V. Saving behaviour and the effectiveness of fiscal policy

OECD Economic Outlook, Dec, 2004

Introduction

Private saving behaviour has important consequences for the effectiveness of fiscal policy ...

Fiscal policy has been used as an antidote to weak activity during the most recent downturn and fiscal consolidation has been delayed in some countries because of its perceived costs in terms of lower activity. However, the impact of fiscal policy on aggregate demand depends on the responses of private saving to changes in fiscal stance. In certain circumstances budget deficit shifts can be offset by simultaneous compensating changes in private saving. This chapter examines the possible extent of such offsets, focusing on the case where co-movements in private and public saving may be related to uncertainties about how long a budget deficit can be sustained and the consequent need to provide against future tax "surprises". (1) Even though conscious "tax discounting" may be rare, experience in many OECD economies suggests that fiscal adjustments made for stabilisation purposes can often be associated with inverse movements in private saving. Depending on their extent, such responses raise important issues for policy-makers.

... and this may be through various channels ...

Identifying the direct offsetting effects of budget deficits on saving is not easy because fiscal actions can be offset by private saving responses through a variety of channels besides tax discounting. The most direct, incorporated into most conventional aggregate demand models, may arise because a fiscal stimulus boosts disposable income and the propensity to consume out of an extra dollar of income is generally significantly less than one in the long term. More indirectly, private saving may rise because higher budget deficits drive up interest rate, which may cause financial "crowding out". In some countries, this effect would be accompanied by the negative effects on asset prices ("wealth effects") accompanying the accumulation of government debt. Because of these other channels of influence, as well as the links running from private to public saving, simple correlations between public and private saving cannot be used as evidence of direct expectations-generated private/ public saving offsets.

The approach used here is to estimate the direct effects of budget deficits on saving from pooled cross-country and time-series data, controlling for income, interest-rate and wealth factors. This allows the identification of OECD-wide behaviour patterns. However, the analysis also investigates whether there are country-specific differences in the behaviour of private agents to changes in the budget and whether the composition of the fiscal action--revenue, current spending or public investment--affects the private saving offset.

The main findings are as follows:

There appears to be a direct private saving offset ...

--The evidence of partial, yet substantial, direct offsetting movements in private saving is strong. The aggregate initial offset is about half in the short term after allowing for income, interest rate and wealth effects (which have an important impact on saving), rising to around 70 per cent in the long term.

... which applies to revenue and current spending ...

--Private saving appears to respond in relatively equal proportion to changes in current revenue and expenditure in the longer run, although the short-run saving offset is greater for changes in revenue.

... but public investment does not elicit a saving response

--Public investment does not elicit an offsetting saving response, consistent with such investment--where properly defined in the public accounts--yielding either a financial rate of return or a social return, accruing to future tax-payers.

The United States may be an exception

--The private saving response to deficits appears, exceptionally, to be positive in the United States over the longer term. Otherwise, there is no evidence of differential country behaviour.

Co-movements in private and public saving

Co-movements in private and public saving tend to be strong ...

OECD countries have experienced considerable swings in private and public saving over time (Figure V.1). However, establishing the direction of causality is complicated by a number of conceptual issues, in particular those related to measurement problems and the need to account for automatic stabilisers (see Box V.1). Because of these complications, raw correlations between public and private saving should not necessarily be taken to indicate the extent to which there is an behavioural relationship through which private saving offsets shifts in public saving. With this proviso, across regions, changes in fiscal stance have often coincided with opposite co-movements in private saving, thus smoothing fluctuations in national saving. This is confirmed by a correlation between changes in private saving and the cyclically-adjusted budget balance which is around -0.5 for OECD countries on average (Figure V.2). A closer look at the 1990s reveals that the countries included in the sample (excluding Japan) experienced a significant improvement in the cyclically-adjusted government balance, with movements in private saving going in the opposite direction (France, Norway and Sweden being exceptions). However, the magnitude of the co-movements differed considerably (Figure V.3).