The most important application of the time-consistency ideas in Kydland and Prescott’s work (with due credit going also to Barro and Gordon and Kenneth Rogoff) is to monetary policy. Consider a central bank that wants low inflation and low unemployment. To keep inflation low the central bank promises to hold down the growth rate of money. Let us suppose that the public believes the central bank’s promise and as a result they plan on low inflation in their writing of contracts. At some point, however, unemployment will increase and the central bank will be tempted to juice the economy with a spurt of inflation. Since the public has planned on low inflation a higher than expected inflation rate will be very effective at reducing unemployment and thus very tempting.
But the situation that I have just described cannot be a rational equilibrium. When the central bank promises to keep inflation low the public will say, ‘this promise isn’t credible – if we take the central bank at their word they will surely try to deceive us later with high inflation rates’. As a result, the public does not plan on low inflation and when the central bank does want to reduce unemployment it must increase inflation even more than when low inflation was expected. The only equilibrium of this game is one with high inflation and no systematic reduction in unemployment.
How can we improve the situation? Surprisingly, a nasty central banker can make everyone better off. A nasty central banker cares only about reducing inflation and not at all about reducing unemployment (think fat-cat Republican living off fixed income bonds). Precisely because a nasty central banker won’t juice the economy to reduce unemployment, the nasty central banker can credibly commit to keep inflation low. The public believes the promise and safely plans for low inflation. Unemployment is the same in both scenarios – because the central bank can never systematically surprise the public with higher than expected inflation – but inflation itself is lower with the nasty central banker and thus the public is better off.
Thomas Schelling once described a similar idea this way: If you are kidnapped who do you want in charge of the negotiations, your loving wife or your nasty ex-wife? Easy, right? But suppose that the kidnappers know in advance who will be in charge of the negotiations – now who do you want? See? Sometimes, nasty people do good things.