The New Normal in Funding University Science
Government funding for academic research will remain limited, and competition for grants will remain high. Broad adjustments will be needed—and here’s a plan.
Science policy analysts have focused recently on the federal budget sequester and the dramatic effects it could have on funding scientific R&D in U.S. universities, certainly a serious problem. But looking only at the sequester misses the larger picture. The sequester simply makes acute a chronic condition that has been getting worse for years. Even if Congress removed the sequester tomorrow and R&D funding returned to pre-sequester levels, university researchers would still face serious and growing problems in funding their research programs, systemic problems that arise from the R&D funding system and incentive structure that the federal government put in place after World War II. This reality dictates that policymakers, research administrators, and the scientific community must adjust to continuing low success rates if scientific research is to continue to flourish on university and college campuses.
Researchers across the country encounter increasingly fierce competition for money. Funding rates in many National Institutes of Health (NIH) and National Science Foundation (NSF) programs are now at historical lows, declining from more than 30% before 2001 to 20% or even less in 2011 (with an uptick in 2009 associated with stimulus funding). The funding rates in some programs are substantially worse, dipping into the single digits. At these success rates, even the most prominent scientists will find it difficult to maintain funding for their laboratories, and young scientists seeking their first grant may become so overwhelmed that individuals of great promise will be driven from the field. As the Chronicle of Higher Education reported in 2013, the anxiety and frustration among principal investigators were manifested in the form of a letter to NSF, signed by more than 550 ecologists and environmental scientists, criticizing the negative impact that new policies, designed to cope with the flood of proposals, would have on the progress of science, junior faculty members, and collaborative research.
Many scientists and outside observers blame these low funding rates on a decline in the federal commitment to funding scientific research. However, the evidence tells another story. The growth of the scientific enterprise on university campuses during the past 60 years is not sustainable and has now reached a tipping point at which old models no longer work and expectations on the part of universities and university-based scientists have to be brought into line with fiscal realities. At the same time, federal funding agencies must work with universities to ensure that new models of funding do not stymie the progress of science in the United States, but instead continue to fund the most deserving research, while recognizing the need to keep a broad portfolio of investigators active in order to hedge the nation’s R&D bet (it is not always easy to recognize the research programs that will bear the most fruit) and to make certain that students at a wide range of institutions can be trained by research-active scientists.
Origins of the crisis
So how did the nation get into this situation? In one sense, the answer is obvious: The demand for research money greatly exceeds the supply. And clearly the demand for research funding has gone up. More universities seek sponsored funding, and individual researchers submit more grant applications. At NIH, for example, the number of research grant applications doubled between fiscal years 1997 and 2011, from roughly 31,000 to 62,000. But that answer—demand exceeding supply—simply restates the question. Why has demand grown faster than supply? Many answers get thrown around, including the fecklessness of politicians who refuse to provide enough money for science, and a lack of understanding on the part of the general public about science and what it does for the country. But those responses clearly do not account for the increased demands that scientists have placed on the federal funding system. The deeper sources of the problem lie in the incentive structure of the modern research university, the aspirations of scientists trained by those universities, and the aspirations of less research-intensive universities and colleges across the nation. These incentives and aspirations date back decades and set up a dynamic that was bound to run into a funding crisis; it was only a matter of when. Perhaps the most surprising feature of the current crisis is that it took so long to arrive.
