The Economics of For-Profit Higher Education

Articles, Microeconomics / Behavioral Economics, culture, Economics, education, entrepreneurship / 0 Comment

Tyler Cowen and Sam Papenfuss released this paper on for-profit education on July 23, 2009. This paper has been reformatted for stylistic reasons and some content, including footnotes, has been removed. Please download the PDF to read the paper as it originally appeared.

Abstract

We consider why some educational institutions choose for-profit corporate status. For-profit educational institutions are prominent where the content of instruction is well-defined, high quality research is of little complementary value, students use education for learning rather than certification, and there are independent means of certifying student quality, such as vocational tests. We consider the explanatory power of some hypotheses that might explain these facts, involving government subsidies, donations to non-profits, agency problems, and the economics of producing reputation.

I. Introduction

For-profit higher education has grown rapidly in the United States over the last twenty years. Examples include vocational schools, technical training, in-house corporate education, the DeVry Institute, Phoenix University, and other new for-profit colleges. In addition, a number of developing countries, most prominently the Philippines, have experimented with for-profit higher education in the past. Adam Smith, writing in 1776, associated for-profit education with high levels of instructional quality, and claimed that endowed, nonprofit education resulted in shirking and poor teaching.

In spite of market experience, for-profit higher education has received little direct attention from economists. The literature treats the non-profit status of higher education as the state of affairs to be explained. We view this presupposition as question-begging and focus on what determines the distribution of for-profits and non-profits in higher education. Unlike many non-profits, for-profits do not offer high-reputation liberal arts education but instead specialize in teaching well-defined vocational skills.

Section II of the paper presents some examples of for-profit higher education and outlines their market niches. Section III consider some reasons why corporate status matters and considers some explanations of the patterns in the data, focusing on the relative role of subsidies, donations, agency problems, and the nature of reputation as a public good. We will see that all four factors play some role but none can explain all of the data or provide a fully satisfactory theoretical account.

Winston (1999) provides a survey of issues relevant to the non-profit status of many colleges and universities; see also Ruch (2001), Morey (2004), and Breneman, Pusser, and Turner (2007). In economic terms non-profits are institutions bound by a non-distribution constraint for their profits. Non-profits do not have owners with rights to residual income, and there are no shares which provide for both control and claims to profits. Net revenue must remain in the corporation, although some revenue will be distributed implicitly in the form of perks to managers, board members, and employees. As with for-profits, however, non-profits must cover the costs of their operations if they are to survive and many in fact earn very high returns. Pauly (1987) provides a dissenting perspective which minimizes the practical differences between for-profit and non-profit institutions. With regard to state and private schools, we treat the difference as one of degree, rather than kind. We view state schools as run as nonprofits, but having state-appointed officials on their boards. It is not even the case that state schools necessarily receive more subsidies than their “private” counterparts; governmental subsidies provide significant portions of the budgets of schools such as Johns Hopkins and MIT.

II. For-profit education

Technical and vocational training

For-profit technical training is common for business courses, barber and beauty schools, nursing programs, vocational and technical institutes, and correspondence courses (Lynch 1992, p.302). Berlitz International offers for-profit instruction around the world in a wide variety of foreign languages. Although Berlitz offers translation services and publishes guidebooks, eighty percent of their revenue comes from language training, through approximately five million lessons a year. Other work-related skills, such as computer programming and specific software classes, are taught frequently on a for-profit basis.

In-house corporate education

Other forms of higher education are provided within corporations. On-the-job training is hard to quantify since much of it is informal learning-by-doing or integrated into work routines. Nonetheless a survey by Training magazine estimated that more than 47 million workers in the United States received formal corporate training in 1994, at a direct expense of $50 billion. Informal training probably runs several times this amount (Hood 1996, p.68). Lynch (1994, p.12) estimates corporate training expenditures at 1.8 percent of the total wage bill in the United States.

One estimate found more than 1,600 corporate “colleges” in the United States. At “Hamburger University,” run in Illinois by McDonald’s corporation, trainees learn how to run a McDonald’s franchise, which includes learning how to cook hamburgers and french fries, how to enforce standards of cleanliness, and how to attract customers. Hamburger University offers a tailor-made education in the managerial skills needed for running a fast food franchise. The school trains 7,000 individuals a year. Ford, Disney, Motorola, and Dana Corporation run institutions of a similar nature.

