Scott Sumner's proposed new political typology chart is getting a lot of attention today:

 

Libertarians obviously love it. I see a few major problems. First, it defines completely out of existence people with pro-government economic views and conservative social views, which are a vastly largely share of the electorate than libertarians or even vaguely libertarian-ish folk, but simply lack any intellectual infrastructure or funding base. Second, it defines out of existence the possibility of libertarian corruption. I see where he's coming from here -- it's hard to be corrupt when you have no power -- but I'm pretty sure a careful study of the Koch brothers political activism would cast doubt on that assumption. (Unless you really think attaching no price to the externality of carbon emissions is a core component of economic freedom.)

And third, he doesn't account for greater or lesser degrees of liberalism or conservatism. Conservatives and liberals can only be more idealistic or corrupt. I suppose you could say Barney Frank is more "corrupt" than Paul Krugman because Frank has to work within the political system and Krugman doesn't, even though that's a very loaded way to characterize people who are working within the system. But how would you plot Barney Frank versus Mark Warner? Olympia Snowe versus Jim DeMint? Each pair would be occupying about the same place in the chart, and I don't think that makes a lot of sense.

I do get what Sumner is trying to accomplish here -- he's trying to account for the difference between issues that are on the political agenda and those that are off it. But I think he needs to go back to the drawing board.