Final National Projections: Major state-level changes, overall somewhere between old & new estimates

Politico health editor Joanne Kenen, yesterday:

@SecBurwell expects a spike in signups near deadline but not as big a surge as last year's enrollment end

— Joanne Kenen (@JoanneKenen) February 13, 2015

OK, let's unpeel this onion. HHS's official target: 10.4 million QHP selections (which has already been reached). My original target was 12.0 million (made on 11/14); I increased this to 12.5 million on 12/12, which is what I've stuck to ever since.

When the dust settles, who will prove to be more accurate...the head of the Health & Human Services Dept...or me? :)

A couple of days ago, I took a look at my own target to see whether 12.5 million still seems realistic or not. I looked at the data from several different angles, and concluded that 11.6 million should be the bare minimum QHP tally, which would essentially split the difference between 10.4M & 12.5M.

After that...

I went back to my 2014 spreadsheet, and as far as I can tell, it looks like total QHPs crossed the 5.4 million mark at midnight on March 24th. They hit 7.1 million as of midnight on March 31st. That means they managed to enroll roughly 1.7 million in the final 7 days of the official enrollment period. My projection assumes around 2 million in the final week of 2015's open enrollment period, which isn't all that much higher (I was actually surprised by this; I hadn't checked those particular numbers when I came up with my 2015 "curve")

Let's look at the individual states again (reminder; in this context, "underperforming" only refers to being measured against my state-level projections, which may not have been realistic in the first place). On February 3rd, I reduced my targets in 9 states, but raised them in 26 other states (albeit not as dramatically). The two pretty much cancelled each other out.

Today, with a couple more weeks worth of data baked in, I'm making some further, final adjustments (and in fact I've tweaked every state except for Iowa and Tennessee):


In the end, I'm cutting my projections in 14 states, with the following caveats/explanations:

  • CALIFORNIA: Ouch. I originally targeted 1.94 million; CoveredCA's own target was 1.7 million. There are strong signs that they're not going to meet their own target at this point; I've dropped my final projection down to 1.6 million. The wildcard here: CoveredCA has announced a 5 day "wait in line" extension for people to wrap up the enrollment process, which could nudge them over the top, but there's no way they'll come close to my original number.
  • COLORADO: In this case, Connect for Health CO was probably just too ambitious in the first place, targeting 194K people. My own target was even more unrealistic; at this point it looks like they'll end up around 143K, which would be a rather disappointing 14% increase over last spring.
  • KENTUCKY: I've knocked my target down to pretty much what HHS is figuring, around 105K.
  • MASSACHUSETTS: NOT THEIR FAULT. I can't stress this enough...I completely misunderstood the situation with MA's weird "temporary coverage" arrangement. I had the correct number of people enrolled in it, but didn't realize that a good 1/3 or so actually qualified for MassHealth (Medicaid), not QHP enrollment. As a result, my 300K original was insanely out of line; 160K is a more realistic goal. Wildcard: Due to a series of massive snowstorms over the past few weeks, the MA Health Connector has announced an 8-day full deadline extension, which means that whatever number they have at midnight tomorrow should not be taken as the final word.
  • MICHIGAN: This is the only state run through Healthcare.Gov where I've had to significantly lower my target. I have no idea what factors are at play here...although the fact that 46K - 69K more people have turned out to qualify for ACA Medicaid Expansion than previously thought might have something to do with it. As a result, I've lowered my outlook for my home state from 425K to 370K.
  • MINNESOTA: You can "blame" MinnesotaCare, the unique "basic health plan" provided for by the ACA that MN has in place, for the majority of their QHP enrollment issues. MNCare is sort of a public/private hybrid; if it didn't exist, just about all of the 24,000 people enrolled in it would otherwise be very likely to have purchased a highly-subsidized QHP instead. Regardless, I've dropped my target for MN from 75K to just 55K...well below their own goal of 67K.
  • WASHINGTON: Between ongoing serious technical problems and an unexpectedly high number of 2014 enrollees who turned out to be eligible for Medicaid expansion, WA has been coming up short all season. It's also important to keep in mind that they're the only exchange which is only reporting PAID enrollments, meaning that we're comparing apples & oranges here...assuming an 88% payment rate, their 139K as of 2/10 likely represents about 158K QHP selections. However, since in their case, it's the paid number which will be used, I'm lowering my target to a mere 153,000 by tomorrow night, well short of their 215K target.
  • RHODE ISLAND: RI"s exchange has actually done everything right: Their site has operated smoothly for both enrollment periods, and they made sure that every 2014 enrollee actually double-checked to see what their options were. Even in terms of reporting, they're making sure to purge unpaid enrollments as soon as the due date passes. Even so, it's a very small state (just over 1 million people), so it's a struggle to beef up their numbers. I'm dropping my target for them to just 34K.
  • VERMONT: VT is a special case. On the one hand, they've been pushing hard for single payer and even legally required people buying individual/family healthcare policies to do so via their ACA exchange. On the other hand, they already had a tiny uninsured population to begin with (only about 45,000 people), and their exchange is still something of a technical mess. I'm dropping my target below their orignal one, to just 33,000 people.
  • Connecticut, Nevada, New Hampshire, New York & Oregon: I reduced my targets for all 5 states, but only slightly for each. Wildcard: New York just announced a 13-day "waiting in line" extension period which should shore their numbers up, even if it takes another 2 weeks for them to be reported.

