Today we are once again treated to a declaration that there is simply no way the crazy Bennett Hypothesis — the theory that student aid helps fuel college price inflation — is true. This time, the end-all-debate pronouncement comes from David L. Warren, president of the National Association of Independent Colleges and Universities, who cites three apparently definitive reports as proving aid is not “driving up” costs.

Aside from the problem that the argument is really that aid fuels price increases rather than driving them — the aid is the gasoline, the colleges the car drivers — what do the studies offered by Warren really tell us?

First is the 2001 federal report everyone who wants to declare the Bennett Hypothesis dead loves to cite: “Study of College Costs and Prices, 1988–89 to 1997–98,” from the National Center for Educational Statistics. As Warren accurately cites, the report does say:

Regarding the relation between financial aid and tuition, the regression models found no associations between most of the aid packaging variables (federal grants, state grants, and loans) and changes in tuition in either the public or private not-for-profit sectors.

But, then, it also says this:

[T]here are considerable data limitations in these models: for example, the availability of only one year of financial aid data and a lack of comparably recent financial data (especially for private not-for-profit institutions). IPSFA data on loans include all sources of student loans; federal subsidizedand unsubsidized, institutional, and private loans cannot be disaggregated. In addition, the IPSFA aid variables focus on the packaging of various forms of student aid in terms of the percentage of students receiving aid and the average amount received, and therefore cannot be used to explore the possibility of a revenue interaction at the institutional level between federal aid and institutional aid. Due in large part to the accounting standards used by the institutions themselves, information on financial aid collected through the IPEDS system for the available years is incomplete, especially regarding student loan volume, which cannot be isolated from tuition revenue in the IPEDS Finance survey data. Finally, financial data such as instruction expenditures cannot be isolated to undergraduate students, making any comparison with undergraduate tuition inexact.

Essentially, the report contains a regression based on a change in student aid for just one year and can’t adjust for a whole bunch of important things. In other words, it tells us little and in no way closes the door on Bennett.

Next, Dr. Warren cites a February 1998 commission report in which the commission purports not to have found any evidence that student grants effect college prices, and no “conclusive” evidence that loans enable rising prices. Then again, the Commission did no meaningful empirical analysis of the question, and as dissenting member Francis McMurray Norris objected, “issues such as tenure, cost and value of research, duplication of facilities, teaching loads, and relationship of student loan programs and rising costs have not been addressed.” [Italics added]

Grounds for putting the Bennett Hypothesis in a pine box? Hardly.

Finally, Dr. Warren cites a 2011 GAO report that looked only at the effect of an increase in the federal student loan limit for first- and second-year students, and only tracked three years of prices and enrollment. It concluded that enrollment and prices rose at rates generally consistent with recent “prior years.” Of course, looking at the effect of such a narrow change in overall aid over such a short time period without controlling for myriad variables that impact prices tells us basically bupkis.

The fact is that several empirical studies do show student aid enabling schools to raise their prices, and I have listed many of them. It is also the case, as most studies point out, that it is very difficult to definitively isolate the effects of aid when so many factors — from school type to student characteristics — are in play. That’s when basic logic also has to come in: People in colleges are like everyone else, and will be happy to take more money if it’s available. Aid makes it available.

One thing that cannot be supported is insisting, as Mr. Warren does, that we know for certain there is no connection between student aid and rising prices. That is something that truly has been disproven.

Andy Matthews, president of the Nevada Policy Research Institute, offers an astute, concise analysis of the Romney campaign’s education plan. In a post titled “Great ideas… from the wrong level of government,” Matthews writes:

When you’re considering the merits of a policy proposal, there are three important questions to ask. One, is the policy constitutional? Two, is this the lowest level of government that can enact the policy? And three, will the policy work?
[…]
If he wins in November, let’s hope Gov. Romney realizes that, while his education plan would be great for a governor to implement, it’s not the right one for a president to push — for constitutional as well as practical reasons.

I offered much the same advice to President G. W. Bush when he was advocating private school choice at the national level, based on education policy precedents from around the world and across history. Those lessons still apply.

I want to thank Kevin Welner for giving the Cato Institute’s education people some pub, and I’m heartened that he thinks our ideas have a lot of influence with the Romney folks. As far as I can tell, though, they don’t. Why? Because if they did, Gov. Romney wouldn’t be offering federal open-enrollment, voucher, or even “neovoucher” proposals. He would be calling to get the federal government out of education, which is exactly what I and my colleagues have long advocated.

