More churn than a butter factory

Two conservative health policy wonks are outlining an auto-enrollment option that is philosophically aligned with what is in the Collins-Cassidy draft bill state option. I don’t have an intrinsic problem with auto-enrollment with an opt-out. I have a major pragmatic problem with their proposal. But let’s look at the core of their program:

Congress should also help policyholders avoid breaks in coverage. Lawmakers could require insurers to offer products whose premiums match the value of the federal tax credits. If the basic, age-adjusted federal tax credit for a 40-year-old man in a given state is, say, $3,000, then every insurer in the state would have to make a policy available for such customers with a $3,000 premium.

Insurers would adjust the upfront deductibles in these plans as necessary to ensure that the premium equals the credit….

As with other benefit programs, many Americans wouldn’t use the credits for which they are eligible, out of inertia or lack of information. To solve that problem, the states could automatically place eligible households into the no-premium option, randomly assigning them to one of several competing insurance plans and then notifying policyholders of their coverage.

They are identifying a real problem with a reasonable way of minimizing the problem. Recent estimates have the Medicaid Woodwork effect contributing to 30% of the total decline in uninsurance i 2015. The Woodwork Effect is when people who were eligible but not signed up for Legacy Medicaid hit the Exchanges as health insurance enrollment became a major topic of public discussion and found out that they were always Medicaid eligible. This happened in Expansion states, it happened in non-Expansion states, it happened in Red states, it happened in Blue states.

I have a major pragmatic objection to their program design.

How do we figure out who is not otherwise covered in any given month?

The entire non-employer sponsored insured, non-government sponsored insured group can be roughly divided into four groups. The first is non-documented immigrants, this plan most likely will exclude this cluster for political reasons. This would not be a major change.

The other three groups are the problem.

The first group are people who will be in good health but need insurance on the individual market for a long time period. This group are the artists,the consultants, the very small business owners, the people who are fairly reliable buyers. If they were healthy, this class was in the individual market in 2009. They may switch carriers for cost and network purposes but they actively participate in the market year over year.

The second group is also a group that was in the individual market in 2009. It is people who just need some coverage until something better came along. When I was laid off in 2009, my wife and I bought a catastrophic policy that covered us for a few months until we were able to get back onto employer sponsored coverage. My daughter was on that policy for two months until we got her on CHIP (best insurance I’ve ever had, great network, low premium, no stress). These are the policies which function as holding areas until something better comes along.

The final group in the ACA individual market are people with significant health conditions who were previously underwritten or priced out of the market. As a cohort, these individuals are also very stable in participation. A hemophiliac or an individual with MS will make sure that they sign up for coverage as they burn through any deductible by the third week of the policy year.

The first and third groups aren’t a major source of my pragmatic objection. The second group is a major source.

For relatively healthy people who are mainly insuring against catastrophic expense (hit by a meteor coverage) they are likely to churn. With an auto-enrollment system some level of government will need to track eligibility of every single potential covered life on a day to day basis. If someone is laid off on the 14th and their employer has a coverage cut-off on the 15th of the month, are they covered if they have a heart attack on the 16th? They probably would be on a retrospective claims basis but coverage churn is a major concern unless there are either massive back ends claims reconciliation for retrospective eligibility, states absorb the first sixty days of costs of a patient who is not part of a prospective assignment, or some other churn identification.

There are work-arounds but the pragmatic challenge of churn is real and it is big.

11 replies
  1. 1
    Barbara says:

    The bottom line is that the way we do health coverage in the U.S. produces an unbelievable amount of friction, which has the effect of decreasing coverage (it’s very easy to slip up and get confused about what the rules actually are) and increasing conflict between insurers and their customers and different insurers (You’re primary; no you’re primary). I can’t tell you how much effort is wasted on this stupid thing called Medicare Secondary Payer, which is Medicare’s way to MAKE SURE that it gets as much third party liability reimbursement as possible. It sounds right, it sounds fair, but it has become the tail wagging the dog, with employers in particular having to build systems and processes to address the possibility that something like less than .1% of their covered employees have both employer provided and Medicare coverage and are actively working. It’s even worse in Medicaid and as David indicated, because there is so much cycling and the potential for discontinuity is so high. This is the price of fragmentation and it’s actually pretty high.

  2. 2
    gene108 says:

    Second problem I see, is what kind of insurance will you get for only a $3,000 annual premium?

    Unless you are 23, fit and single, I don’t see 3k in premium support going very far, and the plans will still have very high deductibles, which is one of the complaints about the current systems: The Low premium plans have high enough deductibles that they are not cost effective for many people to meet chronic or routine healthcare needs.

  3. 3
    Brachiator says:

    Congress should also help policyholders avoid breaks in coverage. Lawmakers could require insurers to offer products whose premiums match the value of the federal tax credits.

    So, the essence of this proposal is to get around the need for any subsidies?

    I appreciate the way these issues are explained, but it’s hard to work up any enthusiasm for any of this since Congress’ bottom line seems to be to do whatever they can to help the insurance industry (and the pretense of the efficacy of the free market). So, for example:

    Second problem I see, is what kind of insurance will you get for only a $3,000 annual premium?

