expenditures conducted at the sub-central tier alter little during consolidation attempts but if
we differentiate between successful and failed attempts the share of general government
expenditure assigned to sub-central authorities is 19% higher on average after a successful
attempt and some 5% lower after failed attempts. Again, this suggests that skewness of
expenditure cuts toward the sub-central tier is not conducive to success. The sub-central share
of general revenues falls by some 18% on average during a consolidation attempt but this
decline is more pronounced in successful (22%) as opposed to failed attempts.
4.3 The composition of adjustment and the extent of decentralisation
As highlighted in section 2.2, the extent to which spending responsibilities are decentralised
differs markedly across countries. In table 5 we split the countries into two groups, the most
and the least decentralised18, we then look for differences in their behaviour of these groups
during consolidation attempts.
Table 5: Expenditure and revenue changes during general govt. consolidation attempts
(each shown as % of GDP)
|
Total Expenditure |
Total Revenue | |
|
Central Sub-central |
Least Most Decentralised Decentralised signif n=33 n=28 -0.18 -1.01 *** -0.27 -0.22 |
Least Most Decentralised Decentralised signif 1.03 0.43 ** 0.23 0.06 |
It is now apparent that there are significant differences in the behaviour of the central
government tier across these two groups. The most decentralised countries rely less on
expenditure cuts, and more on revenue hikes, than their less decentralised counterparts. This
result might be expected given the central tier’s relatively smaller expenditure responsibilities,
although it is interesting to note that a similar pattern does not emerge at the sub-central level
where the average changes in expenditure are insignificantly different from one another. We
can tentatively interpret these results as consistent with binding constraints on the adjustment
choices of the central tiers - the greater the extent to which spending responsibilities are
decentralised the more difficult it is likely to be to achieve co-ordinated cuts. Instead we see a
bias toward reliance on the instrument that the central tier can control most readily i.e.
centrally determined revenues. Since we have already shown that adjustment that focuses on
18 Countries classified as most decentralised (>35% of expenditure conducted by the sub-central tiers) include
Norway, Sweden, Finland, Australia, Germany the USA and Canada. The least decentralised group includes
Belgium, France, the Netherlands, Ireland, the UK, Spain and Austria.
15