Trends in Part-Time Employment
Since the July Monthly Jobs Report came out August 2nd there has been chatter on talk radio and elsewhere saying 2/3rds or more of all new jobs created this year are part-time.
The National Review published an article on the subject August 5th titled:
“Are We Becoming A Part-Time Nation?“
Well… are we? And, if so, why and by how much? BLS data provides some answers.
Part-Time Jobs This Year
The National Review reports, from BLS figures, that there have been 980,000 jobs created this year and 692,000 of them are part-time. If 70.6% of all jobs created are part-time, that is a huge percentage!
National Review makes two conclusions:
- This year’s incredible growth in the part-time to full-time jobs ratio is “disconcerting”
- Obamacare is compounding the problem
There is a flaw in the figures. It isn’t entirely the National Review‘s fault. The BLS deserves most of the blame. The flaw comes from the standard error found in the originating data source. The Household Survey (Table A-1) has a monthly standard error of +-300,000 in total employment, according to the BLS. All other subtotals suffer accordingly.
For example, if the BLS reported total employment from the Household Survey instead of the Employment Survey then May’s job growth would be reported as +319,000 new jobs instead of the +176,000 they reported. That would have made President Obama positively giddy!
Drawing short-term conclusions from data with that kind of standard error is a mistake.
The Household Survey is better suited to years long studies, not months long. It gets periodically recalculated to reduce its error.
But the NR is forced to use it for this year because the Employment Survey, which has 4X less standard error, does not collect part-time data.
It should, it could, but it doesn’t.
A better choice would be to take the more accurate total – non-farm employment – for this year (+1,347,000) and compare that to the part-time figures from the Household Survey. That changes the part-time to full-time job ratio to 51%.
It waves a great big huge red flag when the total employment from the larger Employment workforce (156M) is 1/3rd smaller than the lesser employment survey workforce (136M).
That does not compute!
51% is still WAY to high, but a far cry from the 70.6% and 97.7% figures the National Review was throwing around.
Mining for Meaning
What does the Household Survey reveal when reviewed over a decade where its short-term standard error is smoothed out, rather than over a few months with pronounced error?
This first simple baseline graph reveals:
- 8.5 million jobs permanently lost in the Great Recession
- There are still 2.3 million fewer jobs today than before the Great Recession
- Reasonable job growth before the Great Recession
- Slower job growth during the recovery
The ‘part-time due to economic conditions’ graph (those desperate for work) reveals:
- Part-time due to economic conditions trended down before the Great Recession
- Part-time skyrocketed to 9 million during the Great Recession
- Part-time has declined by about 1 million during the recovery
- Part-time has trended slightly higher in 2013 (see red line)
The ‘part-time non-economic’ graph (folks choosing part-time) reveals:
- Workers choosing part-time for non-economic reasons trended up before the Great Recession
- Part-time voluntarily declined during the Great Recession (understandably)
- Part-time voluntarily went up during the first part of the recovery
- Part-time voluntarily then mysteriously declined in late 2012 (red arrow)
- Part-time voluntarily suddenly jumped dramatically in 2013 (red line)
The most surprising finding in all this is that workers choosing part-time voluntarily is sharply higher this year (+365,000) after declining by -250,000 in 2012.
Some, and perhaps much, is due to sampling error in the Household Survey.
The number of workers taking part-time jobs involuntarily due to economic conditions is pacing slightly higher this year over 2012. It is likely due to Obamacare, as the National Review suggested. That pace will increase as 2015 approaches.
(Businesses were granted another year to implement the employer mandate)
2013 part-time data itself started in a major statistical dip in December 2012. The number of total jobs created in 2013 from the Household Survey is much lower than the Employer Survey that has fewer total workers. That doesn’t make logical sense and is a red flag.
More than likely, the excessive part-time job trend being reported this week is based more on sampling error than reality.