Self-Selection versus Destination based Learning-by-Exporting: Firm Level Evidence from Pakistan

This study examines the self-selection and the learning-by-exporting hypotheses to determine the direction of causality between firm’s productivity and its export status for the textile manufacturers in Punjab, Pakistan. Using disaggregated product level data from the years 2000-2010, productivity is estimated based on the methodology by De Loecker et al. (2016). The study employes the Propensity Score Matching (PSM) and the Multivariate Distance Matching (MDM) techniques based on multiple matching algorithms. In line with the prediction of recent heterogeneous firm models of international trade, the main finding of the paper is that more productive firms become exporters, i.e., there is clear evidence of self-selection. This is mainly due to the large sunk costs associated with the liability of foreignness and a bigger risk cushion needed against uncertainties within the international market. However, the evidence for the learning-by-exporting hypothesis is less conclusive, indicating that exporting activities may not enhance productivity, unless the products are exported to high income economies. We also find evidence that firms exporting to high-income economies indulge in export sophistication as, in addition to the productivity enhancement, they improve on output quality, which is accompanied by capital accumulation and increased labor usage.