The more I read about how regular people interact with the new health insurance exchanges created by the Affordable Care Act, the more convinced I’ve become that most people probably aren’t making the optimal decision.

Whether it is surveys showing a majority of people said it was difficult to compare plans, assessments showing very few Americans have a proficient level of health literacy, the large amount of help people need to use the program, or studies based on actually watching people use the system, I come away with a strong sense most people lack even the basic insurance knowledge needed to understand one policy. So there is little reason to believe they can then do the complex analysis among multiple policies required to actually make the best choice for them.

Obviously, this inherent problem is one of the many arguments for single payer. In the short term though I think there is a partial improvement that should be looked at: having the exchanges automatically assign everyone to a plan when the register based on their personal data. People could still ignore this strongly pushed advice and select their own plan if they really wanted.

To do this most effectively you would need to upload all of a person’s medical information into the system with their income data so a sophisticated formula could determine what is likely the best plan for them. That setup would possiblly run into serious technological and privacy concerns, but I imagine having people even just answer 30 question about their health care/lifestyle would provide enough additional information for a formula to make a good recommendation. That is also something that could be implemented fairly easily.

This could create two improvements: more people would make the best choice and it could make insurers behave better.

The biggest problem it solves is making sure more people make the optimal choice. Given how little most people seem to understand health care terminology and how it all fit together, I think even a well written formula would outperform the majority of exchange users.

It would potentially also have the added benefit of improving the quality of the insurance plans that are offered. Right now insurers may have the incentive to design their plans with the goal of tricking uniformed shoppers. Making plans merely look good to low information users could prove to be a more effective way of attracting customers than actually creating quality policies. If instead most people stuck with their formula assigned planned, insurers would have the strong incentive to focus primarily on meeting the quality metrics of the formula.

It would be interesting to see states experiment with this idea given that it could be implemented easily on a purely voluntary basis.

I have almost no faith in the idea of health insurance exchanges because of the incredible information disparity between the insures and the regular users. At least with this change the dynamic struggle would be between companies and a team of trained government actuaries instead of between very sophisticated for-profit companies and regular people who barely understand what they are being forced to buy. I suspect the former would produce better outcomes for most people.