In which I apologize to Mr. Conover: Yes, if you count fractions of people as human beings, he was absolutely correct.
2018 MIDTERM ELECTION
Time: D H M S
Chris Conover has responded to both of my items from yesterday. Let's take a look:
Let’s start with this howler: “Actually, the Obama administration never said that they’d reach 7 million paying customers by March 31st.”Seriously? Has Mr. Gaba really not read the September 5, 2013 memo from Marilyn Tavenner (chief of Centers for Medicare and Medicaid Services) titled Projected Monthly Enrollment Targets for Health Insurance Marketplaces in 2014? Table 2b of that memo clearly shows an enrollment target of 7,066,000 for March 31, 2014. End of story.
Yes, I'm well aware of that memo, but that's not exactly the end of the story. Not only do I agree that this memo indeed had a 7M target by 3/31, but I'll go Conover one further: It was the same memo in which the CMS Dept. gave rather outlandish exchange QHP "goals" for all 50 states which made little sense--such as thinking that Kentucky (which only had 249,000 uninsured eligible for QHP assistance last fall) would actually enroll more people in private plans (220,000) than New York would (218,000), when NY had 995,000 uninsured...4x as many...eligible for QHP assistance.
This was the same memo which somehow claimed that both Utah and Vermont, which have wildly different demographics, would have not only identical total QHP enrollments, but identical incremental enrollments for every single month along the way...almost certainly due to someone copying & pasting one state's data into the others on the spreadsheet. In addition, this particular CMS memo admits to taking the 7M calendar year 2014 figure issued by the CBO back in May and attempting to shoehorn the state-level projections in so that they would add up to 7 million (I have no clue why they chose to slap on an extra 66,000).
In short, in my opinion this particular CMS memo was kind of poorly put together all around, which I've noted several times in the past...and which might explain why--unless I'm mistaken--it was never intended for public release (it was specifically addressed to Sec. Sebelius and actually stated "This memo is for your information only; you do not need to take any action on this."). I'm sure there are hundreds of internal memos flitting about within the various departments of the Federal Government which included different projections, estimates, goals and so on, which change from time to time. Perhaps there were 3 other versions of this memo a week or two earlier or later which circulated internally but never made it out to the public as varying factors were taken into account and discussed. I presume that only the ones which have been vetted are issued as press releases to the public. In this case, the GOP House Ways & Means Committee got hold of this memo and published it, and that's where the state-level "goals" by March 31st came from.
So, why did "7 million by March 31st" sink into the public consciousness as an end-all, be-all goal or target? Well, mainly because Kathleen Sebelius took the CBO's projection of 7 million in calendar year 2014 and (foolishly, in my opinion) gave the following interview with this exchange:
- Reporter: "What does success look like?"
- Sebelius: "Well, I think success looks like at least 7 million people having signed up by the end of March, 2014"
If you asked me, "Would winning $100 million in the lottery make you rich?", I'd answer "Yes, absolutely." If you asked me, "Would winning $10 million in the lottery make you rich?", I'd answer yes as well. She didn't say that not hitting 7 million by 3/31 would be a failure; she said that hitting 7 million would sure as hell be considered a success.
Now, Sebelius did make a serious wording flub here, because she certainly gave the impression that anything short of 7 million would be disappointing. However, she also referred to 7 million "signing up" which (as anyone who's got into an earlier debate about this knows) is not the same thing as "enrolling", which in turn is not the same thing as a paid enrollment. Christ, you couldn't throw a rock without hitting some Republican yammering about "signing up isn't actually enrolling!!" every 30 seconds.
You can't have it both ways, guys: Either Sebelius' "7 million signing up" was an intractible vow of the 3/31 enrollment number (in which case you're admitting that the two are equivalent), or one does not equal the other, in which case you have to admit that she only stated that 7 million signups was the benchmark for "success" (even though, again, she never said that a smaller number wouldn't also equal success).
