U.S,
Germany
Canada
61,8
57.6
52.5
44.9
40.3
19,1
:)4,9
25.5
16.9
17,6
8,2
9,6
2.5
11.8
4.6
8.8
France
1939
65,0%
46.5
356
377
25,6
6,'
UK
Sour~e:
1998
19.9%
22,6
22,6
Finland
31,5
28.8
22,7
OEeD Telecommunications Qatabase, 61' October 2000 1M U.S. figure exceeded 35 percent
Table A-3. Wireless Subscribers, Inter~et Access, and Wireless Internet Access as a Fraction of
the Population in G7 Countries and Scandinavia
rinland
Norway
Sweden
Mob:le phones,
year-end 1999
65,0%
61.8
57.S
'la!y
Japar
UK
France
U.S,
Germany
Canada
Internet access,
\tJireless Inlemet subscribers,
mid·year 2Q9Q.~~~~~-"2000",,, estimates
42,5
49-4
50.3
1R6
20.S
32.7
26.6
22,7
1.4
11.0
49.8
18.0
41.8
52.5
44.9
40.3
34.9
31.5
3,7%
3,.1
3.5
2.1
7,9 ~
1.5
1,3
1.2
NfA
Sources: 1999 OECG -:-elecommunieations Oatabase: Nielsen Netratings, September 7, 2000; Inlema:tiot,al Data:
Corporation-. S:atistica! Abstracts of the United Slates, 1999; FOt~e$ter Research, Inc., "Europe's Mobile Internet
Opens Up: December 1999; Goldman Sachs; (.} Estima:es for Japan are frem press releases claiming that i-mode
has 10 millIon subscribers,
.
16
�ApP£l\UlX 2. CASE STUnY Of FI:'\NISH WIRELESS CLUSTER
Michael Potter offers a framework for analyzing the sources of compctHivc advamage in geographi
cally concentrated industry cluslers such a~ the Caiifon'lia wine industry, the Dutch flower industry, or
Silicon Valley's high4cch indu51ry.33 !-lis framework identifies 4 complementary (aclors that promote
"Iocational competitive advantage," \\'hich is characterized by aoo\'c-avcrage productivity and profi!ahil~
it, among industry players in a particular region. These facfOrs are (t) the context (or firm strategy and
ri\'alry, (2) ruetOr (input) condition, (3) demand conditions, and (4) the existence of related and supporting
industries. Figure A-J diagrams the framcwork. used here to analyze the emerging Pinnish wireless
applications duster, The Finnish wireless industry displays ad""3J1ced characteristics in each of the four
areas,
Figure A-1. Sourc,es of Locational Competitive Advantage
Context for
Firm Strategy
and Rivalry
Factor
(lilput)
Conditions
Factor (input) quantity and
cost
- natural resources
• hum<ln re$OU(ces
~ capitaf resources
• scienoo and technoJ(}g!eal in.
frastrt,lclu '0
• informatiQn infr.lslrl,l~li(e
• p~ysio::at mfrastrudure
• acministr..live infraslruclure
Factor quality
Factor specialization
A local context that en
courages appropriate
forms of investment and
sustained upgrading
Vigorous competitior.
among locally based rivals
Related and
Supporting
Industries
Demand
Conditions
.' Sophisticated and de
mandlng local cus~
tomer{s}
Customers' needs that
anticipate those else
where
Presence of capable,
locally based suppli
ers
Presence of com
petitIVe related in
dustries
Unusual local demand in
specialized segments that
can be served glObally
Reoroduced from Porte'. "location, Competition, and Ecooomic Development"
)) Porter, The Compel/lire Adt'(Ht;uge vjNafi{)lts anc PO:1er, '"u.cmion, COf'!lpetiiion, and Economic Development."
17
�Finland is a eountry of 5.2 mi!lion people situated between Sweden and Russia, with per capita GDP
of$23,780 (1999), or- 69 per~ent of the U.S. per cllpila ODP in purchasing power parity terms.
Finland has had competition in lelccorilTnUliications throughout the 2011> century, The national Post
and Telecommunications ne\'er enjoyed a monopoly, After the U.K" Finland was the first country to de~.
regulate in several areas related to telecommunications: the manufacture of cnd~uscr terminals. basic tele
communications services, and data services. Today there are one hundred telecommunications operators
in Finland, or two operators per IOD,OOO residents.
Two mobile operators, Sonera and Radiolinja, have actively developed and launched new mobile
services and applications. 111is has created a favorable environmcnt for small companies in related areas,
Currently the Finnish telecommunications and IT seelor is populated by approximately 3,000 firms. A
consortium of more than JO smaller operators has recemly been gramed a license tQ build a competing
mobile network.
Bcsides domestic competition from Finnish caf1i~rS and equipment companies, finnish firms face
staunch comp~tjtion rrom neighboring Swedcn.
Factor (Input) Conditions
Finland was the first country to ,allocate licenses for third~generation wireless net\\'orks. These Ii·
censes wen: granted free of charge to Sonera, RadioHnja, 3G (a consonium of local phone t:ompanies and
Swedish "Kc:com), and Tdia {Sweden).3! Some of the fimls awarded 3G licenses plan to provide mainly
network opel1ltions l leasing their assets to other firms that will provide consumer marketing and service.
Tlte puhlic sector In Finland has bee!} supportive of R&D in telecommunieations. Tekes, Finland's
National Technology agency, 1:3s jointly spoll:;ored a progmm, "TLX: Creating a Global Village," with
1he private sector and Finnish research institutes. 'lb:s program has provided PIM 710 million ($120 mil
lion) over lhree years to fund technology dcvt'::opmcl1t, including d and 4th generation wireless systems
and wireless vuJuewaddcd sc:vicc$. Tckes has al~o funded the "Electronics for the Illformation Society
Programme," .:md the Academy of Finland has sp::J:isorcd a research program in ''Tdctronics.,,36 Tckcs
also l\mds R&D program:. conducted in m:<lll H:1d medium sized cnterprises,31
r
A' recent Financial Times Survcy of Finland indicates that private sector funding--outside of the
major equipment providers and carriers-for mobile applications has become widely availablc as welL In
this survey. a partner at venture capital I1rm Eavitcc claims that $2 billion in venture capital funding has
been made available in the last year .and a half.)
~ The major source for the following section is. Finland Ministry of Transport and Communication, "Telecommuni
cations StatistiC's 2000,"
)3 "Finland opposes auctioning, because it considers this a fonn of indirect taxation, 'slowing down the spread {;if new
tedlllologi~." ibid.
'
)6 Tekes Web site (www.r""kesJi).
31 \,ijay Maheshwari, "Survey-Finland 2000: All Wired Up and Going Many Places," Financial Timt!J, July 10,
2000.
31 Ibid.
1&
�Demand Conditions
Finnish \;Qnsumers r.1ay be the world's most sophisticated consumers of mobile technology.
A~
the
end of 1999, mobi;e phone penctra:iofl in Finland reached 65 percent, and by August 2000 penetration
reached 70 percent::<) TIle average hous'chold has 1.3 5 mobi Ie telephone connections (subscriptions). In
early 2000, 20 perc-enl of all Finnish households abandoned their wi:'ed telephones aitogether and opted
only for mobi:c phones. Wireless revenue exceeded wireiine revenue f9r tbe Erst time in 1997.
