Tuesday, August 5, 2014

How Can Public Research Universities Pay for Research? (UPDATED 04/15)

Higher ed policy is suffering through a long siege  of intellectual gridlock.  The default result is what I've been calling permausterity, a chronic funding shortage for public colleges that now rests on a chronic lack of confidence in the job they're doing.  This has become a vicious cycle that feeds itself.  

Making matters worse, faculty responses are fragmented, when faculty respond at all.  Some of the most eloquent voices are increasingly disenchanted: William Deresiewicz got so much pushback for his recent piece, "Don't Send your Kid to the Ivy League," in part because he seemed to be saying that even our premier universities are turning America's most successful students into mercenary sheep.


(1) Why Can't College be Cheaper?

Dr. Deresiewicz's piece upset many supporters of the college ideal (e.g., Jim Sleeper), and one reason is that it seemed to lend credibility to this year's leading higher ed question: "is college worth it?" If Yale sucks too, why not give up on rebuilding funding and learning and get on with the inevitable consolidation of higher ed into two dozen university-corporations along the lines of the media industries and IT? The Apollo Group could provide the management, Coursera the online platform, Pearson VUE the assessment, and Harvard-MIT-Stanford the quality control.  Three percent of the college population could still go to prestige-brand research universities and liberal arts colleges, which is about the percentage that goes to them now.  Everyone else would, in this scenario, get converted over 10-15 years to varying combinations of blended learning and online-only. In spite of the MOOC ebb that began last summer, tech-based disruption and downsizing remain at the top of the national higher ed agenda.   

There are good disruptions that should be implemented, bottom-up, in universities, and also obvious reasons not to turn universities into digital learning corporations.  One of these reasons has to do with how people actually learn (as opposed to how they receive and replicate information packets). Some of the growth in student services is a market-driven "amenities race," but much of the growth comes from new structural support for better learning. Fixing the country's educational levels is going to require more and not less money for student services, more and not less funding for active learning, and more and not less payroll to hire permanent faculty.  Adjunct Nation has new allies in Congress, which will also support a deeper discussion of educational quality. We need post-contingent education (see, for example, Jennifer Ruth's recent posts (here and here).


Another large cost is research.  The country expects the vast majority of its basic research to come from universities.  And yet few policymakers and general voters understand who pays for research and how much it costs.  The traditional funders have been the federal and state governments, but states have been reneging on their side of the deal for years, leaving the feds in the lurch.  At the same time, the feds have been partners in this decline, having never explained to state policymakers, much less to voters, that they did not fund the full cost of research.  Admitting that research loses money has been taboo, since it conflicts with Washington's demand that science lead directly to economic growth.  States have cut funding in part because they didn't know they were in effect also cutting economically strategic STEM research.


But in the last few years things have been looking up.  Washington D.C. agencies are finally going public with their concern that we don't know how to pay the full costs of university research after all.


(2) Research Shortfalls are Real

In 2012, the National Science Board published Diminishing Funding and Rising Expectations: Trends and Challenges for Research Universities, and in the same year the National Research Council of the National Academies released Research Universities and the Future of America. Both criticized the states' wholesale retreat from public funding.  Both reports noted that universities are increasingly on the hook to pay for research from their own internal funds--even when the research has an outside sponsor. Institutional funds are now the "second largest source of funding for academic R&D, accounting for $11.2 billion of the $54.9 billion of academic spending on S&E [Science &Engineering] R&D in 2009" (NSB p 16).  The NRC report stated that "The institutional contribution to research has been growing faster than federal funding," which, they added, diverts money from necessities like instruction and maintenance (NRC p 125).

Then, this past June, the Council on Governmental Relations, a leading research university lobby, chimed in with the same message and more graphic detail.  Under the title, "Finances of Research Universities," its report offers a good primer on the differences between private and public university funding and then gets into some of the gory details of research costs.  If one of your summer resolutions is to tone up your skill with calculating F&A overhead on MTDC, then this is the report for you.


The big takeaways are that universities' internal funds are the fastest-growing source of research funding, and that universities' share is large.  The total university contribution has grown again since the NSB and NRC reports, from $11.2 billion to $13.7 billion per year.

Over the period from 1976 to 2012, the share of R&D expenditures assumed by colleges and universities has grown faster than any other category. Institutional Funds accounted for 21.6% of all R&D expenditures in 2012 (adjusting out the ARRA effect) as compared to 12.0% of all R&D expenditures in 1976—a growth factor of +80%.
COGR provides a number of interesting tables, using in many cases data from the NSF's Higher Education Research and Development Survery, or HERD). Here is one:

Reseach and Development (R&D) Expenditures by Funding Source as a Percentage of All R&D Expenditures


State support for R&D is a third of what it was pre-Sputnik (1956) (although unadjusted totals continued to grow).  Federal support, though much more stable, is now heading back down towards its pre-Sputnik share. Over the same period, universities have doubled the size of their piece of research funding. Their share has doubled since the 1970s, in spite of excellent growth rates of federal research funding--or actually, because of this federal growth.  In 2011, a useful article in Nature pointed out a further problem, which can be seen in one of its figures: 

Public universities do twice the dollar amount of research that privates do, and yet spend twice the share of their own funds in subsidizing it (24% vs 12%).   Hence the title question, how can public research universities afford to do the research society does in fact want?

