«Budget Deficits and the Intergenerational Distribution of Lifetime Consumption Daniel Shaviro THE LAW SCHOOL THE UNIVERSITY OF CHICAGO January 1995 ...»
JOHN M. OLIN LAW & ECONOMICS WORKING PAPER NO. 29
Budget Deficits and the Intergenerational Distribution of
THE LAW SCHOOL
THE UNIVERSITY OF CHICAGO
This paper can be downloaded without charge at:
The Chicago Working Paper Series Index: http://www.law.uchicago.edu/Lawecon/index.html Budget Deficits and the Intergenerational Distribution of Lifetime Consumption Daniel Shaviro* I. Introduction Few topics in American politics are more discussed and less understood than the federal budget deficit. This article, which is adapted from a chapter in a book in progress entitled The Political Economy of Budget Deficits, discusses one of the main issues that deficits raise: intergenerational equity, or the concern that deferring taxation relative to government spending, and relying on actual or implicit debt financing for current or projected future expenditure, reduces the lifetime consumption of future generations relative to that of current generations. I begin by summarizing the main empirical conclusions that the earlier chapters of the book reached after an extensive literature review. Obviously, I realize that the conclusions I state require greater support than they receive in this selection. I then turn to the normative issues raised by intergenerational distribution.
II. Summary of Empirical Conclusions In the last two centuries, no question pertaining to budget deficits has received closer or more consistent attention than that of their intergenerational effects. Writers since David Ricardo have alternatively decried or denied the tendency of debt financing to shift lifetime consumption from future to current generations, and from younger to older living persons. Prominent decriers include Ricardo, who posited but rejected Ricardian equivalence; Martin Feldstein, who argued that Social Security reduces saving and shifts lifetime consumption from future to current generations; James Buchanan, who challenged the Keynesian “no-burden, no-transfer” orthodoxy, and later helped develop the public choice attack on debt financing; and Laurence Kotlikoff, who sought a broader measure of fiscal policy’s inter-generational effects and advocated “generational balance.” Prominent deniers include Robert Barro, with his Ricardian view of fiscal policy as having no first-order intergenerational effects; and various Keynesians, based *Professor of Law, University of Chicago Law School. I am grateful to Anne Alstott and Richard Craswell for comments on earlie
both on the “no-burden, no-transfer” claim, emphasized by Abba Lerner, that the burden of public spending is borne currently no matter how it is financed, and on the claim, most recently associated with Robert Eisner, that public debt enhances not only current gdp but long-term economic growth.
My review of this debate has yielded a number of conclusions that are crucial prerequisites to attempting a normative assessment of the intergenerational issues raised
by budget deficits. These conclusions include the following:
Conclusion 1: The Ricardian equivalence theorem, stated in strong form to assert that the timing of taxation relative to spending has no first-order intergenerational consequences, is incorrect. Ricardian offsets, or adjustments to bequests and lifetime transfers between members of different generations to achieve preferred distributions no matter what the government fiscal policy, may take place to some extent, but are incomplete, mainly due to fiscal illusion (or a correct understanding that public debt may remain outstanding indefinitely), as well as to the effects of strategic behavior and impure altruism both within multigenerational households and between different households.
Nonetheless, the Ricardian analysis helps to demonstrate a number of important points, including the difficulty of measuring the net intergenerational effects of government fiscal policy, the fact that such policy provides only one of the many arenas in which the generations interact, and the importance of people’s behavioral responses to government policies, reflecting the policies’ perceived as well as actual effects.
Conclusion 2: Holding constant the amount and character of government spending, and assuming no significant actual or perceived risk of default, debt financing, relative to tax financing, tends in the long run to shift lifetime material consumption from future to current generations. This results from debt financing generally yielding greater aggregate perceived wealth than tax financing, both because a bondholder (unlike a taxpayer) possesses a valuable asset that offsets the transfer of cash to the government, and because, in the case of debt financing, people do not weigh the implied future taxes at their present value equivalent in current taxes. Greater perceived wealth should lead to greater lifetime spending on consumption (which is another way of saying that Ricardian offsets are incomplete). One should keep in mind, however, that this is a ceteris paribus conclusion. So many amorphous factors may affect people’s expectations and perceived wealth that one cannot be certain of the effect’s significance and consistency in practice.
Budget Deficits 3 This analysis of debt financing, relative to tax financing, also applies to “pay-as-yougo” financing where future revenue shortfalls are predictable, relative to accruing sufficient reserves to achieve long-run solvency under present tax and spending rules.