In ~ 999, more than 650 million short message services (SMSs) were sent tn Finland, SMSs are value
added mobile services that use the narrow-band data transmission capability of GSM. Examples of SMSs
include Instant news, financial informarion, or sports reports, and online chat
.
.
Because of its high mobile penetration rate, Finland has become a test-market for WAP (Wireless
Application Protocol} applica:iolls. Applications developed in Filtland include using phones to l1:ake
vending m<lchine p~rchases and 10 purchase time at parking lots, sending and recei'ving e-mail. and read"
ing public transportatioll timetables. As a result, major international cQrporations and venture capitalis:s
'have identif~ed Finland as the development site for mobile phone app!icatior:s. Hev.."le!l-Packard has
headquartered its WAP development unit in Hclsinki.4u Germany's largest technology company,
Sicrr.cns, has at'lllounced that it will locate a new mobile data unit in Finland. Extensive venture capik11
l11o:tey has been distributed in Finland to create mobi!e Internet productS. 41
Related Ilnd Supporting Industries
Nokia, the world's largest producer of mobile handsets, is headquartered in Finland. Formerly a
widely diversified company, Nokia has focused exclusively on mobile technQlogy since 1992, and has
shed its non-mobile businesses. Nokia has become onC of the world's most competitivc telccomm~njca*
lions equipment suppliers. Its market capi:alization of nearly $160 billion is second iargest among tele..
com t.'quipment producers al~d exceeds that of Lucent, Ericsson, Siemens, Alcatel, aud Motorola, and rep
rese:;ts about 65 percent o:thc Helsinki stock l!1arket's capitalization,42
.
~Q Mbistry ofTrnospml and Commu~ica!ion5 Press Release, August 17, 2evO (\',ww.mi;\tc.fi),
.n Teto Kllitlinen, "Finland· WAP Pioneer," Octobc: 6,2.000 (v"yv,.wHrlano.c{lI1:)•
.n MjheshwlIri,
41
Source
W\V",
"All Wjred Up and Going Many Plz;:es."
hedl!¢:>g" as of October 6, 20DO. Nokia s~ares also trade 011 the New York Stock exchange.
19
�Philanthropy in the American Economy
A Report by
The Council of Economic Advisers
November 25, 2000
�..
Philanthropy: in the Amcricnn Economy
A report by the Council of Economic Ad.visers
Executive Summary
As a
follow~up
to the 1999 White House Conference on Philanthropy) thiS report provides an
economic:, analysis of philanthropic behavior in the United States. It discusses trends in giving
over the past several decades and highlights the economic explanations behind the observed
increase in donations. The report also di.scusscs possible future directions fOf philanthropy and
how even greater giving might be encouraged. Among its main findings arc:
• CharitabJe giving reached
1I
record high in 1999. In 1999 Amencans donated over $190
billion, This represents an increase of 41 percent since 1995. Furthermorc t giving has increased
sharply as a fraction of the Gross Domestic Product. rising from 1.7 percem of GPP in 199:1' to
2.1 percent in 1999.
• Growth in lhe income and wealih of fhe population explains much of this trend, A \'erage net
worth ihr the sample of families we analyze g:cv.' by an estimmcd 28 percent between 1992 2.ild
1998 and average income increased by 15 percent oVcr the Simie period. BOlli income and
wealth are strongly positively related 10 the probability und amount of giving.
• Indi\'idua! giving accounJs for the largest frac1ion of ail charifable giving. In J ')98, 70 percent
of American households made a cbaritab1e contribution and individual giving accounted for 85
percent of 1111 dor.atlo)ls. Althougb the largest fraction of giving is attributable to indiv:duals) the
fastest growing component of philanthropic activity . .vas giving by foundations, whleh rose by 72
percent from 1995 to 1999.
• The elderly are more generous donors than any other age group. Controlling for differences in
income and wealth, those aged 65 and over arc approximately 25 perccn! more likely 10 make a
charitable contribution than younger individ~a:s, and ..vhen they do give) they give $500~$600
morc per year on average, FurthemlOre~ because these calculations do not include chnritable
bequests, the true difference in the total amounlS give!l by the elderly arid the non~elderly is
likely to be even larger.
• Single women are more likely /0 give than single men. When differences in economic
:;csources me accou:lled for, single women arc significantly more likely to make charitable
contributions than are single men. Within the population of unmarried women, womcn who have
never been married are more likely to give than widowed or divorced women.
• African Americans are morf! like(l' to give !hun whiles, After accounting for differcnces in
income, wealth, and education, Arric3:1 Amc:-icans are more likel)' to make charitable
contributions than whites, and on average give approxim8tcly the same amount as white
Americans. Other evidence suggests that minorities are under~used resources with respect to
philanthropic giving.
�• The New Economy has hrolfgh( changes in the methods of givil1g. The Internet has affected
philanthropy as it ha~ so many aspects of American iife. lnternet sites nov,l provide infonnation
ubout charitable organizations~ help match donors with causes, and provide a convenient way to
make contributions. Lc~sons learned from the venture capital sector are also being applied 10
philanthropy. Although stili in their infaacy, these dtvelopmen1s have the potential 10 increase
the umounl of giving anc 10 impro,:e the efficiency with whieh grants ·are used by the recipients.
• The aging of lite baby boomers is good news for philanthropy. Because both older Americans
and those with greater wealth give more) the aging of the baby boomers and the wealth of that
cohon point to the likelihood of a dramatic gro\\1h in giving, pcrhaps increasing by several
hundred percent over :he next couple of decades.
• The Administration's la>: policies will likely a/so lead In increases ill giving. Both cconomic
theory and empirical studies indicate that Americans respond to financial incentives' to give,
Through the tax deductibility of chari1able contributions} both inler vivos gifts and bequests urc
increased in number and size. Recent proposals 10 extend the deductibility of donations 10 those
who do not itemize on their income tax returns, and to simplify other aspects of the tax code l wiU
likely result in further tncreases in giving. Evidence suggests that eliI1!inating the estate tax will
decrease charitable bequests,
'
2
�INTRODUCTION
The tradition of philanthropy in the United States is as strong as ever.
Americans
donated 0 rtcord $190.2 billion in 1999~ Adjusted for ini1ation, this represents a 41 percent
.incrcDsejus: since 1995, and a more: tban'doubling since )980. Americans also gave generO:Jsly
of their tim\!. although volunteerism is not the focus of this study, In J998, citizens gave an
estimated 20 billion hours volunteering for charitable organizations, In fact) ovcr half (56
percent) of adults volunteered that year, the highest percentage in at least a decade, These gifts
of both time and mOlley help s~pport an estimated '1.6 minion nO;lpl'Ofit organizations and
religious congregations in the Lnited States.
.
Philanthropy is at an important crossroads in its development. As our nation becomes
more diverse) \vomen and minorities will likely piny a more prominent :-ole in charitable giving.
Technology is also ch~mging philanthropy. with the JIl:crnct providing ncw ways to give.
Moreover) new str<1tcgles arc being used to increase the efficacy of charitable work l such as path~
breaking partnerships between nonprofits, government: and business, as well as new-concepts in
giving such as those embraced in "venture philanthropy."
