University of California Gets Bonus from Feds for Selecting Foreign Grad Students

By David North on September 22, 2010

Download a pdf of this Memorandum

David North is a CIS Fellow

The University of California (UC) receives an extra $15,000-a-year payment from the federal government every time it admits a high-tech graduate student from overseas, as opposed to one from the United States.1 The payment comes to UC after it puts the student on the payroll of a federal grant.

We estimate that UC gets a $50 million yearly bonus from this unusual system.

Whether these financial bonuses, in fact, encourage the 10 UC campuses to discriminate against would-be U.S. grad students is a different and very interesting question, but one we are not going to tackle in this report. Here we describe the internal workings of this bizarre program.

UC, like most other state universities, receives substantial amounts of federal research money, notably from the National Science Foundation and various arms of the Departments of Energy and Health and Human Services. Foreign grad students in science and engineering routinely are employed as researchers on these grants.

UC, unlike most, if not all, state universities, charges the feds more for tuition remission for foreign rather than U.S. graduate students working on funded research projects. It does so by charging out-of-state tuition rates for foreign grad students and lower in-state rates for domestic ones. Typically, other state universities charge the federal government the in-state rate for both domestic and foreign grad students.

The difference between in-state and out-of-state tuition at UC is about $15,000 per student, per year. It does not cost any more or less to educate a foreign student compared to a domestic one. The two-tier tuition system was designed by state legislators many years ago so that local tax funds would not, or would not seem to, subsidize out-of-state students.

What the University of California has done is to use, in combination, the traditional two-tier tuition system and a permissive federal funding process and to twist the combination to its own advantage. While I find the system abusive, for reasons noted below, I must admit that the UC billing system is a stroke of genius.

Grad students working on these federal research projects typically both receive a cash stipend and have their tuitions paid by the grant. A foreign grad student at UC gets no more income from working on a federal research grant than an American one does — the roughly $15,000-a-year differential goes to the University of California’s central treasury. (The graduate school departments making the admissions decisions, similarly, do not profit directly from the out-of-state tuition charges to federal agencies.)

While this odd and expensive system appears to be legal, at least at the moment, it raises several public policy questions:

  • Should the federal government fund a system that rewards a university for admitting one type of graduate student (from overseas) as opposed to another (U.S. residents)?
  • Similarly, isn’t it loading the dice against resident applicants to the grad schools?
  • Should a single state system be given a substantial financial break over most other state universities, simply because California manipulates a federal funding formula in a way that other states do not?
  • Should limited taxpayer resources, and research grant funds, send $50 million a year to UC’s treasury, when it might be spent on additional research at UC, on research activities elsewhere, be returned to the Treasury, or not be spent at all?

An immediate reaction from an outsider to a situation in which one university charges considerably more for a similar unit of work — a year of a graduate student’s time — than another university would be: Is not UC pricing itself out of the market? Several people raised that question with us.

The answer is that research funding decisions are not the same as those made when buying a commodity, like a bushel of wheat; many factors other than cost enter the process.

The basic decision to make a grant relates to the apparent utility of the subject to be studied, the reputation of the principal investigator, and, probably to a lesser extent, the reputation of the institution making the proposal. All else being equal, there probably is a tilt, perhaps unconscious, toward UC/Berkeley or Harvard over an institution in, for instance, the Dakotas. (I made a living for decades doing social science research for government agencies, and have a good sense of how these things work.)

Usually it is after the basic grant decision has been made that detailed budgetary questions arise. Typically, the government leans over backward to allow universities to work out their own salary scales and related matters; tuition remission for grad student researchers is a widely accepted part of the cost structure. Apparently, allowing academic institutions to define their own tuition costs is also part of the picture. Further, there is a strong overlap between the people at the top of the research universities and the people who run the NIH and NSF; the ranking officials in those agencies think like, and often have been, top university officials before they came to the government, or want to become such after they leave it. There is no evidence that the government funders of research are upset about the UC practice.

In this setting, where billions of dollars are spent annually, the $50 million extra to one highly regarded major university system can easily be ignored, and has been. That would-be American graduate students might be harmed has also received the silent treatment.

Grad Student Tuition Practices

Another reason for the lack of public attention is the arcane nature of the problem. As we researched this question we found that a large number of university officials were confused by the subject. We often were given lengthy responses about the way that in-state and out-of-state tuitions were calculated, which was not our primary concern.

