Income groups among white voters: $50-$75k vs $75-$100k

[All Along The Watchtower]

Using this website http://www.dimpledchad.info/result/, if one controls for education (so that you are comparing white voters across different income groups, but within the same educational group), one finds, as expected, that, for the most part, higher-income whites who have "some college" or "no college" (high school graduates) are much more Republican than lower-income whites within the same educational groups (Example: McCain got around 46% of whites in the "some college" category with annual incomes of under $30k, but about 56% with annual incomes of $30k-$50k, and over 60% with annual incomes of $50k or above).

However,  the correlation doesn't work at the $75-$100k income level, even among white voters within the same region and the same educational level. McCain got 56% of the vote  in at income level among white high school graduates, for example, but 65% of the vote among white high school graduates at the lower income level of $50-$75k.

Why is this? My hypothesis is that the demographic "white high school graduates with incomes of $75-$100k a year" might be a good predictor of belonging to a unionized job or a small business owner who feels more financially stable, not struggling as much, and so less receptive to Tea Party's right-wing populist appeals.

Anyone else have any theories about this?

[Benj]

The margin of error on an individual demographic group that refined is almost 10%. It's just noise. Also, "small business owners" are uber-Republican.

[All Along The Watchtower]

The margin of error on an individual demographic group that refined is almost 10%. It's just noise. Also, "small business owners" are uber-Republican.

Makes sense, I guess. Though I question how "uber-Republican" small business owners really are. I mean, we hear about the Tea Party types and such being disproportionately self-employed, but there are plenty of Democratic business owners too.  They just don't fit the narrative.

[Indy Texas]

What would the sample size be of people with no college degree who make $75-100K a year? Not very big, I'll bet.

[Franknburger]

Have you controlled the 75-100k income results for age? I could imagine it including quite a number of people that are already retired or close to retirement.

In this case, you could have a lower educational attainment level than for younger people in the same professional group.  Let's hypothetically take the hospitality sector (a sector that is typically assumed to lean Democrat) - older middle-management staff might be high-school only, younger ones would have undergone some college education.

Asides, as I have noted in this thread (scroll to the end),  the white 45-64 age cohort, and especially the 60-64 subgroup (2008 age classes) seems to display an above-average Democrat lean and was also more resilient to the 2012 white pro-Republican swing. This may on one hand relate to political socialisation ("JFK generation"). In addition, this age group may have preferred Democrats due to specific issues (2008: Financial crisis and its effect on pension funds, 2012: Obamacare / Medicare).  

[barfbag]

When you go much more over a $75,000 income, especially $100,000, you tend to run into elitists and people who think they know everything. With that line of thinking comes the mindset of thinking "people like them need people like us to make decisions for them because after all, the average person needs big brother." When people think like that, they tend to vote for the candidate who best matches their level of elitism. People in the real "middle class" are Republicans. They work everyday and know how the world really works.