Firms in international trade: evidence from Indonesia

This thesis consists of four research papers analysing Indonesian firms and their activities in international trade. The first two papers investigate the 'learning-by-exporting' (LBE) hypothesis that suggests an increase in firms' performance once they enter foreign markets due to exposure to new knowledge and experience from abroad. The first paper explores the impact of export experience on firm productivity. Using firm-level data for 2000-12, I apply a fixed-effect technique by incorporating propensity score matching (PSM) in the first stage to control for the self-selection bias. I find that exporters' total factor productivity (TFP) increases with export age, but not linearly. Larger exporting firms and those engaged in particular industries undergo a clearer learning process. However, the LBE effect is only evident for firms that have had high productivity since the beginning, supporting the 'self-selection' hypothesis. The second paper discusses the policy consequence of LBE evidence. If LBE exists, should we endorse export promotion policies? I address this question by investigating the learning channels of firms in the garment industry. Firms in this industry experienced three decades of quota regulations under the Multi-Fibre Arrangement (MFA), which governed world trade patterns before its abolition in 2005. This investigation allows me to conduct a natural-experiment-type study on how export quotas affected the performance of apparel exporters. Applying PSM and difference-in-difference (DID) methods to firm-level data covering 25 years, I find that the impact of exporting on TFP during the MFA implementation period was mixed; but after its abolition, productivity increased by 9-13 percent. This implies that exporters gain a significant LBE benefit from competition (that is, without a special facility such as the MFA), and interventions that protect exporters from such competition might lessen the benefit. The third paper examines the patterns and determinants of apparel exports from Indonesia after the MFA's abolition. Contrary to predictions, Indonesia has not been able to achieve market share gains under competitive market conditions. The analysis of export patterns and Constant Market Share Analysis (CMSA) suggest that the lacklustre export performance was caused by supply constraints that hindered volume expansion to counterbalance the price-lowering effect of the quota abolition and the failure to diversify the product mix and the direction of exports in line with changing global demand patterns. The firm-level analysis suggests that productivity growth, the domestic textile base and access to imported inputs are key determinants of export performance in the post-MFA era. The fourth paper investigates the impact of increasing imported intermediate inputs on firms' TFP and firms' manufacturing exports. To tackle the simultaneity problem between imports and exports, I employ import tariffs and real exchange rates as instruments, using a weighting procedure that utilises each industry's use of imported inputs. Using firm-level data matched with detailed Customs data of exports and imports for 2008-12, the findings suggest that imported inputs raise productivity and export performance. Higher access to input varieties has a larger impact than an increase in import volumes on export performance. The effect is larger for imports from developed countries, suggestive of a positive effect from technology and product quality. Interestingly, the causal relation does not hold for firms in global production sharing (GPS) sectors, suggesting that firms in these industries manage their input and export decisions differently