Total factor productivity in Indian manufacturing: Evidence from a firm-level study

This is a study of the productivity of Indian manufacturing firms. Firm-level total factor productivity (TFP) estimates are computed and compared across various cohort groups to examine the contribution of firm dynamics to industry productivity. It is observed that survivors demonstrate higher productivity than exiters and become more productive with age. Entrants in later years are more productive than exiters in earlier years. Aggregate industry level productivity estimates are then computed from firm-level estimates. These are decomposed into effects of within-plant productivity improvement, reallocation of factors and market share towards more productive plants and entry and exit of firms from the market. Indian manufacturing industries satisfy only one tenet of the creative destruction hypothesis: while most industries demonstrate within plant productivity improvements, they do not experience positive turnover effects. An examination of the impact of trade liberalization on productivity levels reveals that it does not make a significant contribution to the TFP levels of exporting firms.; The relationship between a firm's TFP and its ability to participate in the export market is also examined to ascertain whether Indian manufacturing firms demonstrate learning-by-exporting. Firm-level estimates of TFP are used to compare productivity premia generated from exporting. It is found that exporting firms have higher productivity levels than non-exporters. A series of empirical investigations are carried out to examine whether the contributions of exporting firms to TFP can be attributed to either of two reasons: positive externalities accruing from learning-by-exporting or the inherent self-selection of firms with higher productivity levels into export market. Comparison of productivity levels of different transition groups in the export market does not reveal any self-selection, but it is found that the textiles, minerals and machinery industries demonstrate evidence of learning-by-exporting. Finally, the semi-parametric estimation methodology is used to control for the biases of selection and simultaneity while testing the learning-by-exporting hypothesis. Export share is found to be a positive contributor to productivity growth when all industries are pooled together. However the evidence is mixed when examined separately for every individual industry. It is observed that labor intensive industries demonstrate learning effects while capital intensive industries do not.