Introducing Bennett Hypothesis 2.0

Introduction

A quarter of a century ago, then Secretary of Education William J. Bennett made waves by declaring:

“If anything, increases in financial aid in recent years have enabled colleges and universities blithely to raise their tuitions, confident that Federal loan subsidies would help cushion the increase.”1

From that point forward, the notion that increases in financial aid cause increases in tuition has gone by the moniker of the Bennett Hypothesis, and its validity has been hotly debated ever since.

Many within higher education view the idea as preposterous. Most colleges are public or non-profit, so how could they possibly be greedily seeking “profits”? At the same time, many observers of higher education view it as an accurate depiction of reality. As Arthur Hauptman has noted, “just as one couldn’t imagine house prices being as high as they now are if mortgage financing were not available, it is difficult to believe that colleges and universities could have increased their charges so rapidly over time without the ready availability of students’ ability to borrow.”

Scholars have found evidence that contradicts the notion, but they have also found evidence that confirms the idea, which has allowed both opponents and supporters to claim vindication. In this paper, I argue that all the mixed evidence and subsequent controversy is a consequence of an overly simplified view of the Bennett Hypothesis. Tweaking the concept to account for a more realistic view of who receives financial aid, the actions available to colleges, and the nature of competition in higher education leads to predictions that are more consistent with the data than the original hypothesis or its antithesis. The three refinements are:

  1. All Aid is Not Created Equal
  2. Selectivity, Tuition Caps, and Price Discrimination are Important
  3. Don’t Ignore the Dynamic Story

 

Collectively, these changes to the original theory yield what I call Bennett Hypothesis 2.0. As we will see in the sections that follow, these changes help explain the mixed empirical evidence and offer a more accurate understanding of the relationship between financial aid and tuition. 

The Original Bennett Hypothesis

The Bennett Hypothesis holds that colleges will raise tuition when financial aid is increased, with the implication that increases in financial aid will not improve college affordability. Intuitively, it can be understood as a logical consequence flowing from the following observations:

  • 1. Individually, each college is trying to improve (the pursuit of excellence)
  • 2. More revenue is very useful in the quest for improvement.
  • 3. An increase in the generosity of financial aid gives colleges the option of acquiring more revenue by raising tuition to capture some of the aid.
  • 4. Most colleges will succumb to the temptation to raise tuition.
  • a. Some colleges will exploit (3) immediately to help them accomplish (1).
  • b. To keep from falling behind the colleges in (4a), even colleges that did not exploit (3) initially are pressured to do so in the future.
  • 5. Thus, an increase in financial aid leads to higher tuition (the Bennett Hypothesis).

 

While logical and intuitive enough, it is helpful to examine the idea in a little more detail. Figure 1, which shows three versions of the “market” for a typical college, will help explain the logic behind the Bennett Hypothesis. In all three panels, the horizontal axis measures the number of students (Q), and the vertical axis measures dollars ($). The demand curve (D) is downward sloping, indicating that as tuition falls, the number of students wanting to attend the college increases. The intersection of the supply and the demand curves gives us tuition (T) and enrollment (Q).

The Original Bennett Hypothesis
The Original Bennett Hypothesis

Panel A assumes that the college is capacity constrained, meaning that it has a vertical (perfectly inelastic) supply curve though we vary this assumption in the other two panels.3 Next, the government offers financial aid in the form of a grant of size G to each student. This shifts the demand curve up by G (to Daid). The new intersection with the supply curve gives a new tuition level (Taid). With a vertical supply curve, the change in tuition will be exactly equal to the financial aid grant (∆T = G), and there will be no change in enrollment (∆Q = 0). Thus, in this version of the model, tuition increases $1 for every $1 increase in financial aid.

This is the model most people have in mind when they think of the Bennett Hypothesis, and the version that is most frequently referred to in policy discussions. However, it is worth introducing two other variations of the model as well.

Panel B is the same as Panel A, except that the supply curve is horizontal (perfectly elastic) rather than vertical. Financial aid still shifts the demand curve (to Daid), but with a horizontal supply curve, tuition is left unchanged (T = Taid ), meaning there is no change in tuition (∆T = 0), but there is an increase in the number of students attending the college (∆Q > 0).

