The Centralization-Decentralization Issue

Aggregate Spending

With a very small Commission budget, the EU does not have the main attributes of a federal state. The Commission does not levy its resources; instead it depends on member states to decide how much to transfer from their won tax revenues. As Figure 1 shows, with expenditures amounting to about 2% of total public expenditures (general governments and the Commission), the EU is not anywhere near other federal states.

Figure 1 Size of central governments (% of total public expenditures)

Less than half of the Commission's budget is used for collective spending, mostly the administrative costs of the EU institutions. As noted in Section 4.2, the remaining is used for transfers, not for direct spending. The main exception is scientific research, which is administered and funded by the Commission, and represents about 10% of total spending.

There is a well-understood explanation for the situation. Democracy was born when elected bodies were given the final say on taxation and public spending. European citizens still consider fiscal policy as a key attribute of the State. As long as they consider that the State is fundamentally national, fiscal policy will remain wedded to national sovereignty.

From a fiscal federalism perspective, the central budget is remarkably small. It is as if there are almost no policies that are characterized by important externalities and/or returns to scale. Yet, it is not difficult to list a few policies that would qualify: defense, including border protection, diplomacy, or intelligence. These policy areas are indeed mostly federal in federal states such as the US, Canada, Switzerland or Australia, along with health and social policies. Clearly, defense and diplomacy are seen universally as regal functions and there is little prospect for the Commission to take over beyond the rump diplomatic service already in place. Health and social policies are taken up in the next section. The conclusion is that, at this stage of limited political integration, the EU's center spending power is unlikely to grow.

Policy Allocation

Much of the fiscal federalism literature is devoted to determining the level of government at which each spending catgeory is better located. As noted earlier, this involves balancing various criteria: externalities and increasing returns vs. information asymmetries; additional considerations include possible policy failures (including capture by interest groups) at different levels of governments.

Table 1 shows how public functions are allocated in a number of European countries. The diversity of arrangements stands out, showing how the various trade-offs are dealt with. More surprisingly, perhaps, the distinction between federal and unitary states does not really help interpreting actual choices. Education is the only unambiguous spending category that is clearly decentralized in federal systems, and even in many unitary systems. Housing and public order (which presumably includes police and justice) are often decentralized. These categories all involve strong local preferences, information asymmetries and relatively limited externalities. Income disparities often lead to transfers, either vertical or horizontal.

In the EU, all the categories listed in Table 1 are fully decentralized to the national level, often to the subcentral level. Given the diversity of practices in member countries, there is little scope for the centralization of any of these functions to the EU level. The ambiguity of theoretical prescriptions, the diversity of preferences and the subsidiarity principle, imply that most public spending is bound to remain highly decentralized.

Table 1 Share of subcentral government spending (% of total government spending)

Still, welfare spending has recently been identified as a candidate for some degree of EU centralization in a number of proposals. For instance, Claeys et al. (2014), have suggested that the EU should adopt a common system of partially centralized unemployment benefits, as initially foreseen in the MacDougall Report (1977). This proposal is based on the need for countercyclical horizontal transfers, which act as mutual insurance. Mutual insurance is justified by returns to scale and the existence of an externalities among highly integrated regions (and countries in the EU). Indeed, most federal states rely on the central government for this task. Currently, the EU lacks a central budget that could act as a counter-cyclical instrument. The externality is most clearly felt in the Eurozone where individual member countries have abandoned the monetary policy tool. Their only remaining counter-cyclical instrument is fiscal policy, which is restrained by the Stability and Growth Pact (SGP). Furthermore, the Sovereign Debt crisis has shown the deleterious effect of the existing arrangement. The need for some collective countercyclical policy is indeed an additional argument in favor of centralization in the Eurozone.

The usual counter-arguments emphasize the diversity of unemployment benefits in place. This is why the proposals typically only include a minimum of Eurozone-wide benefits, to be topped up in each country according to existing arrangements. Another aspect concerns eligibility criteria, which also vary considerably. The solution would be to adopt the minimum common eligibility criterion, complemented as needed at the national level.

However, such transfers raise a moral hazard issue. Given wide differences in structural unemployment, care must be taken to avoid permanent transfers from low to high unemployment countries. Differences in structural unemployment rates reflect deep local preferences, often the result of policy failures, which are even more unlikely to be erased if the failures are rewarded through transfers. The natural solution is to focus on deviations of actual from structural unemployment. This is a technically complex approach, however, open to manipulations and controversies. Indeed, part of the reason for these divergences lies in unemployment benefits eligibility criteria, a deeply political issue.

An alternative solution, partial sharing of unemployment benefits, may be politically appealing, but it combines both moral hazard and weaker counter-cyclical policies at the central Eurozone level. The more limited the scheme, the less worrisome is the moral hazard issue but the less effective is the scheme. The trade-off is obvious, and challenging.

At this stage, it is worthwhile to specify what failures are to be treated and which policy responses are justified. Cycles are largely driven by price and wage rigidities, hence the unassailable need for counter-cyclical policies. If these rigidities vary from country to country, centralization is not directly justified. However, in addition to the existence of externalities and returns to scale, noted above, the case for centralization can invoke the idea that limits on the use of national fiscal policies are justified by the need to establish fiscal discipline but, as discussed in Section 5, the proper response is not a centralized solution like the SGP. A further reason for limiting national fiscal policies is the current existence of large public debts, which limit the policy space in a number of countries. Section 5 argues that a temporary situation – large debts – should not shape permanent institutions. Once these two policy failures – improper fiscal discipline framework and large inherited public debts - are taken care of, national governments will be able to fully recover the ability to conduct counter-cyclical fiscal policies, thus avoiding the highly uncomfortable choice of a partially centralized unemployment benefits system.