Saturday, September 17, 2022

Will Medicaid's "great unwinding" when the PHE ends trigger a "great uninsuring"?

Subscribe (free) to xpostfactoid

During the pandemic, from February 2020 to May 2022, Medicaid enrollment increased by 18 million, or 29%, according to administrative data that CMS collects from states. That's mainly because of a moratorium on disenrollments that began in March 2020 and has yet to end. The moratorium will end when the federal government declares an end to the  Public Health Emergency, which will happen in mid-January 2023 at the earliest (the PHE has been extended repeatedly in 3-month increments). 

As noted in my last post, the disruption that may be triggered by the resumption of state "redeterminations" of Medicaid enrollees' eligibility, and subsequent disenrollment of some, is a focus of considerable angst -- and preparations, in states where Medicaid personnel are committed to keeping as many people insured as possible, to proceed with due deliberation and compassion. The Urban Institute has estimated that 15 million people may be disenrolled over the course of a year, the time period that CMS has asked states to devote to clearing the "redetermination" backlog. The Kaiser Family Foundation (KFF) estimates somewhat more modest losses, in a range from 5.3 million to 14.2 million.

This week the Census Bureau released its annual report on health insurance coverage in the United States. Based on the annual supplement to the Current Population Survey, the report shows a more modest increase in Medicaid enrollment from 2020 to 2021 -- 0.9% -- than CMS's administrative data would indicate.  According to CMS, Medicaid and CHIP enrollment increased by 6.6 million from December 2020 to December 2021. That's about 2% of the population.

The Census Bureau also released a second report, spotlighting health insurance changes over two years, and based on the American Community Survey. which interviews people throughout the year about their current insurance status (the CPS, conducted early in the year, asks respondents if they were insured at any point in the past year).  The ACS also shows a gap between Medicaid enrollment gains as reflected in administrative data compared to the survey data. According to the report, the percentage of the population insured by Medicaid increased by 1.3% over two years, from 2019 to 2021 (based, again, on surveys conducted throughout each year). The administrative data records an increase of 11.7 million enrollees from June 2019 to June 2021. That's about 3.5% of the population.

An analysis of the ACS data by KFF attempts to explain this gap. The explanation suggests to me that the disenrollments that will begin at the end of the PHE may not be as disruptive as "15 million disenrolled" might indicate -- at least in states that work in good faith and with due diligence to establish contact with all enrollees, accurately determine their status, and help them consider their options.  My emphasis via yellow highlight below (the bolded subhead is in the original):

Policies adopted during the pandemic to ensure continued coverage in Medicaid were largely responsible for the decline in the uninsured rate. Specifically, provisions in the Families First Coronavirus Response Act (FFCRA), enacted at the start of the pandemic, prohibit states from disenrolling people from Medicaid until the month after the COVID-19 public health emergency (PHE) ends. The ACS data show an increase of 1.3 percentage points in the Medicaid coverage with 69 million covered by Medicaid in 2021. Data from the Centers for Medicare and Medicaid Services (CMS) showed that Medicaid enrollment in May 2022 had increased by nearly 25% since February 2020, much larger than the ACS increase, with 87 million enrolled as of December 2021. There are long-standing discrepancies between survey and administrative data, due in part to different ways of counting people.  Medicaid administrative data are reported by states and reflect enrollment at the end of a given month while the ACS asks individuals about coverage at a point in time. However, many people may not know they have Medicaid coverage, perhaps because their coverage is administered by a private managed care plan, and may misreport their source of coverage on the survey. National survey data also typically undercount lower income people who are more likely to be covered by Medicaid. Because so many people have been kept continuously enrolled in Medicaid during the public health emergency – in many cases without any notification – the disparity between administrative and survey data may be exacerbated.

The first passage highlighted above addresses longstanding disparities between survey and administrative data for Medicaid enrollment. The second addresses conditions particular to the pandemic and the disenrollment moratorium. In the spring of 2020 some 20 million people lost their jobs, and states took a variety of emergency measures to smooth the path to Medicaid enrollment for many of them, such as modifying eligibility criteria,  granting presumptive eligibility to people with sudden income losses, and, in the case of Kentucky at least, reaching out to the newly unemployed.  Job recovery since that vertiginous early drop has been extraordinarily swift, and some people who picked up employer-sponsored insurance may not have informed their state or local Medicaid agencies. 

Subscribe (free) to xpostfactoid

Even in ordinary times, Medicaid agencies are often sclerotic and hard to communicate with, notwithstanding extensive system upgrades in the year's since the ACA's passage. In a previous post, I shared some anecdotal evidence of this:

I know a young man who, during a year of transition, lived in two states and worked at three jobs, with a period of unemployment. At different points in the year he applied for Medicaid in two (blue) states and received rejection notices. From both of those states, months later (and months apart), while insured through a new employer, he was sent managed Medicaid membership cards and informed that he'd be enrolled since shortly after his application was completed.  In both states, it took some doing and some time to get himself disenrolled.

Under the Biden administration, CMS, under the leadership of Chiquita Brooks-LaSure, is urging states to do everything possible to "unwind" backlogged Medicaid redeterminations in orderly fashion and minimize increases in the number of people uninsured. Under the Trump administration, conversely, under the leadership of Seema Verma, CMS encouraged states to throw as many people off the rolls as possible via periodic data matches, which 30 states had enacted by January 2020. States vary, needless to say, in their impulses to pursue these opposing goals -- and in their capacity to minimize disruption if their leadership is so inclined. 

Late last month, CMS announced a proposed rule that, if finalized,

would standardize commonsense eligibility and enrollment policies, such as limiting renewals to once every 12 months, allowing applicants 30 days to respond to information requests, requiring prepopulated renewal forms, and establishing clear, consistent renewal processes across states.

Public health insurance in the U.S. remains administratively burdensome, confusing, and limited, whether by provider network or protection from out-of-pocket costs. On the other hand, the safety net we do have, bolstered by the ACA, pandemic legislation, emergency measures and new Biden administration initiatives, did manage to fend off a swelling of the uninsured population under the economic shock of the pandemic. That is an accomplishment. Whether the unwinding of backlogged Medicaid redeterminations also unwinds a good piece of that accomplishment remains to be seen.

No comments:

Post a Comment