After 5000 respondents, the DailyKos/PPP poll now has enough data that we can see a slight pattern in Obama's approval as a function of income:
Approval is slightly, but significantly, higher among the lowest and the highest earners. Can we make sense of this?
First, we will look at only those older than 30. Many of those under 30 are still in school or training programs and while their income may not be much, they are not representative of the working poor in general in that they have prospects for dramatic improvement in the near future.
Removing young people doesn’t actually make that much difference, but it does make some:
The next obvious thing to do is to separate out racial categories, as income is clearly not distributed equally by race in this country:
And now the pattern resolves itself into something more clear. In each category there is an apparent relationship (but it is not statistically significant) between approval and income.
So, approval increases with income, right? Well, but….just because you don’t approve doesn’t mean you disapprove. Here’s what the numbers look like for those who say ‘Not Sure’ in response to the approval question:
Low income voters are simply less likely to express an opinion. Which messes with the apparent trends we saw in approval. If we express approval as a percent of those who have an opinion, what weak trends there were pretty much disappear:
There is an exception, and that is that the highest earning whites have a significantly greater approval of Obama than three of four lower income categories. (As more data come in, we may see something develop in Hispanic and Other categories, but for now the error bars are too big.) Indeed, if we look at the obscenely wealthy - those with net worth over $5 billion, about 55 individuals according to Forbes - the split between supporters of Democrats and Republicans in general is somewhere around 45%. (Is this a good thing or a bad thing, really?)
But white working class voters look about the same as white middle class and upper middle class voters (median household income is about $50,000).
In comparison, here’s the planned votes reported in last year’s general election in the UK by social grade, which is roughly equivalent to income. (AB is >£35K, DE is <£15K as of 2004) Political behavior in the UK is clearly more tightly linked to class.
So what can we say after all this? More low income voters are unsure of their opinion of Obama, but after accounting for this, and race, approval of Obama is independent of income, except approval of Obama among highest income white voters is a little higher. In other words, politics in this country is, strangely enough, almost completely disconnected from broad economic class differences.