This comparative research project on the present and future of the social federations in Germany, the United States, and the European Union (EU) emanates from the GHI's cooperative program on "Continuity, Change, and Globalization in Postwar Germany and the United States" (see Bulletin no. 24, Spring 1999). Its main goal is to find ways for the EU to become a social federation that safeguards distinct social welfare traditions in its member states while helping them to meet the common challenges from closer market integration. The analysis focuses on social assistance, including unemployment compensation, because this is where the political and economic case for the EU to become involved is most compelling. Both Germany and the United States provide foils for the present study in that they are polar extremes of federal social welfare systems. The German system is characterized by close cooperation among the different levels of government. This cooperation has become increasingly strained since unification, supposedly because Germany has become a more heterogeneous socioeconomic space. In contrast, the American system fosters fiscal competition among the federal government, states, and municipalities due to overlapping jurisdictions. The Welfare Reform Act of 1996, which implemented the so-called New Federalism, provides a rich arena in which to study an extremely decentralized system of social welfare provision.1
It may seem counterintuitive to suggest a role for the EU in social policy. After all, there already exists a crisis of confidence in the European "social model." The usual suspects in this crisis are globalization, demographic change, and long-term unemployment. And even without such doubts on how well (continental) Europe is able to cope with these challenges, the very basis for legitimate social policy in democracies - solidarity - is even more fragile at the European level than at that of the member states. The EU administration has little experience in policy formulation and implementation. National governments are reluctant to give Brussels a say out of fear that this will lead to either upward or downward harmonization of social policies. Finally, the fiscal hands of the various EU governments are presently tied in their collective attempt to support the establishment of the Euro as a hard currency. This situation leaves little room for major social policy initiatives, whether at the national or the European level.
Against these odds, why should the EU want a social federation at all? First, the very formation of the EU implies new risks to employment and income generation that can only be assured at the EU level. Second, and at the same time, European integration places pressure on a variety of national traditions and preferences with regard to social policy, which may lead to weaker and less diversified safety nets. In other words, the EU seems to be in need of a social federation because it is itself the source of risk and because the very process of EU integration tends to weaken the national systems that carry them.
Where do the new risks come from? The integration of asset, labor, and commodity markets as well as a common monetary policy create a common price dynamic (inflation or deflation) and lead to further synchronization of business cycles. Insurance against such risks is the classical realm of automatic or built-in stabilizers, that is, the countercyclical behavior of public net spending built into the tax and spending system. Social assistance and unemployment insurance are prime examples of such automatic stabilizers: Net spending on these social policies rises in recessions, whereas they decrease in boom times. It is one of the characteristics of modern social welfare, ever since it was invented during the New Deal, that it provides for such macroeconomic stabilization, especially preventing recessions from turning into depressions or periods of deflation. The insurance principle establishes that such stabilization is most effectively provided at the aggregate, that is, the EU level because that is where individual national risks can be pooled and thus redistributed.
Where do pressures for downward homogenization come from? First, there are the incentives from fiscal competition, that is, the fear of governments (and resident voters) that their state or locality may become a "welfare magnet." This need not be the result of actual migration. Historical case studies of the United States have shown that virtual migration, media reports about individual cases, may suffice to create that latent "angst" in the public mind that makes for the notorious rush to the bottom.2 Moreover, and apart from fiscal competition, the way fiscal relations are designed has an impact on public spending at lower levels of government. The Stability and Growth Pact, which sanctions excessive budget deficits, has made it imperative to cut spending for most EU governments. Social welfare is certainly one of the prime candidates because it accounts for a sizable share of the budgets of most EU governments.
Obviously, a resume of these considerations produces a dilemma. A role for the EU in social-policy matters is generally not accepted, but such a role is warranted nevertheless. In light of this dilemma, the idea of "laboratory federalism" seems to be worth pursuing. The term was recently forwarded by one of the pioneers of the theory of fiscal federalism in postwar economics.3 The EU could engage in setting up a framework for this kind of federalism, one that enables the different levels of government to engage in horizontal and vertical learning.
The U.S social welfare system after 1996 may have features of such a laboratory federalism, even if it was not consciously designed that way. For instance, the New Federalism is usually identified with administrative and operational devolution of responsibilities for social welfare provision. Yet, welfare finance has simultaneously become more centralized. Programs that are exclusively financed by the federal government, such as the Earned Income Tax Credit (EITC) and Food Stamps, play a more significant role in "welfare to work" or workfare. Moreover, the federal block grants to the states have been fixed at a historical peak level, whereas states must maintain only 75 percent of their historical spending levels. For the time being, both have led to an increase in the federal government's share of welfare spending. This counters several incentives to downsize spending on social welfare that resulted from devolution. Ironically, the counter tendency may well be a product of many state governments' drive to the bottom in social welfare spending, avoiding or at least shifting costs to other constituencies.
What are the lessons of this example for the EU? Above all, the role of the federal level in laboratory federalism is that of a promoter of horizontal and vertical learning processes in an area where there is a consensus that change is inevitable, yet hardly any consensus on what that change should be and how to implement it. More specifically, the European level should prevent the systemically destabilizing effects of social welfare reform at the national level. These effects, which would hurt the member states themselves in the long run, may result from each government competing for wealthy taxpayers by lowering income taxes while at the same time trying to deter marginally employed households by lowering benefits and restricting eligibility.
To a large extent laboratory federalism in the United States is about centralized funding of decentralized experimentation. Given the limited fiscal means of the EU Commission, however, one has to find functional equivalents. One currently fashionable idea is benchmarking: A common framework for comparing various reforms could provide a safeguard against premature harmonization while furthering a systematic exchange and evaluation of learning experiences. Another idea would provide EU grants for rigorous evaluation of reforms. This supplemental funding of research would require comparatively small means but may have a large impact on learning processes and reform momentum, as the historical experience of the United States in the 1980s and early 1990s once again amply shows.4
The first results of this research project will be presented at an upcoming workshop, "Learning from Diversity in Federal Systems: Social Assistance in the United States, Germany, and the European Union," to be held at the GHI on November 17-19, 2000.