The Life Expectancy Myth

It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so.  --- Mark Twain

One of the most often cited reasons in support of socialized healthcare is the fact that the United States spends more per person on medical care than any other nation, yet our life expectancies are lower.

That statement is actually true. However, as Robert Ohsfeldt pointed out in his book The Business of Health, the reason it is true is because Americans have a very disproportionate tendency to kill themselves. If you exclude homicides and auto accidents from the data, US life expectancy is #1 in the world. While our health care system is doing a poor job of preventing shootouts and drunk driving, it is doing a good job of healing the sick.

One of the exercises we had in my first economics class was to write a paper using statistics to "prove" something that was demonstrably false. It tuned out to be an easy assignment (I think I proved that church attendance caused crime rates to increase). In this case, it turns out (not all that surprisingly) that there are many things that affect mortality besides health care. Always look behind the statistics.

Household Income Has Fallen. People's Income May Not Have

This chart appeared in an article in the Wall Street Journal today. What it shows is actually true. The implications that the WSJ writer derives from it are not. Those not statistically literate (which is to say 95% of the population and 99% of journalists) would look at this chart and conclude that people are earning less now than in 2000.



A statistician would look at this and ask the question, "I wonder if the composition of households has changed since 2000?" It has.




The number of people per household who are actually earning an income has dropped quite a bit since 2000. Here's a simple example. Suppose you live in a neighborhood of 20 households. The median income per household is $70,000. The husband and wife in one household separate. Now there are 21 households in that neighborhood that, collectively, are earning exactly the same amount of money as they were before. The median income of that neighborhood has dropped, although nothing else as changed,.

It's important (even for journalists) that in examining data that you at least ask the question, "what else might be going on here?" before you leap to a conclusion.

Worldwide IQ in Decline

This chart was derived from a study at the University of Hartford (Connecticut). While people have tired desperately to come up with all sorts of hypotheses related to pollution and global warming (yes, global warming), the answer is quite simple to anyone who has taken even an introductory course in statistics.

Fertility rates are inversely proportional to intelligence. That is, the more intelligent you and your spouse are, the fewer children you are likely to have. Rudimentary statistics tells you that the average intelligence of the entire population will fall.




Now, compounding this are two additional factors:

1. One of the balances to the fertility rate problems has been the Darwinian one -- low IQ is not a survival characteristic. That is, peoples with low IQ tend to have lower life expectancies. Therefore they reproduced at lower rates. Modern science and technology has intervened to reduce the effects of life-threatening disease -- from malaria to Ebola.

2. As Charles Murray pointed out in Coming Apart, the tendencies for like to marry like are increasing. Since a child’s IQ is highly correlated to that of his parents, this means that instead of reverting to the mean, IQ is becoming more bi-modal – that is, greater concentrations at high levels (low fertility)  and low levels (high fertility).

All of this suggests strongly that we should expect the decline of the last 50 years to continue. Not a particularly positive trend in light of the fact that technology is also rapidly eliminating the sorts of jobs that people with lower intelligence can do. 

Correlation Does Not Equal Causation

Economists frequently issue the caution that correlation does not necessarily imply causation. Chris Blattman at Columbia University came up with this succinct reminder. Even if Chrome and Firefox are more "peaceful" browsers.