Federalism fulfils two major functions:
a) A vertical separation of power by a division of responsibilities between two levels of government. The component units as well as the federation are usually geographically defined, although `societal federalism'8 considers non-territorial units as components of a federation.
b) The integration of heterogeneous societies, while preserving their cultural and/or political autonomy (et pluribus unum).
Both functions imply that the component units and the federation have
autonomous decision powers which they can exercise independently from each
other. Thus, sovereignty is shared or divided, rather than exclusively located
at one level. By no means do we suggest that the EU is, or should become, a
federal state. But even without the legitimate monopoly of coercive force, the
European Union has acquired some fundamental federal qualities. The EU
possesses sovereignty rights in a wide variety of policy sectors reaching from
exclusive jurisdiction in the area of Economic and Monetary Union to
far-reaching regulatory competences in sectors such as transport, energy,
environment, consumer protection, health and social security and, increasingly
penetrating even the core of traditional state responsibilities such as
internal security (Schengen, Europol) and, albeit to a lesser extent, foreign
and security policy.9 In
most policy areas, Community law is not only superior to national law, it can
also deploy direct-effect giving citizens the right to litigate against their
states for violating their rights conferred to them by Community law. This is
part of a second development, which has been addressed more recently. The
European Union is transforming itself into a political community within a
defined territory and with its own citizens, who are granted (some) fundamental
rights by the European Treaties and the jurisdiction of the European Court of
Justice. The European Community was conceptualised as a primarily functionally
defined organisation of economic integration (Zweckverband funktionaler
fixed territorial boundaries and no direct relationship between its
institutions and the European citizens. With the Treaties of Maastricht and
Amsterdam, however, the Single Market has been embedded in a political union
with emerging external boundaries11 and a proper citizenship.
Not only has the EU developed into a political community with comprehensive regulatory powers and a proper mechanism of territorially defined exclusion and inclusion (Union citizenship). It shares most features of what the literature defines as a federation:12
a) The EU is a system of governance which has at least two orders of government, each existing under its own right and exercises direct influence on the people.
b) The European Treaties allocate jurisdiction and resources to these two main orders of government.
c) There are provisions for `shared government' in areas where the jurisdiction of the EU and the Member States overlap.
d) Community law enjoys supremacy over national law, it is the law of the land (Bundesrecht bricht Landesrecht).
e) European legislation is increasingly made by majority decision obliging individual Member States against their will.
f) At the same time, the composition and procedures of the European institutions are based not solely on principles of majoritarian representation, but guarantee the representation of `minority' views.
g) The European Court of Justice serves as an umpire to adjudicate conflicts between the European institutions and the Member States.
h) Finally, the EU has a directly elected parliament (since 1979).
The EU only lacks two significant features of a federation. First, the
Member States remain the `masters' of the treaties, i.e., they have the
exclusive power to amend or change the constitutive treaties of the EU. Second,
the EU lacks a real `tax and spend' capacity, in other words, there is no
fiscal federalism. Otherwise, however, the European Union today looks like a
federal system, it works in a similar manner to a federal system, so why not
call it an emerging federation?
If we accept that the European Union has been developing into a federal system where formal and material sovereignty is divided and shared, federalism offers different alternatives to organise the distribution of power vertically, i.e., between the European Union and the Member States, and horizontally, between the executive and legislature. In principle, we can distinguish two federal models, which differ according to the distribution of competences between the two levels (shared versus divided), the representation of the states at the federal level (strong versus weak), and the fiscal system (joint versus separate).
