Jim Manzi has graciously replied to my inequality post below. Let me just reply to this one bit now, and I'll take up his worries about the threat of the angry mob later.
I think the reason this indicates that inequality poses “more of a political problem for Republican coalition-building than it does for the Democrats” is, to [simplify] greatly, that if inequality of condition causes a preference for Democratic policies, then as inequality increases, you will get more Democrats. If it’s a marker for other causes of Democratic voting, then it’s a marker for things that make more Democrats. The idea of a coalition of aristocrats and the proletariat against the petite bourgeoisie and yeoman farmers is not a new one.
I've see no evidence that inequality tends to produce more Democratic voters. Does Jim have some?
It's funny, because Jim is arguing the opposite of what I'm calling in my paper the Inequality Road to Serfdom (IRS) argument, which just about every left-leaning thinker seems to accept. According to IRS, the wealthy convert their economic resources into political resources deployed to protect their advantages. Past a certain threshold of inequality, the gap in political resources is so vast that the wealthy–by means of insidious think tanks and the Republican Party–basically capture the democratic process and effectively disenfranchise the rest of the population, leaving us with a de facto oligarchy hiding inside a meaningless husk of democratic forms. Serfdom! Which is just to say, the worry on the left has been that as you get more inequality, you get more Republicans. This is superficially plausible to your Paul Krugmans in part because the era of rising inequality, beginning in the early 1970s, corresponds with a resurgence of Republican political power, first with Reagan, then big Republican congressional majorities in the 1990s, and then unified Republican government under Bush. I'd be interested to hear more from Jim about why Krugman et al have been getting it exactly backwards.
There does seem to be a mild increase in support recently for redistribution, but it doesn't look like a steady, long-term trend. Income inequality has increased a lot over the last thirty years but over that time support for redistribution has zig zagged all over the place. If I recall the paper I'm thinking of from Lane Kenworthy, the rise in inequality seems to correlate well with fairly firm gains in demand for education reform–pointing to the typical American focus on opportunity–but that doesn't obviously help the Democratic Party, insofar as it is widely identified an agent of the change-hating teacher's unions.
I think Democratic gains in the last election have almost nothing to do with inequality. Most of the gains are just part of the political business cycle, and so indicate nothing much other than that the candidate from the incumbent party doesn't do well when saddled with an unpopular war and a recession. I think the more likely permanent gains, among the young and rich, for example, have a lot to do with the general (I think salutary) social liberalization of American moral culture. The GOP is intolerably socially conservative for an increasing share of Americans. I predict that a significant decrease in inequality, whether from recession or redistribution, would do almost nothing to stop this flight toward the Democrats.