5 DO MULTINATIONAL ENTERPRISES AFFECT DOMESTIC FIRMS’ PRODUCTIVITY? by Cesare Imbriani * and Filippo Reganati ** 1. Introduction 1 Foreign direct investment (FDI) and multinational enterprises (MNEs) have been traditionally considered an important vehicle in the diffusion of techno- logical and organisational innovations. In this context, several authors (Caves, 1974; Blomström, 1989; Blomström and Kokko, 1998) have stressed that one of the most important effects of FDI on host economies is the increase in pro- ductivity of local firms due to knowledge transferred from parent firms to for- eign affiliates, which then leaks out to host country firms and thus indirectly increases productivity. Following previous papers on FDI spillovers, this work represents our final contribution within the methodological context largely used in the empirical literature on this topic. We utilise sectoral data on Italian manufactures for two reasons: first because this approach permits comparison to empirical studies on spillovers from FDI in other countries; second because, despite the limited data in Italy, this study might be a reasonable starting point for inter- preting some characteristics of the Italian case 2 . * University of Roma, La Sapienza, Istituto di Economia e Finanza, P.le Aldo Moro, 00185 Roma, Italy, Tel: +39 6 490642, Fax: +39 6 49910648, E-mail: imbriani@uniroma1.it. ** University of Foggia, Facoltà di Economia, Piazza IV Novembre 1, 71100 Foggia, Italy, Tel: +39 881 724026, Fax: +39 881 720566, E-mail: f.reganati@unifg.it. 1 Financial support from CNR and MURST is gratefully acknowledged. Although this paper has been jointly developed by the authors, it was written as follows: section 1 and 2 C. Imbria- ni; section 3 and 4 F. Reganati. 2 As it has been recently stressed (Aitken and Harrison, 1999), we are conscious that a secto- ral data approach might create some problems in the identification of the causal meaning of the correlation between FDI and productivity; MNEs could not determine the high levels of pro- Studi economici n. 78, 2002/3