What prevents most firms from exporting abroad? What prevents most exporters from accessing geographically remote markets? In this research, the PI proposes a new theory of international trade where firms worldwide are connected to each other through a global network of importers and exporters. Building on recent discoveries in network theory, the PI generates precise predictions for the entry of individual exporters into foreign markets. The PI uncovers and explains robust empirical regularities on the distribution of the number of foreign markets accessed by firms, the geographic distribution of those foreign markets, and the time-series of entry into foreign markets.
The model is built on the premise that before exporting to a foreign country, a potential exporter must acquire a contact there. The PI postulates that potential exporters meet foreign contacts both at random, and via the network of their existing contacts. This dynamic model generates two main predictions. First, firms that already export to many markets are more likely to enter new markets, so that the distribution of the number of markets accessed across firms is fat-tailed. Second, even if geographic distance impedes the formation of trade links with faraway countries, exporters are able to gradually expand geographically by radiating from countries where they already export. Both predictions are supported by the data.
This project is potentially important for three main reasons:
(i) The model is among a handful of dynamic models of international trade. The proposed research replicates key features of the observed dynamics of firm level trade, and explains how those dynamics are connected to strong empirical regularities for the cross-section of firm level trade. It does so in a tractable way, so that this model can easily be generalized and used as a framework for answering a variety of questions in international trade and beyond.
(ii) The model provides an explanation for the unequal ability of individual firms to enter foreign markets. The predictions of the model regarding the number and location of the foreign markets accessed by individual firms are supported by the data. Existing theories cannot explain those facts quantitatively.
(iii) By generating predictions on the geographic distribution of exports, the model can explain the role not only of size, but also of distance in the "gravity" equations in international trade. Since its discovery in 1962 by Tinbergen, no one has been able to explain the precise (and surprisingly stable) role played by geographic distance. The proposed model will potentially offer the first explanation for this long lasting puzzle.
In addition to enhancing our basic economic knowledge, the project will offer valuable tools for answering two types of questions:
(i) This network model allows to model the propagation and magnification of economic shocks internationally. Identifying the potential source of global shocks, be they firms or countries, is obviously an important goal for policy research.
(ii) The proposed model can be used to study the connections of firms in non geographic spaces, such as the space of industrial sectors in industrial organization or international finance, or the space of cultural, religious or genetic differences in international trade.
