Specialization and openness to foreign trade in the European Union

Abstract

This study explores the relations between inter-industry specialization and openness to foreign competition in the European countries. By controlling for other variables influencing specialization, the study found a negative relationship between openness and inter-industry specialization, meaning that the closer the countries become in their degree of openness to trade, the closer are their industrial structures, as is expected from the Hecksher–Olhin–Vanek framework with equalization of factor prices. Nevertheless, control variables powerful influence specialization, claiming for a more complete theoretical framework.

Notes

¹ However, the authors do not find the same result when looking at trade specialization indexes.

² In contrast with the evidence provided by several researchers (Globerman and Dean, 1990; Greenaway and Hine, 1991) of a slowdown in the growth of the intra-industry trade during the first half of the 1980s, the recent empirical analysis of Fontagné et al. (1997), Brülhart and Hine (1999) and Díaz Mora (2002), referred to the years 1980–1994, 1961–1992 and 1985–1996 respectively, show the reinforced importance of intra-industry trade flows between the economies belonging to the European Union.

³ Only a very general version of the HOV model seems to work reasonably (Davis and Weinstein, 2001).

⁴ Empirical analysis does not offer a strong support for factor price convergence among countries that have established free trade among themselves. For a survey on that subject, see Doroodian and Jung (1995).

⁵ The openness to foreign trade, measured as in this study, is frequently criticized because it is sensitive to the size and geographical situation of the countries. In fact, these variables affect the impact of openness in the production of a country, expected to be higher for the more open of them, more depending of foreign trade, following the results of Dar and AmirKhalkhali (2003). Nevertheless, measures are used in this study for every sector and they are made relative to the manufacturing industry as a whole. This means that the study is correcteed at least by the differences in size among countries, probably the more powerful of the mentioned variables influencing openness.

⁶ To find a negative correlation between specialization and openness in that case of reverse influence, it would be necessary that any increase in production and specialization led to a reduction in imports higher than the rise in exports, in order to reduce the openness ratio. The odds of this sequence are very low, as inter-industry specialization is often accompanied by intra-industry specialization, making difficult the reduction of imports.

⁷ See Rodríguez and Rodrik (1999) for a general view of this problem, and Balaguer and Cantavella-Jordá (2004) for the concrete case of one of the countries here considered.

⁸ Among more recent works, see for example, Dollar and Kray (2004) for a great sample of different kind of countries or Dar and AmirKhalkhali (2003) for OECD countries. A survey can be found in Shan and Sun (1998), Rodríguez and Rodrik (1999) and Temple (1999).

⁹ The construction of this database for 10 European countries (EU-10) was a difficult and laborious task, involving the actual compilation and homogenization of the data provided by Eurostat, as well as making up for the statistical gaps that existed in the original sources. The absence of available data, broken down into subsectors, that could be compared with the statistics one has for the countries selected (Germany, Belgium, Denmark, Spain, France, Netherlands, Italy, Luxembourg, Portugal and the United Kingdom), prevented the rest of the current members of the European Union being included in the analysis and, in addition, conditioned the choice of time period for study.

¹⁰ The exports and imports data were taken from Díaz Mora (2001).

¹¹ For details of the econometric method, see Arellano and Bover (1990, 1995).