Apprenticeships at for-profit corporations provide an indirect means of purchasing training in some skill or vocation. The apprentice typically accepts lower wages in return for a package which combines a job and training. Apprenticeships are rare in the United States but they are much more common in Europe, especially Germany, which has had measured rates of apprenticeship of over sixty percent (Blanchflower and Lynch 1994, p.240, Soskice 1994, p.26). Most of these apprenticeships are with for-profit companies rather than with non-profit institutions.

For-profit higher education in the United States

The DeVry Institute is the most prominent for-profit institution in United States higher education. DeVry is a for-profit, publicly traded company which dates from the 1970s and went public in 1991. The company is traded on the New York Stock Exchange and has private shareholders and other characteristics of a private corporation. DeVry receives no public funds, and student tuition fees account for approximately ninety percent of revenues which totaled over $1 Billion in 2008.

Approximately 68,000 individuals are taking DeVry programs at numerous campuses in the United States and Canada. Full-time undergraduate students pay from $56,000 to 59,000 for a four-year Bachelor’s degree. The DeVry institute offers a no-frills education oriented towards specific technical skills with high labor market returns. The school concentrates in fields such as electronics engineering, computer science, business, accounting, and telecommunications. A related branch of the school, the Keller Graduate School of Management, is currently serving several thousand students seeking MBAs. Numerous other for-profit institutions of higher education operate in the United States, including Colorado Technical College, Laboratory Institute of Merchandising (New York), School of Visual Arts (New York), Bassist College (Oregon), Huron University (South Dakota), and Strayer College (Northern Virginia), all of which are currently accredited. The largest is Phoenix University, which now has about 300,000 students.

For-profit education in the Philippines

For-profit higher education started in the Philippines in the early part of this century and blossomed after the Second World War. Later in the 1980s, the Marcos regime of martial law imposed restrictions and unfavorable taxes on proprietary educational institutions; until that time, for-profit colleges and universities competed on a nearly level playing field. We focus on the earlier period – the 1960s and 1970s – when for-profits competed against non-profits with few legal hindrances.

The private sector has traditionally covered most of Filipino higher education, often more than ninety percent (Zwaenepoel 1975, pp.162-4). In 1969 the Philippines had 36 private universities and 559 private colleges, serving 573,094 students at the graduate and undergraduate levels. 274 or 49.37 percent of these institutions were run as for-profit corporations, 281 or 50.63 percent, were run as non-profits, and another forty could not be classified due to limitations in the data (Zwaenepoel 1975, pp.71-2). The for-profits covered roughly three-fifths of all students receiving higher education in the Philippines (Miao 1971, p.207). Many of the for-profit colleges have been publicly held corporations traded on the Manila Stock Exchange (Geiger 1986, p.58). One observer noted: “The Filipino passion for education has become a lucrative market and private education is dominated by schools, colleges and universities which operate as profit-making stock corporations and which actually declare dividends on their stock. These include the largest universities in the country.”

The Philippines is not the only developing country to have relied upon for-profit higher education. We also find significant numbers of for-profit institutions in Indonesia, Malaysia, and Turkey, at various points in time.

Overall patterns in the data

Two primary features characterize the observed educational for-profits. First, for-profits tend to specialize in highly practical or vocational forms of training. For-profits are especially prominent in areas where student performance can be measured by a relatively objective, standardized test. Nonprofits, in contrast, have a stronger presence in the liberal arts, although they are by no means restricted to that arena.

Second, for-profits offer products of lower academic reputation, relative to non-profits. Educational for-profits tend to sell their services at a relatively low price to students who otherwise would not seek higher education or who are marginal applicants. Students who wish to become automobile mechanics or beauticians are more likely to patronize for-profits, as are older students who wish to learn some specific skill, such as accounting or computer programming.

The relevant distinctions between for-profits and non-profits depend on which part of the non-profit sector we examine. The difference is most marked when we examine research institutions of high academic reputation, such as Harvard, Princeton, and Yale. These schools serve the students with the strongest academic records and hire faculty with the strongest research performance. The difference is less marked when we compare for-profits to community colleges or mid-level regional teaching schools. The community colleges, for instance, resemble for-profits by having relatively loose admissions standards. Nonetheless even in these cases the for-profits are more heavily specialized in vocational education and place less emphasis on the liberal arts. Non-profit business and law schools offer a kind of vocational training, and in that sense they resemble for-profits; non-profit business schools of low academic reputation, of all parts of the non-profit sector, are perhaps closest to for-profits in their orientation. Few if any non-profits, however, offer vocational training for barbers and beauticians, to name just a two areas of for-profit specialization.

