Abstract: I document several facts about contingent work: there is greater dispersion and larger changes in hours worked by contingent workers, and contingent workers' hourly wages are 11.5 percent lower than traditional employees'. I develop a novel model of how individuals and firms choose contingent work or traditional employment. Contingent work offers hours flexibility to individuals but traditional employment earns a higher wage in equilibrium and has the security of unemployment insurance. Firms hire traditional employees before observing their TFP and must pay administrative costs to hire or fire traditional employees. They can hire (less productive) contingent workers flexibly without these constraints. I show that the recent development of apps (such as Uber) that make contingent work easy to find lowered the optimal UI replacement rate from 48 percent to 41 percent, which shows that contingent work provides valuable insurance to all workers in the economy. I also analyze a recent policy change that extended UI to contingent workers. This policy generates welfare losses of 0.08 percent when the UI programs are funded by separate tax rates. Funding both UI programs together with a single tax generates welfare gains of 0.09 percent.
(Accepted at the Journal of Political Economy)
Abstract: This paper provides new insights into the nature of entrepreneurship using a novel panel dataset based on U.S. administrative data from the Internal Revenue Service and the Social Security Administration. These data are used to analyze patterns of income growth and determinants of entrepreneurial choice for a large population of business owners. Earlier studies relying on household survey data have been limited by small samples, short panels, and income top-coding and, as a result, have focused on the typical self-employed individual rather than the typical dollar earned in self-employment. Without these limitations, we find that self-employed individuals have significantly higher average income and steeper, more persistent income growth profiles than paid-employed peers with similar characteristics. Contrary to the survey evidence, we find a much smaller role for non-pecuniary motives in driving entrepreneurial choice and little evidence for inordinately high risk factors or startup costs impeding entry. Linking individual and business filings, we find that business founders have sufficient resources in the initial years of operation to ensure positive individual income despite the fact that most claim a loss on the business.
Abstract: Many developed countries provide housing subsidies to low-income renters. Recipients in the U.S. pay a fixed percentage of their income towards housing, and the subsidy covers the remaining cost. In other countries, the subsidy is instead calculated as a percentage of the contract rent. This paper documents evidence of search and matching frictions in the rental housing market and examines how these frictions impact the costs and benefits of income-based and rent-based housing subsidies. I develop a directed search model of the rental housing market and analyze the equilibrium effects of subsidies. Under rent-based subsidies, higher matching frictions increase costs for both the households and the subsidy provider. On the other hand, when subsidies are based on income, households bear most of the additional costs of increased matching frictions. In general, income-based subsidies distort households' search behavior more than when subsidies are rent-based.
Working from Home and Human Capital Accumulation