The Impact of FDI on Firm Survival and Employment: A Comparative Analysis for Turkey and Italy

This report summarizes the findings of a research project using firm level data on Italian and Turkish manufacturing industries. In this project we study the dynamics of firm survival and growth, and the spillover effects from foreign-owned to domestic firms. First, we investigate the differences in survival patterns of foreign-owned and domestic firms and test the hypothesis that foreign multinational enterprises (FMNEs) display “foot-loose” behavior. Secondly, we analyse the effects of FDI on the survival and growth prospects of domestic firms by disentangling horizontal and vertical spillovers. We use hazard models for the econometric analysis of firm survival and the system-GMM and Heckman selection models for the analysis of firm (employment) growth. In the case of Italy, a comparison of survival rates of domestic and foreign firms shows that foreign firms are more likely to survive than domestic firms, although the survival rates of foreign firms are not much different than those of Italian multinational firms. To check for a more general applicability of this preliminary finding, we estimate the hazard functions for the domestic and foreign firms, controlling for a number of sector-specific and firm-specific characteristics. The results reveal that foreign firms are more “foot-loose” compared to their domestic counterparts while Italian multinationals exhibit lower hazard rates with respect to both domestic non-multinational firms and to foreign multinationals. Besides, the foreign firms’ likelihood of exit compared to domestic firms is higher in sectors with low technology- and knowledge-intensity. In the Turkish case, the simple comparison of survival rates also highlights that foreign firms are more likely to survive than Turkish firms, although the survival rates of foreign firms are not different from those of large domestic firms. Since foreign firms usually start with a larger size, use more capital-intensive technologies, survival rates may reflect the impact of entry characteristics. The hazard function estimates reveal that, when we control for sector-specific variables, foreign firms still have higher survival probabilities, but once firm-specific variables are included in the hazard function model, they appear more “foot-loose” for the 1983-2001 period. Foreign firms are more likely to survive than the domestic firms in the 2003-2009 period even after firm-specific variables are taken into account, but the inclusion of firm-specific variables reduces the impact of foreign ownership on the likelihood of survival considerably. These results for Italy and Turkey indicate that foreign ownership has not necessarily a positive impact on firm survival. Conversely, there is evidence that multinational experience matters for survival because multinational firms have larger size and may employ more capital-intensive technologies thanks to their superior financial strength and experience in other markets. Other firm-level characteristics (size, skill level, etc) are also crucial for survival. The exit behavior of foreign firms is also quite related to the technological environment due to the role played by opportunity costs, which are more relevant in low-tech industries, and by sunk investments costs, which (on average) are lower in more traditional sectors. The mixed results for Turkey across the two periods considered also highlight the importance of the institutional setting for firm survival and growth. Turkey experienced two different policy and growth regimes in the 1990s and 2000s. The 1990s, which is labeled by some researchers as the “lost decade”, is characterized by extreme uncertainty and boom-and-bust cycles, whereas the Turkish economy achieved a high and stable growth performance in the 2000s. In terms of industrial policy, the foot-loose behavior of foreign multinationals should be taken into account in designing investment incentives to attract foreign multinationals also pursuing sector specific policies and institutional reforms ensuring that managers have the right incentives to make long-term investment and to enhance absorptive capacity development. Besides, to improve the likelihood of firm survival, policy makers should target firm-specific characteristics that are crucial determinants of performance gaps in survival, primarily size, productivity and multinational activities. Concerning the issue of how the presence of foreign firms affects the domestic firms’ survival and employment growth, our findings suggest that there is a huge degree of heterogeneity across firms, periods and sectors in both countries. However, positive evidence in favour of positive spillovers is not overwhelming. In the case of Italy, the survival of domestic firms is positively affected by the increased presence of foreign firms within the same industry, but this only occurs in low- and medium-low tech industries. This result may be due to the fact that domestic firms in medium-high tech industries have not enough absorptive capacity to benefit from FDI spillovers. The relevance of domestic firms’ absorptive capacity for spillover effects is confirmed by our analysis: only domestic firms that have smaller technology gap vis-à-vis foreign firms benefit from significant horizontal and vertical (upstream) spillovers on survival. From the system GMM growth estimates we find that, in terms of FDI spillovers, there is evidence of a negative impact on domestic firms employment growth if the foreign firm share in the region employment increases (negative local spillovers), .and a negative employment impact for firms with a higher technology gap is detected if the foreign firm share in the sector increases. For Turkey, the regional share of foreign firms has a weak negative static impact on the survival rate, and an increase in the share of foreign firms in a sector also has a negative impact on survival in the 2003-2009 period. The foreign share of users seems to have positive coefficients, i.e., domestic firms will be more likely to survive if users are foreign, but these results are statistically significant only if firm-specific effects are not controlled for in the 2003-2009 period. Moreover, there is some evidence of a negative effect on survival if downstream firms are foreign in the 2003-2009 period. Regarding firm growth, foreign suppliers and change in regional share of foreign firms have strong negative impact on domestic firms' growth rates, i.e., those firms supplied by upstream foreign firms, and those firm operating in regions with an increasing foreign presence experience lower growth rates. There is also a weak negative impact of sectoral foreign share on growth whereas a weak positive impact is observed for the change in sectoral foreign share. These results do not support the broad conclusion that FDI have positive impact on firms’ indigenous survival and growth dynamics. Conversely, our findings provide not a favorable picture in terms of the balance between displacement/competition versus spillover effects of FDI on domestic firms. We also obtain evidence indicating that the interaction between the presence of foreign firms and domestic firm survival is markedly affected by the technological environment that shapes up domestic firms’ absorptive capacity. The displacement effect in dynamic industries implies that the damage is concentrated on high-tech firms, which should be the higher quality segment of national production. In terms of industrial policy, this implies that the desire to encourage FDI and simultaneously building up a stable supply of indigenous enterprises is more challenging in dynamic sectors, where a trade-off in terms of these objectives appears to exist.