How might this price problem be addressed? Alterman (1991) points out that import and export unit values that can be constructed off of Census Bureau trade-flow data do not adjust for quality changes. In principle, one might be able to compare BLS traded-price indexes with traded-goods unit values to look for evidence of changing product quality.
For example, if domestic producers are moving towards higher-quality products then export unit-values should rise faster than export price indexes. Indeed, Alterman does a direct comparison like this and reports the “puzzling” result that during the 1970s and 1980s BLS import price indexes tend to rise faster than the Census unit values. Alterman discusses that one might expect quality-adjusted price indexes to rise more slowly if import quality is rising over time. But thinking about different countries moving into different cones suggests an alternative view.
Perhaps the average quality of U.S. imports is declining, not rising as is usually presumed, as foreign producers increasingly concentrate on unskilled-labor-intensive products. In this case quality-adjusted price indexes should rise faster than unit values, as the data actually show. Another possible solution to changing product mixes and price indexes is suggested by Feenstra (1994) who develops a methodology for adjusting import-price data to account for new product varieties.
This paper has attempted to provide a comprehensive survey and synthesis of research on how product-price changes have contributed to rising U.S. wage inequality. It has surveyed nine product-price studies which together demonstrate how the methodology of product-price studies has evolved. After surveying each paper individually I synthesized the findings and drew two main conclusions. The first conclusion is that this literature has refined a set empirical strategies for applying the SS theorem to the data from which important methodological lessons can be learned. The second main conclusion is that despite the methodological progress that has been made, research to date still has fundamental limitations regarding the key question of how much international trade has contributed to rising wage inequality.