In response to my post below, Are the elites more polarized? Yes!, a comment:
The high "moderation" of low-income voters tends to make me think that what's measured is low-information voters instead, people don't take an active interest and who answer in harmless generalities. And maybe people who feel totally helpless, as though their input cannot make any difference.
This seems plausible. How to figure this out? First, one thing to remember is that there is a correlation between measures of intelligence/knowledge and socioeconomic status. Here is number correct on the vocab test out of 10 on the Y axis and socioeconomic index divided into quartiles on the X (WORDSUM and SEI):
The correlation between SEI (which is on a 17-98 scale) and WORDSUM is 0.37. In terms of "low information," one has to take into account general levels of education and/or intelligence.
At this point I wanted to query all the variables which ask people how "firm" they are in their views. Unfortunately, the vast majority have very low N's. Additionally, none of these included socioeconomic information. There is one question though, ABFIRM, which asks about abortion, with a decent sample size (2,430 whites answered this question). Since this is the exemplar of Culture War issues, I decided to look at it for education and vocab test.
It looks to me that the low proportion who seem very changeable in their viewpoint on this touchstone cultural issue are those with less education, or lower vocabulary test scores (note that I lumped everyone who scored from 0 to 2 together since the N's here were low). So there does seem to be some suggestion here of polarization and certitude as one moves up the socioeconomic ladder in terms of an issue which is often portrayed as a black & white divide. Unfortunately, it's just one variable, and the effect is weak enough that I would want to look at others (obviously, the overwhelming number of people are rather set in their ways when it comes to their position on abortion; at least at any given time). But as I said the N's for other issues I could think of were just really small. If people can think of a way to use the GSS to expand beyond this issue, or have other data besides the GSS, I will post an update....
- Log in to post comments
Like the last time I commented, it feels like there should be a cusp involved. Perhaps higher strength attracting force to settle them out on the different points of the cusp catastrophe faster? Also, possibly those well-settled tend to already be relatively far from the cusp transition area.
Also, those lower on the socioeconomic ladder have less marginal effort unrelated to survival available to dispose of. They thus expend more effort on immediate concerns. Those who have more time to spare may have put more effort into considering a position, and thus be harder to shift.
Available leisure is unevenly distributed according to SES. There are certainly many low-SES people working two jobs, or single parents, etc., for whom that would be true, though.