4. The Dynamics of Firm-Level Adjustment to Trade Liberalization

Abstract

We build a dynamic model of rm-level adjustment to trade liberalization that jointly incorporates the main salient features highlighted by recent empirical micro-level studies of rms and trade. Our model captures the joint entry, exit, export, and innovation decisions (subject to sunk costs) of heterogeneous rms as they adjust to trade liberalization. We characterize this industrial evolution over its entire transition path to a new steady state with lower trade costs starting from the time that trade liberalization is rst announced (but not necessarily yet implemented). We rely on numerical methods to solve for these equilibrium paths. In order to more accurately capture the dynamics of rm adjustments to trade, we model the sunk nature of market entry costs for both the domestic and export market as well as the per-unit and additional xed costs of exporting incurred in every period. Firm-level productivity evolves stochastically, and innovation involves a trade-o¤ between its cost and a return in terms of a “better”distribution of future productivity draws. Although the empirical micro-level studies of rms and export status initially emphasized the selection e¤ects of more productive rms into export markets, several recent studies have highlighted a separate channel for the e¤ects of trade on productivity operating through rmlevel improvements in productivity. Our model captures both of these channels for the productivity enhancing e¤ects of trade and analyzes their interactions over the adjustment path to lower trade costs. In particular, we highlight how the relative timing and magnitude of rm-level productivity improvements and export market entry decisions are also determined by non-technological factors such as the timing of trade liberalization announcements and the speed of liberalization. Under all these di¤erent trade liberalization scenarios (anticipated versus surprise, gradual versus sudden), we characterize both the distributional e¤ects across rms as well as their aggregate e¤ects for industrial performance. We nd that the anticipation of upcoming liberalization, and a more gradual path of liberalization (once implemented) induces rms to innovate ahead of export market entry. We are grateful to Robert Feenstra, Elhanan Helpman, and Esteban Rossi-Hansberg for helpful discussions and suggestions. We also greatly bene ted from all the discussions at the book conference where the paper was originally presented.