The blurring line: Between merit and need in financial aid – Industry Overview

The blurring line: Between merit and need in financial aid – Industry Overview – Statistical Data Included

Michael S. McPherson

Recently, the phenomenon of “merit aid” or “non-need-based” aid has attracted increasing commentary in American higher education. A number of observers (including the authors) have argued that the growing significance of merit aid is an indicator of the rising competitive pressures on colleges and universities, and that it is one factor undermining commitment to the principle of pricing a college education according to a family’s ability to pay.

The distinction between “need-based” and “non-need-based” student grants is, however, a slippery one. Many students who receive need-based assistance from a college will also receive a “merit award” as part of their overall aid package. Sometimes such a merit award will boost a student’s total grant dollars above those of another student with similar means who didn’t receive a “merit” award; in other cases, the school may simply be putting a “merit” label on dollars the student would have gotten anyway. Similarly, two students at the same college, both receiving only need-based aid, may receive quite different aid packages. The more desirable student may receive either a larger total aid package or a similar package but with a larger component of grant aid and lower amounts of loans and work than the less desirable student receives. And this can happen without any of the dollars being labeled “merit” dollars.

As we will show, a focus simply on dollars labeled as “merit” scholarships misses a good deal of the action regarding the responsiveness of grant awards to indicators of merit, understood as academic achievement or potential. Relatively few students receive awards that are explicitly labeled as “non-need-based” or “merit” awards (in our sample of 7,000 full-time, dependent students who completed the SAT, for example, only 4 percent of undergraduates at public colleges and 15 percent at private colleges received such awards from institutional funds), while many more receive “need-based” awards (22 percent at public colleges and 52 percent at private colleges). If these need-based grants (which throughout this article exclude athletic scholarships) are even moderately sensitive to “merit,” the impact on the overall distribution of aid may be considerable.

EMPIRICAL FINDINGS

The data reported in Table 1 suggest that this is the case. This table examines the distribution of grant aid among fulltime, dependent undergraduate students in the National Postsecondary Student Aid Survey (NPSAS) in 1995-1996. The table examines the joint impact of parents’ income and students’ SAT scores on grant awards. (Throughout this article, SAT scores are used as a convenient and measurable proxy for observable measures of academic achievement and promise.)

It is clear that at both public and private colleges and universities, family income has a significant impact on need-based aid, as one would expect in a system built around a family’s ability to pay. Thus, among students earning high SAT scores, institutionally awarded need-based grants range from $1,255 for the lowest-income students at public colleges to $330 for the highest-income students. At private colleges the range is from $7,123 to $1,364. These per-student award amounts are for all students–that is, they include those with zero awards.

Yet even when focusing on need-based grants, it is clear that SAT scores have a significant impact on the size of grants students receive. Among low-income students, those with high SAT scores who attend public colleges receive on average much more in need-based grants ($1,255) than those students either with middle ($904) or low ($565) SAT scores. The relationship to SAT score is just as strong in the other income bands. A similarly consistent relationship exists among students at private colleges. Students with high SAT scores consistently receive from their institutions, on average, more than twice as much in need-based grants as do low-scoring students, with middle-scoring students receiving an intermediate amount.

Obviously, including “merit” or “non-need-based” grants in the total mix increases the influence of SAT scores on award levels. But because relatively few students receive merit awards, the overall grant distribution is dominated by that of need-based awards. For example, for low-income students at private colleges, the difference between average need-based awards going to high- and low-scoring SAT students is $4,741, while the comparable difference for non-need-based awards is only $1,028.

This tabular analysis clearly indicates the importance of examining institutions’ overall distribution of grants, but many sources of variation underlie the findings in Table 1. The table reflects variation both within and between institutions and includes differences in the fraction of students receiving awards, as well as in the average size of the awards.

It is likely, for example, that part of the difference in institutional grants between students with higher and lower SAT scores comes from the fact that students with high scores probably attend more expensive institutions. This possibility is confirmed by the data reported in Tables 2 and 3, which look at the relationships among SAT score, income, and aid awards separately for students at high-tuition (Table 2) and low-tuition (Table 3) institutions. For private institutions, a positive relationship between the SAT score and need-based grant levels continues to be observed, but the magnitude of the relationship is considerably reduced at higher-tuition institutions (average need-based aid to low-income students is $7,700 for high SAT scorers and $5,978 for low SAT scorers-a difference of approximately 29 percent), while it remains strong at lower-tuition institutions ($2,828 for high SAT scores and $1,701 for those with low SAT scores on average–a difference of more than 66 percent).

