Category Archives: Health Insurance

Election 2026: Day -54

Health Insurance, Politics

With the midterm elections within sight, I’m going to temporarily adopt the “Election 20XX” branding that I’ve used previously, and I will endeavor to blog more frequently over the next several weeks.

In my last post I commented that I believed primary season was at long last over. Turns out I was wrong: two days ago New Hampshire had its primary, and yesterday it was Rhode Island’s turn. In Rhode Island, the incumbent Democratic governor lost his primary, in a rematch between the two leading candidates from the 2022 primary.

Yesterday and today, the GOP have been holding a so-called “midterm convention,” in Dallas. I’d never heard of such a thing before, but apparently the Democrats had tried it in June 1982. However, the timing of the event is slightly odd in that the first 2 NFL games of the new season took place yesterday evening and this evening, opposite the event. Trump spoke for 105 minutes last night, and both Trump and Vance are scheduled to speak tonight. Unexpectedly, Sen. Fetterman (D?!?-PA) made an appearance by video at the conference last night.

The big news from last night is that during his speech Trump vowed that if the Republicans win the mid-term elections, then every adult American will get $5,000. Given that such a promise would cost over $1 trillion, and given that Trump had previously talked about $5,000 “DOGE dividends” and $2,000 “tariff dividends”, I’m not sure Trump is to be taken seriously in this instance. Still, that’s quite a campaign promise for a (checks notes) Republican President to make.

However this morning there was also news about an upcoming $500 payment to a subset of Americans, and this news actually appears to be for real. Moreover, it relates to one of my areas of professional expertise, the individual health insurance policies offered on the federal exchange (Healthcare.gov) as a result of the Affordable Care Act.

What the White House fact sheet says is that “nearly a million” Americans who purchased policies from the federal exchange will each receive a $500 rebate in October, to compensate for the fact that they were “overcharged” by the Biden Administration with respect to exchange user fees.

As background: When insurers offer products on the exchange, the premium build-up includes a %-of-premium fee that the insurer needs to pay to the federal government, as a so-called “user fee” to cover the government’s costs of maintaining the exchange. That fee is set by the federal government, many months in advance of the start of the calendar year. In the waning days of the Biden Administration, in mid-January 2025, a final regulation came out indicating that the exchange user fee for calendar year 2026 would be 2.5% of premiums.

Apparently, the Trump Administration has concluded that the level of user fees collected from 2026 premiums turned out to be more than what is actually needed to run the exchange. In May 2026, the Trump Administration set the exchange user fee level for calendar year 2027 at 1.9% of premiums, instead of 2.5%. As such, when the Trump Administration talks about “overcharges” by the Biden Administration, I think what they’re saying is that 2026 premiums were, roughly, 2.5% – 1.9% = 0.6% higher than they really needed to be, because the exchange user fees collected via those 2026 premiums were more than what was needed.

What should the government do here? Door 1: Just keep the excess money, and perhaps use it to subsidize lower exchange user fees charged for years beyond 2027. Door 2: Return it to the insurers, since they’re the parties that actually remitted the exchange user fees to the government. Door 3: Return it not to the insurers, but to the parties who actually paid the premiums to the insurers. (The reader can analogize to the recent situation with refunds of illegally imposed tariffs.)

So, the government chose Door 3, right? Well, sort of…

One unusual but important feature of the health insurance policies sold on the exchange is that most of the people buying these policies qualify for a government subsidy of part, or in many cases all, of the cost of the premiums. These premium subsidies are administered through what are called Advance Premium Tax Credits (APTC), through which the individual buying the policy doesn’t need to pay the premiums to the insurer and subsequently seek reimbursement from the federal government — instead, the federal government directly pays part or all of the premiums to the insurer.

What that means is that if 2026 premiums really were too high by 0.6%, then the vast majority of that excess cost was borne by taxpayers at large, via higher APTC payments funded out of general government revenues! Only a small portion of the purported excess cost was actually borne by the people covered by the insurance.

Moreover, the math doesn’t really line up here. First, there are vastly more than 1 million people who bought coverage from the federal exchange, although most did so with the assistance of APTCs. Second, if you assume that the monthly premium for an exchange product is $700 per month (a number I admittedly pulled out of the ether, but it’s approximately right and it makes the resulting math pretty) then 0.6% of premium is $50 per year, not $500.

As such, here’s what I suspect is actually going on. I think the government has taken a pot of money equal to roughly 0.6% of all the 2026 exchange premiums, spanning both people with APTCs and people who fully fund their own premiums; and then I think the government is using that pot to make $500 payments to only the people who fully fund their own premiums. And that set of people — working-age individuals who don’t have access to employer-sponsored health insurance but make too much money to qualify for APTCs — consists largely of the self-employed upper-middle-class, who I’ve always thought of as a core Republican constituency.

This is my original analysis, and I certainly could be wrong. But assuming I’m right, then politically it’s a clever albeit disingenuous move: Identify a way in which your predecessor’s policies arguably brought harm to a portion of the population largely lying within your party’s base, and provide those people with payments that vastly exceed the harm they experienced (i.e., the $500 versus the $50), a month before an election, while blaming your predecessor.

Kansas Proposal to Increase HMO Tax: Unsound Policy?

Health Insurance, Tax Policy

Today I read an article in a national outlet about proposed legislation in Kansas that touches on a number of themes I’ve dealt with in my professional career.  (Full disclosure:  Until recently I worked for a major insurer that is identified in the article as leading the lobbying effort against this legislation; however, I had zero involvement in that lobbying effort and in fact was completely unaware of this issue until reading about it today.)