_ In October 1999 the President and the First Lady hos~ed the White Ho'Usc's first
Conference on Philamhropy to highlighl these trends in giving, and to emphasize the importance
of our philanthropic tradition and the responsibility of all Americans 10 teach and sustain that
tradition. At the conference the President asked the Council of Economic Advisers to prepare a
report on the role of philanthropy in the economy ilod on woys to encourage Americans to give
niorc.
This report provides that assessment. It begins with an o'.'e!"vicv.' of recent trends in
giving, pointing to the rise in giving relative to the size of the economy in recent years. At the
same :.ini~; however, th'.! report provides a cautionary analysis of individual giving, suggesting
that Americans may not be inherently more generous than the),' have been in the past 1n
p2r:.icular~ the amount that people give at 'any panicular level of income and wealth appears to be
about the same as in the past, but the sheer amount of wealth has increased dramatically) leading
to an increase in charitable donations. The New Economy and the explosion of wealth have also
fostered new metbods of giving that may result in even gn.::u\er contr:hudons'it. the future, To
ensure that the current high levels of giving continue even in limes of slower economic growth. it
is important that we invest now in strategies that encourage future philanthropic behavior.
I
3
�RECENT TRENDS IN GIVIl'iG
As background 'for further analysis, t~is section proviCes a concis~ overview of important
recent trends in charitable giving.
Overview
Philanthropic giving rose by OVer 40 percent between 1995 and 1999, fa $190 billion
This was JaSler rhan aggregate economic growth and giving as a share of GDP has
inc:.reased to levels close UJ those last seen in the i 960s
Totnl philan~hropic givIng has risen strongly in the past 5 years, increasing ove:- 40
percent from $:34.7 billion in 1995 to $190.2 billion in 1999, using inflation-adjusted :999
dollars (Sf:: Chan 1),' This lncrease is equally impressive when compared .to measures of
'eco!lOmic gro\\1h. Since ]995) 'growth in charitable giving hus oulpaced even ou:' s~rong
cco;)omic growth with the aggregate level of giving rising relative to t~c Gross Domestic Produc1
or GDP (51!>! Cha!1 2). This recent increase has reversed a decline in 6e early 1970s that left
chadtab;e giving fluctuating in l:.I range roughly around 1,75 percent ofGDP for two decades. At
a ratio of 2, 1 pe::cent in 1999, giving as a share of GDP has flea:!), re1ume<:i to the highest levds
of the 196(:$,
.
Chart 1. Total GiVing,
Chart 2. G.virg as iii Stwe Of GDP, 1960-99
1960~99
"
w,r---------------------,
----------------------,
"
1.'
,L-.~
__________________
"
"
"
~
"en
'-----c-~--~~------'---'
191)0
1970
1960
""
Sources of Giving
individual giving remains the primary source of American philanlhropy. w!lile g£fh by
foundalions have shown [he las/est growth
I AAFRC Tr!lSt for Philanthropy, Giving USA, 2000, Giving, by jndividual~ aac COrPOf:ll;ons' tn the years 1998 and
1999.is based on projections from Giving USA. Pr~ie(,Led va;ues are furye!iol:s of the levels ofpe:sonal ir.come and
the value of the stock market at (he end of. the year. Conclusions t:ase~ oa these :'lumbers shOJld therefore oc
viewed with caution. For our calculations in this report,. !10minal figures were infla:ior.·adJus:ed using the CPJ-U·
RS series v.'hcre available (19T(, to 1999) nne the C?[wU series for prior years, Giving USA uses the CP1-U series
only, so the inf1ation·adjus~cd figures presented :n tha: report d:!Ter somewr,at from :hose presented here. When
citing other research, we use the CP:~U it, an altt:mp: to be eOI:s.i~tem Wilh the assump~ions of these studies.
4
�including bo:h in"'/' vivos gifts and bequests, individuals accounted for nearly 85 percen"
of all giving in 1999, with the rest coming from foundations and corporations (see Chart 3), Totai
individual giving also accounted for the majority of the $55.5 billion increase in giving m
between 1995 and 1999, rising by $44 billion over this period, or approximately 40 percent
Ho\',:cve:, the fastest~g;o\.vir.g component of giving was giving by foundations, which increased
by 73 pe:"c-c:1l from $11.47 billion in 1995 to $19.8 billion in 1999. 2 Since 1960~ givir.g by
foundations has increased fivefold.
There has also been a large increase i11 giving by corporations) which increased by more
Ihan38 percent between 1995 and 1999 (see Chart 4),
Chart 4. T(eMS in Rea! Giving by Source
Chart 3. Sources of Giving, 1999
,r-------------------~
Irdlvi(lJa!s
$144 bllUCI'!
Corpolilbort
$'.1nl!.vr
Explanations for (he Inercase in Foundation Giving
ScV{:m! economic factors have contributed to the grow1h in fOllndatIon giv:ng. First)
. stocks comprise a large shure of foundations' portfolios, part:cularly among 1oundations with the
greatest assets. As the value of stocks increased sharply during the 19905, the aSSetS of
foundations did as well, Because private foundations must donate 5 percen~ of the value of their
2.ssets cach year to maimai:l their tax-exempt status, as the value of their endowments grew, the
amount given by foundations also increascd, Sl.:cond, the continued low inflation rate helped
maintain the real value of multi-year grants denominated in nominal dolhtrs. Third, a substantial
number of new foundations were established. Over half of all large foundations currently in
eXiSle:1Ce were fOllnded after 1980, and 30 percent were created after 1990. These newly active
foundations were responsible for 20 percent of the gro\\'th in foundation giving between 1997
I
~1~8
2 Of course. all philanthropy can be traced back to individuals. III addition to providing direct gifts to nonprofirs and
charilics individuals provide the seed money for foundations, and employees or olher stakeholc.ers provide the
resources for Cnrporate giving. To the extel:llhm foundations redistribute the funds received from individuals in the
s,amc year in which they afC r~cei\'ed, :hcre wi:! 00 a double-counting of donntions-both the donu:ion 10 the
foundmion and the fOJ:1d#ton's subsequen: donation 10 a charitable cause will be inch.:.de::l in total giving. An
approxlmanon nrcbe extcm aftois ovcr·couming can be derived by assuming that!<lUnd.nlons give away 5 percell!
of any increase in endowment Subtracting 5 percent of the value of individual .g:ifis to founca:ions fro:n thc total
givin!,; by foundations yields a reduction in giving by foundations in 1998 of approximately S1 billion.
5
�ImllHcations
Philanthropic activity as a percentage of GDP increased sharply in the second half of the .
1990s with the majority of this gro\\-1h corning through increases in individual gi"ing. This rapid
increase has allowed the nation to match the generous rates of giving as a fraction of GDP last
observed in the t 960s, If wetare to maintain these rates and avoid the declines experienced in the
197-0s, it i:; critical that we learn more about both indh'idu;;I! giving and the increasing
prominence of foundations.
