What could explain result 2), namely the decline in the effects of government spending
on GDP and its components in S2, and the larger real interest rate response? This section
explores a few possible explanations.
First, all the economies in the sample have become more open over time; however, the
increase in the export / GDP ratio is probably too small to account for the large decline
in the government spending multipliers. Second, for a large part of the first period the
countries of the sample were on a fixed exchange rate regime: a standard Mundell -
Fleming model would predict that fiscal policy is less powerful under flexible exchange
rates. However, when exports are added to the benchmark VAR (results not shown),
there is no evidence of systematic crowding out of exports by fiscal shocks in S2. Third,
taxation might have become more distortionary in S2: the neoclassical model would then
predict a smaller, or even negative GDP response (see e.g. Baxter and King [1993]). While the intrinsic distortions of a tax system (and in particular of corporate income taxes) are
difficult to measure, in general marginal tax rates on individuals have decreased in S2.
Fourth, we have seen that in several countries the persistence of the government spending
shock falls from S1 to S2. As the wealth effect falls, the neoclassical model can easily
explain the decline in the GDP and private investment responses, but it has the opposite
prediction for private consumption, since the negative wealth effect on forward-looking
consumers also weakens. The real interest rate also increases less. The neo-keynesian
model of Galí, López-Salido and Vallés [2003] has similar difficulties in explaining the
decline in the response of private consumption, because the non-ROT consumers behave
exactly like in the neoclassical model. In addition, note that not only the GDP response,
but also the spending multiplier falls in S2, hence the lower persistence of government
spending is unlikely to be the whole explanation.