Arts United States. Congress. Senate. Committee on Labo.

Education's impact on economic competitiveness : hearing before the Subcommittee on Education, Arts, and Humanities of the Committee on Labor and Human Resources, United States Senate, One Hundred Fourth Congress, first session ... February 2, 1995 online

. (page 8 of 15)
Online LibraryArts United States. Congress. Senate. Committee on LaboEducation's impact on economic competitiveness : hearing before the Subcommittee on Education, Arts, and Humanities of the Committee on Labor and Human Resources, United States Senate, One Hundred Fourth Congress, first session ... February 2, 1995 → online text (page 8 of 15)
Font size
QR-code for this ebook


tripled for holders of a bachelor's degree, and more than tripled for
those who held advanced degrees.

Keeping in mind that the consumer price index was about 2V2
times in 1992 what it was in 1975, these results indicate that the
earnings of high school dropouts did not even keep pace with infla-
tion, and high school graduates barely managed to stay even. Real
wages rose only for persons with education beyond the high school
level.



49

If these patterns continue, earnings differences between low and
high levels of education may become even more dramatic than cur-
rent patterns indicate.

Thank you.

Senator Jeffords. Thank you very much,

[The prepared statement of Mr. Kominski follows:]

Prepared Statement of Dr. Robert Kominski

Thank you for the opportunity to appear before the subcommittee this morning.

Since the Census Bureau began routine measurement in the 1940's, the edu-
cational attainment of the U.S. population has continued to rise, and is cun^ently
at an all-time peak: in 1993 four fifths of all persons ages 25 or older had completed
high school, and over 20 percent had earned a Bachelors degree or more.

When we examine the earnings of individuals cross-classified by their education,
we see that there is a strong relationship between these two factors. Based on data
obtained from the 1993 Cuirent Population Survey, administered by the Census Bu-
reau, we are able to show that mean annual earnings in 1992 ranged from-a low
of $12,809 for high school dropouts to a high of $74,560 for persons with a profes-
sional degree, such as an M.D. Other levels of education fell into line with this rela-
tionship. For example, high school graduates earned $18,737; persons with some col-
lege but no degree made $19,666; Associate degree holders on average made
$24,398; Bachelor's earned $32,629; while Master's degree holders made $40,368;
and persons with Doctorate degrees averaged $54,904.

Data from the CPS confirms that the strong relationship between earnings and
education not only exists in average across the entire population, but within age
ranges or at different points of the fife course, as well. Looking at the working popu-
lation across 10-year age intervals of 25-34, 35-44, 45-54 and 55-64, we see the same
strong earnings-education relationship in each age group.

It is possible, using the average earnings within these age specific groups, to com-
pute the possible returns to a given level of education over the course of a hypo-
thetical 40-year "working life", that is between the ages of 25 and 64. Doing this,
we can see the effects of the education-earnings relationship compounded over time.
Worklife estimates derived using this method show strong differences for edu-
cational levels. For instance, high school dropouts would make about $609,000 over
their life, and high school graduates would make about $821,000, while persons
with some college but no degree would make $993,000. Associate degree recipients
would make about $1,062,000; Bachelor's degree holders, $1,421,000; Masters de-
gree holders, $1,619,000; Doctorates, $2,142;000, and persons with Professional de-
grees, $3,013,000.

This estimation method assumes that 1992 earnings levels would stay in effect
throughout one's lifetime, but the reality is that the value of a dollar has changed
over time. Recent history shows that the value of higher levels of education has
risen faster than lower levels. Comparing 1975 returns to those in 1992, we see that
average earnings doubled for high school dropouts, rose 2.5 times for those with
high school degree only, almost tripled for holders of Bachelor's degree, and more
than tripled for those who held advanced degrees. Keeping in mind that the
consumer price index was 2.5 times in 1992 what it was in 1975, these results indi-
cate that tne earnings of high school dropouts did not even keep up with inflation,
and high school graduates barely managed to keep pace. Real wages rose only for
persons with education beyond the high school level. If these patterns continue,
earnings differences between low and ni^ levels of education may become even
more dramatic than current levels indicate.

