Comment on The Conduct of Monetary Policy with a Shrinking Stock of Government Debt

For the past several decades, the Federal Reserve has relied almost exclusively on open market operations in Treasury securities for supplying liquidity and carrying out monetary policy. Following the string of government surpluses under the Clinton administration, the supply of Treasury debt was (until recently at least) projected to dwindle, raising questions about the feasibility of current procedures for the implementation of monetary policy. To examine the implications of a shrinking supply of government debt for monetary policy, Stacey Schreft and Bruce Smith provide a carefully constructed general equilibrium model of banking and the market for bank reserves. Using their model, Schreft and Smith conduct a number of thoughtful experiments that nicely illustrate the relevant concerns and offer some guidance for the consequences of alternative policies dealing with the shrinking stock of debt. Importantly, they show that a disappearing government debt could have, in their words, "severe consequences" for the implementation of monetary policy under current operating procedures but also point to a rather simple fix that would diffuse this problem.

The Schreft-Smith model is rather elaborate, arguably more than would be essential for the analysis presented in the paper, and one could quibble about some of their specific modeling choices or the confidence with which one should view the quantitative answers obtained from the numerical calibration of their model. In my view, greater emphasis on institutional detail and market structure, perhaps at the expense of the detailed derivation of basic concepts from first principles, might have presented a better balance. The basics of their argument, however, can be described in rather simple terms and the key lessons do not appear sensitive to many of the modeling details one could quibble about. To highlight the basics, consider a conventional money demand curve such that the demand for central bank reserves is decreasing in the nominal interest rate and, therefore, inflation, π, as shown in Figure 1. (I should note that much of the machinery in the paper can be thought of as essentially deriving such a downward-sloping demand curve from first principles within an overlapping generations model with limited communication and spatial separation. But the exact details of how this is done do not appear to be crucial for the argument so, for my purposes at least, I believe it suffices to take the existence of a conventional money demand curve for granted.) Given a desired rate of inflation, π*, this demand determines the level of reserves that the central bank would need to supply to bring the economy to an equilibrium with its desired inflation target (point A). Now suppose that the only way reserves can be supplied by the central bank is through open market operations in government debt. As long as the supply of government debt exceeds the demand for reserves at the desired target (as in line B in the figure), monetary policy with the desired inflation target can be implemented through open market operations with no difficulty.

Fig. 1.

Demand for Reserves and the Diminishing Debt Constraint

With declining debt, however, the supply of reserves will eventually become sufficiently limited that the central bank will no longer be able to provide the desired supply of liquidity required to support its inflation objective. (In the figure, think of line B gradually shifting to the left toward lines B´ and B˝) As long as open market operations present the only way reserves can be supplied, this imposes a binding constraint on the supply of reserves. Suppose, for example, that government debt shrinks to line B´ in the figure, below the level of liquidity necessary to support the inflation objective π*. With this constraint, the central bank must resign to the fact that it cannot supply enough liquidity to meet the demand associated with its original inflation objective. Instead, it must operate with a higher inflation objective (point A´), where the limited supply for reserves can satisfy the reduced demand corresponding to the higher opportunity cost associated with the higher inflation rate. With debt shrinking further (for example, from B´ to B...