Here are some examples of the built-in complications: It is usually impossible for foreign graduate students on temporary visas to move from non-resident to resident status, while, after a year, it is easy for U.S. out-of-state students to become in-state residents. Further, in some cases, including UC, out-of-state tuition charges are levied on graduate students, but after a few years when they have matured to doctoral candidate status (needing only to finish the thesis) they move to in-state status. Finally, in some cases (e.g., Yale), while there may be tuition charged to graduate students, none is charged to doctoral candidates.

That some tuition remissions might be charged to the federal government, that these costs might be charged at very different rates by different universities, that the rates within a given university’s system might vary by the citizenship of the student, and that the federal government apparently did not care about the level of these charges were both central to our research, and, as a group, understandably hard to grasp.

Further, tuition of graduate students at many research universities is little more than a theoretical concept to the students, not a bread-and-butter issue. Tuition is almost automatically waived at some of the better-funded institutions. Once when I was talking about graduate school tuition with an assistant dean at my alma mater, Princeton, I asked if the university ever charged it. She replied: “Well, if we noticed that some prince from Saudi Arabia was driving a Rolls-Royce, we might send him a tuition bill.”

In my volunteer work helping graduate students at the University of Maryland with their income taxes during the springtime, I have sensed that only a small minority of PhD students in the sciences at that institution pay tuition. These tend to be marginal students in their first semester, or foreign students with the full support of their homeland governments, or a few in mild disfavor with the university. In the last instance, I sense the university sends a tuition bill rather than taking more drastic action.

Grad student tuition is more often charged at less-well-funded institutions, and in the arts and the humanities, rather than in science and engineering. It is tough to get federal research grants for art or literature.

In terms of tuition dollar levels, most, but not all, UC entities charge about $29,000 for out-of-state residents, as is noted on Table 1. That level is generally above that of the other state universities. Further, the UC differential between the two rates is generally above the differentials seen at the other institutions.

UC, at the out-of-state level, does not charge quite as much as the University of Michigan; turning to another comparison, to Harvard, Yale, and Princeton, all of these institutions (and other non-state universities) charge a single rate, and do not worry about where people live. The prominent trio all charge tuitions of about $35,000 a year.

This appears to be a “them that has, gets” policy, tilted toward the rich universities. While it might be interesting to puzzle over a federal policy that repays tuition at more than four times as much in the Ivies as it does at the University of North Carolina, for example, and while it would be intriguing to explore a policy that appears to reflect a bureaucratic desire to avoid hard decisions,2 that would be a distraction from the thrust of this report.

The main point of Table 1 is that UC has managed to game the system so thoroughly that it gets two- to three-and-a-half times as much money for the same product — a year’s tuition for a grad student — as other state universities. And it does so by treating foreign graduate students as more valuable than domestic ones.

Minimal, Ineffectual Federal Attention

This strange pattern has been touched upon twice in formal governmental actions, but the key problem — paying the university more for admitting foreign graduate students than for admitting domestic ones — was never addressed. The UC practice, thus, survived intact.

In one instance, the higher payments for the work of foreign grad students was raised as a secondary policy issue in a not-very-satisfactory GAO report, which was subsequently ignored; in the second, a court case was decided on a legal variable that had nothing to do with the appropriateness of the UC policy.

The GAO study, “Compensation Paid to Graduate Students at the University of California,” showed its conceptual problems in the title.3 The GAO’s apparent suggestion is that the foreign graduate students were paid too much.

No, the central problem was that the university was paid too much for working with that population. The GAO study, thorough though it was, is an example of the careful analysis of the bark of the tree while failing to notice the forest fire.The study dealt with three issues: 1) Were some of the graduate students being compensated more than first-year post-docs doing the same kind of work? — something that was prohibited. 2) Were the grad school billing policies vis-a-vis tuition remission consistent with both NIH and IRS financial rules? and 3) Had the “foreign GSRs (graduate student researchers) received a larger proportionate share of the total tuition and fee remission costs charged against federal grants than did resident GSRs”?4

The answer to the first question was “yes,” some grad students were (once the tuition remissions were brought into the equation) paid more than the first year post-docs. No tuition was paid by or for the post-docs, so the basic comparison was flawed from the start.

The answer to the second question, which occupies a great deal of the report, was that NIH regarded tuition remission as a salary-type payment so it was OK to bill the government, while IRS regarded tuition remission as a fringe benefit, so it did not need to be taxed. Neither agency seemed upset that the other agency disagreed with it. (That the NIH decision allowed greater federal spending, while the IRS decision substantially reduced federal tax collections — in other words both were hard on the Treasury and taxpayers – was just another elephant in the room ignored by GAO.)

