September 2012

The best-laid plans: It’s time to move forward

In 1997, the U.S. Senate passed a resolution 98–0 stating that the budget of the National Institutes of Health should be doubled over five years. Through the leadership of Sen. Tom Harkin, D-Iowa; then-Sen. Arlen Specter, then-R-Pa.; and then-Rep. John Porter, R-Ill., this doubling became a reality. From fiscal 1999 through fiscal 2003, the NIH appropriation was increased to $27.2 billion from $13.7 billion. The goal of this dramatic growth in federal support for biomedical research was to increase the scale of the enterprise to facilitate further discoveries, to capitalize on the great scientific opportunities made possible by previous discoveries and to translate these advances, where possible, directly to affect human health.

In 2003, as the time for the doubling to end was drawing near, policy leaders from the academic community published a Science Policy Forum article entitled “The NIH Budget in the ‘Postdoubling’ Era” (1). The authors analyzed some of the challenges of transitioning from annual budget increases of about 15 percent to more modest increases, including maintaining progress into new research areas, incorporating the impact of ongoing commitments of previously funded grants, and sustaining support for new investigators and trainees. They developed models to estimate the effects of different rates of increase in the NIH budget from 4 percent to 10 percent per year. Their analyses revealed that annual increases below 6 percent were likely to lead to tough compromises for the NIH and the research community. They concluded that “policy-makers may feel that the federal government has done its part for NIH-funded research and that the agency can be allowed to coast … at static levels of funding. To the contrary, we emphasize that levels of growth below 6 to 8% will negate many of the advantages achieved by the doubling … They will also severely strain the relationship of trust between NIH and its awardees on which our nation’s successes in biomedical research rest.”

Unfortunately, this effort at thoughtful analysis did not yield the desired implementation. Since fiscal 2004, NIH budget increments have ranged from −1 percent to 3.2 percent, with an average of 1.4 percent. This is substantially below the rate of inflation for biomedical research, resulting in a loss of more than 20 percent in purchasing power over this eight-year period. The NIH did, indeed, have to make many of the tough choices that were foreshadowed in the article to prevent success rates for new and competing grants from falling dramatically.

A substantial amount of the stress on the system arose from events set in motion by the budget doubling. In anticipation of and in response to the doubling, many institutions grew by hiring additional faculty members and constructing new research spaces. However, both faculty hiring and, particularly, building construction take time, so many new faculty members did not come on line until near the end of or after the doubling period. This led to a significant increase in the number of investigators and associated grant applications competing for the available resources from the NIH and other funding sources. Furthermore, institutions had become accustomed to the relatively high success rates (approximately 25 percent for R01 applications) that occurred during the doubling period, so many institutions expected investigators to raise a larger fraction of their salaries from NIH grants or other sources than they had previously. While creating more faculty positions was beneficial to the enterprise and to the researchers who filled these positions, these and other positions became, on average, more dependent on external funding with a lower level of institutional commitment.

In addition, as detailed in the recent Biomedical Research Workforce Working Group report from the Advisory Committee to the NIH Director (2), among other sources, the number of biomedical Ph.D.s awarded remained almost constant at 5,400 per year during the doubling period. However, this number grew linearly to 7,700 by 2009. Given an average of five to six years for Ph.D. training, the post-doubling increase in Ph.D.s reflects growth in the size of Ph.D. programs during the doubling period; that is, most of the students who graduated through 2009 started their graduate training during the doubling period, before the NIH budget had flattened out. This growth was not based on any attempt to increase the number of trainees, but rather it was due to the fact that Ph.D. production is quite strongly coupled to research activity: when more research funds were available, institutions responded by increasing the sizes of their graduate programs.

Thus, many of the factors that are placing the biomedical research enterprise under stress were set in motion by the NIH budget doubling followed by stagnant NIH appropriations after the doubling. Of course, further stress has been imposed by the worldwide economic downturn, which has had substantial effects on federal and state budgets, institutional endowments, and the availability of capital. The support for the NIH and the National Science Foundation in the American Recovery and Reinvestment Act postponed the full brunt of these factors, but we now have reached a critical time.

What can we forecast from the facts at hand? First, it is unlikely that substantial increases in federal support for science will be forthcoming until a more robust economic recovery is under way (although we still must make the case that such investments are important in driving a recovery and that every dollar counts in minimizing the damage that is done in the meantime). Second, the current state of affairs is unlikely to be sustainable. Few laboratories can be confident of sustained support given present and projected success rates. Review groups tend to become quite conservative when success rates are so low, to the detriment of scientific progress. Third, many young scientists have been (and are still being) trained and are ready to move on to different stages of their careers.

Given these realities, what can be done to build a more sustainable future? The Biomedical Research Workforce Working Group report touches on a number of important issues relevant to this question. Some recommendations relate to graduate students and postdoctoral fellows, including those directed toward shortening the training period. The lengthening of this period has been primarily a symptom of some of the factors discussed above, namely the strong coupling of training to research activity and the growth in the number of trainees competing for available positions. Taking steps to shorten the training period is important, but these steps are not likely to be very effective without tackling some of the bigger issues raised in the report. Another key aspect of graduate and postdoctoral training is driven by the appreciation that individuals with Ph.D.s in biomedical sciences go on to pursue a wide range of academic and nonacademic careers. A key question is how training programs need to adjust to accommodate this reality; effective answers to this are likely to come only by energetic engagement with those from nonacademic sectors.

Two of the remaining topics covered by the report reflect fundamental changes in the biomedical work force and the relationship between academic institutions and the NIH. The first of these involves staff scientists. The report notes the important roles that staff scientists can play in supporting the research enterprise. The development of a more robust staff scientist career path has the potential to provide career opportunities for individuals and to provide an avenue for conducting research that is substantially less coupled to training than is the current system. In my opinion, the working group’s recommendations that “NIH study sections … be receptive to grant applications that include staff scientists” and that “institutions … create position categories that reflect the value and stature of these researchers” are not bold enough. The academic community should work with the NIH to explore creative ways to make such positions sufficiently stable so that they become attractive options for individuals interested in long-term careers in research. The second involves the balance of faculty salary support provided by institutions and the NIH. As the report notes, more data are needed, and the diversity of practices at different institutions makes this a challenging policy area, but this issue has so many implications that it must not be ignored and bold options must be explored.

These issues must be addressed by the academic community, government and others working together — and now is the time. Doing so will require careful analysis of desired and potential unintended consequences and creativity to find reasonable policies that balance often-conflicting needs; the future of the biomedical research enterprise in the United States depends on it.

  1. 1. Science 296, 1401 – 1402 (2003).
  2. 2. 


Photo of Jeremy BergJeremy Berg ( is the associate senior vice-chancellor for science strategy and planning in the health sciences and a faculty member in the computational and systems biology department at the University of Pittsburgh.


found= true1955