The credibility or otherwise of government policies is often cited as a cause of the high cost of reducing inflation, in terms of potential output lost and high unemployment. In the United Kingdom during the early 1980s, for example, nominal wages continued to increase rapidly: firms and workers did not believe that the government would sustain downward pressure on inflation through tight monetary and fiscal policy and a high exchange rate. The reason for this credibility problem is well known. In the long term high inflation does not bring any advantage which could not be achieved with price stability; in the short term, however, a fiscal or monetary stimulus may temporarily reduce unemployment and increase output at the price of increased inflation. The private sector's anticipation that a government may yield to political pressures to reduce unemployment demonstrates that the government's stated commitment to low inflation lacks credibility. A government may therefore believe that it is worth acquiring a reputation for being so committed. This reputation will reduce the private sector's expectations of inflation and will thereby raise the level of output and employment consistent with low inflation.

In Discussion Paper No. 159, Research Fellow John Driffill examines the implications of uncertainty about the government's true policy objectives. His earlier work with David Backus has shown that the government can overcome its lack of credibility if the private sector is uncertain about whether the government is 'committed' to achieving zero inflation or seeks to maximize an objective function containing both inflation and output. In each period the government has a 'reputation': the probability with which the private sector believes that the government is committed to zero inflation. As time passes the private sector revises this probability in the light of government policies: this gives the government an incentive to sustain zero inflation. Since a committed government is known never to inflate, and since the private sector perfectly monitors the government's actions, the observation of inflation immediately reduces the government's reputation to zero. The cost associated with losing one's reputation in this way is sufficiently great to deter the non- committed government from causing inflation, mimicking the behaviour of a 'committed' government until near the end of the game, when the incentive to maintain a reputation disappears. The main result from this analysis is that the private sector's uncertainty concerning the government's objectives may induce an uncommitted government to pursue the policy of a committed government.

Driffill then extends his analysis to consider a more general case where the private sector is convinced that the government is 'caring' (about unemployment and lost output) but is uncertain about the relative weights it attaches to inflation and output. He finds that the desire of a 'more caring' government to maintain credibility can lead it to mimic the actions of a 'less caring' government, so that reputation continues to discipline government behaviour. This can break down, however, if the government is so 'caring' that it is not prepared to mimic the actions required to achieve low or zero inflation.

Finally, Driffill considers a model in which random shocks affect the economy and prevent the government from exercising perfect control. As a result the private sector cannot determine the government's intended action: is an increase in inflation the result of deliberate policy or a shock to the economy? Driffill finds that this reduces substantially the discipline on policy- makers which reputational considerations would otherwise impose.