    Probably not any coverage that satisfies the present standard of minimum essential coverage.

    Republican politicians like to pit one group against another. So, they ask why a healthy young person unlikely to get seriously ill should be forced to buy insurance. I would ask is there some combination of low price and high deductible or some other set of features that might make participation in the current system by young healthy people feel less burdensome?

  4. 4
    Barbara says:

    @Brachiator: I have been doing this for a long time, and I can’t tell you how exhausting it is to become an expert in arcane rules that exist solely for the purpose of finding the line between whether you are covered under this program or that program or no program at all; for instance, how to escape losing coverage after you lose your job (were you fired for good cause or just because?) or whether your stepson is actually your legal dependent or how to measure 63 days from the date your last coverage ended (is day 1 the last day you were covered or the first day you weren’t covered?). I could write several pages if not actual books of this kind of stuff. Simplicity is not an end in and of itself but people who have wonkish ideas need to understand that complexity adds cost and confusion and often enough defeats the purpose for which rules have been made.

  5. 5
    Brachiator says:


    I have been doing this for a long time, and I can’t tell you how exhausting it is to become an expert in arcane rules that exist solely for the purpose of finding the line between whether you are covered under this program or that program or no program at all;

    I got nothing but respect for people who do good, hard work in this field.

  6. 6
    Victor Matheson says:

    While Mayhew is correct to be concerned about churn, it should also be noted that there is also a huge amount of status quo bias for people as well. It is a lot harder to get people to sign up for something than to re-enroll once they are in the system. The most famous example is auto-enrollment for 401K plans. Even though it literally takes 15 minutes at most to sign up for a 401K at work places that offer them, the move to auto-enrollment with an opt-out increased employee participation in these plans by 30 to 60 percentage points and there was very little increase in people dropping their plans later (about a 1 percentage point increase).

    I would think that it would be worth investigating whether the government should automatically enroll every person in the Social Security system in a silver plan at moment they turn 26 (and are no longer eligible for their parent’s insurance.) Basically every 26 year old gets a birthday card with a bill for insurance. At that point you can either pay the bill, choose a different plan, return the form stating you have insurance, or a return your card stating that you understand you will be going without insurance and are subject to the penalty and full responsibility for medical costs.

  7. 7
    Victor Matheson says:

    @Brachiator: One thing you could do is expand the rate band from 3:1 to 4:1. That will make insurance more expensive for older people but cheaper for younger (and healthier) people. But add enough younger people (by giving them this incentive) and you might not see too much increase in the rates for older people.

    Another thing you could do is make the penalty for not having insurance larger.

  8. 8

    @Victor Matheson: I like that idea of a 26 year old card — but you have to make it 18 — sign up for the draft, buy a lottery ticket and then sign up for insurance. 26 only applies to en masse to middle or upper class kids whose parents are in “good” jobs.

    Your point on status quo bias is why I am okay with this proposal on the level of ideas. We know people will go with the flow so shifting the flow to be covered instead of not covered is a win.

  9. 9
    tim jost says:

    I would add yet another group that I am not sure is captured in the four categories listed–people who just don’t want health insurance. They rarely need health care and when they do go to the local community health center or free clinic or work out some arrangement with a local doctor or local hospital. They would be quite upset if they found out that someone had enrolled them in coverage even if they could opt out of it There are a lot of people here in rural Virginia whom I think would fit in this category..
    Also, it seems to me that the problems of auto-enrollment in terms of identifying the uninsured, assigning them to a health plan, informing them as to who their insurer was, telling them who was in their network and what drugs were on their formulary, and ensuring that they did not end up owing large out-of-pocket costs for out-of-network providers are insurmountable, or nearly so. A great idea, but impractical.

  10. 10

    @tim jost: I see what you’re saying. My question is does it matter on two grounds?

    First, a paternalistic ground that society says that everyone “should” have insurance unless they specifically choose otherwise. Motivated individuals who really don’t want insurance can opt-out. And I imagine the law could be written in such a way that people with religious objections to insurance (Amish etc) could be presumptively excluded.

    Secondly, on a far more pragmatic basis, they are paying nothing out of pocket for minimal catastrophic coverage. They are not mandated to use insurance when they go to the clinic. I don’t think there is anything in any of the proposals floating around that would mandate that insurance gets billed nor insurance information would have to be presented at the point of service. If someone wants to pay cash for their one interaction with the medical system every few years, they should be able to do so.

    I think there is a difference between a priori dislike of insurance which given utilization patterns won’t be challenged hard by circumstances for most of the people in your proposed bucket and the moment of truth when one of these individuals is told that they have cancer where insurance is now extraordinarily valuable even if it comes with a one time $10,000 deductible.

    I need to think about this some more.

  11. 11

    @tim jost: The second part of my response is I agree with you prospective assignment sounds like a nightmare. Moving towards a retrospective assignment with presumptive eligibility for a network of all Medicare accepting providers is a pragmatic work-around. Besides having an insurer compete to be a TPA, I am trying to figure out why a private carrier adds significant value here over and above the transaction and information frictions that they impose as you outline.

Comments are closed.