Universities and the money fix

Funding woes plague US biomedical researchers. But calls for more funding ignore the structural problems that push universities to produce too many scientists, argues Brian C. Martinson.

Federal funding for biomedical research is a substantial investment in the US science community. Earlier this year, representatives of several major research universities testified before Congress and issued a report arguing that the budget of the National Institutes of Health (NIH) in Bethesda, Maryland, is insufficient to sustain "a strong and vibrant program of basic research"1. They pointed to stifling of innovation and damage to the career prospects of young scientists, ultimately warning that there could be a threat to US pre-eminence in biomedical research if Congress does not increase levels of funding for the NIH. Yet, what is it that poses the most potent threat to the future of biomedical research — a lack of resources, or our failure to manage the level of competition for available resources? The answer to this question is vital if society is to gain maximum benefit from the public money invested in biomedical research.

There is undeniably excessive competition for NIH grants, and we should all be concerned about the negative effects this may have on the robustness of the research engine; by damping scientists' willingness to pursue high-risk projects; by causing them to spend excessive time in pursuit of funding; or by causing talented individuals to shun research careers. Yet, largely because of the structure of the funding flows between the NIH and the universities, there are few checks in the system to keep competition for grant funding at a healthy level. Thus, calls for further increases in the NIH budget may only make matters worse. In my view, it is time to ask the biggest beneficiaries of NIH largesse — the universities and academic health centres — to find ways to balance supply and demand that better reflect their obligations to researchers and society.

University leaders know that when the money gets tight, it's junior faculty members who feel the pinch. They are less established in their careers, more peripheral to their professions and institutions, and often most dependent on obtaining NIH funding as an implicit or explicit condition of their continued employment. As NIH funding becomes harder for junior researchers to obtain, we might expect them to experience the elevated levels of depression, anxiety and job dissatisfaction documented in a survey2 of medical faculty members in 2006. We might also expect the greatest effects to be felt by female scientists and those from minority groups, for younger researchers to leave science, and to see somewhat less ethical behaviour among those who stay. The robustness of the research engine must be judged on more than the level of funding or the number of scientists.

The doubling of the NIH budget between 1998 and 2003 was intended to increase success rates in obtaining NIH grants3, which have been declining since the mid-1970s. Yet, the budget rise did not have its intended effect, and by 2003, grant-application success rates were slightly worse than before. What happened? The budgetary increases were swamped by an equally large escalation in the number of NIH applicants and applications (see graph, below)4. In 1998, there were about 19,000 scientists applying for competing awards; in 2006 there were approximately 34,000.

Since the 1970s academic researchers in biomedicine and the institutions that employ them have become increasingly dependent on NIH dollars5. The financial reasons for this are simple. The 'direct costs' of NIH grants cover the fixed costs of faculty salaries, whereas 'indirect cost recovery', pays for operational overheads, capital equipment and other expenses. Federal training grants also provide revenue streams for doctoral and postdoctoral training, directly stimulating workforce growth. Even before the doubling in funding, the Bayh–Dole act of 1980 created incentives for universities to grow their NIH workforce by permitting employers to own the inventions their employees created with federal funding.
Ageing cash cows

As dependence on NIH grants has grown, they have also become harder to obtain, especially for junior scientists. The average age at which PhD scientists earn their first independent support from the NIH has increased steadily6, from 34 in 1970 to 42 in 2006. The situation has certainly been made worse by the flat NIH budget (declining after taking inflation into account) since the end of the doubling initiative. Yet, the excessive demand for NIH funds predates the recently flat budget (see graph, below). Since the early 1980s new investigators have been entering NIH funding at a more rapid rate than experienced investigators have been exiting4, leading to a population increase.
With academic faculty members seen as revenue generators, they are encouraged in subtle and not-so-subtle ways to expend greater effort on lucrative activities: this has made research a preferred activity over teaching or patient care. It also means they must spend a substantial amount of time writing grants. This arrangement generally works in the universities' favour, but the downsides of the dependence on NIH funding are becoming harder to ignore.

For too long now, financial incentives to the universities have been aligned to promote unlimited growth in the number of biomedical researchers seeking funding from the federal government, despite the realities of finite resources. Some have suggested that a solution lies in biomedical researchers and universities becoming less dependent on NIH money by finding commercial funding sources and philanthropies7, but this approach is not without its own risks, and it avoids dealing with the structural arrangements that keep us from applying sound principles of supply and demand to the scientific workforce.

We need to look at both the supply and the demand sides of the NIH funding equation. Most who worry about these issues have focused on the size or distribution of the pool of NIH dollars. Far fewer have given consideration to the size or dynamics of the population of biomedical researchers living on NIH funding. Few have overtly asked the question — are there too many biomedical scientists?

There are insufficient 'feedback loops' linking the production of biomedical researchers to the availability of resources to support them. Instead, the educational system is replete with incentives to generate ever more PhDs and medical doctors. In the short term these arrangements may benefit universities, but in the longer term, such extreme levels of competition for funding are unsustainable. And they may already be doing harm. Difficult funding decisions are increasing ill will, perceptions of injustice, and eroding the bases of ethical behaviour among academics. Some of my own work leads me to believe that the current situation may be adversely affecting the integrity of research8.
The needle and the damage done

The imbalance between the supply of NIH funding and the potentially unlimited demand for grants threatens the future of US biomedical science. I have argued that because of structural incentives, demand for NIH grants is largely a function of the size of the biomedical workforce. Recent NIH initiatives to increase funding of junior researchers are welcome, and have the best chance of maintaining a pool of new research talent. But without some counterbalance, these initiatives may escalate competition for grants.

Calls for further increases to the NIH budget are a facile response from institutions overly dependent on NIH dollars. But they are an incomplete, and potentially dangerous, answer to the problems of excessive competition. And although short-term NIH budget increases to make up for inflation-related declines since 2003 seem reasonable, further increases risk fuelling, rather than reducing, demand. For now, budgetary increases that simply keep pace with inflation would seem prudent, so as not to reactivate the growth impulse. Regrettably, the current imbalance may be addressed only through a reduction in the biomedical workforce; something that already seems to be happening.
There are two main routes to contraction of the academic workforce today — through tenure failures, and with younger investigators shifting from academia into industry research8. This is worrisome for university research in particular because history suggests that the most dramatic innovations come from the young. So is the only solution to force long-time NIH grant getters into retirement? Perhaps not. Universities have benefited handsomely from the efforts of senior faculty members in securing NIH grants during their careers, perhaps those same universities could now return the favour by taking full responsibility for paying these faculty salaries in their later years. This would serve the dual purpose of getting them off the NIH dole, and encouraging them to share their knowledge with their younger colleagues through more teaching.

This won't be easy. Given the levels of dependency on NIH money, it is akin to asking an addict to give up an easy fix. And not all universities will be in financial positions to employ this strategy, but it's difficult to imagine that richer institutions — some of whom acknowledge that their success lies in capturing an increasing share of the NIH pie9 — could not lead the way in this. Prospective students and their parents may also look favourably on senior faculty members spending more time teaching.

What is needed is not necessarily more people, but more time, space and freedom.

An implicit assumption underpinning the current system of funding is that having more biomedical scientists automatically leads to greater innovation and more breakthroughs. Yet what is needed is not necessarily more people, but more time, space and freedom for existing researchers to ask questions in new ways, to be willing and able to take risks, and to innovate rather than simply writing safe, incremental grants. The excessive competition for NIH funds discourages this kind of risk-taking, and ultimately reduces opportunities for the sort of creative thinking that leads to major scientific breakthroughs.