chapter  10
Foreign direct investment and firm survival in transition countries
Pages 24

The data used in these studies cover countries as diverse as Colombia, Cote d’Ivoire, Ireland, Morocco, Mexico, Sweden and Venezuela. It is true that such studies do not map one-to-one with the TC-specific studies. Nevertheless, there are parallels. This raft of studies similarly investigates issues, such as absorptive capacity, which are just as relevant for Europe’s TCs. What the bulk of these studies document is the ability (or inability) of host country firms to commercialize the new knowledge introduced by foreign firms. These studies also comment on the rates at which local firms get crowded out of the marketplace in response to foreign entry. Some studies look at the timing of the effects for local firms (e.g. Barrios et al., 2005) or others at the absorptive capacity of local firms enabling them to take on board new knowledge. Some of these studies focus on the induced effects for firm productivity (e.g., Globerman, 1979; Aitken and Harrison, 1999). Others look at firm survival or net entry (e.g. De Backer and Sleuwaegen, 2003; Görg and Strobl, 2003; Bandick and Görg, 2010). Both types of studies are relevant to this debate. Our intuition for seeing parallels in studies which look at the firm’s extensive margin (net entry or survival) and intensive margin (sales growth or productivity growth) is as follows: positive productivity changes induced by FDI spillovers are captured (appropriated) by the fittest local firms. The increased competitiveness of local firms, all other things equal, should help their ultimate survival.3 We would like here to single out the panel data analyses among this category of studies. Panel data allow us to capture some fairly static features of firms such as management and ownership structure, which can help to explain differences in overall outcomes such as survival. Generally, these studies with panel data which examine FDI spillovers in less developed countries do not find positive impacts for foreign firms entering the same industry (Haddad and Harrison, 1993; Djankov and Hoekman, 2000; Aitken and Harrison, 1999). The only exception to this consensus is the study by Aitken et al. (1997) for FDI into Mexico. Foreign firms entering Mexico typically used the host country as a manufacturing location for goods which were subsequently predominantly exported (export platform). Domestic firms seized this opportunity to enter the

same export markets due to heightened awareness on what, where to and how to export their produce. This example by Aitken et al. is interesting because the knowledge spilled over is neither technical nor scientific. Rather, the new knowledge can be classified as ‘generic knowledge’, which has more to do with easyto-imitate business practices. So far we have looked at the impacts of FDI at the firm level. But impacts may also take place at a higher level of aggregation: at the level of the industry sector. In Belgium there is evidence that although foreign firms steal market share from domestic firms and generally crowd them out, the survival of existing local firms, paradoxically, is generally higher in industries with a high foreign presence. This suggests the presence of learning, demonstration, networking and linkages effects between domestic incumbents and foreign entrants (De Backer and Sleuwaegen, 2003). However, the De Backer and Sleuwaegen findings are based on cross-section data and may fail to capture some dynamic effects. Görg and Strobl (2003) confirm this finding, but this time for panel data. Using Irish data, they similarly find better survival chances for local high-tech domestic firms situated in the high-technology sector. Moreover, this positive survival bias still holds, even in sectors with a high foreign-firm presence. What we can take from this finding is that the survivability of domestic firms is greatly helped in sectors where domestic firms are able to meet foreign competition head-on and appropriate technology transfers (Görg and Strobl, 2003). Related to this, Bandick and Görg (2010) also investigate whether firm survival is conditioned on the absorptive capacity of local firms. They apply a careful econometric framework (instrumental variables (IV) and propensity score matching) distinguishing between the survival prospects of domestic firms acquired by foreign firms (treatment group) and those domestic Swedish firms acquired by new Swedish owners (control group). Acquisition by foreign firms only fosters the survival of Swedish firms which were taken over and which had already been exporting before the takeover date. Therefore, the firm’s exporting history is the key determinant of a firm’s readiness to take on board new knowledge from foreign entrants. By how much do the firm’s survival chances improve? This depends largely on the nature of the acquisition: vertical or horizontal integration. Specifically, the survival of foreign-acquired firms increases by between 17 per cent and 34 per cent for vertical acquisitions. The corresponding percentage change for horizontal acquisitions is lower at 6 per cent and 8 per cent, respectively. The message to take from Bandick and Görg (2010) is that the survival chances are highest for local firms with a history of internationalization. It is these firms which are best able to appropriate new knowledge from foreign entrants. What these studies does not reveal more explicitly is the nature of foreigndomestic firm cooperation. To answer this question, we need to look at foreignlocal firm relationships through the lens of linkages. Accordingly, Kugler (2006) takes the literature a step further by considering changes in the lagged capital formation (investments made) and productivity growth by domestic firms in Colombia on the basis of same-and different-sector FDI. His co-integration

analysis reveals that foreign entry into the same sector crowds out local firms but promotes growth in other sectors. However, it could be argued that the FDIproductivity relationship represents a ‘black box’ unless we can similarly throw light on the way in which foreign and local firms are connected. In the words of Javorcik (2004, p. 606): ‘Kugler (2000) also finds intersectoral technology spillovers from FDI in Colombia. However, he does not distinguish between different channels through which such spillovers may be occurring (backward versus forward linkages).’ Indeed, most firm-level studies, following the seminal Javorcik (2004) paper which popularized Blalock and Gertler’s (2003, 2008) technique of decomposing inter-firm connections (linkages), have applied the same taxonomy. This taxonomy first suggested by Blalock and Gertler (2003, 2008) distinguishes vertical and horizontal backward linkages. More recently, forward vertical and horizontal linkages have similarly been investigated. In this regard, Barrios et al. (2005) specifically consider the timing of impacts. We can conclude from their analysis that when looking at the overall impact of FDI on the net entry rate of domestic firms (entry rate – exit rate), linkages with foreign entrants can only be set up after some time has elapsed. The same is true for learning, demonstration and network effects. In these cases, domestic firms may have to upgrade production facilities/workforce skills/R&D capability. These efforts take time. Therefore, it is to be expected that, initially, foreign entry induces negative effects on net entry due to the dominance of the competition effect where weaker domestic firms are forced out of the market. These negative effects in time are moderated or even reversed by the positive spillover effects. The result: a U-shaped pattern for the effect of FDI on net entry rates. Their theoretical framework shows how FDI can be overall positive for local firms’ employment growth, especially if the FDI brings with it larger capital endowments and greater efficiency of the local firm. While FDI prompts the exit of some weaker local firms, the remaining firms can enjoy a windfall of spillover effects. The policy implication of Barrios et al. (2005) is that efforts should be made to shorten the transition period by which weaker domestic firms exit the market while supporting the efforts of remaining domestic firms to improve their absorptive capacity and productivity. We will return to the question of absorptive capacity in the next section, because it is a problem that can moderate or even reverse the otherwise positive predicted effect of FDI on the productivity of local firms, specifically in Eastern Europe’s TCs.