Are Congressmen who challenged the election results in thrall of Trump's base?
Jacob Whiton implicitly raised an interesting question on his Medium blog: who are the constituents of the 139 Congressmen who objected to the certification of the electoral college even after the Capital rampage? Ie, to what social class do they belong? I had a long exchange with Whiton on Twitter. I objected to his claim that higher income and wealthier voters are still more likely to be Republicans.
[V]oters with higher incomes and who own more wealth, are still significantly more likely to support Republicans.
Jacob Whiton, "Where Sedition is Rewarded," Medium.
He did not share median income or wealth data by congressional district. But the median home values he reported were inconsistent with higher income voters favoring the GOP. We had a long exchange and I urge you to see the Twitter exchange.
Whiton is not alone in misunderstanding the class basis of Trumpism. From Jeff Sachs to Nikole-Hannah Jones, the entire professional class seems to be living in la la land. Even the rising far-left star, Seattle Congresswoman Kshama Sawant, has no clue about the social basis of Trumpism.
Anyway, Jacob Whiton was kind enough to share his data and sources with me. In what follows, we'll look at a few socioeconomic indicators of the districts represented by the Congressmen who refused to ratify the election. I obtained the objection vote data from Jacob Whiton, who tabulated it by hand. I obtained socioeconomic indicators from IPUMS and aggregated it at the congressional district level. You can find the replication code in python, the graphs, and instructions on how to obtain the data from IPUMS at my GitHub.
We begin with education. The variable EDUC is an ordinal variable that roughly corresponds to education attainment, ie years of schooling. We present both race-blind and race-specific plots for non-Hispanic whites in order to avoid confounding by race. Between Democratic and Republican districts, we can see that there are dramatic differences in educational attainment — the former are much more educated than the latter. These are especially pronounced among whites. There seems to be a difference between districts represented by the objectors and non-objecting Republican Congressmen. Are these differences statistically significant? We'll get to the bottom of this shortly.
Next, we look at median household income. We find virtually the same pattern — Republican-held districts are poorer than Democrat-held districts. Again, we find that there is a visible difference between districts held by objecting and non-objecting Republican Congressmen. Are they significant?
Class partisan polarization has proceeded so far that it does not make any difference whether one uses income or education as a proxy for class — the GOP is the party of the working class, the Democratic party is the party of the affluent. Whiton's following claim is simply inconsistent with reality:
The conflation of educational attainment with class obscures the extent to which many white Americans without a four-year degree receive middle class incomes….
Jacob Whiton, "Where Sedition is Rewarded," Medium.
More importantly, education contains a stronger class signal in the sense that it is a better predictor of life outcomes than income. This is now very well understood by health researchers who have shifted en masse to using educational attainment as the class proxy. Some political scientists have got it; others still haven't gotten the memo.
We now look at occupational prestige, an ordinal variable that ranks respondent's occupation by generally perceived prestige. This variable arguably contains an even stronger class signal than educational attainment, and, a fortiori, household income. Again we see precisely the same pattern. The gap between Democrat-held districts and Republican-held districts is dramatic. And there is a visible gap between objectors and non-objector GOP-held districts. It is also a bit more pronounced among non-Hispanic whites. Are these gaps statistically significant?
Finally, we look at Poverty, a variable that measures a family's total income as a percentage of the local poverty threshold. Yet again, we see the same pattern. GOP-held districts are poorer than Democrat-held districts. And this pattern is more pronounced among non-Hispanic whites. We also see the same small gap between objectors and non-objector GOP-held districts. Is the latter statistically significant?
In order to test the statistical significance of these differences we carry out t-tests. Table 1 reports party-blind means and test statistics for objectors and non-objectors. There is no disagreement at all between the different measures, apart from the fact that restricting attention to non-Hispanic whites makes the differences uniformly more significant, suggesting that race confounds the class signal. Simply put, the constituents of objecting Congressmen are less educated, poorer, and enjoy lower occupational prestige than the constituents of non-objecting Congressmen.
These differences are largely due to class-partisan realignment. Table 2 reports the means and test statistics for Democrat and GOP-held districts. The constituents of GOP Congressmen are less educated, poorer and enjoy lower occupational prestige than the constituents of Congressional Democrats. And the difference is more pronounced and significant once we restrict attention to non-Hispanic whites, suggesting that the nonwhite bias towards Democrats masks some of the class-partisan realignment that has obtained.
But the crucial question is whether there are significant differences in the constituents of objecting and non-objecting Republican Congressmen. We find that, although the differences are uniformly in the direction of objectors' constituents being less privileged than non-objectors' constituents, they are only statistically significant for occupational prestige. The implications are straightforward. A vast socioeconomic gap exists between the parties. The GOP has become the party of the working class; the Democrats have become the party of the educated and the affluent. And within the GOP there is some weak evidence to suggest that more radical Congressmen, as revealed by their objection to certifying the electoral college tabulation, represent more working-class communities.
It simply won't do to pretend that the Democrats represent the working-class of America. They do not. And they haven't for a while. As for Trump's base, it is solidly working-class. And it became more so in the 2020 cycle. Moreover, Trumpism has broken out of the white working-class and attracted significant number of Black men and Hispanics. This does not bode well for Democrats.
Democrats won the election fair and square. But Trumpism has not been defeated. And the present Moral Panic over the storming of the Capital is likely to exacerbate the breakdown of elite-mass relations. Yes, as Nikhil Pal Singh put it, 'we beat them this summer'. That is hardly surprising. We have the money, we have the media, and we have the tech overlords on our side. The advantage always lies with the winners and we, the professional class, are the winners of the class war. But dominance is not hegemony, as I explained in my exchange with Nikhil. See the whole tweet storm.