Co-operative or intra-state federalism, of which Germany is almost a prototype, is based on a functional division of labour between the different levels of government. While the federation makes the laws, the states are responsible for implementing them. The vast majority of competences are concurrent or shared. This functional division of labour requires a strong representation of the states at the federal level, not only to grant an efficient implementation of federal policies, but also to prevent the states from being reduced to mere administrative units. The reduced capacity for the self-determination of the states is compensated by their strong participation in federal decision-making through the second chamber of the national legislature. Major policy initiatives always require the consent of the Federation and the majority of the states. The chamber of territorial representation is organised according to the Bundesrat (Federal Council) principle: the states are represented by their governments in relation to their population, with smaller states usually enjoying over-representation. Representation is not only disproportionate but also indirect. The sharing of policy competences in terms of a functional distribution of labour is backed by a sharing of tax revenue in a joint tax system, which is usually complemented by financial equalisation. Federation and states share the most important taxes. The allocation of joint tax revenue also allows for a redistribution of financial resources between states with stronger and weaker spending power. The functional and fiscal interdependence of the two levels of government not only gives rise to a co-operative federalism, interlocking politics and joint decision-making, it also favours the emergence of a policy-making system in which policies are formulated and implemented by the administrations at both levels of government. Such executive federalism is counterbalanced by a strong and vertically integrated party system, which provides for an effective representation of non-territorial (functional) interests at federal level.
Dual or inter-state federalism to which the US most closely corresponds, emphasises the institutional autonomy of the different levels of government, aiming at a clear vertical separation of powers (checks and balances). Each level should have an autonomous sphere of responsibilities. Competences are allocated according to policy sectors rather than policy functions. For each sector, one of the two levels of government has both legislative and executive powers. As a consequence, the entire machinery of government tends to be duplicated because each level should manage its own affairs autonomously. The sectoral or dual allocation of policy competences is complemented by a rather weak representation of the states at federal level. The second chamber of the federal legislature is usually organised according to the `Senate Principle'; the states are represented by an equal number of directly elected Senators, irrespective of their size and population. As a result, the Senate does not represent the interests of state executives, as the Bundesrat does, but the interests of the electorate or the parties within these states. The states articulate their interests through voluntary co-operation with the central state, usually in intergovernmental conferences. Consequently, there is no need for a strong, vertically integrated party system promoting functional interest representation in order to counterbalance executive dominance. The institutional autonomy of each level of government, in the final analysis, presupposes a fiscal system which grants the states sufficient resources to exercise their competences without the financial intervention of the central state. This should be ensured by a comprehensive fiscal autonomy of the states which allows them to levy their own taxes in order to have an independent source of revenue.
Which of the two models appears most appropriate for a European federation? Fischer's own proposal, while most recently leaning toward the US model, remains, nevertheless, ambiguous as he has not made up his mind how best to preserve the strong role of the Member States in a European federation: either by granting them a strong representation at European level (German model) or by providing the Member States with a strong autonomous sphere of competences (US model). He cannot have it both ways. If Fischer wants strong representation of Member State interests at European level, he will have to opt for the German model-i.e., the executives of the Member States must be represented (the Bundesrat model)-on the one hand, and sovereignty rights will have to be shared rather than divided, on the other. The Member States would have a veto on any major decision and would also be responsible for the implementation of European policies. The comprehensive legislative powers of the European federation, albeit shared with the Member States, would have to be matched by an corresponding tax and spending capacity at European level.
A `senate' type concept whereby the members of the second chamber of a future European parliament are drawn from the national parliaments (Fischer's speech in Strasbourg) provides only a weak representation of territorial interests at European level. As the US Senate provides ample evidence, the senators tend to represent functional and constituency interests rather than territorially defined concerns. It also follows that such a model has to be built on the division of sovereignty rather than on the concept of shared sovereignty, in order to avoid a far too centralised federal state. The EU would need to dispose of legislative and executive competences, which it could exercise independently of the Member State governments. Furthermore, independent legislative and executives responsibilities would have to be accompanied by a minimum degree of taxation and spending autonomy for the European government, if the European federation is not become a mere fig-leaf, veiling a return to the Europe of the nation-states.
Which model, therefore, is the most realistic for a European federation? First, given the current distribution of power, whereby the EU and the Member States share most of the policy competences, the German model of co-operative federalism appears to be most feasible. With the exception of monetary union, the EU cannot legislate without the consent of the Member States, even in the area of its exclusive competences such as foreign trade. There are hardly any areas in which the Member States completely ceded sovereignty to European level and do not directly participate in decision-making.