At public institutions, however, the relationship between the SAT and need-based aid remains about as strong within the higher- and lower-tuition subgroups as it is in the whole group of institutions. Presumably this reflects a weaker relationship between average SAT scores and the tuition level at public universities than at private ones. Much of the variation in public tuition is explained by variation in state tuition policies rather than by differences in institutional prestige or “quality.”

This tabular analysis is strongly suggestive, but it requires stronger statistical verification. In another aspect of this research we have examined the relationship among the grants students receive from various sources and students’ personal characteristics at an individual level. We wanted to know how being a student from a family with a lower ability to pay affects the grant level, and also how his or her test scores affect the award. Here we limit ourselves to summarizing the major results.

Our key findings strongly support the picture provided by the data reported in Table 1. As we would expect, federal grant awards through the Pell and Supplemental Educational Opportunity Grant programs are strongly responsive to need, but show little systematic relation to SAT scores. State need-based grants behave similarly, while non-need-based state grants–a growing component of state awards–are sensitive to SAT scores and not to family resources.

Not surprisingly, merit or non-need-based grants awarded by the institutions are strongly responsive to SAT scores at both public and private institutions. While relatively few students receive such grants (only about 4 percent at public and 15 percent at private colleges and universities), our estimate is that if a student receives such a grant, a 100-point increase in his or her SAT results in more than a $1,000 increase in the expected annual award at a public institution and close to an $1,800 increase at a private institution.

More surprising, however, is our finding that SAT scores are significantly correlated with levels of need-based grant aid at both public and private institutions. As noted earlier, such grants are much more widespread than merit awards, reaching 22 percent of full-time, dependent students at public institutions and 52 percent of all students at private institutions. In both sectors, a student receiving a need-based institutional award can expect more than a $300 increase in her annual need-based award for every 100-point increase on her SAT.

Interestingly, these need-based awards appear to be less responsive to an ability to pay than a pure needs-based approach would lead us to expect. The NPSAS data set constructs from underlying data on family income and assets an estimate of the “expected family contribution” (EFC)–that is, the amount the family is calculated to be able to contribute to college costs based on standard needs-analysis formulas. A pure needs-based system would have grants decline by about $1 for every $1 increase in EFC.

This, indeed, is just about what happens with Pell Grants in our data. However, our data indicate that need-based grants made by institutions fall only by about 17 cents for every dollar of higher EFC at public institutions, and by 35 cents at private institutions. Of course, as family ability to pay goes up, not only institutional grants but also federal and state need-based grants would be expected to fall. However, taking all these forms of variation into account, we find that grants, in total, fall by only about 27 cents for every dollar of higher EFC at public institutions and by 41 cents at private institutions.

In summary, institutional need-based awards, in our data, are less sensitive to need and more sensitive to “academic merit” than the principles of needs analysis would lead us to expect. At this point, the most recent data we have available for analysis are for 1996. We are eager to replicate our analysis with the NPSAS ^2000 data, which are slated for release in the next several months. Our strong hunch is that the merit considerations in awarding need-based aid will have become stronger, and the purely income-oriented component weaker, in the intervening years.

IMPLICATIONS

The empirical analysis summarized here underscores a theme that is familiar both in our work and in that of a number of other researchers: The principle of awarding financial aid according to demonstrated need is, in significant measure, honored in the breach. Our contribution here is to show that even need-based grant awards are measurably sensitive to students’ academic promise (as measured by SAT scores), while the responsiveness of award levels to income is less than a totally need-based system would produce.

Our analysis underscores the point that a focus solely on “non-need” or “merit” aid significantly understates the role that academic promise and achievement play in the distribution of institutional grant aid. A significant caution is that our analysis only characterizes the net results of the existing admissions and financial aid systems, as reflected in the choices made by both students and institutions through offers made and offers accepted. We cannot produce from these data a complete picture of the distribution of the aid offers institutions make to applicants. Nonetheless, our results are so sufficiently strong and consistent that we believe they point to an important merit component in the award of need-based grants, as well as in the award of explicitly merit-based grants.

