XREAP2007-14
The empirical analysis of sustainability of different federal fiscal systems is rather scant.3 Apart from
data problems, the main reason is that disentangling the effect of different fiscal systems on aggregate
fiscal sustainability is fraught with two major difficulties (Bordignon, 2006). First, most fiscal federal
systems are not primarily designed for allocating resources to different tiers of government. Above all,
devolution of political power reflects cultural and economic heterogeneity. Hence, fiscal federal
systems mainly aim at redistributing resources between rich and poor regions. Redistributive and
purely allocative transfers are intertwined and hard to disentangle. Second, the identification of a
sustainability problem at regional level is hampered by the strategic behaviour of both central and sub-
national tiers of government. The basic problem is that we may not observe debt problems in the data
at regional level, even if there are soft budget constraints. The absence of bail out does not imply that
the fiscal system is sustainable.4 The central government can set up the fiscal system to avoid the
direct recourse to the central budget of regions in fiscal trouble. Basically, the federal and regional
governments may anticipate fiscal problems with additional transfers. At the same time, regional
governments may anticipate future adjustments in grants of other governments. Expectations of bail
outs are what matter for sustainability (Bordignon, 2006). These expectations are hard to identify in
the fiscal data we observe.5
Most studies of federal fiscal systems focus on cases of regional default and bail out (Rodden et al.,
2003; Singh and Plekhanov, 2005). However, the identification problem occurs because the focus is
on regional public finances. The consequences for the sustainability of public finances are hard to
detect as we need to uncover the complete incentive structure in the intergovernmental relationships.
However, we can avoid making assumptions on how the fiscal system affects the expectations of bail
out by comparing the evolution of general government deficits and debt to that of the central and
regional government. As additional transfers have to be financed by at least one tier of government,
we can use the aggregate deficit and debt position as an indicator of sustainability. A fiscal system is
then characterised by a soft budget constraint if at least one tier of government does not need to face
the consequences of the creation of public debt and the sustainability of its own public finances.
To that end, we recast a test for debt sustainability in terms of a fiscal policy rule à la Bohn (1998) so
as to account for the interaction between various tiers of government. Fiscal policy is deemed
sustainable when the government obeys to the intertemporal budget constraint. I.e., the sum of the
present discounted value of expected future primary surpluses suffices to pay off current debt. A
3 Darby et al. (2005) stress the importance of fiscal adjustment across all government levels for achieving an
economically successful consolidation. Büttner and Wildasin (2006) are closer to our model, as they examine
how US municipalities adjust spending or taxes to achieve sustainable policies in a panel VEC model.
4 However, a bail out is not necessarily an indication of unsustainable fiscal policies.
5 Bordignon and Turati (2005) isolate these expectations for regional Italian health expenses using EMU entry as
a natural experiment. Heppke-Falk and Wolff (2007) identify moral hazard of investors in the German regional
bond market. Dahlberg and Petterson (2003) test expectations of bailouts for Swedish municipalities.