The Washington post featured a story about life expectancy for women and men by county. The data come from the University of Washington's Institute for Health Metrics and Evaluation, and in particular a paper by Sandeep C Kulkarni, Alison Levin-Rector, Majid Ezzati and Christopher JL Murray.
The Post story contained a fun interactive map, and I could not help but notice that "blue" places appeared to have longer lived people. I couldn't help myself, so I downloaded data on elections from Mark Newman's website at the University of Michigan. Specifically, I downloaded data on the 2008 presidential vote by county.
I then drew a scatterplot in Stata of Obama's share of the Obama-McCain vote against women's life expectancy:
Again, the unit of observation is county. The size of the dots reflects total votes by county. The vote weighted correlation between women's life expectancy and share of Obama vote is .262, which is not insubstantial for a bivariate correlation.
One cannot draw any causal inference from any of this, of course, and my guess is that the Obama vote share is proxying for something else--perhaps educational attainment. But drawing these pictures is fun (BTW, the correlation for Obama vote and men's life expectancy is a smaller .168).
The Post story contained a fun interactive map, and I could not help but notice that "blue" places appeared to have longer lived people. I couldn't help myself, so I downloaded data on elections from Mark Newman's website at the University of Michigan. Specifically, I downloaded data on the 2008 presidential vote by county.
I then drew a scatterplot in Stata of Obama's share of the Obama-McCain vote against women's life expectancy:
Again, the unit of observation is county. The size of the dots reflects total votes by county. The vote weighted correlation between women's life expectancy and share of Obama vote is .262, which is not insubstantial for a bivariate correlation.
One cannot draw any causal inference from any of this, of course, and my guess is that the Obama vote share is proxying for something else--perhaps educational attainment. But drawing these pictures is fun (BTW, the correlation for Obama vote and men's life expectancy is a smaller .168).
No comments:
Post a Comment