The Affordable Care Act is making the U.S. health system much more efficient and fair. One provision of it, however, remains controversial, even among those strongly supportive of the overall law. This is the 40 percent excise tax on the marginal cost of expensive health plans, sometimes very misleadingly referred to as the “Cadillac Tax.” Defenders of this tax, and even many reporters, have claimed recently that the tax will “give Americans a raise” or will “raise incomes.” These claims are wrong. Instead, the excise tax— even in the best case—is an unambiguous cut in after-tax pay for workers.

Beginning in 2018, the tax will be levied on the cost of single plans in excess of $10,200 a year, and non-single plans in excess of $27,500. The point of the tax is to nudge workers into taking thinner health plans—those with lower premiums that stay under the threshold for the tax. But choosing plans with lower premiums will generally lead to higher out-of-pocket costs – higher deductibles, co-pays and/or other forms of cost sharing. This increased cost-sharing is the point of the tax, not a byproduct. By boosting the marginal cost of each new episode of obtaining health care, the theory is that health consumers will shop more wisely and cut back on unnecessary care. We have strong reservations about leaning on this dynamic as effective cost containment, but for now I’ll focus on a side claim made by defenders: that a happy consequence of accepting plans with lower premium costs is that workers will see higher wages.

The theory for this is that if employers cut back on contributions to health insurance premiums as workers choose thinner plans, more money will become available to boost non-health care compensation—wages or other fringe benefits. This presumed increase in wages actually accounts for a significant share of estimated revenue that will be raised by the tax. (I should note that if the compensating wage boost stemming from lower employer premium payments does not happen, this does not necessarily mean that the tax won’t raise money. Lower premium costs and unchanged wages paid by employers imply a rise in business income or profitability, and this higher profitability should mean higher tax payments by employers.)

This potential boost to wages is what leads many of the tax’s defenders to call it a “pay increase.” But of course, even if wages rise in response to cutback s in employer premium contributions, the level of pre-tax pay (where pay includes both wages as well as other employer-provided benefits) won’t increase at all—instead its composition will simply shift, with employer premium payments falling and wages rising. Further, because wages are taxed and employer premium payments are not, this means that the total tax bill on the worker’s overall compensation will have unambiguously increased. So, post-tax compensation is lower with the excise tax, period. It is a take-home-pay cut.

We know that, all else equal, average out-of-pocket costs will rise in line with the average reduction in premiums as firms and workers shift to thinner health plans. If out-of-pocket costs rise one-for-one with the reduction in premiums, this implies that all of the increased wages stemming from this compensation shift will simply be forked over to doctors and other health providers. This would mean that even wages that were pre-tax but post–health-care expenses (PTPHCE) would be no higher after the tax. If the increased cost-sharing (higher out-of-pocket costs) spurred by the tax instead led to workers deciding to consume less health care, then workers’ PTPHCE wages would rise. But their total compensation would remain lower than it was before the excise tax, and their wage increase would be driven by their decision to forego health care.

How confident should we be that wages will rise in response to a cutback in employer premium payments in an effort to avoid the tax? I’ve been indoctrinated enough as an economist to think this will likely be the case over the long run. However, the ability of workers to demand wages as recompense for accepting thinner health plans and lower employer premium contributions will depend on many things, including how close the economy is to full employment. The U.S. economy has spent many long periods of time away from full employment in recent decades, so it is no guarantee that these compensating wage-gains will come quickly. This necessary ingredient for effective workers’ bargaining power and its frequent absence in the U.S. labor market may be a reason why the compensating wage gains stemming from rising health costs has been hard to definitively pin down in the research literature.

Despite many references to a wage/health care benefits trade-off being a completely settled topic in empirical economics, it’s actually far from a robust finding. There are high-quality studies that find a wages/health care cost trade-off, and high-quality studies that don’t. Further, this is a topic where one needs to be worried about publication bias—the possibility that studies failing to find this trade-off are just not published. This has been a real problem in previous economic debates , like that around the disemployment effect of the minimum wage.

An illustration of the importance of labor market conditions on whether we see compensating wage gains is the past eight years. Since 2007, health care costs have decelerated rapidly. This should have opened up a lot of room for faster wage gains in the post-2006 period. Yet inflation-adjusted wages after 2007 have fallen for the vast majority of American workers.

In the end, I do think that wage gains will eventually come if we incentivize thinner health plans, but it could be quite a while, and it doesn’t change the fact that the excise tax remains an unambiguous pay cut for American workers.