While the similarity between explicit public debt and long-term revenue shortfalls most often is made with regard to Social Security, it potentially applies to all government budgeting. Suppose, for example, that the federal government currently is achieving annual budgetary balance, but that future spending increases (say, to meet increased needs in health care or national defense) are expected with near-certainty. A policy of keeping taxes constant, and thus maintaining current budgetary balance but not accruing the surpluses necessary to achieve long-term balance under existing policy, would favor current over future generations relative to a policy of raising taxes now so as to achieve such long-term balance. Obviously, the same point holds for surplus relative to balance as for balance relative to deficit; future generations would benefit still more (all else being equal) from our levying taxes in excess of anticipated long-term revenue needs.
Conclusion 3: Holding constant the amount and character of government taxation, a debt-financed increase in government spending tends in the long run to shift lifetime consumption from future to current generations. This is less definite, however, than Conclusion 2, which was based on the reverse assumption of holding spending constant and changing the level of taxation. What makes it less definite is the lack of specification as to the character of the new government spending. If the spending is for long-term investment—at least, relative to the private uses of resources that it replaces (disregarding the Keynesian case of slack in the economy, where the resources would otherwise have gone unused)—then it may instead have the opposite effect, shifting lifetime consumption from current to future generations. The claim that, as a general matter, increased government spending is more likely to benefit current generations is based on the relatively short time horizon that political decisions, as distinct from private household decisions, may tend to have, given politicians’ disinclination to look beyond their terms in office, and voters’ rational ignorance of complex public issues and incentive to maximize short-term benefits relative to other households.
An unfunded promise of future spending, no less than debt-financed current spending, can shift lifetime consumption from future to current generations if it is credible. A good example is provided by Social Security through the 1970s, which, by promising current 4 Chicago Working Paper in Law and Economics workers benefits in excess of the value of their contributions, may have induced them to save less overall than they would have otherwise—in effect, for the same reason that one might save less if one expected to win the lottery upon retirement. Moreover, windfall increases in benefits to the elderly, such as that occurring in Social Security in 1972, probably reduce saving as well even if they were not anticipated, since they act as a onetime wealth transfer to an age cohort with a relatively low propensity to save.
Conclusion 4: As suggested by Conclusions 2 and 3, a strong form of the “no-burden, no-transfer” claim made by Keynesians such as Abba Lerner is incorrect. Although government spending has a real resource cost to the society without regard to how it is financed, the subjective burden on a bondholder, who voluntarily lends money to the government based on a specific expectation of repayment with interest, generally is less than that on a taxpayer, who pays involuntarily and with no such definite expectation.
Under some circumstances, however, the distinction between the two types of payors may be murky. For example, despite the involuntary nature of Social Security taxes, a contributor to Social Security may be more like a bondholder than a taxpayer, if the program does not force him to save more overall than he would prefer, and if the definiteness of his expectation of receiving future benefits with a present value at least equal to that his contributions approaches that of a bondholder.
Lerner’s “no-burden, no-transfer” argument helps to remind us of two important points, however. First, public debt, in terms of its direct effects, can transfer wealth only between overlapping generations. Only currently living persons can make or receive bond payments of principal and interest. As Lerner put it, public debt does not provide a time machine enabling us to transfer wealth from the future to the present. Second, public debt is a financial asset, not a real asset. Whereas real assets—both tangible, like a factory, and intangible, like a living person’s set of skills—are components of existing societal wealth, financial assets merely evidence people’s relative claims to such wealth.
Thus, in a real sense public debt has no direct relevance to the level of existing societal wealth, although over time it may influence the level of such wealth through its effects on behavior.
Conclusion 5: The traditional Keynesian claim that, short of full employment, stimulative budget deficits do not shift consumption from the future to the present, but instead increase consumption over the long-term, is incorrect, except perhaps under Budget Deficits 5 special circumstances (such as those prevailing during the Great Depression) when capital markets are not well-functioning, and/or a crisis of confidence has caused consumption and investment demand to collapse. No matter how socially undesirable and even destructive one considers a high rate of unemployment, and no matter how blameless the unemployed may be for the quirks in marketplace demand that (among other causes) prevent them from finding jobs, the claim that much unemployment is “involuntary,” in the technical, traditional Keynesian sense, is unpersuasive.
Accordingly, full employment is an artificially defined (however laudable) policy goal, rather than a meaningful description of a distinct state of the economy, and the claim that, when unemployment is regrettably high, there is economic “slack” that makes government spending effectively cost-free, cannot be accepted. Contrary to the traditional Keynesian position, society generally does face a tradeoff between increased consumption today and increased consumption in the future. It may similarly face a tradeoff between present and future employment, although this is less certain given the complexity of causation for employment levels. The new Keynesian case for expanding the money supply to combat the market failures that may deepen and prolong recessions is reconcilable with this analysis, since it does not rely on liquidity traps that make current consumption cost-free (although one could in theory imagine an instance where the only effect of monetary stimulus was to eliminate the waste from failed short-term market-clearing).