Tuesday, April 06, 2021

Is the U.S. uninsured rate at an all-time low?

Subscribe to xpostfactoid

Enough states have reported Medicaid tallies through February to posit that the pandemic has increased enrollment by more than ten million nationally since February 2020. Adjusting for the usual difference* between CMS's official totals (now posted through November**) and my sample below (based on state monthly reports), year-over-year enrollment growth since February 2020 is likely about 14.8%, and total enrollment likely stands at about 81.7 million. A quarter of the U.S. population is enrolled in Medicaid.

I don't think we've fully fathomed the effect on access to health insurance of the pandemic, the battered and flawed but still functioning and funded ACA programs in place as the pandemic hit, and pandemic relief measures.  Consider...

The Families First Act effectively required states to pause Medicaid disenrollments for the duration of the Covid-19 emergency, and enrollment will continue to grow until that moratorium ends.***  Enrollment growth among those rendered eligible by the ACA Medicaid expansion is close to 30%. In the ACA marketplace, average monthly enrollment in 2021 will probably exceed 2019 enrollment by at least a million, perhaps more, spurred by extended Special Enrollment Periods and subsidies enhanced by the American Rescue Plan Act. As the population ages, Medicare enrollment grows by about 1.5 million yearly. Meanwhile, the huge job losses triggered by the pandemic appear to have had only a modest effect on employer-sponsored insurance: enrollment through September was down by just 2-3 million, according to a KFF estimate.

Put it all together, and it would appear that somewhere near 10 million more Americans may be insured at present than on the eve of the pandemic.  While the uninsured rate had ticked up from 2017 to 2019, it may well be at an all-time low at present (the NHIS survey shows a drop in uninsured in the second quarter of 2020). 

     Medicaid enrollment in 34 states, February 2020 to January/February 2021 (click to enlarge)

Totals in green are estimated; totals in blue are in states that have enacted the ACA Medicaid expansion; totals in red are for nonexpansion states. Wisconsin is in purple signifying partial expansion (Medicaid eligibility to 100% FPL).

For more context and explanation of my compilation practices, see this post and this post

Update: I forgot about the Census Bureau's Pulse Survey, an experimental online-only survey updated weekly during the pandemic. Health insurance tallies are recorded here. While volatile and to be regarded with caution, the Pulse Survey shows the uninsured rate trending down. As a benchmark, the survey's headnote highlights that in 2019, 14.7% of NHIS respondents age 18-64 were uninsured at the time of interview. In the most recent two week period recorded, March 3-15, the Pulse found 11.6% of adults in that age range uninsured at the time of interview.  The average for the last three periods, from February 3 through March 15, is also 11.6%. Last July, across three one-week periods, the average was 14.0%.

Update 2, 4/7/21: Today CMS released enrollment numbers for the emergency Special Enrollment Period commenced on Feb. 15. Through March 31, it had generated 300,000 more enrollments than SEPs in the prior year in HealthCare.gov states alone. Marketplace enrollment is probably at an all-time high for this point in the year.


* CMS has posted the preliminary November 2020 total, and as I estimated on back on Dec. 31, it shows an increase of 11.6% since February 2020.  CMS's official tallies for all 50 states plus D.C. generally trail my compilations of states' monthly reports by about three months and consistently show about a half percentage point less enrollment growth than my 34-state tally. As the 24 states in my sample that have reported through February show .08% enrollment growth for the month, my sample will come in at about 15.3% year-over-year growth as of February 2021 when the month's tally is complete. 

**CMS posts preliminary and then final tallies for each month. Since each month's final tally typically  differs from the preliminary total by several hundred thousand (throughout the pandemic, final totals have been consistently higher), CMS compares preliminary-to-preliminary monthly totals. The preliminary total for February 2020 that CMS uses for comparison with the preliminary November total is lower than the final February total by 552,394. In the last month for which a final total is available, the preliminary total in October is lower than the month's final total by 315,894. As I noted in November, from January through June last year, CMS's updated monthly total exceeded the preliminary count by an average of 474,539 per month. 