Since the founding of NSF in 1950, the research enterprise on university campuses in the United States has grown rapidly, especially as measured by the numbers of science and engineering doctorates awarded. The competitive grants system encouraged such growth. Principal investigators need a dedicated, inexpensive, and talented workforce of apprentices (graduate students) for research projects. Therefore, if a university wants to attract a significant amount of sponsored research money, it needs doctoral programs in the relevant fields and faculty members who are dedicated to both winning grants and training students. The production of science and engineering doctorates has grown apace. As NSF has detailed in a 2006 historical study of doctorates in the United States in the 20th century, for the five years of 1920-1924 the nation produced a total of 2,724 such degrees, an average of 545 per year. By period 1955-1959, straddling Sputnik, doctorate production had gone up by more than a factor of 10, averaging 5,662 per year. By 1995-1999, science and engineering doctorate production had gone up by another factor of 5, averaging 26,854 per year. There has been little increase since then, with the nation producing 27,134 such degrees in 2010. The growth in Ph.D. production in the latter half of the 20th century was fueled by the growth in size of science and engineering departments at major research universities, as well as by an increase in the number of universities offering such degrees.
Even though not all doctorate recipients become university faculty, the size of the science and engineering faculty at U.S. universities has grown substantially. This cadre numbered 271,550 in 2006, up from 221,682 as recently as 1993, according to NSF statistics. These scientists and engineers are spread among a much larger number of Ph.D.-granting departments than was the case 60 years ago, and these departments have adopted the norms that their faculty should be active in acquiring sponsored research money and in producing new knowledge and publications. Hence, proposal pressure goes up. These strategies make sense for any individual university, but will fail collectively unless federal funding for R&D grows robustly enough to keep up with demand.
Derek J. De Solla Price, in his prescient 1963 book Little Science, Big Science, saw that the rapid increase in resources going into scientific research could not continue indefinitely. Price put the dramatic events of his time into historical context and showed that they were part of a trend that had been developing for more than a century. At the very time that universities were enjoying rapidly growing budgets, and creating modes of operation that assumed such largess was the new normal, Price warned that it would all soon come to a halt. He pointed out that in the United States and Western Europe, the human and financial resources invested in science had been increasing much faster than the populations and economies of those regions, and that such growth could continue only as long as the absolute number of scientists remained a very small proportion of the total population and R&D budgets a small part of the total economy. Once the number of scientists reached a few percent of the population, the growth would have to slow down; otherwise, he said, “we should have two scientists for every man, woman, child, and dog in the population, and we should spend on them twice as much money as we had.” Since that result is not possible, growth in the scientific enterprise would have to slow down at some point, growing no more than the population or the economy. Science policy built on the assumption of indefinite rapid growth was bound to come to grief.
Many other analysts have perceived a crisis in the federal funding of university science. In 2007, the National Academies published the ominously titled Rising Above the Gathering Storm: Energizing and Employing America for a Brighter Economic Future. In 2013, the Academies followed up with Research Universities and the Future of America: Ten Breakthrough Actions Vital To Our Nation’s Prosperity and Security. Both of these studies sounded an alarm about the potential decline in U.S. global leadership in science and technology and the grave implications of that decline for economic growth and national security. Both also contained numerous analyses and recommendations; at their core, however, was a proposal for the federal government to spend significantly more money on research, especially on basic research. As expressed in Gathering Storm: “Increase the federal investment in long-term basic research by 10% each year over the next 7 years through reallocation of existing funds or, if necessary, through the investment of new funds.” This would mean roughly doubling that budget in seven years. In a footnote, the report recommended that the “reallocation” of money for this doubling could come from any federal agency, not just the research agencies. In addition to this doubling, the study called for an additional $100 million for special early-career investigators and yet another $500 million per year for facilities and instruments. The follow-up report on research universities echoed the call for a doubling of basic research funding.
Although we are not opposed to increasing federal funding for research, we are not optimistic that it will happen at anywhere near the rate the Academies seek, nor do we think it will have a large impact on funding rates. Universities and colleges have demonstrated a remarkable capacity for absorbing increases in federal funding by increasing research infrastructure, as was seen when Congress doubled the NIH budget over the course of five years beginning in the late 1990s. More good projects were funded as a result of the infusion of funds, but success rates for grants remained flat during the doubling and dropped immediately afterward. A more serious problem with the recommendation to double the basic research budget over the course of seven years is that it is devoid of context. It does not acknowledge the current pressure on the federal budget or the historical relationship that R&D funding has had to that budget.
