Monday, August 21, 2017

Should the government raid university endowments? (Michael Simkovic)

Vanderbilt Tax Professor Herwig Schlunk wants the federal government to tax university endowments, preferably out of existence.  He writes:  “In the best of all possible worlds, the federal government could and probably should . . . confiscate[e] all private university endowments . . .”

Toward that end, Schlunk recycles arguments that were discredited years ago.

Professor Schlunk is famous for asserting that law school is a bad investment.  Schlunk’s bold claim—based on back of the envelope calculations and highly unscientific website surveys—was popularized by the Wall Street Journal and echoed by sympathetic media outlets.  Peer reviewed research by labor economist Frank McIntyre and me—using high quality nationally representative government data and well-established econometric techniques—subsequently demonstrated that Schlunk was mistaken. (See here and here).

This post critiques Schlunk’s recent work on endowments for misuse of discount rates, overlooking the importance of educational quality, mismeasuring student earnings and higher education expenditures, selectively targeting higher education, supporting policies that undermine economic growth, and overlooking stark differences between popular votes and political power.

Misuse of discount rates

To arrive at his headline-grabbing law school result, Schlunk relied on some spectacularly unrealistic assumptions.  As Frank McIntyre and I explained four years ago:

“Professor Schlunk’s analysis assumes astronomical discount rates, low earnings growth rates, and zero inflation for thirty-five years. None of these assumptions are empirically or theoretically justifiable.


Most studies [of higher education] by economists have generally used a discount rate between 2.5% and 3%. . . . Compared with the 3% discount rates applied in labor market studies by economists and suggested by the real (net-inflation) costs of financing a law degree . . .  Professor Schlunk applies real discount rates of between 8% and 27%. 


If Professor Schlunk had used comparable assumptions about discount rates to evaluate the value of a college degree compared to a high school diploma, he would have reached the conclusion that few should go to college. Indeed, given a 30% nominal discount rate, whether it makes financial sense to complete high school might be debatable.”


Undeterred, Professor Schlunk once again relies on unrealistically high discount rates and overlooks differences in completion rates, this time to argue that private non-profit universities provide little value when compared to leanly funded, politically vulnerable public universities.  Based on this analysis, he concludes that the federal government should tax universities more heavily than it already does.  Higher discount rates mean that future cash flows have a lower present value.  Thus the value of a lifetime of higher earnings from higher quality education is diminished by choosing a higher discount rate.

Schlunk’s justification for using such high discount rates is that higher education “puts me in mind of income streams I confronted when advising investors in the private equity sector [where] discount rates of as high as 30% were generally applied.”[1]

For the record, peer reviewed research generally finds that private equity returns net of fees are close to or less than those that can be found in the stock market—not remotely close to the 30 percent returns assumed by Schlunk.  (In addition, discount rates are supposed to reflect the weighted average cost of capital, NOT the (higher) returns to equity).[2]  If P.E. investors were applying high discount rates to cash flow projections, this likely means that investors believed that P.E. cash flow projections were over-optimistic.

Overlooking college completion rates

In his latest critique of higher education, Schlunk also overlooks large differences in completion rates.  Four-year completion rates for bachelor’s degrees are almost twice as high at private non-profit universities as at their more leanly funded public counterparts. If one accepts Schlunk’s assumptions of extremely high discount rates, even a modest delay in completion would have a dramatic impact on value.

Overlooking effects of increased educational expenditures and educational quality

Peer reviewed studies that control for differences in student characteristics consistently find that higher expenditures per student lead to significant increases in student earnings and likely contribute to higher completion rates.  (For brief reviews of the literature, see The Knowledge Tax and Populist Outrage, Reckless Empirics; See also here). 

Professor Schlunk overlooks these studies.  

Mis-measuring student earnings and educational expenditures

Schlunk overestimates the difference in expenditures and resources at elite public and private universities, which leads him to over-estimate the earnings premiums necessary for more resource-intensive private education to be worthwhile.  Schlunk assumes incorrectly that all students at elite flagship state universities pay low in-state tuition, when many students at these institutions pay much higher out-of-state or international student tuition.  He overlooks the extent to which expenditures per student at elite public universities exceed in-state tuition because of state subsidies and cross-subsidies from out-of-state students.  He overlooks the extent to which differences in financial aid affect net-tuition—and therefore educational resources and expenditures—at different universities.

The elite public universities that Schlunk presents as controls that he sees as similar to private universities, but without endowments, actually have larger endowments than many private universities.  

To the extent that Schlunk uses earnings data—again from non-representative website surveys—he focuses on median rather than mean earnings, which very likely reduces his estimate of the value of higher quality education to students or society.  He does not control for differences in student characteristics. It is unclear whether he takes into account differences in subsequent graduate school attendance. 

Selective targeting of higher education

Schlunk’s narrow focus on universities is peculiar: his critique of charitable contribution deductions as “inefficient” and “undemocratic” applies even more forcefully to other charities, such as many churches, healthcare organizations, and especially think tanks–which he does not mention. 

Many think tanks are officially recognized as 501(c)(3) educational organizations, notwithstanding the fact that they educate no students, confer no degrees, primarily produce advocacy with little resemblance to scientific inquiry, offer little protection for their research staff from politically motivated firings, and function as de facto corporate lobbies.  Think tanks have collectively attracted billions of dollars in tax-deductible charitable contributions. 

Undermining economic growth

Schlunk believes that the federal government should raid university endowments to pay for its other priorities.  As Schlunk puts it: “any expenditure made by the federal government presumptively reflects the will of all 321 million of us.”

None of these expenditures stand to generate as much economic growth as investment in education.  Government investment in education pays for itself in higher future tax revenues and lower burdens on social insurance programs.  Only around 3% of federal spending and 3% of GDP is devoted to higher education.   Extremely high returns on investment in higher education suggest underinvestment

Overlooking differences between popular votes and political power

Schlunk’s view that our national system of representative government reflects popular democracy is a perspective with which those familiar with the electoral college, the U.S. Senate, and political geography might take issue.  There can be—and currently are—stark differences between popular votes and political representation at the federal level.


[1]  In his recent article, Schlunk presents examples using discount rates ranging from 8 to 10 percent—still too high for higher education earnings premiums, but progress from his earlier article.  Nevertheless, Schlunk references his earlier claims regarding 30 percent discount rates and argues that 8 to 10 percent discount rates are “modest” and even “indefensibly low.” He never discloses that these “indefensibly low” rates are extremely high compared to those used in the established labor economics literature and the actual real (net-inflation) cost of student loan financing.

[2]  The cost of equity financing is usually much higher than the weighted average cost of capital.  Because P.E. transactions (and individual investments in higher education) are mostly debt-financed, the cost of debt is far closer to WACC than the cost of equity.  Moreover, Schlunk’s claim that higher education is undiversified (and therefore risky and deserving of a high discount rate) is implausible in the context of a large university, donor, or government funder.


Update 8/24/2017:  The title of this post was updated to more clearly reflect the topic of the post.

Guest Blogger: Michael Simkovic, Legal Profession, Ludicrous Hyperbole Watch, Of Academic Interest, Religion, Science, Student Advice | Permalink