Trade and Turnover: Theory and Evidence* Carl Davidson and Steven J. Matusz Abstract Is the pattern of trade correlated with cross-sector differences in job turnover? Theoretically, external shocks feed through to changes in domestic employment and cross-sector differences in turnover give rise to com- pensating wage differentials, which feed through to output prices. Using two different datasets on turnover, we find strong evidence that normalized US net exports by sector are negatively correlated with job destruc- tion and worker separation rates. Weaker evidence suggests a positive correlation between normalized net exports and job acquisition. Using sector-specific job destruction data for both Canada and the US, we find confirmation of the theoretical prediction that normalized net exports to Canada are negatively related to the ratio of the US job destruction rate to the Canadian job destruction rate. 1. Introduction A view that seems to be commonly held by a significant portion of the public is that international trade generates forces that threaten job security, particularly in the US manufacturing sector.The basic idea is simple and intuitive—a sudden surge of imports in a sector intensifies competition, drives American firms out of business and destroys American jobs. Of course, international trade could also have a positive effect on labor market outcomes—a sudden increase in the demand for American-made goods could lead to an increase in the number of jobs available as export sectors expand. In either case, changes in the international trading environment create or destroy jobs, leading in turn to worker turnover. 1 Recently, Davidson, Martin, and Matusz (DMM) (1999) have explored the implica- tions that labor market turnover might have for net exports in a general-equilibrium model of trade. In the model that they construct, jobs are continuously created and destroyed. In order to attract and retain workers, sectors with above-average rates of job destruction or where jobs are harder than average to find need to compensate searching workers for their risk by paying higher than average wages. Higher wages drive up production costs. By adding a second country, DMM are able to relate exoge- nously-specified cross-country differences in turnover rates to endogenously-deter- mined trade patterns. In particular, holding all else constant, an increase in the rate at which jobs breakup in sector i in country C raises the cost of producing that good and erodes the country’s comparative advantage in that good (or aggravates its compara- tive disadvantage). In this view of the world, exogenously-given labor market turnover is an independent determinant of comparative advantage. 2 Review of International Economics, 13(5), 861–880, 2005 © Blackwell Publishing Ltd 2005, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA *Davidson: Michigan State University, East Lansing, MI 48824, USA. Tel: (517) 355-7756; Fax: (517) 432-1068; E-mail: davidso4@msu.edu. Matusz: Michigan State University, East Lansing, MI 48824, USA.Tel: (517) 353-8719; Fax: (517) 432-1068; E-mail: matusz@msu.edu. For helpful discussions and comments, we thank Mary Amiti, Richard Disney, John Giles, Steven Haider, Daniel Hamermesh, James Harrigan, David Hummels, Christopher Magee, John McLaren, Douglas Nelson, Peter Schmidt, Richard Upward, Katharine Wakelin, Jeffrey Wooldridge, and seminar participants at Michigan State University, the University of Nottingham, the Upjohn Institute for Employment Research, the Spring 2001 meeting of the MWIEG conference, and the Fall 2001 meeting of the EIIT conference.