What will the Fed do? (Sept. FOMC update)

After last night’s post, I got emails accusing me of being a two-handed economist. While anatomically correct, I take the point, so I’ll add this note saying that my view of likely policy is where it was in the Jackson Hole update.

A quick summary: I’ve been arguing that the consensus of the FOMC, as nurtured by the Chair, has followed a policy of steadfast commitment to gradual tightening, tempered by a willingness to take a tactical pause if credible evidence arises that job gains might soon falter. A more compact summary: Do it (slowly) until the economy says ‘No.’

The FOMC has been in a tactical pause since early this year, but by Jackson hole it looked pretty clear that—by the standards of how promptly the FOMC ended earlier tactical pauses—a rate increase would soon be warranted. As for the Sept. or Dec. question, I still feel this way:

[T]he timing of individual moves doesn’t matter much in terms of the Fed’s dual mandate. Predicting stuff that doesn’t matter is a sucker’s game, and I’ll pass. I know that this doesn’t help those of you who are eager to place bets on this question. Others can hedge the modest timing risk and focus on stuff that really matters.

Forced to give a probability on this question, my view is that a rate increase is very likely by year end (absent disturbing new data). As for the timing, I’ve hovered near 50 percent: I don’t have a clear sense of the minor considerations that might tip the balance.[1] These views have not changed.

What is the most likely scenario in which this view is wrong? Setting aside disturbing data, which obviously could and should change policy, my main caveat is this. While nonfarm payrolls and the headline unemployment rate continue to be consistent with one more interest rate step, other measures of the labor market are less encouraging.[2] The FOMC has always emphasized a holistic look at many indicators, but payroll gains and the headline unemployment rate have been the first among equals in communication on behalf of the consensus. If worries over these other measures were taking on greater importance, Jackson Hole would have been a natural place to surface that idea. Instead, Chair Yellen said the following:

Smoothing through the monthly ups and downs, job gains averaged 190,000 per month over the past three months. Although the unemployment rate has remained fairly steady this year, near 5 percent, broader measures of labor utilization have improved.

Not exactly sounding the alarm, and no hint of an emphasis shift toward worrisome aspects of broader measures.

Might the FOMC, nonetheless, cite these other measures as a reason to wait until December? Certainly, just about anything could justify that. Such a delay could be communicated in much the same way that communication after last September’s delay in liftoff was taken to mean that a December liftoff was very likely (absent an ugly turn in the data). If, however, the FOMC uses these other labor market measures as a reason for more than a modest change in timing, it would look to me like a poorly communicated, dovish turn in the intentions of the consensus.


1. My own belief, for what it’s worth, is that if the consensus has decided that rate increase is very likely by year end, I would do it now. But there will be 17 opinions on the lesser issues that will determine this timing question. If I were on the Committee, I wouldn’t push very hard for one over the other. [back]

2. Improvement in both payrolls and the unemployment rate have slowed, but essentially everyone on the FOMC expected this. Many would have been disturbed if this did not happen. However, U-6 is still elevated and now sits only one-tenth below last October. Total hours growth has slowed considerably. [back]