Cost of separating budgetary policy from control of inflation: a neglected aspect of central bank independence.(Special Issue on the Economics of Central Bank Independence)

1. Introduction

The Bank of England became independent within a few days of the 1997 General Election. This means that, in principle the UK's monetary policy is set independently of its fiscal policy. There may be some informal discussion between the new Monetary Policy Committee and the Treasury(1) of how the Monetary Policy Committee might react to particular fiscal policy changes, and conversely what the fiscal response might be to any particular changes to the interest rate. But there is neither evidence of any formal co-ordination between monetary and fiscal policy nor any recognition that such co-ordination might in some circumstances be desirable.

Indeed the main argument for central bank independence has come from one particular perspective. Central bankers argued in 1920 that they could do their jobs better if only they had greater power and influence (Sayers, 1976). But economists' enthusiasm for the idea has been, in large part, generated by articles by Kydland and Prescott (1977), Barro and Gordon (1983), and Rogoff (1985). These articles stress the benefits of giving a monetary authority the job of controlling inflation independently of the state of the rest of the economy. A government in charge of the monetary authority will want to stimulate output and this will be associated with unexpected inflation. The private sector will come to expect this unexpected inflation. It will anticipate an inflation rate so high that the government will abstain from stimulating demand because it does not want inflation to rise even higher. Thus, if the government controls monetary policy, the expected rate of inflation will be higher than if monetary policy is set by an institution which has no interest in stimulating demand.

These theoretical arguments have been combined with empirical studies (Cukierman, 1992) which suggest that inflation is lower in countries with independent central banks than in those where the government plays a major role in the setting of government policy, without there being any evidence of an output cost. However, as Doyle and Weale (1993) show, the effect is much stronger during periods of generally high inflation (the 1970s) than it was during the period of fixed exchange rates or after 1979. The combination theory and empirical finding led to a powerful case for central bank independence despite the suggestion by Posen (1993) that the empirical studies confused cause and consequence.

Later studies have changed the conclusion slightly, and suggested that an 'override' mechanism may be useful whereby the government regains control of monetary policy during a severe depression (Lohmann, 1992). But the basic framework for analysis has concerned the credibility of inflation control and it is not very surprising that, in this context, central bank independence appears attractive.

An argument based on credibility relies on the private sector having forward-looking expectations of inflation which react to policy pronouncements even if not always in the way that policies authorities would hope. In financial markets there are strong reasons for arguing that traders with irrational expectations will lose money to those whose expectations are rational. These arguments do not hold with the same force in the labour market, and a full analysis of the question of central bank independence should therefore also investigate other aspects of the division of responsibility for financial (monetary and fiscal) policy between two different independent authorities.

One important consequence of the division of responsibility is that the question of relations between the two financial authorities has to be considered. This has not been much discussed. Andersen and Schneider (1986), Hughes-Hallet and Petit (1990), and Spann and Ziemes (1992) consider a situation in which a monetary authority tries to control inflation while a fiscal authority seeks to pursue an output target. Cukierman (1992) gives only a very cursory analysis of fiscal policy. Levine and Brociner (1994) consider the interactions between two fiscal authorities in a monetary union, but do not address the nature of relations between monetary and fiscal authorities.

Our concern is with the situation which arises when a monetary authority seeks to control inflation, while a fiscal authority tries to maintain a budgetary, rather than an output target. This is of obvious interest given the policy structure likely to evolve in Europe as a consequence of the Maastricht treaty. Sargent and Wallace (1981) demonstrated the inter-relationship between monetary and fiscal policy arising from the government's budget constraint; this does not rule out the possibility either that the dynamic policy response may be very different or that the long-term equilibrium may differ when there are two independent authorities from the case where monetary and fiscal policies are set jointly. Meade (1990) was concerned about the disturbances which might be generated by a lack of co-ordination between the financial authorities even when they were in full agreement about policy targets; in his example (see also Meade and Weale, 1995), each authority set policy on the assumption that the behaviour of the other would continue unchanged and the analysis was simplified by the assumption that private sector behaviour was exogenous. Beetsman and Bovenberg (1997) have looked at the strategic issues which can arise between the different authorities, but only in a two-period context; in their model it is not possible to investigate what might happen if each authority is unsure how to anticipate the responses of the other.

In this paper we consider, in a fully dynamic context, the problems generated by interactions between a government which pursues a budgetary target and a central bank which sets monetary policy to control inflation. We are able to investigate the consequences of a period of learning which may arise if each authority is not ex ante sure of the responses of the other.

2. Co-ordination problems

The problems generated by co-ordination (or the lack of it) between the central bank and the government can be illustrated diagrammatically, at least as a means of indicating the questions involved. This provides a starting point for our subsequent …