A comparison of for-profit and non-profit institutions in the Philippines bears out many of the differences noted above. Filipino for-profits tend to charge lower fees, specialize in education of lower academic reputation, spend less on capital equipment, and serve students who plan on pursuing vocational careers or taking a standardized vocational test upon graduation.
We observe similar tendencies in the DeVry Institutes in the United States, but the large number of Filipino for-profits, especially if we look back in time a bit, allows for a more systematic comparison.

Unlike Filipino non-profits, the for-profits typically did not have entrance examinations, and accepted any student who has completed a secondary education and can pay the relevant fees (Zwaenepoel 1975, pp.163-4). From a survey of Manila institutions, the for-profit institution had an average student to fulltime faculty ratio of 27:1, whereas the non-profit religious institutions had an average ratio of 19:1 (Miao 1971, pp.71-2). For-profit institutions tend to invest in classrooms to accommodate large enrollments, rather than investing in library facilities, book holdings, or laboratory facilities. Furthermore, Filipino for-profit institutions tend to limit their class offerings to low-cost, labor-intensive classes, such as teacher education and commerce (Zwaenepoel 1975, pp.322, 342, 348, 587). As of 1970, nonsectarian institutions (typically for- profits) spent four percent of their total budget on sites, equipment, and facilities, whereas sectarian institutions (typically non-profits) spent a much higher 12.41 percent (Isidro and Ramos 1973, p.157). As of 1971, for-profits held an average of 2.58 books per student, whereas non-profits held an average of 8.9 books per student (Zwaenepoel 1975, pp.347-8).

Filipino for-profits also produce a different kind of education. Students from for-profit institutions tend to take standardized vocational exams in much greater number, although they pass them at a lower rate. These facts reflect both the vocational emphasis of for-profits as well as the lower academic reputation of their students. Based on a sample of institutions from the Manila area (from 1963 and 1968), students from nonprofit religious institutions pass these standardized tests at an average rate of 38 percent, whereas students from for-profit institutions pass the same tests at a lower rate of 18 percent. For-profits, however, produce a much greater number of students taking the tests, and therefore pass a much greater number of students through the tests. Students at for-profits are approximately ten times more likely to take the tests. Adjusting for the lower pass rate from for-profits, the for-profits are putting about five times the number of students through the tests as the non-profits, even though for-profits educated no more than three-fifths of all Filipino students at the time (Miao 1971, p. 207).

American experience with the DeVry Institute supports this characterization of for-profit education. DeVry maintains a tough curriculum which focuses on practical skills and vocational training. The school has a good record of getting jobs for its students, but it does not pursue academic prestige, as defined by the standards of research universities. Few joint products are provided with the education. DeVry does not have campuses in the traditional sense, but rather holds its classes in office buildings. The school funds only those sports teams and clubs that are directly related to job placement, and they do not encourage their faculty to do research or publish; many faculty do not even have graduate degrees. Faculty work year round and have very high teaching loads (Glass 1995, Spencer 1995).

III. Why does corporate status matter?
We can think of a few relevant hypotheses as to why corporate status might matter for an education institution; we will survey each in turn.

Subsidies

Subsidies discourage many institutions from achieving for-profit status. In the United States non-profits are exempt from corporate and property taxation. Donations to non-profit organizations are tax deductible but donations to for-profits are not. Many non-profits have the ability to issue tax-exempt bonds or enjoy lower postage rates. In addition, for-profit educational institutions often have faced legal or collusive barriers from governments or from other educators. For-profit education in the Philippines declined when the Marcos regime of martial law instituted unfavorable regulatory and tax treatment. In the contemporary United States, for- profit institutions of higher education have difficulty receiving accreditation; prevailing accreditation bodies are controlled by educators who believe in non-profit education.

