Only limited evidence of positive spillovers from FDI
by Chakravarthi Raghavan
GENEVA: There is only limited empirical evidence that foreign direct investment (FDI) results in positive spillovers for host economies, according to a study presented at a recent expert meeting here organized by the UN Economic Commission for Europe and the European Bank for Reconstruction and Development.
The expert meeting was on Financing for development: Enhancing the benefits of FDI and improving the flow of corporate finance in the transition economies. The paper presented at the meeting on FDI and intra-industry spillovers is by Holger Gorg and David Greenaway (at the Leverhulme Centre for Research on Globalization and Economic Policy, University of Nottingham). Greenaway is a well-known academic and neoclassical mainstream economist, and the paper co-authored by him has surveyed the existing studies and literature.
The Gorg-Greenaway paper concludes that while theory points to several reasons why one might expect positive spillovers from multinational enterprises (MNEs), finding empirical evidence in support of positive spillovers is rather limited.
This could be because the theory is illusory and MNEs are extremely effective in protecting their assets or the wrong data is being looked at. In any event, more systematic research and more discriminating work focussing on firm- and plant-level panel (time-series) data analysis are needed to determine whether the form of investment entry (greenfield or acquisition), ownership characteristics, corporate governance and so on matter.
The consensus from the literature is also clear that general policies aimed at altering fundamentals are more important than specific policies geared to particular investments. The latter seem to affect primarily the distribution of rents.
On the one hand, governments compete in offering investment incentives and in the process dissipate rents to MNEs. On the other hand, they then use (at least some) trade-related investment measures (TRIMs) to try to reclaim some of those rents.
Both econometric evidence and survey/case study work suggest that in general the characteristics of the economic environment are much more important - infrastructure, local labour market conditions, reliability of communications systems and so on, as well as the overall macroeconomic and trade policy climate.
That, of course, does not mean that selective interventions will cease to be extensively deployed.
Governments will no doubt continue to see opportunities for targeted measures and MNEs will stand willing to accept them. This too is therefore an area for potential future work. We know very little about the comparative impact of alternative instruments, adds Greenaway.
(An area that is not covered in the paper is the effect of any proposed investment rules at the World Trade Organization on host-country measures to promote positive linkages.)
Greenaway notes that while most economists would view the outcomes of the globalization process (fuelled by international exchange of goods, services and factors of production) as fundamentally benign in the long run, it does not mean that adjustment to globalization would be costless. The short-run costs may be borne by displaced factors, and it is these that are generally the focus of public hostility to globalization.
While arms-length trade, migration of workers and cross-border investments are the drivers of globalization, cross-border investments are the most publicly visible, which perhaps explains why public hostility to globalization manifests itself as hostility towards multinationals.
According to theory, Greenaway notes, MNEs set up production facilities overseas rather than export directly and/or license their production or technology, and the most persuasive explanation for this is the emphasis on co-existence of proprietary knowledge of some form and market failures in protecting that knowledge.
In general, according to theory, among other elements, the MNEs have firm-specific advantages which may be related to the production methods they use, the way they organize their activities and the way they market their products/services. And once they have set up a foreign subsidiary, they may not be able to prevent some of the benefits of these advantages from spilling over to indigenous firms via imitation, labour mobility, competition or local firms learning to export.
Such spillovers have the potential to raise productivity and their exploitation may be related to the structural characteristics of the host economy, in particular the hosts absorptive capacity.
Surveying the evidence of productivity spillovers in about 31 studies on manufacturing industries in developing, developed and transition economies, Greenaway finds the empirical results on the presence of spillovers are mixed.
Most of the studies undertake an analysis of existence of spillovers in an econometric framework in which labour productivity or total factor productivity of domestic firms is regressed on a number of independent variables assumed to affect productivity.
Many of the earlier pioneering studies that found positive spillovers used cross-sectoral aggregated industry data, while a number of later studies using firm-level panel (time-series) data that can provide a causative linkage have found either a negative effect or no statistically significant effect.
Greenaways survey of the literature found 12 studies that did not find any statistically significant effects of MNEs on domestic productivity, and 16 others that reported statistically significant results.
However, as pointed out by Holger Gorg and Eric Strobl (in the Economic Journal, Vol. 111, 2001), cross-sectional data analysis may lead to biased results, since research design could crucially affect whether or not spillovers are found, while panel studies using data on a firm rather than on industry level appear to be most appropriate to judge the true extent of productivity spillovers.
When this is borne in mind, the evidence on productivity spillovers appears to be even bleaker. The overwhelming majority of studies finding positive spillovers, he says, use cross-sectional data and hence need to be treated with caution. Only two studies (for the UK and Romania) using firm-level panel data find positive results. The UK study also uses industry-level data and thus sub-optimal aggregating data over heterogeneous firms.
Some of the positive spillover effects are seen when MNEs locate in an economy to take advantage of its import-substitution strategy, while those locating for exports do not seem to have such an effect.
In surveying the literature on productivity, wage and export spillovers, and empirical evidence of intra-industry spillovers, Greenaway says: Overall, only limited evidence in support of positive spillovers has been reported. Most work fails to find positive spillovers, with some even reporting negative spillovers at the aggregate level. Evidence on wage and export spillovers is also mixed.
Studies that further disaggregate data however find some evidence for spillovers on firms that have a certain level of absorptive capacity. However, further study is needed to determine the factors that allow domestic firms to build up such capacity, such as access to finance for investments, human capital and management expertise.
In general, notes Greenaway, the precise determinants of economic growth remain far from settled, both in theory and in empirical literature, though one driver persistently identified as important is investment. Most policy-makers in most settings understand this and see investment promotion as a priority.
FDI, more than an equivalent amount of indigenous investment, would be seen by most governments as having the potential to impact on total factor productivity. Foreign firms are seen as more likely to bring in their best practice or better practice technology and/or management than those available to local firms. And this perception, added to the view that MNEs have potential spillovers to domestic firms, explains why attracting inward investment figures so prominently in the list of policy priorities of many governments.
Overall, the evidence on the role of policy in influencing the level and composition of FDI would seem to suggest that in general intervention should be targeted largely at providing a supportive economic environment: education and training policies aimed at upgrading general skills, technology policies aimed at developing clusters, and public investment policies aimed at developing efficient and reliable transportation and communication networks.
However, the evidence on spillovers is not encouraging, whether one takes developing, developed or transition economies, and little evidence in support of the presence of spillovers has as yet been reported.
Until the Uruguay Round of trade negotiations, TRIMs could be used as legal or extra-legal methods to encourage positive spillovers and there had been a proliferation of them in developing countries. But the Uruguay Round agreements have proscribed the use of a number of them and have laid out a range of reporting requirements. Several of them, in particular those for local-content requirements, have to be phased out.
These are, according to the study, second best measures and what little work has been done on them has failed to establish a direct link between their presence and transfer of useful technologies. However, this appears to be because many of the measures are difficult to specify precisely and difficult to monitor, and also because more general policies (to promote spillovers) are in practice rather more important. (SUNS5027)
The above article first appeared in the South-North Development Monitor (SUNS) of which Chakravarthi Raghavan is the Chief Editor.