[1996], Heylen and Everaert [2000] and Von Hagen et al. [2001])8. It is unlikely that the
measured discretionary fiscal impulse will be exactly zero even when there is no discretionary
policy action enacted by governments. So, the key objective is to distinguish between small
adjustments and a category of significant, determined attempts to improve the fiscal position.
The criteria used in the existing literature differ slightly from paper to paper. In what follows
we apply the original Alesina and Perotti [1995] definition9:
Definition 1:
A period of fiscal consolidation is deemed to have occurred in a given year if the
discretionary fiscal impulse is greater than or equal to +1.5 percent of GDP.
3.3 Classifying the Success of Fiscal Consolidation Attempts
Of course, not all consolidation attempts are successful10. The success/failure classification
scheme adopted by Alesina and Perotti [1995 and 1997], Alesina et al. [1998] and Von Hagen
et al. [2001] is based upon observed changes in the debt to GDP ratio in the periods
subsequent to a consolidation attempt. Specifically:
Definition 2:
A fiscal consolidation is successful if, three years after the consolidation attempt, the ratio of
debt to GDP is at least five percentage points below the level observed immediately prior to
the consolidation attempt.
The above definition of success does not take into account the differential effort that would be
required by governments to reduce their debt to GDP ratio given their differing initial (pre-
consolidation) conditions. If a government manages to stabilise its debt to GDP ratio from a
previously explosive path it seems reasonable to view this as a relatively successful outcome.
8 A small number of studies have adopted a “narrative approach” to the analysis of discretionary monetary policy,
notably Romer and Romer [2000], but we are not aware of any cross-country studies of fiscal policy that adopt this
approach.
9 A limitation of Definition 1 is that it does not account for more gradual consolidations involving small changes
over longer time horizons. Any criterion used is necessarily subjective, but it remains possible that we have
ignored a number of smaller but more sustained consolidation attempts. Whilst we recognise this, there is a danger
that the use of more liberal criteria would reduce the discriminatory power of the tests that follow. Alesina and
Perotti [1995] and Alesina et al. [1998] do allow for the possibility of two-year consolidations and it is worth
noting that our key results remain robust to this change. For a discussion of the duration of fiscal consolidations
see, for example, Von Hagen et al. [2001].
10 Discussions of the 'success rates' of recent fiscal consolidations are provided in IMF [2001], EC [2002] and
OECD [2002].