Of course, we know that “average” behavior in this arena hides substantial institutional variation. Certainly there are institutions that award all (or almost all) of their aid on the basis of need, measure all applicants’ need in the same way, assemble aid packages without regard to the desirability of individual students, and thus (thanks to large endowments and affluent applicant pools) are able to operate near the “ideal” of a need-based aid system.

There are also institutions whose aid policies are publicly and openly aimed at maximizing institutional advantage through recruiting the highest-achieving students at the lowest price, and thus view their student aid policies in purely “instrumental” terms. Most institutions operate somewhere in the middle. But the data suggest that this great middle may be oriented more toward the instrumental end than we are inclined to recognize.

This analysis raises several issues that we believe are relevant to policy discussions. First, there is a persistent tendency in state and, to some extent, in federal policy discussions to argue for more “merit-based” aid programs on the grounds that existing programs don’t give students significant incentives to perform well in high school. At least to the extent that SAT scores serve as a useful proxy for achievement in high school, our evidence is that there are considerable financial awards to achievement. (Our data set does not include information on high school GPA or other measures of scholastic achievement before college.)

As Table 1 indicates, even in public institutions, moving from the low- to the middle-scoring SAT category, or from the middle- to the high-scoring-a change that will often amount to between 100 and 200 points-can be worth several hundred dollars per year in institutional grants to a student. And the rewards at private institutions are on average substantially higher. These appear to be significant incentives. Given the popularity of SAT prep courses, students seem to understand that a better high school record will improve their chances of attending a more selective college-one where they are likely to require more financial assistance.

The other side of the coin is whether the existing admissions and financial aid system provides adequate opportunities for students with greater need. As Table 2 shows, a high-income student with a high SAT score attending a public institution can expect to receive on average $371 in need-based grant aid, more than half as much as the $692 received by a low-income student with a low SAT score. Middle-income students with low SAT scores do worse in total institution-based grant aid ($322) than do upper-middle- or high-income students who have high SAT scores ($746 and $558, respectively).

Analogous patterns prevail at private institutions. While low-income, high-scoring students certainly do well (with total institutional grants averaging $8,853), a high-scoring student from an upper-middle income family receives more than half as much in need-based grants ($3,507) as a low-income student with a low SAT score ($5,978).

These data thus make it natural to reflect on whether this distribution of institution-based aid serves the public interest in providing broad college access for all socio-economic groups and, if not, what factors might influence institutions to make different choices in how they distribute aid. We have argued elsewhere that an increasing focus on using the aid system to attract desirable students is one of several factors promoting increased stratification in American higher education.

Now, there may well be cases in which an institution’s decision to offer more aid to more academically attractive students is educationally productive. This would occur if a student’s presence in a school with less-talented classmates brought educational stimulus to those students that outweighed both the loss of her contribution to more-talented classmates elsewhere and whatever loss she might experience by being educated with less-talented, rather than more-talented, colleagues. However, once competition for talented students through the use of differential student aid offers becomes wide-spread, it is likely that much spending will go to offset the aid offers made by other institutions, and relatively little net movement of students will result.

In the latter case, the effect of colleges’ using aid to compete for more able students is either to divert resources from the educational program or from needier students who would otherwise receive grant support. When these circumstances prevail, curtailing the competition for students through merit-sensitive offers of aid may be in the social interest.

What might we consider to bring this about? It is helpful to think of five dimensions along which agreement might occur within a broadly need-based framework of financial aid awards. These dimensions can be described as common standards of ability to pay; common standards of aid packaging; commitment to need-only aid awards; commitment to full-need funding of aid; and commitment to need-blind admission. These different aspects of financial aid policy are often conflated with one another, but sorting them out would clarify public discussion. It is also important to appreciate that agreement can exist on some aspects of aid policy and not others.

Common standards of ability to pay are simply agreed upon ground rules for measuring family resources. A number of difficult judgments must be made in assessing a family’s financial capacity–familiar examples include the treatment of home equity and retirement savings, the contributions of divorced parents, whether other siblings are in college, and so on.

A major issue in the current aid system is that two broad systems for measuring need exist: “the federal methodology” instituted by Congress and the “institutional methodology” developed by the College Board. In most circumstances, the federal methodology yields a lower estimate of a family’s ability to pay for college. By strategically selecting which methodology to use in particular cases, a financial aid officer currently can be more or less generous to particular students, according to how attractive the students are to the school. In addition, financial aid officers have the ability to exercise “professional judgment” to deviate from the standard methodologies in particular cases. The idea is to use such judgment when unusual circumstances warrant; in practice, the unusual circumstance may not lie in the family’s financial circumstances but in the school’s desire for the student.

