How many more enrollees will the Hurricane Special Enrollment Period add?

Now that it appears that the full list of states and counties eligible for hurricane (or windstorm, in the case of Maine) Special Enrollment Periods (SEP) has settled down, Huffington Post reporter Jonathan Cohn asked an interesting question:

How if at all do you allow for the extensions in FL, TX, etc.? Or, to put another way, how many post-Dec 15 signups through https://t.co/bhGNSognZK do you expect?

— Jonathan Cohn (@CitizenCohn) December 20, 2017

The closest parallel to this particular situation I can think of was the #ACATaxTime SEP back in spring 2015. In that case, it was the first year that the ACA's (defunct as of this morning) Individual Mandate was being enforced, and a lot of people either never got the message about being required to #GetCovered or at least pretended that they didn't.

This led to a whole bunch of confusion and anger as millions of people went to file their 2014 federal tax returns. In response, HealthCare.Gov (along with most of the state-based exchanges) added a unique, one-time 6-week Special Enrollment Period for people who "forgot" or "didn't know they had" to enroll.

There was much speculation as to how many additional enrollees this would generate; I spitballed that it could be anywhere between 600K - 1.2 million people, but the truth is no one had a clue, really. In the end, the total turned out to be around 214,000 QHP selections nationally tacked onto the official Open Enrollment Period tally.

Anyway, I decided to take Cohn up on his challenge: Just how many people will enroll using the "hurricane SEP" announced by CMS Administrator Seema Verma back on September 28th?

Well, first, the hurricane SEP lasts from December 16th - December 31st.

Next, let's look at the full list of counties/states eligible. After a lot of updates it appears that the final list includes:

• The entire state of Alabama
• The entire state of Florida
• The entire state of Georgia
• The entire state of Maine (due to a massive wind storm, not a hurricane)
• The entire state of South Carolina
• 53 counties in Texas (Aransas, Austin, Bastrop, Bee, Bexar, Brazoria, Burleson, Caldwell, Calhoun, Chambers, Colorado, Comal, Dallas, DeWitt, Fayette, Fort Bend, Galveston, Goliad, Gonzales, Grimes, Guadalupe, Hardin, Harris, Jackson, Jasper, Jefferson, Jim Wells, Karnes, Kleberg, Lavaca, Lee, Liberty, Madison, Matagorda, Milam, Montgomery, Newton, Nueces, Orange, Polk, Refugio, Sabine, San Augustine, San Jacinto, San Patricio, Tarrant, Travis, Tyler, Victoria, Walker, Waller, Washington, Wharton)
• 32 parishes in Louisiana (Acadia, Allen, Assumption, Beauregard, Calcasieu, Cameron, De Soto, Iberia, Jefferson, Jefferson Davis, Lafayette, Lafourche, Livingston, Natchitoches, Orleans, Plaquemines, Rapides, Red River, Sabine, St. Bernard, St. Charles, St. Helena, St. James, St. John the Baptist, St. Martin, St. Mary, St. Tammany, Tangipahoa, Terrebonne, Vermilion, Vernon, Washington)
• 7 counties in Mississippi (George, Greene, Hancock, Harrison, Jackson, Pearl River, Stone)
• The 200,000 - 300,000 Puerto Ricans and U.S. Virgin Island refugees who've since moved to the U.S. mainland

I went through and added up the total populations of all of the above states/areas and came up with roughly 42.1 million people in the five full states above, 15.8 million in the listed Texas counties, 2.83 million in the 32 Louisiana parishes, 481,000 in the 7 Mississippi counties, and let's call it 250,000 from Puerto Rico/USVI. That's a grand total of around 61.5 million people.

That's around 19% of the total U.S. population. Now, the enrollment total as of last December 15th is a bit tricky to lock down due to how and when auto-renewals are added. The official total as of 12/17/16 was 8.8 million, which included auto-renewals from the state-based exchanges but not the federal exchange. The total as of Christmas Eve, 12/24/16, was 11.54 million, and included all auto-renewals. To the best of my estimates, the total as of 12/15, including auto-renewals was something like 11 million or so nationally.

That means that between 12/15 and the end of Open Enrollment last year, around 1.2 million additional people enrolled. If you 19% of that, you get 228,000 enrollees.

However, that extended all the way out to January 31st (47 more days) whereas the current SEP only runs 16 more days, through 12/31. Furthermore, most of the hurricane SEP stretches across the holiday season--Hanukkah, Christmas and New Year's, when enrollment tends to drop to a crawl. In addition, if the outreach/advertising for Open Enrollment at HealthCare.Gov was anemic in general this year, I'm guessing it's nonexistent for the SEP in these states. Finally, while normal SEPs can be enrolled in via the normal HC.gov website, in this case enrollees will have to go through a rather cumbersome process for reasons unknown:

Our instructions are: Call @HealthCareGov for SEP. You will receive approval by mail. Then enroll. Damage is not necessary. SEP eligibility is based on county of residence which for us is every county.

— Palmetto Project (@PalmettoProject) December 20, 2017

On the one hand, kudos to CMS for not requiring documentation of damage, etc etc...on the other hand, why the heck do you have to call them and fill out the forms by mail? Why not simply list it as another checkbox option alongside "Lost Coverage", "Moved", "Gave Birth" and so on?

(sigh) Anyway, the point is that there's some obstacles and a lack of awareness which will no doubt dampen enrollment for this reason. I'm guessing it won't be more than 25% of the theoretical potential, so let's call it perhaps 57,000 people.

Let's look at it another way: How many people from the effected areas actually did enroll after 12/15 last year? Well, I did my best at calculating that as well and came up with around 442,000 people (I included 100% of post-12/15 enrollees in AL/FL/GA/ME/SC, 60% in LA/TX and 16% in MS). Again, divide by 4 and you get around 110,000 enrollees.

Finally, a third theory: Only look at the enrollments between 12/15 and 12/31 last year, since that's the actual last day of the SEP. As far as I can tell, only around 550,000 people enrolled nationally over those 16 days last year; take 19% of that and you get about 105,000 people. Since the January 31st factor is now out of the picture, let's call it half of that instead of a quarter, and you're left with around 50,000 enrollees.

Basically, no matter how I slice it, I can't see more than 50K - 100K people taking advantage of the hurricane SEPs across all 8 states.

Then again...who knows?