Senator Jeffords. Dr. Schapiro.

Mr. Schapiro. Thank you.

This testimony is on behalf of myself and my co-author Michael
McPherson. We have been invited to comment on the relationship
between tuition charges and college enrollment and to describe the
effect of college enrollment on lifetime earnings.

We considered first economic returns to higher education. Post-
secondary training contributes to labor force skills at a variety of
levels. Considerable evidence — we just heard some a couple min-
utes ago — exists that the economic returns to a college education
have risen in the last 15 years. This is reflected in a widening gap



50

between the earnings of workers whose terminal degree is a high
school diploma and those who continued on to college. High returns
apply at all levels of postsecondary education.

The earnings gap has widened between those with some college
and those with none, as well as between college graduates and
those with some college.

What accounts for these higher returns? Are they likely to be a
transient phenomenon? Although fluctuations in the number of
people entering the labor force plays some role, the main source of
higher returns to education has been on the side of the demand for
labor, a result of the rapid pace of technological change in the U.S.
economy.

Those parts of the economy which rely less on college-educated
labor, such as farming and heavy industry, have declined in impor-
tance, while industries which use more college-educated workers
and have been driven in a major way by technological change, such
as financial services and high-tech manufacturing, have grown.

Continuing rapid technological change implies that this trend is
likely to continue, and thus the economic payoff to higher levels of
education is expected to be substantial for the foreseeable future.
Hence, from the standpoint of economic efficiency and growth, the
Nation will require and continue to require high levels of invest-
ment in human skills.

We turn now to the impact of college costs on enrollment deci-
sions. Our own empirical research and our review of other studies
lead us to stress two points. First, students' decisions to enroll in
college are sensitive to the net price they pay. Higher tuition
charges or reduced student aid have a measurable and nontrivial
effect in discouraging enrollment, while reduced charges or in-
creased student aid encourage enrollment.

Second, enrollment effects are concentrated on lower-income stu-
dents — and here, by "lower-income," we mean below family income
of about $25-$30,000. Our study, which examined the impact of
fluctuations in prices and aid levels on U.S. college enrollment over
an 11-year period, showed that a $100 increase in the net cost pro-
duced about a 2 percent decline in the enrollment of lower-income
students but had no observable effect on the enrollment of middle
or upper-income students. The observation that lower-income stu-
dents are more responsive than others to price changes is consist-
ent with the findings reported in studies by many other research-
ers.

Another aspect of the cost question is the impact of college prices
on enrollment destinations of college students, as opposed to
whether they attend any college at all. Much of the popular discus-
sion regarding where students go involves students from middle-in-
come backgrounds. It is widely believed that increases in private
tuitions have caused a massive exodus of middle-income students
out of private higher education and into public colleges and univer-
sities.

We recently completed a study of the income backgrounds of
American college freshmen over the period 1980-1993 and uncov-
ered a rather different story. We did not find a redistribution of
middle-income students from private to public institutions. In 1980,
21.5 percent of middle-income students were enrolled at private 4-



51

year colleges and universities. In 1993, 21.2 percent were in those
institutions.

The most striking movement among middle-income students has
in fact been within the public sector, with a sharp decline in the
share of middle-income students at public 2-year institutions, offset
by growth in the share of middle-income students at public 4-year
institutions.

Indeed, the increasing concentration of low-income students at
public 2-year colleges suggests a worrisome trend. It implies that
rapid puolic tuition increases, along with limitations on Federal
student aid, are combining to deny many low-income students ac-
cess to public universities and 4-year colleges, in effect limiting
their postsecondary choices to community colleges and trade
schools.

The main role of the Federal Government in higher education
has been to widen the educational opportunities of those people
least able to afford college. With public tuitions on the rise and a
growing economic need for college-trained workers, it is urgent that
Congress concentrate its energies and the Nation's limited re-
sources on fulfilling this role.