Lastly, Panel C repeats the exercise but with a typical upward sloping supply curve. The financial aid still shifts the demand curve (to Daid), but the new intersection of supply and demand imply an increase in the number of students (∆Q > 0) and an increase in tuition, though tuition increases by less than the increase in financial aid (G > ∆T > 0).

It is clear that the effect of financial aid on tuition depends heavily on the nature of the supply curve. Table 1, which summarizes the results of an increase in financial aid of $G based on the different supply curves, indicates that the Bennett Hypothesis effect (∆T > 0) will occur whenever supply is not perfectly elastic.

Results of an Increase in Financial Aid of $G in the Original Bennett Hypothesis
Results of an Increase in Financial Aid of $G in the Original Bennett Hypothesis

 

The Scholarly Evidence

Many scholars have examined the validity of the Bennett Hypothesis. A non-random sample of findings includes:

  • “Of the many studies that have tried to identify whether colleges react to federal financial aid, most find little to no response. While several studies do find a college price response, their overall results are mixed and often contradictory.”4
  • “Previous studies with evidence pertinent to the Bennett hypothesis are suggestive. McPherson and Shapiro (1991), Turner (1997), Li (1999), Netz (1999), Acosta (2001), and Long (2002) all find evidence that tuition rises for at least some segments of the higher education market… we find no evidence in support of the Bennett hypothesis among public or lower-ranked private universities… Among the best private universities, though, we find strong evidence of sharp increases in net tuition associated with increases in Pell aid.”5
  • “We found no evidence of the ‘Bennett Hypothesis,’ [at private institutions]… We did, however, find that public four-year institutions tended to raise tuition by $50 for every $100 increase in federal student aid.”6
  • “Estimates of the size of this ‘Bennett Hypothesis’ at public institutions range from negligible to a $50 increase in tuition for every $100 increase in aid.”7
  • “we find that higher education institutions raise net-price and lower their average institutional financial aid award when their states increase need-based awards, an indication that they are capturing increased state financial generosity.”8
  • “Previous studies of the Bennett hypothesis among public and non-profit institutions have found mixed results… we find large and significant differences between the tuition charged by [aid eligible] and [aid ineligible] institutions… [aid eligible] institutions charge about 56 log points, or 75 percent, more… The magnitudes are comparable to average per-student federal grant aid awards, suggesting that [aid eligible] institutions may indeed raise tuition to capture the maximum grant aid available.” 9

While certainly not a comprehensive review, these excerpts are representative of the typical findings in the literature and lead to three general observations. One, most studies find no evidence of the Bennett Hypothesis for at least some segment of higher education. Two, many studies find support for the Bennett Hypothesis for some segment of higher education, and three, among the second group, the increase in tuition is usually less than the increase in aid. This mixed and often contradictory evidence leads to the obvious conclusion that “the issue of whether various forms of financial aid ‘cause’ tuition increases remains unresolved.”10,11

Bennett Hypothesis 2.0

As the previous section indicates, the empirical evidence on the Bennett hypothesis is quite mixed. Strikingly, there is no evidence that a $1 increase in financial aid yields a $1 increase in tuition, which convincingly rules out the most common version of the Bennett Hypothesis - Panel A of Figure 1 (inelastic/vertical supply). Given the numerous findings that a $1 increase in aid results in an increase in tuition of greater than $0 but less than $1, we can also probably rule out Panel B (elastic/horizontal supply). Thus, to explain the scholarly findings in terms of the model presented above, we’d have to conclude that Panel C (which has an upward sloping supply curve) is the best description of the higher education market.

However, this conclusion is problematic. Most traditional colleges are capacity constrained at or near their current enrollment, especially in the short run, which means that their supply curve should be fairly inelastic (not necessarily perfectly vertical, but close to it).12 This is problematic for the Bennett Hypothesis because if supply curves are inelastic (or close to it), then as shown in Panel A of Figure 1, tuition should increase 1 for 1 (or close to it) with increases in financial aid, but the empirical evidence is clear that it does not.

