Another Hanukkah miracle due in Week 7 of ACA Open Enrollment for 2020?

Subscribe to xpostfactoid via box at top right (requires only an email address; you'll get 2-3 emails per week on average)

Last year, enrollment in the ACA marketplace in the 39 states using HealthCare.gov trailed the prior-year enrollment pretty dramatically in each of CMS's "weekly enrollment snapshots."  By the end of Week 6, enrollment was at 88.3% of the total at the same point in December 2017.  Then lo, at the end of Week 7 the gap had shrunk from 12% to 4%.  What caused the late surge?

The surge was illusory. Week 1 was a day short last year. That was no mystery, and trackers accounted for it, noting that first-week enrollments stood at about 125,000 per day. What was not fully accounted for in advance was that the extra day made up in Week 7, which had seven enrollment days last year vs. just six in 2017, carried a lot more weight than the shorted November day, as enrollment is much heavier at the end of the season. As I noted at the time:

Enrollment season starts slowly and ends fast. This year, enrollments averaged 123,892 per day in Week 1 and 617,493 in Week 7. More than half of all enrollment -- 51% -- happened in the final week (boosted by about 1.7 million auto-enrollments or passive renewals, added in at the end). Last year, 47% of enrollees were booked in the final (6-day) week.

Through Week 6, this year's tally was short a day that roughly equalled 1.4% of the end total. The extra day in the final week, conversely, was worth 7.3% of the total. Thus - hey presto -- the year-over-year apparent enrollment drop shrank from 11.7% to 4.2% -- notwithstanding that per-day enrollments in Week 7 this year were just 89% of last December's per diem total (617,493 vs. 690,661).

This year, we may be in for another December surprise. Like last year, enrollment to this point has had one less day than enrollment at the end of Week 3 last year -- 16 days vs. 17. Also like last year, enrollment is apparently lagging, down 13% -- 1,669,401 to date vs. 1,924,476 at the end of Week 3 last year. (This week is more or less flat.)  Subtract a day (6%) from last year's total and the gap shrinks roughly in half.  Nevada's withdrawal from the federal exchange, and Maine implementing a Medicaid expansion this year, account for a small part of that gap (about 19,000 for Nevada and perhaps 4,000 for Maine).

It looks, however, like the calendar may be setting us up for another Hanukkah miracle. Week 7 will begin on Dec. 8 and end on Dec. 14.  Open enrollment ends on December 15. So the enrollment oil will burn for 8 days in the endgame.

If about 40% of the final-week enrollment surge is due to auto-re-enrollment, that suggests about 375,000 active enrollments per day in that week. All other things being equal (admittedly a big if), the enrollment gap as of the end of Week 6 may close by about that much as of the end of Week 7 (or Week 8, if the final day is counted by its lonely self). The gap as of now is 255,075 -- or about 230,000 if you account for Nevada and Maine.

Other factors further reduce apparent enrollment drops. The twelve state-based marketplaces -- now thirteen, with Nevada added to the ranks -- have outperformed the states on the federal exchange in the Trump era. Enrollment in SBEs was up 1% at the end of OE last year, while it was down 4% in HealthCare.gov states. Further, attrition as of the first quarter and year-end has dropped in recent years, to the extent that enrollment as of December 2018 in HealthCare.gov states was higher than in December 2016 (perhaps in part because of silver loading, which has put a lot of people in free bronze plans or in gold plans that provide better value than silver plans offered at only slightly lower, or even higher, premiums).  Much of the drama that we track in real time is a mirage.

UPDATE, 12/20: Enrollment in HealthCare.gov states is indeed essentially flat this year. Joshua Peck has the lowdown.