January, 1997
Policy Insight #101

Power to the Student
An Alternative to Higher Education Funding Increases

The Use of Direct Student Assistance
to Finance Higher Education Expenditures

An Analysis of the Potential in the State of Oregon

Prepared by
Randall Pozdena, PhD
Portland, Oregon
Cascade Policy Institute
Portland, Oregon

This project was made possible thanks to the generous support of Reason Foundation, Los Angeles, California.


Over the last century, the public sector has come to dominate the higher education system in the United States and in Oregon. Public sector involvement in higher education is predominately through the direct operation of colleges and universities that offer low-tuition enrollment. Publicly-operated colleges and universities currently enroll nearly 80 percent of students attending institutions of higher learning in the country as a whole, and nearly 88 percent in the state of Oregon.

As elsewhere, the State of Oregon assists higher education primarily through direct appropriations to the publicly-operated institutions, with the effect of reducing tuition for all students, regardless of circumstances or need. Although Oregon has small programs to provide needs-based tuition assistance to students in both public and private institutions, they are dwarfed by the scale of direct government appropriations to the Oregon state higher education system. In the 1995-1997 biennium, appropriations to the state university system and the state system of community colleges will exceed $850 million. Student-based funding, in contrast, is trivial, at $44 million, only $7 million of which is available to students outside the higher education public sector.

The current system subsidizes the supply of publicly-produced higher education. It is traditionally defended by those who believe that the resulting low tuition cost broadens the appeal of a college education and, thereby, better serves both the egalitarian and economic goals of the State. This paper argues that directly subsidized, government provision of higher education is neither an equitable nor efficient means of assisting in the higher education of Oregon citizens. In particular, providing tuition assistance directly to the attending student, in the form of tuition assistance grants or some similar mechanism, would result in improved overall access to higher education at a given taxpayer cost. Such a reform would likely stimulate changes in the institutional structure of higher education, making it simultaneously more efficient and responsive to changing demand for educational services.

The outline of the remainder of this paper is as follows: The next section discusses the origins of public (government) involvement in the system of higher education, tracing the path by which the current, public-supply oriented system evolved. This is followed by the factors affecting the demand for higher education services, and how enrollments are affected by tuition and loan policies. After that, the institutional structure and organizational behavior of higher education institutions are examined, and the implications for the efficiency and effectiveness of higher education services. Finally, these various threads are brought together to evaluate the merits of providing education subsidies directly to students, rather than to the suppliers of higher education services. This approach is contrasted with traditional mechanisms of assisting higher education.

The Origins of Public Involvement in Higher Education Finance

Government involvement in higher education, like its involvement in K-12 education, emerged in the middle of the 19th century. Prior to that time, virtually all higher education services had been provided by tuition-funded, private colleges and universities. The private system of higher education responded to the rapid economic development that occurred in the 18th and 19th century, so that by 1860, there were approximately 750 private colleges and universities [1].

Although the first public institution of higher education was founded in the United States in 1816, it was the passage of the Morrill Act in 1862 that precipitated the modern public higher education movement. The Morrill Act offered federal lands to states to encourage college programs in scientific, agricultural, industrial, and military studies. The Morrill Act support could be used to support private, as well as public institutions. However, since the state legislatures administered the Morrill Act moneys, the funds typically went to form or support public, rather than private, institutions, because it was believed that private institutions might not be as responsive to the legislative interests that had backed the Act. (Farm interests, for example, blocked the designation of Yale University as a land-grant university, resulting instead in the creation of the predecessor of the University of Connecticut). The Morrill Act stimulated public provision of higher education on a large scale and, in this sense, invigorated the public university movement. Altogether, 72 land-grant colleges and universities were created under the Act. Over time, the original, narrow education mission broadened, with most of the land-grant colleges offering coursework in areas outside those designated in the Act.

As many chroniclers of higher education have pointed out, the relatively late entry of the public sector into the provision of higher education services had the effect of making public universities a more common phenomenon in newer regions (the west and Midwest) than in the northeast or southeast. This legacy persists today; the observed market mix between consumer-financed (mostly private) and government-appropriation-financed (mostly public) institutions varies greatly from region to region. Public university enrollment relative to private university employment is almost 10 times greater in Arizona (a new, western state), for example, than it is in Pennsylvania, which entered the 20th century with an abundant endowment of private institutions [2]. Oregon's public/private enrollment ratio [3] of 7.6:1 makes it the 16th most public-sector oriented of all the states, but relatively middle-of-the pack for western states.

Table 1 displays the trend, from 1950 forward, in the number of public and private institutions, and enrollment at those institutions, for the nation as a whole and for Oregon. Table 2 presents enrollment levels and the relative public/private enrollment rates --i.e. the ratio of total public university enrollment to private enrollment, by state.

Table 1: US and Oregon Institutions of Higher Education, 1950-1994

1950 1969 1970 1980 1990 1994 Chg 50-94

All institutions 1852 2021 2556 3231 3559 3688 99%
4-year 1312 1431 1665 1957 2141 2215 69%
2-year 540 590 891 1274 1418 1473 173%
Public institutions 636 700 1089 1497 1567 1641 158%
4-year 341 368 435 552 595 605 77%
2-year 295 332 654 945 972 1036 251%
Private institutions 1216 1321 1467 1734 1992 2047 68%
4-year 971 1063 1230 1405 1546 1610 66%
2-year 245 258 237 329 446 437 78%

Oregon, All Institutions



Oregon, Public Institutions



Oregon, Private Institutions




Source: Author, from NCES data, 1995; includes branches after 1970.

Table 2: Fall Enrollment in 2-yr and 4-yr Colleges and Universities, 1993

Ratio of
Public to Private

United States10,011,787 2,312,172 4.3 --

Alabama187,261 20,7589.0 11
Alaska27,351 1,696 16.1 4
Arizona222,075 17,582 12.6 5
Arkansas78,986 11,137 7.1 18
California1,498,921 129,289 11.6 6
Colorado180,466 19,902 9.1 10
Connecticut89,799 38,264 2.3 44
Delaware32,438 5,475 5.9 21
Distnct of Columbia10,020 37,681 0.3 51
Florida476,985 77,677 6.1 19
Georgia210,177 49,541 4.2 33
Hawaii43,707 10,805 4.0 36
Idaho40,988 10,663 3.8 37
Illinois499,601 121,975 4.1 34
Indiana200,110 55,637 3.6 38
Iowa106,567 43,195 2.5 41
Kansas133,249 14,915 8.9 12
Kentucky136,974 26,486 5.2 28
Louisiana151,259 19,936 7.6 16
Maine35,883 14,508 2.5 41
Maryland199,277 23,995 8.3 14
Massachusetts164,351 165,242 1.0 50
Michigan416,853 73,519 5.7 23
Minnesota187,982 43,108 4.4 31
Mississippi99,137 10,822 9.2 9
Missouri178,119 73,542 2.4 43
Montana30,840 5,105 6.0 20
Nebraska84,103 16,945 5.0 29
Nevada56,633 594 95.3 1
New Hampshire32,339 22,545 1.4 47
New Jersey247,862 45,300 5.5 25
New Mexico85,323 2,978 28.7 3
New York539,634 325,418 1.7 45
North Carolina274,413 57,524 4.8 30
North Dakota33,689 3,537 9.5 7
Ohio375,756 108,666 3.5 39
Oklahoma139,847 17,566 8.0 15
Pennsylvania309,249 215,063 1.4 47
Rhode Island35,337 32,261 1.1 49
South Carolina125,911 23,272 5.4 27
South Dakota27,192 6,381 4.3 32
Tennessee172,087 42,162 4.1 34
Texas742,236 80,123 9.3 8
Utah92,357 33,627 2.7 40
Vermont18,976 12,252 1.5 46
Virginia251,723 45,135 5.6 24
Washington227,411 27,219 8.4 13
West Virgina64,131 11,007 5.8 22
Wisconsin231,856 42,424 5.5 25
Wyoming27,013 700 38.6 2

Source: Author, from NCES statistics, 1993.

In contrast to 2-year and 4-year colleges and universities, the noncollegiate institutions of higher learning (usually called vocational and technical or "voc-tech" schools) are predominantly private. Nationally, there are 6,600 such institutions, only 500 of which are publicly-owned and operated. (In Oregon, there are 99 voc-tech institutions; none of which are public.) Over two-thirds of the private institutions are organized as for-profit, proprietary enterprises. The rest are operated as not-for-profit organizations.

It is difficult to compare enrollment measures between collegiate and non-collegiate institutions, since the content and intensity of the programs differ. The federal Office of Vocational Education, however, estimates that in 1991, there we 3.6 million adult individuals enrolled in such programs in 1991. Oregon had only 40,000 students enrolled in these programs [4].

Why should the Public Finance Higher Education?

The reform of higher education finance cannot be discussed without understanding why policy makers have chosen to subsidize higher education. As mentioned earlier, the persistent, wide variation in public enrollment rates across regions that direct public production of higher education services is not critical to the existence of the higher education sector. What, then, is the logic for public finance of higher education, by any mechanism?

Social Returns

One argument offered by education policy makers is that education provides returns not only to the beneficiary of that education (i.e. the student), but also to others in society. To the extent that this is true, the student should be encouraged to invest more in higher education than he might otherwise so as to permit the rest of society to reap the benefits. The strength of this argument depends on the relative size of private versus social returns to education investment.

