Philadelphia Inquirer messes up Medicaid numbers

Hat Tip To: 
Esther Ferington, Xpostfactoid

A few days ago, several people sent me a link to this story from the Philadelphia Inquirer which discussed the impact of the ACA on Medicaid enrollment in the 26 states which expanded the program vs. the 24 which haven't done so (including Pennsylvania), and the positive impact it would have on PA's uninsured rate if they were to do so.

However, the following passage seemed rather strange to me, and I didn't want to post anything about it until I cleared up the mystery:

Pennsylvania and the 23 other states that haven't expanded Medicaid have signed up fewer than three million people for the program. That's in sharp contrast to the almost 10 million people enrolled by the 26 states and District of Columbia that expanded Medicaid when the marketplace opened November.

Using federal statistics, the Institute did another study and found that the 10 million people now on Medicaid represent about half of the total number projected to be in the program in 2016 when enrollment is expected to stabilize.

"We should be over halfway [to the projected total] by the end of 2014," said Buettgens. "That is very good for something that includes such a large new expansion of Medicaid."

Huh. According to the above passage, Medicaid has seen a net gain of 3 million people from non-expansion states, and another 10 million from expansion states. Proportionately, this sounds about right...but the actual numbers seem way too high (and I say this as someone who is obviously rooting for the law to be successful). 13 million nationally?? That makes no sense whatsoever; the official HHS number as of the end of April was a little over 6 million, and even my own estimates only put the total at around 9.6 million...and that includes an unknown number of people who have been "determined eligible" for Medicaid/CHIP but not actually enrolled yet. In addition, all 4 of the major national surveys released recently (2 of them put out just today) give the total net reduction in uninsured people at somewhere between 10 - 13 million, including privately paid QHPs as well as Medicaid.

In short, there's no way that the "3 million + 10 million" figure is correct.

Well, thanks to contributor Esther Ferrington, I now know where the problem is: Those numbers come from the Urban Institute's projection for the end of 2016, not the current enrollment situation. As Esther notes, if you read the actual UI's report, you'll find the following right there on Page 2:

This brief assesses how reported changes in enrollment in Medicaid and CHIP during this period compare with changes in Medicaid/CHIP enrollment projected by the end of 2016 by the Urban Institute’s Health Insurance Policy Simulation Model (HIPSM).

And it's not like that's the only place; the "end of 2016" statement is found easily a half-dozen times throughout the report:

When all states were expected to expand Medicaid under the ACA, enrollment in Medicaid and CHIP was anticipated to increase by over a third by the end of 2016; 

Taking into account current state decisions on Medicaid expansion, Medicaid/CHIP enrollment is projected to increase by 12.8 million by 2016, of whom 9.9 million live in states expanding Medicaid and 2.9 million live in states not expanding Medicaid.

Taken together, this implies that as of April 2014, Medicaid and CHIP programs were almost halfway—or 47 percent (i.e., 6.0 million divided by 12.8 million)—toward reaching the net enrollment increases expected by 2016 (Figure 2, first bars), up from 39 percent as of March 2014.

I should also note that the last two quotes above are where the "3 million / 10 million" numbers come from (actually 2.9 million / 9.9 million, though I guess rounding off is OK in this instance), as well as the "half the total number" quote (47%...again, I guess I won't quibble over a 3% difference).

Anyway, it's a great article overall, and the larger point is certainly valid (that Pennsylvania should get off it's ass and expand Medicaid already and stop screwing over 281,000 of their own residents), but they really mangled those numbers by making it sound as though there were already 12.8 million newly enrolled in Medicaid out of 26 million or so expected to, instead of 6 million out of 12.8 million (even if it's actually somewhat higher than 6 million by now).

HOWEVER, there is another interesting tidbit in the original Urban Institute report:

Recent reports indicate that 2.9 million Medicaid/CHIP applications are currently pending in state Medicaid programs. If these applications contained one person each and most applicants complete the process and are found eligible, this would suggest that states received enough applications to result in enrollment that reached more than two thirds of the expected 2016 increases. At the same time, states need to be ready to tackle the renewals for previously enrolled Medicaid beneficiaries as they come up in order to maintain these numbers; this will be challenging in some states.24 We will monitor states’ progress as states revise their enrollment totals and the year goes on.

So, to review: My estimate is around 9.6 million (5.9 million strict expansion; 1.2 million bulk transfers; 2.5 million "woodworkers"). The most recent CMS report has it at just over 6 million. Add another 2.9 million pending (i.e., backlogged in the Medicaid systems of the various states) and you have 8.9 million. Since the 6M figure was only as of 4/30, it's completely reasonable to assume that another 700K or so have enrolled in the past 2 months.

In other words, again, I'm feeling pretty damned confident in my own estimates at the moment.