As subscG.lIcnt sections of this report will
the strung economy is a key dct:::minant
of thC$C tre:ids. Di;ring the 1995-99 period, the econorny g:e\v at a:1 an:1uul rate of 4, 1 percent,
the unemployment rate averaged 4.9 percent, and inflation rCli.liiined low. Wealth also increased
dramatic~lly. Adjusting for inflation, the average net worth of American families increased f!'Om
$224,800 in 1995 to S282500 in 1998. The~e factors helped give Americans greater financial
resources to spend, invest, and don.ate to the causes they support. The rising stock market also
increased the assets of foundations, and because of legal requirements on distributions, the
amOunt foundations have given.
show~
In addition to assessing the extent 10 which the growing economy has contributed to the
increase in philanthropy, we also examine the possibility that the increase in giving may reflect,
in part, a new, more generous altitude to\vards philanthropy. We find little evidence to support
this hypo1hesis, Because this enormous increase in giving waS accomplished with little if any
change in individual generosity, initiatives designed to instill a greater desire to give remain a
potentially fruitful avenue to explore in an effort to increase giving fI.:rther.
THE CHARACTERISTICS OF INlllVmUAL GIVERS
Economic Mcans and Household Giving
. Perso/la! gh'ing is broadly based, with generosiry displayed hy filmiiies at all lerels of
income ana' \<veallh In terms of the tOlal /'C.'ue of churiwble gil'ing, however, a
disproporlionate share cames/rom those with higll'incqmex or considerabfe wealth
Charitable giving in the United States is a tradition practiced by a broad segment of the
population_ In 1998 an estimated 70 percent of households reported making a charitable
contribution. Even among those with incomes under 510)000, almost half (48 percent) made a
donation and the proportion of givers reached nearly 90 percent for families with incomes greater
than $100,000.
The breadth of charitable activity a'1d the variation in the resources of the donors suggests
that charitable giving can best be understood by examining behavior on an individual level. To
do so, we draw on the Federal Reserve Boa:-d's S'Jrvcy of Consumer Finances (SCF) using data
6
�3
from the years 1989, 1992. 1995. and 1998. We find that a small number of Americans are
responsible for much of the givIng, The 10 percent of respondents' making the largest gifts were
responsible lor 74 percent of the 10131 of all philanthropic contributions,
Not surprisingly. families in the SCF with higher incomes are both more likely to give
and give g:eatcr dollar amounts than lower income families. In 1998: 70 pcrcc.nt of families in
the lOp 20 percent of the income distribution milde a contribution of 5500 or more: and among
those who gave, thc average gift was $5,204. Bcea'Jse :he survey limits rcportltd gifts to those
over $500, the fraction of families making a comribution is undercs~imated, This omission is
likely to be particularly severe for the lowest quintitc where gifts arc expected to be smalle.r on
average. We find that just 9 perccJlt of families i!l the lowest 20 perccru of the income
distribution reportedly made contributions, but given the $50.0 cut~off~ the average amount was
substantial a1 $ 1.287.
Ahhl)ugh many studies of charitable glvmg have focused on 'the role of income, a
family's financial ability to make transfers is obviously detennined by other factors as welL In
particular, One would expect giving behavior to be strongly related {Q weulth. Because of data
limitations, this relationship has been ignored in many studies. By u~ing the SCF we can address
this issue. We find thaI when examined across wealth quinlile, very similar patterns to those for
income exist for both the probability and a:11O~Jlt of transfer: 69 percent of tho$c in the highest
wealth quir:,tile made a cO:ltribution, :;md the mean amount was $5,299; 10 percent of those b
lowest category ga\'e~ and gave $1,686 on average.
Chart 5. GfvilJg as a St'lare of Income and Wealth
"
I "'e
.
,
,
• 2
[h:?C"'h
1C.-40%
.().(jC'¥,
nO-flO%
eO.l00%
Chart 5 assesseS givbg relative to a
family's financial meanS. Both families making a
positive contribution and those making no transfer
are included in the chart. Excluding the lurge spike
for the !Ov.'est Income quintile/ the ratio of
contributions to income rises consistently with
income. Families in the 20~40 pt:rccnl range of the
income distribution on ll\'cragc contribute 1J
percent of their income; those in th.e highest quint:le
contribute 2 percent. If familics are instead grouped
based 0:1 their position in the wea;lh distribution,
we surprisingly find the reverSe pattern, the share of net worth contributed to charities actually
falls as net \vorth increases, Among those with positive net worth, families in the 20-40 percent
range of the wealth distribution gave 1 percent of their wealth to charituble organizations, while
the wealthiest gave just 0.4 percent. 4 Much of the wealth. held by the highest quintile is likely to
.
J The Survey nf Consumer Finances is a national survey conducted every 3 yearS by the Fedt:ral Rese:ve Board.
The data set contains information about household income, weallh, and demographic characteristics. It ulso conlains
information about whether the household made comribulkms: totaling $500 or moOf\' in the previolls year, and if so,
:he amoum. The survey has ihc unformnate drawback that this $500 minin:um or. reporting misses contributions
from families who gave less. ,Ho'h'ever, i! does allow IlS 10 examine the e!Tec!s of bo:h income and wealth for a
representative sample of the V.S popula!ion, lin imponact advamage_ ComriollliohS below the $500 limit are
estimmcd to actoullt for $-10 perc{Ont of lOla! giving.
4 11113 C{)mparison is based (ill;) 0:1 t:'{j~e w:th p05i:ive l()v.:ls of incor,!e and positive IcveJ.~ of wealth. A small
fruction of lumi;ies in the saMple heve n:!glltive net wortll. but make a contribution. 'rher,· ure also families with
7
/
�be in stocks. If these gnins have not yet been realized, wealthy individuals may not ye"t have
increased their giving in response to the~r newfound wealth. This hypothesis suggests tha.t the
giving of the: \vealthiest Americans may increase in time.
Chart S. Income, Weallh. and Giving oy Quintile
'oorT----~~----,_~----_rl
"
'w"""
fop :u,(t"'ft:
'E£J
!
•
'"
"
, ,,
,
, ,,
,
C
S"la'1l' oll"loor.e
ShIFt 01 \t(Ea:1'
Dcspitc the smaller share of net worth given to
charity by the wealthiest, the vas! majority of giving
comes from just such f2.mil:es '(Chart 6). The
wealthiest 2Q percent of families in the SCF made
67 percent of all charitable contributions. They held
81 percent of the wealth, and received about 47
percent of the income. In contrast, the bottom 20
percent contributed 3.2 percent of all donations,
held ~O.4 percent of the wealth, and received 7.4
pcrcen~ of the income,'
Sr,are Of Gi"ng
The Relationship between Income and Wealth; and Cbaritable Giving
Many factors affecT/he decision ofhow much to give. When we take into account tJ broad array
ofpersonal characteristics, we find thar changes in bOlh income and wealllI playa significant
role in defining recent trends in giving. Also, we find no evidence 10 support the view thaI
increases in giving have been driven by changes in fastes or preferences, leaving open lhe
possibililY offutllre changes in this direction.
'
The previous scc,1ion (\cparatcly examined the relationships between charitable giving and
the inc-orne an'd wealth of families, To obtain a mori! accum~e understand:ng of the factors that
influer.ce phila:11hropic behavior. it is nec,cssary to take into aCCoullt simultaneously a broad
array of characteristks, Ignoring important determinants of giving will likely lead to incorrect
conclusions about the observed relationships, For example, giving as a percent of GOP has risen
(Chart 2). Based on this observation alone. one might argue that Americans have become more
generous, in that they are giving a greater fraction of their incomes to charities. Ho\,\,cver, this
simple condusion ignores changes in factors other than in<.;ome that may have contributed to this
fiSC, In particular, it ignores any effect of the recem i;'lcresscs in wealth,
Beyond income and wealth, giving may also be hlOucnced by factors such as g~ndcr,
race, age, and education ~ we control for these factors in our analysis,6 Furthermore, because the
dat3 cover a span of years, we can also examine differences in giving over time, exclusive of
changes in these other facto!'s, lf1here are overall ine-rcasc-s in the propensity 10 give and/or,the
amount of gifts that are not explained by the observable- characteristics~ one could begin to
nega:ive income who ;nukc contribulions. To ;!Void the ditlicultie.& associated with thesc calculations. Ch:.rt 5 uses a
restricted .&k The numbers for the lowest quintile should therefore be viewed w:th caution."
