Commuting and Innovation: Are Closer Inventors More Productive?
May 18, 2020
My dissertation, which is composed of three papers, looks at how geography and housing markets affect economic outcomes at the individual, firm, and industry levels, both within and across regions. One of the papers, co-authored with Andy Wu, is particularly relevant to the study of cities. In “Commuting and Innovation: Are Closer Inventors More Productive?” we aim to understand the causal effect of commuting on inventor productivity.
In the United States, commuting takes up more than 50 minutes per workday for an average worker in 2018. Commuting costs employees time and money, and may also be costly for employers who pay in wage compensation and lost productivity.
Both the direction and size of commuting’s effect on inventor productivity are not well understood. On the one hand, longer commutes could lower total productivity by decreasing incentives to exert effort at work, constraining worker schedules, and raising stress-related costs. On the other hand, these effects could be mitigated by wage compensation that encourages worker effort, more time for brainstorming, and lower stress due to greater mental distance between work and family life.
Using a difference-in-differences design, we compare the productivity of inventors working at the same firm before and after a firm relocation. For every 10 km increase in commuting distance, we find a 5 percent decrease in inventor productivity measured in raw annual patent counts and a 10 percent decrease in inventor productivity for measures of patent quality, such as average scaled citations. These results are robust to a variety of commuting cost measures and model specifications, are not driven by any one metropolitan area with particularly long commutes (such as the San Francisco Bay Area), and show no evidence that firms are moving closer to inventors that are becoming more productive. These results are economically significant given the average pre-relocation commuting distance of around 20 km (the average for inventors in our data). Our findings suggest that any commuting-related wage compensation is far below what is needed to incentivize longer-commute inventors to work with the same level of effort as shorter-commute inventors.
Hongyu Xiao is a 2020 Doctoral Recipient, The Wharton School. After graduating, he will be Senior Economist at the Bank of Canada.
