Entrepreneurship, job search and occupational choice in developing countries

This thesis is about entrepreneurship within the context of occupational choice in a frictional labour market. It unites two perspectives on entrepreneurship: the perspective of the entrepreneur as a firm, and the perspective of the entrepreneur as a self-employed worker. My emphasis is on developing countries, where a large share of the workforce are self-employed entrepreneurs. I combine theory with empirical applications using labour-force panel data from Mexico, Colombia and Ghana. The thesis consists of four chapters. The model in Chapter 1 provides a theoretical framework to capture the nexus of occupational and firm size choices of entrepreneurs. In the model, these two choices are non-separable. Entrepreneurs face a trade-off: they could either make the most of their entrepreneurial potential, or run a smaller firm while waiting to take the next attractive job opportunity. Better opportunities in wage employment increase incentives to keep firms small. Chapter 2 tests this model. To obtain variation in alternative employment opportunities for entrepreneurs, I exploit trade-induced changes to wage employment across local labour markets in Mexico. Using a large dataset built from several sources, I find evidence that is consistent with the theoretical predictions. I use guidance from theory, supplementary evidence, and an event study design to distinguish the occupational choice mechanism from competing channels. Chapter 3 develops an empirical model where the observed allocation of individuals to occupations is a product of both the attractiveness and the availability of job opportunities. This type of selection mechanism is central to each of the other chapters in this thesis. The chapter is concerned with the identification of such a model that combines comparative advantage and search frictions. I estimate the model on panel data from urban Colombia and find evidence for both search frictions and selection on comparative advantage. Finally, Chapter 4 builds and estimates a structural life-cycle model of sectoral choice between self-employment and wage-employment, costly job search, and capital accumulation in Ghana. We incorporate two central types of market frictions: credit constraints for entrepreneurs, and search frictions in the wage labour market. This provides a unified framework to think about career paths in developing countries. We link the estimated structural model to recent experimental literature in development economics, and use simulated experiments to interpret the mechanisms behind behavioural responses to labour market shocks.