I’m hearing a lot of talk these days about how the Affordable Care Act is going to hurt job growth. Some of it is from news stories out today about the one-year delay in the employer mandate. They reminded me of a debate a few weeks ago on CNBCs Squawk Box, where I was mixing it up with some usually level-headed pals (Zandi, Liesman) who I thought were giving far too much weight to ACA’s implementation as a factor weighing down job creation in the monthly employment reports. Then today I read in a data note from economy.com evaluating monthly ADP jobs report that even though the ACA’s employer mandate was just delayed, the law is “…likely to still stifle hiring at firms with around 50 employees…”
This essay (and sorry–it’s longer than usual) evaluates claims like these based on the expected interactions between the law and the job market and some evidence that I’ve pieced together from various sources that bears on the question of the ACAs impact on jobs. While it’s true that there are incentives in the bill that could negatively impact job creation:
–the magnitude of the disincentives are small relative to the job market–the employer mandate in particular will affect very few employers;
–there are incentives that go the other way, i.e., toward lowering the cost of labor;
– it’s highly unlikely that anyone can tease anything out of monthly jobs data that would allow you to reasonably implicate health care reform;
–the ACA is not a jobs bill! Its purpose is to raise access to and affordability of health coverage. Those are the criteria by which it should mainly be judged. Of course, if it should turn out to have even moderately negative impacts on other parts of the economy, like the job market, we would need to get under the hood and try to fix it. But there is nothing in the current or near-term landscape that point in that direction.
Before getting into details, a quick point about the monthly jobs numbers. As the recession fades in the rearview mirror and fiscal headwinds and monetary tailwinds buffet the economy, there are many economic cross-currents in the job market. In addition, the monthly data, especially by firm size, are noisy. Employers’ understanding of the impact of health reform on their labor costs are muddy at best. And the vast majority of employers and employees won’t be affected at all. Given all that, put me down as awfully dubious re my colleagues’ arguments that the law is significantly affecting monthly payrolls, with “significantly” in this case meaning that the signal-to-noise ratio is too low to support their conclusions.
Incentives that could impact job creation
Labor demand depends both on the demand for the goods or services an employer is producing and selling, as well as on the cost of labor. So, holding demand constant, if the ACA raises the cost of labor, we would expect it to have a negative impact on hiring. And one could go further: even if employers expect it to increase their labor costs, they might hold back on hiring at the margin until they have a better feel for the costs of compliance.
Case closed, the ACA is a job killer, right? Wrong! (Who didn’t see that coming?)
Right off the bat, you have to think about the wage/benefit tradeoff. If employers who must now provide health coverage can lower the wage part of their workers’ compensation by something close to dollar-for-dollar, their labor costs don’t change. Though such adjustments can take time, most economists swear by that tradeoff (I suspect it’s less than dollar-for-dollar, especially in the near term). Also, there’s a wage floor—the minimum wage—below which employers cannot lower the wage part of the package.
Even so—i.e., even if the tradeoff is neither instantaneous nor dollar-for-dollar—it’s certainly a factor such that you can’t just assume that new coverage increases the compensation bill by the sticker price of new coverage.
Of equal importance, not every employer faces the mandate to provide coverage nor the $2,000-$3,000 per worker penalty—now slated to begin in 2015—if they don’t provide it. Employers with less than the equivalent of 50 full-time employees (FTE’s, e.g. two half-timers=one full-timer) are exempt, for example. Therefore, there’s a legitimate concern that some firms near that 50 FTE border might be careful to stay south of it, thus discouraging hiring near that margin.
On the other hand, starting in 2014, small firms under 25 FTE’s get a tax credit of up to 50% against the costs of coverage in the health exchanges for two years. So for employers currently providing coverage, there’s a new rule that could lower labor costs and thus incentivize job creation at the margin.
Since the ACA does not require that part-timers are covered, with part-time defined as less than 30 hours/week, there’s an incentive to create part-time as opposed to full-time job slots. Of course, in terms of job counts, this incentive goes the other way—toward more job creation. It also, however, has the potential to lead to more involuntary part-time work.
So far I’ve been talking about incentives faced by employers, but workers will also face incentives that could affect their labor supply. For subsidized workers in the health care exchanges, as their income rises, their subsidy falls, implying an increasing marginal tax rate which could lead some to work less (it could also lead some to work more—the income effect). Pushing the other way, there are many states where under the pre-ACA system working-age people can lose their Medicaid coverage when their earnings exceed a very low threshold. In those states, assuming they accept the Medicaid expansion to cover adults up to 138% of the poverty threshold, a high marginal tax rate will be significantly reduced.
I doubt either of those supply-side incentives will amount to much. The research on labor supply responses to such marginal rate changes finds them to have effects that range from small to immeasurable.
All of which is to say there are many moving parts here, some of which push in different directions. That means we must turn to the evidence. There’s not a lot here, of course, because the ACA is still very much in implementation mode. But what there is largely supports my hypothesis that it’s a mistake to assume the ACA will lead to—already is leading to—significant job losses.
