The showdown over the “fiscal cliff” of tax increases and budget cuts has arrived, and, with it, university-based scientists and their allies are pleading their case for strong and sustained government support.
In some ways, science advocates are in a good position. It’s a rare bipartisan issue, with leading Republicans and Democrats routinely proclaiming their broad support for research spending, particularly in medical fields.
And yet, after years of trying, the science community still cannot answer a simple question, one that has gained the growing attention of policy makers as the economy limps along: How much payback, in real dollars, does science spending actually provide?
“It would be useful” to know, Alan I. Leshner, chief executive of the American Association for the Advancement of Science, said last week after hosting a Capitol Hill briefing for lawmakers and their staff members on the dangers of cutting the science budget.
For as much as lawmakers proclaim their love for science, the budget numbers don’t always reflect that. Congress has cut federal support for biomedical research, as compared with inflation and with gross domestic product, almost every year since 2003. And between 2005 and 2010, Congress promised repeatedly to double spending on nonmedical research, but it hasn’t come close to matching that.
And now, if Congress lets a package of automatic cuts known as sequestration—because the money is being “sequestered,” or taken away from federal agencies—take effect on January 2, federal science budgets will instantly lose about 8 percent, or about $12.5-billion in research-and-development money in 2013 alone.
“We are extremely nervous about what could happen,” Mr. Leshner told the gathering of lawmakers and their staff members. The looming package of $1.2-trillion in automatic budget cuts over 10 years was created by Congress in August 2011 when the lawmakers couldn’t agree on a strategy for deficit reduction.
Part of the problem for universities is that in the current economic climate, even friends of the science budget—many of whom recognize the critical long-term value of research—are increasingly demanding hard numerical proof of precisely what the nation’s $140-billion annual investment in science is buying taxpayers and their struggling economy.
“I’m going to continue to support it,” Senator Barbara A. Mikulski, Democrat of Maryland, told Francis S. Collins, director of the National Institutes of Health, at a budget hearing this year, “but the promise of science needs to have deliverables.”
Dr. Collins has tried. Since taking charge of the NIH in 2009, he has repeatedly cited one compelling statistical measure: For each dollar of the $30-billion it receives annually from taxpayers, the NIH rewards Americans with $2.21 in economic benefit.
The trouble, according to various expert and economists, is that $2.21 is a bad figure, somewhere between misleading and meaningless.
Undervaluing Knowledge
The $2.21 figure comes from a 2008 report by Families USA, an advocacy organization promoting affordable health coverage for Americans. The amount is the nationwide average of the additional business activity each state experienced in 2007 from each dollar of NIH grants and contracts.
The problem is that the figure is what economists use as a “standard multiplier,” not specific to science or research. While there are some variations between industries and localities, the multiplier largely represents the additional spending that virtually any new flow of money from outside a region would generate, as a result of a recipient using the money to buy goods and services.
“It is essentially the same kind of calculation that you would make,” said Joshua L. Rosenbloom, a professor of economics at the University of Kansas and a research associate at the National Bureau of Economic Research, “if the government were to hire a contractor to build a highway, or to build a building, or to operate a prison.”
In that sense, Dr. Collins is potentially far undervaluing the work of his own agency. That’s because the multiplier formula used by Families USA, even though widely accepted by economists, misses what most advocates consider the greatest value of federal spending on science—the additional knowledge that is generated.
That value could be huge. Some of the better-known examples include the hundreds of millions of dollars the government spent developing the laser, mostly through the military, before having it find major and widespread applications in a variety of fields including computer-disk technology, and the $4.5-million National Science Foundation grant to Stanford University that played a role in the development of the ubiquitous Google search engine and its $200-billion corporation.
But such enormous values, the long time frame often required for the payoffs, and the multiple and varied routes through which science travels from initial discovery to commercial application are major obstacles to making reliable assessments of the real multiplier value of federal expenditures on science.
“That’s the challenge everybody’s grappling with,” said Tobin L. Smith, vice president for policy at the Association of American Universities.
It has been tried in the past, without great clarity. One of the earliest studies of the value of research was Project Hindsight in the 1960s, when the Pentagon examined the inventive origins of various weapons systems. It found that only 0.3 percent of the needed innovations came from basic scientific exploration, and virtually all important breakthroughs came instead from researchers trying to fulfill a specific military need.
Before the Pentagon could even finish its analysis, said Joseph P. Lane director of the Center for Assistive Technology at the University at Buffalo, the NSF fired a pre-emptive strike, worried that the Pentagon’s findings could cast doubt on the central value of basic university research. The NSF began a study called Traces, in which it tracked the history of various consumer products and documented major contributions from basic research decades earlier.
