significant, hence it is not obvious that this leads to a build up of public debt. In Germany
and Spain on the contrary, revenues do not change significantly, but government
expenditures shrink considerably, leading to large accumulated surpluses at a horizon of
10 years.
But whether the causality runs from fiscal policy to productivity growth, or vice versa, is
not obvious. Recall that the supply shock contains productivity shocks that may emanate
from the private as well as the public sector. The significant co-movement of spending
and revenues suggests that fiscal ‘supply’ shocks are an important source of the overall
productivity shock.91 If these relations are positive (the case of France and Portugal), this
implies higher spending or tax revenues have contributed to economic growth. In the
opposite case (Germany or Spain), a reduction of spending - and less so a lower tax
burden - would trigger higher potential output growth. But there are a few alternative
explanations. Positive economic shocks that enlarge the tax base would - for a given tax
rate - automatically lead to a larger revenue intake owing to automatic stabilisers. For
reasons of political economy, this could lead the government to directly spend the
proceeds of the treasury. This expansion of the budget could consequently get locked in
and lead to a permanent rise in government expenditure. This mechanism would work for
both permanent and cyclical shocks, if we assume that the government does not
systematically react in different ways to permanent or transitory economic shocks. This
allows us to get some insight in the importance of the private versus public productivity
shocks. The fiscal responses to cyclical shocks, which include business cycle shocks with
transitory output effects that are not related to fiscal policy, can give some indication.
Surprisingly, the effects of cyclical shocks on output are hardly significant and indicate
the small size of temporary economic fluctuations.92 As a consequence, there is not
always an obvious simultaneous rise in tax revenues. In Germany and Portugal,
government revenues do rise in response to a positive output gap, and this effect remains
permanent. Moreover, in both countries government expenditures tend to rise as well.
This gives some support for the ‘ratcheting up’ effect on spending. In France or Spain
instead, government spending does not react in a significant way and tax revenues even
tend to decline.
If we consider in some more detail the two countries in which catching-up phenomena
may be expected to be important, we cannot clearly distinguish between the two
alternative explanations. Both in Spain and Portugal is the reaction of fiscal variables to
temporary and permanent shocks similar. This downplays the importance of productive
fiscal policy contributing to economic growth. A comparison of the impulse responses
shows that only a minor effect would be left in the case of Portugal. Evidence on Spanish
public finances presents a slightly different picture. Positive supply shocks are
accompanied by a strong decrease in total spending, and this effect is much more
pronounced than the reduction in spending after a cyclical shock.
In France and Germany instead, the reaction of fiscal variables to permanent shocks is
opposite to the reaction to business cycle shocks and supports the view that fiscal
variables driving long-term growth in both countries. That spending and revenues go up
91
92
We considered the effect of loosening the long-term constraint on either government expenditures or
revenues in extensions of the structural VAR model in 5.3. We could not reject longer-term effects
of fiscal shocks, endorsing the hypothesis that supply side effects of fiscal policy decisions are part
of the ‘supply’ shock.
This is a likely consequence of the annual frequency of the fiscal data.
136