Essays on Entrepreneurship in Developing Economies

I explore the theoretical foundations and the empirical relevance of the idea that entrepreneurship in developing countries could arise out of poor or non-existent alternatives. In the first chapter, I use an occupational choice model to show that the observed large increase in the rate of business ownership in rural Thailand during the 1997 Asian crisis can be explained by a negative shock to the labor market. According to my GMM estimates, a 47% fall in the outside option of entrepreneurship is required to explain the observed increase in business ownership from 17% to 37% between 1997 and 1998. I find that endogenously starting a business enabled households to offset about 40% of the income loss during the crisis, but also that low entrepreneurial productivity limits the extent to which pro-business policies can stand in as unemployment insurance for the average household. In the second chapter, joint with A. Karaivanov, we explicitly model and distinguish between voluntary and so-called involuntary entrepreneurship, which arises for those who prefer the non-business occupation (e.g., wage work) but cannot obtain it (with some probability that we estimate), due to labor market frictions. We also allow for credit constraints and analyze their interaction with the labor market constraint. We estimate the model via GMM using data from semi-urban Thailand from 2005, and find that 11% of all households in our sample (approximately 17% of all households running a business) are classified as involuntary entrepreneurs. While there are large potential income gains, especially for poorer households, from relaxing either the labor market or credit constraints, involuntary entrepreneurship can only be significantly reduced by addressing the labor market constraint. In the final chapter, I structurally estimate a model in which risk neutral agents maximize total income by optimally allocating capital and labor into entrepreneurship, subject to credit and time constraints. I estimate the model via GMM using 2005 Thai urban data, where about 20% of business owners report a second occupation. I find that while most entrepreneurs that hold two jobs are skill-constrained (the first-best scale of the business does not exhaust the time-constraint), there is a small fraction that are credit-constrained. These two groups within the multiple occupation group are also predicted to be considerably heterogeneous in terms of initial wealth, schooling and entrepreneurial talent.