V. Saving behaviour and the effectiveness of fiscal policy

OECD Economic Outlook, Dec, 2004

(1.) For a full discussion of the methodology and results see de Mello et al. (2004).

(2.) The European Commission (2003) provides a comprehensive study of the effects of past fiscal adjustments in the European Union, as well as a survey of existing studies. Among the episodes of fiscal consolidation identified, this study concludes that around half of them have been expansionary.

(3.)The traditional channel for these non-Keynesian effects is private consumption (Giavazzi and Pagano, 1996, Giavazzi et al., 2000). But non-linear effects may also take place through private investment, as discussed by Alesina et al. (2002).

(4.) Zaghini (1999) suggests that shifting the composition of retrenchment toward expenditure cuts increases the probability of success. Alesina and Perotti (1995), McDermott and Wescott (1996), Alesina and Ardagna (1998), Alesina et al. (1998) report similar findings. Giavazzi et al. (2000) suggest that offsetting saving responses are stronger during large fiscal contractions, but particularly when based on tax measures. Wealth effects are strengthened if interest rates come down as a result of the fiscal consolidation, to the extent that corrective measures, if credible, contribute to reducing risk premia (Blanchard, 1990, and Zaghini, 1999).

(5.) A more conventional approach is to estimate the reduced-form saving equation in a partial equilibrium set-up, in which the lagged dependent variable is included in the set of regressors primarily to deal with inertia in saving behaviour. However, the error-correction specification is preferred because the partial-equilibrium set-up only allows for very simple dynamics, making the estimation of long-term private saving offsets often unrealistically high.

(6.) The control variables affect the private saving dynamics as follows: i) fluctuations in the terms of trade are positively associated with private saving in the short term (an improvement increasing saving); ii) broad money affects private saving negatively in the short term (increased liquidity reducing saving); and iii) the old-age dependency ratio and the equity market index affect private saving negatively in the long term. The proxies for wealth effects appear to have a statistically significant effect on private saving, with an increase in housing and equity prices acting to reduce private saving in both the short and long terms.

(7.) The long-term private saving offset can be calculated by dividing the estimated coefficient on [S.sup.pub.sub.i,t-1] by minus the estimated coefficient on [S.priv.sub.i,t-1].

(8.) In this particular case, the tax reforms with which the deficit was associated may have been instrumental in generating positive wealth effects. See International Monetary Fund (2003), for further discussion and empirical evidence for the United States.

(9.) Based on the parameters reported in Table V.2, the long-term private saving offset is estimated at about three-quarters in the United States (-(-0.22 0.44)/-0.29).

(10.) It may still be the case that a strong complementarity between public and private consumption is overriding a truly Ricardian behaviour in reaction to tax changes. But this would be an original and surprising configuration.