Thank you.

Senator Jeffords. Thank you very much. Dr. Schapiro.

[The joint prepared statement of Mr. Schapiro and Mr. McPher-
son follows:]

Joint Prepared Statement of Morton Owen Schapiro and Michael S.

McPherson

Our charge is to comment on the relationship between tuition charges and college
enrollment, and to describe the effect of college enrollment on lifetime earnings. We
will argue that the strong link between postsecondary training and income, the
strong link between federal financial aid and the price of higher education, and the
strong link between the price of higher education and the enrollment of students
from low-income families, imply that the government policy regarding financial aid
is critical from both an economic and a social perspective.

Certainly no aspect of the evaluation of federal student aid has attracted more
attention than the question of its impact on enrollment levels and patterns. Al-
though it is important to note that affecting enrollment is not the whole justification
for student aid, the aim of promoting the enrollment of targeted groups has been
central to the case for federal student aid throughout its histor>'.

A great many studies over the years have used data for a sample of students in
a particular year to estimate the impact of price or net cost of education on students'
postsecondary education decisions. A minority of those studies have tried to measure
specifically the effect of student aid on enrollment decisions, with the rest focusing
on the impact of tuition price. Although the studies differ widely in data sources
and estimation techniques, they tend to agree on two main points. First, student
decisions to enroll in college respond positively, and nontrivially, to price cuts or aid
increases. Second, decisions about where to attend school also respond non-trivially
to changes in the relative prices of schooling alternatives.

In our own empirical work presented in chapters 2 and 3 of our 1991 Brookings
Institution book, Keeping College Afibrdable, we examine the aid enrollment rate
relationship over time while taking account of differences in the experiences of dif-
ferent income groups and genders. Blacks and other races are excluded from this
analysis because sanipling variability due to small samples from these races in the
Current Population Survey make these data points inappropriate for time-series
analysis at the level of disaggregation we wish to employ. Therefore the results we
report here are limited to whites only.

Our results indicate that increases in net cost over time lead to decreases in en-
rollment rates for low-income students. The magnitude of the coefficient on net cost
implies that for lower-income students a $100 net cost increase results in an enroll-
ment decline of about .68 percentage points, which is about a 2.2 percent decline
in enrollment for that income group.



52

While our estimate of the magnitude of the enrollment responsiveness to changes
in cost differs slightly from the average of econometric studies based on cross-section
data, the imjxjrtant pomt is that our economctrically controlled time series analysis
supports the view that changes in cost lead to changes in enrollment for low-income
students.

We found a very different picture when we looked at the behavior of more affluent
students. We found no e/idence in these data that increases in net cost inhibited
enrollment in these income groups.

This analysis indicates that changes in the net price facing lower-income students
have significant effects on their enrollment behavior. An important policy issue,
however, is whether changes in federal aid in fact wind up changing net cost. If for
example, increases in federal aid were to lead to decreases in the eimount of aid
awaroed by institutions or to increases in tuition, the effect of aid on net cost would
be muted. However, in a separate empirical analysis of the effects of student aid
on institutions (Keeping College Affordable, chapter 4), we found no evidence that
these potential offsetting effects are empirically important. The time series evidence
on net cost further suggests that periods when federal aid is generous coincide with
periods when the net cost facing low-income students is lower. This supports the
view that these potential offsets are not an important factor.

In sum, our analysis supports the hypothesis that federal student aid has signifi-
cantly affected enrollment patterns in U.S. hi^er education over the past two dec-
ades.

Our evidence indicates that student aid matters: when student aid lowers the
costs facing lower income students, it tends to encourage higher enrollment. Policy
makers must carefully consider potential enrollment efiects when determining stu-
dent aid poHcy.

A second implication is that, consistent with what most observers would expect,
changes in net cost have larger effects on the enrollment behavior of low-income
than of higher-income students. Indeed, we did not observe any consistent effect of
changes in net cost on the enrollment levels of middle- and upper income students.
To the extent that federal student aid pwlicy aims to increase college enrollments,
this finding supports the case for targeting student aid on lower-income students.

