Firms’ technological trajectories and the creation of foreign subsidiaries

Davide Castellani§
ISE-Università di Urbino
First Draft: October 2000
This Version: October 2001
Abstract
Multinational firms are traditionally considered as firms possessing some technological lead and exploiting this proprietary advantage in international markets, but a growing literature has been arguing that multinational firms set up foreign subsidiaries not only as a means to exploit their own technology but also to enrich it. This paper provides some empirical evidence in this line of analysis.
The aim of the paper is to assess the effects of the creation of foreign subsidiaries on firms’technological trajectories. The idea is that by setting up subsidiaries in foreign countries multinational firms can achieve some form of reverse technology transfer which can be expected to affect their technological trajectory. The empirical investigation has been carried out using data from 1992 to 1996 on a sample of 2,185 Italian manufacturing firms. Results support the view that the creation of manufacturing subsidiaries have a positive impact on firms’ productivity trajectories and, more interestingly, this positive impact is greater when subsidiaries are created in regions where knowledge spillovers are expected to be relatively higher, such as the U.S
JEL: C23, D24, F23
Keywords: foreign direct investments, total factor productivity, dynamic panel data
* This paper is a contribution to the TSER project on "Assessing the Impact of Technological Innovation and Globalization: the effects on growth and employment" (AITEG), whose financial support is gratefully acknowledged. The author wishes to thank Giorgio Barba Navaretti and Antonello Zanfei for supervision, Roberto Simonetti for providing some of the data used, Massi Bratti, Erich Battistin, Grazia Ietto-Gillies,Alfonso Rosolia and participants at the AITEG Meetings for comments on a earlier draft. Data preparation has been carried out with the excellent research assistantship of Elisabetta Andreani and Elvio Ciccardini. The usual disclaimer applies.

§ Correspondence: ISE-Università di Urbino, Via Saffi, 42 - 61029 Urbino - Italy.

Tel. +39 722 305560, fax +39722 305550, e-mail: [email protected]

 文中提到的其他一些实证支持性结果:

Empirical research on the impact of foreign direct investments on firm’s technology is rather scattered. The most cited example of the technology acquisition rationale for foreign investments are Japanese investments in the U.S., which appear to be driven by U.S. R&D intensity. In particular, the number of entries of Japanese firms in the U.S. by joint venture were positively related to the industry R&D expenditure gap between the two countries, with sectors where Japan was investing relatively less than the U.S. registering a higher number of entries (Kogut and Chang, 1991). Neven and Siotis
(1996) obtained a similar result for FDI from US and Japan directed towards the EU. More recently, still focussing on Japanese FDIs in the US, Branstetter (2000) found that Japanese parent firms having a higher number of subsidiaries in the U.S. experienced an increasing tendency to cite U.S. patents.
Branstetter claims that this result shows some form of knowledge spillovers from U.S. firms to Japanese parent, via FDI. More general examples which are consistent with the idea of worldwide technology sourcing through foreign direct investments come from the increasing dispersion of multinational R&D activities worldwide. Although to a large extent the core technology of the firm still builds upon research and technology developed at home, it has been shown that a process of globalisation of technology is taking place and MNEs are increasingly setting up global networks of affiliates which tap into different scientific and technological basis (Cantwell, 1995, Pearce, 1999).
As far as firm’s productivity is concerned Braconier et al. (2000) test whether, in a sample of 84 Swedish firms, labour productivity is affected by the degree of penetration of outward investments or by R&D performed in foreign countries, weighted by the share of outward FDI directed towards each country. They conclude that outward investments do not work as a channel to transmit R&D spillovers to Swedish MNEs. On the contrary, at a more aggregated level, Van Pottelsberghe de la Potterie and Lichtenberg (2001) testing for the impact of outward investments on home country’s productivity in thirteen industrialised countries find that a country’s productivity is increased when it is home to outward investments directed towards R&D intensive countries. At the sectoral level, Driffield and Love (2001) investigate whether the increase in the stock of knowledge in domestic firms, as measured by the change in domestic capital, represents a technological externality to foreign firms production in the UK manufacturing industries. Results show that such externalities exists and are appropriated by foreign firms but only in R&D intensive sectors.
As anticipated earlier, a stream of empirical work very closely related to the present one addresses the question of learning-by-exporting, that is the impact of export behaviour on firm’s productivity. The basic result of most of these studies is that the export status, i.e. the fact that a firm is in the export market in a given year, does not affect the firm’s productivity trajectory and holds for a number of countries as different as the U.S. (Bernard and Jensen, 1999), Colombia, Mexico and Morocco (Clerides et al., 1998), Germany (Bernard and Wagner, 1997), Italy (Castellani, 2001).
Nevertheless, Kraay (1999) and Castellani (2001) provide evidence that the share of export on total sales affects significantly firm’s productivity in China and Italy. Overall, these results suggest that learning from international operations might be related to the degree of commitment towards foreign markets. Such a commitment is minimum for a firm exporting a small fraction of its output and is expected to increase with export intensity and should be even higher for foreign direct investors.