What is the economic value of research? From the Chronicle a few weeks ago:
Just last week, at a budget hearing on Capitol Hill, a member of Congress cited a well-traveled number: $2.21.
That figure, said Rep. Rosa DeLauro of Connecticut, the top Democrat on the House Appropriations subcommittee in charge of education issues, is how much the National Institutes of Health generates in economic growth for every taxpayer dollar it receives.
"That is an over-100-percent return on the investment," Ms. DeLauro assured her legislative colleagues. Others, including the NIH’s director, Francis S. Collins, the hearing’s lead witness, have been citing the $2.21 figure for years.
Yet researchers studying the question remain in profound disagreement. They’ve been working on the assumption that the true economic value of the nation’s scientific investment, through the NIH and other agencies, is almost surely a lot higher than $2.21 per dollar. Putting real precision on that number, however, has proved a highly elusive goal.
But this week, after five years of trying, a team of analysts announced progress toward solving the puzzle. The group, led by Julia I. Lane, a former National Science Foundation official, published an article on Thursday in Science magazine offering a series of seemingly random findings on the economics of federal spending on science. They include the fact that faculty researchers account for fewer than one in five workers supported by federal science spending, and that universities given federal research money spend about 70 percent of it outside their home states.
The most important value of such information, Ms. Lane said in an interview, still remains several years away. That’s because this week’s data only capture the first step in the life of a federal research dollar, she said. Much more now needs to be done to keep tracing those dollars in a scientifically rigorous manner throughout the economy, she said, to get a firm idea of what benefit the money ultimately brings. ...
The calculation will be far different from the estimates that produced the $2.21 figure, which merely use a standard economic multiplier to measure the immediate stimulative effect of NIH spending on a local economy—in activities such as workers buying lunch at a nearby restaurant—that would apply to any kind of government spending, whether for medical research or road construction.
A major barrier to getting a better number for science has been that the full effects on society of research discoveries often can take decades to be realized. For too long, Ms. Lane said, the scientific community has taken that complexity as an excuse not to try.
The current attempt to establish economic value stems from the federal stimulus measure of 2009, which led to a project known as Star Metrics. Through it, more than 100 universities agreed to electronically collate data related to their federal grant spending, so that researchers can now automatically collect details such as new jobs, journal publications, and patents associated with each grant.
"It’s going beyond this mechanical magic-multiplier stuff," said Ms. Lane, currently a senior economist at the American Institutes for Research.
The logic behind the $2.21 number is sound, but misapplied. There should be no multiplier attached to federal spending because local and regional economic impacts are mostly beggar-thy-neighbor policies. A government spending policy that improves the well-being of a region tends to lower the well-being of competing regions. With federal policies the additional spending is completely offset by the lost spending elsewhere.
The real benefit of research is very difficult to measure with market data since these are mostly spillover benefits. The funding agency and the researcher are not able to capture the results of their efforts. This infuriates politicians who don't like scientists who do research, which leads to nonsensical economic impact studies.