From Jan Hatzius at Goldman Sachs:

The March employment report was a disappointment. Although the unemployment rate fell, this was due to a drop in the labor force as household employment gave back some of its prior big increases. More importantly, the job gain in the establishment survey of just 120,000 fell well short of anyone's estimate. The big question is how much of the slowdown from February’s 240,000 gain was due to special factors, including “payback” for the unseasonably warm winter, and how much reflects weakness in the underlying trend.

We do think the warm weather has been an important driver of stronger payroll numbers over the past few months. As we have shown, all of the acceleration in nonfarm payrolls since the fall has occurred in the (normally) cold states, and our state-by-state panel analysis suggests that weather has boosted February’s level of payrolls by 100k or a bit more (see “Payroll Payback?” US Economics Analyst, 12/14, April 5, 2012). This state-level model suggests that none of the inevitable payback for this boost should have occurred yet, since March was just as warm relative to the seasonal norm as February. That said, weather-sensitive sectors such as mining and building construction did show some weakness, so we would pencil in 10k-20k for weather “payback” in March.

In addition, the 37,000 drop in retail employment was partly related to one-off job reductions in the department store industry, and should probably not be included in an estimate of the underlying employment trend. Taken together, we believe that the underlying trend in payroll employment growth is around 175,000 as of the March report. At this point, we would expect the headline number for April to fall short of this figure, partly because the weather payback is likely to be substantially larger in April than in March and partly because the underlying trend may be decelerating slightly (as suggested, e.g., by the drop in temporary help services employment in March).

Largely because of the weakness in the employment report, our standard metrics for evaluating the US data flow have also started to send a less upbeat message. Our current activity indicator (CAI), which summarizes all of the key monthly and weekly activity data, is showing a preliminary 2.5% for March, down from 3.5% in February. Likewise, our US-MAP, which compares the data with the Bloomberg consensus, has averaged negative readings since late February, after six months of positive surprises. All this reinforces our view that the discrepancies in the US economic data will be resolved mainly via deceleration in the job market indicators rather than acceleration in GDP.

We admit to being puzzled by the twists and turns in Fed communications over the past few months. On January 25, Chairman Bernanke said that under the FOMC’s projections, he saw a “very strong case” for finding “additional tools” to support economic expansion. But in the March 13 minutes, only “a couple” (i.e., two) of the committee’s ten voting members—a number so small that it probably does not include the chairman, whose position makes it unlikely that he would be in such a small minority—thought that additional stimulus could become necessary, and even that only “if the economy lost momentum” or inflation looked likely to undershoot. All this would make perfect sense if there had been a sharp upgrade of the committee’s central forecast over the past few months. But the minutes also said that “…the economic outlook, while a bit stronger overall, was broadly similar to that at the time of their January meeting.” And Chairman Bernanke, in particular, last week went out of his way to cast doubt on the not on that the stronger jobs data through February were indicative of a sharp pickup in growth. Our conclusion is that there has been a shift in the Fed's reaction function back to the hawkish side, and there may be a bit more complacency about the risks to the outlook than suggested by the committee’s decision to retain the assessment of “significant downside risks” in the March 13 statement.

So what can we expect from the Fed? Easing at the April 24-25 meeting looks highly unlikely, although the tone of the statement and the Chairman’s press conference may take a fresh turn toward the dovish side. Easing at the June 19-20 meeting, in contrast, still looks more likely than not, at least under our forecast of weaker activity and benign inflation. Our baseline remains a renewed asset purchase program which involves Treasuries and MBS and whose impact on the monetary base is sterilized via reverse repos or term deposits, but it is also possible that the committee would extend Operation Twist; there is approximately another $200 billion available, and it would only take a small reduction in the flow of purchases to make this number last until yearend.

Stepping back from the tactics, we still see a strong fundamental case for following up Operation Twist with a successor program. First, even under its own forecast, the committee expects to be far from fulfilling the employment side of its mandate by 2013-2014, so it is easy to sympathize with Chicago Fed President Evans’s call for more action. Second, growth could well disappoint the committee’s forecasts, given all the usual uncertainties around the weather impact, the inventory cycle, energy prices, and the “fiscal cliff” at the end of 2012. Third, a failure to do more might imply a tightening of conditions, assuming financial markets are still discounting some probability of easing. In addition, we have found some evidence that at the very long end of the yield curve, where Operation Twist is concentrated, it may be not just the stock of securities held by the Fed but also the ongoing flow of purchases that matters for yields. And fourth, the risk of a material inflation overshoot seems low given the still-large amount of spare capacity, not to mention the Fed’s ability to reverse course and tighten financial conditions substantially via forward guidance, rate hikes, or even asset sales should the need arise.