The studies of Bhagwati, LS, SS, and Krueger assume (explicitly or implicitly) that U.S. product prices are determined in the rest of the world without any influence from U.S. variables. This assumption is a reasonable first-pass at the data but it is almost certainly not correct. The other studies in this survey move away from this assumption in various ways. Leamer decomposes observed product-price changes into two components: technological progress, and “globalization” price changes. However, he does not attribute these “globalization” price changes to anything more specific such as trade barriers or foreign developments communicated through the U.S. terms of trade. BC consider three different forces acting on product prices: international trade, technology, and endowments. Like Leamer, though, they do not specify anything more specific regarding what trade’s force consists of. FH (1997) use regression analysis to explain observed tfp changes with outsourcing and computerization. In estimating the mandated wage changes driven by each of these tfp drivers they allow an endogenous response of product prices. Finally, instead of using direct measures of domestic product prices in tradable industries HB instrument for these prices using data on foreign labor endowments and trade quantities.
These latter studies which attempt to decompose product-price changes among various causes are an improvement on the earlier first-pass studies. Nevertheless, it seems that much more work is needed here. As discussed in Section 2, in theory the question of “how does international trade cause product-price changes?” can be refined with reference to at least four different trade-related forces that are plausibly exogenous to domestic firms at a point in time: U.S. political trade barriers; foreign trade barriers; worldwide natural trade barriers; and developments abroad in parameters such as tastes, technology, and endowments that are communicated to the United States via the international prices constituting our terms of trade. None of the nine studies in this survey uses data on political or natural trade barriers. And only HB uses data on foreign parameters to help explain U.S. domestic product prices.