The evidence does suggest that subsidies matter. In the Philippines, where educational for-profits have been most prominent, for-profit colleges and universities paid a relatively- favorable income tax rate of 10%, at least prior to Marcos’s crackdown. The 10% rate compared favorably to other corporate income tax rates, which typically ran from 25% to 35% (Miao 1971, p.211, Zwaenepoel 1975, pp.311-2). The relatively equal tax treatment of for-profit and non- profit educational institutions accounted for some of the prominence of for-profits in the Philippines. For-profits also had no special difficulty receiving accreditation in the Philippines.

Nonetheless subsidies are not the only reason why many educational institutions choose non-profit status. Subsidies matter most for those segments of the non-profit market that are closest to for-profits in some of their operations, such as local community colleges or nonprofit business schools of low academic reputation. Subsidies have increased the size and scope of non-profit institutions in these areas. Subsidies, however, are less relevant for explaining the end of the market with high academic reputation, such as Harvard University or Middlebury College, a high quality teaching school. The fundamental question remains why educational for-profits, without subsidies, compete successfully with the local community college but not with Harvard or Middlebury. In other words, the subsidies hypothesis does not explain the cross-sectional variation in the data.

Furthermore, non-profit educational institutions predated the existence of personal and corporate income taxes and predated governmental subsidies to the non-profit form. Lock-in effects and path dependence do not suffice to explain the persistence of the non-profit form. Non-profit universities have succeeded in a wide variety of countries and cultures, and it appears that the form is too robust to be explained by history alone. Furthermore, non-profits can switch to for-profit status if the latter is more efficient; many hospitals have made this change in recent times. The switch can go in the opposite direction as well. American medical schools were once largely for-profit, but they switched to non-profit status in the late nineteenth century (Rothstein 1972, 1987).

Donations

The possibility of receiving donations encourages many institutions to choose the non- profit form (Hansmann 1980, 1987, 1990, Fama and Jensen 1983b). In the United States, donations represent approximately fourteen percent of revenues for non-profit institutions of higher education (Okten and Weisbrod, 2000, p.262). Harvard University has a multi-billion dollar endowment and is one of America’s largest institutional investors.

While we can imagine a for-profit that receives donations, most donors would be reluctant to give to an institution whose mission is to enrich the shareholders. The possibility of donations encourages non-profit status, since many colleges and universities have shown they are effective in fundraising on a large scale. Individuals give money to maintain an ongoing connection as alumni, to achieve status in the philanthropic community or in their local country club, or to buy “packaged” goods, such as admission for their children, good seats at college football games, contact with renowned faculty, invitations to exclusive parties, and so on.

Like subsidies, donations play a clear role but do not tell the entire story. Many non- profit educational institutions receive few donations and have little or no endowment. In the United States many higher educational institutions fund themselves primarily through tuition. In other countries, such as Brazil, Japan, and Chile we find considerable numbers of private educational non-profits, even though donations are not usually an important source of funds. Donations help explain the non-profit status of Harvard and Yale, but they do not account for many other non-profit educational institutions.

In the Philippines, when for-profits have enjoyed a nearly level playing field, non-profits nonetheless held nearly half of the market, despite receiving virtually no donations. Although most educational non-profits are run by the Catholic Church, tuition remains their primary source of revenue. Of the private Catholic universities in the Philippines (most of the non-profits are sectarian and Catholic), only two had endowments at all. Ateneo de Manila University had an endowment of $600,000, and the University of San Carlos had an even smaller endowment of $92,000 (Zwaenepoel 1975, p.146). Five private institutions did indicate additional private sources of income, but these were aimed at scholarship grants rather than for operating expenses. Philanthropy towards universities is almost unknown in the Philippines (Respose 1971, pp.107- 8). Even for the church-related non-profit institutions of higher education, gifts and donations account for only 3.08 percent of total income (Isidro and Ramos 1973, p.149).

Agency problems

For-profit and non-profit educators exhibit different advantages at solving agency problems. For-profits appear to be relatively efficient at monitoring the quantity of teaching when the subject matter is well-defined. Non-profits, in contrast, appear to be relatively efficient at monitoring the quality of research.

Faculty governance may play a central role in this distinction. For-profit institutions are controlled by their shareholders. The issue of control is more complex in non-profits, but faculty play a central role in steering institutional priorities, making hiring decisions, and setting pay raises. The faculty have veto power over many proposed changes in non-profits, but not typically in educational for-profits.