Common standards of aid packaging are agreed-upon ground rules on how the components of an aid package will be determined: how much loan, how much grant, how much work-study should be allotted. “Preferential packaging” is a policy of offering a larger grant in the aid package of a more desirable student.

Commitment to need-only aid awards implies an agreement to award assistance only to needy students and only to the extent of their demonstrated need.

Commitment to full-need funding of aid is an agreement to meet the demonstrated need of all admitted students.

Commitment to need-blind admission is an agreement to admit students (perhaps only first-year students or only U.S. students) without regard to their ability to pay.

It’s quite clear that an institution can commit itself to following some of these practices without committing to others. Many schools, for example, fund the full financial needs of admitted students but are “need-aware” in whom they admit. Conversely, some schools that are proud to proclaim themselves “need-blind” in admission don’t offer all the students they admit enough aid to meet their needs, or don’t include wait-listed students in the need-blind group. Similarly, an institution might say that all of its aid is need-based, but if it defines need in eccentric ways, it may give itself great latitude in awarding aid to students it wants. The whole idea of being “need-blind” becomes highly elusive when the aid package for a needy student may include many thousands of dollars of unsubsidized loans, or when admission of a wait-listed student is judged by ability to pay, perhaps without informing the student of this fact.

In principle, competition for students on the basis of aid awards could be eliminated if schools entered binding agreements on all five dimensions. It is very unlikely that such agreements could pass legal muster on antitrust grounds, however, nor is it likely that schools would be willing to give up so much freedom of action. Even in the heyday of “admissions overlap,” when some selective colleges compared in advance the need assessments they had prepared on individual students, agreements on packaging were rare, and agreements to eliminate merit scholarships were difficult to sustain.

In considering the potential for future agreement, we would make two main points. First, there is a very strong case, in principle, for arriving at agreement on common standards of ability to pay, because it would make transparent the procedure used for calculating family financial resources. Schools would arrive at agreement on which methodology to adopt and would work to make the methodology more detailed so as to limit in appropriate ways the discretion provided by professional judgment. They should also agree that professional judgment should be used only to accommodate “non-standard” cases, and not to provide better “deals” to more desirable students. It is important to appreciate that agreement on such calculations, even when enforced through arranging with a third party to monitor schools’ compliance with the agreement, imposes no absolute limitations on the aid awards that schools can make.

Thus, suppose that a college has two students (Sarah and Jane) in similar economic circumstances but finds Sarah much more desirable. Without an agreement on how to measure need, the college can do one of two things:

1) Determine that both students have the same level of financial need, and then offer Sarah a better mix of grants and loans (or a grant above need, recognizing her “merit”).

2) Fiddle with the need analysis to declare Sarah more needy, thus justifying a larger grant award to the more attractive student.

Although either practice may result in similar bottom lines, the two ways of proceeding are quite different. Suppose, for example, that Sarah applied to several colleges charging the same tuition, and one wanted her more than others. If the institutions abided by an agreement on how to measure need, the difference would be clear: Sarah would have the same need level at all three schools, and one would offer her more in grants and less in loans than the others. In the absence of such an agreement, Sarah instead might well receive different assessments of her ability to pay, still resulting in a larger grant from the school that favored her, but with less clarity about the reason.

The greater transparency that results from common measurement of ability to pay has a deeper significance as well. When a school declares that it is awarding (all or some) aid on the basis of need, it is asserting that its aid awards will be constrained by some external standard of measurement. It would plainly be absurd for a college to declare that it would award need-based aid and then, like the Red Queen, declare in the same breath that “need” meant whatever the college said it meant.

This is more than a rhetorical point, because when colleges ask families to fill out financial aid forms, they are implicitly, and sometimes explicitly, declaring that they will use the completed forms to determine what the families can afford to pay. The whole premise of encouraging families voluntarily to provide extensive information on their finances is that this information will be used to help them. Manipulating the use of such information in the institution’s interest has the twin defects of undermining families’ willingness to comply honestly with the system and casting doubt on the implicit contract the school has made with the family in soliciting the information.

