• Subtleties of life expectancy

    There have been a number of interesting posts today (including here) on the subject of life expectancy and its effect on entitlement programs for the elderly.

    There are a number of people out there who believe that we need to increase the age at which people start getting benefits because people are living so much longer than they used to.  They cite data like those below, which show life expectancy for all individuals in the US, from 1950 to 2007:


    When you look at the data like this, it seems only sensible, after all. According to this, almost no one could expect to live much past 65 in 1950. Moreover, there’s been a steep climb for the past few decades. If you take this at face value, then it seems only natural that we would have to increase the age at which these programs kick in.

    This chart is a bit deceptive, however. See, whenever a child dies, it skews life expectancy way down. Anything that increases infant mortality, or care that ends a potentially devastating childhood illness (think vaccines), will increase life expectancy a lot. What we really care about is not life expectancy at birth, but life expectancy at age 65. In other words, if you make it to 65, how long will you be on one of these programs? Let’s add that line in:

    Two things to note here. First, if you made it to 65, even back in 1950, you could expect to be on Social Security for 14 years. In 1970, if you made it to 65 and qualified for Medicare, you could expect to live for about 15 years on the program. So a lot of people were making use of these programs, for a lot of years. The second thing to notice is that life expectancy for someone who lives to 65 and qualifies for these programs, hasn’t gone up as much, or as quickly, as people think.

    The reason that Social Security has become more costly is not nearly as much that people are living longer on the program, as it is that many more people were born into the generation approaching 65. They aren’t getting more benefit individually; as a group there’s just more of them. When you argue that you want to raise the age at which they start to 68, instead of 65, you’re basicly giving them as many years on the program as a person who hit 65 in the mid 1970’s. That’s a pretty big change.

    One more point. Life expectancy at age 65, which is really how many years you can hope to receive Medicare or Social Security, isn’t the same for everyone. Here it is by race:

    So when you propose to raise the age at which it starts to 68, you’re depriving the average black American (fewer of whom make it to 65 already) more of his or her benefit, compared to the average white American.

    Something to think about.

    • I don’t understand how you can say they’re not getting more benefit individually when the chart clearly shows there has been a 25% increase in the number of years receiving social security. In 1970 you could expect about 15 years, today about 19. The shift in demographics is the bigger problem, but that doesn’t mean we should ignore the increase in lifespan.

    • On thing that the charts show is how ineffective Medicare expenditures are. It would have been much more efficient/sensible to give free healthcare to mothers and babies, but politicians know who votes.

    • Just read the huffpo piece, they’re playing coy with the numbers, taking advantage of the absolute vs percentage change slight of hand. “It’s only three more years!”, which sounds so small until you consider the percentage change it represents.

    • I’m guessing that a plot of life expectancy vs. income would show an even bigger difference than the one plotting as a function of race. I’d like to see the data, anyway.

    • Doesn’t the increased percentage of people who make it to 65 in the first place also affect the cost of SS – i.e., further swell the baby boom into a geezer boom relative to earlier eras?

      • They also pay taxes all the way until age 65. That helps those who are older.

        • Is that a help for all, or most workers? Or only workers above a certain income threshold? If I average, say, $35k per year for 35 years, then live to be 85, will I put in more than I take out? Is there an average earnings break point for those reaching average life expectancy, beyond which they take out more than they put in? To what extent do the wealthy fund the benefits of the majority? Considering that you only get back 15% of average lifetime earnings over $64k/year as benefits, I imagine it may be to a considerable extent.

    • What happened around 1976? It looks like black life expectancy was tracking white pretty closely up until then.

    • Forexpostfactoid: I think the point implied in the article but not said explicitly,is that lower infant mortality in 1950 pulled down life expectancy, but that if you lived to, say, 5 or 8, you had pretty close to the same expectation of living to 65 as someone today; with the much lower rates of child birth figured in, there aren’t that many more people, percentage wise, in a population living to be 65 or 77 now versus then, but a greater aggregate number cause the baby boomers were such a large slice of demographics…. I think another way of figuring the infant mortality rates lowering is that back then, a bigger slice of humans born never really got into the SocSec. pool, demographically, or practically speaking. The working pool, so to speak, which the dead-as-infants never got into anyway…..
      I think that makes sense…..of course, abortion numbers probably have to be accounted for in this whole discussion, as to how much the increase in abortion has affected the relative size of the future worker pool, and, ergo, the SocSec pool, since 1973…..
      It seems fair to say there would be, what, 8 million to maybe 20 million more in the worker pool paying into SocSec now if abortion hadn’t been legalized? Maybe symmetrically speaking, about exactly the amount of illegal immigrants working now?? (wouldn’t that be weird?)
      and then, of course,, it would seem, about the corresponding greater number hitting 65, 68 then 77 and beyond, whenever, to collect SocSec…
      I can just imagine what would come out of Alan Simpson’s sailor-style mouth over the above items being added to the discussion……but it seems like real data to me.

      I like the idea of looking at the average years used by SocSec. recipients;
      the factor suggested above,too, of what’s the ratio of contributions versus “take-outs,” and has it changed and does it vary much among various demographic groups.
      And why not add in some rich-people premiums: If you earn more than $200,000, you pay so much more, and each $150,000 level above that, another notch added… Even a 0.25 percent increase in contributions from someone earning $500,000 or $14 million, say, would pay their way pretty dang good… 0.5 percent more from a million-a-year earner would be another $5,000 a year into the SocSec pot, and never missed by the millionaire.

    • This very “subtlety” is also conveniently ignored for life expectancy statistics from WHO and the effect of how different countries count and record live births.

    • A reasonable policy would be to block allocate a fixed number of dollars, then apportion it among all SS recipients. That would auto-correct the problem of a larger cohort, while also giving older voters an incentive to raise the retirement age, since that would shrink the cohort receiving benefits thus increasing their share of the pie.