UPDATE 2/20/20:California and New York have come out with their final, official enrollment numbers, as have both Vermont and Rhode Island, although the latter two are estimates on my part due to the way they're reporting their enrollment data. Vermont is lumping together both on- and off-exchange enrollees, while Rhode Island is only reporting those who had paid their first monthly premium, which is different from the way CMS officially reports QHP selections.
California and New York have both released updated 2020 Open Enrollment numbers, so I figured I'd update my spreadsheet one more time before the final data is released. This time I've included a smaller secondary table at the bottom which adjusts the Federal and State-based exchange numbers for Nevada.
OFFICIALLY, HealthCare.Gov enrollment is down nearly 128,000 people this year, but that's not fair because Nevada broke off of HC.gov onto their own full state-based exchange platform this year. When you adjust for that, HC.gov is only down 119,000 people for the remaining 38 states. Meanwhile, the state-based exchanges are officially down 2,900 at the moment, but again, with Nevada joining them, they're actually down around 8,900.people.
That leaves the missing enrollment data from five states. Rhode Island and Vermont haven't released any data...I'm assuming they'll both be very close to last year (call it at least 33,000 and 24,000 respectively). I'm assuming New York + DC will be good for perhaps 3,000 more enrollees combined in their final days. And California will likely tack on another 30,000 or so in their final 2 days of Open Enrollment.
Incomplete numbers have been released for California, Colorado, Connecticut, the District of Columbia, Massachusetts and New York, all of which still have ongoing Open Enrollment, and I'm still waiting on any enrollment data for Rhode Island or Vermont.
With all that in mind, here's a state-by-state breakout showing where things stand as of today, Jaunary 9th. The states have been sorted from worst-performing to best, although obviously the 8 states with partial or no data are misleading (vice-versa for the bar graph).