The first conclusion we would draw from this analysis, then, is that colleges have good reason to seek agreement on standards for measuring ability to pay, for the sake of preserving a system that serves them well collectively. Moreover, the federal government, we would argue, has an interest in allowing such agreements to be arrived at and enforced. Doing so increases the transparency of the system, makes more meaningful the enforcement of the “overaward” rules the Congress adopted–rules that bar any recipient of federal aid from receiving a total aid package that exceeds his or her need–and helps to sustain a system of financial aid that, on the whole, has served the country well.

At the same time, the prevalence we have documented of institutions rewarding “merit within need”–the considerable influence of academic promise on even need-based grant awards–leads to the second conclusion we would draw: Even an agreement on standards of ability to pay would leave room for great variation among schools in the aid offers they make to students of comparable financial need. The sources of variation would still include explicit merit aid offers, differences in the grant component of aid that meets need, differences in the extent to which need is met, and differences among institutions in the extent to which they admit students without regard to need.

It’s reasonable to feel ambivalent about this. On one hand, it is reassuring for those who worry about restraint of trade among colleges to realize that even agreement on how to measure need still would leave great room for variation in the offers that schools make to students with comparable need. On the other hand, if one judges, as we do, that unrestricted competition for students has negative effects overall on the quality and accessibility of higher education in the United States, it is regrettable to recognize that agreement on measurement of need-or even on confining aid to needy students, as measured by common standards-would still leave enormous scope for competition in aid awards.

Our own conclusion is that the greater transparency provided by common standards of need measurement is a considerable virtue, and is also probably about as far as the antitrust laws might properly allow cooperation on aid policy to go. An agreement of this kind was recently reached by 28 private colleges and universities, including the two that we head. (Currently, enforceable agreements about common standards of need measurement are protected from antitrust prosecution only for schools that meet the standards set in Section 568 of the Education Amendments. That standard requires schools to be need-blind in admission of all U.S. freshmen, including those placed on the waiting list. We see no reason of policy or principle why this exemption should not be broadened considerably.)

Agreements at this level would still allow for a considerable amount of competition in the form of merit aid, as well as “merit within need.” It would, however, make it considerably easier for families to understand the differences in offers from different schools, and it would make schools’ assertions about “meeting full need” or being “need-blind” more meaningful. To move further in the direction of a truly need-based system of student aid, as we think would be desirable, would probably require incentives by the federal government to reward schools and state systems for focusing institutional and state aid on the neediest students.

A promising approach would be to institute a “Super-Pell” award-a higher grant level for students at institutions that devoted a large share of their own aid dollars to supporting needy students. There are some technical problems with implementing such a system, which we have written about elsewhere, but if the nation can muster the political will to take such a step, it would be a good way to begin restoring the historic partnership among colleges, students, and the federal government in helping needy students obtain a college education.

Michael S. McPherson is president of Macalester College and Morton Owen Schapiro is president of Will jams College. The authors wish to thank Richard Dunn for his excellent research assistance on this article.

TABLE 1.

AVERAGE INSTITUTION-BASED GRANTS BY SECTOR AND TYPE OF GRANT, IN

1995-1996 DOLLARS

Public Institutions

Income

High Upper-Middle Middle Low

Need-Based Grants

SAT Score

High $330 $439 $744 $1,255

Middle $272 $203 $413 $904

Low $0 $66 $224 $565

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $413 $192 $155 $37

Middle $0 $51 $70 $119

Low $0 $15 $61 $12

Sum of Need-and Non-Need-

Based Grants

Income

High Upper-Middle Middle Low

SAT core

High $473 $631 $899 $1,292

Middle $272 $254 $483 $1,023

Low $0 $81 $285 $577

Private Institutions

Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,364 $3,118 $5,997 $7,123

Middle $830 $1,988 $3,773 $4,244

Low $604 $1,629 $2,130 $2,382

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $609 $1,005 $1,430 $1,248

Middle $426 $809 $741 $604

Low $38 $247 $175 $220

Sum of Need- and Non-Need-

Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,973 $4,123 $7,427 $8,371

Middle $1,256 $2,797 $4,514 $4,848

Low $642 $1,876 $2,305 $2,602

Notes: Cell sizes below 20 are italicized. Athletic scholarships

excluded. Averages include both recipients and non-recipeints of aid.

Sample restricted to full-time, dependent undergraduate students who

reported SAT scores. SAT ranges: High: greater than 1100; Middle:

900-1100; Low: below 900 Income ranges: High: greater than $100,000;

Upper middle: $100,000-$60,000; Middle: $60,000-$30,000; Low: below

$30,000 Source: NPSAS 1996.

