In the first objective, we explored how productivity of countries are shaped by the related industries. According to the Ricardian theory, countries should produce and export relatively more in industries in which they are relatively more productive. We posit that the distribution and transformation of comparative advantage is not random, but rather a path-dependent process, i.e. what a country effectively produces today is an important determinant of which industries it might be productive at tomorrow. We test whether we can predict the productivity levels in an industry from the productivity in related industries by building different measures of relatedness between industries based on Input Output, Labor and Knowledge tables. We then explore whether this structure has a predictive power in determining which industries would grow after North American Free Trade Agreement. In all our tests, we confirm the importance of the structure in comparative advantage patterns of the countries.
In the second objective, we analyze the interaction between the country-level and product level quality indices and the diversification process. To do that, we first created a novel world trade dataset from the raw trade data. We take advantage of the fact that each trade transaction is, in theory, reported twice by the exporter and importing country which helps us estimate reliability of reporting. We implement this procedure to create databases of trade covering more than 50 years, and show how estimates of trade change for countries. Using this new dataset, we then analyzed whether the country level quality parameters have an effect on the export growth at the disaggregated product level. We showed that having higher years of schooling both sped up the overall growth and also helped countries capitalize on the nearby products. But we could not find consistent results when we used the Rule of Law, Democracy level and Control of Corruption. On the product quality side, having more products nearby a product resulted in an increase in the product quality.
In the third objective, using Diebold-Yılmaz methodology we show that the movements in the stock market indicate large amount of spillovers between industries. These spillovers are explained by the Marshallian externalities like input-output relationships, knowledge sharing and labor-pooling, especially during the epochs with no-crisis. Crisis change the fundamental interdependence between industries. This approach bridges the approaches used in the finance literature to industrial growth. We also analyzed the temporal changes in the exports of countries and defined the ecosystem of products, which consists of products facilitating to jump to a new product. We show that the ecosystem is an important predictor of diversification. Network properties of ecosystem reveal important determinants of diversification.
In the fourth objective, we incorporate the input-output linkages to the Ricardian models of trade. We recently witnessed the extreme fragmentation of the supply-chain and countries often push to move up the chain. We explicitly incorporate IO relationships into the trade models and observe that the inclusion of IO linkages forces non-trivial location choices of industries.
All these results will be published as academic papers in high-impact journals and presented in important conferences. There have been significant efforts to reach out to the public and policymakers.
