In which I make my first major QHP enrollment projection mistake.

Until a few days ago, my predictions about how many people would enroll in Qualified Health Plans via the Affordable Care Act exchanges were, without being boastful, dead-on target.

• I called the New Year's Eve total with 98.2% accuracy and Medicaid/CHIP determinations through 12/31 with 99.7% accuracy.
• I was off by less than 0.02% at the end of January, and by less than 1% through the end of February.
• I projected the 3/31 total to be 7.08 million; the final tally was "around" 7.1 million, and my call of 7.78 million as of 4/15 was off by somewhere between 0 - 2.75% (I'll never know for certain because the only official number given (8.02 million) was as of 4 days later).

However, I'm obviously not perfect at this, and if I misjudge a significant factor, my projections will be off accordingly.

For the off-season enrollments, until a few days ago, I was comfortable projecting a daily average of 7,000 per day...then 8,000, then finally 9,000 or more per day, as the limited off-season state-level data came in from Oregon, Minnesota, Hawaii and a half-dozen other states. I was even encouraged by the latest updates out of 2 Federally-run state exchanges (New Mexico and West Virginia) to bump things up even more, to 9,500 per day.

However, as I explained in my Oregon post yesterday, it looks like I've been seriously underestimating how much the COBRA enrollment extension, which ended on July 1st for most states (and yesterday, July 15, for California) was shoring up those numbers throughout May and June.

Don't get me wrong; I knew the numbers were likely to drop back somewhat, but I figured the COBRA factor was only perhaps 20% of the totals. However, judging from the latest post-July 1st updates out of Hawaii, Minnesota and Oregon, it now looks like COBRA was more like 90% or higher.

The change in all three states is dramatic:

• From 4/20 - 6/29, Minnesota was averaging about 53 additional QHPs per day. The past two weeks? 2 per day.

• From 4/20 - 7/07, Oregon's total QHPs were going up 325 per day. Over the past week that's been cut by more than half, to 156 per day...and the net number, which had been rising by 240 per day through July 7th went up by a mere 10 over the past week.

• From 4/20 - 6/28, Hawaii was averaging 14 per day. In the first week of July the total dropped by 156 people (presumably due to them purging unpaid enrollments, but the bottom line is that for whatever reason, the subtractions outnumbered any additions).

In short, as I stated yesterday:

If that's the case, then I've seriously overestimated the number of additional QHPs which will be enrolled in between July and November.

The numbers should be propped up a bit through today (7/15) because of California's 2-week extra extension, but after that, if I'm correct, the numbers will likely crater to perhaps 1-2% of the Open Enrollment rate instead of the 20-25% that they've been at so far.

If that happens, then the daily rate nationally may drop from around 9-10K/day down to under 1,000/day...or only a couple hundred thousand on top of the 8.7 million or so I have estimated as of early July. That would top things out at no more than 9 million (total) as of 11/15, or 8.1 million paid QHPs.

In other words:

• 10/1/13 - 4/19/14: 8.02 million (7.2 million 1st month paid)
• 4/20/14 - 7/15/14: appx. 750,000 (8.77 million total; 7.9 million 1st month paid)
• 7/16/14 - 11/15/14: 1K/day x 4 months = 120K (9.0 million total; 8.1 million 1st month paid)

I could be wrong, and if the numbers from other states stay as high as they've been (or if OR, MN and HI reverse this trend and start climbing back up again), I'll be more than happy to issue an Emily Litella "Never Mind!" update. Alternately, the national drop might not end up being nearly as dramatic as it has been for these three states in particular. However, I have to go with what the data is telling me at the time, and for the moment it's saying that the end of the COBRA extension also means the end of significant QHP enrollments during the off-season.

UPDATE: Minnesota has been doing a fantastic data-posting job in the off-season period, updating their numbers almost daily...and in fact they just updated their data again this morning:

• July 13: 52,264 QHPs
• July 14: 52,265 QHPs

Yup, that's right...they added exactly 1 more person, which means they've only added 32 between 7/1 - 7/14.

 I should note that there never were any off-season period expectations made by the CBO, HHS, CMS or the White House anyway, so this doesn't reflect negatively on them; this error is mine and mine alone, so I'm owning it.

UPDATE x2: It's been suggested that the 4th of July holiday weekend might be part of the issue. I suppose that's possible, but it seems unlikely that this would account for more than a small portion of the drop.

It's also important to remember that this is based on only 3 states, which collectively only include 10.5 million people, or less than 3.5% of the total population, so I can't tell how representative they are. This is the main reason why I'm so frustrated by the HHS Dept. and the New York State of Health exchange for refusing to release their off-season data...and CoveredCA hasn't chimed in yet either (I was told they'd be doing so "soon" but that was 2 weeks ago).