The Mystery of the Vanishing Minnesota Data


Last night I teased that I'd be providing some context and insight into the Iowa and Nebraska updates posted on the spreadsheet, and I still plan on doing so later today. However, there's been an interesting development over at Minnesota which I need to address first.

Yesterday, Jackie Crosby of the Minnesota StarTribune posted "MNsure grids for late crush of enrollees", which among other things gave an update on the MNsure enrollment numbers, stating:

More than 53,000 Minnesotans have bought coverage as of last Friday, with the biggest gains coming from those buying private health plans on the individual market. By Friday, 19,420 had signed up for individual or family plans; 12,051 had been enrolled in Medical Assistance; and 12,708 were getting coverage through MinnesotaCare.

So far, fair enough. However, in the original version of the article posted yesterday, there were two more paragraphs immediately following this. The first paragraph specified that these enrollments actually represented about 125,000 actual people (i.e., additional family members, etc). It also specified that about 20,000 of those were being held up due to paperwork issues and so forth. The next paragraph stated that around 3,900 people had completed enrollment but hadn't actually paid their first premiums yet.

I used this 125K number (broken down roughly 40% private / 60% Medicaid, based on MN's earlier numbers) on the spreadsheet. As you can see in the previous (Week 12) worksheet, Minnesota's prior update was broken out the same way, since the previous 12/18 StarTribune article (by a different reporter) also gave the "total people represented" number in the last paragraph.

However, when I checked the "MNsure grids for late crush" article this morning, both the "125K people" and "3,900 unpaid" paragraphs are missing (instead, there's a new (I think) paragraph stating the exchange's hoped-for goals of 70,000 private & 135,000 total enrollments). (I should also note that 12,051 + 12,708 + 19,420 only adds up to 44, 179, not 53,000, so I'm not sure what that's about.)

Unfortunately, I didn't happen to get a screen shot of the original version, and Google cache doesn't have it either (presumably because it was changed within just a few hours of being posted), so I have no proof of this, but I assure you, they were there as of around 7:00pm EST last night.

I've already contacted Ms. Crosby to inquire whether she knows what the deal is with those paragraphs being removed, and will update if/when I hear back from her. I'm assuming that either a) the data was incorrect or b) it's correct but wasn't supposed to be publicized, though I can't imagine why that would be a secret. As for the data being incorrect, the total individuals covered as of 12/14 was 97,573 according to the previous article, so 125,000 as of 12/27 is more than reasonable (in fact, given the Christmas rush, it seems to me it should be even higher), so that doesn't seem to make much sense either.

Anyway, if it turns out that I have the Minnesota numbers wrong I'll correct them as quickly as possible, but in the meantime I have a minor mystery to resolve here.

Oh, and it also looks like I'll have to start taking screen shots of every source article going forward, which is unsurprising but depressing at the same time.

UPDATE: OK, I've heard back from Ms. Crosby and she explained the confusion to me.

The "125K people represented" is for people represented by the 72K applications, not for those actually enrolled. She directed me to the latest MNsure Metrics PDF, which (sort of) explains the discrepancy, though it's still a bit confusing. Apparently the applications can include more than one person, but the enrollment numbers are actual people, which means that not only are my "50,000" and "75,000" numbers way too high, but the Week 12 number (97K "people represented") was off as well. The source of the "3,900 unpaid" number is still vague as well, so I've removed that completely; I therefore am still not sure what percentage of MN enrollees have paid yet, though they have until January 10 to do so.

On the other hand, we mutually realized that the "12,051" number was a typo; it should be "21,051" which makes the total of the 3 numbers 53,179, which is indeed "over 53,000". 

Thus, the correct numbers should be:

  • Private Enrollments: 19,420
  • Medicaid/SCHIP: 33,759

Correcting now, sorry about the error.