Do we like school choice? Absolutely, because logic and mounds of evidence strongly suggest that it works. But that does not mean we want the federal government to impose or “incentivize” it.

There are myriad reasons for this, even though it requires resisting the powerful temptation to have the federal government impose a policy we like in one fell swoop rather than going through the hard work of having individual states and districts adopt it.

First and foremost, we oppose federal involvement because the Constitution doesn’t give Washington authority to meddle in education outside of its 14th Amendment duty to prohibit state and district discrimination, and its jurisdiction over federal lands and DC itself. To press for what we know to be unconstitutional just because it would be politically easy would be to subvert the very rule of law, that which protects us from the arbitrary — and as the Founders understood, very dangerous — rule of men.

But the reasons for keeping Washington at bay are not simply legal or to avoid an oppressive dictatorship.

The fact of the matter is that no one person or group of people — even those of us at Cato’s Center for Educational Freedom — are omniscient, or even close to it. There are lots of things we don’t know, and we all make mistakes. That’s why it is never wise to give broad authority to a central government, even if it is utterly benevolent. If it does something wrong everyone goes down, and the human tendency is to do lots of wrong things. It is this understanding that backs the “laboratories of democracy” concept of states. Individual states can try different ideas, but others are free to steer clear of those that fail and run with those that work.

But aren’t state governments and districts prone to the same human failings as Washington, not to mention the same special-interest driven politics, misalignment of political and educational incentives, etc.?

They certainly are, which is why education policy people at Cato support school choice generally and education tax credits — which crucially do not involve public money — specifically. Given the numerous reasons that government fails, as well as the need for the specialization, competition, and innovation that government quashes, we understand that education would work much more effectively if government didn’t control the schools. But it is far better to let fifty states control their own education systems — including getting to experiment with their own school-choice delivery mechanisms — than to empower the central government to dictate one policy for all.

Yes, education analysts at Cato like school choice. But we’re also big fans of federalism and the Constitution, and when it comes to federal policy those things must come first.

C/P from the National Journal’s “Education Experts” blog.

I want to thank John Kirtley for his gracious reply to my criticism of his policy guidelines. He has spilled a tremendous amount of blood, sweat, and tears on the ground fighting to establish, protect, and expand the largest private school choice program in the country, and I, quite simply, have not. I think this kind of policy debate is good for the health of the school choice movement, however, so on it goes …

Andrew Coulson posted a response to many of John’s points, but I think some areas deserve an expanded treatment. One of the primary issues in our discussion is centralization vs. diversification of scholarship organizations. I did not claim there was a “mandated” monopoly, which I take to mean government-mandated. Step Up for Students is, however, the only active scholarship organization in the state. It became the sole scholarship organization through hard work and good performance. John mentions Microsoft in his defense of market dominance, but Microsoft never fully monopolized any product or service. There is, however, a literal monopoly of the education tax credit system that was produced and is maintained by problematic provisions in the credit program that create a very high barrier to entry. The structure of the education tax credit in Florida all but ensures a monopoly in the education tax credit program.

For the first six years of the program, scholarship organizations were required to spend 100 percent of the credit funds they raised on scholarships. In other words, they had no money for overhead, which made establishing and running a scholarship organization difficult and expensive … a non-profit would need to seriously cannibalize its established charitable funding, likely already committed, and/​or fundraise along two separate tracks for administrative and scholarship funding.

To put this in context, Charity Navigator, which rates non-profits, considers it acceptable for a charity to spend close to one-third of its revenue on non-program expenses. Even the 4‑star rated Inner-City Scholarship Fund spends over 13 percent of its revenue on overhead expenses.

Scholarship programs, especially ones with relatively high compliance costs such as requiring detailed checks on a family’s income, require significant but entirely normal overhead spending. Furthermore, local scholarship organizations in a decentralized system act as more than a high-volume processor of financial applications. They act as community organizations that consider the needs and struggles of individual families and children, which requires spending more time and resources on each family. A 10 percent overhead allowance is eminently reasonable, indeed, within the bounds of best practices for such charities. Denying any overhead to non-profits ensured that few charitable organizations would be capable of fundraising and processing scholarships under the law.

