Average Inflation Targeting and Expectations

“If credible and clearly articulated, an average inflation targeting regime would provide a better anchor for inflation expectations than a period-by-period inflation targeting regime. Alas, the Fed’s new policy strategy has not been clearly articulated, leaving short-run inflation expectations unanchored.” ~ William J. Luther

inflation targeting
inflation targeting

In late August, the Federal Reserve revised its Longer-Run Goals and Monetary Policy Strategy. It had previously committed to targeting two percent inflation, which it described as a symmetric target. In fact, the Fed seemed to treat that target more as a ceiling than a symmetric target, as it rarely allowed inflation to exceed two percent. And, as a consequence of continually missing its target, market participants came to expect that the Fed’s actual target was something less than two percent.

Now, the Fed says, it will aim to achieve an average inflation rate of two percent. That means that, “following periods when inflation has been running persistently below 2 percent, appropriate monetary policy will likely aim to achieve inflation moderately above 2 percent for some time.”

What is Average Inflation Targeting?

A central bank engaged in average inflation targeting (AIT) conducts monetary policy to ensure that inflation averages its target rate (e.g., two percent) over some period of time (e.g., a decade). The central bank does not aim for two percent inflation each period, as it would under a traditional period-by-period inflation targeting (IT) regime. Instead, it adjusts its next-period target each period to account for past mistakes. If it previously undershot its target, it will have to aim at an inflation rate greater than its average target until the average inflation rate returns to its target. If it previously overshot its target, it will have to aim at an inflation rate less than its average target until the average inflation rate returns to its target.

In the absence of errors, AIT and IT are equivalent. Each period, the central bank aims for and hits its stated target. The two regimes only differ if the central bank occasionally (or persistently) misses its target. George Selgin provides an eloquent explanation:

…suppose that the economy experiences a series of unexpected shocks that put downward pressure on prices. These could be either negative demand (velocity) shocks or positive supply shocks. Suppose as well that those shocks aren’t matched by corresponding shocks of the opposite sort. The Fed, in other words, faces a “run” of negative price level surprises. In that case, although it continues to set its rates at levels calculated to keep the forward-looking inflation rate at its 2 percent target, the ex-post or “backward-looking” inflation rate will be persistently below that target; and the longer that run continues, the larger will be the gap between actual and intended inflation.

Hence, the Fed’s new AIT regime seems intended to remedy the problem of ex-post inflation persistently falling below two percent, as was the case under its prior IT regime.

IT, AIT, and Long Run Inflation Expectations

A good monetary rule will, among other things, anchor long run expectations. Borrowers and lenders must estimate inflation when financing long term projects. Employers and employees must estimate inflation when agreeing to long-term labor contracts. When inflation is higher than expected, borrowers and employers gain at the expense of lenders and employees because borrowers and employers get to make payments with dollars that are worth less than was expected when the contracts were executed. When inflation is lower than expected, lenders and employees gain at the expense of borrowers and employers because lenders and employees receive dollars that are worth more than was expected when the contracts were executed.

The possibility of transfers due to unfulfilled inflation expectations make long-term contracting risky—and that risk will discourage some long-term contracting. The greater the possibility—and the greater the size of the possible transfers—the bigger the risk. And, the bigger the risk, the more likely it is that long-term contracting is discouraged. Since, in the absence of such risks, long-term contracting would be preferred for many ventures, the possibility of transfers makes us less productive than we otherwise would be.

An IT regime does not anchor expectations very well. Consider the range of possible outcomes if a central bank targeting two percent inflation might miss its target by +/- 0.5 percentage points each period. After ten years, the price level will be somewhere between 16 and 28 percent higher than it was when the contract was signed. After thirty years, the price level will be between 56 and 109 percent higher than it was when the contract was signed. That’s a big range of potential outcomes, and a big risk for those considering a long-term contract.

By requiring the next-period target to be adjusted each period to account for past mistakes, an AIT regime has the potential to reduce that range. In ten years, the price level is much more likely to be 22 percent higher than it was when the contract was signed, as expected. In thirty years, it is more likely to be 81 percent higher, as expected.

IT, AIT, and Short Run Inflation Expectations

While a credible AIT regime provides a better anchor for long-run expectations, it might provide a worse anchor for short-run expectations if not clearly articulated. Under a credible IT regime, forming expectations over the next period is relatively easy because the central bank is aiming at the same rate each period. Under a credible AIT regime, one is potentially left guessing what the next-period target is.

The Fed’s AIT regime has the potential to destabilize short-run expectations because it has not been very clear about its makeup strategy—that is, how it will adjust its target after missing its target. Stephen Williamson explains the issue succinctly in a short tweet thread response to Tony Yates, unrolled here:

The key to a makeup strategy is that you should know at each point in time what the inflation target (or whatever target) is for the immediate future. With average inflation targeting, you need to know what you’re trying to make up, which requires knowing how far in the past you go to determine what average past inflation was. Basically it’s a moving average (could be weighted) with a specified horizon. Then you have to specify how fast you’re going to make up for missing on the high or low side in the past. That could be a specification for a future horizon over which you expect to make up the difference. So, all that will give you a time varying target for the next month’s or next quarter’s inflation. It’s complicated to explain, which is a good reason not to do it. So either you do it, and bear the costs of explaining it. Or you don’t do it, and stick to what you have. The Fed did neither, which is bad.

Whether the Fed will take steps to clarify its makeup strategy remains to be seen. But, as Williamson notes in another tweet, it is apparent “from reading the financial media that people are struggling to understand what the Fed’s new policy statement means.” That’s not good.

If credible and clearly articulated, an average inflation targeting regime would provide a better anchor for inflation expectations than a period-by-period inflation targeting regime. Alas, the Fed’s new policy strategy has not been clearly articulated, leaving short-run inflation expectations unanchored. And, considering that the Fed consistently undershot two percent inflation under its previous inflation targeting regime, one might reasonably question how credible its commitment to this new policy strategy is, as well.