[Disclaimer: I should start this one off with a warning. It's long. And kind of ranty. I haven't slept well in days and it's relatively late here. Sorry.]

With all this crazy news about ObamaCare, a lot of facts (and “facts”) are being tossed around. Often, people will use big fancy health care words like “life expectancy” and while we all kind-of-sorta know what life expectancy is, I’m going to really get into it for you. Why? Because life expectancy is one of those things that seems to get twisted and used in all sorts of ways it was never intended.

So, we’ll start off with the basics. What is life expectancy? Life expectancy^[1] is defined as the the average (expected) number of years of life remaining at any given age.

When people say “life expectancy in _________” they are referring to the life expectancy of that country at birth, but you can define life expectancy in other ways as well. For example, life expectancy at age 1 (or at age 5) is often used in countries with very high levels of infant mortality. Why? Well, I have to get all mathy for this so I’m going to turn it into a footnote. Hover over it if you’re really interested.^[2]

So, we’ve defined it as an average. As with any average, this creates helpful comparisons between countries/cities/regions since it is independent of population size. Also with any average, it doesn’t really define the population. The life expectancy in the US just went up to it’s highest point in history at^[3] 77.9 years. That doesn’t mean you’re going to die exactly .1 years before your 78th birthday. In fact, it doesn’t even mean most people die right before their 78th birthday. It means some people died above 77.9, some below 77.9 and maybe a few at 77.9, but when you added them up and divided by the sample size, you get 77.9^[4]

So here comes the fun part. The United States ranked horribly when it comes to this. Horribly. We are ranked 35th–this puts us at slightly higher than Costa Rica and Cuba^[5] and well, well below Japan (82) or Australia (81.6) and even *gasp* Canada (81.2)! Essentially, this puts us below any industrialized country.^[6]

And that is often what gets touted. We have the lowest life expectancy of any industrialized^[7] country in the world.

This is, of course, true; however, it is also slightly misleading. You see, life expectancy is not the most robust measure of a country’s health care system. Cultural, environmental, geographical, social, structural, or legal differences can all affect this measure and health systems have a hard time countering that. It’s a great crude measure, but horrible for direct comparisons.

It’s this exact weakness that has conservatives pouncing all over the Obama plan. Specifically, there is a presentation^[8][9]–too often (mis-)cited by others–which claims that if we removed all violent crime and automobile accident deaths, we would actually be number 1 in the world in terms of life expectancy.

This is simply not correct. If any of those blogs that cited this presentation actually bothered to read the article, they would find that the regression the authors used removed automobile accidents and violent crimes and the US moved up to 17th. Then they “standardized” the US with the rest of the world by factoring in GDP. Given that the US a huge GDP compared to any other country in the world, this bumped us up to number 1 on the list.^[10]

So the idea is if we removed areas where America has a uniquely higher rate–driving accidents and murder–we’re not as bad as the rest of the world and this is a better reflection of the health care system. Now, I’ve got issues with that already, but my first question is why? Why are we suddenly developing this new metric? We already have a metric designed to reflect a good health care system–it’s called amenable mortality. Amenable mortality is the mortality rate of a country that could have been (theoretically) avoided with timely and proper medical attention.^[11]

So there you have it. Life expectancy is the tool being used in this war of ideologies and it’s about the dullest one in the box. Don’t be tricked by this “if you remove violent crime and automobile accidents” business. Complete garbage. Look for articles that cite something useful like amenable mortality, infant mortality^[12], maternal mortality, measures involving access to care, etc.

1. in the context of human health–since it also applies to animals or manufactured goods or whatever [go back]
2. This is because life expectancy is the mean of a population and as with any mean, it is very sensitive to extreme ranges–the further away from the expected value you go, the more leverage a single point will have. That is, if your mean is 50 in a population of 10, a single person dying at 1/2 years or living to 150 will largely impact your mean. Because humans more often die at young ages than live to extremely old ages, life expectancy is especially sensitive to infant mortality. Hey, that actually wasn’t very mathy at all. Good job, me. [go back]
3. a relatively unimpressive for a developed country [go back]
4. If you still don’t get this and need a mental exercise, imagine a coin and value heads as 1 and tails as 2. If you flipped that coin a million times and then averaged out the values, you would get 1.5. Yet, you never flipped 1.5, did you? [go back]
5. Granted, Cuba is a very special case in terms of public health, GDP, and life expectancy. [go back]
6. I’m sure I could find more recent numbers, but this isn’t a term paper–it’s a blog. Either use my 2006 UN data or do your own research. [go back]
7. read: has socialized health care [go back]
8. See it here (Warning: PDF). [go back]
9. based on this article–again, PDF [go back]
10. Read the “note” here for a more articulate response. [go back]
11. On that scale, the US is still ranked pretty low–17th or 19th if I remember correctly. I don’t have my textbook with me. [go back]
12. There are some issues with this measure as well, but it’s one of the better ones in our limited toolbox. [go back]