In other words, we likely live in a world in which most colleges have fairly inelastic (fairly vertical) supply curves, but the tuition increases we observe in response to aid are too small to be consistent with that. I believe that some simple refinements of the original Bennett Hypothesis could help resolve this dilemma. I will refer to the collective refinements as Bennett Hypothesis 2.0 to emphasize that this is the next generation of the Bennett Hypothesis, and that there are important differences in assumptions, implications, and results.

The three key refinements of the original that constitute Bennett Hypothesis 2.0 are as follows:

  1. All Aid is Not Created Equal
  2. Selectivity, Tuition Caps, and Price Discrimination are Important
  3. Don’t Ignore the Dynamic Story

 

All Aid is Not Created Equal 

There is reason to suspect that different aid programs have different effects on colleges’ ability to raise tuition. In particular, aid programs that are restricted to low income students are less likely to allow colleges to raise their tuition. Intuitively, aid that is only available to low income students will mostly just allow those students previously priced out of the market to pay the prevailing tuition, without giving the college the capability of raising tuition. Figure 2 illustrates this story.

Aid for Low Income Students
Aid for Low Income Students is Not the Same as Aid for Everybody

Figure 2 begins with our standard demand curve (D), but we will make the further assumption that income is the dominant factor in determining willingness to pay, meaning that the top left of the demand curve consists of rich students, and the bottom right portion consists of low income students. When we add a fairly inelastic supply curve (S), the result is tuition of T and enrollment of Q.

Now consider the effect of two different aid programs. The first program is just like the universal grant G we introduced earlier. It is unrestricted (by income), with every student being provided with $G to help pay for college. This shifts the demand curve out (to DUnrestricted), and since the supply curve isn’t perfectly inelastic, results in higher tuition (TU > T) and higher enrollment (QU > Q).

The alternative aid program restricts aid to low income students only. Since low income students are clustered in the lower portion of the demand curve, this means that rather than shifting the demand curve out, it introduces a kink into the curve.13 The top portion of the demand curve remains unchanged because rich students do not qualify for the aid, but the bottom portion pivots at the point where poor students start to qualify for the program.14 The intersection of supply and demand now imply enrollment of QR and tuition of TR.

There are two key implications. First, financial aid that is restricted to low income students will result in a smaller increase in tuition (TR < TU). Second, if the income cut-off is low enough colleges may not be able to raise tuition at all (if the kink is drawn at T or below, T and Q will be unaffected).15

How Does This Help Explain the Scholarly Evidence?

This refinement helps explain why the scholarly evidence generally finds that a $1 increase in aid tends to lead to less than a $1 increase in tuition. As we just saw, we should not expect a $1 to $1 relationship when aid is contingent upon income, since such aid does not give colleges as much room to raise tuition (any room if the income cutoffs for recipients are low enough). 16 This is particularly important because findings that Pell grants do not lead to (much) higher tuition are typically taken as strong evidence against the Bennett Hypothesis. But once Pell grants are modeled as a kink in the demand curve rather than shift of the demand curve, we wouldn’t expect a 1 to 1 relationship, taking away some of the strongest evidence against the Bennett Hypothesis.

What are the Main Lessons?

The main lesson from this first refinement is that aid targeted to low income students (such as the Pell grant and subsidized Stafford loans) kink the demand curve, while universal (or near universal) aid (such as unsubsidized Stafford loans and the education tax credits) shift the entire demand curve. These programs therefore have very different implications when it comes to their impact on tuition.

For policy makers, the key point is that financial aid that is restricted to low income students is much less likely to be captured by colleges, and will therefore be more likely to succeed in making college more affordable and therefore accessible (for low income students). In contrast, universally available programs are more likely to simply fuel tuition increases and therefore more likely to fail to make college more affordable.

From a scholarly perspective, there are three important points to emphasize from this refinement. First, results from one program may not generalize to other programs, particularly when the programs have different beneficiaries. The second, related point is that summing all financial aid programs into one aggregate financial aid variable is inappropriate. If Pell grants and unsubsidized Stafford loans have different effects, then summing them together will not yield results reflective of either program. The third point is that the same program can have different effects at different colleges, leading to higher tuition at some but not others based on the existing level of costs and tuition.

