As you all know, December was an excellent month for the ACA exchanges, federal and state. The official HHS totals at 28 December were 2.15 million for market policies, 1.58 million for Medicaid/SCHIP. Charles Gaba’s running total for all new ACA coverage – including the mass Medicaid baptisms by transfers from state schemes, private policies contracted outside the exchanges, and under-26es added to their parents’ policies – stands at a round 10 million today.

The HHS report rightly ignores the puerile talking point that the private policies aren’t all paid up. So what? Why should anybody shop with great difficulty for medical insurance, choose a policy, and abandon it at the till at the last minute like an excess packet of cornflakes?

It does go in detail into the demographics, as a high takeup by healthier young people will be crucial to the actuarial viability of the plan. I’ve nothing to add on this issue to the full commentary of others. Matt Yglesias’ worries were answered, to my mind conclusively, by Kevin Drum, Sarah Kliff and Josh Marshall. Short take: the ranking expert at Kaiser says the takeup by under-35s is currently at the minimum for viability, with modest premium increases, and on past experience and by common sense the proportion will rise. Above all, the insurance industry is silent. If there were a real risk of a death spiral, they’d be screaming blue murder.

Instead, let’s take a look at the political and social geography. Here is Gaba’s table of the takeup of market policies by state as of 13 January, as a proportion of the uninsured.
He sensibly uses the ratio to all uninsured rather than the iffy CBO target projections which have acquired an undeserved authority. The numbers include in bold his own updates for the state exchanges, not those in the HHS report – which distorts the comparison a little; but since he was 99.2% accurate at end December, he has earned the right to be considered reliable. I’ve inserted a line with the US average.

In some respects the pattern is what you’d expect. The Blue coastal states have higher takeup, the Red southern and Midwest states have lower takeup.

However it’s far from uniform. Alabama, Utah, Florida and Tennessee are close to the national average, North Carolina is well above it. On the Blue or at least purple side, Delaware, Illinois, Maryland, Ohio and New Jersey are well below the average.

What accounts for the very large variance – a range of 28 to 1? This will narrow over time, but the work-in-progress snapshot still calls for an explanation. I think we can discount the websites as an explanation, apart from the very dysfunctional sites of Maryland and Oregon. Four out of five – 83% – of all the signups came in December, when healthgov.com was performing about as well as the better state websites. Most customers faced a pretty uniform environment: the same law, an identical or very similar website experience. The variation must be due to local factors.

The federal marketplace is just an administrative clearing-house, and the underlying markets are statewide not national. The risk pools are distinct and the policies on offer are not the same. But are the prices really very different? The bigger insurance companies will surely be using standard actuarial models, partly merging the risk pools. Pending more data, we should look at the demand no the supply side.

The most economical hypothesis is that it’s down to state cultures and above all state policies. The governors of Blue states have been trying to make ACA work, with positive media messages, help to navigators and so on. Setting up their own websites is an effect of this political support rather than an important independent variable. Those of Red states have been trying to make it fail. Rick Scott’s effort in Florida has flopped. The jury is still out on Rick Perry’s Texas.