Skill and support to globalization in the EU

Abstract

Are individuals’ attitudes towards globalization consistent with the predictions of the Heckscher–Ohlin (HO) theory? Using newly available data from the Flash Eurobarometer 151b of the European Commission, we find that globalization preferences are significantly and robustly correlated with an individual's relative level of human capital, consistent with the HO theory. The impact of skill on globalization preferences is positive in skill-abundant EU countries and strongly negative in EU countries that are well-endowed with low-skill labour. This study is the first to find a robust and significant negative association between skill and support to globalization in low-skill-abundant countries.

Notes

¹ Trade's effect on individual economic welfare depends on the degree of intersectoral factor mobility. If factors are mobile across sectors, as in the HO model, support for free trade depends on factor type. If some or all factors cannot move to other sectors, as in the Ricardo–Viner (RV) model, trade policy preferences are determined by industry of employment (see Scheve and Slaughter, 2001, for a detailed review of the HO and RV trade theory). The Flash Eurobarometer 151b does not provide information on the sector of employment of individuals. Therefore, we only test the HO theory. However, Tombazos (1999) shows that imports have a positive impact on the demand for skilled labour in skill abundant countries, in addition to the conventional Stolper–Samuelson reduction on the demand for unskilled labour. Pyne (2000) also shows that Canadian factor's owner positions on the Canada–United States Free Trade Agreement in 1987 were consistent with the Stolper–Samuelson theorem implications. Finally, Scheve and Slaughter (2001) find evidence that factor type dominates industry of employment in explaining support for trade barriers in the United States. In fact, these authors argue that the RV model can be characterized as a short-run version of the more long-run HO model. All these studies suggest that factor type, being skilled or unskilled labour, is more relevant in determining support for trade liberalization than sector of employment.

² This report provides the distribution of workers with low, medium and high educational attainment across EU regions and countries. We construct the new variable of skill for each region and country following EURYDICE (1997, 2004); assuming 16 years of education for low educational attainment, 18 for medium and 21 for high. We tested other possible constructions and found that the results are fairly similar.