***Conversely, the end of the moratorium is likely to be disruptive, as many states will move aggressively to disenroll those whose incomes have risen above the eligibility threshold.)


Postscript, 4/7/21: Norm Spier, commenting below, is right to keep bird-dogging a major blot on the Medicaid expansion: Medicaid Estate Recovery, to which enrollees over age 55 are potentially subject. That is, a majority of expansion states retain the right to recover the cost of Medicaid enrollment from the estates of all enrollees over age 55, as explained in this post (focus on New Jersey in this post).  As pointed out by commenter Bob Hertz below, MACPAC's 2021 Report to Congress on Medicaid and CHIP was released in March, and Chapter 3 on Medicaid Estate Recovery. While the report is focused almost exclusively on recovery for long-term care services financed by Medicaid (the original target of estate recovery, and also highly inefficient and unequitable), it also relays that 20 states and D.C. currently pursue estate recovery for the over-55 Medicaid expansion population. The map of each state's policy below is from pdf page 47 (report pg. 117) of the MACPAC report. It  reveals that estate recovery in all states yielded a grand total of $773 million in 2019. As Bob Hertz puts it, "This is barely a raised dimple in the overall Medicaid program and not worth the fear and torment that is caused." 

Subscribe to xpostfactoid


  1. I always get a bit upset, (as I have this time at you, Andrew--sorry!), when I see reference to people covered by Medicaid as "insured".

    Though the U.S. Census counts all people with Medicaid (traditional or expanded) as insured, there using a misnomer. Many of them don't have insurance. (From your prior posts, I know you know this!)

    "Insurance" means they pay the bills up to copays, and then they don't ask you, or your estate at you death, for the money back.

    The Medicaids your are listing as covering perhaps 25% of the non-elderly population (for ordinary health coverage--not long term care) in many states is subject to estate recovery for people 55-64.

    The states seem to be able to recover either all medical expenses paid out, or a capitation (the MACPAC document suggests it will be based on whether the state itself self-insured or paid a capitation). In the former case, this is particularly horrendous "coverage" that most people have: there is no insurance at all, people's estates are on the line for all medical bills that come up. The latter case is a bit better--the estates are on the line for a capitation -- a sort of premium, but the treatment is much different than people above 138% FPL, who get highly subsidized (often 100% subsidized now) premiums for real insurance that doesn't require anything be paid back at death.

    By my own examination of estate recovery policies in various expansion states, at least 12 estate recover expanded Medicaid. (With probably more recovering traditional ordinary medical Medicaids. A recent MACPAC report put the number of expansion states recovering some non-long-term-care Medicaid at 20 as on p. 88 here: https://www.macpac.gov/wp-content/uploads/2021/03/Chapter-3-Medicaid-Estate-Recovery-Improving-Policy-and-Promoting-Equity.pdf .)

    Thus in the states that recover (or keep options open to decide to cover retroactive Medicaids) 25% of the population is presumed to have no right to pass on assets.

    This is not what U.S. single payer would do--no one would get just a loan. (E.g. the plans of Sanders and Warren.)

    This is not what Canada single payer does--no one there has to repay anything at death. Nor U.K, nor the rest of the developed world, as far as I can tell.

    The usage of the term "insured" counting Medicaid is a lie. To legitimately describe the numbers in the U.S. Census insured statistics, everywhere used, it needs to be stated that umpteen percent really have just a loan for what may turn out to be all medical expenses, or a capitation, or some combination.

    The mass media, and shamefully, the New York Times and Washington Post commit this sin or misrepresentation and allowing a hoax with "insured" to be perpetrated uniformly and consistently.

    Further, these same sources also don't work to point out a major problem of this kooky ACA contraption contrived from existing parts ineptly: that if a person normally getting on-exchange real insurance loses a job, loses income and gets put on expanded Medicaid, there goes their right to pass on assets. Kaboom thanks to the current ACA combined with state estate recovery policies and what the Feds permit states to recover.