The pressure on the federal budget is obvious. Where R&D fits in that budget is more complex. R&D funds do not come out of the budget as a whole, but rather out of the discretionary budgets, whether defense or nondefense, which are little more than one-third of the total budget. As Daniel Sarewitz pointed out in 2007 in this journal, when examining R&D spending as a percentage of the discretionary budget, a measure the American Association for the Advancement of Science has been tracking for decades, what jumps out is the remarkable consistency of nondefense R&D spending. After a steep rise and fall in the 1960s, mostly due to the Apollo program, nondefense R&D settled down by the middle of the 1970s to make up roughly 10% of the domestic discretionary budget, and there it has stayed for almost 40 years. This has held true during both Republican and Democratic administrations and Congresses and when the White House and Congress have been held by the same or different parties. The defense portion of the R&D budget has been more volatile over the same period, fluctuating from 10 to 15% of the discretionary budget.
What this means for the future of federally funded R&D is that universities should not expect any radical increases in domestic R&D budgets, and most likely not in defense R&D budgets either, unless the discretionary budgets themselves grow rapidly. Those budgets are under pressure from political groups that want to shrink government spending and from the growth of spending in mandatory programs. There is no reason to assume that R&D will be able to make claims on a larger portion of the discretionary budget, because other existing programs can be expected to defend their budgets. Of course, if the total discretionary budget increases due to economic growth, R&D will share in that increase. It is also numerically possible, although it seems politically unlikely, that defense R&D could pick up more of the task of funding basic research, the sort most often undertaken at universities. The basic point is that the growth of the economy will drive increases in federal R&D spending, and any attempt to provide rapid or sustained increases beyond that growth will require taking money from other programs. The solution to the problem of funding university science and engineering research will not come from ever more massive infusions of new money from the federal government. The demand for research money cannot grow faster than the economy forever and the growth curve for research money flattened out long ago.
Path out of crisis
To chart a realistic path out of this crisis, it is necessary to start by reframing the goal. The goal cannot be to convince the government to invest a higher proportion of its discretionary spending in research. For 40 years, R&D funding has competed with a host of other national needs—from road and bridge building to social welfare to public health to education—and for 40 years has come away with 10% of the discretionary civilian budget. Getting more is not in the cards, and some observers think the scientific community will be lucky to keep what it has. Instead, the goal must be to sustain the most vigorous scientific research programs possible on university campuses, given the reality that levels of federal funding are going to grow slowly in concert with the growth of the U.S. economy, at best.
Why should universities and colleges continue to support scientific research, knowing that the financial benefits are diminishing? First, attracting federal research dollars has never been just a financial benefit. A lively research culture makes it possible for universities to attract good students and faculty as well as raise their prestige within the academic community. Second, universities take it as their mission to expand the boundaries of human knowledge, including scientific knowledge. Third, faculty members are committed to their scholarship and will press on with their research programs even when external dollars are scarce. Fourth, the training of scientists, even at the undergraduate level, does not take place in teaching laboratories, but rather in research laboratories. If the United States is to train the next generation of scientists and reap the economic benefits associated with scientific research, it is critical to have active research laboratories, not only in elite public and private research institutions, but in non-flagship public universities, a diverse set of private universities, and four-year colleges. Talented students from a variety of backgrounds are found in all of these institutions, and the nation cannot afford to lose some of that talent by limiting cutting-edge scientific training to a small number of universities.
Universities and colleges have long accepted the reality that the federal government would not underwrite all of the scientific research efforts under way on their campuses. Indeed, institutional funds for R&D are the second-largest source of funding for academic R&D, rising from 12% in 1972 to about 20% in 1991 and remaining at that level through 2009. Increasing, or even sustaining, that level of investment will be difficult, particularly for public institutions as state support for higher education continues its inexorable decline. How then do increasingly beleaguered institutions of higher education support the research efforts of the faculty, given the reality that federal grants are going to be few and far between for the majority of faculty members? What are the practical steps institutions can take?
