IV. Fiscal stance over the cycle: the role of debt, institutions, and budget constraints

OECD Economic Outlook, Dec, 2003

[FIGURE IV.4 OMITTED]

Overall, on the basis of the country experiences reported above, over the past two decades--a period spanning about two full business cycles--the stance of fiscal policy has been characterised by episodes of both counter- and pro-cyclicality. These appear to be dependent on sustainability considerations, particularly with respect to government indebtedness. Restoring longer-term sustainability seems to have played an important role in determining whether fiscal stance is output-stabilising in an immediate sense.

Factors determining pro- or counter-cyclicality

This section tests more formally the extent to which counter- or pro-cyclicality, and asymmetries in fiscal stance between booms and downswings, have been affected by strategic, institutional and political-economy factors. These include high taxes, public sector size and expenditure rigidities, and indebtedness. They also embrace the type of fiscal rule adopted (if any), institutional factors such as the political composition of government, and electoral regimes and cycles. The methodology is described in the Appendix, and is based on regressing movements in the cyclically-adjusted primary balance--measuring discretionary action--on cyclical fluctuations in the primary balance. The intuition is that, if the estimated correlation is negative, built-in stabilisers are being offset by discretionary action, which is pro-cyclical. To test the extent to which fiscal policy responds to sustainability factors, the debt/GDP ratio is incorporated in the equation, while the influence of institutional factors can be gauged by dividing country samples and/or periods according to discrete characteristics. Being based on pooled cross-country and time-series analysis, using a sample of 21 OECD countries during 1980-2002, the aim of the exercise is to highlight OECD-wide trends, rather than individual country experiences.

The results, presented in the Appendix and summarized in Table IV.1, are reported in terms of the sensitivity of fiscal stance to cyclical fluctuations in the budget balance (i.e. the percentage-point change in the cyclically-adjusted primary budget balance associated with a percentage-point change in the cyclical component of the primary budget balance), controlling in each case for the effects of debt on fiscal stance. An important initial finding is that discretionary shifts in fiscal stance tend to be asymmetrical over the cycle. They are counter-cyclical in downturns and there is evidence of pro-cyclicality in upturns, when a one percentage-point increase in the cyclical component of the primary balance is estimated to be accompanied by a relaxation of the cyclically-adjusted balance of about 0.2 percentage point of trend GDP.

While these results indicate a bias towards policy easing in both upswings and downturns, they also suggest the existence of a "sustainability motive" in fiscal policy associated with the need to control public indebtedness. Each percentage point increase in the debt/GDP ratio is estimated to lead to an average year-to-year increase in the cyclically-adjusted primary balance of about 0.06 per cent of GDP. The coefficient seems quite small, but movements in the debt/GDP ratio can be substantial--the experience of Ireland and Australia is that the debt/GDP ratio can be reduced by 4-5 percentage points of GDP per annum, which implies a lower underlying cyclically-adjusted primary balance of about 0.3 percentage point of trend GDP.