• Chart of the day: Cognitive ability and enrollment into public programs

    I’ve seen it suggested in the comments on this blog and elsewhere that low take-up rates among those eligible for Medicaid suggests they do not find the program to be of value. I have not seen any strong evidence to support that view, but I’m happy to look at it if it is suggested to me. Meanwhile, there are other potential reasons for low take-up rates including cognitive limitations.

    A new study by Ifedayo Kuye, Richard Frank, and Michael McWilliams examines the relationship between cognitive limitations and awareness of and enrollment into Medicare Part D’s low-income subsidy (LIS) program, which offers reduced premiums and cost sharing for Medicare’s drug benefit to beneficiaries with incomes below 135% of the federal poverty line and assets below $6600 if single or $9910 if married (2010 figures). The investigators’ source of data was the Health and Retirement Survey (HRS). (Yes, I have switched from Medicaid to Part D’s LIS. They are different and serve different populations. Acknowledged.)

    To assess overall cognitive abilities, the HRS uses a validated survey instrument modeled after the Telephone Interview for Cognitive Status, an adaptation of the Mini-Mental State Examination for use over the telephone. Participants were asked to complete a series of tasks assessing orientation, attention, memory, word recognition and comprehension, and ability to count and perform simple arithmetic. Summary cognition scores could range from 0 (no tasks completed correctly) to 35 (all tasks completed correctly).

    The chart below reports Part D enrollment, LIS awareness, and LIS application as a percentage of LIS-eligible beneficiaries, by quartile of cognition score. The paper includes other results by other cognitive metrics, but this suffices to make my point. Results are adjusted for age, sex, race, ethnicity, health status, chronic conditions, depressive symptoms, and difficulties with activities of daily living.

    LIS cognition

    It could still be true that beneficiaries with lower cognitive skills find LIS benefits to be of less value relative to those with higher cognitive skills. Maybe the forgoing is evidence of revealed preference. This is an argument based on a market model in which consumers are reasonably, if not perfectly, well-informed about their choices. Does it seem likely to you that people with low cognitive skills are as well informed as their otherwise equivalent high cognitive skill counterparts?

    Meanwhile, we have to acknowledge that only 25.5% of high cognitive skill beneficiaries eligible for the LIS enroll in it, which is just under half of those who are aware of it. Almost two-thirds of them enroll in Part D. A large proportion of beneficiaries are forgoing support for a benefit for which they are eligible and for which they are enrolled. Why? Is this revealed preference for, effectively, a lower income? Would a well-informed consumer behave that way? And these are respondents who scored high in cognition!


    • “I’ve seen it suggested in the comments on this blog and elsewhere that low take-up rates among those eligible for Medicaid suggests they do not find the program to be of value. I have not seen any strong evidence to support that view,”

      Don’t you think the fact that so many that won the lottery and didn’t sign up for Medicaid strongly “suggests” (not proves) “they do not find the program to be of value.”? Isn’t that at least in part because we already have a type of universal insurance that makes it mandatory for hospitals to treat any severe illness without a wallet biopsy? Isn’t that also because many have enough money to pay a doctor privately for lower level things? Maybe they recognize some of the problems encountered when one uses Medicaid. Do not cultural values enter the picture? Perhaps some cultures don’t seem to understand the need for insurance or at least first dollar insurance. Maybe they see a benefit in paying small costs with cash.

      I am not sure how your present discussion alters any of the factors listed above or what you are suggesting should be done based upon the study you mention. The lottery in question contained a broad base of people.

      • How does one distinguish between multiple hypotheses that could explain the same information? “Do I not think …? Isn’t that at least in part …? Isn’t that also because …? Do not cultural values enter …?” All good questions. On what basis ought we answer them?

        • ” On what basis ought we answer them?”

          That is based upon what you are trying to prove. In my response I wasn’t trying to prove anything. Rather I was suggesting that the study strongly suggested that many didn’t find Medicaid of great enough value to sign up due to present day circumstances.