Selectivity, Tuition Caps, and Price Discrimination are Important

The second refinement accounts for two common practices within higher education that weaken the link between aid and tuition; tuition caps and price discrimination.

Tuition Caps

Many public universities are subject to tuition caps or growth rate caps by their state legislators as a condition of receiving state funding. Figure 3 illustrates the mechanics of a tuition cap. We start with an inelastic supply curve and no aid, yielding tuition T and enrollment Q. We then give each potential student aid of G, which shifts the demand curve (to Daid). If the legislature does nothing, we’d expect for tuition to increase to TUncapped. But what happens if the state legislature caps tuition at TCapped?

At TCapped, QC students would like to enroll in the college, but because the college is capacity constrained, it can only enroll Q students. With a surplus of applicants at the legislatively capped tuition rate, the college needs to ration enrollment slots in some manner, and the most common method is to use students’ previous academic performance. In terms of figure 3, this implies that among all the students from the origin to QC only the best Q students will be offered admission and the bottom QC - Q will be rejected.

While state governments may force such decisions on some public colleges, this has a much broader implication of illuminating a key tradeoff for colleges between increasing tuition and improving the quality of their student body. Even colleges that are not subjected to legislatively imposed caps on tuition may decide that the benefits of being able to select the best Q students from among the QC applicants outweighs the benefits of increasing revenue by charging the highest possible tuition. Indeed, we do not observe many traditional colleges charging as much as they possibly can, indicating that significant value is placed on the quality of students even when it comes at the expense of lower revenue. 

Tuition Caps and Selectivity
Tuition Caps and Selectivity

Price Discrimination

Up until now, we have been assuming that colleges set one tuition level for all students. While this is typically the case for public colleges, many private colleges engage in price discrimination, which entails charging different students different tuition (accomplished by offering students discounts or college funded scholarships, collectively referred to as institutional aid). Figure 4 will help illustrate the mechanics and implications of price discrimination.

Imagine that we have a traditional private college that is capacity constrained. As we saw in Panel A of Figure 1, a college that doesn’t price discriminate would charge a price of T before the financial aid, and a price of TAid after the aid. But a college that price discriminates does not have just one price, but rather a different price for each student.19 Mechanically, the college sets a high tuition level, such as TMax, but then offers students institutional aid of varying amounts resulting in varying tuition charges. Student J could be offered a relatively small scholarship, resulting in a high price of TJ while student K could be offered a large scholarship, resulting in a low price of TK. By charging each student their maximum willingness to pay, the college can increase their revenue significantly.

For simplicity, we’ll assume that the college has constant marginal costs of C, and that it is unwilling to take a loss on students (for each student, Ti > C). Offering financial aid grants to students at a college that price discriminates puts them in a similar situation as the public university subjected to a tuition cap - the number of students wanting to enroll exceeds the college’s capacity. Just like the tuition capped college, a college which engages in price discrimination could use this surplus of applicants to enhance its selectivity. However, the price discriminating college also has the option to squeeze more revenue out of wealthier students.

Price Discrimination and Selectivity
Price Discrimination and Selectivity

Note that without aid, student K is unable to cover the college’s costs (C), and would therefore not be admitted. But with aid, student K can now cover costs C, and could be a viable candidate. But that does not guarantee that student K will replace a lesser qualified student. To see why, suppose that the university has already admitted Q-1 students, and is now faced with a choice between student J (who is rich but a bad student) and student K (who is poor but a good student) to admit as its last student. For the college, the advantage of admitting J is that he/she would pay more in tuition, and that extra money could be used to improve the college, while the advantage of admitting K is that he/she is a better student whose admission will improve the selectivity of the college. Some colleges would choose student J, but some would choose student K.

How Does This Help Explain the Scholarly Evidence?

