Editor's note: This article was originally published on April 3, 2015 by Menzie Chinn here.
The March establishment series estimate for nonfarm payroll employment clearly fell far short of consensus estimates (release). Here are a couple of quick observations.
Not only was overall employment growth preliminarily estimated to be low; earlier months' estimates were also revised down.
Figure 1: Nonfarm payroll employment in '000s, seasonally adjusted, from March release (blue), from February release (red), and from January release (green). Source: BLS.
One should never put too much weight on one month's preliminary figures. It's important to keep in mind that the mean absolute error of the m/m changes going from 1st release to 3rd release is 45,000.  Still, lackluster employment growth in this month reinforces my view that we should not be overly concerned about an overheated labor market. In my view, we are still some way from full employment.
To me, the more interesting aspect of the release involves the sectoral breakdown. Some have pointed out the drop in leisure and hospitality employment, which seems unlikely in their view to be sustained. For me, the more worrying aspect is the decline in manufacturing employment over time. In 2015 M03, the change is -1 thousand (so doesn't show up in Figure 2 below).
Figure 2: Month on month change in nonfarm payroll employment minus leisure and hospitality and manufacturing, in '000s, seasonally adjusted, from March release (blue), manufacturing (red), and leisure and hospitality (green). Source: BLS and author's calculations.
This suggests worrying trend in tradables sector economic activity, and hence continued concern about the dollar's value.