          I believe there are many reasons for that and therefore the ‘one size shoe fits all’ approach to Medicaid needs to be dramatically altered. My objective is to get better value for the dollar and at the same time make sure that those truly in need are getting what they need. At the same time I also wish to prevent the Medicaid program from providing unintended consequences that negatively impact these same people.

          • But there’s a question here: do people not enroll because the program doesn’t offer them value, or do they not enroll because they don’t understand the value of the program? These need not be mutually exclusive, but we can recognize the difference, right? That difference suggests very different policy remedies with very different implications, costs, etc. For that reason, were it my job to design, evaluate, or vote on policy, I’d like to know more about it. It’s not about proving it’s one way or another. It’s about understanding what is.

            • “But there’s a question here: do people not enroll because the program doesn’t offer them value, or do they not enroll because they don’t understand the value of the program?”

              We have had almost 50 years to figure that out. How much more time do we need? Add to that the studies where health benefits are not adequately demonstrated for the tremendous costs and we have a problem in defending Medicaid as it exists today.

              Are you suggesting that all those lottery winners that didn’t sign up for Medicaid are stupid or even have the same reasons for not signing up? Alternatively do you believe bureaucrats are more competent than all those people and should run their lives? What is the endpoint for that type of paternalistic involvement?

              “I’d like to know more about it. ”

              I’d like to know more about a lot of things, but since money is finite we have to be careful in how we spend it.

            • Good grief. You’re awfully evasive for someone who desired openness to multiple points of view.

              Can you list for me all the reasons why a lottery winner might not enroll in Medicaid? It is longer than “they don’t find the program of value.” What are the other reasons? Do you know them? They’re in the papers published by the study investigators, even ungated ones. You’ve read those, right, so you can tell me the answer. I’ll wait.

            • My initial statement was: “Don’t you think the fact that so many that won the lottery and didn’t sign up for Medicaid strongly “suggests” (not proves) “they do not find the program to be of value.”?”

              Whose being evasive? I made my points clear in the many postings I’ve written. We have gone from examining the mattress to examining the color of the ticking. First the debate is about whether there were health benefits from Medicaid and then why people didn’t sign up for Medicaid. Since the benefits were not shown in the Oregon study we are now discussing the reasons people didn’t sign up once they won the lottery.

              If the lottery was for $1Million once they signed up I am sure they would have done so just like the many lottery winners have done in other lotteries. It wasn’t $1Million dollars so the desire to sign up was less and a huge number didn’t sign up. It wasn’t a few. It was a lot. They didn’t perceive the value we seem to perceive and maybe they are right.

              Now you wish me to list the possible reasons or there is a “Good grief” comment and an accusation that I am being evasive. There are countless of reasons and you have read some of them in the study. All can be contributing factors, but that doesn’t mean that after 50 years of failure we have to spend another 50 years counting them. I brought up some of the reasons I thought important earlier that can have a significant impact but you want more and more when you could just as easily tell us what reason(s) you find more important than the suggestion that “they do not find the program to be of value”. What is the point of this exercise? We could go so far as saying that maybe someone was dying and an individual(s) got distracted. That could be one of the reasons, but I’ll bet that a $1million lottery winner wouldn’t have been quite as distracted. Maybe most are not stupid, but some are. That is another reason, but if they are stupid we better help them with their other affairs and if the stupidity is great enough they probably already have a social worker on the case to fill out the forms.

              Austin, I’m not trying to be difficult here. I want to be practical. I don’t think your desire for a list is important unless you have something significant in mind. We have spent Billions/Trillions on a program that doesn’t seem to work for almost 50 years and when people win a lottery giving them the program they don’t sign up. This secondary issue is really not that important even though other rationals may exist as documented in my original statement (suggests not proves) . Doesn’t all this failure tell you that we need to change the program a bit? That has been my consistent theme along with saying that if we waste too much money on garbage there won’t be any left to do good things.

              What happened to the concept of trade offs?

