UPDATED: Sharyl Attkisson botches her math again but still may be onto a real story

Long-time readers may recall that last June, I tore Sharyl Attkisson to shreds for a rather idiotic story she wrote over at The Daily Signal (a front for the right-wing Heritage Foundation), in which she used some absurd math to try and claim that the ACA had only increased the number of Americans with health insurance by 3.4 million, when it was actually closer to 11-12 million people at the time (now up to around 14-15 million).

A few months later, I again laid into her for a rather lame "Obamacare Raised My Rates 36x!!" tweet that she sent out which, while not actually proven false, certainly didn't include any supporting evidence, details or context whatsoever.

Along the way, in an unrelated incident, Ms. Attkisson also tried to blame a stuck delete key on the Obama administration spying on her computer because Benghazi!!!

Since then, she's become sort of my poster child for "pulling numbers out of their ass" anti-ACA propaganda stories.

Today, however, she may be onto a legitimate story regarding Covered California...even thought she once again completely botches some basic math.

California’s health insurance exchange, established under the Affordable Care Act, has been held out as a national model for Obamacare. In some ways—not all of them good—it is. Whether it’s falling far short of 2015 enrollment goals or sending out 100,000 inaccurate tax forms, Covered California is struggling with its share of challenges.

Now, several senior-level officials integral to the launch of Covered California—who enthusiastically support the Affordable Care Act—are speaking about what they view as gross incompetence and mismanagement involving some of the $1 billion federal tax dollars poured into the state effort.

The first portion of the story centers around a guy named Aiden Hill who was apparently a project manager for CoveredCA's call center, but who left after about 5 months and turned whistleblower. Assuming his story checks out, fair enough.

Attkisson's math problem comes in the next section, regarding the enrollment numbers for 2015 vs. 2014:

As recently as last fall, the official says, California hoped to increase enrollment by 500,000 this year. But only an additional 7,098 have “selected a plan” for 2015.

“Their total enrollment is a step in the right direction but nowhere near what anyone thought it would be for the largest state in the country.”

Covered California would not answer our questions about enrollment figures.

Another telling statistic is Covered California’s poor retention rate. Even though people are required by law to have health insurance, only 65 percent of Covered California’s 2014 customers reenrolled in 2015. The rest dropped off.

OK, hold the phone. Yes, they did hope to increase enrollment by 500K: They had about 1.2 million enrolled towards the end of 2014 and hoped to add another 500K to the tally.

Instead, they ended up re-enrolling/renewing about 944,000 people and adding about 495,000.

In other words, they hit their target for new enrollees, but did completely drop the ball on the retention side. (944K vs. 1.2M). Attkisson screws up the numbers and then makes it sound as though they came up short on both the new enrollee and the retention side, when it's only the latter that was an issue.

In addition, she misstates the retention rate: 944,000 / 1.405 million = 67%, not 65%. Again, that's not a huge distinction, but it's not a rounding error either; that's 28,000 people.

Don't get me wrong: The retention issue is significant, but she still confused the issue tremendously here.

Of course, it's certainly possible that Mr. Hill's accusations are simply false; no other official attacking CoveredCA/supporitng Mr. Hill's claims is named. This may be a "disgrunted former employee" making stuff up, and lord knows, Ms. Attkisson doesn't exactly have the best track record here.

HOWEVER, having said that, there's a couple of reasons why I, of all people, am at taking this story at least seriously enough to approach it with an open mind:

• As I complained about last October, Peter Lee claimed, not once, not twice, but on five separate occasions last spring that a flat 85% of CoveredCA enrollees had indeed paid their first monthly premium. Not around 85%, not "in the mid-80's" but 85% specifically. Then, in October, he stated that oops, it turns out it was only 81.7%, not 85%.

Now, that may not seem like a big difference, but for the largest state in the country, that 3.3% discrepancy adds up to over 46,000 people; a lot more than a rounding error.

• Furthermore, in the same October press release they stated that they had around 1.12 million people still enrolled in QHPs as of mid-October, but somehow expected this to increase by another 180,000 or so (to 1.3 million) by the time the 2015 enrollment period started up. A few other states had managed to increase enrollments over the summer/fall instead of decreasing, so it was possible...but increasing their numbers, by that much, over just a 1 month period seemed, shall we say, unlikely.

• Finally, it's not like some of the allegations made are particularly shocking or unheard of. The debacles in Oregon and Massachusetts last year (as well as the early problems at HealthCare.gov, of course) were filled with similar claims, many of which were later proven fairly accurate.

Anyway, the story goes on to make accusations of gross mismanagement and sloppy and/or misstated numbers all over the place, along with a supposed "coverup" of the mess throughout the 2014 enrollment period. It doesn't indicate whether any of this alleged malfeasance continued on into the 2015 enrollment period.

The irony here is that it's possible that, assuming  they didn't screw up the numbers for 2015, it's actually conceivable that their retention/renewal rate is actually much better than they claim--because the 2014 numbers may have been lower to begin with.

That is, if the real 2014 number of effectuated QHPs as of last fall was only, say, 1.0 million instead of 1.12 million, that would mean 944K = a 94.4% renewal rate instead of 84.3%. That wouldn't make their 2014 reporting any less out of line, but it would help explain why they appeared to only barely squeak above the 2014 numbers this year.

Supposedly there's a Part 2 on the way, and it'll be interesting to see whether Attkisson comes up with any other legitimate criticism of CoveredCA. I just hope she pays more attention to her math and less to her delete key this time.

UPDATE: OK, she posted Part 2 today. There's actually a lot less to analyze here, because most of the 2nd half isn't really data-specific; it's mostly general accusations of "secrecy" and so forth. It even has some surprisingly positive stories about successes via CoveredCA.

The only section which is in my bailiwick is this line, again, about attrition:

However, like a lot of Covered California’s statements, it’s a good dose of spin. In fact, 35% of Covered California’s 2014 customers dropped their Covered California policies in 2015. That’s one of the worst retention rates among states.

Again, that's a bit misleading. First of all, 944,000 / 1.405 million = 67%, not 65%; and again, some of this gets into semantics--since only 1.14 million of those 1.405 million actually paid their first premium, it's really more like an 82% retention rate (it's just that their payment rate was also worse than expected).

Granted, that doesn't make CoveredCA's first year look much better, but it's worth keeping these things straight :)