First, they must change the current model of providing large startup packages when a faculty member is hired and then leaving it up to the faculty member to obtain funding for the remainder of his or her career. The premise for this model is a reasonably high funding rate for grant proposals. The thinking goes as follows: Provide a faculty member with a laboratory and the means to generate preliminary data and anyone worth his or her salt will win a grant and then go on to sustained funding over a career. The new reality of low funding rates calls for a new model, one in which universities and colleges recognize that even the most distinguished researchers will experience gaps in funding and require in-house assistance to maintain a research program. The need for assistance throughout a career can be met only if universities invest less in new faculty members and spread their internal research dollars across faculty members at all stages of their careers, from early to late.
Moving toward smaller startup packages will not be easy. Faculty members in the sciences and engineering see the size of a startup package as a signal of an institution’s commitment to research, and a university that unilaterally reduces its startup packages will find it difficult to attract highly accomplished job candidates. Negative repercussions may be ameliorated by a national conversation about changes in startup packages and by careful consultations with prospective faculty hires about long-term support of their research efforts. Many prospective hires may find smaller startup packages palatable, if they can be convinced that the smaller packages are coupled with an institutional commitment to ongoing research support and more reasonable expectations about winning grants.
Smaller startup packages mean that in many situations, new faculty members will not be able to establish a functioning stand-alone laboratory. Thus, space and equipment will need to be shared to a greater extent than has been true in the past. This will place increased emphasis on the construction of open laboratory spaces and the strategic development of well-equipped research centers capable of efficiently servicing the needs of an array of researchers. The phaseout of the individual laboratory, which is already under way at many universities and medical centers, brings with it enhanced opportunities for communication and networking among faculty members and their students. Collaborative proposals and the assembly of research teams that focus on more complex problems can arise relatively naturally as interactions among researchers are facilitated by proximity and the absence of walls between laboratories. Universities can compete for top faculty members based in part on the collaborative opportunities rather than simply the size of the startup package.
An increased emphasis on team research will place greater demands on research administrators (deans, vice presidents for research, provosts) to be fully engaged with their research faculty members, so that investments in the research enterprise (such as new hires and new equipment) can be directed at projects that have good buy-in from the faculty, have the greatest chance of receiving external funding, and have the potential to lead to important results. Faculty members who are at the beginning of their research careers will require careful mentoring to ensure that they learn how to work both as part of a team and independently. Involvement in multiple projects should be encouraged, and it will be incumbent on senior faculty members to provide junior faculty members with thorough mentoring and with prudent leadership opportunities.
Even with careful mentoring and thoughtful grooming of junior faculty for leadership opportunities, it is the rare assistant professor who will be in a position to lead a research team. The more likely trajectory of a junior faculty member will evolve from contributing team member to increasing leadership responsibilities to team leader. Because contributions to a team and the development of leadership qualities are unlikely to be apparent to outside evaluators, internal evaluations of contributions and potential will become more important in tenure and promotion decisions. Successful leadership of a research team is likely to become an important criterion for promotion to full professor at many research universities.
Low success rates for grant proposals at the major federal funding agencies will not deter faculty members with active research programs from requesting support from those agencies, but relationships with foundations, donors, state agencies, and private business will become increasingly important in the funding game. The opportunities to form partnerships with business are especially intriguing. Many businesses have cut back their R&D infrastructure, so the potential to take advantage of the infrastructure and talent on university campuses may be a win-win situation for businesses and institutions of higher education. Businesses can tap into the expertise of highly skilled scientists and their students on an as-needed basis, while scientists gain insight into the questions important to businesses and the means by which to translate research results into marketable products.
