Intellectual Merit The applied general equilibrium models used to predict the impact of the North American Free Trade Agreement on trade flows failed to capture the explosion of trade that occurred subsequently and the accompanying redistribution of trade across industrial sectors. The analysis of disaggregated trade flows between Canada, Mexico, and the United States shows that much of the increase in trade has been trade in goods not previously exported or exported very little - what is known as the extensive margin of trade. This proposal shows that mechanically incorporating this insight into crude projections of North American trade patterns would have produced significantly better predictions of the pattern of trade following the enactment of NAFTA than did multisectoral applied GE models. The data suggest that capturing the extensive margin of trade is important for understanding the impacts of trade policy, which implies that models that generate extensive margin growth are needed. The proposed research will use advances in the firm-level theory of trade to build models that are suitable for policy analysis. The initial aim of the project is to evaluate the properties of models based on two types of industrial structures: monopolistic competition with entry costs, as in Melitz (2003) and perfect competition, as in Eaton and Kortum (2002). Constructing models flexible enough to be estimated for policy analysis requires generalizing the standard theories. Based on Kehoe and Ruhl (2009b), the models are extended to include multiple dimensions of firm heterogeneity and generalized distributions over underlying productivity. A mapping scheme is developed to match the firm-level decisions in the model to the product-level and industry-level aggregates in the data. Broader Impact To inform the policy debate over free trade and protectionism, economists need to better understand the impact of trade on different industries. The proposed research develops models that transform the firm-level trade models popular in current research into models useful for policy analysis. Methodologically, these sorts of models will allow economists to analyze the impact of policy on the structures of industries. Currently, the United States is negotiating free trade agreements with several Latin American countries, and a free trade agreement with South Korea waits to be approved by the legislatures in each country. Successful trade negotiations require careful evaluation of the impact of trade policy on individual industrial sectors. Without an accurate understanding of the distributional impacts of trade policy, it is difficult to advise policy makers and to inform the public on the merits of such liberalization policies
When trade between two countries grows, does it grow because the countries trade more of the goods they were already trading, because they trade new kinds of goods they were not trading before, or some mixture of the two? In our analysis, we document the importance of "newly-traded" goods in accounting for the growth in trade. We find, for example, that 10 percent of trade growth in post-NAFTA North America is accounted for by countries trading goods which they had previously not traded, or traded very little. In another example, these newly-traded goods account for 30 percent of the growth in U.S.-Korea trade. These results have important implications for the evaluation of trade policy, and other policies that impact the level of trade between two countries. The industries that gain from trade creating policies will not always be industries that were important exporting industries before the change in policy. More research along these lines is needed to develop theories that will yield better predictions of the sectoral impact of trade policy. As part of the funded research, we have begun developing a method to incorporate our findings into a predictive model. In our research, we demonstrate that our preliminary method would have produced better predictions of North American trade patterns following the enactment of NAFTA than did multisectoral applied GE models. We have also applied this method to an analysis of the U.S.-Korea Free Trade Agreement. Our empirical research shows that different kinds of goods respond differently to changes in policy. In theoretical models, we associate a good with the firm that produces it. Differences across firms — firm heterogeneity — then lead to differences in the response to policy that we observe in the data. An important way in which firms differ is in their productivity: Some firms are better at producing than others. As part of this project, we have developed a model of heterogeneous firms who must decide whether to enter a market, how much to produce, and when to exit a market. In the model, a country grows (becomes richer) when less productive firms exit markets and are replaced by more productive firms. We show that when the costs of starting a new business are high or when it is difficult to acquire a loan, this replacement of bad firms with good firms slows down, slowing growth in the country. An implication of this research is that the two barriers to entry — explicit costs of entry and a lack of borrowing opportunities — reinforce each other: If starting a firm is costly, being able to borrow is even more important. Understanding the ways heterogeneous firms are affected by different kinds of policy is important if we are to develop ways to reverse barriers to entry, and improve the growth rate of countries. In a less technical report, co-PI Ruhl and I discuss these ideas while comparing a fast-growing country with a country whose growth has stagnated: China and Mexico. The recent financial crises in the Eurozone and the downgrading of U.S. government debt following political battles over raising the debt ceiling have raised fears of a financial crisis in the United States. Working with Joseph Steinberg, co-PI Ruhl and I have adapted a model that we had previously used to study the Mexican crisis to analyze two possible ends to foreign borrowing by the United States: an orderly, gradual rebalancing, and a disorderly, sudden stop in foreign lending as occurred in Mexico. We find that a sudden stop would be very disruptive for the U.S. economy in the short term, particularly for the construction industry. In the long term, however, a sudden stop would have a surprisingly small impact. As the U.S. trade deficit becomes a surplus, gradually or suddenly, employment in goods production will not return to its level in the early 1990s because much of this surplus will be trade in services and because much of the decline in employment in goods production has been, and will be, due to faster productivity growth in goods than in services.