Non-profit faculty governance militates against high teaching loads. Within non-profit schools, the faculty pressure the administration to make overall teaching loads low rather than high. The non-profit form therefore is inefficient when a high teaching load is called for. When the content of teaching is relatively well-defined, uncontroversial, and requires little independent thought, it makes sense to work the faculty relatively hard as defined by the preferences of administration.

Most educational for-profits have teaching loads well in excess of what even non-profit teaching colleges demand. For-profits keep their instructors on a tight rein and extract maximum teaching effort from them, while discouraging research. Part-time instruction is common and many of the instructors are not expected to invest in the university. In essence, the for-profit is specializing in areas where the relevant output is easily defined and measured, such as teaching hours. A residual claimant needs no particular expertise to monitor this variable, nor is faculty governance required. The traditional efficiencies of for-profit firms dominate in these cases.

Similarly, for-profits appear more prevalent when the content of instruction is well- defined and uncontroversial. As discussed above, most for-profits specialize in vocational training, rather than the liberal arts. The body of knowledge for most vocational skills is concrete and well-defined, rather than abstract or controversial. Schools for auto mechanics, beauticians, and computer programmers are examples here. The for-profit has a smaller advantage, or perhaps no advantage at all, in determining what schools of philosophy should be taught, or in deciding whether students should read Dickens or Garcia Marquez. In these cases, the other advantages of non-profit institutions, discussed throughout the paper, become relatively more important and the for-profit form is less likely.

For-profits do not offer their faculty the teaching discretion that is given to Harvard philosophers or faculty at nonprofit teaching institutions of high academic reputations. When the institution specializes in a well-defined instructional product, academic freedom involves costs but no corresponding benefits. Driving schools do not give academic freedom to their instructors behind the wheel, but instead require that they teach skills of braking, signaling, merging, parking, observing traffic signals, and so on.

Note that the relevant agency problem involves the internal enforcement of quality standards, not a demonstration of quality to the outside world. Outside parties arguably can observe the quality of teaching for an institution as a whole. The U.S. News & World Report rankings, for instance, offer publicly available assessments of university quality. To the extent these rankings are controversial, part of the problem is the intrinsic difficulty of ranking in the area, rather than the poor information of outsiders. In any case, the availability of aggregate rankings does not alleviate the internal agency problem. Inside managers must be able to spur faculty and other participants in joint production to pursue quality. If the for-profit has no comparative advantage at this task, the corporate form is more likely to be non-profit.

McCormick and Meiners (1988) make the related argument that the non-profit form is superior for monitoring faculty research. They claim that the research outputs of university faculty often are difficult for outsiders or non-specialists to monitor or measure. Most university research can be evaluated only by peers, colleagues, and other faculty. Again, outside parties can observe the overall research quality of an institution (we all know that Yale faculty are better than faculty at Podunk U.), but the governance structure must support effective internal monitoring to produce the proper ex ante incentives for individual faculty.

According to the McCormick-Meiners argument, faculty governance serves this end. Faculty and administrators receive a large share of their perks and marketability from the reputation of their institution in the external academic marketplace. Both departments and promotion and tenure committees enforce relatively high standards, based ultimately on peer review in the form of journal articles and outside letters of recommendation. Tenured faculty, by insisting on high standards for their tenured peers, raise the reputational value of the franchise they hold. University faculty usually oppose policies that will lower the academic reputation of the university, even when net university revenues would increase.

Agency problems do explain some of the differences between for-profit and non-profit institutions, but the hypothesis is incomplete. Most importantly, the hypothesis takes the current governance differences between for-profits and non-profits as given. It is true that faculty governance plays an important role in most non-profits today. But it is possible to imagine non- profit institutions of higher education without faculty governance, or with faculty governance of a different kind. Charities, for instance, are non-profits, but they are not typically governed by their employees. If faculty governance prevents non-profits from requiring high teaching loads, why do we not observe the evolution of some non-profits without faculty governance? Such hypothetical non-profits might prove effective competitors with for-profits in areas where teaching loads should be high. The very existence of for-profits implies that faculty governance cannot have a decisive efficiency value in all cases.