! As noted in above, a substnntial fraction of total contributions (5~10 percent) arc not captured in the SCF because
of the $500 minimum, Thus the fraction {)f income and wealth donated is undereslimated_ If these smaller
coniribt:tions come disproportionately (rom lhe lilwest income (or wealt:t) quimiles, lnC'n giving as a share of income
(·.,<calth) wi!! be even lligt:or fractions amung the-se groups re1mive to uw other qt.ljn~ilC$,
(, The regrl.!ssion ar.alysis L1ses l10nlfncar specifications for income, we::l;lh, age, ad edu"Hion, and also il1cludes
con;rols fDr marital status lind number of children.
8
�contemplate the pO$sibility (If an incrt'a~e in gcnc:'osity, However, <1.:1)' obscrn;d change~ in
giving over time could olso be dl~c to changes in facto'rs no! included in our statistical model,
such as changes in govemmem transfers, tax laws, or expectations about future economic
conditions.
One of our key findings is tha: wealth .:;nd income have jndcpcndent e~fects on giving;
cbaritnble donations increase in response to increases in either variable. Because both income \
and wealth have increased throughout the period of analysis~ some of the observed increase in
charitable givjng highlighted in Chart 1 is attributable to a "income effect" and some to a "wealth
effect:'
At the san:e ti!11C. however, we do not find any cvide:lce of an iacrcased preference for
giving over time_ Rather, we find subtle indications thm preferences for giving may have actually
decreased slightly between 1989 and 1995, before rebounding in 1998. Thus there remains the
potential [0 build on recent increases in giving by impro\'i~g attiludes about giVIng,
Beyond the roles of income and wealth, our :f1ndings shed light on the variation in giving
by age, education, race a!1d ethukity, llnd sex. We- find that the relatiollship~ obscn'<.:d in our
multivariate analysis differ substantially from the conclusions drawn based on silnple cross~
tabulations of the data. We now discuss these results,
Older families (in which the family head is aged 65 or ove)~ (J{ every level of income arc
generous givers. Holding constant differences in jinwtcia{ resources, they'are more lfla:ly fo
mala: (1 contribution Ihan younger families, and when doing so, give a larger amounl.
The aging of tbe American population has been highlighted in discussions abo~!t the
future of Sodal Security and the impending difficulty of supporting a large population of retired
individuals. For charitable organizations) however, this trend may provide substantial benefit.
Preliminary'results from the forthcoming Giving and Volunteering 1999 demonstrate a
substantiall:; [ower propensity to give among. those under age 35, but no clear, age trend
thereafter; the probability of making a contribution peaks nt 7S percent f1!f the 65-74-ye'H old
age group. but is nearly identical to the 77 percent for those ages 45-54. I Similar pal1cms arc
found with respect to the amount. While the probability of giving and rhe ievels given are
similar across ages, there is a monotonic increase with age in giving as a fraction of income. This
, figure rises from 1.5 pe:cent for those ages 25-34 to 2.5 percent for those 65-74. Those aged 75
and over contribute an astOunding 4,6 percent of their inco~le.
m
A stronger, relationship exists between age and tile probability of a charitable
contribution in the SCF data. The percent of households making a donation peaks a1 4Q percent
for those ages 45~54 and falls slightlY 10 34 percent for those ages 75 and over. lbe difference
between the age patterns in the Giving and Volunleering data and the SCF data in part refleClS
7
The Independent Sedor, Giving ami Volunteering 1999,
9
�the $500 lower llmit on giving in the SCF, but also may :ef;ect changes over time, Data from
Chart 7, Arr,o;.ml Given by Head Of HOJse,"'old Age
Giving and l-"'olll.nieering J996 show a similar
sccc r - pattern to that in the SCF data. ~ Because our SCF
data arc compiled from the 1989, 1992) 1995, and
,1998 surveys, the patter:1S :n givi:1£, for 1996
reported in Giving and Vulu11leering may more
c;osely match the SCF data. Among those v;ho
make a charitable donatio11 in the SCF cata, the
amount increases with age, fro:n $2,536 for those
ages 35~44 to' $4,423 for donors ages 75 and over
(see Chart 7),
One would not want to conclude solely from
these descriptive results that the elderly an;: more gCI:erous than the non-elderly; there are. also
large differences in income al:d- wealth by age thal need 10 be controlled for. Few elderly are
employed, and ~hey therefore have lower incomes than lhe working age popUlation, while at 6e
same t:me they may have more weali];" Conversely, ymmgcr people in the early stages of their
careers may have relalively high incomes but little wealth, Ho\vever, eve!) conlrolling for these
and other variables, we find thaT the elderly do appear 10 be more generous; bolh the probability
and the amollnt of giv:ng bcreasc$ monotonically with age. Those age 65~74 arc 24 percent
more likely to make a gift than those aged 45-54 and condhional on making a gift, contribute
$46D more on average. Those aged 75 or older are even more likeiy to give (28 percent more
likely, than those aged 45-54) and give $620 l!1ore on average. FurthcmlOre t because reported
gifts in the survey do not include bequests, the total amount of giving in the older age brackets
will be even higher than reponed.
individuals in the age grOU}) 25~34 are significantly less likely to make donations than
those ages 45~54. but those who do give sirnibr amounts, On the one hand, one would imagine
...hat a 30~y(_ar~(l!d with the same levels of income and wealth as a 50 year old should be bener off
in lifetime terms, having many more years over which to cx.pcrier.cc growth in earnings and
wealth acc1.lmulation. On {he other hand, this group is also facing many demands on their
financial n!$ources) including those' of supporting children. Their relative ability to make
donmions i:; therefore unclear.
While the elderly do appea: to be more genc!"ou$ than other age groups, it is impossible lo
infer from these statistics whether U1C current young will be equally generollS when they reach
their retirement years, or \\'hether the current cohort of ,elderly hns ahvuys been excepliomlly
willing to give. However, the relatively low rate of giving among younger families does suggest
that recent initiatives to simulate giving among the young may be ~ell directed.
~ The IndcJ~cndent Sector, GivlI1g and Volunrecrhlg ,u/96.
]0
�Educntiou
Giving is s;gnijicanfl)' Mgher among more educated households in terms of Ihe percentage of
households making cnnlribution-;, the dollar amount of fhose conrribulions, and Ihe percentage
0/h(m~ehold income donated.
