Consulting by RPM || Free Advice Blog

More on the Tax on “Cadillac” Health Insurance Plans

In response to my earlier post on Scott Sumner and Steve Landsburg, David R. Henderson sent me two of his own posts (I and II) on the 40% surtax on so-called Cadillac employer-provided health plans.

I think I’m generally in agreement with everything David wrote, but I suspect he is still more favorable to the tax than I am (…inasmuch as I am very much against the tax, and I don’t think David would be upset if it passed).

So in response to Scott and Steve for sure, and David not as much, let me offer the following points of pushback:

(1) The tax on Cadillac plans does not undo the “subsidy” in the existing tax code. As David himself conceded, it only kicks in at the margin.

(2) As John Goodman pointed out to David, this approach is all pain. (See Goodman’s own suggestion for using a tax credit approach to get the marginal incentives much better.)

(3) I don’t think it was an accident that the original ACA legislation only had the 40% surtax kick in at a high threshold. They weren’t trying to get rid of third-party payment for health expenses. No, in addition to the Cadillac tax the ACA also contained an employer mandate: a stiff penalty for large employers who didn’t provide adequate coverage to their employees. Thus employers were getting hit from both ends: up to a $3,000 per employee fine for not providing a decent plan, but then a 40% surtax on plans that were too good.

So I think Steve and Scott are being naive when they’re acting as if the Republicans have somehow foiled the architects of the ACA who were doing their best to contain health care costs. No, they weren’t. I think they were trying to ram a one-size-fits-all approach onto the employer plans to match the crappy plans people were getting on the exchanges. So (a) people paying out of pocket, (b) people buying plans through the exchanges, and (c) people getting plans from their employer, were all getting crappy coverage where they would still be devastated if they got hit with a really big medical event. Thus keeping health insurance bureaucratic and unsatisfactory, and giving rise to calls for single payer.

(4) In general, if the public is groaning under the burden of stuff costing too much, you don’t help them out by slapping a new tax on them. For an analogy, loose monetary policy led to rampant CPI inflation in the 1970s. I think some commentators actually suggested tax hikes as a way of “sopping up demand” and restraining annual price increases. No, that wouldn’t really help people pay for their groceries. Or for the analogy I used last time: We can point to all the ways the government itself is responsible for skyrocketing tuition costs. But if all the government did were to slap a 40% tax on colleges for every student they admitted, that would hardly help parents pay for college.

(5) I think the underlying assumption in these analyses is that the government holds other spending fixed, and so any new revenue from the Cadillac tax would simply reduce the deficit. But in practice I don’t think that is correct. So to the extent that it brought in receipts, I think we also have to factor in more IRS audits, more drone strikes, more DEA raids, and all the other stuff that the federal government spends money on. It’s because of this element that I am really surprised to see free-market economists lamenting the “missed opportunity” to tax the @)#($# out of employer-provided health plans. (I realize the main response would be for employers to not exceed the threshold.)

(6) As I argued in my Lara-Murphy post, I don’t think the IRS code by itself is what’s driving the absurdities in health care costs. I think the tax code provisions interact with all sorts of supply restrictions and mandates. For just one consideration: Before the ACA, if a person’s child had a medical condition, it might have been impossible for the person to buy individual health insurance. But if the person went and got a normal office job, then he could get himself and his dependents on the company plan without them even asking about his kid’s medical history.

So this is clearly an advantage to employer-provided coverage, and it explains why coverage isn’t portable. Yet this has nothing to do with tax deductibility. I don’t know if it’s because of a regulation saying health insurers have to provide group coverage to everybody, or if it’s a market outcome that insurers are OK with “community rating” because they figure it will all wash out. But for sure, this type of consideration shows that we’re not merely dealing with pre- and post-tax dollars for the same product.

==> In summary, although I understand and agree with the narrow technical analysis of the inefficiencies of the tax code, and how this contributes to an overallocation of resources into health care, nonetheless I was very much against the ACA’s surtax on (the top end of) employer-provided Cadillac plans. I think many free-market economists are fooling themselves if they (a) thought the architects of the ACA were really trying to improve efficiency and (b) predicted that such a new tax would rationalize health spending and reduce deadweight loss, all things considered.