the UK, Ireland, Finland and Norway. However, these countries are not uniformly associated
with low rates of success rates of consolidation attempts: Denmark and the UK record
relatively high success rates; but conversely Finland and Norway combine relatively large
number of discretionary adjustments with relatively low success rates, so only these two
countries within our sample accord with the Alesina and Perotti ‘stop-go’ hypothesis.
Finally, it’s worth examining the behaviour of the European countries in our sample in the
run-up to the launch of EMU. During this period the Maastricht criteria were seen by some as
having helped some of the more lax governments in providing a credible ‘political pre-
commitment’ to the objective of reducing structural budget deficits. We can offer limited
support to the success of this the forced discipline. Of the EMU countries included in our
sample, five made significant adjustments in the post-Maastricht period (Belgium 1994,
France 1997, the Netherlands 1993, Finland 1993 and Spain 1997); Germany and Ireland did
not make significant consolidation attempts during the 1990s, though at this stage they
already had deficit and debt ratios very close to, or within the thresholds set13.
We now turn to analysing the distinct roles of central and sub-central tiers of government in
fiscal consolidation attempts.
4: Analysing Consolidation Effort Across Tiers of Government
In what follows we start by looking at the discretionary impulses attributable to each tier of
government during a general government consolidation attempt and show that sub-central
tiers of government do play important roles in the consolidation process, contemporaneously
with the central tier14. We then look in more detail at the composition of the consolidation
effort, first by separating out the changes in expenditure and revenue and then by breaking
down these aggregates into their key components. In addition to examining consolidations
instigated by central governments we also look at those instigated by the sub-central tier. This
enables us to ask if the sub-central tier behave differently when consolidating alone.
4.1 Analysis of discretionary fiscal impulses by tier of government
Table 3 shows the average sizes of the discretionary fiscal impulse as a proportion of GDP,
across all the identified consolidation attempts (refer back to Table 1 for details), across
successful and failed attempts (denoted “S” and “F”) respectively, and disaggregated by tier
13 In the case of Austria, our sample ends in 1994 and hence we are unable to detect any discernible effects
attributable to the Maastricht criteria.
14 We have looked at whether central government adjustments lead sub-central ones, but the contemporaneous
fiscal impulses at sub-central level are almost uniformly greater. For a more comprehensive analysis of the timing
and duration of various changes see our companion paper, Darby et al. [2004], which employs event study
methodology.
11