Manoj Atolia Indiana University May 1, 2002
Abstract We work with a 4-sector 5-factor model that captures relevant structural characteristics of the developing economies: the unskilled wage is indexed in the manufacturing sectors, the capital in each sector is produced withimported machines and non-traded goods (e.g. construction), and the imported intermediates are used in production. In the long-run, for plausible parameter values, a tari® reduction increases the skilled wage relative to both the competitive and the indexed unskilled wage even when export-manufacturing sector is up to 35% more intensive in the use of the unskilled labor in terms of the cost shares.We derive a very simple analytical condition relating changes in wage inequality to parameters of the model. A long-run increase in wage inequality is accompanied by a steady rise in wage inequality and real skilled wage during transition if the elasticity of substitution of skilled and unskilled labor for other factors are equal. When it is easy to substitute the skilled labor for other factorcompared to the unskilled labor, a short-run rise in wage inequality vis-a-vis the competitive unskilled wage occurs for up to 40 years in cases where wage inequality falls in the long-run. However, the real skilled wage does not rise during transition as empirically observed in the recent episodes of trade reforms. Thus, the paper allows for a reconciliation of the rise wage inequality that hasaccompanied trade liberalizations in past two decades with the factor proportions theory in a dynamic general equilibrium model.
Keywords: Wage Inequality, TradeReform, HOS, Dynamic Analysis. JEL Classi¯cation: F11, J31, F13, F17.
¤ I am grateful to Shouyong Shi, Hugh Kelly and specially to Edward Bu±e for thoughtful comments. All errors remaining are mine. Comments are welcome and may be sentto firstname.lastname@example.org.
It is important to understand the determinants of income distribution as it has implications for the well being of a vast number of people. Hence, the observed rise in the wage inequality that has accompanied the tari® reforms in the past two decades, contrary to the intuition of 2x2 HOS models of international trade is an important unresolvedempirical fact. This paper provides two explanations that can reconcile this rising wage inequality with the factor proportions theory in a dynamic general equilibrium model. One of the earliest positive theories explaining wage inequality is the pure theory of international trade. The cornerstone of the theory is the HeckscherOhlin-Samuelson model and its main explanatory result is theStolper-Samuelson theorem. Brie°y stated the theorem says that in the long run increase in price of output of a sector increases the cost/ price of the factor that this sector uses intensively. Krueger , analyzing the impact of alternative trade strategies on the employment in the LDCs, hypothesized that employment will rise after a trade reform since the relative factor intensities of labor in the valueadded in the export sector is higher than that in the import-competing sector. Secondly, as LDCs are mainly exporters of primary commodities or light manufactures which are more intensive in the use unskilled labor, demand of the unskilled labor is expected to increase relative to that of the skilled labor when the tari®s that protect import-manufacturing industry are reduced. Hence, we canconclude that the wage inequality should decline. However, the overall experience of the last two decades mentioned above is at odds with this basic intuition arrived from 2x2 model, at least in the short run. In an summary of an extensive study covering nine developing countries with high trade exposure from Latin America, East and South-East Asia Robbins  ¯nds evidence of increase in wage...