اتحادیه های کارگری باید از سرمایه گذاری مستقیم خارجی قدردانی کنند

It is often argued that FDI hurts workers in the home country because jobs are moved abroad. Contrary to that view, businessmen often argue that FDI benefits home workers because there will be an expansion in the firm. We show that both views may be correct, and whether home workers gain or lose on FDI depends on which kinds of activities the firm moves to the host country. If there is a big degree of substitutability (complementarity) between activities in the home country and activities in the host country, it is likely that the workers lose (gain) on FDI.

How the labor market in the home country is affected by outward foreign direct investment (FDI) has always been a controversial issue (see e.g. Lipsey, 1994), and there seems to be at least two popular views. One view, often heard from trade unions, is that FDI is an export of jobs. The other view, often heard from managers of firms, is that these investments will be to the advantage of workers in the home country because the lower wage costs in the host country make the firm more competitive. In this paper we show that both views may be correct since the effects of FDI depend on which kinds of activities the firm moves abroad. More specifically, we set up a simple model with one firm which at the labor market in the home country faces a trade union. These two parties bargain over the wage rate in the home country. The production in the firm can be split into several activities, and the firm has the opportunity to become multinational and move one of the activities to the host country. In the host country, the firm has to bargain over the wage rate with a local trade union. In our analysis, we focus on whether it would be optimal for the firm to become multinational and whether the trade union in the home country would appreciate that the firm becomes multinational. We find that, if there is a big degree of complementarity between activities moved abroad and activities remaining in the home country, it is likely that the firm and the home country workers agree whether it is preferable to undertake FDI. On the other hand, if there is a big degree of substitutability between activities in foreign affiliates and activities in the home country, it is likely that the firm gains but the workers lose from FDI. A paper closely related to ours is Horn and Wolinsky (1988). They consider two groups of workers employed in a single firm, and these workers can either choose to be organized in a single encompassing trade union or in two separate trade unions. The firm bargains with the trade union(s) over the wage rate(s). A main result is that, if the two groups of workers are sufficiently close substitutes in production, it is an equilibrium that the workers unite in an encompassing trade union, whereas, if the degree of complementarity between the two groups of workers is sufficiently high, the equilibrium is two separate trade unions. The intuition is that, if the two groups of workers are complements in production, each group is more or less able to paralyse the firm on its own. Therefore, the workers are better off bargaining with the firm in two separate trade unions. Oppositely, if the workers are substitutes, separate groups have a weaker stance in the bargaining with the firm than if the two groups unite. The reason being that the total damage on the firm if both groups stop producing, simultaneously, is bigger than two times the damage if a single group stops producing. The same basic mechanism is at work in our paper, and it partly explains why workers in the home country may gain if the firm becomes multinational. If the activities kept in the home country are sufficiently complementary to the activities moved to the host country, the workers in the home country keep a strong bargaining position relative to the firm. However, there are also important differences between our model and the model by Horn and Wolinsky. First, in our model, it is the firm which decides whether to move activities to the host country and by this splitting the workers into two groups.1 Second, in the Horn and Wolinsky model, it is the same workers who are employed in the firm no matter whether there are one or two trade unions. In our model, a share of the employment is ”given away” to the workers in the host country. Third, in the Horn and Wolinsky model, the firm determines employment before workers decide whether to make an encompassing union, and before the wage bargaining takes place. In our model, employment is determined after the firm has decided whether to move activities abroad and after the wage bargaining has taken place. Hence, contrary to what is the case in the Horn and Wolinsky model, employment consequences of the firm’s investment decision is essential in our model. In the literature, there are a few other papers which analyze the implications of FDI when labor markets are unionized. Two examples are Zhao, 1995 and Zhao, 1998, but the set up and focus of these papers are very different from ours. Zhao finds that FDI gives rise to lower wages, whereas we find that it depends on the degree of substitutability between activities in the home country and the host country.2 There are empirical work that put light on what we consider in our paper. First of all there are results which show that the whole issue about whether home and host country activities are substitutes or complements is empirically relevant. Brainard and Riker (1997) estimate cross-elasticities of substitution between labor in parent US firms and foreign affiliates and between labor in different affiliates. They show that there is a positive elasticity of substitution between parent employment and employment in foreign affiliates but it is relatively small. The largest cross-elasticity of substitution is found between employment in affiliates in developing countries. Contrary to that, it is found that between employment in affiliates in industrialized and developing countries, the elasticity of substitution may be negative (i.e. there is complementarity). Riker and Brainard (1997) and Braconier and Ekholm (2000) get similar results, the last mentioned paper by using Swedish data. The above mentioned papers consider whether employment in different locations are substitutes or complements. However, they do not put much light on how workers are affected by FDI as such. With respect to that, Kravis and Lipsey (1988) and Lipsey (1994) find that FDI by US firms tends to give rise to lower employment but higher wages in the US. In these papers, this result is interpreted as an upgrading in average skills in the parent firm, but our paper may give another explanation, namely that workers in the parent firm get power to bargain for higher wages. Slaughter (2000) documents that affiliates are more intensive in production workers (relative to non-production workers) than their US parents. In spite of that, he does not find that FDI tends to decrease demand for production workers relative to non-production workers in the US. This is fully consistent with our model as we find that, if a firm moves activities abroad, it does not necessarily decrease demand in the home country for labor involved in that activity. Recently, there has been a lot of interest in whether increasing international trade has had implications for relative wages (see e.g. Lawrence, 1994). This interest has been empirically grounded by the fact that the wage gap between skilled and unskilled workers has been increasing. In this discussion, it has also been natural to look at the potential role of multinationals as in Markusen and Venables (1997). They show that FDI always gives rise to an increase in the wage of skilled workers in the skilled labor abundant country, and it is possible that this is also the case in the unskilled labor abundant country. Our work is certainly related to this literature, but we should emphasize that our aim is not to put new light on the role of multinationals in explaining the wage gap. Therefore, in order to simplify our analysis of the role of the degree of substitution between home country and host country activities, we focus on only one kind of labor. In Section 2, we set up the basic model, and in Section 3 we analyze the implications of FDI. Finally, we conclude in Section 4.

In this paper, we have shown that a firm and a local trade union may agree whether the firm should become multinational. The reason why it is possible that the local trade union finds it attractive that the firm becomes multinational is that the total wage cost of the home workers becomes less important for the firm. The trade union in the home country may exploit that to get a higher wage in the bargaining with the firm. It is likely that employment in the home country decreases, but the higher wage may compensate for the employment loss. Moreover, a high degree of complementarity between host country activities and home country activities reduces the employment loss in the home country when the firm expands activities in the host country, and, if the wage costs in the host country are sufficiently low, employment in the home country may actually increase. The higher the degree of substitution between activities in the host country and the home country, the more likely it is that the trade union and the firm disagree whether the firm should become multinational.