The facts about part-time employment

It's fashionable to argue that there have been more part-time jobs created this year than full-time jobs, and that this is therefore a miserable recovery. But a word of caution is in order: the jobs statistics—like all statistics, especially those that are subject to seasonal variations and revisions, and are sourced using different methodologies such as the jobs statistics are—can be manipulated to prove a point, so you have to take claims like this with a grain of salt. The truth is more likely not as bad as it is made out to be. Here's my attempt to be somewhat impartial.

First, let me briefly note that while all jobs statistics come from the Bureau of Labor Statistics, the data are collected in two very different ways. The Establishment Survey relies on sending out a form to known businesses, and it attempts to adjust for new startups that are not known at the time of the survey. The Household Survey relies on a telephone survey of thousands of households for its data. Both have their strengths and weaknesses. Since the Establishment Survey only attempts to uncover the number of jobs, it tells us nothing about who is not working. The Household Survey, in contrast, asks a random sampling of the population if they are working or not, how many hours they are working, and if they are not working whether they are looking for work or not. That's why the unemployment rate comes from the Household Survey, but the most commonly-referenced estimate of total jobs created comes from the Establishment Survey.

As the above chart shows, the Establishment Survey is much less volatile than the Household Survey, although the two tend to tell the same story over time. For example, so far this year the household survey shows the economy has created 975K jobs, but the establishment survey has uncovered 1,347K new jobs. That's quite a big difference—38% more jobs according to the establishment survey. But from the low point in employment (i.e., the end of 2009), the establishment survey shows 6.67 million new jobs and the household survey shows 6.24 million. That's not a very big difference at all (only 7%), especially considering that two-thirds of the time the number of jobs added or subtracted each month according to the household survey can be as much as 300K. The household survey might very well "catch up" to the establishment survey in the next few months.

And if we look just at jobs created by the private sector, the two surveys are very close over longer periods. As the chart above shows, according to the household survey, there have been 7.27 million new private sector jobs created since the end of 2009, while the establishment survey finds 7.29 million.  Note also how much more stable the establishment survey is than the household survey.

So let's assume that the household survey gives us a decent idea of how many full- and part-time jobs there are (that's not hard because it's our only source for this statistic), while the establishment survey does a good job at finding total private sector jobs. The public sector is not apt to rely as heavily on part-time employment as is the private sector, so this is arguably a valid comparison.

The above chart compares the level of part-time employment according to the household survey to the ratio of this same measure to the level of total private sector employment according to the establishment survey.

Several things to note:

1) Since the end of the last recession, the level of part-time employment has risen by about 0.7 million. So far this year, it has increased by 0.73 million. In other words, substantially all of the post-recession rise in part-time employment has occurred this year.

2) So far this year, the household survey says that full-time employment has only increased by about 0.22 million, so from this it follows that new part-time jobs have far outnumbered full-time jobs. But is that a fair characterization of the more important long-term trends? I think not. Consider that the same survey finds that total private sector employment has increased by 1.6 million so far this year. There's a lot of slippage in these numbers, and a lot of variation from month to month.

3) Relative to total private sector employment, the level of part-time employment has been declining (albeit irregularly) since the end of the last recession. The rise this year could be due to statistical noise, or it could be due to employers at firms with over 50 full-time employees reacting to the higher costs that will be imposed on them by Obamacare, and converting full-time employment into part-time employment to avoid the employer mandate. (I'm partial to the latter interpretation.) In any event, the recent rise in part-time employment this year is not likely symptomatic of any major, systematic, or sinister changes in the makeup of the U.S. labor force. We seen levels of part-time employment such as we have now several times in the past, and they usually follow recessions. This is nothing new or extraordinary.

Taking a dispassionate view from a longer-term perspective, the hue and cry over the recent rise in part-time employment seems way overdone. The more important truth is that the economy is generating almost 200K new private sector jobs per month, and most of the new jobs are likely full-time jobs. That's not a very stellar performance by historical standards—if this were a typical recovery we would have seem many millions more jobs created—but it's nothing to disparage.

UPDATE: Mark Perry provides some useful statistics which further explain this. And as he notes here, most of the increase in part-time employment came in just one month: March.