اثرات اشتغال مالکیت خارجی

This paper investigates the employment effects of foreign acquisitions in acquired firms in Swedish manufacturing during the 1990s; a period characterized by a dramatic increase in foreign ownership. We find some evidence of positive employment effects in acquired firms and it seems that the employment of skilled labor increases more than that of less-skilled labor. Our results indicate that the positive employment effects are more pronounced in acquired non-MNEs than in Swedish MNEs. Furthermore, we observe shifts in skill intensities toward higher shares of skilled labor in non-MNEs taken over by foreign MNEs, but not in acquired Swedish MNEs.

During the last two decades, many countries have witnessed a dramatic increase in foreign ownership mainly driven by foreign acquisition of national firms. In the public debate such a development has given rise to mixed feelings. Some fear that foreign acquisitions lead to job losses in acquired firms because foreign owners would be less committed to the host economy. Moreover, multinational enterprises (MNEs) are more footloose owing to their possibilities to relocate production and employment between their affiliates in different countries.1 Others maintain that foreign acquisitions strengthen the competitiveness of the acquired firms due to transfers of technology, knowledge and skills from the acquiring foreign MNE which, in turn, improve the performance in target firms which may involve higher employment. The purpose of this paper is to contribute to this discussion by carrying out a systematic investigation of the effects of foreign acquisitions on employment in acquired firms. To this end, we use a panel of Swedish manufacturing firms between 1993 and 2002. Sweden is an interesting case to study in this context. Ever since 1990s, Sweden has experienced a rapid increase of foreign ownership mainly through mergers and acquisitions (M&As). Well-known former Swedish owned firms, such as Astra, Pharmacia, Volvo Car and Saab Automobile, changed ownership in the 1990s and are nowadays foreign owned. This has entailed that at the beginning of the 2000s, the employment share in foreign-owned firms in Swedish manufacturing was around 40%, which was among the highest in OECD.2 Economic theory on M&A in general does not give any explicit answer to questions related to employment effects in acquired firms. Theories on M&As originate from the industrial economics (IO) and theories for international trade and foreign direct investments. The IO literature is mainly limited to national M&As and gives no special attention to foreign acquisitions of national firms.3 Theories on FDI emphasize comparative advantages, Dunning, 1977 and Markusen and Venables, 1998. A firm needs to possess ownership advantages, like a better product, a strong brand, or a more efficient production etc. in order to become a multinational firm. These ownership advantages may then be exploited in affiliates abroad, Caves (1974). According to this theory increasing presence of FDI in an industry is associated with more advanced technology and an increasing demand for skilled labor.4 The theory on multinational firms doesn't separate between national and foreign MNEs with respect to demand for technology and skilled labor. The skilled labor demand in a Swedish MNE may have been high pre-acquisition and the motives as well as effects on employment may be quite different from acquisitions of Swedish non-MNEs. If so, we might expect significant restructurings of the labor force in acquired non-multinational firms. Separating between foreign acquisition of Swedish MNEs and Swedish non-MNEs is thus a main contribution to the empirical literature on international M&As. A number of papers have examined the impact of domestic acquisitions on employment and the results are ambiguous.5 Other, more recent studies for the UK have focused on employment effects of foreign acquisitions. Girma and Görg (2004) provide some evidence of reduced employment growth in domestic plants taken over by foreigners in the electronics sector but not in the food sector. Girma (2005) finds, on average, no impact of foreign acquisitions on employment in acquired domestic firms. Huttunen (2007) for Finland finds that foreign acquisition has a negative effect on the share of highly educated workers among the plant's employees. In another study for Finland, Eero and Böckerman (2008) find that cross-border M&As lead to downsizing in manufacturing employment, whereas the results for the cross-border M&As in nonmanufacturing are weaker.6 For Indonesia Arnold and Javorcik (2009) and for Italy Piscitello and Rabbiosi (2005) find that foreign direct investments increase productivity, employment and wages in acquired plants. A closely related strand of literature deals with effects of offshoring on labor demand. The empirical evidence suggest that there is small but negative effects on labor demand in the home country, see Görg and Hanley, 2005 and Andersson and Karpaty, 2007. We extend and improve upon these earlier papers in a number of ways. First, we focus on acquisitions by foreign MNEs and by using detailed and unique data for Sweden we are able to separate the acquired firms into domestic MNEs and non-MNEs before a foreign takeover. Hence, we analyze whether the impact of foreign acquisition on employment differs between takeovers of domestic MNEs and non-MNEs. This is an important dimension to the analysis since the acquisition effect on firm performance, i.e. employment, may differ whether the target firms are multinationals or not due to their ex-ante different characteristics as highlighted by, for example Helpman, Melitz, and Yeaple (2004). Second, we use different econometric approaches to evaluate the causal effect of foreign acquisition on targeted firm's employment and to take account of the potential endogeneity of the acquisition decision (for example due to “cherry picking”).7 Specifically, we use an extended version of the matched difference-in-difference (DiD) method suggested by e.g. Blundell and Costas Dias (2000) by in the first step matching, on a yearly basis, the non-acquired and acquired firms with similar propensity scores and in the next step estimate DiD on the matched sample. This methodology allows for different vectors of covariates xt in the matching and difference-in-difference estimations. 8 As an alternative estimation strategy we estimate a growth rate model taking account of the potential endogeneity of the acquisition decision by implementing an instrument variable approach. To preview our results, we find some evidence of positive employment effects in firms taken over by foreigners. If we, like Girma and Görg (2004), divide employment into skilled and less-skilled labor, it appears that employment of skilled labor increases more than employment of less-skilled labor.9 In Girma and Görg (2004), there is a reduction in employment of unskilled labor, whereas in plants acquired by foreign firms in the electronics industry, the employment of skilled labor is unaffected. Finally, we postulate that the impact on employment of foreign ownership differs depending on whether a Swedish MNE or a non-MNE is taken over. We expect the scope for restructuring and changes in employment to be larger in acquired non-MNEs than in more productive Swedish MNEs acquired by foreign firms. Moreover, we anticipate larger potentials for technology transfer from foreign MNEs to acquired non-MNEs than to acquired MNEs. If technological changes are skilled-biased we expect to see more pronounced shifts in skill intensities towards higher shares of skilled labor in non-MNEs taken over by foreign MNEs.10 The results from our empirical analysis are consistent with both these hypotheses. The paper is structured as follows. Section 2 outlines the methodological framework and the econometric specifications. Section 3 describes the data and the construction of the matched sample. Section 4 reports the empirical findings. Section 5 summarizes and concludes.

In the 1990s, the employment share in foreign-owned firms in Swedish manufacturing has grown spectacularly. To examine the employment effects in firms that have become foreign owned in the 1990s, we utilize propensity score matching technique with difference-in-difference estimation and as an alternative estimation strategy we implement an instrumental variable approach. We allowed these effects to be different for less-skilled and skilled employees. Moreover, we argue that the effect may be smaller in firms already engaged in international networks. Therefore, we also allowed for differential impact on employment due to acquisitions of Swedish MNEs and acquisitions of non-MNEs. Our analysis gives no support to the worries that foreign acquisitions may lead to job losses in acquired firms. If anything, there are some indications of positive employment growth in acquired non-MNEs, especially of skilled labor. This may be due to technology transfers from the acquiring foreign MNEs that lead to skilled-biased technical change in acquired non-MNEs. In Swedish MNEs taken over by foreigners, there appear also to be positive employment effects. However this effect is much smaller than in acquired non-MNEs.