Is this petty semantic nit-picking? Absolutely...which is exactly what Conover is doing in the first place.
I'll come back to this word parsing in a little bit. However, I first have to admit an interpretation error on my part:
I correctly noted that the May 2013 CBO report--the one from whence came the "magic" 7 million figure which ended up taking on some sort of mystical quality, even though, as Ezra Klein noted last fall, there's no special power in that particular total number--made no mention of the 3/31 date; it was referring to the entire 2014 calendar year.
However, I incorrectly forgot about an important footnote at the bottom which the Washington Post Fact Checker's Glenn Kessler noted back in April...and which I've actually already written about myself:
Footnote b of the very CBO table Mr. Gaba links to on his website to make his case clearly states: “ Figures reflect average enrollment over the course of a year.”
When I wrote about this back on April 4th, I did a partial back-of-the-envelope analysis of what this might mean. However, now that we have full enrollment data through 4/19, and partial data since then, and we have Mr. Conover's "2-5% attrition" claim which I've already conceded doesn't seem unreasonable, I'm able to plug the numbers into an actual spreadsheet and come up with the following:
So, credit is due to Mr. Conover: If you specifically use the average of the monthly paid QHPs over the full 12-month period of 2014, and if you use this as the specific "expectation" number that you're defining as the only one which counts as a valid one to compare against, and if the 1st month payment rate is 90%, and if the monthly attrition rate is 3.5% each and every month as Mr. Conover claims, then Mazel Tov! A cookie for Mr. Conover! Hell, my chart above even goes beyond Conover's math at the low end (he noted that the "rosiest" scenario still comes up 4% short).
So, does this mean I'm stating that only 5.4 - 5.7 million people enrolled? Of course not. The 8.6 million (of which about 7.6 million have/will have paid their first month's premium) who have enrolled to date, and the estimated 900K - 1.6 million more who will do so over the next 4 months or so are not fractions of people.
This is the thing that ACA opponents either can't get through their heads or aren't willing to admit. If someone enrolls in March, has dependable healthcare coverage for, say, 3-4 months and then moves on to other coverage by getting a job with benefits, aging into Medicare, marrying someone with coverage or so on, that still counts as them having the peace of mind of health insurance. Perhaps they came down with an illness or were injured during those few months; perhaps they didn't. That's the definition of insurance--having coverage in case you need it. These people were covered in the event something unpleasant and expensive happened to them during a certain period of time, and then moved on to something else. This is a good thing.
Now, the CBO does need to know what the monthly average number is for strict budgetary reasons: They need to have an idea of how much the Federal government will be doling out in tax subsidies, so yes, that's important from their perspective. From their perspective, "1/12th of a person" can indeed "exist" on a spreadsheet.
However, from a human perspective, where 1/12th of a person doesn't exist (any more than, say, 3/5ths of a person existed anywhere except the Article I, Section 2 of the original Constitution prior to the 14th Amendment), this is silly and insulting. Someone enrolling in June because they just lost their job, or a woman who enrolled their newborn baby on April 5th because they didn't happen to give birth a week earlier (and therefore "missed" the original, arbitrary 3/31 cut-off) still counts as a person who enrolled this year.
Now, let's move on to Conover's error. In the next paragraph, he states:
Mr. Gaba seems desperate too change the goalposts, perhaps because his own numbers clearly show that as of March 31, paid enrollments on the Exchanges were only 6,372,147 (last row of Week 26 spreadsheet). This simply confirms what I stated in my post. Mr. Krugman was dead wrong to claim signups exceeded expectations. Why is Mr. Gaba disputing this simple and obvious point?
First, let me applaud Conover for actually digging through the earlier worksheets in my spreadsheet; he seems to be one of the very few people who actually bother doing so (most seem content with either the Graph or the Blog sections of the site, but the Spreadsheet is the most important feature, since that's where everything else comes together).
