For a while now, some conservative health-care wonks have been touting the concept of “automatic enrollment.” The idea is that when someone fails to sign up for health insurance, the government could simply arrange (minimal) coverage for them behind the scenes, using the same subsidy money the person would have gotten if they’d signed up. It’s a less coercive alternative to the individual mandate that, depending on how it was set up and whether it was part of a bigger change to the health-insurance system, could ensure universal coverage.
The problem is practical: People can lose or gain insurance as easily as they lose or gain a job, so it’s very hard to tell exactly who is uninsured at any given moment. Over at the Health Affairs blog, in response to these concerns, three health-care experts — Stan Dorn, James C. Capretta, and Lanhee J. Chen — have outlined some ways that this could be accomplished, in particular highlighting various data sources that could help identify the uninsured for auto-enrollment.
This is a possibility worth taking very seriously. It’s a way that states can control any instability that results from the elimination of the individual mandate (set to go into effect next year), and a way a future Congress could ward off calls for single-payer. But here’s a different way to implement the idea that I think would pose considerably less difficulty on an administrative level. Keener minds than mine would need to work out the technical details, but for the record Capretta tells me it’s “not a crazy idea.”
Basically, when we try to identify the uninsured at an individual level we’re wasting a lot of resources. We’re matching people to plans that they might not even use except in the unlikely event something bad happens, and that they might not even need a month from now (e.g. if they find a job with benefits).
Instead, what we could do is estimate the overall size of the uninsured population in each, say, county. (In fact we already have such estimates.) Then we could simply give those folks’ unused subsidies to insurers in exchange for an agreement to cover the small fraction who later needed care. Perhaps insurers could make “bids” in the form of plans, with the credit money going to the insurer willing to provide the best coverage. When the uninsured actually showed up at an ER or a doctor’s office, they could be matched to an insurer at that point — and they would identify themselves so the government wouldn’t have to.
The end result is the same — if you don’t sign up, a bare-bones “default plan” is available to at the very least protect you from catastrophic expenses. But most of the administration would be handled at the level of county populations rather than individuals.