Are Government Bonds Net Wealth?

Barro states at the outset that “the assumption that government bonds are perceived as net wealth by the private sector plays an important role in theoretical analyses of monetary and fiscal effects,” underlying Modigliani’s (1961) argument that debt issue raises perceived household wealth, raises desired consumption relative to saving, raises interest rates, and crowds out capital accumulation, as well as Metzler’s (1951) demonstration of monetary nonneutrality (Introduction, pp. 1095-1096). Barro notes it had “been recognized for some time” – citing Tobin’s 1952 remark asking rhetorically whether the taxes needed to carry a debt’s interest charges reduce other private wealth – that future tax liabilities imply an offset to any direct positive wealth effect, and that the key question is whether this offset is full or only partial (pp. 1096-1097). He identifies the two main arguments in the literature for a partial offset – (i) taxes are borne over a shorter horizon than interest payments because of finite lifetimes (Thompson 1967), and (ii) tax liabilities should be discounted at a higher rate than interest receipts because of imperfect capital markets (Mundell 1971) – as the specific claims the paper interrogates (pp. 1097-1098).

Barro uses a Samuelson (1958)-Diamond (1965) two-period overlapping-generations model with physical capital, in which each generation works and earns wage income while young, and its utility depends on its own two-period consumption and on the attainable utility of its immediate descendant (Sec. I.A, eq. 4, pp. 1098-1100). A member of generation i chooses a bequest A^o_i to a descendant in generation i+1, subject to the bequest being nonnegative. The central result: if the government issues bonds B to the currently old generation, financed by taxes on the next generation, and the solution for the bequest is interior (i.e., the nonnegativity constraint on A^o is not binding), a marginal increase in B is exactly offset by a compensating change in the bequest, so consumption and utility of every generation are unaffected (Sec. I.B, pp. 1101-1104). More generally, the result holds “so long as current generations are connected to future generations by a chain of operative intergenerational transfers” – either bequests from old to young, or gifts from young to old (p. 1106) – a much weaker requirement than every household literally living forever, since it only requires that the marginal transfer decision be interior at every generational link.

If a member of the old generation is already at the corner A^o = 0 (would have chosen a negative bequest if permitted), a government bond issue that transfers resources to that generation creates a genuinely new opportunity: the household raises consumption along with B, generating exactly the kind of excess asset supply that raises the interest rate and crowds out capital formation described by Modigliani (1961) (Sec. I.B, pp. 1103-1104). Barro notes that in a population with heterogeneous households, the aggregate effect scales with the fraction of households at this corner, so the model nests the standard wealth-effect result as the special case where operative transfers are absent (p. 1104, fn. 7).

Yes – Barro shows a social security scheme that pays the current old generation and taxes the young to finance it is “analogous to changes in government debt”: given interior bequests, the current old generation raises its bequest by exactly enough to offset the tax imposed on its children, who in turn adjust their own bequests, so a marginal change in social security benefits has no effect on any generation’s consumption pattern (Sec. I.C, pp. 1106-1107), a result Barro also extends, with the same logic, to other imposed transfer programs such as public support of education (p. 1107). Introducing a proportional inheritance tax (Sec. I.D, pp. 1107-1109) makes an interior bequest solution less likely at any given tax rate, but so long as some intergenerational transfers remain operative even at reduced levels, the paper’s central conclusion – a nil marginal net-wealth effect of government debt – continues to hold (p. 1109).

Introducing a proportional transaction cost γ on debt issuance and tax collection (Sec. I.E, pp. 1109-1110), Barro shows the net-wealth effect of government bonds actually turns negative: the combined two-generation budget constraint (eq. 16) shows total resources fall with an increase in B whenever γ > 0, which is “typically reflected in declines in all terms on the right-hand side” – current and future consumption for both generations. The effect of this negative wealth effect on the interest rate is left ambiguous, depending on how the decline in consumption compares with the increase in resources devoted to the bond/tax transactions themselves (p. 1110).

