Trade, skill-biased technical change and wages in Mexican manufacturing

Abstract

This article looks at the relative importance of competing stories, particularly trade liberalization and skill-biased technical change, to explain changes in the skill premium and the real wages of unskilled and skilled workers in Mexican manufacturing using plant-level data. The channel through which technical change is observed is changes in the domestic price of machinery and equipment due to the availability of new and cheaper machines. The analysis also looks at trade-induced skill-biased technical change by taking into account changes in the price of machinery and equipment caused by changes in the tariff rate specific to machinery and equipment. Instrumental variables, including the price of machinery and equipment in the United States, are used to determine causality between the above effects and wages. Thus, the article provides evidence for some recent findings in the literature that link trade liberalization, skill-biased technical change occurring through technology embodied in machines and increases in the skill premium.

Keywords:

• trade liberalization
• skill-biased technical change
• skill premium
• real wages
• Mexico
• manufacturing

JEL Classification:

• F14
• J30
• L60
• O30

Acknowledgements

I thank Adrian Wood for many helpful comments and discussions. I am also grateful to Francis Teal, Beata Javorcik, Tony Venables, Alasdair Smith, Alan Woodland, seminar participants at the Asia-Pacific Trade Seminar, ETSG conference, EALE conference, University of Melbourne and Australian National University and three anonymous referees. Special thanks go to Abigail Durán and Gerardo Leyva for granting me access to INEGI data at the offices of INEGI in Aguascalientes and to all INEGI employees who provided assistance and answered my questions, in particular to Gabriel Romero, Otoniel Soto and Armando Arallanes. I gratefully acknowledge that part of this research was financially supported by an Australian Research Council grant to Professor Alan Woodland. All errors are mine.

Notes

¹ A more nuanced version of the SBTC hypothesis is the task-based approach, as proposed by Autor et al. (2006) and Goos et al. (2009). This approach suggests that technological progress is complementary with abstract cognitive tasks, a substitute for routine tasks, and has little direct impact on manual tasks. This pattern of complementarity and substitutability, in turn, is the reason for the current job polarization in the United States (US) and in Europe. The SBTC and job polarization hypotheses can be partly reconciled by assuming that skilled workers perform abstract tasks while unskilled workers perform routine and manual tasks.

² The skill premium remained stable in subsequent years (Riaño, 2009), which makes it difficult to identify the effects of reductions in tariffs, due to the introduction of the North American Free Trade Agreement in 1994, on wages.

³ A quarter of the plants surveyed in the EIA are part of multi-plant firms (Iacovone, 2008).

⁴ For more information on the EIA and the cleaning procedure see Iacovone (2008); and Riaño (2009).

⁵ The following formula is used to calculated the Herfindahl–Hirschman index: , where is the share of plant k in industry i and n is the number of plants in industry i.

⁶ Estimation of univariate AR(1) processes gives estimates of the autoregressive coefficients for the relative wage of white- to blue-collar workers and their real wages that are at least 0.8, thus all very persistent (see Blundell and Bond, 1998; Bobba and Coviello, 2007).

⁷ It should be noted that only one quarter of the plants acquire technology by expenditures on royalties and, even among these plants, such expenditures are very small relative to sales.

⁸ Data on imports of M&E excludes purchases of imports via other domestic firms (e.g., specialized importers), which may underestimate their importance.