4. SummaryReferencesTable 1: Order of Integration of the VariablesGlobalization, Roundaboutness, and Relative Wages † Joseph Francois Tinbergen Institute & CEPR Kevin Grier University of Oklahoma Douglas Nelson Tulane University & Leverhulme Centre (Nottingham) January 2005 Abstract: We depart from the trade and wages literature and its emphasis on North-South trade. Examining North-North trade and linkages between trade-based integration and relative wages in an Ethier-type division of labor model, we use this to we identify formal relationships between trade, productivity, and wages. We examine the trivariate relationship between trade, growth in total factor productivity (TFP), and the skill premium in a VAR framework, finding evidence of long-run relationships between growth in intermediate goods and changes in TFP. Controlling for this relationship, we also find a positive relationship between trade and the skill-premium. Our results suggest a significant alternative trade-based mechanism, apart from the North-South trade and skill-biased technical change mechanisms highlighted in the current literature on the evolution of relative OECD wages. JEL Codes: F12, F16 Keywords: Ethier model, trade and wages, globalization, division of labor, monopolistic competition, intra-industry trade † We thank Robert Baldwin and Glen Cain for their help in providing the wage series used in this paper and Tim Lloyd for valuable advice on the econometric analysis. Address correspondence to: Prof. J. Francois, Faculty of Economics, Burg Oudlaan 50-H8-18, 3000DR Rotterdam, Netherlands. Email: [email protected] fax: +31 10 408 9146.Globalization, Roundaboutness, and Relative Wages Abstract: We depart from the trade and wages literature and its emphasis on North-South trade. Examining North-North trade and linkages between trade-based integration and relative wages in an Ethier-type division of labor model, we use this to we identify formal relationships between trade, productivity, and wages. We examine the trivariate relationship between trade, growth in total factor productivity (TFP), and the skill premium in a VAR framework, finding evidence of long-run relationships between growth in intermediate goods and changes in TFP. Controlling for this relationship, we also find a positive relationship between trade and the skill-premium. Our results suggest a significant alternative trade-based mechanism, apart from the North-South trade and skill-biased technical change mechanisms highlighted in the current literature on the evolution of relative OECD wages. JEL Codes: F12, F16 Keywords: Ethier model, trade and wages, globalization, division of labor, monopolistic competition, intra-industry trade 1. INTRODUCTION The recent surge in research on the link between trade and wages has abated, and the collective prior among economists on the empirical magnitude of that link seems to have stabilized around the existence of some statistically significant but practically small effect of trade on the decline in the relative wage of unskilled workers. However, important questions remain. The great majority of both theoretical and empirical work addresses the link between trade and wages within the context of the Heckscher-Ohlin-Samuelson (HOS) model, generalized in a variety of straightforward ways. This tends to focus attention on North-South trade, where the Stolper-Samuelson effects would be most pronounced. That is, given substantial liberalization in North-South trade and falling frictional costs of trade, the substantial differences in relative commodity and factor prices between North and South might cause us to expect large effects. However, as a number of students of this topic have pointed out, the volume of North-South trade is quite small, with the, somewhat controversial, implication that the leverage for 1Stolper-Samuelson effects is also small. Where we do observe substantial and growing trade volumes and deepening economic integration is between North (aka OECD) economies. From its beginning, the literature seeking to explain the increase in the skill-premium (in an environment with an increasing in relative endowment of skilled workers) has considered skill-biased technical change as an alternative to trade shocks. Much of the testing in the Heckscher-Ohlin framework has treated technical change as the alternative hypothesis.1 In this paper, while we retain the emphasis on comparative static analysis, we seek to depart from this literature by focusing on intra-industry trade (specifically, trade in industrial intermediates) in a division of labor model of the sort pioneered by Ethier (1982) and Markusen (1990).2 As with current empirical research on trade patterns, which has examined the implications of imperfect competition models, we provide an empirical analysis of a relatively sparse monopolistic competition model of the link between trade and labour markets.3 Specifically, we examine, theoretically and empirically, North-North trade within a monopolistic competition model, looking for links between globalization and labour market performance. Given the ease with which the HOS model yields an estimating framework, it is not surprising that the empirical literature on the general equilibrium effects of trade and wages and of skill-biased technical change has 1 Leamer (1998) is particularly clear on the problems with the way such testing has gone. More recently, Haskel and Slaughter (2001, 2002) have developed this methodology particularly clearly. 2 Recent work has considered issues of the link between trade and skill bias in models directly focused on technology choice in a variety of environments (Neary, 2000; Thoenig and Verdier, 2003; Ekholm and Midelfart, forth.). We view our work as strongly complementary to the research in these papers. That is, where these papers seek the foundations of technical change in technology choice, we seek the foundations in an expanding division of labor. 3 There is now considerable literature suggesting: 1) that the HOV model does a less than spectacular job in describing directions of trade; and that 2) alternatives including increasing returns and monopolistic competition offer a useful addition to factor-endowments in accounting for directions of trade. Daniel Trefler (1995; Antweiler and Trefler, 2002) has probably done more than anyone to advance this position, but the literature is now