Can anybody with access to General Social Survey data or the like help me out with testing my theory that Republicans tend to have higher standards of living per level of educational credential while Democrats tend not to make as much money as their education levels predict?
Here's a graph using fictitious data that I just made up off the top of my head to give you an idea of what I think is happening. The vertical axis is income in thousands and the horizontal axis is educational level from 1 to 5, with 1 as high school dropout, 2 as high school graduate, 3 as some college, 4 as college graduate, and 5 as some postgrad. Republicans are represented in red, Democrats in blue. I am betting that more Republicans fall above the best fit line, more Democrats below it.
To make this analysis simpler, I'd recommend looking just at non-Hispanic whites.
The biggest problem with this analysis is the huge cost of living differences around the country, with, for example, California 44% over the national average and Texas 11% below it. White Democrats tend to live in higher cost of living locations, so their standards of living are lower than their raw incomes would indicate.
One way to reduce this problem is to look just at a single state with fairly consistent cost of living within the state. Texas might be the likeliest bet because it offers a big sample size of respondents, it's the most urbanized state in the country, and its big cities (with the possible exception of Austin) are not terribly expensive (in contrast, Illinois or New York are pretty rural/small townish except where they are extremely megalopolitan).
Another way is to adjust every individual for his state's cost of living differences. Here's the ACCRA table on cost of living by state that's updated for corporations transferring employees.