Global Sourcing, Imperfect Competition, and the Network of Trade
DescriptionIn the past few decades, the fragmentation of production across national borders has given rise to the global value chain. Firms can now source inputs from the global market. Recent studies using micro-level trade data have allowed us to uncover new features of the underlying buyer-seller networks of trade. One of the most striking features of these networks is the dominance of large firms: the bulk of trade is between large firms which trade with many other firms. These phenomena call for novel frameworks to study firms' sourcing behaviour. This project aims to provide a new model with endogenous buyer-seller networks and imperfect competition in global sourcing. By exploiting panel datasets of firm production and transactional level trade data from China and France, we will estimate the model and use it to address the following new questions. How do firms exert market power in the presence of matching frictions with their buyers or sellers? How does imperfect competition shape the network structure and firms' sourcing behaviour? What are the pro-competitive gains by inducing tougher competition among buyers/sellers and the welfare gains from better matching between buyers and sellers? To that end, we first develop a new quantifiable international trade model with (i) two-sided buyer-seller heterogeneity, (ii) matching frictions, and (iii) oligopolistic competition upstream. Combining highly disaggregated data on firms and trade transactions for China and France, and exploiting exogenous variation in the market entry in China, we will present empirical evidence in line with the model that cannot be rationalized without features (i)-(iii). Downstream buyers import higher volumes and quantities at lower prices when upstream markets become more competitive. These effects are stronger for larger buyers and smaller when input suppliers are very heterogeneous. They also vary systematically with the elasticity of substitution across upstream input varieties and downstream output products. Our quantifiable model permits rich counterfactual analysis. The model allows us to assess the consequences downstream of upstream industrial policy and trade liberalization under different market structures and quantify the contributions of forces (i)-(iii). It also allows us to study the impact of lower barriers to entry upstream, lower matching costs and lower trade costs on firm productivity, firm size distribution, network structure, and aggregate welfare. These effects operate through a combination of matching of buyers and sellers, changes in the number of input varieties, and variable mark-ups.
|Effective start/end date||1/01/21 → …|