Policy

How central bankers can better control inflation

By Robin Mordfin     
February 25, 2014

From: Magazine

Janet Yellen testifies at Senate Banking Committee nomination hearing for Fed Chairmanship. Photo by Andrew Harrer/Bloomberg via Getty Images.

Since the 1990s, when several countries adopted inflation targeting, communication has been recognized as a key tool for central banks. Assistant Professors Kinda Hachem and Jing Cynthia Wu have been studying those communications, and how market participants react to them. The researchers find that when trying to significantly reduce the rate of inflation, banks must make announcements that gradually lower the expected inflation rate. But the reverse is true when combating deflation: in that case, banks must take an aggressive approach.

The researchers created a computer-simulated model that uses tournament selection to predict market reactions to central-bank announcements. The model tracks 1,000 imaginary firms that must establish prices for their goods. The firms, all competing with each other for a greater share of household budgets, make decisions after the central bank releases its inflation predictions, but before knowing the actual inflation rates. Each firm either relies primarily on the bank’s predictions or on commercial forecasts.

The model uses an algorithm to determine the credibility of each forecasting source. Credibility is an essential part of Hachem and Wu’s research because if it were easy to persuade firms to rely only on a central bank’s inflation predictions, all decisions would be based on bank announcements and inflation targeting would be much simpler. But central banks, due to their own forecasting inaccuracies, can lose credibility. Their communication becomes less relevant when there is prolonged divergence between what is communicated and what is achieved.

Central to the model’s credibility-building algorithm is stubbornness. Say, for example, firm A has a forecasting source that turns out to be wrong. Firm A could meet with firm B, whose forecasting source turns out to be right. If firm A has a stubbornness level of one, it will switch loyalties and become loyal to firm B’s source. But if firm A has a stubbornness level of four, it will have to meet with four firms that use an accurate forecasting source before switching loyalties.

The researchers’ model suggests that the pacing of target announcements is crucial when firms are stubborn. Hachem and Wu find that when central banks are trying to combat high inflation, abruptly introducing a low inflation target often causes actual inflation rates to eventually fall below the target rate. When that same low inflation target is introduced gradually, through a series of interim announcements over time, the bank builds credibility more quickly and inflation goals are reached more smoothly.

When the goal is to raise inflation, however, the research suggests a bank should be more assertive, even if that causes firms to temporarily question the bank’s credibility. Banks can eliminate deflation more quickly by communicating an ambitious increase in their short-term inflation goals. That can induce central-bank followers to set high prices for goods, which pushes realized inflation upward.

This creates a gap between realized and targeted inflation, the researchers acknowledge, which in turn hurts the bank’s credibility. So central banks have to gradually lower their initially aggressive goals to regain credibility once inflation has sufficiently risen.

Hachem and Wu have uncovered further empirical evidence to support their thinking. For a sample of 19 countries where inflation targeting was adopted to lower inflation, the researchers compare central banks’ inflation targets with actual inflation data from the International Monetary Fund’s International Financial Statistics database (see graph). The evidence demonstrates that when central banks abruptly announce low inflation targets, they often cause actual inflation rates to overshoot—but when they gradually announce inflation targets, they are able to avoid this.

Work cited

Kinda Hachem and Jing Cynthia Wu, “Inflation Announcements and Social Dynamics,” Working paper, January 2014.