The name is absent

would attribute too much of the variation in fiscal policies to the economic cycle and too
little to the offsetting systematic discretionary adjustments.

To illustrate this phenomenon for Germany, France, Portugal and Spain, we follow Lane
(2003) in estimating the output elasticity of the main budgetary items. I.e., we regress in
(5.7) the main budget items on economic growth for the sample period 1970-2004,

dlogX_i,t =ω_i +γ_idlogY_t +μ_i,t                                                  (5.7)

where X_i,t is total spending, government investment, current spending (consumption and
wage spending), or interest payments, and Y_t is real output. Likewise, we estimate model
(5.7) in which where X_i,t contains either total revenues, current revenues or (in)direct tax
revenues. The estimates are also repeated for the decades 1970-1980, 1981-1990, and
1991 to 2004, as we have reasons to expect quite some time-variation. Table 5.7 reports
the results of an OLS estimation of (5.7), with a correction for first-order autocorrelation.

Table 5.7 Budget Elasticities from OLS on Equation 5.7

Note: */**/*** denotes significance at the 10/5/1 % significance levels, respectively.

The switch from small negative spending elasticities in OECD (Girouard and André,
2005) to a strongly positive elasticity is very strong in Germany and Portugal, where it is
significant for all budget items. Government investment is the most procyclical budget
component. But the main category driving this result is - in absolute terms - government
consumption. In Germany, a large role is also played by wage spending in the last decade.
Fiscal spending expansions under positive economic growth are strongly concentrated in
increased wage spending in Portugal. In contrast, Spain, and in particular France, have
not been subject to a similar bias. No expenditure item - except for interest payments -
shows significant signs of procyclicality.

151

<<   <   143   144   145   146   147   148   149   >   >>