Exacerbating this problem, scholarship organizations are not allowed to target the use of scholarship funds they raised to particular kinds of educational environments. What this means is that a non-profit would have to a) cannibalize money raised from other sources and for other purposes, and b) possibly fund educational environments that directly conflict with their conscience, mission, or best judgment. For instance, a Catholic charity would be required to fund an atheist, Wiccan, Protestant fundamentalist, Lutheran, Islamic, or any other school which met the basic requirements of the legislation. Even a non-sectarian scholarship organization is required to issue scholarships to any school, regardless of quality, as long as it meets the basic legal requirements.

In addition, the Florida tax credit applies only to corporate taxes, the vast majority of which are paid by large corporations based outside of the state of Florida. This means that fundraising is relatively difficult and time-consuming, not to mention extremely volatile, as large corporations shift revenue and expenses to minimize their tax burden year to year. It can take two years for a large corporation to begin disbursing funds after first being solicited. And fundraising requires expensive out-of-state traveling.

The corporate-only credit acts as an additional barrier to entry that grows over time and with centralization. Step Up for Students entered this constrained market efficient and well-capitalized, and spent the next decade bringing on the biggest corporate taxpayers in Florida as donors. A new entry into the credit scholarship realm would need to raise very substantial funds for fundraising for years before they saw a return in credit donations. Even should the very high quality of Step Up decline in the future, its relationships with the biggest donors, scale, and general dominance would pose a very formidable wall to climb for any non-profit. Indeed, it is far more likely that the state government would intervene long before any non-profits entered the market to impose the discipline of competition.

With extremely high start-up costs, low return for many non-profit missions, a fully established monopoly, and no profit motive or access to investment funding, the Florida education tax credit scholarship organization opportunities are all but nonexistent under current law.

Among the many things that were wrong or at least grossly misleading in Stephanie Saul’s hit piece on education tax credits is her claim that credits were invented in the late 1990s as some underhanded work-around for political and constitutional problems with vouchers. Here’s Saul’s creation myth for tax credits:

Vouchers … were unpopular among many voters and legislators, and several state courts had found them unconstitutional. Proponents decided to reposition themselves, and in 1997, Arizona’s Legislature adopted the first tax-credit scholarship program.

Credits are much more popular and legally protected than vouchers, but these characteristics are the result of important differences in function and principles. Moreover, the credit concept goes back more than forty years (more for deductions) and dominated the school choice landscape for over a decade. What changed in the late 1990s was an equity-focused twist on the concept, bringing the benefits of education tax credits to families without significant tax liabilities.

Reporters often mistake this relatively recent policy innovation for the origin of the credit approach and overlook credit advantages in policy principle and function in favor of the practical/​cynical drivers of support for the policy. Sadly, Saul and many other reporters skip serious research in favor of collecting anecdotes that help drive their predetermined narrative.

So, a bit on the origins of education tax credits that Saul and even many school choice proponents miss entirely. (Note: Much of this material comes from my dissertation, which relies heavily on an excellent chapter in The Future of School Choice by Martin West, 2003.)

Despite the pedigree of vouchers as the first school choice proposal and the preferred mechanism of Milton Friedman, and despite the rise of progressive voucher plans in the late sixties and early seventies, education tax credits were arguably the dominant school choice policy from the 1970s through the mid-1980s.

Education tax credits first entered the federal agenda during the Nixon administration, during which he assigned the problem of the struggling private school sector to be studied by a panel within the President’s Commission on School Finance. The attention came in response to growing financial problems among the private school sector—in particular, among Catholic schools pressed by the rapidly declining numbers of priests and nuns who had traditionally provided a quality, low-cost labor pool.

Voucher policies, despite their libertarian lineage, were typically framed in progressive equity terms. Tax credits were typically been framed as a fiscal matter, a way to bring tax relief to the middle class. In taking up a proposal for a federal education tax credit, Nixon explicitly framed the issue as a matter fiscal prudence in support of middle-class families, placing “particular emphasis on the dire fiscal consequences should the nonpublic sector be allowed to collapse” (West 2003, 161). A subsequent panel report released in 1972 included a recommendation for a Federal income tax credit for private school tuition, but legislation would not be introduced until four years later. In 1976, a maximum $1,000 education tax credit for tuition at any school from elementary school to college and vocational school was defeated in the Senate by a surprisingly close 52–37 vote that included support from over a third of the Democrats and half of the Republicans. The fiscal argument for tax credits lacked the power it might hold at the state level, as the overwhelming burden of financing education was shouldered by state and local governments.