As to the third question, the report said “yes,” the compensation to the foreign students was higher than to domestic students, but it failed to notice three things: that this made no difference to the actual incomes of the students, that it made a lot of difference to UC’s income, and that it might have created an uneven setting for making admission decisions regarding foreign vs. domestic applicants.

The report had two useful notations about the UC practice:

However, in reviewing an internal university document prepared in 1996, we noted that the university had found that no other public universities charged the cost of out-of-state tuition to federal grants. We also noted that the state of California does not allow the university to charge state grant awards for nonresident tuition remission.5

So it was OK for UC to charge the nonresident rate to federal grants, but not to state-funded projects; but the federal funding sources did not seem to object.

Nothing apparently happened as a result of the GAO report, perhaps, at least partially, because it was narrow in its focus, and at least partially because the federal agencies paying the extra expense did not mind.

The court case was equally frustrating.6 It was decided at the federal district court level, and involved a whistle-blower (qui tam or relator in legalese) suing against the system.The case never touched the merits because the university’s lawyers argued successfully that an individual under the pertinent federal law had no standing to sue a state entity like UC. The decision was made on October 28, 1999.

Current Billing Practices

So what are the current practices of UC and the other state universities regarding charging federal agencies for nonresident tuition remissions (NTRs)?

We have it on good authority from three people deep within the UC system that the practice continues as before — people we will not name because UC might be miffed at someone on their payroll endangering a $50-million-a-year windfall. The three are from three different units of the far-flung UC system.

There is also a supporting document from the UC Davis Office of Graduate Studies titled “Nonresident Tuition Remission Policy.”7 The pertinent section says: “The cost of nonresident tuition remission [NTR] is based on the rate of nonresident tuition applicable during the term of a qualifying appointment as a GSR [graduate student researcher] and is charged as a direct cost to the fund source that supports an eligible GSR’s salary or wages.”

As to the practices in other state universities, finding out what such universities charge to federal agencies regarding NTR is a bit like learning the sex of an unborn baby in the womb of a woman not known to you. It is potentially knowable, but only after a series of some awkward and not-always successful conversations.

We started a blizzard of queries, some by phone, more by email, with dozens of state universities on this subject. In five other cases we asked friends, such as full professors in appropriate places, to intervene on our behalf, and in three of these we got the information we wanted, as we did at seven other universities.8

The 10 universities for which we have solid information on their practices are listed in Table 1; in all of these it was clear — after some prodding in many cases — that each and every one of them used the same practice: they charged a single tuition rate (the in-state rate) when they were billing federal research grants for NTR for foreign grad student researchers.

Similarly, one of our UC sources, someone who had paid attention to this issue, said that to that person’s knowledge no non-UC institution followed the UC practice.

There was one possible exception. A dean at the University of New Mexico told us that sometimes, but rarely, NTR would be charged to a federal grant. We could not get any details or confirmation of this practice, but he said “it happens with some of the big funders, but never the smaller ones.”

So it is possible that there are others following the UC practice, but we cannot be sure. If they are, the same policy questions could be asked of them as we ask of UC.


This strange tilt, this odd bonus system for picking foreign rather than domestic grad students in science and engineering, seems to have persisted for at least 15 years, seems to favor the University of California alone, and surely is bad public policy. It also appears to cost the U.S. Treasury $50 million a year.

End Notes

1 The author is grateful to Marcie Gard, a graduate student at Texas State University at San Marcos, for reminding him of this issue. Though I had written an essay about it for the Center for Immigration Studies a decade ago, I had forgotten about the matter and had no idea that the practice was continuing. She was also helpful in this research, and for this I am thankful, but the text, the point of view, and any errors are mine alone.

2 Simply allowing tuition levels, set by the universities, to be used as an automatic part of all research project cost structures means wildly varying costs to the government, and excessive expenses in some cases, but it also means that there is no need for government officials to worry about what an appropriate tuition reimbursement amount should be.

3 “Federal Research Grants: Compensation Paid to Graduate Students at the University of California,” General Accounting Office, Washington, DC, 1999 (GAO/OSI-99-8),

4 Ibid. p. 10.

5 Ibid.

6 U.S. ex rel. Robert Baum v. the Regents of the University of California (No. CIV. S-96-1054 WBS JFM).

7 Available at

8 The author is grateful to two leading immigration policy persons, Professors Philip Martin at UC/Davis, and Vernon Briggs of Cornell, for the information secured for us. Also very helpful (vis-a-vis Virginia Tech) was Virginia State Sen. Edd Houck, chair of the Virginia State Senate’s Education and Health Committee.