            • “What is the point of this exercise?”

              To engage new ideas rather than continuing to push the same ones we’ve already tread. I’ve read your prior comments. I found them stimulating enough to try to expand the conversation. Sorry!

              Anyway, reasons one might not sign up for Medicaid despite winning the lottery include (but are not limited to): income >100% FPL, moved out of state, acquired other coverage, aged out of eligibility, did not receive or failed to notice invitation to apply, was unable to do so within 45 days due to other life circumstances (including, yes, cognitive function), correctly or incorrectly judged the value to be too low despite having gone to the trouble of entering the lottery.

              I have no idea to what extent each of these possible factors (and others) played a role. Isn’t it reasonable to expect that some people were affected by each of these? I, for one, would love to know the breakdown. I’m interested in research relevant to this question, which is the entire point of this blog. I do not conclude that the final possibility in the list is or suggests the full answer. Nor do I think any of these imply precisely how Medicaid ought to be reformed.

              If you have any studies in mind that shed light on this area, please suggest them. No, I do not agree that the answer to why people didn’t enroll in Medicaid is settled just because we waste a lot of money on health care.

            • Austin, I would have seen the point if you discussed any of the reasons why you thought they were strong contenders to “they do not find the program to be of value” not because I am interested in the exercise of listing alternative explanations rather that discussion could lead to an improvement in Medicaid itself.

              Empirically speaking Medicaid should have health care value. Maybe that is wrong, but that is the question that remains in my head. Are the actions of government and the abundance of money diluting the positive effects of Medicaid so much that we cannot see them? This does not mean I support or don’t support such a program rather my mind is open to all sorts of ways to provide the needed care at a reasonable cost. That is the goal

              The reasons you provide are all good reasons, but some mean that the person should not have been enrolled in the Medicaid program in the first place. Perhaps if they applied they would have been rejected, but if they weren’t rejected that demonstrates a hole in the system and that would be of great interest. I don’t think any combination of these reasons would be more important than “they do not find the program to be of value”, but there is always that possibility. I do, however, worry about cognitive function problems.

              You say: “I’m interested in research relevant to this question” and so am I because that could lead us to the goal, but I don’t know that the answer lies in more government, more money. I think treating Medicaid in an isolated environment is causing some of the problems. I know and have discussed this problem with some people that fit the criteria for Medicaid. One of the answers we might be looking for was “why should I bother, I can always get the care when I need it” Add to that a disinterest that some have of integrating into what we consider normal society and we see a big problem. Though I don’t know how I can integrate the following into this conversation it swims in my head. Why are so many generations of the same family getting Medicaid. Have we created a new type of culture?

              As far as studies shedding light, I think we have to look outside of the pure healthcare environment looking at history and all sorts of disciplines. Why do people act the way they do where sometimes from our view it appears they are acting irrationally yet they are not mentally ill. I don’t think the best answers are in the health care policy field where too many answers seem to be clones of one another. A lot of my reasoning comes from totally different areas of thought that I integrate with whatever knowledge I have of health care.

            • “if you discussed any of the reasons why you thought they were strong contenders”

              The study investigators report that most of those who failed to enroll in Medicaid did so due to incomes above the threshold. Since it is a program for low-income people, this is by design. It is also very different from the suggested reason of rational assessment of low value.

              Happy to read any studies you suggest. That’s what we do here.

            • When you say ‘most’, are you talking about a plurality or a majority? Perhaps the exact numbers would be more enlightening if you have them. From my recollection of the article I don’t think it mentioned the breakdown or the explanation though it did say more information with regard to the details of the study was posted on the NEJM and NBER websites. My concern at the time was whether or not significant health benefits were found which didn’t happen so I didn’t pay that much attention to anything else.

              If almost all who failed to enroll did so because of incomes above the threshold then the point becomes moot. If however a significant number didn’t enroll for other undefined reasons or reasons that would have been surmounted if a significant desire existed then the question of whether or not they found significant value becomes an issue, maybe more important than the credit we are giving. People do all sorts of things to obtain something they feel is valuable.

