ABSTRACT
Do New York and Nashville face the same pressures from increased trade? This paper considers the role of international trade in shaping the product mix and relative wages for regions within the US. Using the predictions from a Heckscher-Ohlin trade model, we ask whether all the regions in the US face the same relative factor prices. Using the production side of the HO model, we derive a general test of relative factor price equality that is robust to unobserved regional productivity differences, unobserved regional factor quality differences, and variations in production technology across industries. Using data from 1972-1992, we reject the the hypothesis that all regions face the same relative factor prices in favor of an alternative with at least three distinct factor price cones. Sort regions into cones with similar relative factor prices, we find that industry mix varies systematically across the groups. Regions that switch cones over time have more churning of industries.
1. Introduction
Do workers in New York and Nashville face the same benefits and costs from increased international trade? Does a firm’s survival depend upon where in the US it is located? If the US is a single, tightly integrated market, then the answer to these questions should be no: with workers and capital receiving the same returns everywhere, and with technology and inputs free to cross state and county borders, a shock to any part of the US affects the whole country. However, if relative wages vary across regions, then the possibility exists for differential dislocation in Nashville and New York due to trade and globalization.
Our research challenges the view of the US as a seamless, integrated economy. Using the implications of the Heckscher-Ohlin trade model as a framework, we derive a general test of relative factor price equality for US regions that is robust to unobserved regional productivity differences, unobserved regional factor quality differences, and variations in production technology across industries. Using regional plant level manufacturing data from 1972-1992, we reject the hypothesis that all regions in the US face the same relative factor prices. We also find that regional product mix varies with relative factor rewards. In the language of trade theory, we find that the US contains multiple cones of diversification.