Tuition caps and price discrimination (and the resulting trade-offs concerning selectivity and revenue) help explain the scholarly evidence because they weaken the predicted relationship between increases in aid and increases in tuition. Intuitively, public colleges are often subject to tuition caps, and private colleges typically practice price discrimination (and some public colleges do too). Both result in a tradeoff between increasing revenue and increasing the size of the applicant pool which allows the college to become more selective. Because selectively is also valued by colleges, we should not expect for tuition to increase 1 for 1 with aid even when supply is inelastic, meaning that this refinement helps make Bennett Hypothesis 2.0 more consistent with the empirical evidence.

What are the Main Lessons?

For policymakers, the first lesson is that capping tuition at public universities will encourage those universities to become more selective. This may be a good thing in some respects, but it does have drawbacks as well. The second lesson for policymakers concerns private universities. Price discrimination allows these colleges to raise tuition in response to aid at an individual level (this is just the Bennett Hypothesis at an individual level). But in order for colleges to price discriminate, they must know each student’s ability to pay. This means that providing colleges with students’ financial background will lead to more aid being captured. Bizarrely, the government currently provides colleges with this information, thus encouraging and facilitating price discrimination. Ending the counterproductive practice of providing colleges with information on the financial background of students and parents would curtail price discrimination, which would increase the effectiveness of aid in improving college affordability.

For scholars, the main lesson is that it is highly unlikely that traditional colleges’ actions are consistent with simple objective functions such as profit or revenue maximization. While higher revenue is undoubtedly viewed as positive, ceteris paribus, other objectives, such as boosting selectivity, may be hurt by a single minded pursuit of higher revenue. This makes modeling university behavior more complex.

Don’t Ignore the Dynamic Story

The first two refinements of the Bennett Hypothesis can be considered minor tweaks of the model (that nevertheless have important implications). The third refinement is a more substantive change. The Bennett Hypothesis is generally understood in terms of the models presented above, which are essentially snapshots in time (static). But there is reason to believe that changes over time (dynamic) can be just as if not more important than the static considerations.

A Short Detour to Introduce Bowen’s Rule

Before going further, we first need to take a short detour to discuss Bowen’s Rule. In 1980, Howard R. Bowen introduced the five laws of higher educational costs:

  1.  “The dominant goals of institutions are educational excellence, prestige, and influence.”
  2. “In quest of excellence, prestige, and influence, there is virtually no limit to the amount of money an institution could spend for seemingly fruitful educational needs.”
  3. “Each institution raises all the money it can.”
  4.  “Each institution spends all it raises.”
  5. “The cumulative effect of the preceding four laws is toward ever increasing expenditure.”

These five laws have been summarized as “Bowen’s Rule,” which holds that colleges raise and spend all the money they can in the pursuit of excellence. Virtually everyone who has studied the issue has verified Bowen’s Rule as an accurate description of colleges’ behavior:

  • Charles Clotfelter: “the operational objective of the research university is simply to ‘be the best.’ At the same time, each research university is locked in continual battle with its competitors… Expenditures on salaries, facilities, and amenities are crucial to this competition, and therein lies the source of an ongoing, unsatisfied demand on the part of universities for more revenue… every private research university worth its salt always has a list of worthwhile projects to fund.”
  • Derek Bok: “Universities share one characteristic with compulsive gamblers and exiled royalty: there is never enough money to satisfy their desires.”
  • Ronald G. Ehrenberg: “maximizing value to these administrators means making their institutions the very best that they can be in almost every area of their activities. These administrators are like cookie monsters… They seek out all the resources that they can get their hands on and then devour them.”
  • Robert Martin: "higher education finance is a black hole that cannot be filled."

 

The Implications of Bowen’s Rule 

Bowen’s Rule provides powerful insights into the effects of financial aid over time. To explore this dynamic story, consider Figure 5.

Suppose that the higher education sector consists of just two colleges, D and E, and that in time period t, they both have constant marginal costs of Ct . The first college is capacity constrained (inelastic supply), while the second is not (elastic supply). Now suppose all students are given a grant of G, shifting the demand curve out (to Daid). As we saw (in Panel A of Figure 1), we would expect to see this aid lead to higher tuition for college D. But we also saw (in Panel B of Figure 1) that we would not predict higher tuition for college E. We will even go a step further and assume that college E will always set tuition equal to costs, to further emphasize the lack of any way for aid to lead to higher tuition.

