What to Target? Insights from a Lab Experiment

We compare the central bank’s ability to stabilize inflation under three different approaches to monetary policy using a laboratory experiment. In this experiment, participants are tasked with predicting inflation over time in an artificial economy governed by a simple economic model. In this economy, the central bank adjusts the interest rate based on changes in inflation, which result from:

• external shocks that tend to lower inflation
• the participants’ predictions of inflation

The three monetary policy regimes are:

• inflation targeting—which seeks to achieve a given inflation rate in each period
• average inflation targeting—which targets the average of past inflation rates over a given period, allowing inflation to temporarily rise above its target to make up for previous periods of below-target inflation
• price level targeting—which seeks to keep the price level on a growth path that is consistent with the inflation target

We find that average inflation targeting allows the central bank to stabilize inflation around its target almost as well as price level targeting, which is not a plausible approach. This result holds because participants can coordinate and form inflation expectations using simple rules that incorporate past inflation rates.

But average inflation targeting is sensitive to the exact specification of the policy framework. In particular, if the central bank seeks to smooth inflation over a longer period, it becomes more difficult for participants to form expectations consistent with the average inflation over the entire period. This results in more inflation volatility.

For inflation targeting, we also explore the effects of greater communications around the target. We find that increasing communication around the target—especially during shocks that temporarily move inflation away from its target—seems to have only a modest effect on anchoring inflation expectations and only in the longer run.