This post is an attempt to explain and offer suggestions toward an impressive working paper by the Equality of Opportunity Project, which you can find here. In short, the paper goes on to explain the various factors behind inventors in America. The results that the paper establishes are:

• Children belonging to high-income families are ten times as likely to become inventors as those belonging to below-median income families.
• Exposure to innovation during childhood has significant causal effects. For example, growing up in a neighborhood or family with a high innovation rate.
• The financial returns to innovation are extremely skewed and highly correlated with their scientific impact as measured by citations.

As much as the other sets of results are fascinating in their own way. The set that jumped out at me was the middle one – the Exposure Effect. Unsurprisingly, the exposure effect is also the hardest to empirically analyze, primarily due to distinctions between causal effects and correlations. Here’s a snippet from the paper detailing why neighborhoods are so important, “Children who grow up in commuting zones (CZs) with higher patent rates are significantly more likely to become inventors, even conditional on the CZ in which they work in adulthood.” Bell, Chetty, Jaravel, Petkova & Reenen (2017)

The other key insight from the paper was that if girls were exposed to female inventors as boys are to male inventors in their childhood CZs, the gender gap in innovation would be half as large as it currently is. (Bell et al. 2017) The most important realization from reading the paper is that children from low-income families, minorities (in some cases) and women are less likely to ever have the exposure to have higher rates of innovation. The paper though does offer three types of policies to remedy the situation.

• Increasing exposure (Internships etc)
• Reducing barriers to entry (Subsidies for certain groups)
• Increasing private financial incentives

While a remarkable paper in many aspects, the exposure effect I feel doesn’t factor in a different kind of exposure. An exposure that’s not conditional on commuting zones/school district or any geographic space, but the kind of exposure that’s the most prevalent in children’s lives today- electronic. While the paper does provide empirical proof that moving a child from a CZ that is in the 25th percentile of the distribution in innovation (eg., New Orleans, LA) to the 75th percentile (e.g., Austin, TX) would increase a child’s probability to become an inventor by at least 17% and a maximum of 50% [Not suggesting/endorsing moving to Austin though. It’s a shell of its former self and has enough people moving that we may just need to build a wall 😉 ] Unsurprisingly, the most innovative zones are cities. It’s consistent with previous research done into cities where they act as a great network for diffusion of ideas. The power of scaling has been well documented from the Kleibers rule/Allometry, where the larger the organism (and brain), slower the heart beat, and greater the lifespan of the organism to Geoffrey West’s work with urban planning. Geoffrey West’s recent book about scaling offers an extremely nuanced view about why cities are major hubs for innovation. West’s research mostly involves understanding that cities are largely the solution to problem of cities and their need for constant growth fuels innovation. Cities act as great person to person networks helping bring various ideas together. (Picture cities acting as neural networks)

The inventors paper though discounts the impact technological exposure can have on a child’s ability to innovate. They’re obviously trade-off’s that need to be made with electronic exposure, as you have to with any kind of exposure, but if carefully designed for children, the benefits are quite great. A child who doesn’t belong to a specific geographic or income distribution can go online and largely have the ability to immerse himself in the same concepts that a child from a higher distribution does. Electronic exposure could go some way in bridging the information asymmetry that might exist between children in different geographic distributions. I must admit though that with the data currently available its extremely hard to prove causality in terms of electronic exposure. But in purely theoretical terms, electronic exposure should help narrow the gap created by income and geographic distributions.

The benefits of agglomeration and scaling are well documented and quite obvious, and this post isn’t about highlighting those. It’s about figuring out how the lives of those who haven’t benefited from agglomeration could be better off. We’ve clearly seen that certain cities and neighborhoods are better to move to not only because of their immediate financial incentives but also due to the fact that they can increase the chances of a child going on to innovate. The discouraging fact is that the U.S. currently lacks the required infrastructure to sustain such an exposure.

(This blog post first up though started out by trying to understand if there was any causal effect between broadband connectivity and the share of inventions as categorized by the paper. It’s something that I’ll continue looking for as I evaluate various data sets.)

While urban centers are largely immune from poor internet connectivity. Broadband and fiber optic connectivity in rural America is abysmally poor. The states that experts tend to refer to “flyover” states do even worse among states with rural populations as witnessed below.

States that have huge rural populations also have the lowest fiber optic and broadband connectivity (finally a causality!). Rural internet connectivity is today where rural electrification was in the 1930’s when FDR’s stimulus under the new deal helped solve that problem. Most of rural America is still stuck in the dial-up age and alternative internet technologies such as satellites are unable to bridge the deficiency. Connectivity in rural areas today is inadequate and costs too much. While I admit reaching thinly populated areas is harder and thus expensive to build. Unfortunately, this also means that children in rural areas are not only living in areas that are geographically isolated to ideas, but also electronically isolated. The result is that we’re currently hurting the prospects children have of increasing their mobility. Simply put, children today don’t have a better chance of succeeding compared to their parents. An effect that’s even more pronounced in rural areas. (For more on income mobility I’d highly recommend another one of Professor Chetty’s papers)

For America to maintain its current innovative zeal it needs a constant stream of different ideas. Differentiability that might just depend on a different kind of exposure.

Sources

https://www.wsj.com/articles/rural-america-is-stranded-in-the-dial-up-age-1497535841

https://apps.fcc.gov/edocs_public/attachmatch/DOC-349074A1.pdf

http://igeographer.lib.indstate.edu/klier.pdf

http://pages.cs.wisc.edu/~pb/tubes_final.pdf

http://www.equality-of-opportunity.org/assets/documents/inventors_paper.pdf

https://www.google.com/search?q=Co-Mo+Electric+Cooperativ&oq=Co-Mo+Electric+Cooperativ&aqs=chrome..69i57j0l3.644j0j7&sourceid=chrome&ie=UTF-8

https://broadbandnow.com/Fiber

https://thefinancialbrand.com/36981/bank-branch-closing-trends/

http://www.pewinternet.org/fact-sheet/internet-broadband/