What happens if the output gap is wrong?

One of the key foundation for the current monetary easing in the US is that there will be persistent deflation risks because we have a large output gap. Chairman Bernanke is a strong proponent of this view, but there are others who are arguing hat the gap is much smaller. hence, there is a risk of higher inflation if growth starts to pick-up.

From a speach by St Louis Fed president James Bullard.

First, the key is the QE program.

On current monetary policy, “the key issue is how to think about the asset purchase program,” Bullard said. “Liquidity programs are shrinking, but the asset purchase program is only partially complete.”

Second, the output gap issue. The link is weaker than we think.

“I am concerned about a popular narrative in use today—the narrative being that the output gap must be large since the recession is so severe,” he said. “And so, any medium-term inflation threat is negligible, even in the face of extraordinarily accommodative monetary policy. I think this narrative overplays the output gap story.”

He added that measuring the gap is very difficult, both theoretically and practically. He cited research that shows much of the inflationary run-up in the 1970s can be attributed to a misreading of the output gap at the time.

“Even if economists were to accept a particular measure, the empirical relationship with inflation is not robust,” he said. In addition, traditional output gap measures do not account for the concept of bubbles.

“It has been popular to describe recent events as a collapse of a bubble in housing. A look at the housing data makes a convincing case,” Bullard said. “But when it comes to calculating traditional output gaps, there is no notion of a bubble. If part or most of the fall in output was a collapsed bubble, then today’s output gap would be smaller than it appears.” This would mean that inflation risks in the medium term are higher than otherwise thought.

The inflation may be a greater issue than expected over the medium term.