Bennett Hypothesis & Tuition Inflation (Economic Data)
The Bennett hypothesis holds that colleges may raise tuition when students receive more federal aid, because some aid expands their ability to pay. Rising aid and tuition alone do not prove this effect. To test it, compare published prices and net prices, separate grants from loans, adjust for inflation, and use a research design that can identify whether aid changes caused tuition changes.
Digital dashboards and downloadable federal data have made college-price research faster. They have not made it simple. IPEDS, College Board reports, and research papers may use different years, student groups, and price measures, so figures that look conflicting can describe different things.
When I review tuition analyses, I start by asking a basic question: What price is being measured? A college’s published tuition is not the same as the amount a student pays after grants. That distinction is central to judging whether federal aid fuels tuition inflation or helps students afford college.
What the Bennett hypothesis claims
The Bennett hypothesis is the idea that colleges may respond to increased student aid by raising tuition. It is often linked to a 1987 statement by then-U.S. Secretary of Education William Bennett. The claim is testable, but its effects may differ by aid type, college sector, and time period.
The mechanism is straightforward: if students can draw on more aid, colleges may have room to charge more. But that is a hypothesis about cause and effect, not a rule that applies to every campus. Tuition can also rise for other reasons, including changes in state funding, wages, facilities, or student demand.
The word “inflation” also needs care. In everyday speech, it can mean any price increase. In economic analysis, researchers usually mean a price increase after adjusting for general inflation. A nominal tuition increase may look large while its inflation-adjusted increase is smaller, or even absent.
How to define the outcome
A sound analysis states which price it studies before comparing aid and tuition. Published tuition and required fees are the listed charges for attendance, often reported for a full-time student. Net tuition is the price after grants are subtracted. Loans are borrowed funds, so they do not reduce net price in the same way.
This difference matters to students and policymakers. A college may raise its published price while also increasing grants. In that case, the sticker price rises, but the average net price may rise less, stay flat, or fall.
A study of published tuition can help explain colleges’ pricing behavior. A study of net price speaks more directly to what students pay after grant aid. Neither measure alone answers every affordability question, such as how much students borrow or how costs compare with family income.
Key takeaway: Always label the outcome. Do not use published tuition as a stand-in for a student’s net price.
How to measure tuition and student aid
A useful comparison needs consistent definitions across colleges and years. IPEDS, the U.S. Department of Education’s core higher-education data system, reports tuition and required fees, enrollment, and other institutional measures. Researchers should check each data item’s coverage and definition before combining it with aid data.
Begin with one institutional unit and one period, such as annual tuition and aid per full-time-equivalent student. Then document who is included: public, private nonprofit, or for-profit colleges; undergraduate or graduate students; full-time or all students. A change in the student mix can shift an average even when prices for individual students do not change.
Keep aid categories separate
Federal aid is not one measure. Pell Grants are grants for eligible students and do not have to be repaid. Federal loans must be repaid. Other grants may come from states, colleges, or private sources. Combining them can hide different effects on both college revenue and student costs.
Also state whether an aid figure is offered, awarded, or disbursed. An offer is not necessarily accepted, and an award is not always the same as the amount ultimately received. Report aid per student or per full-time-equivalent student, and clarify whether the denominator includes all students or only aid recipients.
Adjust prices for inflation
Show nominal prices, the dollars charged in each year, and constant-dollar prices, which account for changes in general prices. Name the deflator and base year so another reader can reproduce the calculation. The Consumer Price Index for All Urban Consumers (CPI-U) is one possible measure, but it is not the only one.
For example, an analysis could convert all tuition figures to 2024 dollars using CPI-U. That does not mean tuition and general prices move in the same way. It simply helps separate a college-price change from the broader rise in prices.
| Measure | What it tells you | Main caution |
|---|---|---|
| Published tuition and required fees | The listed charge for a defined student group | Does not account for grants |
| Net tuition after grants | Tuition remaining after grant aid | Must define which grants and students are counted |
| Federal loan aid | Borrowed funds available to students | Is not a discount and must be repaid |
| Aid per full-time-equivalent student | Aid relative to a consistent enrollment measure | Results depend on how aid and enrollment are counted |
| Constant-dollar tuition | Price change after general inflation is removed | Depends on the chosen deflator and base year |
Key takeaway: Use consistent years, student groups, and units. Keep grants and loans distinct, and report the inflation method.
What the evidence can and cannot show
Research does not support a universal yes-or-no verdict. Findings depend on the aid policy, sample, time period, and method used to estimate cause and effect. A tuition-and-aid trend can identify a pattern worth studying, but it cannot by itself show that aid caused tuition to rise.
A key challenge is that both variables may change for other reasons. For example, a state may cut funding at the same time that federal aid expands. Colleges could then raise tuition to replace lost revenue. A simple comparison might wrongly attribute the increase to student aid.
Cellini and Goldin’s study of for-profit colleges examines tuition changes around access to federal student aid. It reports that gaining access to federal aid was linked to higher tuition in the colleges studied. This is a distinct sector and policy setting; it should not be treated as a direct estimate for public universities or private nonprofit colleges.
These studies help explain why the debate persists. Their designs focus on particular policy changes and institution groups, rather than every aid program and every college. A careful summary names the sample, years, aid measure, and outcome instead of turning one result into a general law.
A causal test, not just a trend line
A stronger analysis asks what tuition would likely have been without the aid change. One approach is difference-in-differences: compare tuition changes at colleges more exposed to a policy with changes at less-exposed colleges. The key assumption is that, without the policy, the two groups would have followed similar trends.