Efforts at the state level during this period were sporadic and uncoordinated. The attention of school choice supporters was directed toward federal policy and nothing approaching a school choice “movement” had yet developed. Many of the state-level efforts throughout the 1970s attempted to enact tax credit and voucher policy through ballot initiatives and referenda (Catterall 1984).

The tax credit issue came back at the federal level during the Carter presidency with a bill introduced amidst discussion of a similar proposal for higher education. Although the bill passed narrowly in the House, it was defeated 56–41 in the Senate and faced a veto by Carter in any case. It was during this battle over k‑12 tax credits that opposition from the government education establishment coalesced and began to flex to full extent its considerable power. Literally hundreds of lobbyists descended upon Capitol Hill to argue the interests of the recently formed National Coalition to Save Public Education and the National Education Association (West 2003).

The school choice issue was a relatively stable fixture on the national education agenda during the 1980s, first in the form of tax credits and shifting back to vouchers again in the mid-1980s. The hopes for school choice success at the national level peaked with Reagan’s arrival in the White House. Reagan and the Republican Party gave strong and explicit support for education tax credits throughout the 1980’s—with tax credits, but not vouchers, mentioned specifically in the Republican Party platforms of 1980, 1984, and 1988. Tax credits were the primary focus for the conservative policy world as well. The Heritage Foundation, an influential conservative think-tank, published a book in 1983 assessing the failures and accomplishments of the Reagan administration that had an education chapter proposing tax credits to help the middle-class supplemented by vouchers to help low-income parents (White 1983).

The Supreme Court has waded back into the affirmative action thicket, taking up the issue of the proper role, if any, of race in college admissions, in the case of Fisher v. University of Texas at Austin, which it will be hearing this fall, likely in October.

Abigail Fisher, who is white, was denied admission to the University of Texas at Austin even though her academic credentials exceeded those of many admitted minority applicants. She challenged UT-Austin’s consideration of race in selecting its incoming freshmen but lost before the district court in light of the Supreme Court’s 2003 ruling in Grutter v. Bollinger.

In Grutter, a divided Court held that using race as a factor (but not one tied to a set number of points or quotas) was justified in the name of diversity. But UT-Austin’s admissions program treats race in a different way, and gets different results, than did the admissions program Grutter upheld at the University of Michigan Law School.

The Fifth Circuit panel nevertheless affirmed the district court, but Judge Emilio Garza specially concurred to say that while he was bound by Grutter, that decision seemed to conflict with other precedent and with the Fourteenth Amendment’s Equal Protection Clause. The Fifth Circuit then voted 9–7 against rehearing the case en banc (before all judges on the court), over a sharp dissent from Chief Judge Edith Jones that emphasized how the ruling would allow states to play fast-and-loose with Grutter’s narrow-tailoring requirement.

Now before the Supreme Court, Cato filed an amicus brief supporting Abby Fisher and arguing that the Fifth Circuit showed blind deference to UT’s policy rather than the constitutionally demanded strict scrutiny. The lower court explicitly declined to evaluate the merits of UT’s decision to consider race, instead assuming the institution’s good faith. Under this rule, a public university’s mere assertion of a diversity interest, irrespective of the university’s precise circumstances or actual motivations, trumps an applicant’s right to be treated as an individual rather than a racial specimen.

We also point out that the Fifth Circuit ignored the Supreme Court’s requirement (from the 1989 case of City of Richmond v. J.A. Croson Co.) that the government demonstrate a “strong basis in evidence” for racial classifications in order to smoke out the illegitimate motivations that can underlie such schemes. That is, Grutter upheld Michigan’s racial preferences because the school showed that minority enrollment would have plummeted to almost nothing without them, while UT had already achieved real diversity (beyond even that created by Michigan’s preferences) with a race-neutral law that guarantees admission to anyone graduating in the top 10 percent of a Texas public school.

Finally, we note that even if UT could show that racial preferences were necessary for some legitimate reason, its chosen paradigm for applying such preferences is arbitrary. For example, UT justifies preferences to Hispanics by pointing to the need for a “critical mass” of such students, even as it denies preferences to Asians, who comprise a smaller portion of the student body.

We urge the Supreme Court to reign in UT’s unbridled use of race in admissions decisions and take an important step toward ensuring that young Americans are judged on their qualifications and character rather than their skin color.