A new study claims that women, ethnic minorities, people from lower income backgrounds and from certain geographical regions are less likely to become inventors in the US. Increasing the rate of innovation among these groups could bear significant fruit, it states.

Released last week, the study – Who Becomes an Inventor in America? The Importance of Exposure to Innovation – was conducted by the Equality of Opportunity Project, a group of researchers led by Stanford economist Raj Chetty. It tracked the lives, from birth to adulthood, of over one million US inventors by using a new de-identified database linking patent records to tax and school district records. It determined who becomes an inventor based on patents granted between 1996 and 2014, and published applications between 2001 and 2012. The goal of the study was to identify the determinants of innovation and, thus, the effectiveness of current policies – such as investments in STEM education and financial incentives – aimed at increasing the rate of innovation in the US.

The research uncovered significant disparities in the numbers of inventors by sex, ethnicity and socio-economic class. Notably, the differences do not appear to be linked to innate ability. As seen below, early childhood maths test scores do little to explain disparities; children at the top of their class in maths and science are much more likely to become an inventor, but only if they come from a high income background. Indeed, Chetty and his team have found that children with parents in the top 1% of income distribution were 10 times as likely to have applied for a patent compared with their peers whose parental incomes are below the median.

It is suggested that the correlation between parental income and the likelihood of becoming an inventor is linked to three factors: endowments from parents may enable children a greater ability to innovate; lower income children may prefer other occupations, for example because of financial constraints causing higher risk aversion; and higher barriers of entry to becoming an inventor or lack of exposure to invention-based activity.

Race and gender also have a considerable effect on current levels of innovation. White children are more than three times as likely to file for a patent compared to black children; while Asian children are more than twice as likely to file for a patent than white children. Interestingly, when the data is reweighted to account for levels of parental income, the white-black gap narrows by a significant margin, while the white-Asian gap widens noticeably.

Differences in maths and science abilities also fail to explain the gender gap in innovation: the same third grade maths scores only account for less than 3% of the difference between the number of male and female inventors. While the percentage of female inventors has gradually increased with each passing year – at an average growth rate of 0.27% (see graph below) – the study notes that 82% of 40-year-old inventors today are men and that it would take 118 years at the current rate to reach gender parity.

Exposure to innovation is found to play a substantial role in the likelihood of children becoming inventors. Not only that, but it can heavily influence the type of inventions these individuals go on to produce. The study points out that among people living in Boston, those who were raised in Silicon Valley are far more likely to file a computer-related patent, while those who were brought up in Minneapolis were especially likely to apply for a patent in medical devices. This applies in a gender-specific manner too. Girls who grow up in an area with many female inventors in a particular technology class will be much more inclined to produce an invention in the same technology class. Chetty and his team estimate that the gender gap in innovation would fall by half if girls were as exposed to female inventors as boys are to male inventors.

The chart below shows the density of inventors per 1,000 children in the United States. Exposure to better neighbourhoods – in terms of mentoring, transmission of information and networks – is linked to higher levels of innovation. Note also that, with the exception of Austin, Texas, far fewer children from the South of the US grow up to become inventors.

In light of a decline in innovation in the US over the last few decades, the paper’s findings are especially important. It estimates that if women, minorities and low-income children invented at the same rate as high-income white men, the level of innovation in the country would quadruple. While the study does not offer specific guidance for policies, it highlights the fact that further financial incentives or the reduction of tax rates may only have small effects. Instead, increased innovation exposure for under-represented groups could have a much larger impact. The key takeaway here is that programmes should be targeted at women, minorities and those from low-income backgrounds. And, the background of individuals should be considered; for example, taking into account the fact that women are more influenced by female inventors.

Chetty concludes that it is important “[to create] more opportunities for exposure to innovation for kids who are very talented, excelling in math and science at early ages”. He adds: “And then you have systematic programmes to connect them with people from similar backgrounds who become mentors. I think you really need a role model that you can see yourself in that person’s shoes. That’s the type of programme I think could be effective here if done at scale.”


This article first appeared in IAM. For further information please visit www.iam-media.com.