Saturday, November 3, 2012

October Jobs Report: First Impressions


Posted: 11/02/2012 9:05 am

Well, the big jobs report is out showing payrolls grew by a more-than-expected 171,000 last month and the unemployment rate ticked up slightly, as expected, to 7.9%. Job growth for the prior two months was revised up by 84,000, and the average monthly pace of job growth over the past four months-a useful way of smoothing out monthly noise in the data-is 173,000, a sharp acceleration over the second quarter's pace of 67,000 per month (see figure).
The uptick in unemployment was expected after September's 0.3 percentage point drop, but a few things are worth noting. First, the 0.1 point increase is statistically indistinguishable from no change at all-the unemployment rate has to rise or fall about 0.2 points to be significant. At 7.9%, the jobless rate is down significantly-by one full point-from its rate one year ago. Second, one reason for the slight uptick was more people coming into the labor market seeking work. We'll need to see how this development evolves in coming months, but we may be seeing early signs of an improving job market pulling more job seekers in from the sidelines.
All told, given the acceleration in payroll growth, the upward revisions to prior months payroll gains, the trend decline in unemployment, and the pick-up in labor force participation, today's report is generally pointing to job market that's showing signs of improvement.
Obviously, a report like this just a few days before a tight election is going to be a very big deal, and both campaigns will use the results in predictable ways. But if there's anyone out there who's making up their mind based on this one report, please don't. Yes, the monthly employment numbers provide important information about the part of the economy that matters most to people, but that information must be considered as but one relatively noisy set of indicators amid a sea of others.
I always stress, as I did above, the importance of smoothing out some of the monthly noise by averaging over the past few months. As noted, employment growth slowed notably in the second quarter of this year, increasing by only 67,000 jobs per month, but has since accelerated up to an average monthly gain of about 170,000 over the past four months.
Some details:
  • - The effects of hurricane Sandy are not in these numbers as the October surveys were fielded well before the storm hit.
  • - Manufacturing added 13,000 jobs last month, after shedding 14K and 13K jobs in the prior two months. Over the past year, factory employment is up by 189,000, and up about 500,000 since the sector began to recover in early 2010.
  • - Both hourly and weekly earnings are up, before inflation, by about 1.5% over the past year, trailing the recent trend in prices, up around 2% since last September. Average weekly hours have also been flat over the past few months, suggesting employers are meeting increased labor demand by adding workers rather then extending shifts.
  • - Most industries added jobs last month; professional services led with 51,000 jobs, retail stores added 36,000 jobs in October, compared to 27K and 18K in the prior two months, perhaps reflecting stronger consumer activity and confidence.
  • - Construction was up 17,000 last month, driven by both residential and commercial contract work. The sector is showing some early signs of the formerly moribund housing market coming back to life.
  • - Government employment, however, fell again in October, down 13,000. After posting large losses since the downturn, over the past state and local payrolls have been essentially stagnant.


This post originally appeared at Jared Bernstein's On The Economy blog.

On next blog post

Data Notes Part 1: Are Sideliners Coming Back Into the Labor Market?

On shpilkes before the election so a good time for some data work.  Part 1, here, is a deeper dive in the job market.  Part 2 is a little bit of Sandy economics, with an emphasis on net vs. gross domestic product.
An interesting dynamic from the Household survey from yesterday’s report was the 0.1% tick-up in both the unemployment rate and the employment rate, a function of more people coming into the job market looking for work than actually found jobs.* Even though the numbers are large (578K growth in labor force; 410K growth in jobs), small ticks in the rates like these are statistically insignificant.  Still, the patterns are nevertheless worth exploring to see if there’s any sign that the improving labor market is pulling more folks into the labor force, a good thing given that the labor force participation rate has been depressed of late.
The flows data are helpful is this regard.  The first chart shows the flows into unemployment.  As you can see, the rate is falling largely due to people getting out of unemployment, i.e., moving from unemployment into either jobs or out of the labor force.  If you squint you can also see a bit of a downward move in people moving from employment to unemployment, a sign of diminished layoffs.  Also, as noted, last month’s tick up in unemployment from 7.8% to 7.9% was a function of more people looking for work.  In fact, the flows data show that people coming from out of the labor market into unemployment added 0.2% to the rate in October.

Source: CPS Flows data.
Do the flows data shed any other light on this question of folks coming off the sidelines and getting into the improving job market?

One suggestive trend would be if fewer of the unemployed were dropping out of the labor force altogether.  That’s certainly the paradigmatic person we have in mind here, right?  They’re stuck in unemployment, get discouraged with their lack of opportunities, and check out for a while.  Well, as the figure shows, after growing steeply in the downturn, over the past few months that group has been getting a bit smaller (solid line is a 6-month moving average).

Source: CPS Flows data; solid line is 6-mo moving avg.
So, some rough, early indication that if current trends continue, the improving job market will likely pull more folks back in.  Like last month, that may put some upward pressure on the jobless rate, but in the longer term, it’s a positive development.
*I was interested in whether there’s a correlation between this type of dual uptick in unemployment and employment rates and job growth.  It turn out that in months when this occurs, civilian employment grows 0.3% faster than otherwise, all else equal.  (I regressed the log change of civilian employment on its own lags and a dummy variable for months when both unemp and emp rates increased.)

No comments:

Post a Comment