    For references for readers unfamiliar with the estate recovery issue on expanded Medicaid and other non-long-term-care Medicaids in numerous states:

    From this esteemed blog:



    And from me, adding info to Wikipedia in 2019, when I first found out about the issue (coincidentally, same year as Andrew first found out):


    (assertions can be verified from the references, which are all online)

    Backed up, in case someone pulls content from the article:


  2. Thanks for keeping the Medicaid estate recovery issue in play. It is a terrible injustice in most cases.

    You can find a table of the actual recoveries in several states. I believe that the total $ recovered in 2019 was $773 million in all states. This is barely a raised dimple in the overall Medicaid program and not worth the fear and torment that is caused.


  3. Well, Norm, I can always count on you to add the estate recovery caveat to every post focused on Medicaid. It's outsourced.

    Bob, thanks for finding that MACPAC report! While the report is focused almost entirely on estate recovery for LTSS services (also mostly a waste), on pp 17-18 of the pdf, it tells us that 21 Medicaid expansion states (including DC) potentially subject the expansion population to estate recovery. A map on pdf pg 47 shows status in all 50 states.

    1. Yes, Andrew, I am obsessed with the issue!


      How can the mainstream press; the Washington Post and New York Times keep avoiding mentioning the issue? (Washington Post only once, back in 2014, and NY Times never as far as I can tell.)

      If you're describing ACA coverage in a decent newspaper, a most fundamental question is: is it insurance which pays bills without requiring the bills be paid back by (either by the person, by by an estate), or does it require medical bills paid out be paid back by the estate (making it a loan)?

      How can the so-called best of the press fail to report this fundamental aspect of what is going on, and that the ACA, as expanded Medicaid, is the latter (can be a loan for medical expenses), not the former (real insurance)?

      What kind of inability to understand detail among the health reporters at the Times and Post, or perhaps desire to suppress the truth, is behind the Times and Post lack of coverage on this?

      It astounds me.

      Andrew: I appreciate your posting my comments and I know you don't have to.

      And I also appreciate that you covered the issue in your blog twice: in 2019 and again subsequently, unlike probably every every other health reporter or health blog. (I hope the other reporters and the media don't blacklist you for your honesty and clear thinking on the matter.)

      Bob: Your "This is barely a raised dimple in the overall Medicaid program and not worth the fear and torment that is caused." is exactly right.

      Thus, a sample state where I have the estimate, ME, estimated that the estate recovery going beyond the federally mandated long-term-care only accounts for 11-14% of estate recovery revenue.

      ( https://bangordailynews.com/2020/02/05/politics/maine-wants-to-roll-back-estate-recovery-to-spur-medicaid-expansion-sign-ups/ )

      Meanwhile, the MACPAC report, p. 89, indicates the most aggressive recovery state, Iowa, recovered only 14.49 percent of long-term-care expenditures. For non-long-term-care, which includes more actually poor people, the recovery proportion has to be less that 14.49% and therefore combined long-term-care and non-long-term-care recovery in recovery-aggressive Iowa must by no more than 14.49%.

      So, at most, the elimination of the estate recovery on non-long-term-care in recovery-aggressive states like Iowa should result in a loss of < 14.49%*14% of total Medicaid spending, which is 2% of the total Medicaid spending! (This further backs up your statement.)

  4. There are quite a few exemptions from estate recovery, thank goodness. A person who receives Medicaid benefits while they are under age 55 should not be subject to recovery. Also if recovery would leave the inheritors in poverty, there will be an exemption. Of course it is the USA, so some exemptions do not apply in all states.

    Here is a brief outline of exemptions:

    Medicaid Estate Recovery Exemptions

    While state Medicaid programs are very interested in recouping the funds that they spend on their residents’ care, there are some exemptions to the MERP. One of the simplest rules is if a recipient was under age 55 at the time they received Medicaid benefits other than institutional care, then they are exempt from estate recovery.

    Secondly, if a Medicaid beneficiary is survived by a spouse, a child under age 21, or a blind or disabled child of any age (according to the SSI definition of disability), they will also be exempt from estate recovery. Technically, the federal law states that recovery can be made only after the death of the Medicaid recipient’s surviving spouse (if any). For example, if the surviving spouse dies a month after the Medicaid recipient, a state could file a claim for recovery at that time. Many states, however, have taken a more liberal reading of this law, and so long as there is a surviving spouse, no recovery will be made regardless of when they die. Once again, you’ll need to check specific state laws to find out which rule applies.