Bast!d on preliminary figures from Giving and Vo!unreering 1999, 61 percent of
respondents with a high school degree or less reported making a dona,jon in 1998, up from 59
percent in 1995. Among donors, the average household in this group gave $584 or 1.7 percent of
household inCOll'.c. A higher share (72 percent) (If households with some college educationr but
not a degree, rcported making a donation~ with an average level of giving among donqrs of $963,
or 1.9 percent of household income. HO\\'ever~ the fraction, of thls group making a donation was
lower th.m in J995. Giving was even higher for those with a college "degree: 82 percent of
hous.choJds gave nn average of $1,748, or 2.5 percent of hOl.:sehold income-nearly 50 percent
higher {han for those with only a high school degree. An identical trend of higher giving among
those with more cducntlon is found in the SCF data,
Of course~' more educated households also tend to be higher income, hig.her wcahh
households. So much of the relationship between giving ilnd education may be picking up the
cflccts of financial statu;; rather than those of education, Whl.Ctl holding income, wealth; nnd other
factors constant l we continue to find a strong positive relationship between education and giving.
If the head of the household has a collegc degree, then the family is both more likely to give and
tends Lo give greater arno;.Jnts, -This result indicates that there is a separate eiTcct of cducut:on in
addition to the effects of income and wealth. l\vo hypOtheses arc consistent vvith this finding.
First, education itself may instill a greater "preference l ' for giving as one learns more about the
world. Alternatively, coilege graduates may k.ve Ilfctimc camings prospects that arc not fully
represemed by their current income and wealth. Holding income, wealth, and age constant, a
co!!ege. grnduale may expeet greater future i,ncome than a high school grdduate and may
therefore t;.; more cot:nfonable giving <l lorgcr amount. Unfortunately, our data do not allow for II
test of either explanation.
Gender
"yomf/J1 dljfer suhsfantially by marital statu.\', ./\lever-married lHm1en give
more ofien and greater amount'.' than single males, while widovred or divorced women gNe less,
Paflerns
0/ gf\'ingfor
Examining differences in giving by gender is complicated hy the fact that surveys report
a single response for a household. This is true in both the survey used to collect data for the
Giving and Vo/un/cering reports and in the SCF, In these cases it is impossible to distinguish
between the ph:tanthropic behavior of husbands ;lI1d wives. To identify the relationship between
gender and giving we thererore categorize respondents as married couples, single maies t or
single females, and compare the actions of single males and single felna!es. Our results show that
single females are equally likely to make a transfer as married couples, but Single males nre
significantly less likely to do so. \\lith rcspecllO the amount given, single females give less than
single males; but the difference is not significant.
II
�Much ilttcntiol1 bas recently foc'Jsed on women and philanthi·opy (see box). We therefore
look morc closely at the pa:tcrns of giving for women. \Ve subdivide the category of single
\vomen into two groups, those who are widowed or divorced. and those who have never been
married. The results show thm never~marrjed women are aCHlnlly somewhat mOre likely to give
than couples although the estimated difference is not significantly different from zero, while
widows are lcss likely to give. Similar patterns arc cvident for (he amounts: single women give
slightly more than widows, but the difference is not signific~iln. While these differenccs cont:ol
for the number of chi!dre:1, there 1113Y be other factors in family relationships that affect giving
for \\'hicb we do not control. Depending on the underlying cause of the observed diJTcrcncc;
outreach programs designed to encourage gh'ing by women might be most effective if targeted at
widowed women.
Women & Philanthropy
American women have the potential to become leaders in philanthropic giving, Currcntly
they con~ro! morc than 51 percent of the persona! wealth in the United States and own a third of
aU privately held businesses. Furthermore) becal,:,se women typically outlive their husbands, they
are projected to inherit many trillions of collars in the comiag dec;1dc, This untapped potential
represents an important opportunity for the philanthropic community,
\\/o01C11'$ giving pattems have traditionally mirrored those of their busbands; however, a
shift in behavior has begun, with women becoming more involved in giving, Younger women )n
particular are more likeiy to make thc:r own choicl;;':s with respect to charitable gh'jng and there
is reason to believe that as older women become more confident in financial ma:mgcnicnt skills,
they too will begin to act independently.
j
Organizations are beginning t() reach out to encourage philanthropy among women l with
many colleges and universities leading the way, Institutions are expanding efforts 10 n01 only
cr.courage women's giving, but also to study, understand, and support women's philmllbropy, [n
1992 the University of Caiiromia, Los Angeles (llCLA) established "Women & Philanthropy at
eCLAt', As pan of its mission it cncoarages women to give, helps women tailor their giving to
areas that suit their own intcrests~ and helps women develop the skills necessary to as-sume
leadership positions on campus, tv1cmbers of the Women & Philanthropy program arc provided
opportunitits to meet with UCLA researchers and to attend special campus events, The program
has been D success, raising over $20 miflion in the ]999 fLseal year alone. Many other
institutions including Oklahoma State UniversitYl the University of Mis-souri at Kansas City~ and
Purdue Univc:-sity, lmve similar programs.
j
I
Anecdotal evidence suggests that women arc motivated to give by different factors than
men, in particular, women appear to be less interested than men in giving publicly. A recent
survey by the Committee of200, a group of successful business women, found that just 3 percent
of women donors indicated that they would be interested in having a buildhg named in their
bonor or even a plaque engraved with their mHIW, while 40 pcrccm preferred no :-ccognitiort
12
�Instead of publicity, women often C~1e personal gratification 0: having ae impact as the
for givir.g. In accordance with these principals, WOlnen prefer to maintain contact
with the prj)~iects they fund and with tbe people involved.. They also tend to cirect their gifts
moti\'a~ion
I
towards specific purpo$es such as athletic reruns, \vomen's scholarships, and specific facHides (a
new concert hall or a women's gymnasium, for example) rather than 10 endowment campaigns. :
However, as women take on new roles in the business world and !earn,more about philamhrop),>
this d'istiaction between men and women appears 10 be fading. In a 1999 follow~up to a 1992
1
survey~ UCLA s Women & Philanthropy group found that women are becoming less likely to
give soleiy because they are interested in the cause, and are beginning to treat phIlanthropy in a
more business~like manner.
Race and Ethnicity
Evidence suggests lltol col'lfribuJion.l' dijTer by race and elhnicily, When adequately lJ(;crnmling
for d(fferenceJ' in economic status, African Americans are mOte likely to give than whites, and
the amounts given are similar for tile two group...', This conclusion contrasts sharply with resuiIs
obtained whr!n diJferences in financial resources are ignored.
Despite significant increases over the past B years in the household incomes ofminoritlcs,
therc remains ~~ substantial gup bc"twecn the economic resources of African Amcricans and
Hispanics and those of whites, '1l1is gap is likely to be reflectcd in the amount of donations that
families CUll afford to make, Whcn ignoring these important differences in resources it docs
uppear tlwt both African American· and Hispanic families are less likely to give and give Jess
than white families. Preliminary results from Giving and Volunteering 1999 show that 75 percent
of whites reported making contributions in ] 998 compared to 52 percent of African Americans
und 63 percent of Hispanics. Because African Americans and Hispanics on average have lower
income and wealth than whites and nrc thercfo!c likeiv make smaller donations, one would
expect the )500 J'!linimulTI on giving in the SCF to miss n~ore minority giving. 9Jn facl~ giving by
all groups is Jov>'cr in the SCF than in Giving and Vo!uweering 1998, and the percentage change:
is the largest for Hispanics. Thirty~five percent of whites in the SCF made a donatio))'l compared
to 21 percent of African Americans and 12 percent of Hispanic-s. The dollar 'amounts given in
the SCF also differ by racc/clhnici!y. Whiles gave an average ofS3,356, African Americans gave
$2,459, and Hispanics $1 ,627.