Similarly, when faculty governance is desirable, why do not for-profits try to replicate those incentives by giving faculty a greater voice in institutional management? Commercial for- profit corporations use employee governance in a variety of manners, including company unions or simply heeding employee feedback for higher-level decisions.

The agency problems hypothesis requires some broader account of the differences between for- and non-profits, and a treatment of which organizational differences are flexible at the margin and which are immutable. The agency hypothesis has micro-foundations only if faculty governance, and its concomitant benefits, can somehow be demonstrated as less costly in non-profits than in for-profits. While this supposition accords with the evidence, we do not yet have a satisfactory theoretical account of why it might be true.

The agency hypothesis also explains only part of the cross-sectional variation between for- and non-profits. It explains why research-oriented liberal arts institutions of high academic reputation do not covet for-profit status. It is less effective in explaining why community colleges, which have high teaching loads and do not subsidize research, choose non-profit status and to answer that question we probably must look to the legal status of community colleges as government-supported entities. Like many of the other hypotheses considered, it explains some of the corners of the distribution rather than the distribution of institutions in the middle of the spectrum.

Reputation as a public good

When reputation is a public good, non-profit institutions may produce that public good with greater effectiveness. Consider Halls of Fame and scientific prizes. The Nobel Prize, like university attendance, provides quality certification. The Nobel Prize is given out by a non- profit committee, rather than being sold to the highest bidder by a for-profit. A for-profit scientific prize would yield less prestige than the Nobel Prize; in dollar terms, Bill Gates may value a Nobel Prize more than did John Hicks. We can imagine that a profit-maximizing version of the Nobel Prize would award it to Samuelson and Arrow, to build up initial prestige, and then sell it to Gates for a high return. The for-profit Nobel Prize would care only about its reputation with its paying customers, not about its reputation with the outside world per se. We therefore find that the most prestigious prizes tend to be awarded by non-profit institutions.

An analogous mechanism may contribute to the non-profit status of colleges and universities of high academic reputation. Faculty governance implies that for-profits and non- profits place different relative weight on reputation and profits. The for-profit selects students and faculty on the basis of how easily their reputational benefits can be captured by shareholders, whereas the non-profit places greater weight on the reputational benefits that are kept by faculty. The for-profit pursues “reputation as valued by students in dollar terms” and the nonprofit pursues “reputation with the external world,” or “reputation as a public good.” In the resulting equilibrium, for-profits achieve lower status.

For-profits pursue only those reputational benefits that they can charge for and convert into profit. They seek to capture as much of the reputational surplus of students and faculty as possible, even if the total reputation of the institution falls. A for-profit version of Harvard, if we can temporarily imagine such a counterfactual, would look at how much potential students and faculty would pay to reap the private reputational benefits of being at Harvard. Non-profits seek out the students and faculty who reflect the greatest reputation back on the university, even if those same individuals are not willing to pay much for their private reputational benefits.

Not all forms of career success – and thus reputational payback to the university – translate into high incomes and high student demands for university services. Training great moral leaders, politicians, and philosophers, for instance, enhances a school’s reputation, but these individuals do not necessarily have a high willingness to pay for university slots, given their relatively low lifetime incomes. Some individuals seek to become Supreme Court Justices, while others shoot for more lucrative positions as firm partner. The non-profit, which is ruled by reputation-conscious faculty, will pursue academic status as an end in itself.

Note that faculty governance need not account directly for the entire reputational difference between for-profits and nonprofits. Reputation often possesses increasing returns to scale or “snowball” properties. Once a given set of institutions has a reputational advantage over others, that advantage may be self-reinforcing. Top faculty and students will be attracted to the institutions with highest reputation, which will in turn support their reputations even further. Consistent with this mechanism, the list of top universities has changed little over decades, whereas there has been considerable turnover in the largest or most profitable for-profit corporations.

The superior reputations of non-profits will feed back into their broader admissions and fundraising strategies. Non-profits can offer potential status and status-related goods to their donors, which strengthens their comparative advantage in raising donations. Furthermore, the non-profits of highest academic quality will adopt exclusive admissions standards but then “sell” admission to the children of the wealthy on a case-by-case basis. Some non-profits will become increasingly driven by the pursuit of reputation and status, whereas the for-profit will drop out of these market segments, which is precisely what we observe.