Back to the COGR report, which concludes with some bureaucratic fighting words:
The university subsidy is a legitimate issue and one that use be addressed honestly and constructively by all stakeholders.  [Forcing] universities to fund real, unreimbursed costs through non-federal revenue sources [makes them] potentially reduce investments in core missions and infrastructure. Ultimately, this impairs a university's ability to strategically plan and invest in its future research enterprise. (23)
In other words, concealing true research costs hurts the overall university while also hurting research.

I'm happy that a high-level organization is now explicitly saying that unrecovered research costs "are a financial burden with severe implications for the future productivity of research universities" (19).  This is progress.


(3) How Much of the Research Shortfalls are Recoverable?

There's a big wrinkle we now need to consider.  What kind of research costs are universities covering through their Institutional Funds? 

Universities need to support extramural research with outlays for facilities and administration (F&A), whose reimbursements have been capped at 26% since 1991, though only for universities. They also need to build and renew overall infrastructure and pay for research that isn't supported by outside sponsors (which includes nearly all research in the arts, humanities and qualitative social sciences).  They must help start new labs, sometimes build new buildings for them, seed new projects that may attract outside funding at some future date, and provide bridge funding for faculty who are in-between grants but have labs to run and grad students to train.  A combination of these and other activities accounts for the $13.7 billion that universities spent of their own money on research in fiscal year 2012 (out of a total of nearly $66 billion).  (The NSF breaks down costs by university in this table.)

The NSF tries to figure out how much money goes to various research categories through the HERD survey mentioned above.  The COGR report cites its findings as follows:
Of the $13.7 billion, 56% ($7.7 billion) was in the form of direct funding for faculty or student research projects, 9% ($1.3 billion) was devoted to cost sharing, and almost 34% ($4.6 billion) represented unrecovered indirect costs. (2012 HERD Survey)
In other words, somewhat over half of university research expenditures supports the research of their own faculty and students. A third goes to cover costs incurred by sponsored research that are not covered by the sponsors. Another tenth goes to cost sharing, which always involves sponsored projects. Summing up these figures, we might conclude that 44% of Institutional Funds subsidize extramural sponsors, while 56% cover internal research projects.  All of these costs are within the normal scope of research university activity--and, to get pious for a second, form part of its obligation to society.

But is this breakdown correct? The COGR report suggests it is by singling out the $4.6 billion as the main subsidy burden universities bear. It equates, in the report's terms, "to a staggering multi-million dollar obligation per university," and raises a "widespread concern as to the sustainability of the significant investments made by research universities" (19). COGR thus implies that only about one-third of universities' research outlays could be recovered by fixing reimbursement policy.


Other documents tell different tales.  The COGR report itself offers a case study (Chart 13, p 20) of a "Private Research University, Southeast."  This university spent $505 million on research but received $390 million in revenues, which required it to chip in $115 million of its own money.  So nearly 23% of this university's total research costs came from Institutional Funds.  The line-item breakdown of expenses lists University-Funded Research at $33 million, or  a bit over 28% of the Institutional Fund contribution.  This is half of the average for "direct funding for faculty or student research projects" in the HERD survey.  (It is also only 6.5% of this university's total R&D expenditure.)


To take a further case: when the University of California's Commission on the Future tried to get a handle on the university's costs, they summarized research losses like this:

In recent years, the University has received over $3.5 billion per year in extramurally-sponsored research grants, of which over $780 million per year is designated for indirect costs such as facilities support and research administration.  But the actual indirect costs of extramurally-funded research are estimated to be $1.5 billion. (page 111)
UC was thus losing $720 million a year on a research gross of $3.5 billion. This meant that 20.6% of its R&D expenditures came from internal funds, which is very close to the national average.  But this statement suggests that sponsored research caused the entire shortfall.

So we have three stories about the extent to which research universities must spend more money than the public understands in order to cover costs on behalf of research sponsors. 

  1.  A third (or at most 44%) of Institutional Funds go to subsidizing costs of sponsored research, costs that the private sector would likely insist be paid in full.  About 56% goes to non-sponsored or "internal" research for faculty and students.
  2. Something like a quarter of Institutional Funds go to non-sponsored research.  That leaves three-quarters supporting extramurally sponsored research. 
  3. More or less all Institutional Funds go to filling in these shortfalls in sponsored research funding.
Which story is correct? I think the best answer at the moment is all of them, depending on the university. Wealthy private universities may well be close to (1), spending most of their internal funds on their own faculty's non-sponsored projects.   Less wealthy privates and some major public research universities may be close to (2). Both of these stories are about major research losses of somewhat different sizes.