    Notwithstanding the above, even in a state where recovery may be made after a surviving spouse’s death, there is typically a statute of limitations that bars claims against an estate that are made more than a certain number of months after the beneficiary’s death. In many states, that limit is one year. So, in a state with this rule, if the surviving spouse dies more than a year after the Medicaid recipient, it will be too late for the state to file its claim for estate recovery.

    If Medicaid cannot file a claim when a recipient has a surviving child under 21, can they wait until the child turns 21 and then file their recovery claim? Once again, this must happen within the statute of limitations period, assuming the state does not have a blanket exemption for a surviving child under age 21.

    Finally, there will be no recovery made against the exempt home of a Medicaid recipient (i.e. it will not have to be sold to pay back the state) in either of these two scenarios:

    A sibling of the Medicaid recipient was living in the house for at least one year immediately prior to the date the recipient was admitted to the nursing home and has continuously lived in the house since their sibling’s placement in long-term care.
    An adult child (of any age) of the Medicaid recipient lived in the house for at least two years immediately prior to the date the recipient was admitted to a long-term care facility, has continuously lived in the house since then and provided care to the Medicaid recipient prior to their placement, which may have delayed their move to a facility.

    If all else fails, there’s an exemption against estate recovery if such recoupment would place undue hardship on the surviving family members. For example, if the asset in question for the MERP process is a business, and a forced sale of that business would put surviving family members out of work, then it is considered exempt.

    1. Thanks for the discussion, Bob. I know the exceptions.

      Note that I am focusing my concern on non-long-term-related Medicaids, which compose ordinary non-long-term-care coverage post ACA. Because: post ACA, everyone is supposed to have affordable insurance for ordinary health coverage (but not extending to long-term-care) like they do in the rest of the developed world, and also like U.S. single payer would provide, (like say Canada single payer).

      (Long term care is different. The country hasn't attempted a national long-term-care insurance system like say Germany has, and so it has left the issue as a problem for another day. But the ACA, failing on several accounts now, is supposed to get us to universal affordable not-a-loan ordinary non-LTC medical coverage, like the rest of the developed world has.)

      (If the ACA does not provide that now, it should be evolvable into that. Just as the removal of the 400% FPL "subsidy cliff", for 2 years, gets the ACA for those 2 years substantially closer to affordable insurance or everyone, like U.S. single-payer, or the rest of the developed world. Getting rid of the nonsensical poor-engineering "subsidy cliff" where going from below 400% FPL to over shoots premiums up to 25% of more of income.)

      There is discussion, say, in the Health Affairs "Following the ACA" blog and the NY Times of some of the remaining impediments: the "family glitch", and the out-of-pocket maximums, which are commonly up to $7000 a person and $14,000 a family.

      However, outside of xpostfactoid, there is no discussion (after 2014) of the remaining impediment in the ACA to universal affordable "they don't ask for the money back" coverage (like the rest of the developed world has): Medicaid estate recovery on expanded Medicaid and other non-long-term-care Medicads in at least 12 expansion states for people 55-64.

      It's a bizarre handling of people below 138% of FPL vs those above: possible bill for all medical expenses paid out vs real insurance with high and often 100% subsidies above 138% FPL only as on-exchange subsidized. It's a goof in enginerring just like the "subsidy cliff".

      (Not hitting just poor people with little assets. Before people take Social Security, pensions, and tap 401K's, it is easy to hit the expanded Medicaid criterion <L138% FPL no asset test, blocking ACA subsidized on exchange: if: pausing work, no contracts for a few months for self-employed, early retired, unemployed. It's a financial bomb for people in this kooky hybrid system. It can be patched, as Andrew pointed out, by prohibiting estate recovery on expanded Medicaid: I would go farther: on non-long-term-care-related Medicaids.)

      Health affairs Tim Jost did sharply catch the problem in 2014:


      and reading it, he isolated the fundamental goof: a non-asset-tested component (expanded Medicaid) is subject to estate recovery.