First off, the ACA is not big enough as a share of the economy or workforce to have large effects. Holahan and Garrett emphasize that its expenditures amount to well under 0.5% of GDP, and that’s once it’s fully phased in. So it seems implausible to argue that a program of that relative magnitude could be significantly disruptive to job creation.
To be fair, the more common claim, like that of my CNBC colleagues, is that the ACA’s negative jobs impact will be at the type of margins discussed above. Yet here too, the potential magnitudes are small relative to the job market.
While most people work for firms with more than 50 people (about 72%), the majority of those workers—about 95%–are already offered health coverage and most of them take it up. One study that tried to get at the number of workers who could trigger employer penalties found 2.3 million workers who could be vulnerable to reduced hours because they work close to the 30 hour cutoff at larger firms, they’re currently not covered, and their income is such that they’re eligible for a subsidy. But that’s less than 2% of current payrolls and, as noted above, if they’re willing to trade wages for benefits, it’s not obvious employers of even this relatively small group will face higher labor costs.
Still, clear incentives exist and Phillips (no link, but figures below) looks at the share of workers in eligible sectors (full-time, 50 or more workers, not currently covered by employer) and finds their hours of work have grown more slowly of late.
But, as the figure below reveals, Phillips also finds that “employment in sectors with the greater share of workers out of compliance with the upcoming mandates appears to have grown slightly more quickly than sectors with higher levels of coverage,” [my italics]. In fact, the lines in the figure seem to be growing pretty much at the same rate in recent years, with the 20-49 size firms not obviously picking up growth from the others (that’s the group you might expect to be growth faster relative to the 50+ category if employers are trying to keep under the 50 FTE borderline).
Source: Phillips, GS Research
One study I found helpful regarding this question of ACA and jobs comes from the Massachusetts health reform law, which has many similarities to the ACA in terms of employer mandates (though the penalties are considerably smaller) and has been in place since 2006. What’s particularly useful is that the authors were able to do a pseudo experimental design of sorts, comparing employment outcomes in MA to those in states with similar employment patterns. Those states form a sort of control group, because they presumably experienced the same economic trends but not the health-care-law change.
The authors first confirm that coverage increased significantly in MA relative to the control states (establishing the intervention) and then examine whether this difference appeared to reduce employment in MA relative to the comparison state or the nation. The figure below shows that employment rates in MA tracked those of the control group states quite closely in both the few years following the adoption of the new health care law and the recession. Other numbers in the report (see their table 3) show no differences in part-time employment growth either.
The authors conclude: “Massachusetts has achieved its goal of near-universal health insurance coverage under its 2006 health reform initiative, with no indication of negative job consequences relative to other states as a result of health reform.”
Source: Dubay et al, 2012
The fact that most employers—around 95%–subject to the mandate already offer coverage, that the wage/benefit tradeoff exists, that expenditures on the ACA amount to well under 0.5% of GDP, that the smallest firms will get a subsidy to offset the costs of coverage all suggest that the law will do very little to change employers’ hiring incentives. A small group of employers in larger firms—the 5% or so that don’t already offer at least minimal coverage—could face a penalty if their workers end up getting subsidized coverage in the exchanges and that rule could lead some employers near the 50 FTE threshold to reduce hiring or workers’ hours, but for reasons elaborated above, I expect those impacts to be small, if not unmeasurable.
The evidence, most notably from MA where a similar law has been implemented quite strongly support this conclusion. True, MA is not the US, but in one regard that makes the evidence even more compelling. If the law put MA businesses at a competitive disadvantage relative to businesses in the comparison states, we should have seen those effects in that study. But we clearly do not. Because the ACA is a national program, it places no state at a comparative disadvantage.
Two final points. First, it is a serious analytic mistake to complain about the ACAs potential job impacts without stating your “counterfactual,” i.e., your program to address our highly inefficient health care system. As Vice-President Biden used to say in these cases, “don’t compare me to the Almighty, compare me to the alternative.” Unless critics are willing to accept a status quo that is fiscally and economically unsustainable, they need to tell us their health reform plan and how it manages to have a better impact on jobs than the ACA.
Finally, the ACA is not a jobs program. If it succeeds in one of its central goals of slowing health care costs by squeezing out current inefficiencies, it will undoubtedly help the economy by lower cost growth to businesses and improving the fiscal outlook. In fact, recent health care savings, some of which appear related to ACA-type reforms, are already having this effect, though whether these more favorable trends have staying power is yet to be seen.
But to grade the program on job creation is inconsistent with its central purpose. For example, most would agree that if the ACA turns out to increase health care access, affordability, and coverage while lowering cost growth, it will be considered highly successful, even if it is found to be entirely neutral regarding job growth. If, on the other hand, the ACA has the unintended consequence of notably hurting job creation, policy makers will need to quickly take action to try to correct that problem. While the analysis and evidence presented above suggests that this latter outcome is not likely, we should vigilantly watch for and monitor it.