• Shrinking private LTC insurance options

    Prudential announced this week that they will stop taking applications for private Long Term Care insurance (LTCI) at the end of March according to a WSJ story by Leslie Scism. Several other insurers have also exited this market in recent years, which means that private insurance options are shrinking as the baby boomers move toward needing LTC.

    Medicaid exists as a de facto nursing home insurance plan with the deductible essentially being your wealth (Medicaid pays for ~40% of all NH costs in the U.S.). Before that, families provide a tremendous amount of informal long term care, which is both expected by many while also being very burdensome and costly, both in explicit financial terms as well as in other ways. That is the default system.

    Prudential notes two primary reasons for their exit:

    • claims being higher than predicted
    • interest rates being very low

    The first reason they put down to “increasing life expectancy”, but I am not sure I buy that. Life expectancy has been increasing for quite a while. The more likely culprit for higher claims is adverse selection, which just means the people signing up had higher risks than average, which you might expect to be particularly bad for a type of insurance that is so rare (less than 10% of those over 50 have any). Interest rates may seem unrelated to LTC, but effect the return on investment that insurance companies can easily obtain with the premiums for a type of insurance in which persons may pay in for many years with no claims before having large claims later. Prudential says they will honor existing contracts, but premium increases for all members are likely in spite of such policies typically being sold with flat premiums (State regulators have consistently approved such increases, believing default of insurers to be the only other option).

    In short, it doesn’t appear that purely private LTCI markets can work, even with tax credit purchasing incentives that we have had for years. In the aftermath of the CLASS program demise, some noted the biggest problem within CLASS was the inability to assign an actuarially fair premium, which is important in a purely private insurance market, and allowing for simple underwriting could improve CLASS. However, the Prudential story and the move out of this industry by other private insurers shows they haven’t been able to assign actuarially fair premiums either.

    Far more important than focusing on how premiums are set for a small pool is forced risk pooling of some sort, to get all those at risk of needing LTC into the pool. In one sense, if you are OK with families providing care and Medicaid picking up much of the nation’s NH bill then you could argue we have a system. If you don’t like this system, it is amply clear that a purely private insurance-based one won’t work, and some sort of forced risk pooling will be required. The question remains, how will we insure LTC?

    DT

    Share
    Comments closed
     
    • I see this also as a behavioral phenomena. Wouldn’t it just be simpler to institute a forced savings system in general rather than trying to figure out who may be using LTC or not? DeLong has proposed this (if I recall correctly) and it would seem to me a good idea, particularly since a good deal of our lifetime medical expenses are known in advance – we just don’t know exactly how we might spend them.

    • @wintercow20
      yes, it would be simpler just to build LTC into Social Insurance via Medicare and/or Social Security. About 7 in 10 attaining age 65 will use some, about half of the ‘users’ will do so for ~8 months or less, but about 1 in 10 of them for more than 5 years. Yet 3 in 10 have 0 use. That risk profile cries out for social insurance.

    • Dear Don –

      Here’s a little quick back-of-the-envelope calculation. Let’s say we have 10 people, 3 of whom end up in LTC for 8 months (= 2/3 year), 3 who end up in LTC for 16 months (=4/3 months), 1 who has 5 years of LTC,, and 3 who have no LTC. These numbers are undoubtedly not representative of the real world, but they are consistent with the basic statistics that you have reported, and this calculation is supposed to be a quick approximation. That comes out to an average of 1.1 years of LTC for each person, and the denominator includes those who didn’t actually use LTC.

      The big question is: what is the average yearly cost of LTC? A nursing home costs about $50,000 per year, while assisted living can cost roughly $25,000. Presumably the people who use LTC for a short period of time are likely to be in nursing homes, while those who use it for longer periods of times might not be in a nursing home the whole time. Let’s just use the number of $25,000 per year, and remember that it’s possibly a low estimate of the real average yearly cost of LTC. So, 1.1 years of LTC at $25k / year comes out to about $28,000 as the expected benefit per person.

      Now the average Social Security benefit is $1,200 per month, or about $14,000 per year. So this LTC benefit is equivalent to 2 years of additional Social Security checks (and remember that this might be a low-ball estimate). Considering that many plans for SSI solvency entail *raising* the eligibility age by a year, creating a benefit plan that effectively *lowers* the eligibility age by two years would seem to be very, very difficult.

      Admittedly this is just a quick, approximate analysis, so I would put much faith into the specific estimates. But it does suggest that funding LTC through SSI or Medicare might be very expensive.

    • I’d disagree one one point: I do not think this is about adverse selection against Prudential. Adverse selection is when one group of insureds is riskier than another group because the former had prior information about how risky they were. For example, if a lot of people knew they would probably end up in nursing homes, they’d be more likely to buy LTCI.

      The thing is that just about all LTCI carriers have complained that interest rates have been low and morbidity has been worse than expected (that’s how many people get disabled and how long they’re disabled). That’s been exacerbated by the fact that many insurers used to offer lifetime benefits.

      Again, I don’t think that Prudential in particular experienced adverse selection. While most research on this area is going to be proprietary to the industry, some earlier work by Finkelstein and McGarry does not find adverse selection against the LTCI industry in general. They used the Health and Retirement Study, and found that individuals with LTCI are actually less likely to use nursing home care than individuals without.

      http://emlab.berkeley.edu/users/burch/e231_sp03/Finkelstein.pdf

      So, if there’s adverse selection, it must be against the entire LTCI industry. I think it would be hard to tell that, as we don’t have good data on the claim experience of equivalent non-insureds – precisely because they probably get family members to care for them and they’re not on Medicaid. There’s one paper which finds that people who are at risk for Huntingtons based on genetic tests are far more likely to have LTCI. But it’s not clear to what extent people use such tests.

      http://www.nber.org/papers/w15326.pdf

      I think that most likely the industry in general under-predicted mortality because it’s difficult to make very long-range predictions, and the research on disability is limited. If we covered long-term care under Medicare, the CMS actuaries would have the same problem. In addition, it could also be that the industry is simply consolidating. LTCI products aren’t very differentiated, at their core. That should mean that the biggest insurers will gradually crowd out everyone else.

      I don’t mean to sound like a shill for the industry. I think that while the product they have probably works as well as we can make it, it will most likely be a market that’s limited to wealthier Americans. And for the record I’m still a single payer, LTC included guy at heart.

    • Jonathan Pond, Financial Planner, says that 90% of estates are spent this way: 1) nursing home, 2) IRS, 3) children, 4) grandchildren, 5) charity.

      More people are worried about the IRS taking their money than about having to spend it on a nursing home.

      This tells me that most people do not understand what may happen to them and their families if they need care. You cannot control when you will need care but you can control how you pay for it by planning for it.