Who Becomes an Inventor in America?

The U.S. Patent & Trademark Office (USPTO) collects only limited information about inventors –  typically only names, cities of residence, and mailing addresses. Research on who inventors are, and what factors might influence or encourage inventorship, must thus match the USPTO’s records with other sources of demographic information, or rely on inferences, such as an inventor’s gender based on his or her name.

A recent study by the Equality of Opportunity Project, Who Becomes an Inventor in America? The Importance of Exposure to Innovation, provides a wealth of inventor demographic information by linking inventors listed on patents and patent applications with federal income tax records. The authors define an “inventor” as someone listed as an inventor on at least one patent from 1996-2014 or patent application from 2001-2012. Inventors are matched by name, city, and state to income tax returns (1040 forms) and third-party information returns, such as W-2s. A subset of individuals are further matched to their parents (and thus to their parents’ tax and inventorship records) based on tax filings claiming the individuals as dependents. The linked (and subsequently de-identified) dataset allows the authors to examine the relationship between patent inventorship and income, parental income, location, college attendance, and parental employment and inventorship. For individuals who attended New York City public schools, the authors also link inventorship data to race, ethnicity, and math and language arts test scores. Using those scores as an (imperfect) proxy for ability, the authors evaluate whether such differences can help explain gaps in inventorship.

The study finds disparities in inventorship based on parental income, gender, and race and ethnicity that are not explained by differences in ability. Rates of inventorship increased with parental income; children born to parents with incomes in the top 1% became inventors at 10 times the rate of children of parents with below-median incomes.  Consistent with others’ prior research, the study finds substantial differences in male and female inventorship. Though the percentage of female inventors increased in each birth year cohort, the rate of increase was so slow that reaching gender parity would take over a century. Of inventors born in 1940, for example, 7% were female; of inventors born in 1980, 18% were female. Also consistent with others’ prior research, White and Asian children were more likely to become inventors (1.6 and 3.3 per 1000, respectively) than Black and Hispanic children (0.5 and 0.2 per 1000, respectively), even adjusting for income differences between the groups. (This finding was based only on the New York City sample, since race and ethnicity data are not collected in tax returns.) When the authors looked at whether ability could explain the inventorship gaps, they found that early test scores for the New York City sample accounted for only a portion of the gap by parental income (3rd grade test scores explained 31% of the gap), and very little of the gaps by race or gender.

The study also finds that children who were exposed to innovation were more likely to become inventors. In particular, children whose fathers were inventors were 9 times more likely to become inventors themselves (18 out of 1000, versus 2 out of 1000 for children of non-inventor fathers), regardless of parental income. (The study did not look at the effect of having a mother who was an inventor, which, in the study’s age cohorts, would have very unusual.) Even when fathers were not inventors themselves, their children were more likely to become inventors when the fathers worked in industries with high patenting rates. Moreover, children were particularly likely to be listed on patents in the same type of technologies as their fathers. Among children of inventor fathers, almost 1 in 1000 were inventors on a patent within the same narrowly-defined technology class as the father (even excluding instances where a child and father were co-inventors). In contrast, less than 0.2 in 1000 were inventors on a patent within the next closest of the study’s 445 technology classes. As technology classes became less similar to the father’s, the rate of inventorship gradually declined. The authors argue that this discontinuity suggests that inventor fathers make children more likely to become inventors through exposure to innovation, rather than genetics.

The location of a childhood home was also found to affect children’s inventorship rates and the technology areas in which they were inventors. Children growing up in “commuting zones” with more patents per working-age resident were more likely to become inventors, even among those children who moved away from their childhood commuting zone. For instance, children who grew up in the commuting zone containing Silicon Valley were among the most likely to become inventors (5.5 out of 1000). As with the influence of fathers and fathers’ industries, the study finds a positive correlation between inventorship rates of children and of residents in their childhood commuting zone, within both broad technology categories and narrowly-defined technology classes. Among commuting zones having at least 10,000 children, for example, children from Rochester, Minnesota – whose economy is dominated by the Mayo Clinic – were most likely to become inventors in the area of drugs and medical technology (1.1 in 1000).

Drawing from these findings, the authors conclude that there is considerable unmet potential for innovation. More innovation could be achieved, they reason, by interventions that increase exposure to innovation for children from under-represented groups who demonstrate early quantitative ability. Equating inventorship with innovation has its limitations (which the authors acknowledge). But the authors’ findings regarding the influence of childhood environment on inventorship help point us in the right direction for addressing the disparities identified by them and others.

The paper is available here, and the data is available here.

Amy Motomura is a CLB and SPINS fellow at Stanford, from which she received her JD degree in 2012 and a masters degree in Bioengineering in 2010.


Alex Bell, Raj Chetty, Xavier Jaravel, Neviana Petkova, & John Van Reenen. Who Becomes an Inventor in America? The Importance of Exposure to Innovation. NBER Working Paper No. 24062, December 2017.