There is relatively little dispute over the general range of private returns. The effects of higher education on the lifetime path of the student's income can be estimated by comparing the incomes of those with and without such education. Then, weigh these income effects against the cost of the education (including the income foregone while in school). From this comparison comes an internal rate of return (IRR) to the education investment. Like the interest rate on a bank account, the IRR of higher education reflects the economic yield of the investment in education. A college education continues to be significant and, indeed, the gap between the incomes of college and high school-only educated individuals not only persists, but is increasing. The consensus in the economics literature is that the private IRR of an undergraduate degree is on the order of 11.5 to 16 percent [5] , with returns to graduate degrees (other than professional degrees) generally being lower [6]. Thus, from a purely private point of view, an investment in higher education pays a fairly high rate of return, and purely private motives would stimulate college attendance even in the absence of public subsidies.

There is much greater debate whether there are social returns beyond the private returns -- i.e. benefits to someone other than the student. Some of the social effects of higher education that analysts have considered as social benefits include lower crime levels [7], lower burdens on welfare and public medical care systems [8], and the societal benefits from the inventions and other innovations of higher-educated people.

Upon closer scrutiny however, most such effects are less significant than they appear. Even if lower crime does accompany higher education, for example [9], there may be more economical ways to address crime than subsidizing college admissions. If so, the reduction in crime resulting from education should not be counted as an exclusive benefit of education. Similarly, although all of society benefits from scientific and medical research, the question is whether the individual, brilliant researcher responsible for such innovations needs more than the prospect of his own income enhancement to be encouraged to do such research; that is, will subsidizing higher education create more brilliant researchers or dissipate resources training mediocre researchers?

Economic Growth

It also has been argued that public subsidy of higher education serves to attract or retain professionals whose skills have value to the business community and the economy of the state, or the substate region in which the university or college is located. A variation of this argument is that the business activity at a university or college generates "multiplier" effects on jobs and income in the affected state or community. A series of studies in the early 1980s, for example, found income and job multipliers of 2.0 to 2.5 resulting from spending and employment at public and private colleges and universities [10]. Economists put little weight on such arguments, however, because they beg the question of what the economic effects would have been had the tax revenues supporting the institutions been put to alternative use. In addition, interregional competition for professionals, though appearing valid from the standpoint of any one state, has the ultimate consequence (if all states pursue the policy) of having no effect but to preserve the status quo, at considerable expense.

Some research has taken the more useful tack of studying the effect of education on aggregate economic growth. These studies confirm that there are economic growth benefits to education generally, but the specific benefits of higher education have proved more elusive. One researcher estimates that a quarter of the 18 percent contribution that education made to economic growth came from higher education [11], whereas another researcher put it at less than 17 percent [12], and others have failed to isolate an effect. In sum, it is fair to say that the empirical record only weakly supports the notion that higher education returns have a social as well as private component.

Social Equity

Another argument made in favor of subsidizing higher education is a social equity argument. Specifically, lowering the cost of access to higher education permits those whose means would otherwise not allow them to attend college, and enjoy the demonstrated enhancement in their income and careers. This argument has somewhat broader acceptance among economists, because it is recognized that financial markets do not provide an efficient means for individuals to borrow against their future human income potential, without effectively creating peonage relationships [13]. This limits acceptance of this type of lending activity by private financial institutions. Guaranteed student loan programs provide a partial solution, since the government has limited means of enforcing prompt payment. But without some form of assistance, credit market imperfections might otherwise keep qualified individuals from low income households from entering college. We will return to this argument later in this paper.

Other, less charitable explanations of the public sector's motive for directly producing higher education services has been hypothesized, including one considered by Quigley and Rubinfeld.[14] They pose a rather pessimistic model of public sector growth which says that politicians wish to develop subsidized public universities so as to create loyalty among their constituents. Quigley and Rubinfeld speculate that the community college movement may be an illustration of this phenomenon because these smaller, local institutions conveniently are able to be located in individual political districts.

Recent Trends in Education Finance

Whatever the cause for the significant public involvement in the production of education services, trends in higher education finance do not bode well for continued reliance on the current model of broadly subsidized, public education. Although state, local, and federal government support of public institutions of higher education more than doubled between 1980 and 1993,[15] to over $55 billion, the traditional sources of government revenue proved less elastic than total costs. State and local government support, which represented 49.4 percent of current fund revenue in the 1980-81 academic year, was only 40.5 percent in the 1992-93 academic year. Federal government support, too, fell from 12.8 to 10.8 percent over the same time period. (See Table 3.) The result was somewhat greater reliance on tuition revenue in public colleges and universities. Between 1980-81 and 1992-93, tuition as a fraction of total current fund revenue grew from 12.9 percent to 18 percent.

Although some state and federal aid is provided to private colleges and universities, that aid is relatively small, and decreasing rapidly. Approximately 2.3 percent and 14.9 percent of current revenues at private institutions in 1992-93 were from state and federal governments, respectively. Federal aid declined 20.7 percent between 1980 and 1993 (versus a decline of only 15.6 at public institutions). Though state aid grew slightly, the retreat from the public finance of higher education has been faster at private institutions that at public ones over the last fifteen years or so.

The overall effect of these trends has been to increase reliance on tuition generally, but also to widen the gap between private and public institution tuition. Between 1980 and 1994, tuition, room and board at private institutions increased by 197 percent, versus the 175 percent tuition inflation experienced at public universities and colleges. Nationally, the absolute tuition, room and board premium for a private education was a daunting $10,260 in 1994, since private institutions financed over 41 percent of their expenditures from tuition, versus only 18 percent for public institutions [16]. The ratio of tuition, room and board costs at private versus public institutions, which has hovered around 2.0 since the 1950s, is now over 2.7. A similar picture emerges for Oregon, where the differential between private and public tuition levels more than doubled from about $5,000 in 1980 to $11,300 in 1995 on an inflation-adjusted basis.[17]

The spending patterns of both the public and private higher education sectors have been affected by the increasing reliance on tuition. At both public and private colleges, for example, the most rapidly growing cost has not been instruction costs, but the cost of scholarships and fellowships. In constant dollar (inflation-adjusted) terms between 1982 and 1992, such financial aid increased 85 percent and 100 percent at public and private institutions, respectively.

The private higher education sector has been significantly affected by the growing tuition advantage enjoyed by its public sector counterparts. It's response to the sharpened price competition by increasing its emphasis on the quality dimension of its education. For example, between 1982 and 1992, instruction expenditures per student increased by only 13 percent at public institutions, but increased 41 percent at private colleges and universities. Private institutions, of all types, spend significantly more, in absolute dollar terms per student, on both instruction and research (with the exception in the latter case of 4-year colleges). (See Table 4.) Spending per student on libraries and student services is nearly twice the level observed at public colleges and universities, and the higher operations and maintenance expenses imply more extensive educational plants.

In contrast, public universities, over time, have tended to reduce relative spending on instruction and libraries in favor of research and affiliated enterprises such as hospitals. Between the 1980-81 and 1992-93 academic years, for example, the share of instruction spending fell 10 percent and library spending fell over 20 percent, while spending on research increased 10 percent, and spending on affiliated hospitals increased over 30 percent. At private universities, in contrast, the instructional spending share was largely unchanged, and research spending fell. (Library spending declined at private institutions as well, however.)

Increasingly, therefore, the policy of subsidizing public higher education institutions has had the effect of causing private institutions to compete more aggressively on quality. In some sense, as the public higher education sector has grown through a policy of low prices coupled with a lower instructional commitment, the private higher education sector has had to raise the ante, as it were, by increasing spending on scholarships and instruction and, by necessity, by increasing posted tuition levels.

Table 3: Current Fund Revenues and Sources, Public and Private Institutions, 1980-1992

Public Institutions


Revenue ($B)$43.2 $65.0 $69.6 $74.8 $81.9 $88.9 $94.9 $102.2 $108.2


Tuition and fees12.914.514.715.015.215.516.117.118.0
Federal government12.810.510.410.310.310.310.310.610.8
State governments45.645.043.743.442.541.740.338.336.8
Local governments3.
Pnvate gifts, grants2.
Endowment income0.
Sales and services19.620.
Other sources2.


Private Institutions


Revenue ($B)$22.4 $35.4 $39.2 $42.6 $46.6 $50.7 $54.9 $59.2 $62.7


Tuition and fees36.638.639.639.139.439.640.440.741.2
Federal government18.816.51716.616.115.915.415.314.9
State governments1.
Local governments0.
Private gifts, grants9.39.39.398.
Endowment income5.
Sales and services23.323.421.722.622.722.622.923.323.2
Other sources4.


Source: Author, from US Department of Education IPEDS database.

Table 4: Composition of Expenditures, by Type of Institution, 1992

Other Public
Other Public
Private 4-yr
Private 2-yr

Expenditures ($ per student, 1992)

Educ. and gen.
    Public service
    Academic support
    Student services
    Institutional support
    Operations and maint.
    Scholarships, etc.
    Mandatory transfers
Non-ed enterprises2,7004,8803,7244015,90513,5813,5895,905


Expense category:As Percentage of Education and General Expenses
    Public service
    Academic support
    Student services
    Institutional support
    Operations and maint.
    Scholarships, etc.
    Mandatory transfers


Source: Author, from U.S. Department of Education IPEDS database.

