• Life Expectancy Wonkishness

    I’ve gotten a number of emails asking me why I keep stressing that we need to pay attention to life expectancy at age 65, when it comes to Medicare, instead of life expectancy at birth. It’s a fair question.

    People who want to raise the eligibility age for Medicare like to cite the increase in life expectancy at birth. That’s been going up fast, according to them, and, again according to them, Medicare was never intended to support people for so long.

    There are problems with this belief.

    Let’s take a theoretical cohort of 100 people. Let’s stipulate that the life expectancy at birth of this group is 74. In this cohort there is a baby who dies soon after birth, and there are also 25 people who live past the age of 65.

    Scenario A: I manage to find an excellent treatment for what kills the baby in the cohort. It’s not perfect, but it allows the baby to live to age 50.

    Scenario B: I manage to find a treatment for a common illness that afflicts elderly people, and can extend their life an average of 2 years.

    In both these scenarios, I have managed to add 50 years to the cohort. So the average life expectancy at birth has increased to 74.5 years. But this, hopefully, shows you the enormous power of saving one baby or child, in that saving one infant for fifty years is the same as saving 25 adults for only 2 years. The take home message here is that treating children well does far more to increase life expectancy at birth than treating adults for illnesses that kill them.

    And we’ve had remarkable improvements in the treatment of children and chronic diseases in the last few decades. Vaccines do wonders. We can treat many illnesses that used to kill children far into adulthood now. Many childhood cancers can be cured. Saving those children has resulted in relatively big increases in the overall life expectancy at birth.

    In Scenario A, I have not increased the life expectancy at 65 at all! In Scenario B, however, I have increased the life expectancy at 65 a full two years. This has enormous policy implications. The increase in Medicare spending in Scenario A is $0.00. The increase in Medicare spending in Scenario B is comparatively huge. We would need to provide two additional years to all twenty-five elderly, for fifty additional person-years of Medicare.

    Now, this is a fictional cohort. It explains, however, how saving a child can significantly increase life expectancy at birth, but result in little extra Medicare spending. What matters for Medicare spending is the years we have to provide it, which is exemplified by the life expectancy at age 65.

    And, as I’ve shown many times, life expectancy at age 65 is not going up nearly as fast as life expectancy at birth:

    Nor is it going up the same for rich and poor alike:

    In fact, it’s been pretty flat for the bottom half of earners. Remember this when someone claims that “people” are getting so many more years of Medicare than they used to.

    • Aaron a couple of points:

      a) I don’t know that I’ve seen you address this actual argument “… according to them, Medicare was never intended to support people for so long.” That is, what was the intent of the original Law?

      b) You say “life expectancy at age 65 is not going up nearly as fast as life expectancy at birth”, but your graphs shows the following increase:
      i) life expectancy at birth improvements ~ 78 yrs / 68 yrs = +14.7%
      ii) life exp. at age 65 = (84 – 65) / (79 – 65) = 19/14 = + 35.7%

      So really, given that I’m 65 in 2007, I can expect to live 36% more years than I would have in 1950, but if I were just born, I would only have 15% longer to live.

      This % increase may be as big, or bigger, of a concern than the actual absolute number of years left, which is what your graph demonstrates. For instance I have a pool of money designated to pay for a certain benefit (simple, hypothetical). If my benefit period is now 36% larger than I anticipated, that’s a bigger concern than if it is only 15% larger. Thoughts?

      • You’re using relative gains in percentages. What matters is the actual number of years. That’s what determines the increase in money, When people talk about the increases in life expectancy, they cite increases in years (using life expectancy at birth) not percentage gains. Find me a single politician who argues with the percentages you cite and I’ll look into it.

        As to your first point, the point of Medicare was to provide universal health insurance to the elderly. That hasn’t changed.

    • Aaron, thanks for your response. The years may be what people or politicians cite all the time, but I disagree that they are useful at all in what determines the increase in money. If I’m funding my child’s college education which I think is going to last 4 years and in reality they stay for 5, I can tell myself that this is 1 More Year, and talk about that with other people, but what really matters is that I need to save an additional 25% over what I had originally planned on saving.

      This blog discusses relative amounts all the time (medical cost trends, utilization trends, doctors per 100,000, NHE as % of GDP, etc) and for the purposes of discussing the economic impact of additional years of life on the funding of Medicare benefits (life expectancy at age 65) I think it is misleading to leave out the relative impact of an additional year beyond the eligibility age.

      • No. Of course we sometimes use relative percentages, because sometimes those are the relevant statistic. Just like sometimes survival rates are the most relevant, and sometimes they’re not.

        But in this case, when you’re talking about how much more Medicare spending is going to be, what matters is the person years above the age of 65. That’s all. And what tells you that is life expectancy at age 65. Everything else is smoke and mirrors.

    • I don’t think increasing the Medicare eligibility age is all that important right now, Medicare needs structural reform, not simply a onetime reduction in cost. This is not the answer to our Medicare problems, and it is silly that some people are presenting this as the next great idea to fix Medicare.

      But that being said, in general as life expectancy increases well beyond what it was when programs like Medicare and SS were designed, I do think it is worthwhile to increase the age. I don’t see the relevance of the split between income cohorts. Is it a program to provide universal health insurance to the elderly, or is it a welfare program for low-income elderly? We shouldn’t automatically reject a legitimate reform idea because the bottom half of earners have not seen life expectancy increase as much as everyone else. And even if that is a relevant thing to consider, the bottom half has seen an increase in lifespan since 1965. Index eligibility age to life expectancy at age 65, and means test it based on income. We need a lot more means testing in Medicare already, no reason why eligibility age shouldn’t be part of how we do that.

      But all of this is moot until we implement some meaningful reforms to Medicare first.

    • Do you have comparative international data for life expectancy at 65 with or without the rich/poor graphs? Your informative FAQ series has the historical life expectancy at birth by country; it would be a useful reminder here, even if not 100% relevant: http://theincidentaleconomist.com/wordpress/how-do-we-rate-the-quality-of-the-us-health-care-system-%e2%80%93-population-statistics/
      My questions: (1) are life expectancies at 65 on the same increasing curve elsewhere and (2) is the rich/poor divergence a U.S. phenomenon or is it more widespread, representing more physical hardships in life (even if health care is equally accessible in those other countries).

    • “And what tells you that is life expectancy at age 65. Everything else is smoke and mirrors.”

      Not quite. You also have to consider the number of individuals reaching that important age milestone.

      (person years takes into consideration both the number of people reaching 65 and how long they live thereafter.)

      this does not change you core argument that increasing the eligability age is a bad idea for obvious reasons.

      • You are correct, as an indicator about how much Medicare will cost as a whole. But to look at how much it has increased per person (ie is each person getting more benefit), I stand by life expectancy at age 65.