We’ve been extending our work on income and voting to include religion as well. For example:
MS, OH, and CT represent poor, middle-income, and rich states, respectively, and the red, blue, and gray lines on each plot represent frequent church attenders, occasional church attenders, and nonattenders.
Rch people vote more Republican and church attenders vote more Republican, but in addition, the difference between church attenders and nonattenders is greatest among rich people in poor states (as in the Mississippi graph above).
To get more understanding of the patterns of income and religion by state, here’s state average religious attendance vs. state average income, with the colors indicating states that went for Bush or Gore in 2000:
There’s a negative correlation, but also regional patterns, with the western states (except for Utah), some northeastern states, and some others below the main line.
And here are the within-state correlations between income and religious attendance (again plotted vs. state average income):
Religious attendance is positively correlated with income in Mississippi (for example) but negatively correlated in Connecticut. Thus, the zero correlation between income and Republican voting in Connecticut (“What’s the matter with Connecticut”) is partly explained by the fact that poor people in Connecticut are more religious than rich people in Connecticut. (But this doesn’t tell the whole story, as we can see from the graphs at the top of this entry.)
Anyway, we’re thinking more about this–other factors to consider are religious denomination, urban/suburban/rural, and ethnicity. Our original goal was to understand the pattern that richer voters go Republican, but richer states support the Democrats, but now there are many more patterns to figure out.
should you adjust the average income by the cost of living?
I would bet that the MS results for income are highly correlated with Race. I am conjecturing that most poor churchgoers are African-American and therefore more likely to vote Dem. while the opposite is true for the white rich churchgoers. You should control for race.
"Religious attendance is positively correlated with religion in Mississippi"
I think you meant correlated with income.
Thanks,
Jim
David,
Yes, we can do this, although it won't change much, I think, since this will affect the intercepts but not the slopes.
Steven,
Yes, I agree. For our income and voting paper, we did a whites-only analysis and found similar results as with the entire population, and we'll be checking this with our income/religion/voting analysis too.
Jim,
Thanks.
"I think you meant correlated with income."
Wouldn't the results than be somewhat skewed due to the bias?
Two comments:
First, you note that "the difference [in Pr(Republican)] between church attenders and nonattenders is greatest among rich people in poor states (as in the Mississippi graph above)." That's not how the graphs look to me. I would say that the difference between church attenders and nonattenders is greater (0.8 vs. 0.4) among rich people (income = +2) in Ohio than among any other (income,state) combination shown in the graphs. Or I would note that the difference between poor and rich is greatest (0.9 vs. 0.3) among frequent church attenders in Mississippi.
If your statement is correct, then the graphs you show are a nice example of the synecdochic fallacy. The relative differences in rich people's voting between Mississippi, Ohio, and Connecticut need not match the relative differences between poor states, middle-income states, and rich states. To support an observation about poor, middle-income, and rich states, show graphs that display data about (all) poor, middle-income, and rich states.
Second, although some states are richer than others, and some states are more church-going than others, you seem to measure income on a state-relative scale, but church-goingness on an absolute scale. But I'm guessing. You don't explain (that I can find) the five-point income scale. What does +1 mean? Does it mean "between $X and $Y," "between the particular state's 3rd and 4th quintile," "between the 3rd and 4th quintile nationally, after adjusting by a geographic salary differential," or something else? If your income scale does differ from state to state, I'd be interested to see what your graphs look like with a state-relative church-goingness scale (red, blue, and gray representing the top, middle, and bottom thirds of church attendance for the particular state – or even for the particular state and income category) instead of an absolute one. It would also help if you explained your five-point income scale (or point me to the explanation, if I missed it).
Each of your state graphs contains 15 points on an (income,voting) plane, one for each pair of attendance category and income level. But those 15 points represent different numbers of people each, and more importantly, the way the population is distributed across those points differs greatly between states. The churchgoing poor point of the Mississippi graph represents a larger fraction of that state's people than the churchgoing poor point of the Connecticut graph. You could represent this if you printed numbers, drew different-sized bubbles, or adjusted the line widths to show the fraction of the population accounted for at each of the 15 points.
If you can provide the raw data, I'd be happy to create these alternative views.
Steve,
Thanks for your comments.
In our first graph, it would probably help to have little circles at each of the 15 spots, with the size of the circle proportional to the relative number of people in each category.
Regarding the absolute church attendance and relative income scale: the incomes are relative to U.S. percentiles. So, for any given year, you can think of it as absolute income. We are using the same income meausure (and the same church-attnedance measure) in each state.
See also here. It's also possible the graphs will look different if we break them down by ethnicity.