John McDonough, professor of health policy at Harvard, lays out in Politico what he is hearing from Republican health policy wonks on what they want to do about health policy if they have a trifecta again. There is one sentence that I want to really pull apart:
Doug Badger, who doubles at the Heritage Foundation and the Galen Institute, offers a list: “Republicans still believe that ACA premiums are too high, networks are too restrictive, and cost-sharing too burdensome.
The ideal plan from a patient’s perspective is a no premium plan with absolutely no restrictions and no cost-sharing. That is also a fantasy if we have to vaguely think about costs.
Let’s assume that we have a system where the sick and likely to be sick/expensive individuals can get and maintain coverage as well as a system with little to no governmental price setting. I’ll relax that assumption later but right now, there is a trilemma where we are at a choose two (at most) situation.
** We can have low premiums. Premiums need to be sufficient to pay for claims. Claims are a function of both utilization which is a function of population health interacting with provider treatment preferences interacting with benefit design and the price per unit of service rendered. Finding ways to drive down either utilization or the price per unit will lead to lower premiums.
** We can have plans that have broad networks and few restrictions. In that universe, the clinicians have all the leverage and insurers can’t get good pricing per unit. These plans will attract high need patients.
** We can have low cost-sharing. This means the insurer is paying through the premium channel a higher percentage of allowed claims. This also means that people will use more services (both high and low value services.)
All of these things have tensions.
We can have low cost-sharing and very broad network plans with few if any restrictions. Those two incentive sets means premiums are likely to be high as hell as insurers won’t have any ability to credibly threaten “NO” in a pricing dispute and utilization will be high.
We can have low cost-sharing and low premiums. In order to get low premiums this means claims have to be rare and the per-unit pricing has be low. To get low per-unit pricing, the insurer needs to be able to credibly threaten “NO” in a pricing dispute which means the ability to assemble a massive network with 95% participation and not have government price setting like in Medicare is a massively unreasonable assumption.
We can have low premiums and a broad network as long as the patients are picking up a good chunk of the tab through high cost-sharing. Claims will be heavily borne by the insured population and high cost sharing may deter some needed and plausibly unneeded care as well.
It is also quite plausible to get a broad network with high cost sharing and high premiums. In almost all of the United States, in the commercial markets, the insurer and consumer is in, at best, a CHOOSE TWO situation.
Now if we are to assume that the policy environment allows for the private insurance markets to bifurcate the market by active and passive underwriting, then the segment of the market that serves mostly healthy people will have dirt cheap premiums, broad networks and low cost-sharing as this is a group that barely uses health care in a given year. The coverage might have significant holes in it. Good luck finding maternity coverage and mental health coverage but the premiums will be low to the low chance of receiving a cancer diagnosis in any single year at which point reclassification risk occurs.
This schema will carve out 10-20% of the population that drives most of the predictable costs. The favored conservative solutions are either tax incentives or high risk pools that are high cost sharing and underfunded but in this domain, carving out most of the expensive people away from the “normal” insurance system. Now if these carve-outs were adequately funded this could work if we are committed to an insurance market that bears only actuarial and idiosyncratic risk but making sure that these carve-outs are adequately funded means serious money needs to come from somewhere that is not the patients. And that is a huge question in all of these schemes.
So wrapping things up; if we have a guaranteed issued market without administrative price setting then we have a trilemma of premium-cost sharing-network restrictions. If we have a bifurcated underwritten market, the carved-out minority is either bearing massive health costs or they need massive explicit governmental subsidies.
Everything else is a detail.