The following essays examine the nature of product co-production patterns in India—what factors cause these patterns to emerge, and why they are valuable to study. The first chapter establishes a motivation. It takes a measure of product co-production established in the literature—the “proximity” matrix of Hidalgo et al. (2007)—and shows that this measure is an excellent predictor of new product additions by firms and states, even controlling for other potentially relevant explanatory variables. The following chapter employs a reduced-form approach with regression analysis to uncover the factors that could be giving rise to these patterns of co-production. Using this approach, demand complementarities and patterns of input similarity seem to have the most explanatory power for the observed patterns. The final chapter improves the estimation by incorporating product paths and firm profitability into a structural model. We adapt the gravity model of Morales et al. (2015) to our setting to identify costs associated with adding new products based on characteristics of the relationship between the firm and its potential products. In the model, firms seek to expand their product scope into the most profitable products, where this profit is diminished by “distance” they would have to traverse through a characteristic space. Using the moment inequalities method of Pakes et al. (forthcoming), we are able to estimate which dimensions in that space have the greatest effect on firm profits. We find the physical distance to the nearest location of production had the greatest impact, followed by input similarity between their products and potential products.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8TT4QM1 |
| Date | January 2015 |
| Creators | Flagge, Matthew John |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0019 seconds