The hypothesis therefore predicts a segmented market for higher education. Students who seek the highest levels of certification and reputation will attend non-profit institutions, which are run by faculty and use their prestige to raise donations. Students whose quality can be certified by an outside vocational exam do not need the non-profit reputational endorsement. They will pursue the more efficient instruction offered by for-profits.

This hypothesis does not explain why non-profit local community colleges, and other non-profits with indiscriminate admissions standards, can compete with for-profits. These institutions appear to offer little prestige or certification. Nonetheless we do observe some confirming patterns of evidence in other areas. Education for corporate purposes is conducted on a for-profit basis, rather than by non-profit contractors or subsidiaries. If a multinational wishes to instruct its employees in another language, the hypothesis predicts that it will send them to Berlitz. They seek efficiency of instruction and do not need external certification. Similarly, we expect that a small business owner who needed to learn Japanese would prefer Berlitz as well. In the U.S., one sampling from the 1990s found that sixty percent of Berlitz students are employed full or part time and only ten percent were attending colleges or universities. Seventy percent of Berlitz students cited work requirements or relocation as a reason for studying a foreign language.

In addition, older individuals are more likely to purchase for-profit education, whereas the reputation-conscious young will prefer to attend non-profits. Older individuals already have established reputations through their vita (for better or worse), and cannot produce reputation so easily through university attendance. Given their more advanced age, older individuals also face lower benefits from producing reputation. We therefore expect the ambitious young to learn languages, and other skills, in non-profits such as colleges and universities. The old, in contrast, will learn languages by going to Berlitz.

Again, this prediction corresponds to the facts. At least half of all Berlitz students are over the age of thirty (the actual percentage may be higher, as roughly twenty percent of Berlitz students did not give their age). Data from Strayer College, a for-profit in northern Virginia, illustrates a similar pattern. Roughly sixty percent of Strayer students are over the age of thirty, with almost twenty percent over the age of forty. In contrast the vast majority of students are below the age of twenty-two in most non-profit universities and colleges of high academic reputation.

Some predictions of the reputation hypothesis, however, appear to be falsified. Specifically, the hypothesis predicts that for-profits will proliferate in sectors where income and status are closely linked. Unlike in philosophy and many of the liberal arts, most of the high- status businessmen are the richest businessmen. When money and status coincide, pursuit of for- profit goals should maximize institutional status as well. Yet the top business schools and MBA programs are at non-profit universities, not at for-profits.

It is true that MBA programs are run more like for-profit institutions than are most other non-profit educational programs. In this sense the prediction is partially confirmed. Non-profit business schools demand good teaching, they collect tuition rather than offering scholarships, and historically they have placed less stress on faculty research (although this has changed in the last fifteen years). The same cannot be said for nonprofit graduate training in philosophy. Nonetheless the ability of non-profit business schools to take on certain for-profit characteristics again raises deeper difficulties with any explanation of corporate status. A complete account of for-profit education cannot take observed differences between non- and for-profits for granted, but rather must derive the observed inflexibilities of these forms from more primitive assumptions. We still do not have a full account of why non- and for-profit institutions cannot easily “morph into each other” when circumstances dictate, or for particular segments of their market.

IV. Concluding remarks

We have examined the cross-sectional distribution of for-profit and non-profit institutions of higher education. A complex mix of factors, including subsidies, donations, agency problems, and the production of reputation, explain some of the observed patterns in the data. Unlike earlier approaches, which focus solely on the non-profit side of the ledger, we have stressed the need to explain the observed differences across corporate forms and governance structures. At the same time, however, we have found that no single explanation covers all of the ground or can respond to all objections. In that regard significant puzzles remain outstanding.

In terms of practical implications, the analysis suggests limits to the future of for-profit higher education. Non-profit institutions appear to have basic advantages in their ability to reap subsidies, raise donations, produce high quality research, generate academic status, and certify students and faculty in non-vocational areas. For-profit institutions, although they have grown in recent years, do not show serious signs of breaching these barriers.

The properties of the current equilibrium do not always yield accurate predictions about what would result from more dramatic changes in institutions. Nonetheless it is our belief that many forms of higher education are non-profit or not-for-profit for good reason. The reduction of government subsidies, or the privatization of many educational entities, probably would not bring the universal reign of the for-profit form. We expect the non-profit form to remain robust in many sectors of the higher education market.