This dissertation studies the relationship between entrepreneurship, labor markets, and business cycles. Chapter 1 studies how the labor market conditions affect the formation and growth potential of new businesses over the business cycle. I develop a dynamic occupational choice model with labor market frictions and joint firm and worker dynamics, in which heterogeneous individuals choose between being employed/unemployed workers, subsistence self-employed, or entrepreneurs with potential to grow. Using U.S data, I provide support for the following predictions. First, unemployment makes more people willing to start a business because of the lower outside option. Second, a lower job finding rate reduces the value of the fallback option when the business fails (harder to find a job), deterring entry from employment, especially for high-skill workers for whom the wage loss is greater. Third, high-growth startups are mostly started by high-skill workers. Then, with the calibrated model, I study the dynamic response of the economy to a negative productivity shock. I find that (i) entry due to stopgap motives increases while entry into entrepreneurship declines, and (ii) the composition of founders shifts toward fewer employed high-skill workers, making new cohorts to have fewer high-growth startups. Both features hinder job creation recovery, keeping the labor market depressed longer and the entrepreneurship entry persistently low.

Chapter 2 studies the short- and medium-term effects of employment spells in startups on future earnings of employees. Using Chilean Unemployment Insurance data, we find that those who moved toward startups earned over the next five years 21% less than those who transitioned to established firms. However, when taking selection into account, the difference is reduced to -14%, which implies that an important part of the observed effect comes from sorting. We further decompose this effect and find that 11 percentage points of the overall 5-year effect come from lower average wages. The remaining 3% arises from more unemployment periods. We also find that the contemporary differential, measured as the effect over the first year after the transition, is smaller than the medium-term effect. This suggests the existence of a scarring effect on earnings from working for a startup. Finally, we also find heterogeneous effects over the business cycle.

Chapter 3 provides a characterization of the business formation process in the U.S. from an individual-level perspective using the Survey of Participation Program (SIPP). The goal of this paper is twofold. First, it performs a detailed description of businesses and owners in the U.S. in terms of their characteristics. It also performs a detailed analysis of the survival probabilities and labor income differences between before and after self-employment spells. Second, it illustrates the potential of the SIPP as a source of information to study the early stage of business formation.