Quite interesting. I disagree with the conclusion (more on that in a moment), but (a) this is a great thing to investigate with the data; (b) as soon as I saw

Quite interesting. I disagree with the conclusion (more on that in a moment), but (a) this is a great thing to investigate with the data; (b) as soon as I saw the time-series data, I immediately wanted to plot it on the 2-axis / 4-quadrant Nolan chart ... and that's exactly what the author did next! Now, why I'm not convinced that the data actually shows what the author thinks it shows: there are two possible explanations of the data: 1) that people change their views over time 2) that views are a function of when you were born The author accepts point 1 without even thinking about point 2. Two examples: 1) data shows that the older you are, the more your joints hurt. 2) data shows that the older you are, the more you disapprove of homosexuality. I assert that fact #1 is explained by "old age leads to creaky joints", and fact #2 is explained by "social norms changed between 1930 and 2000, and people, to a good degree, retain the social norms of their youth". Instead of concluding that "youth makes you a libertarian and old age makes you an authoritarian", I suggest that the truth might be that "being born after the Berlin Wall fell and globalization occurred makes you a libertarian; coming of age in the hippy era makes you a leftist, and living through the Great Depression and WWII makes you an authoritarian". I don't endorse everything in it, but the book _Generations_ ( http://www.amazon.com/Generations-History-Americas-Future-1584/dp/0688119123/ http://www.amazon.com/Generations-History-Americas-Future-15... ) argues something pretty much in this direction. Given that OKC's data is all from a very short period of time, there's no way to prove one hypothesis over another. Given a longer time series, though, we can disambiguate.