Friday, July 21, 2006

How to Decentralize Monetary Policy

Today's Wall Street Journal reports:

Federal Reserve policy makers raised interest rates last month in part because markets expected them to do so, and they figured failure to act might hurt their credibility as inflation fighters, minutes of the meeting suggest.
Some people might view this response as wimpy--doing what financial markets want rather than showing real leadership. But one can view this approach as a step toward decentralizing monetary decisionmaking.

Suppose the Fed has a long-term inflation target. And suppose the Fed followed this rule:

Look at the market's forecast of interest rates and inflation over the next few years. If the market expects inflation above target, set a path for interest rates a bit higher than the market expects. If the market expects inflation below target, set a path for interest rates a bit lower than the market expects. If the market expects inflation to come in on target, set a path for interest rates equal to what the market expects.

This might seem circular: The Fed is responding to the market, and the market is responding to the Fed. But there is nothing wrong with that. Economists are used to simultaneity.

Of course, the market will catch on to the policy, but that's okay. In fact, it is ideal. We end up in a fixed-point equilibrium in which the market expects the Fed will hit its inflation target. In this equilibrium, the market's forecast of interest rates will tell the Fed what it needs to do to accomplish what it wants to accomplish.