Trends in gross price, need-based aid, and merit aid aU affect the affordability of
different types of hi^er education for different types of students. While a great deal
of attention has been paid by higher education researchers to the question of "ac-
cess" — whether students from various economic backgrounds attend college, less at-
tention has been paid to the question of "choice" — where do these students go.

We have just completed a study on the distribution of college students by income
background in an attempt to address the often elusive issue of choice in higher edu-
cation. *

Much of the popular discussion regarding where students go involves middle in-
come students. It is often suspected that students from middle income backgrounds
have been most affected by the considerable real increases in tuition at private col-
leges and universities. Students from lower income backgrounds qualify for need-
based financial aid, lessening the chance that these students experience an afford-
ability problem. Students from upper income backgrounds receive a different but
analogous form of financial aid — parental contributions that do not require major
proportions of available annual incomes. But, the story goes, when tuitions rise fast-
er than other economic indicators, students from middle-income backgrounds are
forced to switch to less costly educational alternatives.

Our study of the income backgrounds of American college freshmen over the pe-
riod 1980-1993 casts doubt on parts of this analysis. In one sense, there has been
such a melt: the share of mid(De income students (defined as the group with real
family incomes of $30,000 to $100,000 in 1992 dollars) in all of higher education has
declined. Our data cannot tell us whether this overall decline represents changes
in national income distribution (fewer families in the middle class and more either
rich or poor as a result of the Reagan-Bush years) or differential changes in enroll-
ment rates (middle income students increasing their college enrollment rates less
than students from richer or poorer families). But what most people seem to mean
by middle income melt is something different from this: a redistribution of middle
income students among categories of institutions, and especially from private to
public institutions. Our data do not find middle income melt in this sense over the
1980-1993 period. In 1980, 21.5 percent of middle income students were enrolled at



1 Michael S. McPherson and Morton Owen Schapiro, "College Choice and Family Income:
Changes Over Time in the Higher Education Destinations of Students firom DifTerent Income
Backgrounds," November 1994.



53

private four-year colleges and universities; in 1993, 21.2 percent were in those insti-
tutions.

The most striking movement among middle income students has in fact been
within the public sector, with a sharp decline in the share of middle income stu-
dents at public two-year institutions, offset by growth in the share of middle income
students at public four-year institutions. Indeed, one of our most interesting find-
ings is the increasing representation of low income students at public two-year col-
leges, and the declining representation of middle and upper income students there.
It is of course important to remember that the relatively young, first-time full-time
freshmen represented in our survey are not the predominant clientele at community
colleges. Nonetheless, these data do seem worrisome. They suggest that the com-
bined effects of tuition increases and limitations on federal student aid may be im-
pairing the ability of low income students to gain access to institutions other than
community colleges.

It is revealing also to look at changes in the enrollment patterns of upper-income
students. Higher income students in 1993 were more likely to attend universities
(either public or private) than they were in 1980. These increases for universities
(and for public four-year colleges as well) came largely at the expense of private
four-year colleges, whose proportion of high-income students fell from 26.7 percent
to 23.6 percent. Meanwhile the proportion of middle-income students who attend

f)rivate four-year colleges has been a basically stable from 1980 to 1993. Althou^
eaders at these schools have been vocal in talking about middle income melt, it ap-
pears that what they have experienced is in fact upper-income melt. It seems likely
that this loss of full-pay students is a significant part of the explanation for the
growing investments of these schools in tuition discounting and non-need based aid
discussed above.

Among the implications that can be drawn from these various trends, one stands
out that is of special importance for higher education's future role. This is a pattern
of increasing stratification in several senses. Low-income students are increasingly
concentrated in community colleges, perhaps because a pattern of rising tuitions not
matched by student aid increases is pushing the cost of other public alternatives out
of reach. Private universities are pulling away from private colleges in their ability
to attract hi^ income students. Other research we have done suggests to us further
that the most selective and prestigious among private universities and liberal arts
colleges are increasing their distance from most other private institutions, which are
locked in a tense struggle for enrollment and resources.

