The importance of childhood education and the birth lottery for U.S. innovation

The U.S. economy could use more innovation these days. Despite the proliferation of apps and concerns about a robot takeover of the labor market, U.S. productivity growth just isn’t very strong these days. The exact reasons for weak productivity growth aren’t fully understood by economists, but a jump in innovation would seem likely to help increase it.

Getting such a boost, of course, is easier said than done. Most policies that seek to increase innovation in the United States are focused on getting as much out of current innovators as possible. But because the first seeds of new innovation sprout in many different ways perhaps policymakers should instead be focused on creating more innovators. New research shows that the key to increasing the ranks of innovators may reside in the childhoods of potential innovators.

The new research is from a working paper by Alex Bell of Harvard University, Raj Chetty of Stanford University, Xavier Jaravel of Harvard, Neviana Petkova of the U.S. Department of the Treasury, and John Van Reenen of the Massachusetts Institute of Technology. The paper takes a look at the background of innovators, as measured by individuals who filed for a patent between 1996 and 2014. (An assumption here being that patents are good indicator of innovation.) The authors have access to administrative data collected by the federal government on these patent holders that lets the economists take a look at the patent holders’ family backgrounds.

Unsurprisingly, children from low-income backgrounds are much less likely to end up getting a patent than children from high-income backgrounds. Patent holders also are more likely to be white and male. What explains these two gaps? The economists seek to answer this question by looking at the standardized test scores for all individuals who passed through the New York City public school system from 1989 and 2009. The test scores cover from the 3^rd grade to the 8^th grade. By linking this data with the data on patent holders, the economists can see how much of this innovator gap is related to a test score gap.

The co-authors find that only 30 percent of the difference in patenting rates between low- and high-income children can be explained by the gap in math test scores in the 3^rd grade. The amount of the gap for women and children of color explained by test scores is even smaller, at 3 percent and roughly 10 percent, respectively. So education, as measured by standardized test scores, clearly plays a role in patenting later in life, at least when it comes to explaining the income gap. But a significant portion of the income gap can’t be explained by educational differences.

So that leaves the advantages of growing up in rich households. Those advantages include wealthy family and community connections, better educational opportunities, and exposure to people working in innovative fields.

Current U.S. innovation policy is focused on getting more out of the innovators we already have, such as tax credits for research and development, which increases innovation on the intensive margin. But policymakers also should be focused on increasing innovation on the extensive margin: raising more children to become innovators. This would mean less focus on tax policies that reward innovation in the short term and more on policies that can help expand the pool of next-generation innovators to include more women, people of color, and those from low-income backgrounds. The upshot might be not only more economic growth but also more equitable growth.