The Affordable Care Act (ACA) presents employers and potential employees with a variety of new rewards and penalties. These are, in part, exactly what the law intended: by penalizing potential employees for not purchasing health insurance, and employers for not providing it, the law aims to increase the fraction of the population with health insurance. Yet these same rewards and penalties have additional effects, including on the incentive to work; Mulligan (2014), for example, suggests that the ACA may reduce employment by 3 percent on average and have a range of positive and negative effects on average hours worked.
In the work summarized here, I quantify the number of people who will have essentially no short-term financial reward from working more than 29 hours, since this would either render them ineligible for the ACA’s assistance or increase the penalties that may be owed by their employer. This is the first paper to show that the ACA will put millions of workers in the economically extreme situation of having zero short-term financial reward (or less) to working full-time rather than part-time.
In economics jargon, this means the ACA creates marginal tax rates on labor income that exceed 100 percent. Even when helping people who are out of work or who otherwise have low incomes is a primary policy motivation, and even if labor supply does not respond much to the after-tax wage, labor income tax rates that equal or exceed 100 percent are bad policy because they discourage work effort without raising any revenue.
From a strictly predictive perspective, economists expect that full- time employment rates will be low, if not zero, in groups of people who are aware that they receive no financial reward from working full-time (defined here to be working at least 30 hours per week).
Two separate ACA provisions can fully eliminate the reward to full-time work. The first, which is scheduled to be in full force in 2016, pertains to full-time employees of firms that do not offer health insurance: by cutting weekly work hours to 29, they save their employer the annual salary equivalent of more than $3,000, or, they save their employers the threat of even larger penalties. Women workers, young workers, and persons already working 30–35 hour schedules are especially likely to have their short-term financial reward to full-time work erased by the ACA. By my estimates, three to four million workers overall will fall victim to this penalty provision.
The second provision pertains to full-time employees at firms that do offer health insurance. Over 60 million workers obtain health insurance from their employer, not including workers who obtain health insurance from a family member’s employer. About half of them (26 million) are in families between 100 and 400 percent of the poverty line and therefore satisfy the income criteria for exchange subsidies. And 11 million of those are unmarried — so by definition cannot be covered by a spouse’s plan — and another 8 million of the married have a spouse that does not work or otherwise cannot obtain coverage through a spouse.
In other words, almost 20 million workers are ineligible for exchange subsidies solely because their employer offers coverage to full-time employees: these are the workers subject to the ACA’s implicit full-time employment tax (FTET). A 29-hour work schedule, on the other hand, would make them eligible for subsidies without creating any penalty for the employer.
In about four million cases (of the 20 million facing an implicit FTET of some magnitude), the dollar amount of subsidy gain can exceed the after-tax income that is earned for working beyond 29 hours per week. A distinguishing feature of almost 90 percent of these workers is that their family incomes are below 250 percent of the federal poverty line. The four million disproportionately consist of working unmarried household heads because, as noted, unmarried heads are especially likely to be ineligible for exchange subsidies solely because their employer is offering coverage to full-time employees.
Older (but not elderly) workers are also disproportionately represented among those facing an implicit FTET rate of 100+ percent, since older workers are more likely to have employer-sponsored insurance and are more expensive to insure. The 100+ percent FTET from the employer penalty has the opposite age pattern, which means there may be little age pattern for the propensity to face one of the 100+ percent FTETs.
The prevalence of 100+ percent FTETs is an important indicator of its effects on incentives to work, but it is not the only one. There are other ways to avoid the FTET, such as working more hours per week for fewer weeks of the year. If employers are unwilling or unable to adjust work schedules, the FTET may affect the equilibrium relationship between hours and earnings (i.e., compensating differences) rather than changing the distribution of hours. At the other extreme, employers may be able to substantially adjust measured work hours without changing the actual work that is done (e.g., require employees to “punch out” during break periods, and then adjust their hourly wage so that weekly earnings are the same), in which case the ACA will reduce the measured hours for quite a large number of workers.
In effect, millions of workers are becoming eligible for fully federally funded paid days off, akin to the sick leave policies in Western European countries. Because the Western European data suggest that paid sick days really do result in fewer days at work (Lusinyan 2007), we should expect the act’s FTETs to reduce days worked as well, at least for the segments of the workforce that do not avoid the ACA’s taxes in other ways.
These effects of the ACA are only part of its impact; people who believe government should provide health insurance for everyone may regard these costs — the disincentive effects discussed here and elsewhere — as worth paying.
But everyone should recognize that the ACA’s costs are likely to exceed the budgetary expenditure.
This Research Brief is based on Casey B. Mulligan, “The ACA: Some Unpleasant Welfare Arithmetic,” National Bureau of Economics Working Paper No. 20020, available at http://www.nber.org/papers/w20020. All works cited are provided there.