Table 5: Tuition, Room and Board Charges, per student, 1965-1994


Public Institutions

All$983 $1,287 $1,666 $2,373 $3,571 $4,757 $5,962
4-yr -- -- --$2,550 $3,859 $5,243 $6,674
2-yr$670 $998 $1,386 $2,027 $2,981 $3,467 $4,149

(Tuition only)
All$257 $351 $433 $635 $1,045 $1,454 $2,057
4-yr -- -- --$804 $1,318 $1,888 $2,689
2-yr$109 $187 $245 $391 $641 $824 $1,194

Private Institutions

All$2,005 $2,738 $3,663 $5,470 $8,885 $12,910 $16,222
4-yr -- -- --$5,594 $9,228 $13,237 $16,645
2-yr$1,557 $2,103 $2,711 $4,303 $6,512 $9,302 $11,059

(Tuition only)
All$1,154 $1,684 $2,272 $3,498 $5,789 $8,772 $11,128
4-yr -- -- --$3,617 $6,121 $9,083 $11,522
2-yr$768 $1,109 $1,427 $2,413 $3,672 $5,570 $6,865

Ratio, Private to Public

All2.0 2.1 2.2 2.3 2.5 2.7 2.7
4-yr -- -- --2.2 2.4 2.5 2.5
2-yr2.3 2.1 2.0 2.1 2.2 2.7 2.7

(Tuition only)
All4.5 4.8 5.2 5.5 5.5 6.0 5.4
4-yr -- -- --4.5 4.6 4.8 4.3
2-yr7.0 5.9 5.8 6.2 5.7 6.8 5.7

Source: Author, from NCES data. Public institution tuitions are for in-state students.

Factors Affecting the Demand for Higher Education

In order to craft effective proposals for reforming higher education finance, it is important to understand the characteristics of the higher education marketplace. Central among these characteristics is the nature of the demand for higher education: who is in the market for higher education, and how do they respond to tuition levels and other factors? Without an understanding of the nature of higher education demand, it is not possible to predict the effects of changes in higher education policy and finance.

Underlying Demographic Changes

The starting point in examining the demand for higher education services is the trend in the demographic characteristics of the population. Higher education seems to be of interest to two key segments of the population: the recent high school graduate and the older, returning or continuing student who failed to acquire higher education when younger. Although national annual enrollments of men under 25 have been relatively steady at 3.5 to 3.8 million per annum between 1970 and 1993, the enrollment of women under 25 increased by 60 percent to 4 million over that period, and the enrollment of women over 25 increased by over 350 percent to 3.5 million; the enrollment of men over 25 has grown less rapidly over the same time period to only 2.3 million. Consequently, women are currently the largest (55 percent) and most rapidly growing gender in terms of enrollment. The traditional, under 25 segment is larger (at 56 percent of total enrollments) than the continuing or returning student segment, but it is growing less rapidly.

The growth in enrollments has been particularly marked for part-time students, especially women students. (See Table 6.) This trend is a combined reflection of the returning student phenomenon and, again, the female enrollment phenomenon. Private 4-year institutions have experienced very rapid increases in both full-time and part-time enrollments of women at 78 percent and 155 percent, respectively, between 1970 and 1993. Enrollments at 2-year institutions, public and private, have grown even more rapidly than 4-year enrollments, though public institutions have had significantly higher enrollment (at 143% and 84%, respectively, for the same time period).

Going forward, the growth in enrollments by older students can be expected to slow since a smaller proportion of the population will be reaching age 25 without having experienced college enrollment. Nevertheless, in the short run, demand for college education will be driven, in scalar terms, by women, and continuing women, students. Looking farther out, the relatively low elementary school enrollments of the early and mid-1980's will be translating into a relative slowing of college enrollments. The classes enrolling in 1998 through 2005 have the greatest potential of being affected by the "bust" in kindergarten enrollment in the 1980s. Oregon is projected to have a ten-year growth of high school graduates of only 20 percent between 1994-95 and 2004-05,[18] though the slowing of college enrollments likely will be offset partly by the effects of recent in-migration (much of which has involved families with young and teenage children who will be demanding higher education in the next decade) and by increases in female enrollment.[19]

Table 6: Enrollment Growth, 1970-1993, by Sex and Enrollment Status

1970 1980 1990 1993 % Chg


8.58 12.10 13.82 14.31 67%

5.82 7.10 7.82 8.13 40%

Men 3.50 3.69 3.81 3.89 11%

Wornen 2.31 3.41 4.01 4.24 83%

2.76 5.00 6.00 6.18 123%

Men 1.54 2.19 2.48 2.54 65%

Wornen 1.23 2.81 3.52 3.64 197%

4-yr Total

6.26 7.57 8.58 8.74 40%

4.59 5.34 5.94 6.08 33%

Men 2.73 2.81 2.93 2.96 8%

Wornen 1.85 2.53 3.01 3.12 68%

1.67 2.23 2.64 2.66 59%

Men 0.94 1.02 1.12 1.12 20%

Wornen 0.74 1.21 1.52 1.54 108%

Source: Author, from IPEDS data.

The implication of these demographic trends is that higher education systems, in Oregon and elsewhere, face long-term but slower growth, with considerable near-term oscillation between periods of rapid and slower growth. Any reform of the financing of the public higher education sector must take this volatility into account, and be able to adjust to the pending demographic cycles

The Demand for College Enrollment

A key aspect of the demand for higher education is its sensitivity to income and price. Whether or not one subscribes to the "social returns" argument for subsidizing higher education, subsidy does not make sense unless it can be demonstrated that the subsidy actually stimulates additional higher education investment. In the parlance of economics, the justification for subsidizing education requires that college enrollment be sensitive or "elastic" with respect to both the income of the student and the price of education (i.e. tuition).

A very large number of studies have estimated the income and price elasticity of demand for a college education. In most of this research, demand for college enrollment is in general moderately sensitive to family income, and somewhat less sensitive to price. In addition, the sensitivity to price and income vary not only with the academic ability of the student, but also with the characteristics of the institution of higher education.

The elasticity estimates vary somewhat from study to study[20], but show generally that a ten percent increase in household income can be expected to increase the demand for college enrollment by approximately ten percent. In contrast, a ten percent reduction in tuition can be expected to increase the quantity of enrollments by only about one to five percent. In economic terms, the income elasticity is moderately high, suggesting that the demand for education is fairly sensitive to the pattern of household income, and the price elasticities are low. Nonetheless, responses to enrollment can be expected from tuition increases or decreases.

Variations by Income and Ability

The elasticities vary, however, depending upon other characteristics of the student or the school. An early, but respected study by Hight [21] found, for example, that an increase in average household incomes tends to increase the demand for private education by 70 percent more than it will increase the demand for a public education. Economically, the market perceives private higher education as a "superior good", consistent with private institutions' efforts to differentiate themselves by quality. It also suggests that, absent scholarship or other policies, there would tend to be a sorting of students, by income, between public and private institutions, with public institutions tending to enroll students from somewhat lower income households. In general, this sorting does not occur, however, because of scholarship and other policies; indeed, in Oregon at least, the median income of students at private institutions is lower than that at the 4-year public institutions of higher education in the state.[22]

Economists also have detected differences in response to tuition levels among students of different economic and academic backgrounds Bishop found, for example, that students from high-income households or with high academic ability were much less tuition price sensitive than students from low income or low ability households. The elasticities differed by a factor of four in the case of income, and a factor of ten in the case of ability.[23] Increases in overall tuition levels will tend to more significantly reduce the enrollment tendencies of low income or low ability students than high income or high ability students. In other words, if the public sector has finite resources to underwrite the higher education of its population, the funds will have the greatest impact on enrollment if targeted at low-income and low-ability students.

Differentiating tuition on the basis of need is a strategy that private institutions have used successfully to maximize the enrollment of their targeted student populations. The public sector generally utilizes little in the way of need-based assistance, with the result that very large subsidies are conferred on students whose incomes or native scholastic urges would be predisposed to enroll in any case. In Oregon, for example, only about $14 million of the $850 million state higher education budget is used for need-based tuition waivers at the state system (and $4 million at private institutions). The vast majority of the subsidy to the public higher education system goes to underwriting the tuition of individuals who do not need a tuition incentive to seek out higher education.

Whether the State should seek to encourage enrollment in higher education by low-ability students is an issue that cannot be addressed here, since it depends on the balance of educational outcomes that results, for all students, by having low-ability students enrolled with high-ability students. All that can be said here is that the current policy of subsidizing tuition at public institutions has the effect of increasing the enrollment of low-aptitude students in the university system.

Variations by Age and Sex

The different age segments also respond differently to their income and tuition prospects. Corman's seminal study in the 1980s, for example, found that students over 25 years of age were nearly twice as sensitive to tuition levels as students under age 25, and that higher family incomes tended to reduce the likelihood of college enrollment by older males, but increase the likelihood of college enrollment by females.[24] This pattern likely reflects the different historical earnings patterns of men and women, and the relative difficulty women experience interrupting other activities for higher education. This will likely change as increases in female labor force participation increase their average contribution to family income.

