Labor Market Side Effects of Health Reform in Massachusetts and America

Massachusetts began a near-universal healthcare system circa 2007. The entire United States will begin one twelve months from now, pursuant to its "Affordable Care Act" (ACA). Can a labor economist interested in the ACA learn from the MA experience?

The ACA design is said to have been influenced by the MA program. I agree with that for the purposes of discussion.

The CBO and other analysts of the ACA have jumped to the conclusion that the major MA labor market effects, if any, of the MA reform are informative about the US labor market effects of the ACA.

It may be true that the HEALTH MARKET effects of the ACA will be similar to the MA health market effects of the MA reform, but economic theory contradicts the assumption that the two laws have similar LABOR MARKET effects.

From a labor market perspective, two margins are important: the margin between working and nonworking, and the margin between employer-sponsored insurance (ESI) and other sources of insurance. The ACA and the MA situations are entirely different in these dimensions.

THE REWARD TO WORKING

The ACA is the first piece of federal legislation that gives essentially free health insurance to people not working but not technically poor (with the exception of 18 months or so during the "stimulus" when COBRA payments were 65 percent subsidized for people on unemployment insurance). In this way, the reward to working in the US will be significantly lower after 1/1/2014 than it was before.

The MA reform did not change this part of the reward to working in MA because health insurance was already close to free for the unemployed. By the time MA had implemented its health reform, its "Medical Security Program" (MSP) was almost 20 years old. MSP gave (and still gives) people on unemployment insurance (UI) the option to have the state (actually, I think MA calls itself a Commonwealth) pay for 80 percent of their private health insurance premiums (up to $1200 per month!) or to receive health services directly from the Commonwealth, for as long as their UI lasts. MA's MassHealth Essential program provides benefits for the long-term employed (whose UI benefits have presumably expired). Like UI itself, these two programs are not asset tested. The employer financing of the two HI-for-unemployed programs is not experience rated (i.e., employers who layoff more do not have to pay more).

The MA reform even served to reduce UI a bit because the UI benefit amount is based on cash compensation, and the MA reform was encouraging the labor market to tilt a bit away from cash compensation and toward fringe benefits.

[Note that unemployed with UI benefits is not the only form of not working, or even the only form of unemployment (although a large majority of the unemployed since 2007 received UI).  In those cases, the two programs above do not apply.  Other MassHealth programs might, but I am not yet aware how similar or different MassHealth's programs are from Medicaid programs in the rest of the states.]

Massachusetts also encouraged using the federal tax exclusion for ESI, whereas the ACA tries to wean people off of it.  MA told employers: offer tax-excludable ESI -- make employees pay 100% if you want, it doesn't matter if it's affordable or if employees accept your offer -- or else we'll send you the bill for 100% of your employees' uncompensated care.  The ESI tax exclusion helps employed people save money, but does nothing for someone not employed.  [this para was added Feb 24]

Another part of the reward to working comes from the sliding scale subsidies to people working for employers not offering ESI. Both MA and US reforms have subsidies like these, but the MA subsidies apply to a much smaller slice of the population. For one, the frequency of no-ESI employers is much greater in the US pre-ACA than it was in MA pre-MA-reform. Second, the MA subsidies applied to persons up to 300 percent of the federal poverty line (FPL), as compared to the 400 percent threshold in the ACA; there's a lot of people in that 300-400 percent range. Third, the ACA subsidies will be more like cash than the MA subsidies because ACA beneficiaries will use them for pretty much any health insurance plan (for example, the same health plan that their Congressman and his family will use) whereas the MA subsidies can only be used for one of five state-sponsored plans. Judging by their pre-subsidy costs, the MA Commonwealth-sponsored plans appear to offer less than private sector plans. This MA practice enhances the reward to working: as one moves above 300 percent of the poverty line he loses his subsidy, but he also gets access to better health plans.

The end result of the MA sliding scale subsidy is that only 0.16 million people (including dependents) receive them in a Commonwealth with a population of 6.6 million. The ACA subsidies, on the other hand, are widely expected to be received by a much greater fraction of the US population. Naturally, a subsidy hitting a larger fraction of the population has a larger labor market impact.

So the only ways that MA health reform can be informative about the effects of the ACA on the quantity of labor (e.g., the fraction of the population employed) is that: (a) the reward to working doesn't matter that much or (b) analysts have made a correction or adjustment for this fundamental difference between the two reforms. So far, I doubt that either condition holds.

EMPLOYERS' COMPARATIVE ADVANTAGE IN PROVIDING HEALTH INSURANCE

The MA reform is also quite different from the ACA in terms of how it changes the incentives for employers to offer health insurance. There are many differences in this regard, but I begin by naming one or two.

The MA reform distinguishes offering health insurance to employees from helping employees pay for it. In MA, employers can offer health insurance to employees without an employer premium contribution by setting up a "125 plan," which allows employees to use their own pre-tax dollars to buy health insurance (for employees below 300% FPL, I think this opportunity includes the purchase of subsidized plans). Under the 125 plan, the employer only assists a bit in the administration of the premium payments/withholding.

The MA employer penalty for NOT offering a 125 plan (or ESI narrowly defined) can potentially be large: the employer can, in effect, be liable for all of the health costs the Commonwealth of MA incurs in caring for its employees. A small to medium-sized employer with the bad luck of having two employees get triple-bipass surgery in the same year may find himself wiped out by this "employer free-rider penalty." Admittedly, as of 2010 the Commonwealth of MA had yet to collect a single dollar of the free-rider penalties, but employers may nonetheless be scared to death of that liability, which may have led them to adopt 125 plans in mass (and thereby the Commonwealth gets no revenue from the penalty).

The MA employer penalty for not paying for any/enough of their employees' health insurance is just $295 per employee per year.

The ACA does not attempt to encourage 125 plans. In fact, MA may have to eliminate this part of its health reform when the ACA goes into effect. The ACA penalizes employers $2000 (and growing) per full=time employee if the employer does not offer affordable ESI.

One issue here is measurement. Does an employee who buys insurance through a 125 plan consider himself as having ESI? My guess is that he does, and that some of the population surveys are not well suited to detect the distinction, but more research is needed on this. Until then, I'm not sure how to interpret findings that ESI increased somewhat in MA after its reform.

Second, if we include 125 plans as ESI, the MA employer penalty for not having ESI in one form or another is potentially much larger than the $2000 ACA penalty.

Third, as noted above the MA subsidies to persons without ESI are infrequent as compared to the expected frequency of ACA subsidies. Moreover, the MA subsidies are less because MA restricts subsidy recipients to one of five less costly plans. Thus, MA employers have relatively few employees who would gain if ESI were dropped.

In summary, ACA proponents have likely been mistaken in taking comfort in the MA experience: the MA and ACA reforms are not comparable from a labor market perspective.

• A fully implemented and enforced ACA will significantly add to distortions on the margin between unemployment and working, whereas the MA reform did not.
• Via its sliding scale subsidies, a fully implemented and enforced ACA will distort other decisions that enhance family incomes, whereas the MA sliding scale subsidies hit a much smaller slice of the MA labor market.
• A fully implemented and enforced ACA will, on a variety of margins, move people out of jobs offering ESI, whereas the MA changes in these margins (if any) were significantly less, and different.

Surely, if it cares about its labor market, America was imprudent to adopt such a sweeping law before these issues could be acknowledged and better understood.

[the above was edited a few hours after posting to reflect comments by John Cochrane -- please don't blame him for mistakes that remain]