effort in collecting information directly from a large number of businesses and organizations1 . A
simple mechanical concept like the Taylor rule is hardly compatible with such a decision process.
Already Taylor (1993) mentioned that policymakers do not follow policy rules mechanically.
Central banks need more than a simple policy rule to conduct policy. Particularly it requires
judgment to deal with special scenarios, which are not captured in a mechanical formula like
the Taylor rule. But in opposite to pure discretion, the settings for the instruments are not
determined from scratch each period. In this sense, policy rules are not more but also not less
than a tool in identifying the basics behind policy actions, as it is neither desirable nor likely
that a central bank starts from scratch each period.
The reason, why we start this literature review with monetary policy issues, is because policy
rules have so far found less application in fiscal policy analysis. Nonetheless, they offer a way
to think about fiscal policy systematically. Numerous papers deal with the question of cycli-
cal properties of fiscal policy and its ability to stabilize the economy using simple policy rule
specifications. Examples would be Gali and Perotti (2003), who assess the cyclical properties of
fiscal policy in EMU before and after the introduction of the Maastricht Treaty, Taylor (2000),
who investigates the reaction of automatic stabilizers in the United States, and Fatas and Mihov
(2001), who analyze the relationship between government size and business cycle volatility in
OECD countries.
Another strand of literature uses fiscal rules to test for the link between prices and public
debt, as induced by the fiscal theory of the price level (FTPL) and for the sustainability of fiscal
policy in general. Bohn (1998) finds out that U.S. fiscal surpluses have responded positively to
debt. He argues that this provides evidence that U.S. fiscal policy has been sustainable. For
the EMU, Afonso (2002) demonstrates, applying a panel data approach, that the FTPL is not
supported for the EU-15 countries during the period 1970-2001, as Member States tend to react
with larger future surpluses to increases in government liabilities. A recent paper by Davig and
Leeper (2005) analyzes regime switches in fiscal policy for the U.S. They show that there have
been periods of time, when government revenues have been positively and negatively affected by
changes in the debt-output ratio.
This paper follows the approach of Davig and Leeper (2005). We investigate the relationship
between fiscal instruments and public debt in a Markov-switching model, as a crucial difference
between the analysis of monetary and fiscal rules arises from the heterogeneity of fiscal policy.
In contrast to monetary policy fiscal policy is substantially affected by political flavors. With
different political responsibilities we may expect at least some change in fiscal policy behavior.
For this reason, we propose that any econometric analysis of fiscal rules should allow for changes
in the underlying coefficients, as these are generally unlikely to be stable over time.
The remainder of this paper is organized as follows. In section 2 we introduce a simple frame-
work for fiscal policy analysis using policy rules. We then give a brief survey of the methodology
that is applied in this paper. After a description of the data used in the analysis, we provide
the reader with the results in section 2.3. Finally, we make a systematic comparison of the
country-specific results and check for their plausibility in section 2.4 and 2.5. Finally, section 3
summarizes the results and concludes.
1For further details the interested reader is referred to Svennson (2001).