|While the agency's halt to `downward negotiations' may promote fairness, it fails to brighten overall research support vistas|
WASHINGTON--The National Institutes of Health has officially ended its highly unpopular practice of applying unilateral, across-the-board cuts to the grants of those researchers it funds. But most scientists will find that the varying approaches adopted by individual institutes to replace what were euphemistically called "downward negotiations" do not yield any more money for their research.
"We have begun to do away with across-the-board cuts," John Diggs, NIH deputy director for extramural research, told the House appropriations committee's labor and health and human services subcommittee last month at hearings on the agency's 1992 proposed budget. "Instead, we're examining specific categories of cuts, to determine whether they are reasonable and allowable."
The dramatic shakeups that businesses undertake to solve problems aren't an option for the federal government, in which political considerations can sometimes stifle decisive action. Instead, what the government does when something's not working right is assemble a panel of outside experts to examine the troubled unit and gently prod it back into shape.
In 1989, the National Research Council of the National Academy of Sciences issued a report that took the National Science Foundation to task for the way it gathers information about scientific personnel. The report, entitled "Surveying the Nation's Scientists and Engineers," criticized NSF for the quality of the data it collects and for not putting enough resources into the job. It was especially critical of NSF's continued use of a three-pronged test to determine who is a scientist or engineer, calling it unwieldy and pointing out that the results were at odds with numbers collected by other federal agencies.
To that end, in the past year NSF has formed two oversight committees to check up on how the foundation manages the data it is required to collect on U.S. science. The first is an advisory committee to the Directorate for Science, Technological, and International Affairs (STIA), comparable to the panels that exist to advise the foundation's five research directorates. The second is a subgroup of that advisory committee, overseeing the heads of the divisions of policy research and analysis and science resources studies.
"It's becoming clear that you need to keep data collection completely separate from policy analysis," says Judith Liebman, chairwoman of the second advisory committee and vice chancellor for research and dean of the graduate college at the University of Illinois. "It's not clear how separate they were kept in the past." Karl Willenbrock, assistant NSF director for STIA, believes that quality and clarity are the key areas in which improvements are needed. "We need to do a better job than we are doing now" of collecting and analyzing data on scientific personnel and trends, he says. Referring to a controversial working paper by NSF policy analyst Peter House, he adds, "We also need to publish these analyses in such a way that they are not so easy to misunderstand."
One of the problems in achieving that goal, according to Liebman, is money. "There is some concern within STIA that they aren't budgeted to collect the necessary data," she says. "It was felt that an outside committee advising them on how to prioritize their resources would be very useful."
Another issue is how the foundation's efforts should be carried out. Any hint that agency officials have sought to assemble certain types of data, or set about to analyze the data in ways that would prove a predetermined view of the labor market for scientists, could doom the process of data collection, everyone agrees. "If they start to mix the two," says Edward Tufte, a professor of political science and statistics at Yale University and a member of the Liebman panel, about the need to separate the roles of data collection and policy analysis, "then people worry that data collection is being affected by the agency's agenda of policy issues."
Willenbrock is optimistic that NSF can make the necessary changes. He notes that many of the improvements in data collection and analysis require only a new approach to the task at hand, not a bigger budget. At the same time, he acknowledges that a recent NSF decision to eliminate seven unfilled STIA slots as part of a transfer of personnel into the Education and Human Resources Directorate "certainly wasn't a positive move" in strengthening the directorate.
That new approach is a response to a congressional mandate to eliminate arbitrary reductions, amounting to 20 percent in many cases, in the initial recommended budgets of grants to individual investigators. The initial figure is determined by peer review panels known as study sections, and then reviewed by an institute-wide advisory council, as well as by program staff, before institute officials decide on the actual level of funding. This year, for example, the average annual size of a research project grant is $246,000, compared with an average initial request of nearly $300,000.
Replacing these arbitrary reductions, NIH officials have promised Congress, will be a system that gives researchers enough money to do good science and, at the same time, allows NIH to make more accurate estimates of funding levels in the "out years"--all the years beyond the first--of a multiyear grant. In addition, NIH has pledged to reduce the average length of all new and competing grants to four years, from the present 4.3 years.
The new rules vary among the 13 institutes that make up NIH according to each institute's research portfolio and its own priorities. Officials at the National Heart, Lung, and Blood Institute (NHLBI), for example, have decided to exempt brand new awards from any cuts in the recommended level of funding, while applying a formula to continuing grants and those up for renewal that will result in reductions of up to 20 percent.
At the National Institute of Allergy and Infectious Diseases, program managers will be applying a sliding scale to successful grants, trimming 10 percent from the top-rated proposals and 16 percent from those closest to the cutoff. The institute also has set aside a small pot of money to restore some of those cuts. And at the National Institute for Child Health and Human Development, NICHD officials have decided to stretch a 25 percent funding shortfall by supporting less expensive grants and cutting the budgets of higher-ranking research proposals by approximately 15 percent to 20 percent. All institutes in which the average grant length now exceeds four years have begun to lop off the final year of an application, usually reducing to four years what had been a request for five years of support.
The bottom line for most investigators, say NIH officials, is still a smaller budget than what they had requested and what the initial review panel had recommended. That's because of continuing pressure from the scientific community to fund as many researchers as possible. And when there's too little money to go around, NIH believes that everybody should share the hardship.
