5 Conclusions
Recent evidence on the impact of fiscal policy on long-run growth in OECD countries has adopted the Barro (1990) framework to distinguish between ‘productive’ and ‘unproductive’ expenditures, and ‘distortionary’ and ‘non-distortionary’ taxes. Using estimated long-run growth effects from these fiscal variables from BGK (2001), this paper simulated the effects on growth rates of observed fiscal policy changes in the EU. With two exceptions (Finland and Spain - where long-run data appear unreliable) the individual country growth effects of actual changes in taxes, expenditures and deficits are plausible at around –0.3 to +0.3 of a percentage point per annum. Few common policy scenarios are apparent in the data however, with key sources of differences between countries being the extent to which distortionary taxes or deficits were used to fund public spending increases and whether additional spending was focussed on ‘productive’ activities. The paper also considers whether our growth regression model, which imposes parameter homogeneity across countries, is justified. The evidence suggests this is the case, with a high degree of uniformity across countries. One implication of these results is that changes in the overall share of taxes or spending in GDP or the annual budget surplus/deficit are not good guides to whether the growth effects of fiscal policy are likely to be positive or negative.
Finally the paper considered whether there is any evidence of ‘fiscal convergence’ across the EU. That is, are growth-affecting fiscal variables becoming more similar over time for the EU? Though data are limited, the answer to this question appears to be negative, with little evidence of unconditional convergence but countries generally reverting to their own steady-state paths. Some EU countries appear to have chosen different long-run values for key growth-affecting fiscal/GDP ratios from their EU (or OECD) neighbours. However, as might be expected, budget deficits show a higher degree of long-run uniformity.
