monetary union pays transfers to the regional governments or the households in the individual
regions (Kletzer and von Hagen, 2001). Transfers to regional governments directly affect the
demand for goods and services in regional markets. Transfers to households do so only
indirectly, as they operate through the households’ budget constraint and distort their choices
between current and future consumption one the one hand and between consumption and
leisure on the other hand. The resulting effects on employment and savings may destabilize
regional employment even if they reduce the impact of asymmetric shocks on regional
consumption (Evers, 2006).
As it turns out, in a more general macro framework, the optimal design of a fiscal insurance
mechanism depends crucially on the type of shock hitting the economies. It is now well
understood that, in the context of dynamic general equilibrium models and in the presence of
aggregate productivity shocks, output stabilization is not an efficient policy.16 The reason is
that such shocks move the economy’s efficient (flexible-price) equilibrium level of output.
They should be accommodated, since households want to adjust their levels of consumption
and investment accordingly (Rotemberg and Woodford, 1997). Carrying over this insight to
asymmetric productivity shocks implies that full insurance against such shocks is undesirable
in a monetary union. Pure relative demand shocks of the kind considered by Mundell (1961)
can be offset completely by transfers paid between regional governments, provided that the
governments use these transfers to finance the purchase of goods and services in the local
markets. In the case of productivity shocks, however, neither transfers between regional
governments nor transfers between private households alone are sufficient to achieve optimal
insurance. A combination of both is required to stabilize consumption and employment
(Evers, 2006).
As argued above, the scope for fiscal insurance among the participants of the European
Monetary Union depends on the correlation of income and employment fluctuations among
the states and regions of the union. Recent empirical work that has investigated the correlation
of business cycles in the EMU sheds some light on this issue. Traistaru and von Hagen (2004)
find that the correlation between country-specific and the euro-area business cycles is positive
for all EMU member countries. Correlation coefficients vary between 0.30 and 0.50. Montoya
and de Haan (2007) consider NUTS-1 regions in the EMU. They find that the average
correlation between regional business cycles and the euro-area business cycle is above 0.60
and has been growing over the past decade. This is consistent with earlier results by Artis and
Zhang (1997) and Fatas (1997) who find that business cycles became more correlated among
the member states of the ERM during the 1980s and 1990s. Overall, this literature suggests
that monetary integration has lead to more strongly correlated business cycles without
eliminating the scope for fiscal insurance altogether. Traistaru and von Hagen (2004) show
that the business cycle correlations between the new EU member states and the euro area have
increased but remain much weaker than the correlations among the EU-15 states. Future
enlargements of the euro area by Central and East European countries will, therefore, increase
the scope and desirability of a fiscal insurance mechanism.
4.2. Political Economy Considerations
For a fiscal insurance mechanism in the euro area, additional considerations arise (Hammond
and von Hagen, 1998). Since the euro area is not fully politically integrated, the political
acceptability of a fiscal insurance mechanism is an important constraint on its design. A first
point is that the mechanism must be fully automatic and tied to a formula determining the
transfers. If the latter were left to the discretion of the governments, they would soon become
politicized, e.g., by paying transfers to governments facing reelections. Since a fiscal transfer
mechanism would create opportunities for politicians to spend monies raised from tax payers
in other countries, it would set up a classical fiscal commons problem, allowing politicians to
16 See Canzoneri (2006) for a summary of the relevant discussion.
13