The Demand for Vocational School Enrollment

The same study measured the effect of income and tuition levels on attendance at vocational and technical schools. This is a very important aspect of Corman's study, because whether higher education finance reform targets vocational and technical schools or not, it will have an effect on those schools. As stated earlier, it is difficult to compare enrollment measures across collegiate and non-collegiate programs. In addition, in states like Oregon, public community colleges offer curricula that competes with voc-tech institutions, blurring the distinction between collegiate and non-collegiate training. Using the available measures, however, approximately 25 percent of enrollments nationally are enrollments in vocational and technical schools (the percentage is virtually the same in Oregon).[25] Hence, the vocational and technical school segment represents a quantitatively important component of the higher education market. In addition, since the vast majority of voc-tech enrollment is in private institutions (100 percent in Oregon), the relative treatment of the public and private sector also is important in this market.

In general, the vocational and technical school enrollment rate displays sensitivity to voc-tech tuition levels that are similar to those observed in the market for college enrollment, for both men and women. However, higher family income has a strongly negative effect on voc-tech enrollment rates; indeed, the income effect is essentially as strong (but of opposite direction) as the effect of income on college enrollment. Enrollment in vocational and technical schools clearly is of greater appeal to households with lower income, perhaps because of the generally shorter financial and time commitment to voc-tech education.

Interestingly, Corman also measured the effect of college tuition on voc-tech enrollment, and the effect on voc-tech tuition on college enrollment,[26] permitting some further insights into the interdependency of the policies in these two markets.[27] Her results suggest that, for younger students, there appears to be a clear preference for college education, and voc-tech school enrollment rates are affected significantly by college tuition policy, but not vice versa. Among older households, there does not appear to be as clear a preference for college education. For both males and females, voc-tech enrollment and college enrollment are each sensitive to the tuition levels at the other, although a slight preference persists for college enrollment, and women appear to have a stronger preference for college enrollment over voc-tech enrollment.

The implications of Corman's analysis for reform of higher education finance are important. Her analysis suggests that public policies that have the selective effect of reducing college tuition levels will tend to lower voc-tech school enrollment by a greater percentage than they will increase college enrollment, especially among younger students. Further, to the extent that higher education policy fails to address the issue of access to college by lower income individuals, the demand for voc-tech training will be increased. These are, perhaps, obvious phenomena, but Corman's study is important because it demonstrated that the magnitude of the effect is likely to be quite strong.

Race and Ethnic Aspects of Higher Education Demand

Oregon's population is largely white, with non-white racial minorities totaling only about seven percent of the total population. Quantitatively, therefore, Oregon will have less difficulty providing higher education services equitably to blacks, Hispanics, and other minorities than many other states. Nonetheless, it is important to determine whether there are special issues surrounding minority demand for higher education services that are not already accounted for by family income, educational attainment, and other factors common to all race and ethnic groups. In many places in Oregon, the minority population is growing more rapidly than its non-minority population, so the overall effect of minority educational attainment will affect the future workforce composition significantly.

The most prominent facts regarding minority enrollment in higher education concern the recent changes in the relative enrollment rates among whites, blacks, and Hispanics. Specifically, although enrollment rates have been increasing generally for all three groups, the relative enrollment rates of blacks and Hispanics (relative to whites) has declined. For example, although the percent of black high school graduates who enrolled in college rose from 45.7 percent in 1978 to 50.9 percent in 1994, the percentage of whites enrolled rose over the same time period from 50.1 percent to 63.6 percent. Hence, the relative rate of black enrollment declined from 91 percent of the white rate in 1978, to 80 percent in 1994. Similarly, the Hispanic enrollment rate, though increasing from 46.3 percent in 1978 to 55.5 percent in 1993, suffered a decline in relative enrollment rates with respect to whites from 92 percent of the white rate to 88 percent in 1993 [28]. The question of importance to higher education policy makers is whether this relative decline can be explained by changes in relative incomes, etc, or whether there are group-specific factors at work as well.

Recent research by Hauser,[29] using Current Population Survey data through 1988, isolated the group or race/ethnic effects on college entry rates from changes in students' general economic and other family circumstances that may have affected the groups differently.[30] Some of the non-group specific factors considered by Hauser include the age of the student, dependency status, family income, family size, family type, education attainment of the parents of the student, and various metropolitan and regional indicators. These background conditions vary fairly significantly across the three groups. For example, in 1988, nearly 50 percent of black households were female-headed, versus 35 percent for Hispanics, and approximately 20 percent for white households. Similarly, mean years of schooling of the mother in Hispanic households was only 10 years, versus 11.5 years for black households and about 13 years for white households.

After statistically controlling for these factors, Hauser finds that most of the difference in college enrollment behavior between blacks and whites disappears, and that black enrollment rates, in effect, never fell far below the rates of whites. When the background of Hispanics is controlled, in fact, the Hispanic enrollment rate actually exceeds that of whites, a phenomenon attributed to the relatively more selective nature of Hispanic college applicants[31]. Hauser concludes from his analysis that, after controlling for social background, trends in net tuition costs are the most likely factor explaining recent, declining relative enrollment rates of black students. Once again, therefore, of tuition policy emerges as an important aspect of the reform of higher education finance.

Factors Affecting the Supply of Higher Education Services

We turn now to the supply side of the higher education market. In particular, we are interested in understanding how the institutions themselves might respond to changes in the mechanism of higher education finance.

It is much more difficult to study the supply side of the higher education market than the demand side. By the very nature of these institutions, change occurs slowly, and it is difficult to definitively associate particular economic or social factors with these changes. In addition, as we will see, the organizational structure of colleges and universities is very unlike the model of business enterprise that underpins the economics of industrial organization. This makes it difficult to predict the behavior of higher education institutions.

Basic Characteristics of the Higher Education Industry

As of the academic year 1994-95, there were 3,688 total institutions of higher education in the United States. Of these, 1,641 are public institutions (or branches), and 2,047 are private institutions. In Oregon, there were 21 and 23, respectively, in these categories. At an organization level, therefore, Oregon has a higher percentage of public institutions than the average state, and it has no large, private universities. Approximately 1,470 of the total institutions in the United States are two-year institutions, 30 percent of which typically are private. In Oregon, in contrast, there are 14 such institutions, but only one is private.

The smaller number of public institutions conceals the dominance of the public sector in the higher education marketplace. As of 1993-94, over 78 percent of all enrollments nationally were in public institutions. The dominance is sharpest in the two-year college segment, in which the public sector serves over 95 percent of all enrollments; of institutions offering 4-year degree programs, 67 percent of all enrollments are at public institutions. Oregon has even higher public enrollments, at nearly 90 percent overall, 99 percent at 2-year colleges, and 76 percent in four-year undergraduate programs.

The mix of public and private institutions across states is highly variable, as indicated by the wide variation in relative public and private enrollments cited earlier. The fact that such variation exists suggests that the mix of public and private institutions is a matter of political preference and historical accident, and not because marketplace economics predispose any given level of involvement of the public sector. Oregon, with a public/private enrollment ratio of 7.6:1 is very near the extreme of preferences for public provision, suggesting that political forces and the relatively limited historical size of the private college sector have conspired to produce a greater reliance on the public higher education institutions.

Similar choices are implicitly reflected in the relative levels of provision of four-year versus two-year public education. With almost two community college enrollments for every four-year enrollment, Oregon has the seventh most intensive community college program in the country. (California, at a ratio of 3:1, has the most intensive, and South Dakota and Alaska, with ratios of 1:100 and 2:100 respectively, the least intensively developed community college systems.)

Clearly, there is no single formula for public involvement in the production of higher education services. What is clear, however, is that Oregon appears unusually committed to the public production of higher education services.

The Nature of Competition in the Higher Education Marketplace

Economists use their understanding of the competitive forces in a marketplace to predict the response of that marketplace to changes in demand, public policy and so on. Market behavior for most goods and services is well articulated and can be used to easily predict the effect of changes in the market context. The market for higher education, however, displays characteristics that are atypical of the typical market. Consequently, policies that might be expected to work in conventional market settings may not work at all, or may work perversely, in the market for higher education services. From an economist's perspective, there are four particularly important features of the market for higher education that must be factored into thinking about financial reform.

First, in neither the public nor the private sector of higher education is the revenue collected from tuition equal to the full cost of producing those services. In 1992-93, the average total cost nationally of producing a private education in a four-year institution was $25,400 dollars per year per student, while the average tuition, room and board fees, at $14,600 was only 60 percent of that; at public institutions, cost was $17,900 and tuition, room and board fees averaged $6,000, or only about one-third of the total cost. (Oregon costs and fees were slightly less than the national average, but in a similar relationship to one another.)

At private institutions, the difference is made up from endowment and grant income. (In 1993-94, there were $3.4 billion in alumni giving alone, which represented only a little more than 25 percent of all giving.) At public institutions, the difference is made up by tax revenues collected from the general public. In both cases, the services provided to the current generation of students is subsidized by either a different generation of students (in the case of private institutions) or the general taxpayer. The existence of this behavior suggests that the market for education services has some of the features of a club, rather than a typical business enterprise. That is, people other than the students themselves perceive benefits to the higher education of others, perhaps as a way of preserving a lifestyle or subcommunity from which they have benefitted. In any case, predicting the economic behavior of clubs and other such altruistic enterprises is more difficult than predicting the behavior of the typical business.

