Fiscal Policy and the Mehra-Prescott Puzzle: On the Welfare Implications of Budget Deficits When Real Interest Rates Are Low
Bohn, Henning, Journal of Money, Credit & Banking
This paper examines the welfare implications of government budget deficits. The real return on U.S. government debt has historically been far below the average return on equities and substantially below the average rate of economic growth. If the interest rate is below the average growth rate, a government may roll over its debt with interest--that is, run a primary deficit and not provide debt service--and still expect a declining ratio of debt to GDP. Are persistent budget deficits therefore unproblematic?
An analysis of deficit financing strategies that rely on low interest rates faces the problem that we do not have a fully convincing explanation why the real return on safe assets has been so low. As Mehra and Prescott (1985) have pointed out, standard asset-pricing models have considerable difficulties in explaining the equity premium. Complicating the puzzle, the relatively high return on equity capital suggests that the low safe interest rate is not a sign of dynamic inefficiency (see Abel et al. 1989; Zilcha 1992).
This paper examines the fiscal policy implications of different "explanations" of the equity premium puzzle. The various explanations are best organized in two groups. One group of authors maintains either that a high risk aversion parameter should not be viewed as puzzling or that the puzzle can be resolved by more general models within the framework of complete, frictionless markets. A second group of authors argues that a solution to the Mehra-Prescott puzzle can only be found in models with transactions costs that deviate significantly from the complete markets paradigm. [Specific references are in section 1; see also the survey by Kocherlakota (1996).]
Using two simple models, I show that a welfare assessment of persistent budget deficits depends crucially on which of these two lines of argument is correct. If low safe interest rates are due to risk aversion, the government can sustain frequent primary budget deficits by rolling over safe debt, but such policies generally involve an inefficient allocation of risk. This is because safe debt implies high future tax rates in states of nature with low consumption, that is, it creates systematic risks for the taxpayers who implicitly back the debt. A welfare assessment of government debt policy in a stochastic setting must therefore include an analysis of how government debt affects the allocation of risk, and not just focus on expected values. In practice, government policy plans--for example, administration or CBO budget projections--are unfortunately almost always expressed in terms of expected values. Hence, the paper raises serious questions about the interpretation of such projections. One may even wonder to what extent the widespread use of expected value projections provides political incentives to issue debt securities that carry a low interest rate.
Very different welfare implications can be obtained in models that explain low interest rates on government debt by market imperfections. I demonstrate this point in a model with intermediation costs in which the government can issue debt at an interest rate below the interest rate on private loans. Then permanent primary deficits are possible and they do not have adverse implications for future taxes.
Overall, the contrast between the two types of explanations for the Mehra-Prescott puzzle suggests that we do not have definitive answers on how to assess budget deficits in a situation with low real interest rates. If low interest rates on government debt are due to risk aversion, a policy that exploits low real rates to sustain budget deficits will impose risks on future taxpayers. But if low interest rates on government debt are due to special advantages of government borrowing over private borrowing, persistent budget deficits may be sustainable without creating risks and they may be part of a welfare-maximizing policy. Since we do not have a definitive answer as to which of these lines of argument is empirically more relevant, we have to conclude that the Mehra-Prescott puzzle has created a puzzle for assessing fiscal policy. …