When differences in income, wealth, schooling. and other obsef\:ahle characteristics are
lakcn into account, these conclusions change dramatically. With adequate controls for economic
status, our nnalysis shows that African Americans are actually significantly more likely to give
than whites, Hispanics remain significantly less likely than whites to have made a contribution, 10
~~
..
~~~~~-
We caution the reader that o1.lr conclusions "bom racial and c~hnic ciITerences arc based on: analyses or rclalivcly
small samples,
ll} However, Y-:cf.;use lhe difference bctwcc::l the SCF and the Giving and Vohmfeering numbers for :he fruction of
the population making II gin is significanlly larger for Hispunit:s tharr for whi,~s {Jr Al1icao Americaos, l! appears
thal measures of Hispanic givi:Jg are mQre wnsitive to the $500 cmooff, and giving may be substantially
underrcported There:o:c, \\'~ vicw [he resulls fOe Hispanics w:th cBL:tion.
9'
13
'.
�Among those who do give. African Americans g!\'e slightly larger amounts than whites, but the
ciiTctence is nOl sHdstically significant. This dmmatic change in the patten 01 giving by race
when a broad array of characteristics is controlled highlights the importance of multivariate
analysis. In particular, if we ignore the differences in wealth levels in our analysis, as is done in
many studi\~s~ but control for the other variables~ the slight difference between African American
and white giving is atten'Jated.
'
F~ctors
other than differences in fina!1cial resources and scbooling are also likely to be
important in explaining the c.ifferencc if: giving across racial and etlmlc groups. Differences in
exposure to opportunities lO give and the desired beneficiaries of charitable giving are also likeiy
[0 matteL Giving and VQlunteering }996 repons that African Americans and Hispanics are asked
to give much less often than whites. One study reported that H!spanies reee:ve an average of 15
to 20 requests for donations per year compa:ed to 300 for other grotlp~, And yet, another £!;udy
found that when asked, African Americans and Hispanics were more likely to respond positively
to a request than whites, If solicitntions serve to increase giving, tben organizatioliS arc
overlooking, an importont rcsou'rcc by llot sollciting donations from Arrican Americans and
.1"
•
·,lspamcs al greater rates. 1I
1
EXptlrts on philan:hropy in minority communities hm'e also a;gtlcd that minority giving is
less likely to be induded in existing d.ata on ·giving because much of it is done infomlally. For
instance, m~lny Hispanics send remittances to extended family in other countries.; estimates are
that Hispanics living in the United States send at least $3 billion per year to Mexico alone, By
excluding this type of giving, the survey data \.Is::d :n published studies may underestimate giving
by Hispanics relative 10 othe~ groups. J2
'
Even within the formal philanthropic 5..:(;tO:, !:linoritics choose different reciplcn1~ than
do whites. They arc less. likely to contribute to endowment campaigns, and instead focus their
giving on religious institutions and organizations or on efforts that meet pressing needs, The low
partidpation of minorities. in formal philanthropic giving i!i potentially n result of the services
provided by many charitable organizations, In 1997 just undcr 8 percent of ali roundation grants
went to minority concerns. Thus one way to motivate- minorities to give more is to provide
increased oppoTlunltie-s 10 give to organizations that more directly address their concernS. One
organiza:io!11hut docs nppeur to attrtlct a significant amount of gifts from minorities is thc United
Way. In a 1996 survey 26 percent of whites, 30 percent of African Americans; and 24 percent of
Latinos reponed making contributions to the United Way in the previous: 12 months. 13
Minorities, like women, have less of a history of giving to forma: philanthropic
organizations. This implies that outreach activities aimed particularly at these communities, like
those discus::cd earlier that target women, may yield increases in giving. Some organizations are
beginning to understand the potential. Fundraisilm programs using Spanish language matcnals
and foundmions using their grants 10 fund programs in Hispanic communities bave had
P The direction of tile rdailonship b1!!ween wlicilalio:1S Md gi\'ir.g is no: clear. As llOt~d later in the paper, it may
be that organizations ask for contributions from those wbo a:-c more likely 10 give (perhaps because of financial
capabilities) or ·...·110 have given g.encrously in {he past.
l! Giving USA Update, issue 2, 1999.
:, ihid.
)4
�increasing success. These efforts indicate that by addn:ssing the needs and concems of
minoritlc:::r charitable institutions ca:1 substamially incfe,ls~' the wntr:butlons of the,sc groups. 14
Bcquest~
Among lite! very \feu!rhy a large fruction of charifabJe giving is done through beque;;ts.
. Because the stl.ltis~ics ir', the ilfcvious section are cienvcd from survey respondents, there
is no informat~on nn the distribution of beq;JeSlS, Wc therefore ,know much less nboat trends in
charitable tcquests for representative samples of the population than we do about inter vivos
giving, Instead of surveys, past studies ~f bequests have drawn on data from estate'tax returns.
These tax retuDl5 are filed only by estateS that have made bequests and taxable gifts which total
to more than the taxwexempt limit ($675.000 in 2000). While these estates are !lot repn~sentative
of the estates of all decedents, they arc likely to be rcsponsi.bic for the vaSt majority of charitable
bequests.
For many wea}thy individuals, bequests are an import:mt mode of philanthropic giving.
In 1992 only 19 percent of estates filing tax returns made a charitable ~quest, but the lotal
amount given was $10.0 billion (inf1atioll-adjusted 1999 dollars), equal to 8.5 percent of the total
net worth of the estates and significantly higher than the fraction of income or wealth given in a
pal1~cclar year, Furthermore, among those who did make a charitable beques!, a significa:1t
fraction of the estate was donated with bequestS equal· to 27 percent of net worth,15
The magnimde of chari,table bequests made by the wealthy ofien surpaNses their imcr
vivvs giving in magnitude, Using estate and income tax rctum data, a reeent stud), finds that, on
average, v;eahhy decedents gave $8.9 mHlion (inflation-adjusted 1999 dollars) to charities at
death, compared [0 just $3.1 million during the 10 years prior to their death. For' the wealthiest
segment of the population, those with estates valued at more than $100 million) bequests
accour.ted fi}r 78 percent of charitable giving over the final 10 years of life,16
As with infer vivos giving, there arc, differences in bequest behavior by gender Bcca;Jse
Wives typically oUllive their hl'sbands~ female decedents arc much less likely than male
\0 be married .at the time of death, and to leave assets to a surviving spouse, Likely
because of 1his difference, females leave a greater fraction of their estates to charity. (11 1992
female decedents left 10.1 pen;:ent of their net worth to c.harit), on average; compared to 7.S
percent for males, This difference in philanthropic. behavior between males and females reverses
when onc examines only tlle behavior of those who have alxudy lost tileir spouse; widows left
12.1 percent of their net worth to charity while widowers Ic;-t l2,6 percent!7
decedents
14
Henry Ramcs, "Ltnino P!'!ilar.thropy: Expam:!ing US Mode!s of Giving and Civic Participation," 1999.
15 Internal ReV::nll~ Service, "Statistics ofIncome Bulletin:'
16
Rethinking Eslale lind Gift Taxation (2000).