Second, the published tuition charges are seldom paid by any given student; rather, most private institutions implement systems of scholarships and other financial aid. This aid is typically provided to students with weak economic backgrounds (for whom the price elasticity of demand is generally very high), or whose skills (scholastic or athletic) cause them to be more aggressively competed for by a large number of institutions (thereby making their demand for any one institution more price elastic). Although frequently described as an altruistic behavior in support for diversity, such practices are equally well explained as a rational form of price discrimination, whereby students with relatively price-insensitive demand for higher education (e.g. well-to-do students) are charged more than students with more elastic demand (e.g. poor students or especially-competitive scholars and athletes).

The practice of price discrimination also suggests certain things about the cost structure of higher education enterprises. Price discrimination is most successful as a net revenue-maximizing strategy when the cost of producing another enrollment slot is small relative to the average total cost of production -a condition that is likely to prevail when there are economies of scale. In such a case, even if charging a uniform, average cost price would not bring revenue to support the enterprise, price discrimination can be used to enlarge output (i.e enrollment) while still operating a solvent enterprise. Hence, the price discriminating behavior of higher education institutions implies not only that there are different elasticities of demand among student clienteles, but also that there are economies of scale and high average costs relative to the marginal benefits of higher education at observed enrollment levels.

The existence of economies of scale in higher education supply have been confirmed by direct study.[32] In addition, the existence of economies of scale, in turn, implies that large enterprises will tend to dominate small enterprises in the higher education marketplace. The market does, indeed, display fairly high levels of concentration. At both public and private institutions, for example, the largest 15 to 20 percent of the institutions enroll over 60 percent of the students. However, the private higher education sector has many small enterprises; over 60 percent of private institutions, for example, have 1000 or fewer students (serving 14.5 percent of enrollments). In contrast, only 10 percent of public institutions are that small, and they serve only one percent of total enrollments.

This diversity of institution size, within the general context of a concentrated industry, results from the fact that size itself is a differentiating feature of the quality of higher education. Whether it is because of variations in student preferences for school community size, or because of variations in perceived relationship between institution size and educational quality, relatively cost-inefficient enterprises are able to coexist along side institutions able to exploit scale economies. The implication of this is that policies that have the effect simply of producing enrollments at the lowest unit cost may have adverse effects in other dimensions. Hence, at least from this narrow perspective, increasing student sovereignty in the marketplace would be expected to yield a preferred educational product.

Third, the quality and quantity of services provided by a particular institution are difficult to assess until they have actually been experienced, at which point it is relatively costly to switch to a different provider in the hope of improving the services received. As a consequence, economists expect to see individual colleges and universities expending resources to "signal" the quality of the services they provide, much as banks historically have expended resources on lavish branches and facilities in an effort to convey safety and soundness. The tendency to develop elaborate campuses and research facilities, promote institutional reputations through sport programs, and by attracting the most qualified student bodies can be seen as signaling exercises. Some of expenditures of this type obviously actually do improve educational quality; certainly many aspects of academic research, for example, improve the quality of teaching and learning. Similarly, having a high quality student body likely improves the education experience, by mutual association, of other students.

It is also likely, however, that some of the signaling expenditure is wasted. That is, some of these expenditures do not actually improve educational quality, any more than the money expended on the marble columns of a bank improves the quality of the banking services themselves. Indeed, it has even been argued that in a marketplace in which quality is difficult to measure, it may be logical to create high costs and/or advertise high tuition levels as way of signaling quality. Viewed from this perspective, increasing the general level of resources available to institutions can be expected to result in increases in both productive (i.e. quality-enhancing) and "wasteful" (i.e. pure signaling) expenditures. Consequently, reform needs to be focused on mechanisms that are more likely to stimulate quality improvements rather than pure signaling expenditures.

A fourth unusual aspect of the behavior of the higher education marketplace likely is that management authority is very diffuse. Unlike enterprises elsewhere in the economy, the chief executive office of most colleges and universities shares decision making authority with faculty. In effect, decisions about the allocation of a university's resources tend to be made not just with the needs of the student in mind, but also to serve the preferences of faculty and staff.

This behavior would not persist in a typical, competitive marketplace, because the associated waste of the enterprise's resources would ultimately make it unprofitable, and unable to compete successfully with its rivals. This, too, suggests that competitive forces may be insufficiently strong to protect against wasteful behavior of the institution. One possibility is that blanket assistance to educational institutions will be dissipated, at least partly, on perquisites of faculty and staff. Indeed, in the 1980s, for example, average faculty compensation levels rose, in inflation-adjusted terms, and average teaching loads[33] declined at a time that has been otherwise characterized as financially difficult for universities and students. This is at least suggestive that some management factors limit successful cost cutting or quality improvement in higher education institutions, especially at public institutions.[34]

Finally, it should be noted that the higher education industry is one of the few sectors in the economy in which public and private producers coexist. Although quantitatively minor in terms of enrollment, the private sector institutions are highly regarded and, some would say dominate, in the provision of high quality education at both graduate and undergraduate levels. Given the formidable price competition offered by publicly-operated institutions, of course, quality of education and the on-campus experience is the only dimension left to the private sector to differentiate itself.

In addition, the far greater risk of institutional failure undoubtedly stimulates more entrepreneurial and market-responsive behavior on the part of private institutions. Over a thirty-year period, the closure rate of private institutions has averaged ten times that of public institutions; since 1975, the rate has accelerated, with over 160 closures of private colleges and only two closures of public institutions (both of which were two-year institutions). Given these higher failure rates, it is no surprise that innovations such as returning student programs, evening degree programs, and other student-oriented differentiations have tended to originate in the private sector.

In summary, the nature of the higher education enterprise, and the market in which it operates, suggests that there will be benefits to increasing competition through student-based assistance programs. As one analyst observed:

" . . . quality is a problem in public higher education because student demands matter too little, and cost is a problem in private higher education because student demands matter too much."[35]

By giving greater power to the consumer (student) in the higher education marketplace, the public institutions will be encouraged to provide a better educational product, while increasing the pressure on private institutions to provide a cost-effective, as well as high-quality, educational product.

The Impact of Government Aid to Institutions and Students

The general observations above about the economic structure and performance of the higher education industry have important implications for the efficiency and effectiveness of government aid. A few economists have argued that the inefficiencies of the market for higher education services warrants withdrawal of subsidies to higher education.[36] This author does not share that view because imperfections in capital markets alone suggest that a completely unassisted higher education market would be economically inefficient. Specifically, there would be an inefficient quantity of higher education consumed by lower-income individuals if some form of assistance in defraying education costs were not available. Moreover, assuming that the observed tendency for society to subsidize higher education persists (right or wrong), it is important to speculate on the appropriate mechanism for that subsidy if the aim is to enhance educational attainment.

To date, the main forms of subsidy have been threefold: (1) general subsidies, through private endowments or government tax revenues, provided to the institution, (2) student loan programs, and (3) scholarships and other aid provided to the student. How significant have these programs been, and what has been the response of the higher education marketplace to this financial assistance?

The Current Structure of Higher Education Financial Assistance

Disentangling the current structure of subsidies to higher education is difficult, because of the way the statistics are kept, and because subsidies can take the form of direct grants, grants through students, payments through research and work-study programs, and so on. Data on some of these latter programs is available only by survey of the student, and dollar amounts are unreliably reported and often do not correspond with known program totals. In Table 7, the various sources are aggregated to provide a general picture of the pattern of higher education subsidies, and the relative roles of the various fund sources.

Federal grants and loans are the most important aid sources directed at students. (The total of scholarship and fellowship support is less than either of these. However, tuition waivers and other such quasi-scholarship adjustments may not fully appear in the available statistics.) The most important sources of non-institutional student aid are the various student loans; most grant sources are paid to the institution rather than the student despite being considered student aid.

General Aid to Public Universities

The term general aid assistance is used here to mean budgetary aid to institutions that makes it possible for institutions to charge an average tuition price that does not cover average total costs. General subsidies to higher education currently are the quantitatively most important type of subsidy. Of the $55 billion in state, local and federal aid to institutions of higher education, over 85 percent is provided directly to the institution in the form of general aid, with the remainder provided through Pell-type grants, and other portable and institution-specific aid to students.

Table 7: The Pattern of Public Assistance to Higher Education, 1992-93

Public Private All

Sources of Institutional Revenue ($B, percent of total)

Tuition and fees

$19.5 18% $25.8 41% $45.3 27%
Federal government

$11.7 11% $9.3 15% $21.0 12%
Memo: Pell Grants
$4.7 3%
State governments

$39.8 37% $1.4 2% $41.3 24%
Local governments

$4.0 4% $0.4 1% $4.4 3%
Private gifts, grants
$4.3 4% $5.3 8% $9.7 6%
Endowment income
$0.6 1% $2.9 5% $3.6 2%
Sales and services

$25.3 23% $14.5 23% $39.9 23%
Other sources

$2.9 3% $2.8 4% $5.7 3%

$108.2 100% $62.7 100% $170.9 100%

Aid to Undergraduate Students ($B, percent of students receiving)


na 29% na 57% $15.30 37%

na 19% na 36% $8.20 23%
na 18% na 34% $4.70 21%
Supp. Educ. Opp'ty
na 4% na 10% $3.50 5%

na 15% na 42% $12.70 20%

na 14% na 41% $12.30 20%
Direct student loans na 2% na 5% na 3%
Guar. Student Loans na 13% na 38% na 18%
Parent loans for students na 2% na 3% na 3%
Supp. loans for students

na 1% na 7% na 2%
Work study

na 3% na 11% $1.20 5%

na 5% na 10% $2.40 6%

Memo: Scholarships, etc. $3.70

Source: Author, from various U.S. Department of Education sources.
The scholarship estimate includes Pell grants.

