the idea that recessions are economically and/or politically more costly than output
booms, and that the fiscal policy response to them is accordingly stronger. But it is also
consistent with the idea that some elements of the fiscal structure, such as unemployment
compensation, are relatively insensitive to the business cycle at high levels of economic
activity, but become larger in deep recessions.
As pro-cyclicality contrasts with the stabilisation function of fiscal policy, a number of
explanations are offered for these results, including conflicting policy goals, information
problems (real-time data problems), complexity of decision-making and implementation
lags. Talvi and Vegh (2000) offer a model rationalising pro-cyclical fiscal policies
primarily in developing countries but also in the industrialised world - for countries with
a large variability of the tax base in general. If the latter is the case, tax smoothing would
require large deficits to be run in economic downturns, and high surpluses in upswings.
But finance ministers may be tempted to avoid large surpluses knowing that they will
nurture political pressures to spend public monies, and prefer to run a pro-cyclical policy.
Tornell and Lane (1999), on the other hand, argue that the degree of political competition
increases during upswings. After all, each group or power block competing for public
resources knows that governments will not run surpluses during economic expansions,
but that other groups will increase their appropriate share by an even greater amount.
Therefore, they will compete more intensely for resources during expansions, and less so
during recessions. As a consequence, fiscal policy becomes more pro-cyclical the more
fragmented and open governments are to such pressures.
Yet a range of literature also points to possible asymmetries in fiscal responses to
recessions and upturns. Mayes and Virén (2004) find strong evidence of asymmetric
cyclical behaviour of government deficits, with these asymmetries mainly relating to the
cyclically adjusted deficit. Structural deficits increase when output shrinks, but structural
deficits (structural surpluses) also tend to increase (decrease) when output expands.
According to Mayes and Virén, the different cyclical effects show up in both revenues
and expenditures. Revenues seem to be more sensitive to output growth in depressions
than in booms. Thus, in booms, the revenue/trend output ratio remains more or less
constant, while in depressions it decreases quite markedly. Expenditures seem to increase
in depressions and decrease in booms. They conclude that from the viewpoint of counter-
cyclical fiscal policy, the main problem appears to be behaviour in “good times” when
discretionary action does not seem to help smooth the output growth path.
Also the OECD (2003) concludes - on the basis of a panel estimate - that, overall,
countries conducted pro-cyclical fiscal policies in cyclical upturns and counter-cyclical
policies in downturns. However, sustainability problems associated with indebtedness
seem to be a key determinant of whether the fiscal stance is pro-cyclical during
downturns.
Forni and Momigliano (2004), using real time data, find that fiscal policy was generally
counter-cyclical during adverse economic periods. They conclude that fiscal policy was
more counter-cyclical at the beginning of the 1990s than during the recent downturns.
Balassone and Francese (2004), too, highlight that fiscal policies in OECD countries have
been counter-cyclical mainly in downturns. While automatic stabilisers are left free to
operate during downturns, during expansions their effect is compensated by discretionary
loosening, which implies that budgetary balances are not improving in upturns.
102