Since World War 11, postsecondary education on a massive scale has made itself
indispensable to the normal workings of our society — gatekeeper for the professions,
training ground for all manner of skilled work, core generator of scientific and tech-
nological advances, and locus of scholarly endeavor and cultural critique. At the
same time, the principle of universal access to higher education has become an es-
sential symbol of the nation's commitment to equality of opportunity.

Postsecondary education contributes to labor force skills at a variety of levels.
Considerable evidence exists that the economic returns to educational investments
have risen in the last fifteen years. This is reflected in a widening gap between the
earnings of those with high school educations and those with higher levels of edu-
cation. These high returns appear to apply at all levels of postsecondary education —
the earnings gap has widened between those with some college and those with none
as well as between college graduates and those with some college.

What accounts for these higher returns? Are they likely to prove a transient phe-
nomenon? One source of the higher returns is temporary - a result of changing de-
mographics. Returns to higher education were depressed in the late 1960's and early
1970's as a result of the very large cohorts of college-educated workers who ap-
peared in the labor force a that time, as the baby-boomers matured. Since then, the
decline in numbers of young persons entering the labor force has produced some-
thing of a shortage of young college-level workers, and this has contributed to in-
creased returns. The impact of this force can be expected to diminish as the "echo"
of the baby boom leads to larger cohorts of young people in the decades ahead.

This supply side effect, however, does not appear to be the main explanation for
higher returns. Rather, most of the action has oeen on the side of the demand for
labor, and appears in fact to be a result of the rapid pace of technical change noted
above. Katz and Murphy have shown that rising demand for better educated work-
ers has been driven by the relative expansion of industries which have higher de-
mands for educated labor. That is, those parts of the economy which rely less on
college-educated labor (farming, heavy industry) have declined in importance, while
industries which use more college-educated workers (financial services, 'liigh-tech"
manufacturing) have grown.



54

Continuing rapid technical change implies that this trend is likely to continue,
and thus the economic payofT to higher levels of education is likely to continue to
be high for the foreseeable future.

Hence, from the standpoint of economic efficiency and growth, the nation is likely
to require high levels of investment in human skills.

The impact of rising demands for human skill and knowledge has important im-
plications for economic inequality. As we have noted, the high returns to education
over the last two decades have produced a growing gap between the wages of more
and less educated workers. Many observers have worried that this tendency is pro-
ducing a dangerous social divide between a small class of highly trained analysts
and professionals and a growing group of undereducated and underemployed work-
ers. There is a lot of evidence from international comparisons that our educational
system does relatively well by the highest achieving students and that our largest
failures, cognitively and later economically, are with students in the bottom half of
the distribution of academic achievement. One way of combating this inequality is
to invest more of the skills and knowledge of those people who do not under present
arrangements attend four-year colleges and universities and attain bachelors' de-
grees. Concern for this group calls for improvements at the high school level for stu-
dents who are not college-bound as well as for improving the availability of high-
quality non -collegiate postsecondary alternatives, such as vocationally oriented pro-
grams in community colleges and trade schools. It is important also to improve ac-
cess to four year colleges and universities for those among disadvantaged students
who have the motivation and qualifications to attend them.

The main role of the federal government in higher education has been to widen
the educational opportunities of those people least able to afford college. With public
tuitions on the rise and a growing economic need for college-trained workers, it is
urgent that Congress concentrate its energies, and the nation's limited resources,
on fulfilling this role.

Senator Jeffords. Dr. Bishop?


1 2 3 4 5 6 8 10 11 12 13 14 15

Online LibraryArts United States. Congress. Senate. Committee on LaboEducation's impact on economic competitiveness : hearing before the Subcommittee on Education, Arts, and Humanities of the Committee on Labor and Human Resources, United States Senate, One Hundred Fourth Congress, first session ... February 2, 1995 → online text (page 8 of 15)