The Centralization-Decentralization Issue
Taxation and direct transfers
Taxation
The principle of "no taxation without representation" sets a high hurdle for centralization. Until citizens recognize the European Parliament as representing their interests, there is little scope for EU taxation. How far are we? For the EU as a whole, the latest Eurobarometer survey reports that 37% of respondents trust the European Parliament, a low number, but trust in national parliaments is even lower at 28%. The situation varies considerably across EU member countries, as Figure 2 shows. The correlation between trust in the national and European Parliaments (0.58) suggests that citizens apply domestic lenses to look at European politics. Even so, its better performance probably does not provide the EU Parliament with sufficient support to claim a transfer of some taxing power. It stands to see its legitimacy further eroded the minute it acquires and exercises this power.
Figure 2 Trust in European and National Parliaments (%)
Transfers: A Proposal
Fiscal federalism principles argue in favor of matching taxation and spending, possibly along with transfers
intentionally designed to be redistributive. The reason is that one of the arguments in favor of decentralization is
proximity to citizens and easier internalization of externalities, including sharing the tax burden to finance local
public goods. Any imbalance between spending and revenues, which requires transfers across the federal units, breaks this logic. Such transfers are also often controversial and prone to pork barrel politics.
Here again, the EU lies at the extreme end of decentralization. Vertical transfers, from the EU to national
budgets represent about 0.5% of EU GDP. There are virtually no horizontal transfers, from state to state. As
Figure 3 shows, in most OECD countries transfers from the central governments are typically small, but not
negligible. The EU vertical transfers are mostly limited to the Cohesion Policy and the Common Agricultural
Policy. These policies are explicitly redistributive, and both are controversial. Controversy is quite natural
because redistribution is a zero-sum game and because it is unlikely to enhance growth, one of the official
objectives of the policy.
Figure 3 Gross transfers from central government (% of GDP)
Many reforms of the Cohesion Policy and of the Common Agricultural Policy have been attempted over the
years, resulting in changes that have been limited and disappointing to most countries. This is largely the result
of a zero-sum game combined with pork barrel politics. Sadly, as the main expenditures of the EU, these policies
have become a negative symbol. Because the costs and the benefits are unevenly balanced across countries,
doing away with these policies is politically impossible. A reform, therefore, would have to recognize that those
countries that tend to loose are properly compensated.
A solution is possible, however. It would shrink the program budgets to their net balances, country by country,
thereby freezing existing transfers. Ideally, these transfers would be subject to a sunset clause on a long enough
horizon to be acceptable by the governments currently in place. Governments that are currently net recipients
would direct the transfers to current beneficiaries, possibly topping up these amounts to protect, fully or partly,
the beneficiaries. Governments that are net contributors could redistribute the savings through other spending or
tax reductions. This proposal has four interesting features.
First, it decentralizes the associated subsidies to the national level, where taxpayers are better able to judge the
merits of these subsidies. In addition, decentralization breaks the confusion between subsidizing specific
activities and inter-country transfers. The logic behind supporting farmers is distinct from the logic behind
transfers from richer to poorer countries, an issue that successive reforms have failed to deal with effectively.
Second, it would remove a perennial irritant when the EU budget is discussed. The stakes of pork barrel politics
would be lessened.
Third, it would mechanical reduce sizably the Commission's budget. This would be an opportunity to thoroughly
rethink the resources and functions of the Commission, which never really happened since the Treaty of Rome.
If it were decided to keep the resources about unchanged, this would be a great opportunity to re-allocate them
towards policies that are explicitly justified on the ground of fiscal federalism principles. As noted above,
defense, scientific research and networks could qualify, as they are subject to important returns to scale.
Fourth, sunset clauses are an appealing way to deal with the legacy of existing arrangements, which were and are poorly justified, as noted in Tanzi and Schuknecht (1996) and Wyplosz (2006). This would create a useful precedent.
Tax Evasion
A by-product of globalization is global "tax optimization". The existence of the Single Market obviously makes
tax evasion even more attractive. Potential beneficiaries are firms that operate in several countries; they can use
internal prices to locate profits where taxation is lowest. More generally, corporations can use a host of features
of the Single Market to reduce taxes. Consumers too can arbitrage tax exposures through simple border or
internet shopping, but also through the right to establish wherever they want. The very existence of the EU is an
externality that leads to (mostly implicit) tax competition.
There is nothing new as federal states face the same situation, in fact a more complicated one as there is also
vertical tax competition. The recommended solution, to associate taxes and public goods and services, is de facto
implemented in the EU as nearly all taxes finance public goods and services delivered by member countries, in
addition to transfers. Transfers, however, are more troublesome. Since they represent about half of public outlays
in most EU countries, tax competition is a difficult issue. The solution cannot be to avoid taxing mobile sources.
This has led to much soul searching, still under way.
In addition to an agreement on VAT ceiling and floor, this has led to calls for tax harmonization. Taking to its
logical end, this externality would require a common tax structure throughout the EU. Against the externality
argument, there is a massive local preference argument. Taxation is deeply redistributive, therefore profoundly
political. National tax structures reflect both national preferences and histories. They represent a fragile
equilibrium that is very difficult to alter. The case for decentralization, therefore, is overwhelming. More modest proposals call for the harmonization of tax bases, leaving tax rates in national hands. Yet, local preferences
matter considerably when it comes to defining tax bases. The balance between the externality and local
preferences is very uneasy to determine. Subsidiarity would then suggest to not seek centralization.
This has led to the more modest project to harmonizing tax bases. In principle, this would allow for more
transparent comparisons of what member countries do. The intention seems to be to then use peer pressure to
very gradually move to some degree of tax harmonization. However, it is hard to see how such a process of
backdoor centralization could succeed. First, defining comparable tax bases is likely to be a challenging
endeavor. This is an area where details crucially matter. The intricacies of national tax systems have produced a
great many idiosyncrasies that will be difficult to meaningfully circumvent and impossible to eliminate. Second,
what to do the tax rate differences that will emerge? In the absence of any acceptable prescription on what is an
appropriate tax structure, peer pressure – already a weak process – is unlikely to produce much result. Taxation
is bound to remain fully decentralized