My understanding of the situation, based largely on reporting from the Wichita Eagle, is this:

  1. It is well known that Gov. Brownback has been pursuing an aggressive tax reduction campaign in Kansas.  As part of this, he is looking to find a new revenue source to replace about $80 million of general state tax revenues that are devoted to support Kansas’ Medicaid program.
  2. His proposal involves increasing a particular Kansas tax that applies only to HMOs (and not to other health insurers) and that has a peculiar name:  the privilege fee.  This tax is currently defined as 1.0% of annual premiums, but under the legislation it would increase to 5.5% of annual premiums.
  3. Three of the HMOs to which the tax increase would apply are contractors under the Kansas Medicaid program, KanCare.  If I understand correctly, the privilege fee applies to all premiums written by HMOs, including both KanCare and non-KanCare premiums.

This proposal strikes me as suboptimal public policy, for a number of reasons:

The tax increase on HMOs will ultimately get passed through to other parties, so what exactly is the point?  Look, it’s potentially very appealing for a politician to be able to say, I’m reducing the general tax burden, and in exchange I’m increasing taxes paid by a special interest group – particularly when the group in question is one of the few industries that is almost as unpopular as Congress.

But, it’s not a particularly intellectually honest position.  The privilege fee paid by Kansas HMOs is one part of their cost of doing business; and it is rational to expect that those HMOs will ultimately adjust their pricing – that is, the insurance premiums they charge – to reflect any increase that legislators enact in that aspect of their costs.

So, at the end of the day, this tax increase will be borne broadly:  by individuals who purchase insurance policies directly from HMOs; by employers who purchase insurance from HMOs for their employees; by the employees of those employers, to the extent that the employers react to the premium increases by ratcheting up the portion of healthcare benefits that they ask the employees to fund themselves; and by the consumers of those employers’ products, to the extent that employers react to the premium increases by increasing the prices they charge for their goods and services.  I think it will also be borne by one other party, which makes this even odder…

Won’t KanCare end up having to pay for part of this tax increase?   As far as I can tell, the 5.5% privilege fee will apply to all the premiums that Kansas HMOs write, including those for KanCare business.  But what the HMO contractor calls revenue, the state Medicaid program calls an expenditure.

So, to the extent the HMOs need to increase the premiums they charge KanCare in order to cover the privilege fee increase, doesn’t that automatically lead to increased Kansas Medicaid expenditures?  A recently-adopted Actuarial Standard of Practice reiterates that, in order to be actuarially sound, the rates that Medicaid agencies pay to HMOs need to take into account any taxes or fees that the HMOs need to pay out of the revenues they receive.

Thus, we have some dollar-trading here:  The HMOs that participate in KanCare will pay higher privilege fees, but then they’ll receive correspondingly higher revenues from KanCare.  This doesn’t strike me as a model of efficiency.

Having said that:  I’m not a Medicaid expert, but the fact that Medicaid is a joint state/federal program may mean that it’s not purely dollar-trading viewed from the standpoint of the state of Kansas.   If Kansas collects an additional dollar in privilege fees from a KanCare contractor, and then as a result that contractor needs an additional dollar in revenues from KanCare, does the federal government end up subsidizing some portion of that new dollar in revenues?

Differential tax rates on market participants can create market distortions.   The privilege fee we’re discussing applies only to HMOs, who pay this fee but in exchange are exempt from the Kansas premium tax paid by all other insurers.

In my view, this is an accident of history.  While there are exceptions, today most HMOs operate in a manner very similar to most other health insurers, in that the underlying healthcare services are provided not by the HMO’s own employees but by an unaffiliated network of providers; the main difference today is that, typically, an HMO network will be narrower than a non-HMO network.  This convergence of business models over the past couple of decades calls into question historical practices in some states that treat HMOs differently from other health insurers.

In the Kansas case, today HMOs enjoy a modest competitive advantage relative to other health insurers:  HMOs pay a privilege fee of 1.0% of premiums, whereas other health insurers instead need to pay premium taxes equal to 2.0% of premiums.  The proposed legislation would turn this around, and give non-HMOs a very significant competitive advantage.

This in turn can be expected to have ramifications on choice in the broader health insurance market in Kansas.  It’s a fair generalization that HMO products tend to be cheaper than non-HMO products, to the extent that HMO products generate cost savings via the usage of narrower provider networks.  By placing HMOs as a tax disadvantage to non-HMOs, the inherent cost advantage of those HMO products will shrink, which arguably will make consumers worse off.  (This is the argument being advanced by Aetna, according to the Wichita Eagle article.)

The tax increase will, in a sense, be retroactive.  I argued above that, ultimately, HMOs will react to this tax increase by passing it through to their customers via increased premiums.  And while that may be true in the long run, it would likely be far less true in the immediate term.

The statute calls for the increased privilege fee to apply to all premiums written by HMOs in calendar year 2015.  The problem is that insurers in general (including HMOs) price their products well in advance of when the premiums take effect, and then by regulation hold each customer’s premiums steady for an agreed policy term (typically 12 months for health insurance, as opposed to 6 months for auto insurance).  So, presumably all the HMOs in Kansas had priced their 2015 premiums under the assumption that the privilege fee would remain at 1.0%, and cannot go back and change those premiums if the statute is enacted.

Thus, from the HMO’s standpoint, this proposal represents a material change in the rules in the middle of the game.  An unanticipated expense in 2015 equal to 4.5% of premiums (the new rate of 5.5%, less the existing rate of 1.0% already reflected in pricing) could absorb most or even all of an HMO’s expected profits for the year, health insurance being a relatively low-margin business.  Regardless of how one may feel about the HMO industry, there’s something untoward about a mid-year tax change that can be expected to have such a dramatic impact on the short-term financial prospects of an industry, given that by regulation the industry cannot take immediately effective action to react to the changed circumstances.