Embracing the confusion

In Health Affairs, three researchers set up a simplified Medicare Part D choice experiment that used the plan finder to make better choices.  This is a big problem of choice on the Medicare Part D (and I expect on some ACA exchanges):

A sizable body of evidence indicates that suboptimal plan choices are pervasive. The vast majority of Part D enrollees do not choose the lowest-cost plan available to them.⁵ After the amount of protection that plans provide against unanticipated drug costs is accounted for, people are still estimated to overspend by about 30 percent of their total drug spending each year.^6,7 Beneficiaries rarely switch plans, despite annual changes to plans’ costs and benefits,^8,9 and it appears that plan selection has gotten worse over time despite older adults’ gaining more experience with the Part D program.⁷ …

When we examined the cumulative proportion of plans chosen by their total cost rank across the four study arms, we found that people in two of the experimental arms (the total cost only and total cost, premium, and out-of-pocket cost groups) selected the lowest- or second-lowest-cost plan about 65 percent of the time (exhibit 2). In comparison, people in the third experimental arm (the total cost and premium group) selected such plans 54 percent of the time, while control-group participants did so just 44 percent of the time.

Picking an insurance plan is tricky. And the business case of providing easy to make and near optimal choices is weak for an insurer.

If there is optimal choice defined as the patient minimizing their total expenditures and we assume that the prescribed drugs are going to be prescribed no matter what insurance policy, then an optimal choice means, from an insurer’s point of view, that they get some combination of lower premium revenue or higher share of drug expenditure as the out of pocket expenses that split the cost of the drugs are minimized. If everyone goes into the optimal plan for them, the insurer will be having lower premiums and higher expenses than if there is confusion and non-optimal and barely satificing choices made.

If one insurer offers exceptional decision support tools and another insurer offers mediocre or no tools to help with decision support (instead of merely marketing tools), the first insurer selects for a whole lot of uncompensated risk. Good decision support offered by an insurer is a risk selection magnet.

Now if we want people to have good decision support tools to help them make a complex choice, those tools have to be both effective and come from an outside third party. The current CMS tool for Part D plan choice does not seem too effective if less than half of the users pick a near optimal plan. This needs to get better.

If insurers know that there is a good, reliable, third party decision support tool that is widely used, they will adapt. Some of that adaption will be raising premiums for certain groups while other changes will be different benefit structures to capture small segments of the market.

I’m interested in this paper as I’m slowly putting together a research project that looks at ACA plan choices in an even simpler environment than the one that was tested. The TLDR from this project is that buying insurance is hard to do.