Research Discussion Paper – RDP 8003 Debt Versus Tax Financing of Government Spending: On Barro's Neutrality Theorem

Abstract

One of the longest-standing debates in economics is whether or not substitution of debt for tax finance of government spending has any real effects on the economy. The earliest argument, usually attributed to Ricardo, is that debt will only alter individual behaviour if it is regarded as a positive contribution to net wealth. Subsequent debate revolved around the question of whether or not individuals perceive debt as wealth. In the most recent contribution to this debate, Barro showed that debt will not be regarded as net wealth in a system characterised by operative intergenerational transfers. The essence of his argument is that, provided the individual cares enough about his children to leave them a bequest, he will also calculate the tax liability that falls on them due to debt issued in his lifetime. In other words, the use of deferred taxes in the form of debt should not fool the population into believing that they are better off for having avoided current taxes.

The objectives of this paper are twofold: to extend and reinterpret Barro's result, and to re-examine the various objections to the neutrality proposition. Before summarising my findings it is appropriate at this point to stress the definition of neutrality as relating to the effects of substituting debt for tax financing of a given level of government spending. Neutrality in these terms does not imply that an increase in debt coupled with an increase in spending will have no real effects. It simply means that the effects are no different to those that would follow from a tax financed increase in spending.

Using an overlapping-generations life-cycle framework this paper extends the neutrality proposition to include long-run neutrality, perpetuities as well as finite-maturity debt, and gift as well as bequest behaviour. Two insights are provided by this extension. The first is that neutrality is shown to be completely independent of the net wealth content of debt. The second is that neutrality is shown to hinge crucially on the perfect budgetary substitutability of debt and intergenerational transfers. The first result suggests that the theoretical and empirical literature on debt neutrality have focussed on the wrong issue. This assertion requires some elaboration.

The traditional neutrality argument considered the economy as being adequately represented by a single individual. As viewed by this individual the issue of debt generates an asset and a liability of identical value. That is, in every period over the life of the debt the government gives with one hand exactly what it takes with the other. Apart from being a zero contribution to his net wealth, the debt leaves the individual's budget constraint unaltered. With the exception of some minor distribution effects that could arise from having many individuals, rather than one, it was generally agreed that rationality should be sufficient to generate neutrality. This argument sounds convincing but it has a flaw. As soon as a life-cycle framework is used it becomes apparent that debt involves a particular type of redistribution effect that is sytematic and fundamental to the nature of debt as a financial asset, and therefore should not be ignored. Since debt is typically acquired in youth and sold in old age it involves intertemporal and intergenerational redistribution; that is, acceptance of the debt issue implies a change in the pattern of planned lifetime consumption from youth to old age.

The rearrangement of lifetime consumption patterns associated with debt will generate a behavioural response from individuals. The essence of the neutrality question is whether or not they can restore their pre-debt consumption patterns without altering the aggregate demand for capital. At this point we come to the second result of the paper. So long as individuals observe intergenerational transfers and debt as perfect budgetary substitutes their optimal response to debt will be to substitute bequests for debt as a means of restoring their desired comsumption patterns. As the demand for bequests rises equally with the supply of debt there will be no influence on the demand for capital and neutrality will follow.

Note that with recognition of the intertemporal implications of the debt the case for neutrality becomes more restrictive. In the traditional case individuals had only to recognise that they were no better or worse off with debt and therefore to keep doing what they were already doing. In the life-cycle model they must recognise that their lifetime consumption patterns are being changed, realise how to respond optimally, and actually be faced with the perfect budgetary substitutability necessary for them to neutralise the consumption effects. It is the last of these that is the weakest link in the argument. While it is true that the economy as a whole faces perfect substitutability, particular individuals do not face the same transformation rates as the economy due to corner solutions, illusions, distortions, indivisibilities and distribution effects associated with taxes and family size. Since aggregate neutrality depends on individual neutrality, the theoretical case for full neutrality is not seen as being particularly strong. Neutrality is nevertheless argued to be a plausible long-run result. The paper also contains some suggestions for empirical testing of the neutrality hypothesis.