California has convinced me to walk back "How many have PAID??" a bit: 88%

The massive data dump from this afternoon out of Covered California was mostly excellent news, both for the ACA as well as myself personally:

  • I nailed the total QHP enrollment figures during the off-season almost precisely (99% accurate on the total number, 96% accurate on the daily average)
  • The gross attrition rate has been only 1.6% per month, and the net attrition rate since April is an astonishingly low 1.8% over 6 months, or just 0.3% per month, which is far better than anyone (including myself) had imagined. Basically, people being added are almost entirely cancelling out people dropping coverage.
  • The list of technical improvements, new features, beefed-up staffing and ramped-up outreach efforts for the 2nd open enrollment period included in the press conference was extremely impressive (see the lower half of this entry from earlier today).

HOWEVER, there was one key data point which was quite surprising and disappointing to me: The initial payment rate.

As I noted earlier, Peter Lee, the director of CoveredCA, stated point-blank no fewer than 5 times last March and April that 85% of all QHP enrollees on the exchange had paid their first month's premium. Not once, not twice--5 times over about a month-long period. There was no "roughly"; it was 85%, period.

Since all of these pronouncements were made in the midst of the enrollment period, when a good 1/3 of the people signing up hadn't even started their coverage yet, it was natural for me to assume that the eventual payment rate would end up a bit higher; probably around 90% even.

Since other states like Minnesota, Rhode Island, Connecticut and even red states like North Dakota and Wyoming were all reporting payment rates ranging from 90-95%, this seemed quite reasonable, and right in line with my expectations.

However, today's announcement that the actual payment rate in California specifically was only 82% (Lee said 81%, but the actual numbers he cited divide out to 81.7%) requires me to backtrack that a bit. I'm not exactly eating crow here; I can only operate on the best data presented, and the head of the exchange had repeatedly given the 85% figure for California without any hard evidence being presented to the contrary.

However, after shouting "90%!!" from the rooftops for so long now, I do have to admit when I make an error. If it was a smaller state like South Dakota or Delaware this correction wouldn't impact the national average at all.'s California. Largest state in the country, and responsible for a full 18% of the total QHP enrollments nationally.

When their payment rate drops from an estimated 90% down to 82%, it has a significant impact. Result: The national "1st month paid" rate on the spreadsheet has dropped 2% overall, from 90% down to 88%.

Does this matter all that much in the greater scheme of things? Probably not. 88% is still a hell of a lot better than the absurd 67% figure that the GOP was trying to claim back in May. And it could be argued that the fact that the net attrition rate is proving to be far lower than expected is more significant in the long term.

In addition, my guess is that the average payment rate will increase a bit in the 2nd enrollment period (possibly right back up to...90% overall), since more people will have gotten used to how the system works, most of the exchanges will have vast improvements in their technical and billing systems and so on. For example, while only 2 states were really set up for payment-at-point-of-enrollment last year (Washington and Massachusetts), and only one of those really actually worked (MA's, of course, was a mess), this year, CoveredCA is adding the ability to pay the first premium at the point of enrollment. It won't be required, and it won't be set up for every insurance provider, but if even half of the 2015 enrollees utilize this feature it should improve the payment rate in CA noticeably.

All that being said, the fact remains that I have indeed been overestimating the "But how many have PAID???" figure by about 2%. I hereby concede the point (or two, technically).