In which Investor's Business Daily completely mangles my data.

Investor's Business Daily published a story today by someone named John Merline which references an Inspector General report stating that the enrollment data from isn't being reconciled properly with the data from the insurance companies on the federal exchange, something which is apparently required by the ACA law itself. Now, if this is true then that's certainly an issue which needs to be resolved. However, my problem with the article comes a bit further down, where my site and data are referenced repeatedly:

The administration hasn't released updated enrollment numbers since May, which covered the entire open enrollment period at the federal exchange. An update would shed light on how many are keeping up with premium payments.

But 15 states have separately reported paid enrollment numbers, and according to data compiled by, paid enrollment is 322,000 fewer than the last official White House count — which means nearly 13% of those counted haven't paid their premiums.

That website estimates that if the payment rate in the remaining states averaged 90%, actual enrollment would be less than 7.2 million, not the 8 million touted by the administration.

OK, this guy is mixing together several numbers from here and misunderstanding (or understanding but misstating) what they mean. Don't get me wrong; I do appreciate being touted as a reliable resource, really...but if you're going to do so, please don't misstate my data.

Yes, I estimate that the paid enrollment figure is indeed around 7.6-7.7 million at this point, which is indeed around 322K (somewhere between 300K - 400K, anyway) less than the last official 8 million figure released, which runs through 4/19. And yes, an updated report would indeed shed light on the updated enrollment situation; I've been upset about this for a month now. And yes, 90% paid out of 8 million total is indeed 7.2 million.

HOWEVER, once again, 8 million people did enroll in QHPs via the ACA exchanges as of April 19, and an additional 600-700K have enrolled since April 19.

Out of that total (around 8.64 million by my count), around 90% of the 8 million have paid by now, plus around 80% or so of the additional 600K - 700K should have by now as well (remember, around 200K of these policies don't start until August 1st). Thus, around 7.6 - 7.7 million people out of the 8.6 - 8.7 million total enrollments have paid their first month's premium, give or take. That's where his "nearly 13% haven't paid" number comes from, although he used it in a confusing passage (I actually have it around 12%, but whatever).

Does that mean the administration is lying about "8 million enrolled"? No, not at all. Again, at the time of the 8 million announcement on May 1st, 33% of those 8 million total (around 2.6 million) policies had just started coverage, as in that same morning...and the report itself had been written up several days earlier. Most of the insurance companies have some sort of grace period for payments to show up, many of their billing systems were horribly backlogged and so on. When the dust settled a few weeks later, it did indeed start to look like around 10% of those hadn't paid their first month's premium.

So, again, yes, about 800K should be lopped off of the original 8M figure...but that doesn't mean that they "lied" about it. If a car company says they sold 100,000 units last month with a "no money down" promotion, and later it turns out that 10% of the customers were deadbeats, had their credit denied, their checks bounced or whatever, that doesn't mean that the car company was guilty of "lying" about how many they sold; at the time, it was accurate information.

In addition, again, with around 8,000 or more* people enrolling each day even in the off season via qualifying life events, the paid number should hit 8 million around the end of July anyway.

Next come two more interesting points: IBD claims that...

Aetna says that out of 720,000 sign-ups, only about 580,000 were paid up by May 20, a payment rate of only 80.6%.

Hmmm...interesting, since according to Bloomberg News, Aetna's official testimony before Congress gave the paid number as in "the low to mid-80 percent range" of May 7th. "Low to mid-80's" would suggest around 83%, or 598K. Not sure how that number dropped by 18,000 over the next 2 weeks. Clerical error on Aetna's part?

More interesting to me is the next passage, about the attrition rate (people stopping coverage in the 2nd month, 3rd month and so on). Until now I've been operating on Chris Conover's claim of a "2-5% attrition rate" based on a March 5th entry by Robert Laszewski which claimed that "Carriers are telling me that another 2% to 5% of those January enrollments never paid their second month's premium."

However, according to this IBD article, Laszewski's ongoing attrition rate range estimate has apparently dropped a bit:

Widely followed industry analyst Robert Laszewski of Health Policy and Strategy Associates, said that "generally, carriers are saying that 85% paid the first month and that enrollment deteriorates after that at a rate of 2% to 3% per month."

