Dominated plans are plans which on all relevant characteristics are no better than or clearly inferior to a single alternative. A person choosing a dominated health insurance plan offered by the ACA is making an objectively inferior choice where they are paying more in premium or being exposed to more cost-sharing or are buying a smaller network when a larger network that includes the entire smaller network as a subset is available at the same premium and cost-sharing. Under a rational choice framework, no one should make a dominated choice. We don’t live in a rational choice universe. Dominated choices in health insurance are wicked common. Recently published research from California that we talked about on this blog showed that even with modest nudges, three out of four people who were exposed to a dominated/non-dominated choice where the non-dominated choice required active intervention ended up in the dominated choice.
Why is this the case?
And more importantly, why is this systematically the case instead of a matter of individual inertia or failings?
I think the short answer is that there is no one making money on good choice selection.
An agent either gets a flat commission per month enrolled or a percentage of sold premium. If they are given a flat commission, then the extra work to move someone out of their default plan and into a new plan only pays off if the person first recognizes that the agent did them a solid and then comes back in future years. This is a highly discounted revenue stream for the ACA individual market as a lot of people move in and out of the marketplace. If the agent is paid as a percentage of premium, their short run incentive is to place people into the highest premium plan possible. There are very few agents who are explicitly paid on the basis of avoiding dominated choices.
An insurer makes money when people are making dominated choices. In the California example above, the dominated choice was likely an 80% AV Gold plan with a higher premium than the 87% or 94% CSR Silver plans. When someone chooses a dominated plan in this setting, the insurer sees both higher revenue and lower claims pay-out due to both the static effect of more cost-sharing and the dynamic effect of less utilization due to the additional cost-sharing. There is no insurer side incentive in the short run to call members to get them to switch to the superior, dominating plan.
Exchange boards are trying to make their marketplaces work. There is significant interest in improving the active shopper experience and I know many state based marketplaces are actively thinking about how to give moderate to strong choice architecture nudges to keep people out of objectively bad for them plans –> one example will be that for anyone who earns under 150% FPL will be shown CSR-94 Silver zero premium plans first before any other zero premium plans. These are likely to be effective nudges and architecture choices but people will still make errors. However, their funding is not dependent on maximizing objective choice quality and minimizing or eliminating the possibility of dominated plans being chosen.
Right now, there are modest incentives on the part of mission-focused marketplaces to improve active chooser choice quality with nudges and a great amount of expectations that individuals will be able to navigate a complex choice environment. Entities with the expertise, experience and ability to navigate clear cut cases of dominated choice are either financially indifferent to choice quality or actively benefit from poor choice quality.
I am not sure how to fix this problem, but we know that dominated choice is a pervasive problem on the ACA individual marketplaces, and the current financial structures don’t help solve that problem.
PaulB
Based on my experience with the Washington state exchange, there isn’t enough information provided to me to tell me whether the plan I just selected is a dominated plan or not. You mention, for example, the network but aside from being able to check local options one at a time, there isn’t much I can do to determine the network.
Lobo
Entities with the expertise, experience and ability to navigate clear cut cases of dominated choice are either financially indifferent to choice quality or actively benefit from poor choice quality.
I am not sure how to fix this problem, but we know that dominated choice is a pervasive problem on the ACA individual marketplaces, and the current financial structures don’t help solve that problem.
Yes, it is the classic case of profit motive versus cost savings in health care. Do we want to profit maximize or client maximize. These two are in tension in our present system. This is a case where possible machine learning or AI could be better, but biases could be built into these. Make this your Ph.D thesis.
Lobo
@Lobo: My solution would be to create a “good enough” model plan on paper(a psuedo-public option). The system would then show the person how their current plan compared and offer maybe three others that might compare better. The people could ask to compare on one or all three of these factors premiums(cost), benefits, and service. This is where AI/machine learning would come in. Hopefully, this would balance out all the constraints. It would not look to perfection but to good enough.
David Anderson
@Lobo: I would not be surprised if dominated choices ends up somewhere in my doctoral work as either a 2nd year paper or part of the dissertation. However, yesterday, I was chatting with a friend, co-author and faculty at the department where I will be studying and I received some excellent advice from her — don’t worry about a dissertation just yet…. :)