Dear HHS: Your numbers are perfectly fine, so stop jerking me around.

Remember back in September, when HHS Secretary Sylvia Burwell announced that as of August 15th, 7.3 million people were enrolled and paying for private healthcare policies through the ACA exchanges?

Here's what I said at the time:

This is FANTASTIC news.

It means that either...

a) the actual number who've been enrolling daily in the off-season is higher than my 9,000/day estimate, or
b) that the attrition rate after the first month is lower than my 3%/month estimate, or
c) more than 90% of enrollees eventually pay their first month's premium, or
c) some combination of all three.

If my 9K/day additions / 3% per month subtractions / 90% payment estimates were all accurate, the currently enrolled number would only be about 7 million, not 7.3 million.

Now, I think it's unlikely that the "first month paid" percent is any higher than 90%, which means that it's probably a combination of the daily additions being higher than 9K and/or the attrition rate being slightly lower than 3% per month.

At the time, I was very, very happy to have underestimated the mid-August number. The fact that this meant I was off by about 4% was certainly fine with me.

However, it turns out that this wasn't exactly the case. Alex Wayne of Bloomberg News just posted a pretty big scoop which, if accurate, has me pretty upset.

There's actually some good news in it: The actual total exchange enrollment number as of August 15th was actually 7.37 million...some 70,000 higher than announced.

So why am I upset?

BEFORE I CONTINUE, an important caveat: Keep in mind that last April, the GOP House Energy Committee issued a press release regarding payment rates which was so full of crap that it took me two separate entries to document all of the garbage. So, believe me, I'm not posting this entry lightly.

However, the following appears to have been confirmed by Mr. Wayne.

Because it turns out that roughly 400,000 of those policies appear to be standalone DENTAL plans, not healthcare plans (QHPs).

Now, there's absolutely nothing wrong with 400K people getting dental coverage. That's a good thing.

And there's nothing to be embarrassed/ashamed about if the August QHP number was "only" 7.0 million instead of 7.3 million. I said that 7.3M in August would be fantastic...but 7.0 million is still pretty damned good.

However, assuming this claim is accurate, the HHS Dept. absolutely owes me and everyone else an explanation about why they would mislead about something there was no need--even from a political spin POV--to mislead about.

7.0 million QHP enrollments as of mid-August would have been a perfectly reasonable attrition rate. Furthermore, there's nothing necessarily wrong with people dropping their policy, depending on their reasons for doing so.

And yet, (again, assuming the story is true), HHS chose to frame the number in such a way that make everyone--myself included--think that the "7.3 million" number referred to Qualified Healthcare Plans alone.

As Wayne notes, here's the exact wording of Burwell's 9/18 statement (emphasis mine):

“As of August 15, this year, we have 7.3 million Americans enrolled in Health Insurance Marketplace coverage and these are individuals who paid their premiums. We are encouraged by the number of consumers who paid their premiums and continue to enroll in the Marketplace coverage everyday through special enrollment periods.

And again, here's the official press release (again, emphasis mine):

As of August 15, 2014, 7.3 million Americans were enrolled in Marketplace coverage and had paid their premiums. This number represents a snapshot of a point in time, not the cumulative enrollment data from October 2013 through August 2014. For example, it does not include individuals who:

• Enrolled in Marketplace plans and paid their premiums but discontinued coverage before August 15, 2014.
• Signed up for Marketplace coverage but did not pay their premiums by August 2014.
• Have signed up for Marketplace coverage and paid their premiums after August 15, 2014.

Note the wording there. "Marketplace coverage". That doesn't necessarily mean Qualified Healthcare Plans, and is therefore not a false statement...but it sure as hell makes it sound like it is.

Now, compare this with the official enrollment report from May 1st, 2014 (which covers all enrollees from 10/1/13 - 4/19/14), which specifically separated out the 1.1 million standalone dental plans (as opposed to QHPs which also included dental), and made it clear that they were separate from the 8,019,763 QHP total.

So, what does this mean? There's three ways of looking at it: Personal, Policy (that is, what does it mean about the ACA's effectiveness) and Political.

• On a PERSONAL level, it means that I was correct about both the August and October numbers in the first place.

Here's the attrition estimate chart that I had been going with prior to the 9/18 announcement, which assumed a 90% payment rate, 9,000 people enrolling per day and about 3% dropping coverage per month:

I've already noted several times that my 90% payment rate was a bit high--it turns out it's more like 88% nationally. This simply means, however, that the number being added during the off-season was a bit higher, perhaps around 10,000/day (my off-season estimate range was between 7,500 - 10,000/day). Here's what the same chart looks like when you make those 2 adjustments:

As you can see, either way, the August number comes in at around 6.9 - 7.0 million.

When Burwell announced "7.3 million on August 15th", I assumed that my assumption of a 3% attrition rate was too high; dropping that down to 2% gave roughly 7.3 million as of August/September. Now it appears that I actually was right about the 3% rate.