Illustrating How Aid Fuels Cost Increases
Illustrating How Aid Fuels Cost Increases

Intuitively, the financial aid increases the ability of students to pay for both colleges. Since college D is capacity constrained, they raise tuition, become more selective, or do some combination of both. For simplicity, assume that it raises tuition to Taid (though as we saw in Figures 3 and 4, this is unlikely). In contrast, college E will leave tuition unchanged (at Ct = Tt) and simply enroll more students. Thus, we should not be worried about the Bennett Hypothesis at all at college E.

But this is only the immediate (static) story. What happens the next year, and the year after that (the dynamic story) is also relevant and much less reassuring. To understand this dynamic story, we return to college D. It raised tuition, which gives it more revenue. But what does it do with that revenue? Because most colleges are public or non-profit, they cannot distribute the money to shareholders, which means that the extra revenue will be spent to improve the institution. It may hire more professors to conduct more research, to lower class sizes, or to allow teaching loads to be reduced. Or perhaps it builds new laboratories or classrooms, or expands student services to improve its graduation rate. Each of these may be an appropriate expenditure in some cases, but each will also raise the college’s costs in the future. Tenured faculty are difficult to get rid of, new labs and buildings must be maintained, and new bureaucracies become entrenched. What it spends the money on is irrelevant for our purposes; the important point is that the college spends it, and virtually regardless of what they spend it on, it will result in higher future costs. So at college D, costs in the next time period (t+1) are higher than starting costs (Ct+1 > Ct). This is not necessarily a problem for college D, since it is already charging students enough to cover its higher costs.

But what about college E? We know that initially, aid does not affect tuition at all at college E. But as college D spends more money, college E needs to spend more to avoid falling behind. If it wants to attract the best professors, it needs to increase pay, lower teaching loads, and build state of the art labs when college D does. And if it wants to recruit good students, it has to offer the same amenities that college D does. Thus, the same things that lead to higher future costs at college D lead to higher future costs at college E, so Ct+1 > Ct , which for college E means that Tt+1 > Tt .

The story told in Figure 5 is truly remarkable. For college E, we have stacked the deck against the Bennett Hypothesis as much as possible, assuming both a perfectly elastic supply curve and that it mechanically sets tuition equal to cost. Either of these assumptions should be enough to rule out the Bennett Hypothesis completely, and, in the static case, they do. But as soon as we look past the immediate (static) term, and think about the future (dynamic) impact, we find that Bennett Hypothesis 2.0 applies.

There is Almost No Escaping Bennett Hypothesis 2.0

The lesson from Bennett Hypothesis 2.0 is that there is an overwhelming danger, especially over time, that higher financial aid will lead to higher tuition. While the first two refinements reduce the threat of this result in the short run by weakening the tie between aid and tuition, the third indicates that in the long run, the two are tightly related even in situations where there is little immediate danger.

Is there any way to circumvent Bennett Hypothesis 2.0 - to avoid financial aid leading to higher tuition? Yes, but before we get there, we first need to explore the driving force behind Bennett Hypothesis 2.0.

The Nature of Competition Drives Bennett Hypothesis 2.0

In Figure 5, we saw that in higher education, even with an elastic supply curve, a subsidy (financial aid) leads to an increase in price (tuition) over the long run. In a typical market with an elastic supply curve, subsidization does not have this effect. For instance, suppose that the supply curve for bread is perfectly elastic, and that the government starts subsidizing bread. We would not expect bread producers to simply spend more making bread until they had captured the entire subsidy, so why does this happen in higher education?

The key difference between higher education and the bread industry is the nature of competition: In higher education, colleges essentially compete in a zero-sum game for relative standing. Due to the lack of measures of output and outcomes, colleges cannot compete on quality, and instead compete based on reputation/prestige/excellence. Essentially, they use high quality inputs as proxies for quality because there is no way to demonstrate high quality directly. Since high quality inputs are costly, and colleges are playing a zero-sum game of relative position, there is no limit to what college will spend in the pursuit of excellence. Thus, they will spend as much as they can, meaning that revenues drive costs (an implication of Bowen’s Rule). 

