October 9, 2013, Testimony by Dean Baker before the U.S. Committee on Small Business Subcommittee on Health and Technology.
I want to thank Chairman Collins and Ranking Member Hahn for giving me the opportunity to address the subcommittee. I will use this opportunity to discuss work that I did with my colleague at the Center for Economic and Policy Research, Helene Jorgenson, examining the extent to which the Affordable Care Act (ACA) may have led to more part-time employment as employers cut work hours in order to avoid the employer sanctions in the law.
The ACA includes a provision that requires employers of 50 or more full-time workers (those working at least 30 hours per week on average) to either provide affordable insurance coverage directly to workers, as defined in the law, or to pay a penalty for each full-time worker who is not covered and subsequently buys subsidized insurance in the health care exchanges. There are two obvious ways to avoid this penalty.
The first route would be to keep the total number of workers under 50, either by not hiring workers for businesses near the cutoff or shedding workers for businesses just over the cutoff. The second route would be to reduce the number of hours that employees work so that they fall under the 30 hour a week average that would have them count towards the firm’s penalty under the ACA.
There are good reasons for questioning the extent to which the employer penalty provisions in the ACA would affect employment. Most importantly, the overwhelming majority of firms that employ at least 50 already provide health insurance coverage to their workers that would meet the standards of the ACA. According to a survey by the Kaiser Family Foundation more than 94 percent of firms that exceed the ACA’s 50 workers cutoff already provided coverage to their workers voluntarily. Clearly these firms consider it a good practice to offer a valuable benefit to their workers or they would not do so. Since most firms in this category already provide coverage voluntarily, it is difficult to believe that requiring the remaining firms to provide coverage or pay a penalty would create such an onerous burden.
Furthermore, the penalty for not providing insurance of $2,000 per worker (excluding the first 30 workers), is relatively modest. If the pay of full-time workers averaged just $10 an hour, this would be an increase in annual compensation of less than 10 percent. A considerable amount of research has found no measureable employment impact from considerably larger increases in the minimum wage.
For these reasons it seems unlikely that the ACA would have a large negative impact on employment. However, there have been numerous accounts of employers claiming to reduce employment or adjust hours in order to avoid the requirements and penalties in the ACA. If this is the case, we should have first begun to see evidence of the impact of ACA in January of 2013, since under the original law employment in 2013 would serve as the basis for assessing penalties in 2014. (The Obama administration announced on July 2, 2013 that it would not enforce sanctions in 2014 based on 2013 employment, but employers would not have known that sanctions would not be enforced prior to this date. Therefore we can assume that they would have behaved as though they expected to be subject to the sanctions.)
Some employers claim to have reduced employment because of the provisions of the ACA as soon as its passage in 2010, and many have blamed the ACA for the slow pace of employment growth in the years from 2010-2012. This is not plausible. There is enormous churning in the labor market, with close to 3 percent of employees leaving their job every month (half voluntarily and half involuntarily). If an employer felt the need to hire additional employers in 2010-2012 to meet the demand for labor they were seeing at the time, they would have no difficulty getting their employment. Just the normal churning in the labor market would bring a firm with 52 or 53 employees below the 50 employee threshold in a few months. Since employers generally have the option to dismiss workers at will (unless they have a union contract), there is no reason that they could not have added employees in the years prior to 2013 to meet their demand for labor at the time, and then reduce employment in 2013 to avoid the ACA penalties.
Evidence from 2013
It is too early to assess the claim that employers are staying just below the 50 employee limit since we do not yet have data available on employment by firm size in 2013. However as a practical matter it is implausible that the behavior of these firms could have any noticeable effect on employment growth. It is unlikely that more than 1 percent of potential employment growth would be in firms that are near this cutoff. Furthermore, most of these firms would already be providing health care insurance for their employees and therefore need not be concerned about the sanctions in the ACA. If some number of firms actually are limiting or reducing employment to stay below the 50 worker cutoff then the impact would be too small to be noticed in the economy as a whole.
The alternative course of evading ACA penalties, reducing average hours of work below 30 per week, could at least plausibly have an impact on employment patterns. In fact, several large employers have claimed that they would deliberately keep workers’ hours below 30 hours per week in order to avoid having them count toward the number for whom they would have a $2,000 penalty.
It is possible to test whether employers are actually reducing hours below the 30-hour threshold. The Current Population Survey (CPS) provides monthly data on workers usual weekly hours. We used the CPS to compare the first six months of 2013 with the first six months of 2012. Our original focus had been on the group of workers who reported working 26-29 hours a week. We considered this range a reasonable cutoff for an ACA effect. Presumably if an employer would have a worker put in more than 30 hours a week in the absence of ACA penalties, they would require a worker to put in close to, but less than, 30 hours in order to avoid the penalties.
In an analysis done on the first four months of data from 2013 we found a modest drop in the percentage of workers who worked this number of hours compared with the corresponding months of 2012. This suggested that concern over the employer sanctions in the ACA was leading firms to reduce work hours below the 30 hour cutoff.
We repeated the analysis after the July data became available, giving us all six months for which employers might have acted on the belief that they would be subject to the ACA’s sanctions based on 2013 employment levels. With the full six months of data we instead found a modest increase in the percentage of workers who are putting in 26-29 hours. The share went up from 0.61 percent of the workforce in 2012 to 0.64 percent of the workforce in 2013, an increase that corresponds to slightly more than 40,000 workers who have work schedules that put them just below the threshold as shown in the table below.
