外商直接投资,技术寻求和逆向技术溢出效应【外文翻译】

外文翻译原文FOREIGN DIRECT INVESTMENT, TECHNOLOGY SOURCING AND REVERSE SPILLOVERSMaterial Source:The Manchester School Vol 71 No. 6 December 2003 Author: NIGEL DRIFFIELD Business School, University of Birmingham And JAMES H. LOVE† Aston Business School, Aston University Recent theoretical work points to the possibility of foreign direct investment motivated not by ‘ownership’ advantages which may be exploited by a multinational enterprise but by the desire to access the superior technology of a host nation through direct investment. To be successful, technology sourcing foreign direct investment hinges crucially on the existence of domestic-to-foreign technological externalities within the host country. We test empirically for the existence of such ‘reverse spillover’ effects for a panel of UK manufacturing industries. The results demonstrate that technology generated by the domestic sector spills over to foreign multinational enterprises, but that this effect is restricted to relatively research and development intensive sectors. There is also evidence that these spillover effects are affected by the spatial concentration of industry, and that learning-by-doing effects are restricted to sectors in which technology sourcing is unlikely to be a motivating influence.1 IntroductionTraditional models of foreign direct investment (FDI) have been heavily influenced by a framework which suggests that where a company has some ‘ownership’(i.e. competitive) advantage over its rivals and wher e, for reasons of property rights protection, licensing is unsafe, a company will set up production facilities in a foreign country through FDI (Dunning, 1988). Since much of the discussion of ownership advantages is couched in terms of technology and/or m anagement expertise, there is a strong a priori assumption that this ‘technology exploiting’ FDI will be an important method by which technology is transferred internationally. Indeed, there is a growing literature concerned with the extent to which FDI contributes to technological advance in host countries. Much of thisanalysis is based on estimations of externalities from inward FDI, with the evidence generally pointing towards positive effects of FDI on domestic productivity (Blomström and Kokko, 1998).However, the literature is increasingly turning to the possibility that FDI may be influenced by multinational firms’ desire not to exploit an existing ownership advantage abroad but to acquire technology from the host country, i.e. that ‘technology sourcing’ may be the motive for FDI. Kogut and Chang (1991) and Neven and Siotis (1996) point out that this possibility has exercised the minds of policy-makers in the USA and the EU, with concerns that host economies’ technological base may be undermined by technology sourcing by Japanese and US corporations respectively. These studies examine the effects of host versus home country research and development (R&D) expenditure differentials on FDI flows between Japan and the USA and the USA and the EU respectively. Both studies find a positive relationship between these measures, and interpret this as evidence of technology sourcing. The literature on the internationalization of R&D also contains an increasing amount of evidence that technology sourcing may be a motive for FDI (Cantwell, 1995; Cantwell and Janne, 1999; Pearce, 1999).This literature stresses a range of reasons for FDI in R&D, much of which is concerned with the relative technological strengths of the capital exporting (i.e. ‘home’) firm or country v ersus that of the host. For example, Kuemmerle (1999) distinguishes between ‘home-base exploiting’ FDI and‘home-base augmenting’ FDI. The former is undertaken in order to exploit firm-specific advantages abroad, while the latter is FDI undertaken to access unique resources and capture externalities created locally. And in an analysis of inward and outward FDI in 13 industrialized countries, van Pottelsberghe de la Potterie and Lichtenberg (2001) find positive spillover effects from outward FDI arising from accessing the R&D capital stock of host countries, leading them to conclude that FDI flows are predominantly technology sourcing in nature.Recent theoretical work represents an important step forward in this area, with Fosfuri and Motta (1999) and Siotis (1999) both presenting formal models of the FDI decision which embody the possibility of technology sourcing. They show that a firm may choose to enter a market by FDI in order to access positive spillover effects arising from close locational proximity to a technological leader in the host country. Because of the externalities associated with technology, these spillovers decrease the production costs of the investing firm both in its subsidiary operations and in its home production base. Siotis (1999) also shows that the presence ofspillovers may induce firms to invest abroad even where exporting costs are zero.The theoretical and empirical work reviewed above hinges crucially on the assumption that foreign firms investing in a host economy are able to capture spillover effects from the domestic (host) industry. The purpose of this paper is to test for the existence of this ‘reverse spillover’ effect for a panel of UK industries. If there is some evidence of productivity spillovers running from the domestic to the foreign sector of UK industry, this would suggest that the necessary condition for technology sourcing FDI does exist in practice. In addition to testing empirically for reverse spillover effects we also test for two elements which are implicit in the theoretical analysis: first, that the spatial concentration of production has an