That article made me wince because it was apparently written by someone who had never heard of Simpson's Paradox[1]. Just because Texas as a whole tends to sup

That article made me wince because it was apparently written by someone who had never heard of Simpson's Paradox[1]. Just because Texas as a whole tends to support Romney, and just because Texas as a whole has a lot of people who pay no income tax does not give you the data needed to conclude that people who do not pay income tax tend to support Romney, or even that they are more likely to do so. And in fact, the exact opposite is true. It's been a staple of American political analysis for decades that although poor states tend to vote Republican, poor people tend vote Democratic. If you take Mississippi, you will find that it is poor and thus (as expected) it votes Republican. But within Mississippi the poorest groups overwhelmingly vote Democratic (regardless of skin colour, in case you think racism plays a role), while the richest groups vote Republican. Look at a rich state and the same trends hold - the poorest groups within that state will vote Democratic, while the richest will vote Republican. (A ton of research exists on this; this blog post seems to sumarrise it well[2].) To quote from the article: "So Romney appears to be wrong about these voters. But if calling for a less progressive taxation system was enough to alienate poor voters in the Deep South, the Republican Party would have already lost its stronghold there." This is something that could only be written by someone who has no grasp of American politics. You only have to glance at an exit poll - any exit poll - to realize that Republicans alienated poor voters LOOOONG ago. Let's say 2004, presidential election, Mississippi again[3]: Kerry beat Bush by 24 points among people making less than $15k a year, but by only 9 points among people making $15k - $30k. Look at people making $30k - $50k and he lost by 20 points, a massive 36 points among people making $50k - $75k, and an eye-popping 43% (71% to 28%) among people making over $75k. And, again, the same trends exert themselves in every state, every race, and every year, over and over and over. Texas in 2004? Kerry wins every income band up to $30k, and loses everything else by increasing landslides (losing 92% to 8% for people making over $200k). California in 2004? Same deal; Kerry wins every income band up to $100k, and loses everything else by ever-increasing margins as you climb the income scale. Congressional elections in 2010? Same thing: [4] TL;DR: Today I learned that you don't need to have a good grasp of statistics, logic, or American politics to be an associate editor of The Atlantic. [1]: http://en.wikipedia.org/wiki/Simpsons_paradox http://en.wikipedia.org/wiki/Simpsons_paradox [2]: http://super-economy.blogspot.co.nz/2011/04/red-state-rising.html http://super-economy.blogspot.co.nz/2011/04/red-state-rising... [3]: http://edition.cnn.com/ELECTION/2004/pages/results/states/MS/P/00/epolls.0.html http://edition.cnn.com/ELECTION/2004/pages/results/states/MS... [4]: http://4.bp.blogspot.com/-JGX1ch5CkN8/TZ9mEU7lnJI/AAAAAAAAAeU/KIF8RJuRY0g/s1600/cnn.png http://4.bp.blogspot.com/-JGX1ch5CkN8/TZ9mEU7lnJI/AAAAAAAAAe...