I gave a short talk earlier in the week to a bunch of ACA nerds. As I was prepping my slides about choice complexity, I created a graph showed something that I sort of had realized but not actually seen or fully internalized:
We know that picking insurance is really tough. The Medicare Advantage literature suggests that choice quality goes down hill rapidly after people see 15 or more choices. Some of my ongoing research looks into administrative burdens and choice frictions that lead to sub-optimal choices and dominated choices. We’ve talked about the increasing proliferation of large choice menus before. The work that I did with Patrick O’Mahen looked at the population with at least 50+ choices and in 2021, that was about 45% of people were exposed to 50+ choices and when I did a short off the cuff update, 2022 saw 67% of people seeing 50+ choices.
I had not realized that the count of extreme menu lengths were becoming so much more common over the past couple of years. I had known that big cities like Miami and Houston had ridiculous menu lengths but in 2023, about 10% of all Healthcare.gov counties have very large menus.
When I poked at the causes of this menu length expansion, it is a bit of everything. More insurers are entering markets on average. Each incumbent insurer is adding a couple of extra plans per year per county. And some of the insurers that are expanding their footprints have a historical strategy of “stacking the shelves” with a lot of nearly similar plan variants.