XREAP2007-14
cyclical response of the budget in the fiscal rule (6).18 We consider a much more homogeneous sample
period than Bohn (2005). Nevertheless, even over the period 1963-2000 there is some evidence of
significant breaks in the debt response (Table 1). The Andrews tests indicate a significant break in
1967, which coincides with increasing defence spending on the Vietnam War. However, if we control
for possibly varying volatility in the subsamples before and after this break (Stock and Watson, 2003),
the test locates the break in 1994. This heralded the start of the long economic boom and the
consolidation of public debt by the Clinton Administration. As these breaks are located at the ends of
the sample, we cannot explicitly model the differences in debt response.
For Germany, the response of the government budget to debt is insignificant instead. With a control
for a cyclical response of the budget, there is some weak evidence that the budget of all government
tiers together becomes sustainable. This is in line with the weak or insignificant responses that several
other studies have found on general government data (Ballabriga and Martinez-Mongay, 2005). One
might suspect that Reunification profoundly changed the fiscal system. Break tests on the fiscal rule
also confirm the relevance of this shift. However, if we condition the break test on the Reunification in
1991 (and in additional control for changes in volatility in the subsamples), then there is evidence of a
change in debt response in the early eighties. This break is only precisely estimated for the fiscal rule
(6) that allows for a cyclical reaction. This break is due to the large changes in fiscal policy in the
seventies. Government spending (mainly on social transfers) boomed under the Brandt government in
1976. Fiscal policy in the eighties was instead aimed at a gradual consolidation. We henceforth
continue with a split up of the sample in the periods 1970-1990 and 1991-2005. Estimating the fiscal
rule over these subsamples shows a significant difference in debt consolidation. The debt response is
insignificant for West-Germany, whereas there is some weak evidence that German fiscal policy
became more sustainable afterwards after Reunification.19 The explanatory power of the fiscal rule is
rather weak: we find a weak debt response only if we include a cyclical reaction of the surplus. Note
also that the finding of weakly procyclical policies has been a common finding for German fiscal
policy.
The aggregate debt responses can be attributed to rather different reactions of the central government
and the state governments in the US and Germany. We first test the sustainability of federal
government policies and each US state in a system of fiscal rules.
18 The results seem to be robust to the inclusion or not of the cyclical variable. We have also used IV estimation
techniques to take into account the potential endogeneity, and results are quite similar to the ones shown in the
paper. Results are available from the authors upon request.
19 In contrast, Gall and Perotti (2003) or Greiner and Kauermann (2007) find weaker responses to debt over the
nineties.
14