Open Collections

Abstract

The spatial distribution of economic activities is shaped by the way in which firms operate. I study firms' intra- and inter-national decisions and how their responses to policies affect the social welfare. First, I develop a quantitative model of multi-plant oligopolists where firms decide on plant locations, export and pricing, taking into account competition and cannibalization. I advance the existing trade literature on multinational firms by allowing for interdependent entry, oligopolistic rivalry and variable markups. Despite having a combinatorial discrete choice problem, I provide a toolkit to estimate the model in three steps. I present simulation-based evidence to show that neglecting interdependencies among plant locations within a multi-plant firm introduces quantitatively relevant differences in estimation. Second, I estimate the multi-plant firm model developed in the first essay focusing on the cement industry in the US and Canada. Estimates reveal important features of this industry, especially fixed costs as a key determinant of the market structure. I apply the estimated model to quantify firm-level responses to changes in environmental, trade and competition policies, and highlight the welfare implications of having multi-plant production. Results indicate that a carbon tax of $50/tCO2 in Canada will generate carbon leakage to the US which offsets 18% of the domestic emissions abatement, and also exacerbate the market distortion. A unilateral tariff increase of 20% reduces domestic welfare and does not lead to significant increase in number of plants in the US. I also analyze a hypothetical acquisition in which the efficiency gains of multi-plant firms offset the effects from competition-lessening. Last, I study firms' sourcing decisions through rules of origin (RoOs) in regional trade agreements. They are a set of criteria that define the origin of a product to qualify for preferential access. This chapter distinguishes trade diversion through RoOs from tariff reduction on intermediate goods, focusing on the automotive industry. I find that RoOs significantly distort where to source auto parts and the effect is nonlinear depending on the restrictiveness of rules. The shift from foreign to regional inputs increases before a stringent content rule backfires.