In contrast, bread producers compete based on value (roughly defined as quality divided by price, both components of which are observable to consumers). Costly improvements to bread making will only be undertaken if they improve quality enough to compensate for the increase in costs (and therefore, in the long run, price). Bread makers seek to make the bread with the highest value, they do not seek to spend as much as possible in the pursuit of the highest quality bread regardless of cost. In other words, there is no Bowen’s Rule for bread making. And because there is no Bowen’s Rule for bread making, there is no need to worry that subsidizing bread will lead to higher bread prices and capture of the subsidy by bread makers (assuming a perfectly elastic bread supply curve, as is likely in the long run).

Escaping from Bennett Hypothesis 2.0

Because the nature of competition in higher education is the driving force behind these dysfunctional results, the clearest way to escape Bennett Hypothesis 2.0 is to change the nature of competition. Colleges compete in a zero-sum game based on prestige because they cannot compete based on value, and they cannot compete based on value because measures of both quality and price (net tuition) are obscured. If information on those two were available, the pursuit of excellence would be replaced by the pursuit of value, Bowen’s Rule would break down, and Bennett Hypothesis 2.0 would no longer be a concern.

Progress is being made in making pricing information available (colleges are now required to publish net price calculators), but there is no progress regarding quality. A good start would be to publicize employment outcomes, value-added pass rates on certification exams, etc. Until information on such outputs and outcomes is available, we will be stuck in a world where competition is based on prestige, which in turn means we will continue to suffer from Bowen’s Rule and Bennett Hypothesis 2.0.

Barring an overhaul of the nature of competition in higher education, there are a few other ways to avoid Bennett Hypothesis 2.0. While Bowen’s first two laws hold for most public and private non-profit colleges, for-profit colleges don’t care about prestige, they care about profits. Thus, if competition in higher education worked like it does in other industries, the Bennett Hypothesis would not apply to forprofits. Unfortunately, competition in higher education is broken, so we can expect the Bennett hypothesis 2.0 to apply even to for-profits,  and indeed a recent study finds that the Bennett Hypothesis does apply to for-profits.

Another way to avoid Bennett Hypothesis 2.0 was pointed out earlier - restrict aid to only the very poorest. If aid is only provided to those that were previously priced out of higher education, then colleges cannot raise prices when aid is given without again pricing them out of the market. Of course, this implies giving aid only to the very neediest students and even then only a limited amount of aid. 

With minor exceptions, all of the potential escapes from Bennett Hypothesis 2.0 are unlikely to apply to higher education in the near future. This means we should be worried about Bennett Hypothesis 2.0 in most situations.

Conclusion

Original Bennett Hypothesis + a couple refinements + Bowen’s Rule = Bennett Hypothesis 2.0.

The original Bennett Hypothesis held that increases in financial aid will lead to higher tuition, but the empirical evidence testing the hypothesis is inconclusive. The next generation of the concept, Bennett Hypothesis 2.0, adds three refinements.

  • All Aid is Not Created Equal
  • Selectivity, Tuition Caps, and Price Discrimination are Important
  • Don’t Ignore the Dynamic Story

These three refinements not only help explain the mixed empirical evidence, but also provide a better understanding of the relationship between financial aid and tuition. While the first two refinements weaken the link between the two (lessening our concern about Bennett Hypothesis 2.0), the third refinement strengthens the link, implying that we should almost always be concerned about financial aid leading to higher tuition.

Given the current structure of the higher education system, Bennett Hypothesis 2.0 implies that the government will always be fighting a losing battle to increase access to college or improve college affordability since “additional government [financial aid] funds keep providing revenues that, under the current incentive system, increase costs.”54 As higher financial aid pushes costs higher, it inevitably puts upward pressure on tuition. Higher tuition, of course, reduces college affordability, leading to calls for more financial aid, setting the vicious cycle in motion all over again.

Bennett Hypothesis 2.0 exacerbates rather than causes out of control spending by colleges, the ultimate cause of which is Bowen’s Rule. Nevertheless, that is no excuse for ill-designed financial aid programs to pour fuel the fire. As Bennett noted:

“Federal student aid policies do not cause college price inflation, but there is little doubt that they help make it possible.”

