Looking at Japan and the US, which are respectively the countries with the lowest and highest obesity rates, and also noticing that Japan had nearly the highest lifespan of all nations in the world where the US has a non-impressive value although we spend the most on health care per nation, it seemed worth studying the relationship between these various values.
Firstly, here is a plot of the amount spent per individual versus the average lifespan of that nation. As you can see, there’s quite a bit of a diminishing return over $1000 per year. There is a rather obvious relationship though.
Now, here is a comparison of the percentage of the population that is obese versus the average lifespan. The relationship here is much less strong. Japan is in the upper left, the US on the far right.
Now, theoretically, if the percentage of obese people in the nation has a particular link to the average lifespan removing that difference should make the graph of health spending to lifespan a smoother curve. It should reduce overall noise.
My first guesstimate (by eyeballing the graph) was that there’s something like an average of a 0.3 year loss of lifespan per 1 percentage of obesity, but on applying this correction to the spending graph, the results weren’t pretty. Figuring that the spend-light countries were messing me up, I then tried drawing a line straight from Japan to the US, which is about a 0.167 year loss of life. Again, applying this to the spending graph, the results were more jumbled than before. Looking at the original spending graph, I figured that if the US has “average” health care based on where it should be based on spending, it would be about 2 years higher. To achieve that, I’d need to apply a 0.067 year loss of life due to obesity. The result of that transform is below:
Overall, this does seem to make there be fewer outliers. Looking at the obesity percentage graph, the 0.067 rate of decline seems to be a line which follows the top of the scatter. Perhaps that makes sense?
Well so, assuming that we should trust that–that the “corrected” graph is more accurate than it was pre-correction–then overall we could expect up to a two year increase in overall national lifespan by reducing obesity. And either way it seems likely that spending more than ~$1500 per year per person on health care is mostly wasted.
A food tax (for instance, one which was based on calories per volume) would likely be profitable for the government, and would probably be the most surefire way to reduce the national rate of obesity.
http://www.nationmaster.com/graph/hea_obe-health-obesity
http://www.nationmaster.com/graph/hea_spe_per_per-health-spending-per-person