Our suggestion is hardly new, and some universities have already developed extensive relationships with businesses. In order for such collaborations to expand, leaders in both sectors need to rethink how and why they can take advantage of joint work and the diversity of forms that such collaboration can take. If universities wish to replace a significant portion of their federal funding with money from the private sector, businesses will need to greatly expand their university work. Industry has funded only a modest portion of university R&D, falling to about 5% in 2011 from 7 to 8% in the early 1950s, according to NSF’s National Patterns of R&D Resources. How much industry might be able or willing to increase that funding is difficult to predict, but even a large percentage increase in such funding will represent only a small increase in university R&D funds, considering the system as a whole. Obviously, for a few specific universities, industry funding could provide a considerably larger fraction.
Further complicating university collaborations with business is that past examples of such partnerships have not always been easy or free of controversy. The biotechnology revolution created many deep relationships between firms and universities. This led to considerable controversy, as some faculty members worried about firms dictating the research priorities of the university, pulling graduate students into proprietary research (which could limit what they could publish), and generally tugging the relevant faculty in multiple directions. To whom did faculty members owe their loyalty, and what were appropriate constraints on faculty members whose work attracted the biotech firms? The large research universities had to grapple with these problems and developed rules and guidelines to control them. The results of that work should be widely disseminated to help head off some of these problems.
University faculty and businesspeople often do not understand each other’s cultures, needs, and constraints, and such gaps can lead to more mundane problems in university/industry relations, not least of which are organizational demands and institutional cultures. University researchers need to be sensitive to the intense deadlines firms may face, and firms need to realize that they cannot approach a university during the last two weeks of spring term and ask that six faculty members drop everything and spend the next month working full-time on a new project. That said, both sides have good reasons to work out those differences. In addition to funding for research, universities can receive indirect benefits from such relationships. High-profile partnerships with businesses will underline the important role that universities can play in the economic development of a region. Perhaps the biggest barrier to genuine collaboration will be adjusting the expectations of both sides. Universities have to see firms as more than just deep pockets, and firms need to see universities as more than sources of cheap skilled labor.
These extramural relationships can also extend beyond the private sector. University researchers and their students can take on problems relevant to local and state governments or nonprofit organizations that operate in their area, possibly with assistance from foundations or other philanthropy. An article in the June 21, 2013 issue of Science pointed out how important private philanthropy is to university research programs, although some people quoted in the article doubt that it can replace large sums of federal or state money. According to NSF’s National Patterns data, nonprofit entities provided almost 10% of university and college R&D funds in the early 1950s but they provide only 7 to 8% today. Universities can encourage new outreach efforts on the part of their faculties by honoring all successful efforts to generate research dollars, not only those that result in NSF and NIH grants. Greater local involvement would provide other benefits to universities and colleges. Institutions of higher education that become scientific and engineering resources for the localities and regions in which they are located can make a stronger argument for financial support from state and local governments.
We do not believe that research proposed and supervised by individual principal investigators will disappear anytime soon. It is a research model that has proven to be remarkably successful and enduring, and one that appeals to talented junior and senior faculty members. In addition, initiatives such as NSF’s Experimental Program to Stimulate Competitive Research have succeeded in making serious research opportunities available to a geographically diverse set of institutions, encouraging the development of scientific talent throughout the country. However, we believe that the most vibrant scientific communities on university and college campuses, and the ones most likely to thrive in the new reality of funding for the sciences, will be those that encourage the formation of research teams and are nimble with regard to funding sources, even as they leave room for traditional avenues of funding and research.
The new reality of low funding rates for most proposals that university scientists submit to NSF and NIH results not from a lack of support for science on the part of legislators or the public, but from a buildup of the scientific infrastructure in universities and colleges that has now outstripped the funding capacity of the federal government. This budgetary landscape is unlikely to change in the foreseeable future, and universities need new models of supporting scientific research on campus if they are to remain centers of training, research, and economic development. The new normal need not be bleak, but it requires science policymakers, university leaders, and the scientific community itself to rethink models of the scientific enterprise on university campuses.