A useful model should consider institutional type, enrollment, state appropriations, and other relevant cost or revenue changes. Researchers should test whether results change when they use other time periods, comparison groups, or price measures. They should also report uncertainty, such as confidence intervals, rather than only a point estimate.
| Evidence pattern | What it supports | What it does not prove |
|---|---|---|
| Aid and tuition rise together | The two measures moved at the same time | Aid caused the tuition increase |
| More-exposed colleges raise tuition more after a policy change | A causal effect may be present if comparison assumptions hold | That the effect applies to all colleges or aid programs |
| Published tuition rises but net tuition does not | Grants may offset some of the listed-price increase | That every student’s cost stayed unchanged |
| Results differ by sector | Effects may depend on college type or policy setting | That one sector’s result represents the whole system |
Key takeaway: Look for a credible comparison group and clear assumptions. Correlation is a starting point, not a causal conclusion.
How to interpret prices for students and policy
For a student choosing a college, the Bennett hypothesis does not predict an individual financial-aid offer. It addresses possible system-level responses to aid. A student should compare the actual grant offer, likely net price, expected borrowing, and program costs across colleges.
The following examples show how sticker-price and net-price changes can differ. They are arithmetic scenarios, not estimates from a research study. Each assumes the grant increase goes to the same student group used to calculate net tuition.
| Illustrative change per student | Published tuition change | Grant aid change | Net tuition change |
|---|---|---|---|
| Tuition unchanged; grant rises | $0 | +$1,000 | -$1,000 |
| Some tuition pass-through | +$600 | +$1,000 | -$400 |
| Tuition increase exceeds grant increase | +$1,200 | +$1,000 | +$200 |
The middle case shows why the phrase “aid raises tuition” is incomplete. Even if tuition increases, a student’s net tuition may fall when grant aid rises by more. Yet averages can conceal differences: students who receive no grant, attend part-time, or face other fees may experience another outcome.
For policy analysis, report tuition and net tuition together where possible. Compare sectors and periods, and state whether the estimate applies to grant aid, loan limits, or another policy. College Board’s Trends in College Pricing and Student Aid can provide context on prices and aid, while NCES and IPEDS supply institutional data for more specific analysis. Their measures should be checked before they are merged.
Next step: Match the statistic to the decision. Students need their own net price and borrowing terms; researchers need a defined policy, sample, and causal design.
A practical checklist for evaluating a claim
A claim about aid and tuition is easier to assess when its evidence can be traced. I use this checklist to separate a descriptive headline from a causal result:
- Does the analysis name the aid policy and outcome?
- Does it separate published tuition, net tuition, grants, and loans?
- Are the institutions, student groups, and study years stated?
- Are prices adjusted for inflation, with the deflator and base year given?
- Is aid measured as offered, awarded, or disbursed, and per which student group?
- Does the study use a comparison group or another strategy to estimate causation?
- Does it account for state funding, enrollment, institutional type, and other relevant changes?
- Are uncertainty, limits, and differences by sector reported?
If several answers are missing, treat the result as descriptive rather than causal. That does not make the data useless. It means the conclusion should be narrower.
Key takeaway: A strong claim is specific about where, when, and how an effect was measured.
Conclusion
To make sound comparisons, keep published tuition and net price separate, distinguish grants from loans, adjust for inflation, and inspect how aid and enrollment are measured. Then ask whether the study identifies a causal effect or only reports that aid and tuition moved together. Those steps turn a crowded set of statistics into evidence that can guide decisions.
Frequently asked questions
These answers summarize the core terms and cautions in the tuition-and-aid debate. They are brief by design, but each depends on the same discipline: identify the price, aid measure, institution group, and period before drawing a conclusion.
What is the Bennett hypothesis?
It is the claim that colleges may raise tuition when students receive more financial aid, because aid can increase their ability to pay.
Does federal student aid always cause tuition to rise?
No. Evidence depends on the aid policy, institution type, time period, and research design. A rise in aid and tuition at the same time does not establish causation.
What is the difference between tuition and net tuition?
Tuition is the published charge for a defined student group. Net tuition is tuition after grant aid is subtracted. Loans are not grants and do not reduce the price in the same way.
Should student loans count as tuition discounts?
No. Loans provide money to pay costs, but they must be repaid. Treating them as a discount can make the student’s price look lower than it is.
How can I tell whether tuition rose in real terms?
Compare prices in constant dollars using a stated inflation measure and base year. Nominal prices alone do not show how tuition changed relative to general inflation.
What data can I use to study college tuition?
IPEDS provides institutional data on tuition, enrollment, and other measures. College Board’s Trends in College Pricing and Student Aid offers broader context. Check definitions and coverage before combining sources.
Why might two reports show different tuition figures?
They may cover different years, college sectors, student groups, or price measures. One may report published tuition, while another reports net price after grants.
What makes a study of the hypothesis more convincing?
A strong study identifies a specific aid change, uses a credible comparison group, states its assumptions, and tests whether results hold under alternative choices.
Can a tuition increase still leave students better off?
Yes. If grant aid rises more than tuition, net tuition can fall. But averages may not describe every student’s actual price.
Does evidence from for-profit colleges apply to public universities?
Not automatically. Sector differences matter, so findings from one group should not be generalized to another without supporting evidence.
(This article was written by one of our staff writers, Kevin Marlowe. Visit our Meet the Team page.)