The extreme case of (3), in which nearly all Institutional Funds subsidize sponsored research, may be right for the case for which it was developed, the University of California.

To check whether this could possibly be true, I offer some seat-of-the-pants numbers for one campus, UCLA.  It has formally recorded Institutional Funds expenditures from at least two sources, its Academic Senate Committee on Research, and the Office of the President's Research Grants Program Office (RGPO). The former, in the pre-cut year of 2007-08, dispensed about $2 million in travel and research support.  The latter, over a three-year period 2010-13, spent $44 million per year (  Annual Report page 25).  (I apologize for mixing years but here I'm just going for scale).  I'll assume that UCLA got about one-fifth of RGPO system resources based on its large size.  That means the campus spent $11 million of Institutional Funds through formal channels on faculty and student research projects in a period when it was grossing around $1 billion a year in extramural research funding.  In other words, UCLA spent 1.1% of its Institutional Funds on designated faculty research. 

(UPDATE 04/15: Having looked again at the RGPO awards for the relevant period, I think UCLA's share may be half of my estimate here. It's hard to say because they do not publish dollar amounts.  In addition, newer COR reports are online. Award totals are the same in 2013-14 as in 2007-08, my baseline here. The 1996-97 COR award $1.9 million, which is about $2.84 million in 2013-14 dollars: UCLA's COR now awards about half the amount it did 20 years ago.  "Half" seems to be the theme today: RGPO has awarded about half the number of multi-campus grants in this cycle compared to the last.)

This is obviously not the whole picture of internal research funding, but we don't have public information on the use of discretionary funds retained at various administrative levels--but also no reason to think a large percentage of this unknown figure goes to non-sponsored faculty research.  Throw in the fact that Committee on Research funds go to some extent to top up extramural grants. You can then see why the UC Commission report rounded up to the claim that essentially 100% of Institutional Funds go to paying for unreimbursed indirect costs of extramural research.

The implication of all of these stories, especially 2 and 3, is that public universities can pay for research, but, as we go forward, only if federal, state, and private funders stop asking them to subsidize a large chunk of indirect research costs.


(4) A Few Steps Towards Improvement

Regardless of which story is correct for a given university, they all point towards the following list of to-do's.

A. University administrations should say openly and often that research loses money. It must be publicly supported because it loses money.  The more fundamental the research, the greater its long-term social potential, the more likely it is to lose money for years if not decades. The Internet provides an easy example of this point. 

B. Point out that effectively freezing public funding to hundreds of research universities is undermining the country's research ecosystem.  Converting higher ed to online, in whole or in part, will wreck that ecosystem.

C. Act on these NSB, NRC, and COGR calls "to cover the full costs of research projects and other activities they procure from research universities in a consistent and transparent manner" (NRC Recommendation 6, p 122).  (It is already official University of California policy to charge sponsors enough to "cover all expenses, direct and indirect" (APM-020 Revised Regulation No. 4, II. 3)  Set up a multi-year plan for fixing at least the one-third of the problem that all agree is attributable to sponsors' underpaying of indirect research costs.

D. Sort through Stories 1-3 above. Get clean numbers, campus by campus, for "indirect indirect" costs--all the set-up costs that support extramural research rather than research that is ineligible for extramural funding. This will mean distinguishing clearly that research which, for historical and institutional reasons, depends wholly on Institutional Funds. It will also mean campus admins publishing those numbers to their communities, so that they can be understood and discussed.

E. Identify and quantify the needs of the large, complicated sphere of this (mostly) qualitative and/or truly experimental research that cannot receive external sponsorship. Explain why its ineligibility to receive external sponsorship follows from the historical shape of Western scientific, military, and industrial development rather than from a lack of merit or social value.  (This needs to be done for a society that doesn't generally understand market failure, spillover effects, or noncommercial social value.) Then make sure that this research has equal or superior claim to Institutional Funds.

Universities need finally to get ahead of the curve on research costs. If they don't, the "unbundling" pressures will only increase.

1 comment:

  1. Just a clarification to a great article--the "A" portion of "F & A"(administration) has been capped at 26%. The F portion (facilities) is not capped in the same way. Thus, the incentive to expand by construction, which the F portion can help to pay for. But expansion may create additional costs in administration that are not covered by F&A, and thus can produce more long-term shortfalls, just not ones booked directly against research.

    Given that public universities have no constraints on non-federal indirect costs, why don't they charge the full cost of research to, say, industry sponsors of research? Instead, they use the federal rate, and expect the state to subsidize that research, too. From the state's perspective, the state is forced to subsidize university research without ever getting a say in what sort of research, or how much. That research could be in anything--not necessarily meaningful even to the industry in the state--and somehow university administrators and faculty alike expect the taxpayer to write a check to make up what the sponsor is not willing to pay, or that the university is unwilling to charge.

    ReplyDelete

Note: Firefox is occasionally incompatible with our comments section. We apologize for the inconvenience.