Social Push and the Direction of Innovation
How do barriers to innovation/entrepreneurship affect consumers, growth, and inequality?
There's a growing body of evidence that women, minorities, and individuals from low SES backgrounds are extremely underrepresented in the innovation system ("unequal access").
Children whose parents were in the top 1% of earners were 10 times more likely to be inventors than those whose parents were in the bottom 50 percent (Bell et al., 2019).
Women represent a small fraction of patent inventors, venture capital backed entrepreneurs, venture capitalists, tech industry workers
Further research has highlighted various mechanisms driving these differences
Preferences, discrimination (esp. in obtaining financing), exposure effects
Missing from the literature:
Implications for direction of innovation
Implications for growth and inequality under a general equilibrium framework
App-level consumer gender ratios, broken down by gender of the founders of the venture-capital-backed company that created the app (red bars are companies with any female founders, white bars are male-founded). Source: Nielsen EMM data and Crunchbase
Descriptive Evidence on the Implications of Unequal Access for Direction of Innovation
Methodology: we bring together data on innovator characteristics and consumer characteristics:
Micro data: venture-backed entrepreneurs and who they sell to (phone apps, consumer packaged goods)
Macro data: industry level data on patent inventors and entrepreneurs matched to CEX data on average consumer characteristics by industry.
Broad patterns
Women sell more to women; same patterns hold for age, parental income, geography
Holds within narrow product categories and across industries
Additional evidence
Women entrepreneurs are much more likely to be funded by female venture capital partners
There are very few venture-backed African-American entrepreneurs in the consumer packaged goods space.
Quasi-Experimental Evidence: Does Personal/Social Experience Matter for Innovation Direction?
The descriptive evidence suggests that personal background seems to matter for the direction of innovation pursued by an inventor/entrepreneur.
Obtaining variation in personal background is difficult. We leverage data from Finland, where conscripts are quasi-randomly assigned to roommates. In this setting, we have sufficient information on the background and subsequent innovative activities of individuals.
Key Findings
Individuals from high-income families exposed to roommates/classmates from low-income families are more likely to pursue entrepreneurship in sectors catering to low-income individuals.
Lessons from a Product Variety Growth Model
Model is based on the product variety model of endogenous growth in Romer (1990). The model has a few advantages
It incorporates general equilibrium effects (balance between research workers and production workers)
We extend to a two-sector version: allows us to map our empirical findings to implications for growth and inequality.
Key Findings
Access barriers do not matter if individuals have homogeneous research productivity. With heterogeneous ability, access barriers reduce long-run growth.
In a two-sector model with taste heterogeneity and inventor specialization (people are better at innovating in the sector they prefer), access barriers reduce growth rates and increase inequality across the two groups.
A second-best solution to access barriers could be to have more social interaction across the two groups. Increased interaction generally increases growth and reduces inequality, holding access barriers fixed.