2.5 Plausibility of the Results
Figure 4 relates the run of debt/GDP ratios in the two countries to the underlying fiscal policy
regimes9 . The graphical representation shows that the regime changes also translate into the
process of debt/GDP ratios, which is absolutely reasonable, as a weak or insignificant reaction of
the revenue/GDP ratio to increases in the expenditure or the lagged debt/GDP ratio would lead
to an increase of government debt in the current period and vice versa. We see a stabilization of
public debt in relation to GDP in Germany, during the more sustainable periods of fiscal policy,
while the debt/GDP ratio even starts shrinking in the case of Spain.
Besides this rather intuitive and informal plausibility check of our estimates, one can find formal
evidence for the robustness and plausibility of our results. Thams (2007) analyzes the relevance
of the FTPL for Germany and Spain using data for the period 1970-1998. The analysis provides
evidence for non-Ricardian fiscal behavior in Spain, while the opposite is true for Germany.
These results coincide with those given here. In Spain we find a very little reaction of revenues
to changes in public debt during that period. Instead the path of public debt is stabilized
by a one-time shift in autonomous government revenues. Although Germany shows an almost
continuous increase of public debt over the sample period, autonomous revenues are considerably
larger than in Spain leading to higher revenue/output ratios given everything else, combined with
the fact that we find a positive probability for a switch to the more sustainable regime 1 during
almost the entire sample period10 . This would militate in favor of a Ricardian fiscal policy.
9 The choice for one regime or the other is determined by the corresponding regime probabilities. We say that
fiscal policy is in regime 1 at time t, if the probability for regime 1 at time t is larger than 0.5. In the case of
Germany we summarized the period of multiple regime changes at the end of the 1990s to a single regime change.
10Davig and Leeper (2005) state that one may use the regime probability as a proxy for individuals’ expectations
about future fiscal policy.
12