Macro Policy
No U Turns

The issue of credibility has been an important element in recent discussions of government policy. In macroeconomics, credibility problems arise naturally when the effectiveness of government policy depends not only on current policy, but on expectations of future policy as well. In a wide range of theoretical models, the government has an incentive to announce one type of future policy but to follow another. The implications for the credibility of such announcements are obvious.

A common example is monetary policy. If a policy of zero money growth, now and in the future, is announced in an inflationary situation, it may reduce current inflation partly by reducing expectations of future inflation. Wage demands, too, are likely to be smaller if workers anticipate low rates of inflation over the near future. But once expectations of inflation are reduced, the government has less incentive to sustain the policy, and it may even be tempted to use monetary expansion to stimulate output and employment. The initial announcement of zero money growth may therefore lack credibility unless the government can somehow commit itself to carrying out its announced policy. Otherwise workers might rationally demand wage increases consistent only with higher rates of inflation than the government claims will occur.

In Discussion Paper No. 63 David Backus and Research Fellow John Driffill explore the twin themes of commitment and credibility, first as a general theoretical problem, and second in the context of disinflationary policy. They approach the problem as a 'game' between the government and a decentralized private sector. The game is a repeated one, and at every stage each player can choose among alternative actions. Each player is assumed to make a rational choice based on his knowledge of how his actions affect the other player.

Backus and Driffill compare the effectiveness of policies in two distinct situations: when the government is able to commit itself to some pre-announced plan, and when it cannot. The ability to make commitments, is valuable, of course, since it makes announcements of future policy credible. In an illustrative numerical example, the authors find that the plan without credibility achieves disinflation at the cost of a more severe recession than does the optimal plan when commitment is possible.


Credibility and Commitment in Economic Policy
D Backus and J Driffill

Discussion Paper No. 63, June 1985 (IM)