Those words remain just as true today as they were a quarter century ago.

References: 

  1. Basken, Paul. "Financial-Rescue Plan, in Shift, Could Aid Student-Loan Providers." Chronicle of Higher Education. November 13, 2008.
  2. Bauerlein, Mark. "Professors on the Production Line, Students on Their Own." Future of American Education Project Working Paper. American Enterprise Institute, 2009.
  3. Bennett, William J. "Our Greedy Colleges." The New York Times, February 18, 1987.
  4. Bok, Derek C. Universities in the marketplace: the commercialization of higher education. Princeton University Press, 2003.
  5. Bowen, Howard R. The Costs of Higher Education. Jossey-Bass Publishers, 1980.
  6. Breneman, David W. "Rising cost of college tuition and the effectiveness of government financial aid: hearings before the Committee on Governmental Affairs, United States Senate, One Hundred Sixth Congress, second session." United States. Congress. Senate. Committee on Governmental Affairs, February 2000.
  7. Cellini, Stephanie Riegg and Goldin, Claudia. “Does Federal Student Aid Raise Tuition? New Evidence on For-Profit Colleges.” National Bureau of Economic Research working paper 17827, February 2012.
  8. Clotfelter, Charles. Buying the best: Cost escalation in elite higher education. Princeton University Press, 1996.
  9. Cunningham, Alisa F., Jane V. Wellman, Melissa E. Clinedinst, and Jamie P. Merisotis. "Study of College Costs and Prices, 1988–89 to 1997–98." National Center for Education Statistics, December 2001.
  10. Curs, Bradley R., and Luciana Dar. "Do Institutions Respond Asymmetrically to Changes in State Need and Merit Based Aid?" November 2010.
  11. Ehrenburg, Ronald G. Tuition Rising: Why College Costs So Much. Harvard University Press, 2002.
  12. Gillen, Andrew. “Financial Aid in Theory and Practice: Why It Is Ineffective and What Can Be Done About It.” Center for College Affordability and Productivity, April 2009.
  13. Gillen, Andrew, Matthew Denhart, and Jonathan Robe. "Who Subsidizes Whom? An Analysis of Educational Costs and Revenues." Center for College Affordability and Productivity, March 2011.
  14. Greene, Jay P., Brian Kisida, and Jonathan Mills. "Administrative Bloat at American Universities: The Real Reason for High Costs in Higher Education." Goldwater Institute, n.d.
  15. Hauptman, Arthur M. "Class Differences in Room for Debate, Rising College Costs: A Federal Role?" The New York Times, February 3, 2010.
  16. Long, Bridget Terry. "What Is Known About the Impact of Financial Aid? Implications for Policy." National Center for Postsecondary Research, April 2008.
  17. Martin, Robert E. "The Revenue to Cost Spiral in Higher Education." The Clarion Call. John William Pope Center for Higher Education Policy, July 2009.
  18. McPhearson, Michael. "Comment to Bridget T. Long, "The Impact of Federal Tax Credits for Higher Education Expenses"." In College Choices: The Economics of Where to Go, When to Go, and How to Pay For It, by Caroline M. Hoxby, 165-168. National Bureau of Economic Research, 2004.
  19. McPherson, Michael, and Morton Owen Schapiro. "The Student Aid Game: Meeting Need and Rewarding Talent in American Higher Education." Princeton University Press, 1998.
  20. Rizzo, Michael J., and Ronald G. Ehrenberg. "Resident and Nonresident Tuition and Enrollment at Flagship State Universities." Working Papers. National Bureau of Economic Research, September 2004.
  21. Singell, Larry D, and Joe A Stone. "For Whom the Pell Tolls: Market Power, Tuition Discrimination, and the Bennett Hypothesis." University of Oregon Economics Working Paper, April 2003.
  22. Smith, Gordon C. S., and Jill P. Pell. "Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials." British Medical Journal, 2003.
  23. Wolfram, Gary. "Making College More Expensive: The Unintended Consequences of Federal Tuition Aid." Cato Insititute Policy Analysis No. 531. CATO Institute, 2005.