Check Out the Graphs on Andrew Gelman

Why does it have to be moral conflict? Why can it not be that a percentage of the well off realize that they have a duty as 'well off people' to give back to the system which allowed them to become well off? And so don't mind progressive taxation? And so judge between the parties on matters other than pocketbook? On perhaps ... just to pick an example ... "the good of the Republic" ... or "seperartion of church and state" .... or pick one ... when you stop looking at your own pocket book the playing field changes ... not all in one direction certainly ... but it changes.

It's really not very hard to understand: Democtrats reached equality to Republican voteshare in people who make 6-figures. I believe Obama may have even won this Demographic outright. Look at NY, CT, and CA - some of the wealthier states - all blue.

My guess would be that for "SuperRich" people - eg. people making $1 million or more - morality is conflicted with the fact that they are SuperRich. There's a point in the income distribution where you have to sacrifice some of your morals to become as rich as you are. I would think that there are some people in this group that feel very guilty, no matter how often they go to church.

Measuring church attendance has long been used as a highly imperfect surrogate for religiosity, and it's highly imperfect, particularly among Catholics. I wouldn't read this much into it other than the GOP has a problem if the country gets more secular.

It's curious that at the third quartile, the change in slope is so dramatic, and most dramatic in the people in the middle (neither always nor never attend church). Other than this section of the graph, each of the lines is fairly similar to each of the others, and each could almost be approximated to a straight line. A few of the other graphs posted at 538 show a similar pattern.

Geography.

The top quintile disproportionately can be found in urban centers. Political preferences are strongly influenced by the people one associates with, which explains why church attendance (mostly) increases the tendency to vote Republican. (Going to church with social conservatives makes you more conservative.) However, in large enough cities, the population will lean Democratic strongly enough that church attendance does not suggest increased association with social conservatives.

I really should check myself rather than asking (I would plead busy morning, but then why am I looking at a blog?), but are they controlling for other variables, or is this just the straight data? I'm specifically thinking about education, but maybe race & gender too. IIRC playing around with GSS, holding race & gender constant, income makes one slightly more conservative (holding education equal) and education makes more one slightly more liberal (or at least socially liberal), holding income constant. I wonder what happens to the bends in those curves if they control for education (assuming they haven't) or if don't (assuming they have)?

This is the straight data. In Andrew's post, he makes the case for the virtues of eyeballing plain data.

You can go back to Kevin Philips and "American Theocracy" for a pretty good answer. I personally think Philips can go over the top, but has nonetheless been a shrewd observer for the last couple decades.

My quick take on where a socially liberal orientation translates into policy (please tell me where I'm wrong).

Gay marriage. OK, but then you should be equally as enthusiastic for polygamy as well. Meaning, my thumbnail sketch of a properly ordered libertarian social philosophy would view it as a marketplace where consenting adults would be allowed to experiment (the entrepreneurs) with different social arrangements and that the 'invisible hand' of the marketplace would allow for the successful arrangements to succeed while the losing arrangements would disappear (barring subsidies or bailout from the gov't of course!). In this understanding, the right to be a polygamist should be fought for and defended just as much as the right to be in a same sex marriage.

Drug policy - Obama admin made some steps in toning down drug war but I am not convinced that this is an issue which is either right or left. Proponents of both sides of argument are on both sides of aisle (with nobody willing to stick their neck out on the issue terribly far)

Immigration - Probably the strongest argument liberaltarianism has, IMO. However, there needs to be a reckoning of both national security concerns and how welfare-state benefits are conferred within such a framework.

At any rate, I wish you success in your liberaltarian adventure. If we could get both parties (puh-leeze don't go the 3rd party route) to fight with each other over who could be the most free-market-ey-ist, no one would be happier than I.

Is that broken down at the level of states or ethnic/racial groups? I also recall Larry Bartels pointing out that the idea of the white working class going for Republicans applies a lot less to women.

Could you find this graph with error bars on it? Chances are that whatever their sample, unless they deliberately oversampled the very rich there aren't very many of them, so the variances on the points are probably higher further to the right in the graph than they are in the left portion of the graph.

Well, there seems to be a slight dip across the board between the second highest income section and the highest, but in all cases it's a very slight dip. Are we sure that it's statistically significant?

If you are confident that you will gain from more government intervention either because
(1) it is the business in which you work (e.g. lawyers) or
(2) you have good access to regulators
then you are more likely to vote for it. (Or, at least, less likely to vote against it.) Interventions can, after all, discourage competition--see Latin American with its elite-patronage mercantilism.

Depends on the what the definition of "rich" is. It makes sense if the break is where one starts finding those who are high earning professionals who are still employees rather than owners. I'd expect that a lot of small business owners fall on the left of the break in the chart, whereas lawyers and doctors fall on the right.

I count five points of data, so he's probably plotted quintiles.

That would mean the final, downward-leaning lines represents households with income starting at about $100,000 and ending at Bill Gates.

I think that works for my little theory, though 100K in New York is not rich whereas it is in Jackson, Mississippi. At the end of the day, that may be all that is being captured. If you restricted the data to New York City, voters might just keep getting more conservative as they get richer, though overall less likely to vote Republican than in the more conservative lower income areas.

EDIT: That is, nationally, the more conservative areas of the country are less likely to have as many people in the top national quintile of earners. So if you broke it out by state or region, the local highest quintile locally would always vote more Republican than local lower quintles, though that particular highest quintile would likely appear more liberal than other quintiles of other parts of the country.

Could be, Ben.

The chart for the 2004 election didn't have the final, downward-sloping lines in the frequent and occasional church attender plots, though.

I'd guess either Bush soured some of the rich on the Republicans or they thought Obama was better for them than McCain.

For some of them, guilt. As you make significantly more money, some people (regardless of affiliation) are going to feel bad about it.

For some of them, it's that government hurts less. Who cares about 60% tax rates when you make 700k a year? This is part of the limousine-liberal issue and I'm sure you have a number of limousine-Republicans as well.

Most of my friends who grew up rich or well-off (dems, repubs, libs, etc.) are nowhere near as angry about taxes and government spending as I am, but those taxes had a bigger impact on my lower-middle class family.