Data issues in analyzing agri-food trade in BIMSTEC: Challenges and recommendations

The focus of research on international trade has recently shifted from industries and countries to firms. Firm heterogeneity is shown to be a determinant of trade at both the intensive margin (increase exports per firm/product) and extensive margins (the number of firms exporting – new products, new partners, new varieties, and new prices). It is now widely accepted that exporting firms are larger, comparatively productive, more skilled, and capital-intensive, and pay higher wages than non-exporting firms. The innovations in international trade literature that explains both the emergence as well as levels and the nature of trade flows through value chain integration necessitates examining trade-based exchanges at the highest possible levels of product disaggregation. Developments in trade theory emphasize that it is individual firms not countries that trade and analysis needs to incorporate firm characteristics in decisions and ability for exporting and importing. Firms are the appropriate unit of analysis for trade flows. It helps several paradoxes once the import of firm heterogeneity is understood. Despite the substantive importance of granular level data and the significant level of disaggregated product-level bilateral trade flow data and enhanced computing power that are becoming available, most studies have tended to rely on analysis with high level of aggregation. Recent research on firm heterogeneity in international trade highlights the importance of extensive margins i.e., new products, new partners, new varieties, and cumulative of these i.e., new prices in trade patterns and firms' responses to trade liberalization and other policy changes. However, the high dimensionality of the data and the large number of responses to changes can easily overwhelm researchers. Additionally, bigger data sets may contain more noise, which can mask important systematic patterns. In analysis of trade flows, notwithstanding the rising incidence of differentiated products (varieties) and value chains that transcend national boundaries, methods in agri-food trade analysis in particular have not kept pace in terms of empirical methods and suitable data.