دامپینگ اجتماعی و جابجایی: آیا موردی برای اعمال یک بند اجتماعی وجود دارد؟

Public opinion in Europe seems worried about the relocation of production plants toward low wage countries often accused of practicing ‘social dumping’. To reduce the incentives for relocation trade unions proposed the adoption of ‘social clauses’ protecting domestic markets from commodities produced in countries where minimal labor condition are not met. We analyze the effects of the adoption of a social clause in a vertically differentiated Bertrand duopoly. We assess how such a policy affects firms’ relocation decisions in order to be able to assess its welfare implications. We also characterize the optimal social clause policy, both under domestic welfare maximization, and from an efficiency point of view. While we show that a social clause policy cannot be dismissed on domestic (or world) welfare grounds, its case is weaker the higher is the domestic wage and the lower is the foreign wage.

It is well known that free trade and financial liberalization induce structural changes which have important redistributive effects. It is thus natural that the groups which suffer from such policies lobby for measures aiming at mitigating their losses. A striking example of this is the protest of the labor (and environmental) organizations against ‘globalization’ during the November 1999 WTO meeting in Seattle. On that occasion, tens of thousand of trade unionists participated in protest rallies and marches organized by the AFL-CIO.1 According to the labor organization, ‘fundamentally flawed trade policies have resulted in the loss of hundreds of thousands of high-paying manufacturing jobs and [have] undermined domestic measures designed to protect human rights and the environment. Trade agreements have […] empowered multinational corporate giants while leaving workers and communities to fend for themselves in an increasingly bitter global competition for scarce jobs and investment.’2 Moreover, according to the AFL-CIO, the reduction of tariff and non-tariff barriers, together with the fall of transportation costs, has magnified the appeal of relocating production plants toward low-labor cost countries practicing ‘social dumping’. Under these circumstances, ‘the only winners are multinational corporations that […] move factories, jobs and investment capital at a moment’s notice to escape national laws, collective bargaining agreements and simple standards of human decency’. 3 The term ‘social dumping’, though very widely used nowadays, has not yet found a precise economic definition.4 This is not surprising since the very concept of ‘dumping’ is still quite controversial among economists.5 If it is true that the distinction between dumping and price discrimination hinges on the razor’s edge, this is even more so for the distinction between social dumping and comparative advantage.6 Barros and Cabral (2000) present a model in which peripheral countries may have an interest in subsidizing foreign firms to induce them to serve domestic demand by local production, and this in order to reduce unemployment. ‘Social dumping’ is there treated in the same spirit as fiscal competition among jurisdictions. In such a framework, the welfare effects of ‘social dumping’ – defined as the lowering of domestic wages to attract foreign investments – concern exclusively the countries that engage in the ‘social dumping’ game, and the workers are those who bear the burden of the race to the bottom. Another way of addressing the ‘social dumping’ issue, i.e. the ‘flight of capital to low-cost areas’ (Ehrenberg (1994), p. 12), is that of studying how the possibility of relocating production plants out of a country affects that country’s welfare. Under this perspective, it seems natural to define ‘social dumping’ as the decision of a home firm to serve the domestic market through a plant located in a foreign country, where workers’ protection does not meet home standards and labor costs are thus significantly lower. The aggravation of the unemployment problem in Europe, the increased income inequality in the US, together with the media coverage of the labor conditions in the sweatshops located in developing nations, made trade unions, public opinion, and politicians particularly sensitive to this aspect of ‘social dumping’. 7 An illustration of the position of European labor unions with respect to the threat of ‘unfair competition’8 arising from social dumping is the proposal to adopt a particular type of social clause in international trade agreements. Such a social clause would be different from the ones previously considered and adopted,9 consisting of ‘socio-degradable taxes on imports produced without respecting the ILO rules regarding human and workers’ rights. These taxes would be given back in the form of aid programs to the development of social and economic democracy’. 10 This view is similar to that of Ehrenberg (1994) that suggests that ‘The United States might also consider mechanisms that allow the Americans who benefit from free trade to partially bear the costs of improved standards […]. This might be done if the US partially pays for the cost by remitting back to the trading partner funds raised through general tax revenue or through a tax on traded goods’ (p. 95). It is worth noting that the spirit of such proposals has been encompassed in the EU regulation concerning the application of generalized tariff preferences (1995 to 1998), where it is stated that ‘special incentive arrangements in the form of additional preference may be granted to beneficiary countries […] that have adopted and actually apply domestic legal provisions incorporating the substance of the standards laid down in ILO conventions Nos 87 and 98…’.11 A social clause, by mitigating the effect of this ‘unfair competition’, would, at the same time, be an instrument to promote workers’ rights in many of the third world countries. This is the position of the General Secretary of the International Confederation of Free-Trade Unions: ‘A social clause in international trade agreements […] would certainly persuade reluctant governments to pay more attention to their international obligations and respect their workers’ right to freedom of association’ (The Economist, Sept. 25, 1994 (Letters)). In this paper, we analyze the effects of using a ‘social clause’ policy, in a Bertrand duopoly model with vertically differentiated products. By using a vertical differentiation model we are able to understand the technological pattern of location, and thus to infer which firms are the ones with greater incentives to relocate. A social clause can be thought of as the imposition, by the domestic authorities, of a minimal wage paid abroad by the domestic firms, in the case they decide to serve the domestic market while producing abroad profiting from lower labor costs.12 This minimal wage can be thought of either as a ‘socio-degradable’ tax, or as a rule preventing importation of goods produced without respecting minimal working conditions. In such a set-up,13 the geographical market structure depends upon the size of both the relocation cost and the social clause. By choosing certain levels for the social clause, domestic authorities may influence firms’ decisions and sustain a specific configuration.14 We first regard the social clause as a policy instrument aiming at domestic welfare maximization. Domestic welfare is here defined as the sum of domestic consumers’, producers’ and workers’ surpluses. As in Brander and Spencer (1987), the presence of unemployment is taken to be an economically important feature of the domestic country, and its welfare impact is taken into account through the workers’ surplus. Furthermore, assumptions similar to those in Brander and Spencer (1987) allow the workers’ surplus to be measured by the wage bill. We show that, depending upon wage differentials and the fixed costs of relocation, the level of the social clause maximizing domestic welfare can be either dissuasive, or effective or nil. When imposing a dissuasive social clause, the government chooses a minimal wage high enough to induce firms not to relocate. When imposing an effective social clause, the government chooses a level of the minimal wage that keeps one of the firms from relocating, while inducing the relocation of the other. Finally, when setting a zero level of the social clause, the government opts for no intervention. We also show that no intervention is more likely to maximize domestic welfare the higher the wage differential between the countries. The policy implication of these findings seems to be that no intervention is more likely to be recommended when relocation occurs toward countries with significantly lower labor costs. Nonetheless, the possible case for safeguarding labor conditions of foreign workers (effective clause) cannot necessarily be dismissed on the grounds of this model. In the last part of this paper, we study which levels of the social clause would maximize world welfare (efficient outcome). Our main findings concerning this issue can be summarized as follows: also with respect to the efficient outcome, no intervention is more likely to be the optimal policy the higher the wage differential. When comparing the domestic welfare maximizing policy with the efficient policy, we find that the cases for no intervention coincide, while the cases for imposing a dissuasive social clause are less likely to occur when the efficient policy is implemented. In Section 2, we present the model underlying each firm’s location decision. Section 3 is devoted to the study of a social clause aiming at domestic welfare maximization, while in Section 4 world welfare is taken as the objective. Finally, in Section 5 we provide concluding comments. In the paper, we have chosen to present the results in an intuitive way, and we refer the reader to the Appendix for the more formal arguments and the proofs of the propositions.

