Essays in Labor Economics and Innovation

Abstract

This collection of essays applies empirical techniques to questions in labor economics, with a focus on innovation. A common theme among these articles is that changes in technology and institutions can have impacts on the functioning of the labor market and innovation.

Chapter 1 explores an understudied question in innovation: what is the relative importance of patent term (i.e., duration of patent protection) across different industries? This article (co-authored with fellow Princeton Ph.D., Neel Sukhatme) measures term sensitivity through patent applicants’ response to the passage of the TRIPS agreement, which changed how term was calculated. We find significant differences in term sensitivity across industries, some of which follows conventional wisdom (patent term is important in pharmaceuticals) and some of which does not (it also matters for software).

Chapter 2 investigates whether the emergence of ride-sharing companies like Uber and Lyft lowered the wages and employment of taxi drivers and chauffeurs. To evaluate the impact of ride-sharing services, this paper uses the timing of the rollout of Uber along with the company’s degree of popularity across cities to estimate the effects of competition from Uber on taxi and limousine drivers’ employment and wages. In contrast to the literature on disruptions in other industries, including the deregulation of the trucking and airline industries in the 1970s, I find no evidence that the advent of Uber has had a negative impact on wages in the taxi and limousine industry thus far.

Chapter 3 documents that it was the behavior of the record number of long-term unemployed during and after the Great Recession that explained the initial movements away from the Beveridge curve, which puzzled economists. I extend the calibrated model originally estimated in Krueger, Cramer, and Cho (2014) and focus on the channels through which duration dependence could be a factor. Incorporating the cyclical patterns in the labor force flows of the long-term unemployed (lower labor force withdrawal rates during recessions and higher labor force withdrawal rates as recoveries continue) improved the accuracy of the model’s forecast of the unemployment rate and produced the counterclockwise loops as predicted in Blanchard and Diamond (1989).