Friday, January 9, 2015

Are full-time jobs up 427,000 or down 47,000 in December?

Not-seasonally-adjusted full-time is in red.
The latest Employment Situation Report for December 2014 shows full-time jobs either up 427,000 in the seasonally-adjusted measure, or down 47,000 in the not-seasonally-adjusted measure, both from the respective November levels.

Which to believe?

Since 1968 the not-seasonally-adjusted count of full-time jobs between November and December has gone down 33 times vs. 13 times going up, with one year flat (1992). This is consistent with the historical record of cyclicality in full-time vs. part-time.

Full-time typically peaks in the summers and troughs in the winters while part-time does the opposite. Full-time tends to peak in the summers with work related to seasonal and student employment, while part-time tends to peak in the winters with holiday additions to the workforce.  Therefore it is consistent with this pattern to expect part-time jobs to be peaking right now (they already did last month) and full-time to be near its lowest point in the current cycle, which usually happens in January, for which measure we will have to wait another month.

So full-time down 47,000 is obviously more in keeping with the generally expected pattern than the seasonally-adjusted figure.

It is noteworthy, however, how low that negative full-time figure is relative to the recent past and to the historical average.

The 30-year average of the subtractions to full-time between November and December (excluding the outlier years in 2007, 2008 and 2009 when employers panicked and fired 1.4 million on average, 1.7 million on average in 2008 and 2009 alone) is a subtraction of nearly 244,000 full-time jobs.  Add to that that we haven't had this low a subtraction since the year 2000 between November and December and you get the feeling that things are indeed improving.

Unfortunately what we don't see yet is the kind of addition to the full-time rolls which occurs rarely at this time of year and typically after recessions. The last time we saw this in the November-December data was in 2005, 2004 and 2003 when we had three back to back years of full-time gains averaging 290,000, well above the average gain for the 13 up years of 145,000.

What we'd like to see right now, but don't, is a similar strong recovery of full-time after a recession like we've seen in the past.

For example, after the recession of 1970, full-time recovered between November and December of both 1971 and 1972, adding an average of 105,000 full-time jobs for those two months. Similarly after the recession of 1974, full-time jobs recovered for three straight years, averaging an addition of 195,000 full-time jobs between November and December of 1975, 1976 and 1977. And of course after the recession of 2001 we've already pointed out the three years of November-December additions to full-time averaging 290,000, double the average.

Even the long drought of additions to full-time jobs at this time of year which began in 1978 and lasted through 1992 was broken for two back-to-back years in 1986 and 1987 when an average of 66,000 full-time jobs were added between November and December. This was the rather delayed recovery of full-time after the recession of 1982, which cast a long shadow over employment much like the most recent recession has done.

As things stand, the current brutal drought of full-time additions at this time of year now stands at a record nine years, one more than the previous record posted between 1978 and 1985. The average subtraction to full-time then between November and December was 202,000. Now it has soared to 586,000 on average, almost 3x worse.

That's the scale of the trouble we've been in, and so far there's been no sign of leadership out of this mess, except that the pain right now is well-below its average level for this time of the year.

The simple fact remains that full-time is still far below its 2007 peak, no matter how you measure it.