Second, the European Council and the Council of the European Union could be easily transformed into a Bundesrat-type second chamber of the European Parliament, while the Commission would become the European government (with or without a directly elected European president). One can still think about whether the members of the first parliamentary chamber should also be members of the Member State parliaments. Finally, the German and European federal systems share a consensus-oriented political culture which helps to prevent political stalemate and allows the smaller members to have a fair chance of being heard, even if their voting power is curbed, which seems to be unavoidable given the prospect of EU enlargement.
But the European Union lacks one important feature of the German federation, which is unlikely to be replicable at European level. German co-operative federalism corresponds to a clear political preference for equal living conditions enshrined in the German Constitution and widely shared by German society. Instead of preserving and accommodating socio-economic and cultural plurality, the post-war German federal system was to provide similar living conditions for all German citizens, irrespective of the state they lived in. Due to the largely divergent spending capacity of a big state like Bavaria compared to a small city state like Bremen, legislative and fiscal competences have been increasingly transferred to federal level. The German states were compensated by means of the strong participatory rights of the Bundesrat in federal decision-making. While a further transfer of policy competences to European level does not seem completely unrealistic, the real issue is the weakness of the `taxation and spending' power of the European Union. Its redistributive capacity is currently limited to 1.27% of the GDP generated by all Member States, whereas the redistributive capacity of the individual Member States amounts to approximately 50% of their GDP. In Germany, the federation receives about 40% of the overall tax revenue. Thus, a comparable spending power for the European federation would correspond to a share of about 20% of the European GDP. An almost twentyfold increase in the EU's spending power might strengthen the legitimacy and effectiveness of European governance, but it is inconceivable that the Member States would agree to such an enormous decline in their revenues.
The US model of dual federalism, in turn, would allow for a weaker European federation. It is grounded in a deep suspicion of a strong central state and, hence, resonates with the French and British distrust of what they perceive of as an emerging European federal state and with the corresponding claims for a strict application of the principle of subsidiarity. The restriction of European jurisdiction to a clearly defined area would also leave the Member States their autonomous taxation powers. A directly elected European president and a stronger European Parliament would significantly increase the legitimacy of the European federation. Finally, as state executive interests are less dominant at European level than in the German model, a vertically integrated party system, which is still missing in the EU, is of lesser importance. Yet, the introduction of the American model of federalism may be even more demanding than the German model.
First, divided sovereignty would require that most Europeanised legislation be dis-entangled in the areas for which the EU would have to hold exclusive competences as opposed to those in which the Member States are solely responsible. This is an almost impossible task given that the current EU is based on shared competences. It is also likely to meet with resistance from smaller Member States with low institutional and economic capacities. Second, the Member States would have to give up their strong representation at European level in order to grant the European federation independence in exercising its, by then, considerably curbed competences. The European Council and the Council of the European Union would be replaced by a senate representing the citizens rather than the governments of the individual Member States. The European Commission with a directly elected president would become a truly federal bureaucracy, which then, however, would have to be considerably strengthened (including field services in the Member States) in order to execute European policies effectively. Finally, given the strong, and with enlargement even increasing, socio-economic heterogeneity of the Member States, the European federation would need a minimum of redistributive capacity. The example of the American federation which started off with hardly any `taxation and spending' capacity is rather instructive.
The European Union seems to be stuck between a rock and a hard place. None of the two federal models are without flaws. The introduction of either requires profound institutional reforms, which are far more demanding than a redistribution of voting powers in the Council or the reallocation of national slots in the Commission, as are currently being considered in the Intergovernmental Conference. Yet, as Fischer correctly points out, such reforms are indispensable if European governance is supposed to remain effective in the light of enlargement.
8 Hueglin (1991).
9 von Bogdandy (1999).
10 Ipsen (1972:196).
11 Article 11 of the European Union Treaty refers to the protection of the integrity of the Union and its external boundaries.
12 Burgess (2000); Wheare (1963).