This is ancient so far as internet attention goes (i.e., last week) but the Center for Economic and Policy Research released an analysis of Current Population Survey data on people working between 26 and 29 hours a week. The economists, Helene Jorgensen and Dean Baker, were looking for macro-evidence that employees have had their weekly hours reduced to comply with the employer-mandate provision in the Affordable Care Act.
The findings (my emphasis in bold):
An analysis of data from the Current Population Survey shows that only a small number (0.6 percent of the workforce) of workers report working just below the 30 hour cutoff in the range of 26-29 hours per week. Furthermore, the number of workers who fall in this category was actually lower in 2013 than in 2012, the year before the sanctions would have applied. This suggests that employers do not appear to be changing hours in large numbers in response to the sanctions in the ACA.
I’ve been on this hobby-horse for a few weeks now so I apologize if it’s getting old. However, moving forward towards implementation and beyond we’ll hear many, many anecdotes about the ACA’s effects so the big data picture is only going to become more important.
As it’s been reported elsewhere, the mandate that requires firms with more than 50 full-time employees to start offering health insurance or pay a penalty when subsidy-eligible people enter the exchanges affects relatively few of either. Details of that slice of the working population Baker and Jorgensen studies yields further insight:
- Well under 1 million workers, roughly 0.6 percent of the labor force, typically work between 26-29 hours a week.
- It is also important to remember that many of these workers choose to work less than a full-time job. More than two-thirds of the workers who report working less than full-time jobs say that they are doing so by choice.
- If this ratio also applies to the workers who usually work between 26-29 hours it would mean that less than 300,000 workers, or roughly 0.2 percent of the workforce, are working this number of hours as a result of their employer’s decision.
To summarize; according to CPS data employees working in the range have actually decreased between 2012-2013 (wherein 2013 hours would’ve applied to the mandate’s scope in 2014 before the delay), which is the opposite direction one would assume if massive underemployment were occurring, and the share of workers who are in that range not of their own volition represent about 0.2 percent of the labor force.
That being said, there is a number greater than zero for which this provision will have a negative impact. I’ve had personal conversations with those that have bosses explicitly telling them that their hours are being cut because they don’t want to give them health insurance. It’s going to continue happening and it will make some folks very upset. It’s also true that the same segment most likely to risk having their hours reduced are also most likely to greatly benefit from the law. Yet opponents of healthcare reform will seize that nugget of discontent and claim an epidemic of devastating macro-proportions ruining the economy. That’s where the big data proves them wrong so far, and that’s why it’s important to continue looking at the big picture moving ahead.
Again, and I’ll repeat this ad-infinitum, no one is claiming that the small proportion of workers disfavored by the employer mandate makes it an unimportant policy concern. Many of the ACA’s advocates, including myself, consider this provision of the law comparatively unimportant and would be completely open to changing it in a way that benefits everyone. To be blunt, the only thing standing in the way of such improvements are Republicans.
***Update: Over Twitter Adrianna McIntyre says there’s an important caveat to this analysis, that the “2013 ‘lookback’ window has six month option.” Ergo, the CPS analysis comparing 2012 to 2013 ending in May of both years could be less relevant if such an option meant only hourly data post-June were used in applying 2014 penalties. This is a tentative caveat — some quick searching didn’t immediately yield any extra information about this — but it’s worth putting out there. Irrespective of this possibility I would just note that there is still a decided lack of evidence from the (much more current) BLS monthly data.