Abstract
In October 2013, the Affordable Care Act introduced a new insurance market — state and federal exchanges where people can purchase health insurance for themselves or their families. Although the rollout of the exchanges was disastrous, around-the-clock efforts fixed many of the biggest technical problems, and nearly 7 million people purchased insurance in the new market. The second round of enrollment exposed some new problems with the exchange websites — for example, Colorado's website had difficulty determining whether people were eligible for tax credits — but these problems paled in comparison with those encountered when the exchanges were first rolled out. In short, we have a largely glitch-free system of health insurance exchanges that present millions of people with a robust set of health insurance choices.
Which means that it will soon be time to tackle the much more challenging job of designing exchange websites in ways that maximize the chances that consumers will choose plans best suited to their needs and preferences. If the first round of open enrollment was primarily about avoiding catastrophe and the second round was about ironing out wrinkles in the underlying programming code, then version 3.0, in our view, should focus on redesigning the way exchanges present their insurance choices, to avoid features known to bias people's decisions.