I've already shown, repeatedly, that the actual 1st month payment rate is closer to 90% than 85% after the grace periods are over, so I'm not going to rehash that. As for the attrition rate, the other day I posted a chart which assumed Conover's 2-5% range was reasonable, splitting the difference to 3.5%.

However, if the same source (also an ACA critic) now claims that it's only 2-3% per month, fair enough: I'll drop that down to 2.5% on average going forward.

If those estimates turn out to be true, the administration's 8 million would fall to around 6.1 million after six months — or right about where the Congressional Budget Office predicted enrollment would be in ObamaCare's first year.

(sigh) And here we come to the crux of the matter. NO, this is not correct, because IBD has forgotten to add the roughly 1.9 million additional enrollments which will have come in between 4/20 - 11/14. Again, here's how that math works--and again, this assumes that you "can't count" someone who is only enrolled for 10 months, 9 months, 8 months etc, which is insane on the face of it:

  • 8.02 million (10/1/13 - 4/19/14)
  • ADD appx. 1.9 million (4/20/14 - 11/14/14)
  • GROSS total = appx. 9.9 million
  • SUBTRACT appx. 10% for non-payment = appx. 8.91 million paid enrollees (1st month)
  • SUBTRACT 2.5% attrition = appx. 8.69 million (2nd month)
  • SUBTRACT 2.5% attrition = appx. 8.47 million (3rd month)
  • SUBTRACT 2.5% attrition = appx. 8.26 million (4th month)
  • SUBTRACT 2.5% attrition = appx. 8.05 million (5th month)
  • SUBTRACT 2.5% attrition = appx. 7.85 million (6th month)...much higher than the 6.1 million that IBD claims.

*(As an aside: Up until now I've been cautiously using 8,000/day as an off-season estimate, even though my projection chart has it ranging from 9K - 12K/day; for purposes of this entry, I'm using the 9K/day lower-bound number).

Now, it's true that using this calculation, by the time you get to the 10th or 11th month, you're "down" to around 7 million or so. In fact, I'll go you one better: If you actually chart it out and only "count" the people who pay for the full 12-month period of 2014, as argued by Chris Conover and as demonstrated in the chart below, you'll see that the final number as of the end of the year is likely to end up being "only" around 5.8 million:

Some ACA supporters might think that I'm "damaging" my case by "admitting" this, but I have no qualms about it...because it's completely beside the point.

The administration didn't say that 7 million people would be enrolled in time for January 1st, 2014 coverage, and pay for the full 12 month period through December; they said that they hoped that 7 million would have enrolled as of March 31st (later bumped out by another 2 weeks to mid-April). Sure enough, 7.1 million enrolled as of 3/31, of which about 6.4 million have paid. Another 900K had enrolled as of 4/19, of which about 810K have paid...bringing the total paid number up to around 7.2 million, as the article states earlier. Another 600K or so have enrolled since then, of which around 540K will have paid by mid-August or so, and an additional 1.3 million or so will enroll between now and November 15th, of which around 1.17 million will pay, for a total of around 8.9 million paid enrollments.

Whether some of these 8.9 million (around 22% or so, according to Laszewski) don't stick it out for a full 12 month period is a completely separate question as to whether they enrolled in (and paid for) an ACA exchange QHP...which they did/will have done by 11/15.

Maybe they got a job with benefits and didn't need the QHP anymore. Maybe they got married to someone who was already covered. Maybe they moved out of the country. Maybe they turned 65 and moved over to Medicare. Maybe they fell on hard times and moved onto Medicaid. Maybe their spouse or child passed away. Maybe they passed away. Who the hell knows?

The point is that while the attrition rate is important from a budgetary POV (the CBO needs to know this so they have an idea of how much money they'll be doling out in subsidies), from a human perspective, this aspect of the law is working exactly the way it's supposed to: Giving people the peace of mind of decent healthcare coverage for however long they need it before they move onto some other coverage. For some this may only be a month or two; for others it may be years. This is a good thing.