Barro next considers infinite-lived individuals split into low-discount-rate (r_l) and high-discount-rate (r_h) types, where r_h = (1+λ)r_l and λ > 0 represents the private transaction costs of lending to the “bad collateral” high-discount-rate group (Sec. II, pp. 1110-1111). If a new perpetual government bond is purchased by low-discount-rate individuals and its proceeds are transferred (with associated future tax liabilities) partly to the high-discount-rate group, the net-wealth effect for the high-discount-rate recipients is positive if and only if the government’s transaction cost γ is smaller than the private-market transaction cost λ – that is, only if the government is a more efficient intermediary of this specific loan than the private market (eq., p. 1111). Barro stresses that if the public-choice process pushes government debt issuance to the point where this efficiency gap is exhausted at the margin, the marginal net-wealth effect returns to zero “despite the continued existence of ‘imperfect private capital markets’” (Sec. II, p. 1112) – imperfect markets are a necessary but not sufficient condition for a positive marginal wealth effect.

If a bond yields liquidity services valued at L per year in addition to interest, and government incurs marginal cost c per bond in providing them, Barro shows the marginal wealth effect of new debt equals -(rL - c)/r, i.e., proportional to (L - c) (Sec. III, pp. 1112-1113). This is nil if the government acts as a competitive producer (public choice pushing L = c “as it should on efficiency grounds”), positive if the government under-supplies liquidity services relative to a monopolist’s optimum (L > c), and negative if it over-produces them (L < c) – and Barro notes this same logic extends directly to non-interest-bearing government money, which yields a zero explicit interest rate but analogous liquidity value (p. 1113, fn. 26).

Barro directly engages Tobin’s (1971) claim that debt issue “forc[es] on taxpayers a long-term debt of some uncertainty while providing bond-holders highly liquid and safe assets,” arguing Tobin’s framing neglects that the future tax liabilities are themselves part of household balance sheets (Sec. IV, pp. 1113-1114). If relative tax burdens across households are known with certainty, holding a fraction of government bonds matching one’s expected tax share is a perfect hedge against the variability in real bond values, so a simultaneous rise in interest payments and offsetting tax liabilities produces no net change in household portfolio risk (pp. 1114-1115). If relative tax liabilities are instead uncertain, the net risk effect of a debt increase depends on whether that uncertainty is purely random (raising risk, absent perfect private insurance) or is correlated with income (in which case taxation functions partly as income insurance, potentially lowering risk) – so “there seem to be no clear results concerning the effect of government debt issue on the overall risk contained in household balance sheets” (pp. 1115, 1188 [Summary]).

Barro’s summary conclusion is that “there is no persuasive theoretical case for treating government debt, at the margin, as a net component of perceived household wealth. The argument for a negative wealth effect seems, a priori, to be as convincing as the argument for a positive effect,” which he argues undercuts the common view (citing Patinkin 1964) that the marginal net-wealth effect lies somewhere strictly between zero and one (Sec. V, pp. 1115-1116). If the true marginal effect is close to zero, Barro spells out three specific implications: (1) the Metzler-type argument for nonneutrality of changes in the stock of outside money would not be valid; (2) a change in the stock of government debt would have no effect on capital formation; and (3) fiscal policy that shifts the mix of tax versus debt finance for a given level of government expenditure would have no effect on aggregate demand, interest rates, or capital formation (p. 1116).

The paper does not use the term “Ricardian equivalence” – that label was attached to Barro’s argument only afterward (notably in the exchange with Feldstein and Buchanan published as a 1976 JPE reply, cited within the ledger’s related literature). Barro’s own framing throughout is narrower and more conditional than the later shorthand “debt doesn’t matter” suggests: the zero-net-wealth-effect result is explicitly conditional on an operative chain of intergenerational transfers (Q2), does not hold at corner solutions (Q3), turns negative with transaction costs (Q5), requires relative government efficiency in the imperfect-capital-markets and liquidity-services extensions (Q6-Q7) to be exhausted at the margin, and leaves the risk-composition channel’s sign genuinely ambiguous (Q8) – the paper is best read as identifying the specific, falsifiable conditions under which government bonds are and are not net wealth, rather than asserting debt neutrality as a general law.