Revenue obtained through generalized, organizational subsidies have very different incentive effects than revenue obtained through "sale" of products, such as educational services. Economic theory predicts, for example, that some portion of such subsidies would be dissipated on so-called X-inefficient behavior of the enterprise -essentially, indulgence of the perquisites of the management and staff of the enterprise, rather than the quality or cost of the service. Although some offsetting incentives can be built into the granting mechanism (such as making the subsidy capitation- or enrollment-based), it is difficult to craft an incentive formula that replicates consumer incentives.

The aid formulae employed in most states thus do little more than prorate aid to the size of the institution. Hence, they would not be expected to lead, necessarily, to increases in instruction expenditures of direct benefit to students, per se. Rather, some of the subsidy would be expected to be spent on expenditures of interest to faculty and staff, rather than students. This intuition is borne out by a variety of studies of the effect, mostly in the public sector context, of increased subsidies. These studies indicate that only about one-third of each dollar of direct appropriations to public institutions goes to instruction-related expenditures[37]. Clearly, expecting increases in generalized aid necessarily to increase the quality or quantity of instruction is unwarranted.

These issues are not unique to public institutions; although public institutions receive generalized aid through general tax sources, private institutions receive similar aid from alumni, corporations, and other private donors, in addition to some public aid. Disentangling the incentive effects of private gifts can be particularly difficult because they are more likely to have specific restrictions that vary considerably from gift to gift.

Institutional vs. Student-Based Aid

Direct subsidies to higher education institutions are frequently justified as egalitarian. That is, by subsidizing the operation of public universities, a broader range of individuals can find their way to higher education. It is far from clear that the actual redistributive effect of tax-financed educational institutions is egalitarian, however, since the utilization of public university services is often more progressive with income than the effective federal and state tax systems that provide the appropriation. Hence, the motives for providing generalized subsidies to higher education institutions may originate partly in intergenerational "club" incentives, whereby each subsequent generation agrees to underwrite the education of the younger generation as a way of preserving an opportunity for their children or others in their social or economic class to pursue a lifestyle that they feel benefitted them.[38] In this view, subsidizing institutions, rather than individuals, has a certain elitist or establishmentarian aspect to it.

More importantly, the lesson from the earlier theoretical discussion and the observed competitive response of the private university sector, is that higher education institutions will compete and innovate in response to student demands. Superficially, at least, this suggests that institutions who need to attract students to survive financially are likely to respond by providing (or signaling that they provide) high quality education experiences. From this perspective, providing financial assistance through students (rather than through the institution) should stimulate, on the part of college and university managers, increased emphasis on student educational outcomes rather than outcomes that benefit faculty or staff. A wide range of studies confirms that student grants stimulate enrollment (particularly by lower income individuals), enhance persistence, and sharpen school choice decisions,[39] as this view would suggest. Studies of the largest applications of student-based financing of higher education, the GI Bill, also has illustrated that vocational education vouchers were superior to publicly-subsidized training programs in elevating the earnings of those who received them.[40]

The only risk to this conclusion is that some resources may be dissipated by signaling (but not necessarily providing) quality if quality is difficult to discern prior to enrollment. Hence, systems to measure improvement and reporting of educational outcomes, and perhaps some linking of subsidy mechanisms to institutional performance, may be justified to limit this effect.

Merit versus Need-Based Assistance

Making aid available to students, rather than institutions, raises the issue whether such aid should be merit-based, need-based, or a simple, flat amount per student. Simplistic arguments can be made in favor of each: merit-based aid rewards prior and continued achievement, and thereby stimulates achievement; need-based aid is egalitarian, and gives students equal access despite family circumstances; and the single, flat amount per student is fair because need-based structures would unfairly compound the progressivity of the tax structure with further progressivity in distribution of tuition assistance.

Each of the arguments has an element of truth, but oversimplifies a complex issue. From an economist's point of view, a system of higher education finance must balance at least three goals: (1) overcoming the market imperfections that limit students' ability to borrow against their future income; (2) encouraging efficient behavior of individual educational institutions, and (3) doing the above in the manner least burdensome to taxpayers. If one believes strongly in the social returns argument, a fourth goal, of stimulating educational attainment above the privately-desired level, might be added.

Viewed from this perspective, merit-based assistance has little to be said for it as a general policy. First, if most of the benefits of education are private, achievement will be adequately stimulated by private motives. Second, those with high achievement records are relatively insensitive to the price of higher education, and thus assistance is less necessary to assure their attendance. Finally, although the presence of higher quality students likely improve the quality of education at any given institution, it is unclear how this fact would be implemented in student assistance policy. Would higher merit students be paid more to attend institutions without many other such students (as a way of improving the school's quality)? The incremental benefit to school quality is likely to be low relative to the loss of opportunity to the individual, high-merit student. Although competition for high quality students among individual institutions will (and does) occur, it is not obvious that a universal policy of merit-based assistance has a basis in public policy except, perhaps, to stimulate better performance in secondary education.

In this author's view, equally little can be said for the policy of a uniform subsidy per student. The basic market failure in the higher education market -that is, the lack of a good mechanism for borrowing against future income- is that it mostly burdens students from low-income households. Individuals with high accumulated family wealth and income are less likely to be constrained by this market failure. Hence, some progressivity in student assistance with family wealth and income is justified on economic grounds, unless this market imperfection is resolved in some other way. In addition, the policy of uniform subsidy is fiscally inefficient. Any given enrollment level can be achieved more economically by price discriminating (i.e. charging in inverse relationship to the price sensitivity of demand).

On these grounds, needs-based assistance seems to make sense as the most efficient policy, quite apart from whether it also may be interpreted as more equitable. Not surprisingly, this is also the mechanism most widely used in the private sector of higher education. The policy has permitted that sector not only greater solvency than it otherwise might have enjoyed, but also resulted in diversity in enrollments, by family income, that matches (at least in Oregon) that of the public system.

Student Loans vs. Grants

Not surprisingly, surveys of student reaction to grants versus loans suggest that grants are more stimulative of enrollment and persistence in enrollment than the availability of loans.[41] This is simply another manifestation of the price-sensitivity of higher education enrollment. A grant is tantamount to a dollar-for-dollar reduction in tuition costs; even a subsidized loan, with extended repayment terms, is tantamount to only a partial reduction in tuition costs.

The theoretical argument in favor of grants is that historically, the private loan market has provided little access to individuals who do not have assets with which to secure their borrowing. Even if such access is available, students have relatively limited ability to forecast their future income with any precision. The uncertainty of future income, coupled with risk-aversion results in "underborrowing", and underinvestment in education.

Both of these issues can be resolved with proper public policy and, to some extent, have been addressed in recent years with programs for government-guaranteed lending. It is true that private lenders have relatively limited ability on their own to collect on delinquencies in relatively small loans in a cost-effective manner. (The average undergraduate loan outstanding at graduation is $11,000.)[42] As a result, conventional lenders historically have not found student lending to be profitable except at lending rates that likely exceed the private IRR of higher education. A government agency, with access to a tax system with strong, inherent garnishing capabilities has the theoretical potential of lending at lower rates because of lower collection costs. This simultaneously addresses the access issue, and reduces the import of uncertainty about future incomes.

Student borrowing through government-guaranteed or assisted lending programs exceeded $25 billion in 1995, three times the level of ten years earlier. Perkins loans (formerly National Student Direct Loans) carry an interest rate of 5 percent as of 1996, and Stafford Loans (formerly Guaranteed Student Loans) a rate of 8.25 percent, with a lower accrual rate while the student is in school or when deferred for unemployment or other qualifying event. Hence, federal student loan policy has evolved in directions that should help resolve the credit market failure that affects higher education policy. However, federal lending programs remain capped at relatively low levels ($2,625 for freshmen, rising to $5,500 for seniors). This leaves room for innovative state policy to (at least partly) replace appropriations to the Oregon higher education system with expanded loan programs, if the management of delinquencies can be done economically.

Stimulating Institutional Efficiency

The most difficult question to address is the issue of institutional efficiency. It is clear theoretically that a system that finances institutions, rather than individual students, is likely to be dominated by X-inefficient organizations with somewhat muted focus on student achievement. To the extent that student-based subsidies stimulate competition among educational institutions, they may also stimulate wasteful signaling, rather than true education expenditures.

Two approaches to resolving this issue present themselves. One is to improve the measurement and dissemination of "accurate" educational outcome information, so that signaling is less necessary and less effective. Precisely how this would be done is not clear; measurement of educational outcomes is notoriously difficult (hence, the signaling phenomenon), but there may be ways to improve such measurement that have not been undertaken because individual institutions have an incentive to not be compared on equal grounds.

The second method is to link student support levels to some gross indicator of educational quality, such as the percentage of the budget devoted to education services. Though crude, and though it runs the risk of stimulating spending, but not outcomes, it may be sufficient to mitigate wasteful spending on signaling.


