>someone who is very rich is much more likely to be a republican (even if they still only have a 50% chance) than someone who is very poor. This is correct, an

>someone who is very rich is much more likely to be a republican (even if they still only have a 50% chance) than someone who is very poor. This is correct, and I think you would also be correct to say political affiliation varies with income level, but that is still a far cry from income level being predictive of party identification. If you are interested in more accurate predictors of political affiliation, you may want to read Jonathan Rodden's excellent work on geography and political identification.[1] To be clear, this work addresses the populations of areas, not individuals. [1] http://www.stanford.edu/~jrodden/wp/shadow.pdf http://www.stanford.edu/~jrodden/wp/shadow.pdf