Some scientists are skeptical that the changes are real. "What's happened is that they've eliminated the word `arbitrary' from their vocabulary," says Thomas Edgington, an immunologist at the Scripps Clinic and Research Foundation in La Jolla, Calif., and the president of the Federation of American Societies for Experimental Biology. "A different euphemism will be used, but it won't make any difference to the PI [principal investigator]. They [NIH] still don't have enough money."
Others are slightly more generous in their assessment. "The objective is the same--to fund more grants with the same amount of money," says Dave Moore, an assistant director of the American Association of Medical Colleges. "The good news for researchers is that they'll have a better idea of how much they're going to get. The bad news is that it's not as much as they'd hoped for."
The impetus for the changes comes from language in a 1991 House appropriations committee report (The Scientist, Sept. 3, 1990, page 1) recommending that NIH "aggressively eliminate any costs which it believes are unnecessary ... [after which] a grant should be funded at the full level, without arbitrary reductions." That report reflected growing concern about recent fluctuations in the number of new and competing grants that NIH awards each year, from a peak of 6,400 in 1987 to a low of 4,600 in 1990 and a current level of 5,785.
Those fluctuations generated complaints from the research community that the process of obtaining an NIH grant had become a lottery. Those protests were heard by federal legislators, who have given NIH budget increases far beyond what the current and previous administrations have requested. With the persistent federal deficit having dimmed prospects for future large increases, however, even NIH's staunchest supporters began to wonder whether the agency was getting the biggest bang for the taxpayers' bucks.
Out of that concern has come a cost-management plan (The Scientist, Jan. 21, 1991, page 1). It was presented last December in draft form, with a final version due out shortly.
Even so, it's not the last word on the subject. NIH director Bernadine Healy testified last month that the plan was only "Phase 1" of a larger effort to examine all financial matters. The second part, she told chairman Rep. William Natcher (D-Ky.), will address such long-term issues as modeling the effect of varying the size, length, and number of grants on budgets; anticipating future health crises that require a swift and massive response; and tackling the controversial topic of indirect costs (The Scientist, April 15, 1991, page 1).
"Those will take a little more time," she testified. "We have no firm timetable for Phase 2. But I can tell you that it will get into dicier issues that will be harder to resolve than those addressed in the current plan."
NHLBI recently put in place a complex system for giving money to individual investigators.
To begin with, NIH divides awards to individual investigators into three major categories. Type 1 grants are new awards to scientists who don't have existing NIH grants. Type 2 grants are competing renewals to researchers whose grants are expiring and who want to continue their work. Type 5 grants are noncompeting renewals to scientists in the middle of multiyear grants.
Starting this past winter, the heart institute decided to award Type 1 grants at the budget level recommended by the study section, the outside group of peers who weigh the merits of every investigator-initiated proposal. It is the only institute to adopt such a role, according to NIH officials.
Researchers in the other two categories won't be treated as well. For Type 2 grants, the institute will award either 110 percent of the amount awarded in the grant's last noncompeting year, or 80 percent of the level recommended by the study section, whichever is larger. For Type 5 grants, the institute distinguishes between those grants that competed against other grants prior to 1990 and those that competed last year. The former will receive a maximum of 4 percent more than their current level of funding. The latter will see their 1991 recommended level cut by no more than 11 percent. In the future, these grants will receive an annual increase of 4 percent.
Institute officials emphasize that these rules are merely guidelines, and that the budget of each grant will still be examined carefully by program staff to see if it should be treated differently. Overall, however, NHLBI's director of extramural affairs, Ron Geller, says that the new system is a significant improvement for grantees by offering them more stable funding than in the past.
"For one thing, at whatever level we issue a Type 5 grant, future years will be based on a projected 4 percent increase," he says. "That's a big change from the past, when each year the grant was subject to a downward negotiation of 15 or 20 percent. That means scientists will be in a better position to plan the scope of their research, based on a more realistic assessment of the dollars that should be available."
The heart institute is also cutting one year from the length of most new and competing grant proposals. "Only a few proposals, at the highest percentile levels, aren't being cut," says Geller. The goal is to bring the institute's average grant length to four years, in keeping with NIH's promise to Congress. Not all institutes will feel the lash of that directive; at the National Eye Institute, for example, the average length of the investigator grants in its research portfolio is already four years.
The eye institute, in fact, is a case in which business as usual seems sufficient to meet the requirements of the new cost management plan. "We've never engaged in the arbitrary, across-the-board procedures that other institutes have used," says Jack McLaughlin, associate director for extramural programs. "It's not that we've always had enough money to pay for everything we want to fund, but we've always encouraged our grant managers and our advisory council members to participate actively in setting the budgets of individual grants."
In the aggregate, according to McLaughlin, the average eye institute grant winds up being cut by 10 or 11 percent from recommended levels. But he says that the cuts are handled on a case-by-case basis, with no need for overarching rules to be applied to each successful proposal.
Donald Clark, chief of the office of grants and contracts at NICHD, notes that the new system at his agency places a greater emphasis on the cost of any particular grant. NICHD has adopted a policy that gives its advisory council leeway to fund any proposals in a wide range above and below the suggested pay-line (that percentile ranking above which all grants could be funded if the institute went strictly in rank order).
"If a couple of grants are cheaper, then picking them will mean that we'd have to cut other grants less," he says. "That's certainly going to be an important consideration, along with program relevance and program balance."
Clark acknowledges that these new approaches won't solve the root problem, which is too many good ideas chasing too few dollars. In fact, he says the changes may scarcely be visible to the typical grantee. "Will the average PI notice any difference in his or her award? Not really. The big change," he says, "is how we get there."