Productivity growth in developed countries has been slowing down, threatening to cause a long-term stagnation in living standards. One contributing factor could be that it’s getting more expensive to find new scientific and technological ideas. Economists Nicholas Bloom, Charles I. Jones, John Van Reenen and Michael Webb recently tried to measure innovation in various fields, and found that more and more researchers were needed to produce the same pace of discovery.

That’s an ominous trend. Governments and companies can pour ever more resources into science and technology, but if this spending becomes less effective, the costs of innovation will eventually swamp the benefits, causing scientific progress to grind to a halt. To avert this scenario for as long as possible, governments and companies need to find ways to make each dollar of research spending go further.

One idea for how to do this comes from an interesting new paper by biologist Wayne Wahls of the University of Arkansas for Medical Science. Wahls looked at the way National Institutes of Health research grants are awarded, and finds a major inefficiency -- too many dollars are flowing to big, famous laboratories.

Prestigious institutions had 65 percent higher grant application success rates and 50 percent larger award sizes, whereas less-prestigious institutions produced 65 percent more publications and had a 35 percent higher citation impact per dollar of funding.

Wahls compared medical schools, and found that -- unsurprisingly -- prestigious institutions like Harvard and Stanford get more funding than low-ranked institutions like the University of North Dakota or the University of Mississippi.

The high-ranked schools also tend to be significantly more likely to get projects funded.

In order to measure whether that money is being well-allocated, Wahls looked at citation impact -- the number of citations given to papers published as a result of the grants, adjusted for the quality of the journals in which the citations occur. This measure is equivalent to the one used by the NIH to measure research effectiveness. He found that the less-prestigious institutions tend to do more with less.

Wahls examined some other measures of research productivity -- for example, the number of publications per dollar of funding -- and found the same result. Nor is he the only researcher to notice this unpleasant correlation -- a recent paper by Michael Lauer, Deepshikha Roychowdhury, Katie Patel, Rachael Walsh and Katrina Pearson found the same.

In other words, granting agencies are watering the green spots -- funneling money to prestigious laboratories, while leaving humbler labs starved of funds. The result is inefficiency and waste, with the big famous labs failing to find productive ways to deploy the government’s largesse.

Wahls attributed this to the so-called Matthew Effect -- i.e., granting agencies simply being star-struck by famous names like Harvard. Another reason might be that granting agencies tend to reward research projects that are similar to efforts that have been successful in the past, eschewing blue-sky ideas in favor of follow-up investigations. Prestigious institutions are also more likely to hire professors who have done successful research in the past. But if ideas have diminishing returns, this method of selecting promising research could be backfiring. The NIH has attempted to compensate for this by creating so-called high-risk high-reward grants, but even those awards may simply flow to famous institutions.

A better idea would be to funnel more funding to young researchers who are just starting out, instead of paying famous people to follow up on their successes. Granting agencies should consider reserving some of their funds specifically for younger, less established researchers and lower-ranked institutions.

In general -- not just in granting, but also in salaries and hiring -- the US academic system tends to heap rewards on professors who have done great things in the past, instead of those who might do great things in the future. But while that might create great incentives for young researchers to work hard, it might also be starving them of the resources they deserve.

Rich donors, too, should rethink how they allocate their money. Getting a building or a school named after you at a famous university is fun, but huge gifts like hedge fund billionaire John Paulson’s $400 million donation to Harvard’s engineering school in 2015 might benefit human progress more if they were given to scrappy, struggling low-ranked schools.

Ideas may get more expensive to find, but that doesn’t mean we have to surrender to the trend.


Noah Smith
Noah Smith is a Bloomberg Opinion columnist. -- Ed.

(Bloomberg)