VINDICATION!! ProPublica.org confirms 480,000 additional HC.gov QHPs since 4/19!! (UPDATED)

Hat Tip To: 
Charles Ornstein

Yes, I'm quoted in this breaking story; Mr. Ornstein contacted me earlier today to ask my thoughts on his scoop, and I agreed to keep mum until it went public:

For months, journalists and politicians fixated on the number of people signing up for health insurance through the federal exchange created as part of the Affordable Care Act. It turned out that more than 5 million people signed up using Healthcare.gov by April 19, the end of the open-enrollment period.

But perhaps more surprising is that, according to federal data released Wednesday to ProPublica, there have been nearly 1 million transactions on the exchange since then. People are allowed to sign up and switch plans after certain life events, such as job changes, moves, the birth of a baby, marriages and divorces.

Now, before everyone jumps on the "1 million" figure, a few important things:

  • First: The actual number, from 4/20 - 7/15, is exactly 960,649 transactions
  • Second: "Transactions" is very important; that doesn't necessarily mean people enrolling, since someone changing their policy (for instance, moving from Michigan to Illinois and switching from BCBSM to BCBSI or whatever) counts as two transactions. An "insurance industry official" estimates that perhaps half of the 960K are actual new enrollees; the other half would be split (240K dropping the old policy, 240K replacing with the new policy), for a net gain of around 480K.
  • Third: Remember, this is only for the federal exchange; it does not include the 15 state-run exchanges (that's right...assuming 87% of them receive subsidies, that's an additional 420,000 people who will be screwed over by Halbig if the ruling stands)

So, how does this line up with my own estimates? Well, I've been assuming gross additions of around 9,000 people per day nationally. During the open enrollment period, the Federal exchange made up an overall average of around 68% of the total...but that proportion increased every month, as HC.gov started working properly.

In October, HC.gov made up only 25% of the total enrollees; in November, 43%; 59% in December, 65% in January and 72% in February. I'm pretty sure this stablized in March and April (I'd have to do some digging to be sure) and since then, probably at around 75% of the total. That would mean around 6,700 people per day (Ornstein quotes me as saying "between 6K - 7K").

An insurance industry official estimated that less than half of the transactions are new enrollments. The rest are changes: When an existing member makes a change to his or her policy, two 834s are created — one terminating the old plan and one opening the new one.

Charles Gaba, who runs the website acasignups.net that tracks enrollment numbers, estimates that between 6,000 and 7,000 people have signed up for coverage each day on the federal exchange after the official enrollment period ended. Gaba's predictions were remarkably accurate during the open enrollment period.

"That doesn't account for attrition. That doesn't mean that they paid," Gaba said. "That's been based on limited data from a half dozen of the smaller exchanges, extrapolated out nationally."

Now, if his "insurance industry official" is correct, it's somewhat lower (5,500/day x 87 days = around 480,000). If my estimate is closer, that'd be around 6,700 x 87 = around 580,000 people.

As long as the actual number is somewhere in that ballpark, I'd say I've been pretty damned well vindicated.

Of course, the HHS Dept. could save everyone an awful lot of these estimates if they'd just release the actual reports like they did during open enrollment. I still haven't heard a good reason why they won't do so.

UPDATE: I can't believe I forgot about a 4th major factor: As noted by Ornstein, "each report can cover more than one person in the same family"...so we're not just talking about individuals, we're talking about households (or at least partial households). According to the U.S. Census Bureau, the average household size is around 2.6 people, although I've typically used a more conservative estimate of just 1.8 people per household. My guess is that it's even lower than that in this case--perhaps no more than 1.5x, since even if 3-4 people live in a house, not all of them will necessarily need to be added to the policy; perhaps both parents are covered already and their kid turns 19. Perhaps a woman in the household gives birth, and so on.

Even assuming a 1.5x factor, that opens up the possibility that both Mr. Ornstein's industry insider estimate and mine could be correct: It could indeed be "less than half" which represent new additions to the QHP total (5,500 per day or less), but those transactions could total 6,700 people per day...or potentially even more (1.5 x 5,500 = 8,250).

Again, it would save Mr. Ornstein, myself and others in the healthcare journalism field a lot of trouble and guesswork if the HHS Dept. would just released the damned data. Yeesh.