Empirical studies of international trade and multinational firms in Argentina, Brazil and China

This dissertation consists of three empirical studies, grouped in three chapters. In Chapter l, I develop and estimate a model of demand, supply and trade in cars for Argentina and Brazil. Trade is subject to two non-tariff barriers: a quota on bilateral net imports and a trade balance constraint, which are modeled as additional costs imposed to firms. I estimate the shadow value attached to them together with the cost of production of each car model. By modeling the behavior of firms in two countries, rather than in one, I am able to capture the strategic interaction across markets. In addition, it allows me to estimate the supply parameters by minimum distance instead of by a perfect fit method, and without the need to make functional form assumptions on the cost side (as is the usual practice in the literature).; In Chapter 2, I apply the methodology and results from Chapter 1 to the evaluation of a change in trade policy. I simulate a change in policy that will occur in 2006 when trade in cars is fully included in MERCOSUR. The change in policy implies the adoption of a common external tariff and the removal of the non-tariff barriers studied in Chapter 1. I estimate the separate effects of these changes in policy on prices, trade flows and welfare.; In Chapter 3, I study the introduction of new varieties of goods in the manufacturing sector in China. I compare the performance of foreign and domestic firms using firm-level data. Once observed firm characteristics are accounted for, I find that firms with more than 50 percent of foreign ownership introduce on average twice as many more new goods as private domestic firms; while fully foreign-owned firms introduce three times more new goods. Foreign firms are subsidiaries of multinationals and have more experience in the introduction and production of these new varieties. I measure these two advantages with the cost of innovation (R&D and purchases of technology) and total factor productivity. These measures of efficiency explain between 13 and 33 percent of the difference in the number of new goods.