Fiscal Requirements for Price Stability

The paper opens by noting a “worldwide movement toward greater emphasis upon the achievement of inflation targets” alongside increased central-bank independence, which “would seem” to imply that monetary policy can properly be separated from fiscal policy (Introduction, p. 1). Woodford identifies the two theses this separation requires – “first, that fiscal policy is of little consequence as far as inflation determination is concerned, and second, that monetary policy has little effect upon the government budget” – and states his central claim up front: “neither proposition is true, for reasons that are related,” because the standard calculation dismissing fiscal effects of monetary policy “neglects a more important channel…namely the effects of monetary policy upon the real value of outstanding government debt” (p. 1).

The two grounds are that “inflation is purely a monetary phenomenon” and that “Ricardian equivalence” implies fiscal policy cannot affect aggregate demand under rational expectations (Introduction, pp. 2-3). Woodford argues fiscal shocks affect aggregate demand, and fiscal policy specification matters for the consequences of monetary policy, “in rational expectations equilibria associated with policy regimes…that I shall call ’non-Ricardian’…even when the monetary policy rule involves no explicit dependence upon fiscal variables of any sort” – these effects operate “through the effects of fiscal disturbances upon private sector budget constraints and hence upon aggregate demand,” and are neutralized by Ricardian equivalence only under a restrictive class of fiscal policies.

The familiar textbook story assigns the central bank a seignorage target dictated by the size of the fiscal shortfall, so that “inflation is still a ‘purely monetary’ phenomenon,” just one driven by a fiscally-dictated money-growth rate (Section 2, pp. 3-4). Woodford argues fiscal policy can affect the price level even when the central bank pursues a fully “autonomous” interest-rate rule with no dependence on fiscal variables at all, and that this possibility “continue[s] to exist even in…the ‘cashless limit’” where seignorage revenues are negligible – so the threat to price stability from fiscal policy “cannot be so easily dismissed, even for advanced economies” with sophisticated financial markets and independent central banks (pp. 4-5).

The U.S. Treasury-Fed interest-rate control program from April 1942 through the March 1951 “Accord,” under which the Fed pegged the yield on 90-day Treasury bills at 3/8 of a percent (with the bill price “completely fixed” through June 1947) and supported the prices of 1-year certificates and 25-year bonds at fixed yields (Section 2, pp. 8-9). Woodford highlights that “the commitment to supporting the price of long-term bonds seems to have been the central element of Fed policy in the late 1940s”: when bond prices rose in early 1949, “the Fed sold over three billion dollars of its bond holdings” to defend the price floor, “in the face of criticism at the time, over the contractionary consequences” – direct evidence that price-level-relevant Fed policy was driven by debt-management goals rather than by a seignorage target imposed by the Treasury.

“A fiscal policy commitment [is] Ricardian if it implies that the present-value relation…or equivalently the transversality condition…necessarily holds for all possible goods-price and asset-price processes” (Section 3, p. 26). His example is a rule setting the primary surplus each period as a fixed fraction of existing government liabilities (adjusted for interest savings on the monetary base); Woodford shows this rule mechanically guarantees the transversality condition regardless of the price path, so that fiscal variables drop entirely out of the equations that determine the price level, and Ricardian equivalence – invariance of the equilibrium price process to the choice of fiscal policy within this Ricardian class – follows as a consequence (pp. 27-28). He notes this definition is a deliberate refinement of his own 1995 usage, chosen because it isolates exactly the case in which “the transversality condition ceases to play any role in equilibrium determination” (footnote 26, p. 27).

If the primary surplus instead follows an exogenous process not automatically adjusted to fiscal news, then “most paths for the price level and the nominal interest rate…will not imply dynamics for total government liabilities that satisfy the transversality condition”; only those price-level paths consistent with the specific fiscal expectations in question will do so, “making fiscal expectations relevant to price-level determination” (Section 3, pp. 27-28). This is not a violation of the government’s true intertemporal budget constraint – Woodford addresses directly the objection “Mustn’t fiscal policy satisfy an intertemporal budget constraint?” – but rather reflects that under non-Ricardian policy, the constraint is satisfied by an equilibrium adjustment of the price level itself (as in the bond-price-support example), not by an automatic fiscal adjustment.

Following Loyo’s (1999) analysis of the Brazilian hyperinflation of the 1980s, Woodford shows that with a Taylor rule i_t = phi(Pi_t) satisfying the “Taylor principle” (elasticity of the rule’s response exceeding one at the target inflation rate), the target inflation rate Pi* is a locally determinate equilibrium only if fiscal policy is locally Ricardian (Section 4.1, pp. 51-54, Figure 3). If instead fiscal expectations are described by an exogenous surplus sequence, there is a single initial inflation rate consistent with those fiscal expectations, and if expected future surpluses are “too small” so that this rate exceeds Pi*, “the only possible [perfect foresight equilibrium] is one in which the inflation rate grows without bound over time” – an inflationary spiral in which higher inflation raises nominal rates, which raises the growth rate of nominal government liabilities, which raises inflation further; if surpluses are instead “too large,” the same logic produces an unbounded deflationary spiral. Woodford stresses this can occur “even if” a bounded, locally-Ricardian-looking fiscal rule is nominally in place, so long as the fiscal expectations that actually prevail are inconsistent with the Taylor rule’s target.

Woodford notes that even when both fiscal and monetary policy are jointly consistent with a stable-price equilibrium, that equilibrium may be only one of several rational-expectations outcomes, and “there may be good reason for people’s expectations to coordinate upon” an alternative – including self-fulfilling deflationary spirals that a Taylor rule alone does not exclude (Section 5, Conclusion, p. 71). Excluding these undesired equilibria requires a fiscal policy that is “locally Ricardian, and not merely globally Ricardian,” since a policy that is only globally Ricardian (e.g., an exogenous primary surplus process bounded only by eventual debt limits) leaves room for exactly this kind of self-fulfilling instability.

Woodford proposes “a target for the real value of the conventional budget deficit (inclusive of interest on the public debt)” as a fiscal-policy commitment to accompany a Taylor rule (Section 5, Conclusion, pp. 71-72). Such a target is locally Ricardian, so it “does not frustrate the central bank’s use of a suitably ‘active’ monetary policy” – avoiding the Loyo-style inflationary/deflationary spiral – and it simultaneously “plac[es] a floor on the path of the nominal value of total government liabilities, which can be useful as a means of excluding self-fulfilling deflations that would otherwise be possible equilibria under a Taylor rule.” Woodford notes the practical timeliness of the proposal: “commitments to budget balance or to deficit limits have achieved new prominence in macroeconomic policy in the same period” as increased central bank independence, in both the U.S. and the European Union.

“A central bank charged with maintaining price stability cannot be indifferent as to how fiscal policy is determined. Commitment to an anti-inflationary monetary policy rule, such as a Taylor rule with a low implicit inflation target, cannot by itself ensure price stability” (Conclusion, p. 70). He is careful to distinguish his diagnosis from the more familiar “central bank versus fiscal authority” framing associated with Sargent and Wallace (1981): the problem is “more subtle” than a simple inconsistency in which one authority must accommodate the other, because an equilibrium can exist in which both the monetary and fiscal commitments are maintained forever, yet that equilibrium involves an unbounded inflationary or deflationary spiral rather than the intended stable target rate – a distinction that matters because it shows central bank “independence” or “credibility” alone cannot solve the problem; only an appropriately chosen, locally Ricardian fiscal-policy commitment can.