ABSTRACT
This chapter explores the use of structural models as an alternative to reduced form econometrics when decomposing observed joint trade and technology driven wage changes into the components attributable to each source. Many papers have been written on the principal sources of increased wage dispersion in the form of an elevated premium paid to skilled labour in OECD countries in recent years, and most focus on increased trade and skill-biased technological change as the two principal causes. Lawrence and Slaughter (1993), Krugman and Lawrence (1993), Learner (1996), Baldwin and Cain (1997) and others conclude that the role of trade is small; Wood (1994) points to a dominant role for trade. Conclusions in this literature rest largely on reduced form regressions. Some, such as Murphy and Welch (1991), and Borjas, Freeman, and Katz (1991) estimate the factor contents of trade and use these estimates via exogenous (literature based) labour demand elasticities to infer the wage change attributable to trade. They then compare this to observed wage changes. Others, such as Learner (1996) and Baldwin and Cain (1997) use estimating equations derived from explicit general equilibrium models.
