In which Megan McArdle cites me as a major source but ignores what I actually say.

I haven't posted much the past few days, partly because of a lack of major ACA news, partly because I'm busily playing catch-up with my day job (I do have one, believe it or not), and partly because all the oxygen has been (rightly) sucked out of the news atmosphere the past week or so by the slow-motion trainwreck called "Ferguson, Missouri".

So, I was a bit surprised to see my name and this site figure heavily as a source by Megan McArdle over at BloombergView, in an article called "More Bad News for Obamacare". McArdle has decided to pick up the "Massive Attrition!!" ball which I debunked last week and run with it, citing me as a source several times (including using me as her source for the news of the HHS Dept. dropping off-season enrollment reports)...and yet completely ignoring my entire point:

on net, they expect enrollment to shrink from their March numbers by a substantial amount -- as much as 30 percent at Aetna Inc., for example.

Um, no. Aetna is not reporting a 30% "shrinkage" from their March numbers; they're reporting a 15% gradual attrition rate from their paid enrollee numbers, spread out over a 7 month period.

The former makes it sound like 30% of the enrollees were somehow "made up" when in fact half of them just didn't pay up (yes, this is important financially, but it's completely different from deliberately "cooking the books" which seems to be the implication here) while the other half are gradually dropping their policies for a variety of completely legitimate reasons (aging into Medicare, getting a job, etc).

The very same article notes:

Cigna (CI) said that it expects its individual market customers, including more than 100,000 in the exchanges, to "move from 300,000 down to 280,000 in that range," Cigna CEO David Cordani said in a conference call.

Hmm...they don't specify what portion of that 20K loss they expect to come from the 33% in the exchanges, but assuming it's proportional, that's a loss of around 7,000 out of 100,000 by the end of the year...or just 7% attrition, or just over 1% per month. Huh.

McArdle doesn't mention this data point, nor the 3rd one out of Washington State which comes in at around 2.5% per month.

How much does this matter? As Charles Gaba notes, this was not unexpected: Back in January, industry expert Bob Laszewski predicted an attrition rate of 10 to 20 percent, which seems roughly in line with what IBD is reporting. However, Gaba seems to imply that this makes the IBD report old news, barely worth talking about, and I think that’s wrong, for multiple reasons.

Interestingly, even though she mentions me by name twice, she fails to actually link to the very article she's citing...which is understandable considering that a) it debunks the Aetna claim and b) nowhere did I say any such thing about it being "barely worth talking about". Yes, it's worth talking about...which is why I've devoted not one, not two, but around a dozen entries about the subject. My point about it not being "new" was more of a point of personal pride than anything; the IBD article (which, amusingly, also quotes me as their source for the Washington State data...which McArdle then fails to mention) made it sound like no one had been discussing the attrition issue when in fact Chris Conover of Forbes and I have been debating/discussing it for several months now.

It’s always valuable to have actual data rather than guesstimates (and we should remember that we’re still getting data; we won’t know the final attrition rate until December).

I agree 100% with her on this point.

In fairness, the Barack Obama administration conveniently stopped issuing enrollment reports after that March peak, so there isn't a better hard number to use. But if enrollment is indeed declining, then we should include that caveat when we discuss the figures we have.

Here's where she does link directly to my earlier piece in which I read the riot act to HHS over dropping the enrollment reports. However, she's wrong about something else here: I do believe that the overall enrollment number is indeed declining...just not nearly as much as the IBD report made it sound like. To the best of my calculations, it's increasing by around 3% per month and dropping by around 4-5% per month (non-payments + dropped policies) simultaneously...resulting in a net loss of around 10% by the time mid-November rolls around, or around 7.3 million enrolled in exchange-based QHPs by that point.

This is something which I do indeed "include" when I "discuss the figures I have".

UPDATE: Link fixed.