Finding and Training the Missing Inventors

In “How Some Universities Translate Inclusive Innovation Into Regional Growth,” (Issues, Winter 2025), Mercedes Delgado and Fiona Murray stimulate a potential research agenda to understand the factors affecting women’s movement into the STEM fields (science, technology, engineering, and mathematics) and innovation. They are surely correct in focusing our initial attention on universities for their role as agents of change.

University effects on their local economies can be difficult to isolate and measure, with numerous confounders. Still, my work has shown that innovation from universities generates increased corporate innovation, employment, wages, establishment entry, and venture capital investments in local industries most closely related to the university’s technological strengths. When a university has stronger effects on its local economy, then, we observe more extensive local agglomeration of related industries, clustering nearby to take advantage of the university’s ideas.

Universities also feed skilled workers into the local ecosystem, and to the extent that they train women for inventive careers, they may increase the female share of the local innovation economy. In regions with weaker ecosystems, the strongest graduates are likely to leave for more innovative places. Data on post-graduation placements could help assess the contributions of universities to the STEM workforce by geographic area and help interpret the link between university and industrial female shares.

Because women’s occupations feed back into girls’ career decisions, the effects of training women compound themselves across generations. Exposure to female inventors in the local labor force (and at home) increases girls’ propensity to go down the path of invention themselves. We know from a study by Alex Bell et al. that the nation has many “missing inventors”—individuals with the intelligence to invent but without the childhood experiences to nudge them in that direction. Learning how to elicit these (generally female and minority) missing inventors would promote economic growth by increasing production of new ideas that raise our standard of living.

So why do some universities seem to be better than others at drawing women into invention at high rates?

To structure further research addressing this question, a useful guide can be found in the work of Harvard University economist and Nobel laureate Claudia Goldin on the dramatic changes that occurred during the late nineteenth and twentieth centuries in women’s education and labor force participation. In medicine, for example, women initially entered lower-ranked institutions at higher rates due to some combination of quotas, discrimination, and outright exclusion at top schools, as well as geographic and financial barriers women faced in attending non-local universities. Similar patterns have played out in engineering and science, as well as in law and business. Changes in job flexibility, structure, and pay affect men’s and women’s career choices. Considering these factors can help us organize a variety of testable hypotheses for why women may be entering certain university programs and areas of invention fastest. For example, federal and industry-sourced research funding is substantially more plentiful at the top-tier universities and in larger ecosystems. A gap between men and women in how much funding they require to forgo employment and enter a PhD program may in part generate this divergence in enrollment across universities.

One need not be pro-woman or pro-inclusivity to support programs that find and train the nation’s missing inventors. Evidence on the distribution of skills suggests merit-based programs will find them, so long as we can keep them increasingly interested in STEM. Our standard of living rises to the extent we don’t miss their future inventions.

Naomi Hausman

With their choice of title, Mercedes Delgado and Fiona Murray put a hopeful touch on what arguably is a task that still requires significant effort: fully leveraging the potential of talent across all segments of society for innovation. The authors look in particular at female inventors, a group that might be indicative of challenges faced by others as well.

Their analysis is driven by concern that the United States is currently failing to fully leverage its available talent. The basic statistics on the gender perspective of this issue are striking. Since 2006, every year more PhDs in total are awarded to females than males, with females now outnumbering males by 3:2. In the STEM fields this ratio is the opposite, with males outnumbering females 3:2. In the new data on STEM inventors that Delgado and Murray report, the male to female ratio is 3:1 for inventors at universities and close to 5:1 for inventors overall. At least so far, the conversion of advanced education into patenting is not happening at the same rate for women as for men.

What are the reasons and what should be done? This is a discussion that comes at a time when a framing of more diversity being automatically better will meet headwinds. Some observers express concern that efforts to increase diversity may lead to a reduction of quality, not raise quality by reducing barriers for qualified individuals. The data shown will not convince these critics. To be fair, assessing this point goes beyond the essay’s scope, which is more focused on showing the heterogeneity across universities and regions on the female inventor share. But it will take such evidence on how diversity indicators relate to performance, rather than reporting diversity itself as performance, to advance the debate. For concrete policy advice, it will then also be critical to assess which measures to enhance diversity are indeed likely to enhance performance by removing barriers. Here, one might assume, the “how” will matter a fair bit.

The authors note that the female share of inventors is much lower in the broader economy than in universities. This could reflect differences in the barriers they face outside academia, or in the type of patenting happening there. There are also intriguing comments on the importance of the match of universities’ patent portfolio with the structure of the surrounding regional economy. This is a topic that deserves attention even before digging into the relation to diversity.

In sum, the essay does what a brief discussion such as this should do: present new facts, stimulate further thinking, and raise additional questions. At least for me, it created the appetite to read more from these authors—and more broadly as well—on this issue.

Christian Ketels