There have been almost no subsequent large-scale studies aimed at resolving such questions, said Mr. Rosenbloom, who began a two-year stint this summer as a program director with the NSF’s Science of Science and Innovation Policy program. “And I think it reflects the degree of difficulty and the cost involved in assessing those payoffs in a scholarly, rigorous, and defensible manner,” he said.
Firmer Numbers
Nevertheless, a new generation is trying to put firm numbers on the economic value of research. The NSF now spends about $8-million a year on the Science of Science and Innovation Policy program, a unit that began in 2007 to finance grants exploring the overall value of federally sponsored research.
And the NIH began its Star Metrics effort in 2010. For its first phase, Star Metrics encouraged universities to supply data that would let federal officials track the number of new jobs attributable to each grant. In the more difficult, second phase, project leaders are trying to develop more-comprehensive measures of the contributions of federal science spending to the nation’s overall economic productivity, and to human health.
Both the new NSF division and Star Metrics stem from a 2005 call by John H. Marburger III, then the chief science adviser to President George W. Bush, for a “scientifically rigorous, quantitative basis” for supporting research.
Seven years later, those trying to meet the challenge acknowledge success won’t be simple, cheap, or quick. One of the best efforts was a report last year by the Battelle Memorial Institute, a nonprofit science-and-technology-development company, that found the $3.8-billion that the U.S. government invested in the Human Genome Project from 1988 to 2003 could be credited with generating $796-billion in economic growth and $244-billion in personal income.
That study, however, was laborious and costly, and left its authors with little confidence that such an effort could be replicated on a wider basis. It involved tracking long lines of detailed expenditures by numerous companies that benefited from genomic research, and it cost more than $100,000 to produce, said Simon J. Tripp, a Battelle analyst who helped write it.
It’s not clear whether the economic multipliers attributed to the Human Genome Project are typical for all federal spending on science, or whether such an ambitious survey could be applied more broadly with reliability, said a developer of the Families USA study, Richard P. Clinch, director of economic research at the University of Baltimore’s Jacob France Institute.
Rewards for Revenues?
Then there’s the question of what better information about economic payoffs might really mean for university scientists. The obvious benefit could be bigger federal research budgets. But a future in which scientific fields can be ranked according to their apparent contributions to economic success could lead Congress to try its hand at predicting—and rewarding—the fields that might have the greatest odds of a financial windfall.
Mr. Lane, of the University at Buffalo, welcomes the notion. He has argued that countries like China are far more interested in finding the commercial applications of basic science discoveries than in pursuing basic science itself. In that regard, “considering just the potential for sequestration is a really important exercise,” he said, as it might help the United States better understand how to make the most efficient use of its research dollars.
Many others disagree. Morteza Gharib, vice provost for research at the California Institute of Technology, said the unpredictable nature of the value of research can be seen in examples like magnetic resonance imaging, the now-common medical diagnostics technology that for decades was considered useful only for chemical and physical analyses.
The NSF now routinely asks grant applicants to assess the economic value of their work, “giving the idea to the investigator that you have to push your research to something more applied,” Mr. Gharib said.
Mr. Smith said he resists the idea that all scientific value must be convertible into dollars. “I don’t think we have to necessarily fall into the trap of oversimplifying or dumbing down things because some policy makers want to focus only on the immediate economic value of research,” he said. “We have to explain the complexity involved.”
Still, greater attention to the economic value of science seems inevitable. During the Capitol Hill briefing hosted by Mr. Leshner, Steven J. Fluharty, senior vice provost for research at the University of Pennsylvania, described in detail the dire effects threatened by sequestration. Yet he also said “this impending crisis provides us all an opportunity to step back” and take a big-picture look at how better to assess the value and models of federally financed research.
Mr. Fluharty said the science community has to “try to develop common frames of reference for what matters, and it’s got to be evidence-based, it’s got to be based on the return of investment.”
The cost of actually doing that, however, might prove in the end to be the greatest obstacle to moving from anecdote to arithmetic in policy debates. At a hearing in March of a House Appropriations subcommittee, the panel’s chairman, Rep. Denny Rehberg, Republican of Montana, made clear that he didn’t see why the NIH should be diverting its money—intended by Congress for medical research—to evaluate the economic effectiveness of its work.
“This isn’t research,” Mr. Rehberg told Dr. Collins. “This is an economics study.” He promised the NIH director that “it’s something I’m going to be looking at as chairman.”
That, at least, may be one less worry for Dr. Collins. Mr. Rehberg is leaving Congress in January, after losing his bid this month for the seat of Sen. Jon Tester.
