In this section I cover in chronological order the following product-price studies: Bhagwati (1991), Lawrence and Slaughter (1993), Sachs and Shatz (1994), Feenstra and Hanson (1995), Leamer (1996), Baldwin and Cain (1997), Krueger (1997), Feenstra and Hanson (1997), and Harrigan and Balaban (1997). Together, these papers demonstrate how the methodology of product-price studies has evolved. Where appropriate, other related papers will be mentioned. By surveying these nine papers chronologically I can relate each to the relevant work preceding it. For each paper I refer the interested reader to exact pages with the key results.
Bhagwati (1991) is the first researcher I am aware of to link rising U.S. wage inequality to international trade working through product prices. He plots (Figure 7, p. 51) quarterly observations from 1982 through 1989 of the U.S. price indexes for exports and imports of all manufactures aggregated together. Observing that import prices rose more quickly than the export prices after 1986, Bhagwati concludes that “the trade-focused explanation [of rising wage inequality thanks to declining relative prices of imported products which presumably employ less-skilled labor relatively intensively] … therefore carries little plausibility, at least at first blush.” Link
Lawrence and Slaughter (LS) (1993) are the first researchers to use price data disaggregated by industry and also direct measures of industry factor usage. They use three different sets of U.S. manufacturing prices: imports, exports, and domestic production. The export and import prices cover all industries at the two-digit and three-digit SIC level for which the Bureau of Labor Statistics assembles the data. These data do not cover all manufacturing industries: for example, the import prices cover 18 of the 20 two-digit SIC industries but only about 50 of the 143 threedigit SIC industries. LS assume (fn. 55, p. 195) “that the price movements in these industries are reasonably representative” of all industries. The domestic prices cover all three-digit SIC industries. LS refer to the traded prices as “international prices” whose “changes … [are] prompted by international trade” (pp. 198-199). Thus, they model the United States as a small price-taking economy facing product prices determined exogenously abroad. For domestic prices LS assume (fn. 63, p. 202) that “changes in these domestic price deflators tracked changes in international prices. This is a weaker assumption than the law of one price: it allows prices to differ across countries by some fixed constant,” presumably due to U.S. trade barriers. Because LS assume that all product-price changes come from foreign developments, when analyzing industry differences in total-factor-productivity (tfp) growth, they assume that U.S. technology changes do not influence product prices (p. 199).