Thread by @bjdickmayhew

🚨🚨 NEW RESEARCH ALERT 🚨🚨

@petrasmussen and I have published "When All that Glitters is Gold: Dominated Plan Choice on @CoveredCA for the 2018 Plan Year" in the @MilbankFund Quarterly

We look at consumer choices after #Silverload

TLDR: People are confused

1/13

Dominated choices happen when on all relevant attributes Item A meets or beats Item B.

In a rational framework no one should ever choose Item B as it is dominated by A.

Our research is part of a growing literature that shows many people choose Item B for health insurance

2/N

Institutional context: Plan Years 2014-2017 #CostSharingReduction subsidies were directly paid to insurers by the Federal Government. CSR only applies to silver plans and there are 3 flavors (high, mid, low) from 100-150 FPL, 151-200 FPL and 201-250 FPL respectively.

3/N

4/N Fall 2017 The Federal government stops paying CSR to insurers. @CoveredCA building on ideas by @UrbanInstitute Blumberg and Buettgens respond by embracing both the #Silverload and #SilverSwitcheroo.

CSR costs are placed into the premium channel for only silver plans.

5/N This produces a huge relative price shift. Silver plans become way more expensive compared to same-insurer/network Bronze and Gold plans. In California, 2 insurers had 2018 Gold plans priced below comparable Silver plans (same network/plan type/insurer triad)

What happens?

6/N @petrasmussen in previous research in @Health_affairs found lots of people moved around and switched plans in the face of a price shock. I asked for a Sankey and this is what motivated our paper. 2018 looked WEIRD lots of movement out of Silver
healthaffairs.org/doi/full/10.13…

7/N We identified that for folks earning over 200% FPL the two triads Silver plans would be strictly and transparently dominated by the Gold plan -- Gold offered same network, insurer unobservables, plan type for less premium and less cost-sharing for any level of utilization

8/N from a policy POV, ideally no one earning over 200% FPL in these triads ever buys Silver.

That sooooo did not happen

9/N Our sample was ~175,000 individuals who earned over 200% FPL and enrolled in one of these two potentially dominated insurers in 2018.

~35,000 people enrolled in a dominated plan in 2018

20% enrolled in a dominated plan

$450 more annual premium + lot more deductible
#OUCH

10/N #Inertia is a big story --- Enrollees who were in a 2017 Silver plan were disproportionally likely to be enrolled in a dominated plan in 2018.

Part of this is #AutomaticReEnrollment part is active choice inertia

11/N However even people making active choices still chose dominated plans
7% new enrollees chose dominated plans
~10% switchers from other insurers chose dominated plans

Active choice in a transparent domination scenario helped but still a lot of objective errors were made

12/N So what can be done?

Feher and Menashe use soft informational nudges
healthaffairs.org/doi/10.1377/hl…
to get modest effects

But I think that is too limited....

13/13 In this paper, we propose better #ChoiceArchitecture --- either show the dominant plan first or just hide choices that are strictly dominated altogether.

Choosing insurance is hard -- let's make it a bit easier.

14/13 Post-script --- this paper is one of the two projects that really motivated me to go back for my doctoral studies at @duke_pophealth in a few weeks. It changed my priors a lot. We also have some cool and related work coming out next week, so stay tuned.

Research link is here: onlinelibrary.wiley.com/doi/10.1111/14…