How do firms establish new export relations and what determines the dynamics of exports at the firm level? Finding convincing answers to these questions is important for trade theorists and policy makers alike. While the former try to assess which trade model adequately describes export dynamics, the latter would like to understand which policies are effective for promoting exports.
The most prominent models of export dynamics rely on sunk fixed cost to enter the export market. Such costs can explain why only few very productive firms export (Melitz (2003)), why firms’ export status is very persistent over time and why the probability that a firm exports is determined primarily by its past export status (see Roberts and Tybout (1997) among others). However, a growing number of micro studies on export dynamics (Eaton, Eslava, Kugler and Tybout (2007), Buono, Fadinger and Berger (2008), Lawless (2009)) has revealed evidence that is at odds with this traditional view.
First, export values are usually small when a firm breaks into a new market. Second, most export flows have a very short duration (one or two years), few survive for a longer period and these grow fast. This leads to hazard rates that are sharply decreasing with duration and fast growing export values conditional on survival. Finally, a novel stylized fact, which we uncover in the present paper, is the positive relation between persistence of export flows and the quality of legal institutions in the destination country.
In order to explain these observations, we believe that it is crucial to consider that exports at the firm level are relationship-specific. Most exporters and especially those who try to penetrate a new market neither sell a perfectly homogeneous good that can be sold in an organized exchange nor do they own a distribution network in the export destination. As a consequence, exporters need to rely on partners in each market. These are either trade intermediaries, distributors that locally market the exporter’s product or firms that import the good to use it as an input in their production process.
In our model, firms that want to start exporting to a specific country have to search for a partner in that destination. If an exporter is matched with an importer, she is initially uncertain about the importer’s reliability. Contracts are incomplete, so that some partners may try to hold-up the exporter. Whether an importer has incentives to do so depends on the value of short terms gains from holding up the partner relative to the value of maintaining a long term relation. This depends among other things on the exporter’s productivity, the importer’s type (patient or impatient), the extent of sectoral contracting frictions and the quality of legal institutions in the destination country. Patient importers value future profits from any relation sufficiently in order to respect contracts with all exporters. Differently, impatient importers try to renegotiate contracts ex post if contracting frictions are severe (the payoff from renegotiation is large), legal institutions are weak (the opportunity to renegotiate is big) and exporters are relatively unproductive (the expected value of future profits is low). Since exporters have to learn their partners’ type through experience, uncertainty is initially large and thus export values are small. As an exporter observes that the contract is respected she becomes more confident that her partner is reliable and the value of exports grows.
The combination of these ingredients leads to several interesting patterns. Here, we focus on the more important ones. First, matching frictions generate persistence (state dependence) of export decisions, even though there are no sunk costs in the model. Exporters are unwilling to give up a partner unless they are sure that she is unreliable. Second, better legal institutions make it more likely that a given relation survives from one period to the next. As a consequence, higher legal quality implies more state dependence and reduces hazard rates. Moreover, this effect is larger the more severe contracting frictions are in a given sector. Similarly, larger destination market size or higher productivity of the exporter implies that a given relation is more valuable for importers and thus makes it less likely the she will not honor the contract. Hence, state dependence is larger (and hazards are lower) in destinations with larger markets and for more productive exporters. Moreover, hazard rates decrease with the age of the relation because partnerships involving unreliable importers are sorted out, while relations with reliable partners survive in the long run.
Download
Learning, Incomplete Contracts and Export Dynamics: Theory and Evidence from French Firms
