Non-farm payrolls rose by 162,000 versus widely dispersed expectations that varied from -40,000 to +360,000. This was a particularly challenging number for the economists to predict (though I suppose none are easy) due to census hiring and a rebound from uncertain weather effects in the previous month. However, it appears that after sterilizing the numbers for census hiring, private sector jobs grew by 123,000. We should point out that the birth/death model added 81,000 of these jobs, but that seems a little petty. This may account for some of the strangeness when comparing March non-farm payrolls to the ADP numbers from Wednesday, which showed private jobs falling by 23,000. Nonfarm payrolls showed a gain of 41,000 goods-producing jobs, while ADP reported a decline of 51,000 goods-producing jobs. Interesting.
The unemployment rate, though appearing in the headlines to remain steady at 9.7%, actually rose from 9.687% to 9.749% (which reminds us of this recent “Quadrophobia” story). The rate seems to have been driven higher by a decline in the number of people not in the labor force, many of which moved into the unemployed category. Ultimately, more people in the labor force is a good thing. The U-6 unemployment rate, which includes workers who are discouraged/marginally attached/part-time for economic reasons continued its climb, rising to 16.9% from 16.8% in February.
These jobs numbers, they remain a bit of a mystery.
In an effort to shed some light on the employment situation, we will borrow a graph from a recent Cleveland Fed paper by Murat Tasci that asks the question “Are Jobless Recoveries The New Norm?” Among other topics, the article addresses a simple idea: the unemployment rate is a level, not a flow. An unemployment rate of 9.7% represents a snapshot in time and tells us the number of unemployed people as a percentage of the labor force. Weekly initial jobless claims are a flow, but only a flow in one direction (into unemployment, obviously). It’s like measuring the volume of water in a pool; you need to know the rate at which liquid is being added, but also the rate at which it is being drained away. This was a very roundabout way to get to the point that the unemployment rate is also dependent on the outflow of workers from unemployment, meaning workers finding new jobs.
I’ll let Mr. Tasci define a few terms for you:
“In any given month, some unemployed workers find jobs and some employed workers lose theirs, leading to a flow of workers out of and into the unemployment pool. We define the average rates at which these flows occur as job finding and separation rates, respectively.”
Now let’s take a look at the chart.
Spikes in the unemployment rate tend to happen because of both increases in inflows into unemployment (a rising separation rate) and decreases in outflows from unemployment (a decreasing job finding rate). What is so interesting about job losses in the previous three recessions is that the driving force seems to be the impressive decline in the job finding rate, meaning that workers are staying unemployed for longer. The separation rate has reversed course recently, which is showing up in the falling initial claims numbers, but we have yet to see a material rebound in the hiring activity.
Compare the above chart with another favorite of ours, jobs lost and the pace of recovery in past recessions:
During the previous three recessions, in which unemployment was largely driven by the inability of people to find jobs, the pace of the jobs recovery has been slow and U-shaped instead of quick and V-shaped. While it is possible that today’s nonfarm payrolls release marks the beginning of a jobs recovery, and the sheer size of the government stimulus during this recession (as opposed to the prior two) will change the shape of that recovery, these possibilities have yet to manifest themselves in the data.