Unemployment Rate Drops

Just a quick note on this morning’s jobs report: the US unemployment rate dropped below 8%, it was 7.8% for September, down from August’s 8.1%, and the lowest we have seen since January 2009.

The economy added 114,000 new jobs during the month, a rate, which as I have said many times, not sufficient to hold unemployment down for too long. Indeed the two headline numbers – the unemployment rate and the number of new jobs – appear inconsistent. That low level of new job creation is not usually enough to bring the unemployment rate down, least of all by 0.3%.

So there is something in the details we need to look for.

First, recall that there are two concurrent surveys from which the figures are drawn. A survey of households produces the unemployment number, and a survey of businesses – the establishment survey – produces the job creation number. So it is possible for them to diverge in the story they are telling, and such divergence is quite frequent, especially at inflection points in activity.

Second, the household survey produced one eye opening figure in September: about 873,000 people reported that they had found new work. That is the highest level for this statistic since 1983, and if sustained could mark a dramatic turning point. It also helps explain why we have been seeing better levels in consumer confidence despite many of the aggregate reports on the economy continuing to paint a malaise like picture. Things may be much better than is being currently captured by those aggregate figures. Or, of course, this September surge in work may be a blip. We have to wait for the next month or two to find out.

Third, and also positive, the number of jobs created in both July and August were revised upwards: from 141,000 to 181,000 for July, and from 96,000 to 142,000 in August. This revision, which is not abnormal, also supports the notion that, at a grass roots level, things are improving more rapidly than we would conclude were we only to see the headline numbers.

Fourth, one source of divergence between the unemployment rate and the rate of job creation is workforce participation. If people drop out of the workforce, which has been a major depressing factor during this crisis, their absence from the pool of workers makes the unemployment rate fall even if few new jobs are being created. This effect doesn’t seem to be the cause of the divergence in September. The participation rate looks to have been stable, implying that the apparent inconsistency between job creation and the unemployment rate must have a different cause. Most likely this is simply that the household survey is telling a story in advance of the establishment survey. This is not unusual and is something to watch closely.

Lastly, both the average weekly wage and the average work week rose very slightly, by seven cents an hour and by one tenth of an hour respectively, which also supports the theme of lower level improvement. But only just.

Overall the jobs report is a decent one. The rate of job creation might be understated and thus subject to more revision in coming months.

Either way, today’s report is better than expected. Normally that low level of job creation would elicit damning commentary from me, but the household survey suggests we ought be more upbeat. The message is mixed, but in a good way.