This paper connects trade flows to deviations from the law of one price (LOOP) in a structural model of trade and retailing. It accounts for the observed cross-country dispersion in prices of goods, based on retail price survey data, by focusing on two sources of goods market segmentation - (i) international trade costs, and (ii) non-traded input costs of distribution. I find that a multi-sector Ricardian trade model, ala Eaton-Kortum, augmented with a distribution sector, can account for the average price dispersion for a basket of goods fully and generates 70 percent of the variation in price dispersion across goods within the basket. While tradability of goods is important in explaining the average price dispersion for the basket of goods, distribution costs are important in explaining why, within the basket, some goods show more price dispersion than others.

We develop a quantitative, two-country, three-sector model to evaluate the impact of trade and trade liberalization for structural transformation. We provide an analytical characterization of the employment and GDP share of each sector in terms of domestic and international factors. We calibrate the model to data from South Korea and the OECD, and ask whether the contraction in observed agriculture employment and GDP shares, and growth in employment and GDP shares of the industrial sector, in South Korea between 1962 and 2000 would have occurred in the absence of trade. Our preliminary results show that while a closed economy model with non-homothetic preferences and differential productivity growth across sectors can account for much of the contraction in the employment and GDP share of agriculture in South Korea, it cannot account for the growth in the employment and GDP shares of the industrial sector. Our two-country model, calibrated to match import ratios from the post trade liberalization era in South Korea, can account for all of the growth of the GDP share of industry in South Korea, and a substantial portion of the growth of the employment share of industry. The model also out-performs a variant based on calibration to pre-liberalization import ratio data.

Most "macro" analyses of the gains from trade employ frameworks in which the entire economy is aggregated into one sector. In this paper, we investigate the role of heterogeneity across sectors, focusing on the key elasticity used to evaluate the gains from trade. We employ a multi-sector Ricardian trade model and derive implications for the country-by-sector costs of international trade, conditional on the sectoral trade elasticity. We estimate the sectoral elasticities by employing the simulated method of moments methodology developed by Simonovska and Waugh (2010). Our data include OECD cross-country micro data on prices for 1410 goods - each of which is mapped into one of 21 manufacturing sectors - as well as bilateral trade flows for each of these sectors. Our estimates reveal significant heterogeneity in trade elasticities across sectors - the range is 2.7 to 9.8, with an average elasticity of 4.6. We use the estimated elasticities and estimates of the trade costs to evaluate the gains from trade and assess the importance of sectoral heterogeneity. Heterogeneity matters in at least two ways. First, the gains to trade obtained using the sectoral elasticities and trade costs are different from the gains obtained when the average elasticity is used. Second, when all manufacturing sectors are aggregated into one overall sector, the elasticity estimate is 2.6. Employing this elasticity instead of the vector of sector elasticities leads to an over-estimate of the gains from trade.

There has been a dramatic increase in the wage of college graduates (skilled labor) relative to the wage of high school graduates (unskilled labor) in the U.S. after 1980, especially during the period 1980-1990. At the same time, the quantity of skilled labor relative to that of unskilled labor has also increased. The data also reveals that in the post-war U.S. economy there has been a shift of resources (labor and capital) from the industrial sector to the service sector. Using individual level data from Current Population Survey (CPS) and disaggregated industry level data from Bureau of Economic Analysis (BEA), this paper investigates the link between the rise of service sector and the rise in wage inequality. The objective of the paper is to develop a three factor, two sector general equilibrium model to generate the structural change observed in the U.S. economy along with the rising wage inequality.