Foreign Direct Investment Spillovers in Portugal: Additional Lessons from a Country Study

Abstract

This paper investigates the impact of foreign direct investment (FDI) on the productivity of Portuguese manufacturing sectors. It improves on previous studies by considering the choice of the most appropriate interval of the technological gap for spillovers diffusion. Sectoral variation in the coefficients of the spillover effect is also allowed for. Idiosyncratic factors are identified by means of a fixed effects panel model and positive inter-sectoral spillovers are also examined. Significant spillovers require a proper technological differential between foreign and domestic producers and favourable sectoral characteristics. They may occur in modern industries in which the foreign firms have a clear, but not too sharp, edge on the domestic ones. Agglomeration effects are also a pertinent specific influence.

Ce texte examine les répercussions des investissements directs étrangers (IDE) sur la productivité des secteurs manufacturiers au Portugal. La spécification du modèle est améliorée par le choix de l'intervalle le plus approprié de l'écart technologique pour la diffusion des spillovers. Nous tenons également compte des variations sectorielles dans les coefficients de spillovers; des facteurs sectoriels idiosyncrasiques sont identifiés au moyen d'un modèle à effets fixes. Les spillovers intersectoriels positifs sont examinés. Des spillovers importants requièrent un véritable écart technologique entre producteurs étrangers et domestiques ainsi que des caractéristiques sectorielles favorables. Ils peuvent se produire dans les industries modernes où les entreprises étrangères jouissent d'une supériorité nette, mais pas trop marquée. Les effets d'agglomération exercent aussi une influence spécifique pertinente.

Acknowledgements

The authors are indebted to participants in the 3rd International Workshop on European Economy, ISEG/Lisbon and the Irish Economic Association Fourteenth Annual Conference (Waterford). Comments and suggestions from Frank Barry, German Calfat, Rod Falvey, Isabel Proença and Frances Ruane, as well as from two anonymous referees and the Editor-in-Chief, are heartfully acknowledged.

Notes

1. For a clear and well-documented survey of the preparations for joining the EU and the post-1986 period see chapters 5 and 6 in Mateus (1998).

2. We are indebted to an anonymous referee for making this sharper point.

3. The Inquérito às Empresas Harmonizado is the responsibility of the Serviço de Estatísticas Estruturais das Empresas, Departamento de Estatísticas das Empresas/INE.

4. As DP is a ratio between (the dependent) PROD and the productivity of the related foreign firms, concerns about lack of independence with the error might be raised. However, not only PROD enters as a numerator in DP, as alternative estimation methods yielded qualitatively similar results.

5. Model (4) does not include DP itself as an independent variable. Though a possible specification, we favoured those without it to characterise that it is indeed a ‘new variable’ – the gap within a restricted range – which plays a role in the phenomenon we try to explain.

6. It has been recently emphasised (Cantwell and Molero, 2003; Narula, 2004; Álvarez and Molero, 2005; Molero and Garcia, 2006) that sectoral specificities appear to be a central factor for the ‘virtuous circle’ of innovation.

7. This is also an asymptotic test to be taken with care, given our sample size.

8. Similar to the respective one obtained from model (4).