This also likely means that the 7.1 million October enrollment figure is off by several hundred thousand as well. Here's the press release from November 10th:

Since October 2013, individuals have been able to shop for and buy insurance in the Marketplace. By the end of the first open enrollment period, over 8 million people selected Qualified Health Plans (QHP) through the Marketplace. As of October 2014, 7.1 million people were enrolled and paying for health coverage (“effectuated” enrollees) through the Marketplace.

Again, note the distinction: 8M QHPs as of April 19th, 7.1M health coverage as of October.

How many of the October 7.1 million number are standalone dental plans? I don't know, but let's assume that the ratios stayed the same from August to October (400K/7.4M = about 5% of the total). That should mean that the October QHP number was actually down to around 6.75 million or so.

Well, guess what? Here's my post from 4 days earlier:

In any event, there you have it. IF the other 2/3 of the states follow similar retention patterns to the 1/3 I have data for, the current number enrolled in exchange QHPs as of this writing should be somewhere between 6.8 - 7.0 million people.

So...yay, me. (sarcastic golf clap with a tinge of bitterness).

• On a POLICY EFFECTIVENESS level, this means...not much at all.

As I've noted many times before, "churn" in the healthcare marketplace is a normal and common thing. People drop their policies for all sorts of reasons--they turn 65 and move to Medicare; they get hired by a company providing ESI; they fall on hard times and move to Medicaid; in some cases, they die, of course. Without knowing the reason why these folks dropped their policies, this doesn't mean much. In addition, you're still talking about upwards of 28 million people having solid, comprehensive health coverage, 11 million or more of whom didn't have it before.

That's a record to be proud of even if that number doesn't go up one person more (and believe me, it will, as long as the Supreme Court doesn't torpedo the law next summer).

• From a POLITICAL point of view, however, this is probably pretty bad.

The Obama administration has already been taking a lot of heat from not just opponents but supporters as well for being somewhat squirrelly/cagey about releasing ACA enrollment data.

Hell, just yesterday the Washington Post had a story about the dearth of enrollment data so far this time around.

As I said in my Open Letter to HHS back in May, when they announced that they were discontinuing even providing monthly reports during the off-season:

In addition, aside from the "can't/won't keep paying" crowd, none of those other reasons for leaving the exchange QHPs reflects negatively on the ACA in any way! If someone gains a job with benefits, good for them! Their exchange QHP tided them over for a few months while they improved their lot. If they age into Medicare, again, that's two different "Big Government" programs working exactly the way they're supposed to. The same holds true if their situation goes negative and they find themselves on Medicaid. And if they do happen to pass away, at least their ACA QHP gave them a fighting chance at medical care that they otherwise might not have had.

In other words, even if there is a significant amount of attrition, the administration still has plenty to be proud of with the first year's enrollment numbers. I find it difficult to believe that this is enough to cause HHS to scrap the reports.

The Big Magic Number which everyone was obsessing over last year was 7 million. It doesn't really matter that "7 million QHP enrollments the first year" was never necessary for the law to succeed (it was simply the CBO's projection of how many they expected to do so in 2014); thanks to Burwell's predecessor Kathleen Sebelius' foolish response in an interview last fall, that was the number which stuck, for good or for bad.

Reaching 8 million total, 7.1 million of whom paid and 7.0 million still enrolled 4 months later is perfectly fine from a PR point of view...or it would have been.

So now, you've taken data you should be proud of and turned it into yet another attack point for the GOP, right when the ACA is still on shaky ground.

Oh, and you did one other thing in the process:

You made me--someone who's specifically focused on aggregating accurate enrollment data--look like either a chump or a shill.

I started this project as a lark, but over the past year I've developed a bit of a following, and people have come to rely on me for reliable, accurate data when it comes to ACA enrollments.

I posted the following in the F.A.Q. for this site nearly a year ago (and yes, I know it's horribly outdated by now...):

Does your data have a liberal bias, then?
Well, first of all, the ACA--or at least, the Private Exchange Enrollments portion of it--is hardly a "liberal" law. It requires millions of people to agree to pay private, for-profit corporations potentially billions of dollars each year, which isn't exactly part of the "progressive" political philosophy.

Having said that, my data comes from a variety of sources, including the HHS, the State-Run Exchange press releases and news outlets ranging from small local newspapers to major national news outlets. I do not guarantee that any of their data is accurate, but I do guarantee that any of their data that I enter into the spreadsheet is as accurate as I can make it given my limited time and resources. Occasionally I may make an honest mistake; when this happens, I attempt to correct it as quickly as possible and will also post an explanation of what happened in the blog as appropriate (see The Mystery of the Vanishing Minnesota Data for a case study).

I'm going to continue updating the site because I believe that the Affordable Care Act, with all of its flaws and shortcomings, is still an important and substantial improvement over the system that we had before it.

I just wish that you guys had enough confidence in your own policies not to embellish the metrics...especially when they stand on their own just fine.