However, using Conover (and Cannon's) standard, you could have theoretically had 30 million people enroll at 11:55pm on March 31st, and even if 100% of them put a check in the mail the moment they completed the process, not one of them would "count" as having "enrolled by 3/31" since their payment checks (or even their automatic credit card transfer) wouldn't have cleared the bank until after midnight. By that logic, Conover/Cannon would claim that not a single person enrolled in an exchange QHP by the administration's deadline, and therefore none of them count.
Mr. Conover claiming that I'm the one "moving the goalposts" is the height of hypocrisy. He seems to think that Krugman's "exceeded expectations" refers only to the unofficial 7 million estimate/projection figure, only as of 3/31/14, and only those whose payments had fully cleared the bank.
Now, perhaps that's what Krugman meant, but the "expectations" that I'm referring to can be found in Jonathan Cohn's fantastic roundup of the endless litany of doom & gloom predictions by everyone from John Boehner and Orrin Hatch to the Daily Caller and FOX News, in which they were mocking the idea that HC.gov would be able to enroll 6 million, 5 million, 4 million, or even enough people to fill a football stadium during the early, "broken website" days of October and November.
Oh, one other thing. Conover then says:
First, 10 million is not paid enrollments: it’s total enrollments. In his eagerness to prove me wrong, he’s forgotten to deduct the 10% (at minimum) unpaid enrollments. So if he truly believes there will be 10 million cumulative enrollments, the correct figure for paid enrollments is only 9.1 million.
On this, Conover is flat-out wrong. I clearly stated the monthly totals as follows (emphasis added to the 2nd line)...note that he even overstates the "paid enrollment" number by 100K (90% of 10 million is 9 million, not 9.1 million):
- 10.0 Million enrolled
- 9.00 Million paid 1st month's premium (90%)
- 8.69 Million paid 2nd month (96.5% of 9M)
- 8.38 Million paid 3rd month (96.5% of 8.69M)
- 8.09 Million paid 4th month (etc...)
- 7.80 Million paid 5th month
- 7.53 Million paid 6th month
- 7.27 Million paid 7th month
- 7.01 Million paid 8th month
However, this is just a quibble. The reality is, the end result of all of this he said/she said gobbldygook is that we now have a wide range of exchange QHP numbers to "choose" from, ranging from a low of around 5.4 million to a high of around 9 million, all depending on your perspective and what political point you're trying to score.
NONE of this back-and-forth silliness changes the fact that by the end of the year, at least 7.6 million, and likely another million or so, will have enrolled and paid their first month's premiums on one of the federal or state exchanges...plus another 7-10 million or so who will be newly enrolled in Medicaid or CHIP, plus a few million more 19-25 year olds who will be covered by their parents' policies due to the law, plus millions more who can no longer be dropped or denied coverage due to a pre-existing condition or be forced into bankruptcy due to endless medical expenses.
Mr. Conover may or may not update his post yet again, depending on how long he wants to keep this pissing match going. If not, fine. If so, unless he says something truly insane, I'll let it go here. I already spent half my birthday writing my original rebuttals (I turned 44 on Saturday); there's no purpose to be served by lobbing this level of minutiae back & forth endlessly other than to muddy the waters so badly that the average Joe ends up hopelessly confused. There's been enough of that by the likes of FOX News and the Koch Brothers already.
Still, congratulations again to Mr. Conover on the "2014 monthly average" issue; you win on that point. Well played.
Oh, one more thing: Mr. Conover included a typo from my posts to which he added a [sic] to indicate that the typo was my error, not his. Fair enough (the typo has since been corrected). However, this does raise one more point: Mr. Conover is being paid by Forbes and presumably has a professional copyeditor on the payroll as well. I'm not on anyone's payroll to run this site, and while there is a copyeditor who generously donates her assistance to me, she can only do so sporadically. So, yes, I may have the occasional typo or grammatical error from time to time.