The increasing magnitude taken by the phenomenon of relocation in the recent years has created a growing concern with its consequences on labor market. It is clear that the imposition of a social clause would tend to protect workers’ revenues in high wage countries28 by either impeding firms’ relocation or, at least, by reducing the number of national jobs lost. Moreover, under the assumptions of our model, the imposition of certain levels of social clause also benefits foreign workers (when the social clause is effective rather than dissuasive). Does this mean that we should expect this form of protection to be supported by economic agents and institutions other than international labor organizations? Even if both domestic and foreign workers may benefit from the imposition of a social clause, the case for such a measure cannot be made without considering its impact on other economic agents’ welfare. In particular, we have shown that consumers lose from such a policy while firms can either gain or lose. A first step toward ascertaining the overall effect of the social clause policy is to look at its effect on the sum of agents surpluses. We have done this both at domestic and at world level and identified efficiency with world surplus maximization. Though the main conclusion of our analysis is that a social clause policy cannot be dismissed on efficiency grounds, we have shown that the absence of a social clause is more likely to be optimal (both on domestic and world welfare grounds) towards countries with significantly lower labor costs, the natural candidates for such a policy. Results presenting a similar flavor can be found in Drèze and Sneessens (1994). In a simple competitive general equilibrium framework, where transfers take the form of unemployment benefits, they show that a protectionist policy can be defended on domestic welfare grounds only when labor conditions do not differ too much between countries. Our result sheds some doubts on the case for a social clause put forward by trade unions. In fact, such a policy seems more appealing towards countries presenting relatively small labor cost differentials, while it can be hardly defended towards the countries which trade unions would like to see it imposed on. Of course, this does not imply that actions aiming at imposing the respect of workers rights cannot be defended as such or that the use of a social clause cannot be justified on purely moral grounds.29 Clearly, the fact that we used stylized assumptions and worked in a partial equilibrium set-up may have led us to misperceive or ignore some effects. What are these effects, and how would they affect the welfare implications of a social clause? There are at least three issues that deserve to be discussed. First, our model is characterized by a fixed total demand and one might argue that a more realistic description would allow for a demand expansion induced by the lower costs associated with the firms’ relocation. By not having allowed this effect, we somehow underestimated the positive welfare impact of relocation, and thus left more room for defending the imposition of a social clause. Second, by taking domestic workers’ surplus to be the whole wage bill, we may have over-estimated the negative impact of firms’ relocation. Again, by doing so we created a more favorable situation for the imposition of a social clause. On the other hand, in studying world surplus maximization we assumed that only the rent from the social clause enters the foreign workers’ surplus. Had we assumed the total wage bill as the rent, we would have expanded the case for a social clause. Third, because of the partial equilibrium analysis, we ignored two types of income effects: one the one hand, the income effects on other sectors’ domestic demand; and on the other those on foreign demand, possibly including the potential expansion of the domestic export sector induced by this foreign income effect. Were the effects of increased consumption taken into account the gains from relocation would probably be greater, and the scope for the imposition of a social clause would be reduced.