EU enlargements and U.S. FDI flows in the six core states

ABSTRACT

This paper addresses an under-researched question of whether regional integration with lower-income countries results in displacement of foreign direct investment (FDI) from high-income countries. Difference-in-differences (DID) estimations are used to, respectively, analyse the effects of the 1986 enlargement and the 2004/2007 enlargement of European Union (EU) on U.S. FDI flows into the EU core states. The results reveal a negative impact of the 1986 enlargement on EP FDI and positive effects of the 2004/2007 enlargement on EP and vertical FDI. The comparative analysis demonstrates that the impact of regional integration depends on the relative strength of a few factors determined by the circumstance. High-income countries do not necessarily lose FDI following an integration with lower-income countries. While the integration tends to divert labour-cost sensitive EP FDI away from the high-income members, as long as the high-income members adaptively go through industrial restructuring and take full advantage of the expanded production networks in response to the newly integrated regional environment, they are likely to become more resilient and competitive in the regional and global markets and reap the benefits of integration.

KEYWORDS:

• Regional integration
• EU enlargement
• EU core state
• foreign direct investment (FDI)

JEL CLASSIFICATION:

• F15
• F21
• F23

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Studies on trade agreements between countries of similar development levels are not discussed here, as the paper concerns integration of dissimilar countries.

² In 1995, three of them (Austria, Finland, and Sweden) joined EU.

³ The lecture can be found at https://www.mailman.columbia.edu/research/population-health-methods/difference-difference-estimation

⁴ Blonigen and Piger (2014) point out that there is very little consistency in the empirical FDI literature about the covariates one should use when empirically modelling cross-country FDI activities. They use Bayesian statistical techniques to examine the robustness of a comprehensive set of FDI determinants proposed by previous studies on FDI, and suggest using a fairly parsimonious FDI specification comprised of traditional gravity variables, cultural distance factors, relative labour endowments, trade agreements, host-country remoteness, and parent-country real GDP per capita. Variables with little support for inclusion are multilateral trade openness, most host-country business costs, host country infrastructure (including credit markets), and host-country institutions.

⁵ The proxy is a human capital index, based on years of schooling and returns to education.

⁶ BEA does not have data for Iceland and thus Iceland is excluded from the estimation. Finland is excluded, as Finland was not a full member of EFTA until 1986.

⁷ As comparison, the EP share of all FDI is 17% in Italy, 26% in France, 30% in Germany, 53% in Belgium or Netherlands in 1986.

⁸ Of course, in the long run, things can change. As income levels of the lower income members converge to the high-income members and as consumption preferences change due to the available advanced products at more affordable prices, demand structures in the lower income members can change. While this study indicates that high-income members can reap even more benefits from integration in the long run, the long-run analysis is beyond the scope of this study.