The Federal Reserve says it aims for 2 percent annual inflation. It is considering shifting to a target of 2 percent average inflation over the business cycle. If in one year it missed the mark and inflation runs at 3 percent, then it would aim for 1 percent inflation the next year. Matthew O’Brien explains the thinking behind the idea in the Washington Post. Among other things, an average target would make it easier to predict what the price level will be in a few years.
But there’s an ambiguity worth thinking about, highlighted by O’Brien’s comment that an average target “would mean, for example, that the Fed should be doing a lot more to stimulate the economy right now.” There’s a good case for the Fed to announce that from now on it will correct course when it misses its target – for the Fed to say that it won’t “let bygones be bygones” going forward. But there’s no good case for the Fed to raise inflation now in order to correct for undershooting inflation over the last decade, at a time when wasn’t targeting an average – no good case, that is, for imposing the policy retrospectively.
Let’s say the Fed adopts the policy now. So if today’s price level is 100, it will aim for 102 in 2020, 104 in 2021, 106 in 2022, and so on. Let’s say, further, that it undershoots the target and the price level in 2020 ends up being 101. The case for targeting average inflation is the case for getting back to that path (104 in 2021, 106 in 2022, etc.) rather shifting to a permanently lower level for all future years (103 in 2021, 105 in 2022 . . .) because of one year’s mistake. By having a memory of how it has missed one year’s target, the Fed sticks to its target over the long run.
We wouldn’t get any advantage, on the other hand, if the Fed decided today that it will try to correct for undershooting inflation from 2009 through now. Trying to undo the shortfalls of the last decade — which have accumulated to about 7 percent, as O’Brien notes — wouldn’t give us any added ability to predict the future price level. Boosting inflation significantly for that purpose would reduce monetary stability rather than increase it.
If the Fed wants to adopt this policy, the right way to do it is probably to say that it will apply it starting in the next recession. That’s when to start making memories, so to speak. And if there’s a case for looser money at this moment, the need to shift to average targeting isn’t it.