      I insist the press and whatever experts, in missing to describe the additional problem of estate recovery on non-long-term-care Medicaids is making a logical error, if not intentionally covering up, a needed and serious ACA defect.

    2. Bob:

      By the way, send me regards, if you see him around town, as you report you often do, to your neighbor Rick:


      He's become my friend, because he's still active on the issue, after apparently being much of the impetus for Minnesota to stop the estate recovery on expanded Medicaid. (Which they did, in about 2017.)

      Actually, I am in contact with him, and he has told me that the state of MN told him that it could decide to retroactively estate recover expanded Medicaid for any year, including prior years to now. (So that it is not even clear that there won't be an estate recovery on expanded Medicaid 2017 to 2021.)

      To this problem, Rick is trying to get legislation through the MN legislature to have all prior expanded Medicaid records destroyed.

      Such seems to be necessary in our U.S. health insurance system, in which it often seems to me the only sane response is to develop an anxiety disorder.

      (Compare to Ontario, Canada. Everyone covered automatically. There is a designated tax of about 1% of income for health care. Virtually everything is covered. The exception" out-of-hospital Rx's are not covered until a program call Trillium kicks in, paying for everything over 4% of post-tax income, up to minor copays.)

      Thanks, I enjoy the discussion.

  5. Thanks Norm for your comments. I moved away from the rural area where Rick Rayburn lives. He built a remarkable log house, heated with wood and grew a lot of his own food; my house was more like a yuppie antique that you would find in Hudson Valley NY.

    I share your concerns, but I think you are incorrect on Canadian health care taxes. Based on a quick Google search, it appears that a middle to upper income Canadian pays about 10 per cent of gross income in the various taxes that fund health care. Canada also has sales taxes to help pay for health care.

    What makes reform so hard in the USA is that millions of us pay more than 10% for health care, but there are other millions who currently spend much less than 10%. When you start counting the people on Tricare, the recipients of generous employer plans, and the uninsured who are currently healthy, there would be an awful lot of resistance to a 10 per cent tax.

  6. Yes, yes, you are correct in the sense that total taxes in Canada would pay 10% of GDP towards healthcare, being most of the total 11% Canada spends (last 1% being out of pocket). (Whatever sources for the 10%--corporate, individual income tax, etc, with a small part this being a "designated healthcare tax".) It has to be paid for somehow! Of course, through general revenue, most of it, in Canada's single payer.

    What I was referring to as 1% is a designated tax for premium they have in Ontario.

    Let me clip from my frequent NY Times comments on the matter:

    In Ontario, Canada, there is 5% of AFTER-tax income max in total expenses for medical designated premium plus out of pocket. Plus everyone has real insurance, no one gets just a loan.

    This cap on total expenses is as follows:

    There is a small designated tax (like a premium) of at most about 1% of income.

    The main Ontario plan covers everything except out of hospital Rx's, with negligible tiny copays.

    Out of hospital Rx's are not covered at all until about 4% of after-tax income is hit, but then are covered fully up to tiny copays of like $2.


    The tax premium:


    (Premiums are on the last page of the form 6006-c-19e.)

    Main insurance system (missing just out-of-hospital Rx's):


    and, the "Trillium" program, capping expenses for out-of-hospital Rx's at about 4% of after-tax income.


    So my points about Canada are:

    a)everyone is automatically enrolled in the main plan, and everyone who needs it can enroll in the Trillium plan (at no cost).

    b)The Canada single payer system doesn't ask for the bills paid out to be paid back at death from anyone's estate.

    c)Out-of-pocket plus premium for a family (the designated tax for healthcare) is capped at 5% of after-tax income. (In U.S., before the subsidy cliff was temporarily done away with, a family could need 50% of pre-tax income for same. Still here in the U.S., the $14,000 copays for on-exchange are well over 20% of pre-tax income, plus 8.5% for premium. And, for some people, we ask for the bills to be paid back at death from estates,

    So, yes, of course, the rest of Canadian health care spending comes from general revenue. (They have progressive taxation reasonably and fairly distributing the cost to the people in the country--holding down the exposure during illness and for health premiums to 5% of after-tax income.)