This may look like it is confirming exactly what opponents of the ACA warned against, that employers are responding to the threat of sanctions and cutting back workers’ hours, exactly as several prominent business owners had promised they would do.
However a closer examination shows that the data don’t support this story. The percentage of workers putting in 25-29 hours is up, but so is the percentage of the workforce that puts in 35 hours a week or more. In fact, the share of the workforce that reports working just over the limit, either at 30 hours a week or 31-34 hours a week, is up also.
It turns out that the big declines are in the percentage of workers who put in 1-19 hours a week, 20-24 hours a week, or who report that their hours typically vary. The data indicate that fewer workers are in these low or “hours varied” categories and more workers report falling into all the categories at 25 hours a week or above.
These changes are all small and mostly not statistically significant. They also reflect the influence of many factors other than the ACA. But the data certainly provide no evidence supporting the claim that the shortening of workweeks has been a widespread phenomenon.
Just to be clear, it is likely that the 30-hour sanction cutoff will have a modest but measurable effect on hours through time as employers adjust schedules and new businesses open. And any movement away from employer-based insurance will eliminate an important overhead cost that discouraged firms from shortening hours and hiring more workers.
There are three points that should be kept in mind about any possible movement to shorter hours that may result from the ACA. First, the vast majority of people who work part-time do so voluntarily. In many cases they have family or other obligations that make part-time employment desirable. Even with the current weak labor market more than two-thirds of the people who work part-time report that they do so voluntarily. In more normal times this share would typically be close to 80 percent.
Second, the United States is an outlier in that workers put in far more hours each year on average than they do in other wealthy countries. One of the factors that has prevented the same sort of decline in the length of the average work year that we have seen in other countries is that health insurance is typically seen as an overhead cost by employers. Typically they pay the same amount for workers’ health insurance regardless of how many hours they work. This gives employers an incentive to get more hours from each worker instead of hiring more workers.
Insofar as the ACA leads to a movement away from employer provided insurance, it will take away one of the factors leading to longer work hours. The result is that employers may more often decide to hire more workers at fewer hours per worker than would have been the case in the absence of the ACA.
Finally, the impact of the ACA’s sanctions is only likely to be seen over time. It is very disruptive to a workplace to overhaul work schedules, especially if workers see the goal as being to deny them a benefit that they would otherwise receive. For this reason, it is not likely that many employers would restructure work hours immediately after the ACA employer sanctions took effect. However over a longer term, new businesses are established and employers do periodically redo business plans, including restructuring of work hours. It would stand to reason that when employers have the option of having a worker put in a number of hours that is just over the cutoff for sanctions or just under the cutoff, they will opt in many cases to give workers a schedule that has them put in just under the cutoff.
In this respect, it is worth noting that the analysis of the impact of Hawaii’s employer sanctions for firms not providing coverage found no effect on the percentage of workers employed at less than the 20 hours cutoff for the period 1992-1993, more than ten years after the law had been in effect. The impact only appeared in an analysis of work patterns for the years 2002-2005; more than twenty years after the law had been in effect. Based on the findings of this analysis of Hawaii’s law, any possible effect of the ACA’s employer sanctions will not be large enough to be picked up in the data for many years into the future.
The employer sanctions in the ACA do provide a modest incentive for employers to reduce work hours below the 30 hour cutoff in the law. However, research suggests that the size of the sanction is relatively modest compared to other interventions, like the minimum wage, which have generally been found to have little or no effect on employment. Our analysis of hours data from the Current Population Survey for the first six months of 2013 actually finds a small increase in the percentage of workers who usually work more than the 30 hour cutoff. This suggests that the number of employers who may have actually cut hours to avoid the sanctions is too small to have a noticeable impact on the labor market.
2 The Kaiser Family Foundation and Health Research & Educational Trust. 2012. “Section 2: Health Benefit Offer Rates” in 2012 Employer Health Benefits Survey, 33-46. Menlo Park, CA: The Kaiser Family Foundation. http://kff.org/report-section/ehbs-2012-section-2/
3 Schmitt, John. 2013. “Why Does the Minimum Wage Have No Discernible Effect on Employment?” Washington, DC: Center for Economic and Policy Research. http://www.cepr.net/index.php/publications/reports/why-does-the-minimum-wage-have-no-discernible-effect-on-employment
5 Jorgenson, Helene and Dean Baker, 2013. “The Affordable Care Act: A Hidden Jobs Killer?” Washington, DC: Center for Economic and Policy Research. http://www.cepr.net/index.php/publications/reports/the-affordable-care-act-a-hidden-jobs-killer.
6 An analysis of the impact of Hawaii’s employer mandate, which first took effect in the early 1980s, found that it led to an increase of 3.7 percentage points in the share of workers in the bottom wage quintile who worked less than the 20 hour a week threshold specified in the state’s law (Buchmueller, Thomas C., John DiNardo, and Robert G. Valletta. 2009. “The Effect of an Employer Health Insurance Mandate on Health Insurance Coverage and the Demand for Labor: Evidence from Hawaii.” FRBSF Working Paper 2009-08, April). This implies an increase in the percentage of the whole work force working less than 20 hours of 0.74 percentage points.