TABLE 2.

AVERAGE INSTITUTION-BASED GRANTS BY SECTOR AND TYPE OF GRANT, 1995-1996

DOLLARS, FOR INSTITUTIONS WITH TUITION ABOVE THE MEDIAN

Public Institutions, Tuition At or Above $3,128

Income

High Upper-Middle Middle Low

Need-Based Grants

SAT Score

High $371 $498 $1,040 $1,847

Middle $318 $292 $640 $1,358

Low $0 $123 $250 $692

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $187 $248 $146 $0

Middle $0 $65 $98 $302

Low $0 $35 $72 $28

Sum of Need- and Non-Need-Based

Grant

Income

High Upper-Middle Middle Low

SAT Score

High $558 $746 $1,186 $1,847

Middle $318 $357 $738 $1,660

Low $0 $158 $322 $720

Prviate Institutions, Tuition At or Above $14,666

Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,337 $3,507 $7,443 $7,700

Middle $1,286 $2,862 $6,004 $6,809

Low $1,363 $2,930 $3,834 $5,978

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SATScore

High $459 $757 $1,118 $1,153

Middle $281 $575 $740 $527

Low $0 $378 $134 $516

Sum of Need- and Non-Need-Based

Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,796 $4,264 $8,561 $8,853

Middle $1,567 $3,437 $6,744 $7,336

Low $1,363 $3,308 $3,968 $6,494

Notes: Cell sizes below 20 are italicized. Athletic scholarships

excluded. Average include both recipients and non-recipients of aid.

Sample restricted to full-time, dependent undergraduate students who

reported SAT scores. SAT ranges: High: greater than 1100; Middle: 900 –

1100; Low: below 900 Income ranges: High: greater than $100,000; Upper

middle: $100,000 – $60,000: Middle: $6,000-$30,000; Low: below $30,000

Source: NPSAS 1996.

TABLE 3.

AVERAGE INSTITUTION-BASED GRANTS BY SECTOR AND TYPE OF GRANT, 1995-1996

DOLLARS, FOR INSTITUTIONS WITH TUTION BELOW THE MEDIAN

Public Institutions, Tuition Below $3,128

Income

High Upper-Middle Middle Low

Need-Based Grants

SAT Score

High $272 $344 $455 $663

Middle $224 $84 $206 $471

Low $0 $21 $196 $472

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $79 $104 $164 $75

Middle $0 $33 $43 $39

Low $0 $0 $51 $0

Sum of Need- and Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $351 $448 $619 $738

Middle $224 $117 $249 $510

Low $0 $21 $247 $472

Private Institutions, Tution Below $14,666

Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,469 $1,974 $2,976 $2,828

Middle $422 $1,340 $2,259 $2,218

Low $335 $1,348 $1,754 $1,701

Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $1,209 $1,735 $2,083 $1,460

Middle $556 $983 $742 $664

Low $51 $219 $185 $159

Sum of Need- and Non-Need-Based Grants

Income

High Upper-Middle Middle Low

SAT Score

High $2,678 $3,709 $5,059 $4,288

Middle $978 $2,323 $3,001 $2,882

Low $386 $1,567 $1,939 $1,860

Notes:

Cell sizes below 20 are italicized. Athletic scholarships excluded.

Averages include both recipients and non-recipients of aid. Sample

restricted to full-time, dependent undergraduate students who reported

SAT scores. SAT ranges:

High:greater than 1100

Middle:900-1100

Low:below 900 Income ranges

High:greater than $100,000

Upper middle:$100,000-$60,000

Middle:60,000-$30,000

Low:below $30,000

Source:NPSAS 1996.

RELATED ARTICLE: RESOURCES

* McPherson, Michael S. and Morton Owen Schapiro, “Financing Undergraduate Education: Designing National Policies,” National Tax Journal, September 1997.

* —–, Keeping College Affordable: Government and Educational Opportunity, The Brookings Institution, Washington, D.C., 1991.

* —–, “Reinforcing Stratification in American Higher Education: Some Disturbing Trends,” paper prepared for the Macalester Forum on Higher Education, May, 1999.

* —–, The Student Aid Game: Meeting Need and Rewarding Talent in American Higher Education, Princeton University Press, 1998.

COPYRIGHT 2002 Heldref Publications

COPYRIGHT 2002 Gale Group