A Proposal for Reform: The Implications for Oregon

The review of our understanding of the demand for higher education, and the behavior of suppliers of higher education leads to several obvious conclusions:

1. The sensitivity of demand for education to price is not so high that a greater reliance on tuition funding would seriously impair enrollments in colleges and universities. It would be unlikely to impair at all the college enrollment rates of students from high income families, and students with high ability;

2. The arguments in favor of stimulating higher education beyond that justified by private returns' calculus are weak. Overall, therefore, greater reliance on tuition as a source of revenue for higher education is justified;

3. Students from poor households are the most constrained by the failure of the marketplace to provide a means of borrowing against future earnings. Consequently, they are much more sensitive to tuition levels. This factor alone suggests that a need-based tuition subsidy mechanism should be employed to differentially lower tuition costs for poor students, or an inefficiently-low number of such students will enroll in colleges and universities.

4. The nature of the primary market failure in education finance (i.e. the difficulty in borrowing against future income) suggests that public intervention should primarily be directed toward addressing this failure. That is, some of the funds currently used to provide grants or institutional subsidies might be more appropriately and economically used to underwrite student loan risks and to develop efficient methods of collecting on delinquent loans.

5. The nature of the higher education product is that competition among institutions can result in resources being used to "signal" quality characteristics that may, in fact, not improve the quality of educational services. As a consequence, general subsidies (to public universities through tax appropriations, or private universities through endowment donations) likely are at least partly dissipated on expenditures that do not improve instruction and education, but do increase overall education costs. This finding suggests that a mechanism must be developed to link outside financial assistance to spending on education, rather than signaling activities;

6. The lack of reliance on tuition for revenue at public universities results in reduced attention to student needs and, as a result, relatively inferior instructional quality on average. A symptom of this phenomenon is the overly important role of faculty and staff in university decision making, leading to lack of focus on student, versus self-serving institutional, goals. The converse is the case at private institutions, where the higher tuition costs relative to public institutions require aggressive differentiation from the public institutions by providing quality, or at least signaling that quality is provided. The combination of a relative lack of cost consciousness at public universities, and the need to compete via quality signals at private institutions, likely has contributed to the spiral of prices and costs in higher education This is consistent with the finding that university salaries generally have progressed at rates in excess of general rates of compensation, despite theoretical reasons why they should be expected to decline over time, in inflation-adjusted terms

Four recommendations to reform higher education finance

(1) Generalized public aid to public institutions of higher education should be largely eliminated, and replaced with aid to students attending Oregon institutions of higher education, public or private. To maximize the effectiveness of the aid, the amount of aid above a base amount should be need-tested.

(2) A student loan program will be more cost effective than outright tuition grants or vouchers. Hence, expanded student loan programs should augment or replace the function of outright student grants mechanisms. Effective implementation of this recommendation requires development of means for economical administration and delinquency management.

(3) To assist students in evaluating education quality at Oregon institutions and to encourage efficient competition among higher education institutions, the State should facilitate provision of cost, educational program and accreditation information to prospective students. Eligibility for receipt of student assistance grants or loans also should be withheld from institutions with low commitment to instruction and other activities directly related to education.

(4) As a result of the prior recommendations, the total non-tuition sources of assistance to higher education can and should be reduced. This should be accompanied by granting public university and college presidents greater authority for managing their institutions, including setting tuition rates and standards for admission.

Implementing the Proposed Reform in Oregon

The proposal above has features similar to those found in recent proposals in other states. In a special report to the Governor of Minnesota, for example, a blue-ribbon commission recommended "that the current practice of allocating 90 percent of the state's appropriation to institutions [of higher education] and 10 percent to students be nearly inverted."[43]

What would be the effect of such a reform on higher education in Oregon? In order to answer these questions precisely, very detailed modeling is required, an exercise beyond the narrow scope and resources of this paper. Nonetheless, rough indications of the effect can be obtained by using the results of other studies. In addition, the considerable variation in higher education sector characteristics observed across states can be used to infer how the public and private education marketplace in Oregon might respond to differences in public funding policy. The author used some of the empirical data cited in this report, coupled with a series of simple statistical models, to simulate the consequences of the reform.

Effects on College and University Enrollment Rates by Income Class

The proposed reform modeled here replaces institution-oriented support with tuition vouchers. As a starting point of analysis, it is helpful to think through the simplest, hypothetical reform. One such reform would be to apply the policy only to public institutions (i.e. the vouchers could be used only at public institutions) and to give each admitted student a voucher worth exactly the amount that the public system currently receives on average per student (approximately $5500 currently in Oregon). If public tuition, in turn, were grossed up by that amount, enrollment clearly would be little affected in the short run. Any long run effects would come from the competitive effects as the various public producers vied for students and their vouchers.

If, instead, the flat $5,500 voucher per admitted student was extended to include its use at private colleges and universities in the state, in the short-run an additional 12 percent increase in higher education support would be required. (This would be necessary to both maintain public sector enrollments and revenues, and provide the vouchers to the 12 percent of total enrollments currently represented by the private institutions.) Even though some substitution of private for public financing would occur in the private sector, more total revenue likely would available to the higher education sector under this scenario, and total enrollment would increase over time. In addition, more vigorous competition between the public and private sectors might be expected to increase quality or reduce price over time.

Finally, instead of a simple, flat voucher payment level, the payment could be set inversely with the likely tuition elasticity of demand of enrolling students. This has the effect of a price-discriminating tuition policy, similar to the one in place today at private institutions. Drawing on the estimated elasticity measures, by income level, available in the literature, this suggests that the effective tuition level might range by as much as a factor of 4-to-1 from the highest to lowest family income strata. Performing this exercise in the context of a demand model implies that either (a) the same level of aggregate subsidy could support over 30 percent more enrollments or (b) the same level of aggregate enrollments (and very similar shares, by income class) could be supported with 30 percent lower total costs.

Even greater enrollments (or lower total public costs) could be effected if the funds were used to underwrite student loans, rather than providing outright grants, in a progressive (i.e. needs-tested) way. To the extent that enhanced competition improves price-adjusted educational quality, or otherwise encourages institutional efficiencies, the effects in the long run could be greater still. Conversely, to the extent that the needs-testing is less aggressive, or competition involves costly signaling exercises, the fiscal efficiency and/or enrollment effects will be less positive.

Effects on Public and Private Institution Capacity and Finances

Predicting the effects on relative public and private university enrollment and finances similarly can be approached in a rough way. If the existing subsidy to higher education were shared pro rata with private institutions, public institution tuition in Oregon would have to increase, and perceived (net of voucher) tuition in the private sector would fall. As students shifted from public to private sector institutions in response to this price realignment, the posted tuition price at private universities likely would rise in face of the stronger demand. Evidence from cross-state comparisons suggests that somewhat less than half of the stimulating effect on private institutions would be taken up by posted tuition increases, with the rest serving to increase the market share of the private sector.

Precisely how large a market share change would occur is complicated by a number of factors. First, the capacity of the private sector to accommodate larger enrollments is greater in the long run than it is in the short run; indeed, one expected effect would be for the private sector to enlarge institutionally, either by new entry or by privatization of public institution capacity. A gradual phase-in of the program is possible simply by starting with a lower level of grants or loans. The consequences of the policy in the long run, however, can be inferred roughly by using data from studies of the cross-elasticity of school choice with respect to relative tuition. Calculating the relative tuition changes induced by sharing the existing state subsidy per student across all students, public and private, the ratio of public to private posted tuition would rise from 0.39 to approximately 0.95. A change of this extent would reduce the public to private enrollment ratio from 7.6 today to below one. In essence, the policy proposed here likely would reverse, significantly reducing Oregon's historic 88 percent public enrollment share.

There is no inherent threat to the viability of a public education sector of the policy proposed here. The public colleges and universities very well may continue to operate as public enterprises. Unless there is some residual advantage to being a public enterprise, however, it seems likely that the institutions may select the more flexible private form. In this scenario, although the role of the public sector in the long run would be much smaller, it is hardly uncharted territory; public systems of higher education on such scales exist in many places in the United States (indeed, in most places in the eastern United States). Even the transition to such a structure need not be particularly disruptive since the process would consist primarily of changes in ownership and management of facilities; although there would be some changes in program mixes, in overall terms, the public and private sector staffs, libraries, dormitories, and other facilities are likely to called on to provide similar services. Nor is there any inherent reason not to include vocational and technical schools in such a proposal. As earlier discussion in this report suggests, voc-tech schools are viable substitutes for collegiate education for certain students at certain times in life. Although there may be a logic to differentiating the stipend size somewhat to account for voc-tech's different products and inherent cost structure, the underlying rationale for assisting this type of educational product is essentially the same as for collegiate enterprises.

Implementing these outlined reforms in Oregon would reduce the total amount of government assistance to universities with no ill effect on the quality of education. Further, these reforms would increase the quality and quantity of education provided. They would also improve access to higher education among those Oregon families lacking private access to the necessary financial resources. Reform, rather than increases in higher education funding, improves the system overall and worth consideration in current and future budget debates.

About the Author

Randall J. Pozdena, PhD is Managing Director of ECONorthwest, an economics and finance consulting firm based in Portland, Eugene and Seattle. He is a former research officer in the Federal Reserve System, and a former professor of economics and finance at the Graduate School of Business, University of California, Berkeley and the Graduate School of Administration, University of California, Irvine. The views expressed in this paper are his own.


1. U.S. Department of Education, National Center for Education Statistics (NCES), Digest of Education Statistics 1995, Table 166.

