The effects of sunk costs on market structure, specialization and welfare

This dissertation examines the relationship between market structure, welfare and average productivity. In doing so, two distinct questions are integrated into a unified framework. The first question addresses the role of concentration indicators in accessing the welfare properties of industry equilibrium. The second question focuses on the way incentives for upstream specialization decisions are shaped by the downstream market structure.;The first chapter analyzes the relationship between market structure and welfare by developing a general equilibrium model in which exogenous and endogenous sunk costs determine market concentration, market power and productivity levels. In this framework, horizontal and vertical differentiation interact to uncover a nonmonotonic link between concentration and welfare. The nature of this relation is also shown to be sensitive to the type of sunk cost driving it. Exogenous sunk costs are positively correlated with market concentration and market power, but their relation to welfare approaches an inverted U-shaped curve. Endogenous sunk cost parameters are positively correlated with concentration, but negatively with market power, in the absence of collusion. Their relation to welfare approaches a U-shaped curve. These results have major policy implications, detailing new reasons why simple concentration indicators are a misleading statistic for welfare evaluations. The model suggests new tools in order to improve these assessments, making use of observable variables to proxy for unobservable cost determinants.;The second chapter extends the previous general equilibrium model with the purpose of identifying new interaction channels between downstream market structure and upstream incentives for specialization. A symmetric setting is established where final producers access a variety of intermediate goods and suppliers are also able to sell their inputs to multiple users. A specialization choice is made by suppliers, enabling more specific inputs to enhance the productivity of a particularly targeted final user. However, this also increases adaptation costs for all other downstream firms, the more so the larger their technical heterogeneity. The main results suggest that when stronger heterogeneity is associated to higher concentration in the final production stage, average productivity is negatively affected in large economies, but the impact on small economies is uncertain. Conversely, when higher downstream heterogeneity is associated to lower concentration levels, average productivity is negatively affected, regardless of economy size.