انگیزه های امنیت اجتماعی، سرمایه گذاری سرمایه انسانی و تحرک نیروی کار

Migration between countries with earnings-related and flat-rate pay-as-you-go social security systems may change human capital investments in both countries. The possibility of emigration boosts investments in human capital in the country with flat-rate benefits. Correspondingly, those expecting to migrate from the country with earnings-related benefits to a country with flat-rate benefits may reduce their investment in education. Allowing for migration may generate an intertemporal Pareto-improvement with cross-border transfers, and the contribution rates satisfying certain conditions. However, these conditions are not satisfied with those contribution rates that would arise if the governments maximize the welfare of their citizens without migration.

Pensions are the single largest government transfer program in the European Union. In 2001, the then 15 EU member states spent on average 8.8% of their GDP on public retirement benefits (OECD, 2003). Pension policies have traditionally been viewed as a domain of nation-states and, correspondingly, there are wide differences in how benefits are determined. In Continental Europe, like France, Germany, and Italy, pensions are viewed as postponed wage income and their aim is to smooth lifetime consumption. In such “Bismarckian” programs, retirement benefits are linked to past earnings. In the competing “Beveridgean” tradition, retirement benefits are used to protect the elderly against poverty. Benefits are then rather flat, with the link to past earnings weak or even non-existent. These rather flat-rate systems dominate, or at least play an important part, in public pensions in Denmark, Ireland, the Netherlands, and the United Kingdom (Disney, 2004). Both earnings-related and flat-rate pensions are mainly organized according to the pay-as-you-go (PAYG) principle, implying that the benefits of the current retirees are paid by current workers.1 Migration may endanger both earnings-related and flat-rate pension systems. With migration, the intragenerational redistribution of flat-rate systems may generate an adverse selection problem. Earnings-related systems, on the other hand, may benefit from the inflow of high-income contributors. The intergenerational redistribution component, on the other hand, poses a more severe challenge to earnings-related systems as these are larger.2 In 2001, public spending on retirement benefits was on average 6.4% of GDP in OECD countries with flat-rate benefits, and 9.4% in countries with earnings-related benefits (Disney, 2004 and OECD, 2003). Social security rules and migration possibilities also influence incentives to invest in human capital. An option to migrate to a country with a less redistributive social security system increases the expected private return to human capital, thus boosting such investments. When young people do not know beforehand how mobile they will be, the investments of those who finally remain change. Such uncertainty about one's own future mobility may result from uncertainties related to family formation, future partnership status, and on random events related to employment. This paper takes a dynamic view of economic integration and the challenges it poses to social security systems of different types. There are two countries, one with earnings-related benefits, and another with flat-rate benefits. The two countries may differ also in social security contribution rates, and both have organized their social security system on a PAYG-basis. At the starting point, there is no migration. Then labor becomes mobile, corresponding to tighter integration. Since production technologies are identical in both countries, migration does not affect the productivity of the human capital stock of the migrant. This paper asks three questions. The first one is how the possibility of migration affects incentives to invest in human capital and economic well-being in the two countries. The second question is what are the welfare effects of migration. The third and final question is whether both social security systems can be maintained after the labor markets have been integrated. The main findings are the following. Assume first that neither country has a social security system that would be preferred by all citizens. Then allowing for migration would increase the investment in human capital at the upper echelon of the productivity distribution in the country with flat-rate benefits, and decrease investment in human capital at the lower levels of productivity in the country with earnings-related benefits. Thus, allowing migration would increase human capital formation in the country with flat-rate benefits, and reduce human capital formation in the country with earnings-related benefits. The effects on the stock of human capital in both countries after migration would depend on the productivity distribution, and on the social security contribution rates. Allowing for migration may, but need not, generate an intertemporal Pareto-improvement. Assume next that migration goes in only one direction. Generations living in the stationary equilibrium in both countries may still gain whichever the direction of migration. The welfare effect on transition generations depends on the direction of migration, and on the relative contribution rates. If unilateral migration goes from the country with flat-rate benefits to the country with earnings-related benefits, and the latter has the same or a higher contribution rate, then cross-country transfers allow securing that all generations in both countries gain from allowing migration. However, if migration would tend to go from the country with earnings-related benefits to the country with flat-rate benefits, there is no cross-country transfer scheme that would allow for an intertemporal Pareto-improvement. Endogenizing the contribution rates that prevail in the absence of migration suggests that, in the political equilibrium, the migration would go from the country with earnings-related benefits to the country with flat-rate benefits, precluding an intertemporal Pareto-improvement. This paper is organized as follows. Section 2 reviews the literature. Section 3 develops the model. Section 4 presents the results concerning investment in human capital and migration. Section 5 analyzes welfare effects. Section 6 endogenizes contribution rates in the absence of migration, and presents the implications on the possibility of a Pareto-improvement when mobility is allowed. Section 7 concludes.

This paper derives some expected and some rather surprising results on the effects of allowing migration between countries with different social security systems. As expected, there is some cutoff productivity level above which citizens prefer earnings-related systems, and below which they would rather have flat-rate benefits. This cutoff level may also be associated with a corner solution, with all citizens preferring one of the competing systems. Moreover, the possibility of migration from a flat-rate system to an earnings-related system stimulates human capital formation. Conversely, those willing to migrate to a country with flat-rate benefits may reduce their investment in education. Allowing for free migration may, but need not, pose problems for either system. Perhaps the most surprising result is that allowing for migration could generate an intertemporal Pareto-improvement even in the absence of any productivity differences between the two countries and with unilateral migration. This requires that migration would go from the country with flat-rate benefits to the country with earnings-related benefits, and that the latter country would have either the same or higher contribution rate. Moreover, generating an intertemporal Pareto-improvement with unilateral migration, if feasible in the first place, would require that the destination country with earnings-related benefits transfer part of the tax revenue of immigrants to their country of origin, during the transition period. Even the possibility of such a Pareto-improvement would not arise in the absence of endogenous human capital formation, highlighting the importance of taking into account investments in education when analyzing the welfare effects of labor market integration. This paper focuses on the efficiency implications and welfare effects of migration between two countries with given social security contribution rates. This is only a first step in exploring the welfare effects of migration in democracies. An important topic for future research would be to endogenize contribution rates. This paper takes a first step in this direction, by solving for the political equilibrium contribution rates that the governments maximizing the weighted sum of the welfare of the elderly and the currently young generation would choose in the absence of migration. Already this suffices to prove that the potential Pareto-improvements identified may not materialize with contribution rates that arise from an endogenous political process. It remains for future work to identify the equilibrium contribution rates that would arise from a political process over the full transition path, as well as to evaluate empirically the magnitude of the different theoretical effects that this paper identifies.