Tuesday, May 12, 2009

Life expectancy

I still see a lot of people trotting out life expectancies statistics as a proxy for how inferior U.S. health care is compared to other countries with a larger government role. I don't find such comparisons terribly useful owing to different demographics and behavior patterns. As an example let's compare Norway with North Dakota, the U.S. state with the greatest percentage of Norwegian-Americans:
Life expectancy in Norway: 79.78 years
Life expectancy in North Dakota: 79.8
A difference of .02 years -- in favor of North Dakota. Would Norway then be well advised to model its health care system after this U.S. state?

1 comment:

Paradigm Shifter said...

I have always found the comparison of means in the absence of other sample population characteristics to justify sweeping costly policies to be disigenuous. Analysts within the government already capture the point data of age for each death. It is so very simple to perform a statistical test, like I did earlier on unemployment rates, to calculate the mean AND spread of data. Doing so takes a few extra clicks of a mouse button, and allows for far more powerful statistical tests to determine whether a statistically significant difference exists. The same types of tests could show whether the Dakotas show a significant difference compared to the rest of the country.

These simple analyses should be the minimum expectation of policy makers when they debate trillion dollar health care plans.