Friday, November 27, 2020

State Medicaid enrollment totals in light of CMS's (lagging) reports

Subscribe to xpostfactoid 

Update, 12/21/20: CMS has just posted its preliminary August Medicaid/CHIP enrollment total: 76,489,912, up 8.3% since February (per below, CMS compares preliminary totals from each month for consistency; the final total for August will probably be 300-400,000 higher than the prelim). My own August estimate, based on state reports, was 8.7% growth since February. As I'm showing an 11% increase from February through October, CMS will probably come in at a bit over 10% - probably about 78.5 million total enrollment.

CMS's monthly tally of state-by-state Medicaid enrollment growth in this pandemic year is currently posted through July* -- reflecting five months in which enrollment was affected by the Covid-19 pandemic, and four-plus months in which the suspension of disenrollments mandated by the Families First Act was in effect. 

Most states post provisional Medicaid totals on state websites on a monthly basis, revising at varying intervals, and I have been tracking enrollment in 32 states in advance of CMS's 2-3-month lag. The available state reports vary somewhat from the the CMS tabulations, and both are adjusted retroactively. February-through-July is a large enough sample for me to compare my monthly tallies with the  CMS figures.

From February through July, in 32 states, I recorded a 7.4% increase in Medicaid enrollment.  Since CMS has so far only released its preliminary report for July, an apples-to-apples February-July comparison requires using CMS's preliminary rather than the updated February totals. With each month's new preliminary report, CMS reports updates the prior month state-by-state totals, and the updated totals are higher, as they include retroactive enrollment and enrollees whose applications were not finalized in the prior month. The February updated total is 405,491 higher than the preliminary February tally.** Comparing the preliminary totals for both months, as CMS does here, suggests a 6.9% increase nationally from February-July, and 6.8% in the 32 states I track, as the chart below shows. Links to the state-issued reports are provided at bottom.

The most recent published update to my 32-state sample derived from state-published supports suggests about 9.7% increase from February through September. Reducing that estimate proportionately to the July estimate would peg growth from February-September at 8.9%, suggesting total enrollment of about 77.5 million. [Update, 4/5/21: CMS's updated total for September is 77,682,409.] Totals look to be up more than 1% in October, putting enrollment above 78 million.

Subscribe to xpostfactoid

Factors that may affect the difference in tallies are discussed below. Links to the state reports I've used are provided at bottom.

    Medicaid enrollment in 32 states, February-July 2020 
    State tallies compared to CMS preliminary monthly totals

Factors that may affect the difference between the state tallies and CMS's:

1. The set of programs that states include in their tallies differs to varying degrees from those counted by CMS -- sometimes including more programs than does CMS (which counts only those programs providing Minimum Essential Coverage) and sometimes fewer. Probably the largest source of difference is in the  states that tally only enrollment in Medicaid managed care programs, leaving out those enrolled in fee-for-service Medicaid. Since states have been transitioning steadily to managed care, I suspect that the FFS Medicaid tallies have probably remained mostly flat, resulting in lower overall average increases in CMS's totals than in the state managed care tallies.

2. Some states reports revise totals for prior months, in which case I revise the earlier totals. Minnesota, for example, reported its March total ("thru February" in my tally) as 841,241 in March, and has revised it to 848,098 in its November report.

3. Minnesota shows a particularly sharp variance between state report and CMS report: the state reports indicate an increase from February through July of about 110,000, compared to about 77,000 for CMS. I don't know why this is; I will query the state. 

4. CMS tallies are as of the "the last day of the reporting period." In the database, however, the report date for each month is the first of that month. In my tallies,  for states that report early each month, and label the report in the month reported -- e.g., "August" results reported August 3 -- I back-date, labeling the August report "thru July."

This post is perhaps mostly for my own benefit: in my estimates of total enrollment growth derived from state reports going forward, I'll be able to take into account the difference between the state-published tallies and CMS's totals. We'll see where CMS's updated July total lands, and going forward I will take both preliminary and updated new CMS reports into account.

Below are links to most of the state reports I access.  I found many of them through an analysis of enrollment through May published in Health Affairs by Chris Frenier, Sayeh S. Nikpay, and Ezra Golberstein, to whom much thanks are due. The article is open source; click on the "supplemental material" link for a list of state websites, and specs as to which programs are included for general and ACA expansion totals. 

Expansion states 
California - by eligibility group - thru March as of Aug. As of 60 days after elig. month
California - up to date. Massive database. Certified eligibles by aid code
Maine  - expansion only

* The Kaiser Family Foundation relays CMS totals in a convenient format that makes month-to-month comparisons easy -- keeping in mind that the latest month logged will be a preliminary rather than updated total.

** From January through June this year, CMS's updated monthly total has exceeded the preliminary count by an average of 474,539 per month. The difference between my monthly estimates of the percentage increase and CMS's official tallies, which lag by 2-3 months, accordingly may be still narrower than 7.4% vs. 6.8%  (or, conversely, wider; the June increase in CMS's updated total was just 273,117). 

Subscribe to xpostfactoid 

No comments:

Post a Comment