17
Wimer 1996.97,
David Joulfaian, "ChnrilabJe Giving in Life and Dealh," forthcoming in Wi!1j~m Gale and Joel Slcmrod. tditors,
1RS, Winter 1996·97.
15
�Summary
Income and wC\ll!h arc important detem:inants of charitable giving, Over the pust decade
income and wealth have increased substantlal1y wHh charitable giving following suit Because
wc cnnnot guarantee equally strong ccoaomic gro\\'th into the indcfinite futurc, fUr',her inc::eases
in philanthropic bchavior can perhaps best be attained by tapping undcr~u$cd groups:
encouraging giving among the young; widQws~ sin£le men) and minorities. We return to this·
issue in the final section of the paper.
RECIPIENTS OF CHARITABLE GIVING
Contempora:-y philanthropy sup-ports a wide range of activities and causes. As the amount of
charitable giving hus grown, there have also been changes in the way in which money is
distributed to the various 'lypes of phibmhropic o:'ganil.ali::ms.
Religion is by far the single largesl recipient
of contributions, while gifts to /mmdalions
Chart 8. Recipients of Givinq. 1998
Gre file Jastesr growiJ7g"
Based on data from. Giving USA 1999.
Religion is by far the single largest recipient of
chariwblc giving, with about 44 percent of giving. in
1998 going, to religious organizations. 'S Education
is the second largest r~cipienl but trJ.i1s religion
significamly, with rcc'cipts of $27 billion in 1999,
compared to $82 billion for religion in the same
year.
N~_'
Although :l1c amount gIven to religion
Chart 9. Char.ge in Share of Giving, 1990·98
,
EdUU:.on
Publo~elf ~"
E'W"tm"""nt
r
foundations have t"xperienced tremendous
~
-
total giving
·10
·5
0
~fW1\l1~~ Ch~llP~
bCl\VCCll
1990 and 1998 c:.1Jne from
greater gifts to founilillions, while st:ch gifts were
responsible for on!)' 6 percent of the increase during
...a\
•
gfO\\'th~
increasing by 3) 3 percent in real terms between
1990 and 1998. Indeed, 29 percent of the rise in
I,
R~
remains large, its relative. importa:1ce has
diminishl-d recently.
In contmst, gifts to
the 1980,< In 1998 approximately $10 oul of every
5
$100 donated in the United Stales went to
foundations, up from Jess: than $4 in 1990. As a
result of the relatively slow grmvth rate ir. religiOl.:s
13 Link of wha1 is giVC1110 religious congregations is rediSlrlblll~d else\>'hcre, While approximutely 86 percent of
the revenue ofreligiolls congregations is derived from comributio['ls, 82 pcrccllI OfCXpC:llditures are for operating
expenses or capital improvcmcl1ls.
16
���������������������������������
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
Clinton Administration History Project
Creator
An entity primarily responsible for making the resource
Cinton Administration History Project
Council of Economic Advisers
Department of Commerce
Central Intelligence Agency
Department of the Interior
Department of Defense
Corporation for National Service
Council on Environmental Quality
Department of Justice
Domestic Policy Council
Department of Education
Department of Energy
Environmental Protection Agency
Federal Emergency Management Agency
General Services Administration
Small Business Administration
Social Security Administration
United States Agency for International Development
National Economic Council
Office of Management & Budget
Office of National Drug Control Policy
Office of Personnel Management
Office of Science & Technology Policy
Office of the Vice President
United States Trade Representative
Date
A point or period of time associated with an event in the lifecycle of the resource
1993-2001
Description
An account of the resource
<p>The Clinton Administration History Project describes in detail the accomplishments of President Clinton's Administration for the period 1993-2001. The records consist of the histories of 32 agencies or departments within the Executive Branch. In general, each organization associated with the Project submitted a narrative history along with supporting documents. These narrative accounts are primarily overviews of the various missions, special projects, and accomplishments of the agencies. The supplementary records include substantive memos, press releases, briefing papers, and publications illustrated with photos and charts.</p>
<p>Agencies:<br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Council+of+Economic+Advisers&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Council of Economic Advisers</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Central+Intelligence+Agency&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Central Intelligence Agency</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Commerce&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Commerce</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+the+Interior&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of the Interior</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Defense&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Defense</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Corporation+for+National+Service&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Corporation for National Service</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Council+on+Environmental+Quality&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Council on Environmental Quality</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Justice&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Justice</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Domestic+Policy+Council&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Domestic Policy Council</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Education&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Education</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Energy&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Energy</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Environmental+Protection+Agency&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Environmental Protection Agency</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Federal+Emergency+Management+Agency&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Federal Emergency Management Agency</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+General+Services+Administration&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the General Services Administration</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Health+and+Human+Services&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Health and Human Services</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Housing+and+Urban+Development&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Housing and Urban Development</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Labor&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Labor</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+National+Economic+Council&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the National Economic Council</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Office+of+Management+and+Budget&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Office of Management and Budget</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Office+of+National+Drug+Control+Policy&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Office of National Drug Control Policy</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Office+of+Personnel+Management&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Office of Personnel Management</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Office+of+Science+and+Technology+Policy&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Office of Science and Technology Policy</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Office+of+the+Vice+President&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Office of the Vice President</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Small+Business+Administration&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Small Business Administration</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Social+Security+Administration&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Social Security Administration</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+State&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of State</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Transportation&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Transportation</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+the+Treasury&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of the Treasury</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+United+States+Agency+for+International+Development&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the United States Agency for International Development</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+United+States+Department+of+Agriculture&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the United States Department of Agriculture</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+United+States+Trade+Representative&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the United States Trade Representative</a><br /><a href="http://clinton.presidentiallibraries.us/items/browse?search=&advanced%5B0%5D%5Belement_id%5D=39&advanced%5B0%5D%5Btype%5D=is+exactly&advanced%5B0%5D%5Bterms%5D=History+of+the+Department+of+Veterans+Affairs&range=&collection=21&type=&user=&tags=&public=&featured=&exhibit=&submit_search=Search+for+items">History of the Department of Veterans Affairs</a></p>
Is Part Of
A related resource in which the described resource is physically or logically included.
<a href="http://clinton.presidentiallibraries.us/items/show/36051">Collection Finding Aid</a>
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
Clinton Presidential Library & Museum
Format
The file format, physical medium, or dimensions of the resource
Adobe Acrobat Document
Extent
The size or duration of the resource.
1474 folders in 111 boxes
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
[Council of Economic Advisors] [7]
Creator
An entity primarily responsible for making the resource
History of the Council of Economic Advisers
Clinton Administration History Project
Date
A point or period of time associated with an event in the lifecycle of the resource
1993-2001
Is Part Of
A related resource in which the described resource is physically or logically included.
Box 1
<a href="http://clintonlibrary.gov/assets/Documents/Finding-Aids/Systematic/Administration-History-finding-aid.pdf">Collection Finding Aid</a>
<a href="http://catalog.archives.gov/id/1224798">National Archives Catalog Description</a>
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
Format
The file format, physical medium, or dimensions of the resource
Adobe Acrobat Document
Publisher
An entity responsible for making the resource available
Clinton Presidential Library & Museum
Medium
The material or physical carrier of the resource.
Reproduction-Reference
Date Created
Date of creation of the resource.
6/24/2011
Source
A related resource from which the described resource is derived
1224798-council-economic-advisors-7
1224798