The spatial organization of multinational firms


  • Fabrice Defever is also affiliated with GEP and the Centre for Economic Performance at the London School of Economics and CESifo. I thank an anonymous referee for thoughtful comments and suggestions. I am also grateful to Pol Antràs, Andrew Clark, Keith Head, Laura Hering, Philippe Martin, Thierry Mayer, and Jean-Marc Siroen for helpful comments on early versions, as well as participants at the Second CEPR Conference of GIST ‘Globalization, Investment and Services Trade’ (University of Ljubljana, June 2010), the GEP conference on Trade Costs (University of Nottingham, June 2009), the ‘Agglomeration and Growth in Knowledge-Based Societies’ conference (Kiel, 2007), and GDR Urban and Spatial Economics (Lille, 2007) workshops. I also thank Nicolas Lambert and Celine Rozenblat for helping me to find the maps, and Thierry Mayer for helping me to calculate the distances between regions. An analysis of the quality of the database and its management was carried out at the Invest in France Agency (AFII) under the supervision of Fabrice Hatem and Edouard Mathieu. I thank them for their help with this task. I am also grateful to the consulting group Ernst & Young and more specifically to Barry Bright and Mark Hughes. The usual disclaimer applies. Email:


Abstract Using six years of firm-level data covering 224 regions of the enlarged European Union, we evaluate the importance to a firm of locating its activities (production, headquarters, R&D, logistics and sales) close together. We find that, after controlling for regional characteristics, being closely located to a previous investment positively affects firm location choice. However, the impact of distance is dependent on the type of investment (production or service). The impact dies out faster for service activities. Finally, we show that a surprisingly positive effect comes from locating a new production plant close to an existing production investment, but in another country.


A l’aide de données au niveau des firmes sur une période de six ans pour 224 régions de l’Union européenne agrandie, on examine l’importance pour une firme de localiser ses activités (production, siége social, R&D, logistique et ventes) tout prés les unes des autres. On découvre que, tenant compte des caractéristiques régionales, être proche de la localisation d’un investissement antérieur a un effet positif sur le choix de localisation. Cependant, l’impact de la distance dépend du type d’investissement (production ou service). L’impact s’atténue plus vite pour les activités de service. Finalement, on montre qu’il y a un impact remarquablement positif qui vient de la localisation d’un nouvel établissement de production prés d’un investissement de production existant, mais dans un autre pays.