2. J. Quigley and D. Rubinfeld, "Public Choices in Higher Education," in C. Clotfelter and M. Rothschild, eds. Studies of Supply and Demand in Higher Education, National Bureau of Economic Research Project Report (University of Chicago Press, 1994).

3. Public university enrollment divided by private university enrollment.

4. U.S. Department of Education, Office of Vocational and Adult Education, "Adult Education Program Facts, Program Year 1990-91."

5. Excellent meta analyses are presented in L. Leslie and P. Brinkman, The Economic Value of Higher Education (American Council on Education, 1993), Part II.

6. The estimated returns to graduate education have been much more variable, ranging from a negative 2 percent to a positive 20 percent. See Leslie and Brinkman, ibid.

7. Robert Spiegleman, "A Benefit-Cost Model to Evaluate Educational Programs," Socio-Economic Planning Sciences, Vol. 1, pp. 443-460.

8. I. Garfinkel, and R. Haveman, Earnings Capacity, Poverty, and Inequality (Academic Press, 1977).

9. The direction of causality is not clear. It may be the case that low-crime tendency individuals seek higher education, rather than higher education creating low criminality individuals. If this is true, of course, there may be no benefits in the form of reduced criminality.

10. See, for example, L. Olson, "The Economic Impact of Independent Colleges and Universities on Massachusetts in 1979-80 and 1980-81," (Data Resources Incorporated, 1981).

11. One of the earliest estimates is from G. Psacharopoulos, Returns to Education (American Elsevier, 1973).

12. T. Schulz, The Economic Value of Education (Columbia University Press, 1963).

13. Peonage was abolished with the 13th amendment to the US Constitution as a form of slavery because of the egregious abuses associated with debt peonage. However, the common 17th and 18th century peonage practices, such as apprenticing for fixed periods to masters, were the solution of that time to the problem of imperfect capital markets.

14. Quigley and Rubinfeld, ibid, p. 244.

15. NCES, ibid, table 318.

16. NCES, ibid tables 319, 320.

17. Oregon Independent Colleges Association, Fact Book, December 1995, page 7-1-10.

18. W. Zumeta and J. Fawcett-Long, "Access Policy in the Western United States: Is There a Role for Private Higher Education?" (University of Washington, Graduate School of Public Affairs, 1995), working paper.

19. Oregon's female labor force participation rate is 20 percent below the national average.

20. For a recent synopsis and critique of education demand analyses, see W. Becker, "The Demand for Higher Education", in S. Hoenack and E. Collins, The Economics of American Universities (SUNY Press, 1990), Chapter 7.

21. J. Hight, "The Demand for Higher Education in the United States 1927-1972: The Public and Private Institutions," Journal of Human Resources, Vol. 10, 1975, pp. 512-520.

22. Human Capital Research Corporation, Oregon Family Resource Study, August 1995, pp. 6-8.

23. J. Bishop, "The Effect of Public Policies on the Demand for Higher Education," Journal of Human Resources, 1977, Vol. 12, pp. 285-307

24. In fact, the income elasticity of older females was essentially the same as that for the younger, recent high school graduates.

25. This percentage uses full-time equivalent enrollments for colleges and universities, and nominal enrollments for non-collegiate institutions.

26. In economic terms, this is known as the "cross-elasticity" with respect to the other price.

27. Corman found that the interdependency relationships differed noticeably by age group. Specifically, younger students' demand for college enrollment was little affected by the tuition levels at technical schools, but technical school enrollment was significantly affected by tuition levels at colleges and universities, particularly for females. Arguably, this is because the vocational training historically sought by women (office and medical assistant training, etc) provides a career that is a closer substitute to the professional careers available with college education than is the case for men. Corman also found that the demand for technical school enrollment falls elastically with decreases in college tuition; the elasticity of technical school enrollment with respect to college tuition, in fact, is about 50 percent higher, in absolute value, than the elasticity of college enrollment to college tuition. See Corman's paper, "Postsecondary Education Enrollment Responses by Recent High School Graduates and Older Adults," The Journal of Human Resources, 1983.

28. US Department of Labor, and US Bureau of the Census. The Hispanic enrollment rate is a three-year moving average.

29. R. Hauser, "Trends in College Entry among Whites, Blacks and Hispanics," in C. Clotfelter and M. Rothschild, eds. Studies of Supply and Demand in Higher Education, National Bureau of Economic Research Project Report (University of Chicago Press, 1994).

30. Hauser also investigated gender differences. However, these have been discussed in detail above and, in any case, are fairly consistent across race and ethnic group.

31. Only 60 percent of Hispanics tended to complete high school at the time of Hauser's study data. Hence, the available pool of college applicants is more aggressively self-selected than is the case for whites or blacks.

32. For a convenient review of the education cost literature, see Brinkman, "Higher Education Cost Functions," and s. Hoenack, "An Economist's Perspective on Costs within Higher Education Institutions," in Hoenack and Collins, op cit., Chapters 5 and 6, respectively.

33. Author's calculations from NCES data. The average student-faculty ratio is used to proxy the average teaching load.

34. It may also be a manifestation of the perverse effects competing through signaling.

35. M. McPherson and G. Winston, "The Economics of Cost, Price, and Quality in U.S. Higher Education," in M. McPherson, M. Schapiro, and G. Winston, eds., Paying the Piper: Productivity, Incentives, and Financing of U.S. Higher Education (The University of Michigan Press, 1993), Chapter 4.

36. A. Daniere, Education in the American Economy (Random House, 1964).

37. M. McPherson and M. Schapiro, "The Effect of Government Financing on the Behavior of Colleges and Universities," in M. McPherson, M. Schapiro, and G. Winston, eds, op cit., Chapter 10.

38. See, for example, McPherson and Winston, op cit.

39. See Chapter 8 of Leslie and Brinkman, op cit. for a meta-analysis of studies on the effect of student aid on educational choices.

40. D. O'Neill, "Voucher Funding of Training Programs: Evidence from the GI Bill, Journal of Human Resources, Vol. 12, No. 4, p. 425-445.

41. Chapter 8 of Leslie and Brinkman, op cit.

42. Sallie Mae, "A Loan at Last: New College Graduates should know their Loan Repayment Options," press release, June 18, 1996, p. 1.

43. J. Brandl and V. Weber, "An Agenda for Reform: Competition, Community, Concentration," A Report to Governor Arne H. Carlson [Minnesota], November 1995, pp. 24-26.

Selected Bibliography

G. Becker, Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education (Second Edition), (National Bureau of Economic Research, 1975).

J. Brandl and V. Weber, "An Agenda for Reform: Competition, Community, Concentration," A Report to Governor Arne H. Carlson [Minnesota], November 1995.

C. Clotfelter and M. Rothschild, eds. Studies of Supply and Demand in Higher Education, National Bureau of Economic Research Project Report (University of Chicago Press, 1994).

E. Dean, Education and Economic Productivity, (Ballinger, 1984).

C. Finn, Scholars, Dollars and Bureaucrats (Brookings Institution, 1978).

I. Garfinkel, and R. Haveman, Earnings Capacity, Poverty, and Inequality (Academic Press, 1977).

D. Garvin, The Economics of University Behavior (Academic Press, 1980).

S. Hoenack and E. Collins, The Economics of American Universities (SUNY Press, 1990).

Human Capital Research Corporation, Oregon Family Resource Study, August 1995.

L. Leslie and P. Brinkman, The Economic Value of Higher Education (American Council on Education, 1993).

M. McPherson and M. Schapiro, Keeping College Affordable: Government and Educational Opportunity (Brookings Institution, 1991).

Oregon Independent Colleges Association, Fact Book, December 1995.

G. Psacharopoulos, Returns to Education (American Elsevier, 1973).

M. Schapiro, and G. Winston, eds., Paying the Piper: Productivity, Incentives, and Financing of U.S. Higher Education (The University of Michigan Press, 1993).

S. Rose-Ackerman, The Economics of Not-for-Profit Institutions (Oxford University Press, 1986).

H. Rosovsky, The University: An Owner's Manual (W.W. Norton, 1990)

T. Schulz, The Economic Value of Education (Columbia University Press, 1963).

U.S. Department of Education, National Center for Education Statistics (NCES), Digest of Education Statistics 1995.

Washington Office of the College Board, Trends in Student Aid.

Cascade Policy Institute, Oregon's free market think tank, is a 501(c)(3) not-for-profit, independent public policy research organization. With a respect for individual liberty and market forces, Cascade's mission is to explore voluntary, market-oriented answers to Oregon's public policy questions. Cascade works with a wide range of advisors, researchers and writers to produce easy-to-read articles and studies. Research and educational programs are directed towards the present and future public policy of Oregon. Cascade calls on local and national scholars, business people and policymakers to generate creative studies and publications. Financial support is based solely on private contributions. Tax deductible donations are welcome to support the work of the Institute.

Nothing written here is to be construed as necessarily reflecting the views of Cascade Policy Institute, as an attempt to aid or hinder the passage of any legislation, or as an endorsement of any candidate or initiative. Permission to reprint granted provided proper credit is given.

ISSN: 1078-4896

This project was made possible thanks to the generous support of Reason Foundation,
Los Angeles, California.

Cascade Policy Institute 813 S.W. Alder, Suite 450 Portland, OR 97205
Phone: (503) 242-0900
send mail to info@CascadePolicy.org

Return to Cascade home page