              Yes, I recognize you are here for studies and a lot of other things as well rom what I have seen, but with regard to this question I doubt you are going to find something satisfactory and fits your needs.

            • The breakdown you seek is precisely the question I raised many comments ago. This was my entire point. Let us not presume the answer. Let us seek it. When time permits, I will try to find a breakdown in their published work. Failing that, one could email the authors. If you do so, please let me know what you find.

            • Of course, if the discussion point is whether or not those that didn’t sign up saw value in the program one would want to know the breakdown. It was an unexpected secondary feature of the article that bodes poorly (until explained away) for those that assume Medicaid is doing the job intended.

              From your earlier statements it seemed you had that information and it was easily available from the study.

              “Do you know them? They’re in the papers published by the study investigators, even ungated ones. You’ve read those, right, so you can tell me the answer. I’ll wait.”

              Now I find that I was not crazy and they are not in the study and you do not have them either. I guess the question remains open as to whether or not the Medicaid group finds the program valuable enough to put away other reasons and sign up when the program is offered. That is not a good sign.

    • Austin has already acknowledged that the study above represents people age 65+ enrolling into Medicare, whereas with the Oregon Medicaid experiment, we’re concerned over different types of cognitive limitations. But just to lay it out there explicitly:

      In seniors, we’re concerned about cognitive decline related to dementia. People forget stuff. Their ability to manage money and conduct other types of planning declines.They probably rely more on their kids to help them.

      In the population enrolling to Oregon Medicaid, which is uninsured adults under age 65, we are concerned about lower education, perhaps mental illness, perhaps being a highly mobile population. There will be some folks with severe mental illness, but I would imagine not a lot of them (as you can often quality for Medicaid and/or Medicare on the basis of SMI). The types of cognitive limitations in this population are probably quantitatively less severe than those among seniors with dementia. They may also be qualitatively different. However, dementia lowers executive function (iirc), and mental illness probably also does, and education may also imply lower executive function.

      Two side notes. The Health and Retirement Study, which was the above study’s data source, includes only seniors. Medicare covers people under age 65, and at least some of those folks would have developmental disabilities or other cognitive disabilities, but the HRS doesn’t cover these individuals. Also, not all HRS respondents actually take the cognitive test above. Those who do not respond to the cognitive test have a very high prevalence of cognitive impairment. I know this from working with the HRS and CI. Either way, the study does likely exclude many seniors with the lowest cognitive functioning.

      • I was thinking a bit more about what I said re cognitive function in seniors versus general education and executive function in younger adults. Among the people in the study, a lower level of education could make someone less likely to comprehend the value of insurance. Or, we are all vulnerable to hyperbolic discounting (where we discount the value of future benefits by much more than we discount the benefits of near-term or immediate benefits), but lower education might make you more vulnerable to hyperbolic discounting.

        Among the seniors, there is also that problem. However, the cognitive limitations measured by the HRS are different. They would be more along the lines of, Mrs. Chen decided to sign up for the LIS. But her memory is failing and she forgot when the deadline was. Or Mr. Rodriguez would have signed up for the LIS, but he isn’t able to manage his own finances because of cognitive limitations, and he couldn’t locate any of the necessary paperwork to fill in with the application.

        I think these are qualitatively different from the cognitive limitations that the Oregon Medicaid people would have experienced. So, while I accept the stated premise, I don’t think this is the research to prove it. It is a really good piece of research, though!

    • two things:

      1) the Medicare Part D LIS population probably had a lot of overlap with Medicaid – dually eligible patients are common.

      2) States have an incentive to make enrollment as annoying as possible. About 6 years back I was working for a Medicaid Managed Plan in Oregon and nobody reading this blog would believe the hassle that members had to go through to get and stay enrolled in Medicaid. Asset tests, original birth certificates, weekly re-applications. You have to be much more organized about paperwork if you want to be poor, apparently.