effect on productivity spillovers; and second, that learning-by-doing effects are linked to the investing motivations of foreign firms.2 THE MOTIV ATION FOR FDI, SPILLOVERS AND FIRM GROWTHFosfuri and Motta (1999) present a simple model in which two local (i.e. single country) firms are endowed with different technologies and are given the option of exporting to the other country, engaging in FDI or not entering. They show formally that an investing firm which is a technological laggard (i.e. has unit costs of production above those of its competitor) will find it profitable to invest abroad despite having an efficiency disadvantage, as long as the probability of acquiring the leader’s technology through productivity spillovers is sufficiently high. In other words, ‘technology sourcing’ rather than ‘technology exploiting’ FDI may occur. Siotis (1999) develops a similar model, but allows for the possibility of two-way spillovers between foreign and domestic firms. He too finds theoretical support for technology sourcing as a motivation for FDI.It seems plausible that the probability of benefiting from productivity spillovers will at least in part be dependent on the actions of the firms concerned, and that the scope for spillovers, particularly in the context of technology sourcing investment, will vary with the research efforts of domestic firms. Thus technology sourcing is most likely to occur where the scope for productivity externalities to be assimilated by foreign firms is greatest; this in turn is a positive function of the R&D intensity of domestic industry. We therefore anticipate reverse spillover effects being most apparent in those sectors in which domestic industry has invested heavily in R&D; these are the sectors in which the probability of acquiring technology through spillovers is greatest and in which technology sourcing FDI is most likely to occur. However, traditional explanations for FDI based on the exploitation offirm-specific‘ownership’ advantages shou ld not be ignored. Siotis (1999)shows that where a foreign firm has an ownership (i.e. efficiency) advantage relative to domestic firms, FDI will only occur if spillovers are likely to be small (the ‘dissipation effect’). We therefore anticipate technology exploiting FDI to be most likely where there is little scope for reverse spillovers, i.e. where domestic industry does not invest heavily in R&D. Reverse spillover effects should therefore be most evident in relatively research intensive sectors, but absent or less evident in sectors which are relatively non-research intensive.Two further and related hypotheses can also be tested. The first relates to the growth paths exhibited by firms that have different motivations for FDI. To the extent that it is possible to make the distinction between technology sourcing and technology exploiting FDI, then it is also likely that the patterns of development arising from these forms of investment will be different. This is likely to be important in the study of the development of total factor productivity in the foreign owned sector, following the theory of the multinational enterprise dating back to Dunning (1958) and more explicitly outlined in the seminal papers by Vernon (1966), Buckley and Casson (1976) or Dunning (1979). The traditional explanation of the existence of multinational enterprises is that firms transfer firm-specific assets across national boundaries but internalized within the firm (technology exploiting FDI). Firms operating in the foreign country then have to undertake the process of adapting this technology to a new environment, to take account of local working practices, available human capital and customers’ tastes for example. This is neither costless nor instantaneous, and so total factor productivity of foreign investment motivated in this ‘traditional’ manner is likely to demonstrate experience effects and significant learning-by-doing effects. By contrast, firms motivated by technology sourcing are less likely to undergo this adaptation of internal technology: their concern is not with adapting existing technology but in assimilating knowledge generated externally, in this case by local firms.Of course, in some cases the extent of adaptation by technology exploiting firms may be minimal in certain markets, while technology sourcing subsidiaries may undergo some degree of adaptation, so that the relative extent of learning by doing is ultimately an empirical issue. On balance, however, we expect significant learning by-doing effects among technology exploiting foreign firms, but perhaps not in the technology sourcing firms, where spillovers from domestic investments are likely to contribute more to total factor productivity in the foreign sector.The second subsidiary hypothesis relates to the extent to which technological externalities are constrained spatially. The theoretical analysis of Siotis (1999) depends on the existence of geographically localized spillovers to provide an incentive for technology sourcing FDI; Fosfuri and Motta (1999) also acknowledge this geographical dimension to spillovers. Empirically, there is significant evidence that technology spillovers are indeed limited geographically within countries, as well as between them (Head et al., 1995; Driffield, 1999). This suggests that reverse spillovers may be linked to the spatial distribution of industry; we therefore test whether the spatial concentration of production has an effect on the scale of productivity spillovers running from domestic to foreign industry.译文外商直接投资，技术寻求和逆向技术溢出效应资料来源:曼彻斯特大学学报71卷第6期作者：奈杰尔•德里菲尔德英国伯明翰大学商学院；詹姆斯H.爱阿斯顿商学院，阿斯顿大学近期的理论研究表明，外商直接投资的动机可能不是“所有权”优势，而是跨国公司希望通过直接投资积极利用东道国的先进技术。