And yes, I did note that only 4 of the 14 states (MI, NH, NV & OR) are being run through 2 of those, Nevada and Oregon, were only brought onboard this year. I'll let others ponder the significance of this.

So, that's a total reduction of 765,000 QHP selections. If that was the whole story, I would reduce my national projection down to around 11.7 million and call it a day.

HOWEVER, I've also increased my final targets for practically every other state. I won't go into detail on all of them, but here's the major ones:

  • FLORIDA: Holy I noted in my most recent exclusive, Florida was already 41% ahead of their 2014 tally as of 8 days ago, and their enrollment outreach machine shows no sign of stopping. I've increased my final target for FL from 1.5M to 1.67M...a 70% increase over last year's tally.
  • VIRGINIA, TEXAS, GEORGIA, MISSOURI, NORTH CAROLINA, INDIANA, ARIZONA, LOUISIANA, SOUTH CAROLINA, PENNSYLVANIA, ALABAMA & OKLAHOMA: I've increased my target fro each of these states by 20,000 - 66,000 apiece. Notice a pattern here? Again, aside from  purple Pennsylvania and Virginia, every one of these is Republican-held as far as the eye can see. The fact that they're all outperforming my expectations is a testiment to the strength of the ACA in the very states which resent it the most. Also note that both VA & OK have already reached my original target, so increasing my final projection is kind of obvious for them.
  • IDAHO: This should be included in the list above, of course; I bumped it up by 30K even. I've listed it separately purely because ID is the only state which broke away from Healthcare.Gov this year to start their own full exchange...under a hard-GOP leadership. So far they seem do to be doing very well, although they haven't updated their data since 1/17.
  • Illinois, Utah, New Jersey, Kansas, Mississippi, Nebraska & Maine: I bumped each of these up anywhere from 10 - 20K. Again, note that 4 of the 7 are solidly red states.
  • Hawaii, Maryland, DC, Alaska, Delaware, Ohio, West Virginia, Wyoming, Arkansas, North Dakota, Wisconsin, Montana, New Mexico & South Dakota: I nudged each of these up a bit (from 2K - 9K). Again, 7 of these states have been confirmed to have already reached my original target.

Now, obviously the final number is not going to be exactly 12.5 million, and the overperforming states aren't going to exactly cancel out the underperformers.

Still, I'm pretty confident that when the dust settles, it'll be somewhere between my original projection (12.0 million) and my official one (12.5 million)...although I suspect it'll still be at the higher end of that range; let's call it between 12.2 - 12.5 million.

In other words, yes, I'm finally backing off of my until-now steadfast 12.5 million figure...but that could just as easily be proven to be right on the nose, or even undershooting.

On the other hand, at the moment the following message is showing up on the Massachusetts exchange website...