A few weeks ago there were a bunch of stories on how white the audience was at Glenn Beck’s rally. That’s empirically true, and the Tea Party movement as a whole is overwhelmingly white. So is the American conservative movement. This in a nation which is ~65% white in a colloquial sense (i.e., white Hispanics are excluded from the class of “white”). It makes one’s eyebrows go up I suppose when you see a very unrepresentative set of people. But what irritates me about media observation of this statistical reality is that the elite media is also disproportionately white. Much of the elite media and the up & coming pundit class reside in a majority black city, but if you check out their Facebook photos or flickr accounts you would be totally surprised at the fact that they reside in a “Chocolate City.” Why are the social circles of elite media types, liberal or conservative, not representative of the city in which they reside? There are pretty clear reasons of confounds of class and socioeconomic affinity with race. The demographics of one’s social circle don’t necessarily lead one to prima facie accusations of bias, rather, they’re embedded in a set of causal assumptions and conditionals. So, from a liberal perspective the whiteness of the SWPL milieu is situational, while that of the right-wing milieu is essential. The demographics of conservative political movements themselves are interpreted through a particular historical frame of racism for most liberals implicitly. In contrast, the white demographics of elite liberals, including the Netroots, are often “contextualized” as emerging out of a whole range of historical and social processes, which if not just in and of themselves, are structural factors which elite white liberals are not responsible for and are attempting to change.
It seems a pretty robust social science finding that white liberals have less racialist sentiment than white conservatives. My main beef, as a non-white conservative, is that a quantitative difference of degree gets collapsed into a qualitative difference of kind. Transforming a quantitative variable into a dichotomous categorical one totally changes the inferences one makes from facts. The whiteness of conservative movements and classes then entails the casting of particular aspersions, while the whiteness of liberal movements and classes tends to go under the radar as having a sociological cause out of the control of white liberals.
To explore the quantitative, as opposed to qualitative, difference between white non-Hispanics of varied political stripes I decided to look at the GSS data set. There are a variety of questions on racial issues, though I focused on the ones related to white opinions/attitudes/relations with blacks since they are more numerous. For example, in 1974 23% of white liberals and 36% of white conservatives favored a law banning interracial marriage. In 2002 the values were 8% and 13% respectively. In both cases you can see that white conservatives have more racialist feeling, but the difference is not dichotomous, but one of degree. Below is a table of responses to a set of questions by white non-Hispanics in the 2000s. I broke out the data set by liberal and conservative, and Democrat and Republican. Additionally, in addition to the raw frequencies I also calculated absolute and relative differences between liberals and conservatives and Democrats and Republicans.
| Ideology | Party | Ideology | Party | |||||
| Response | Lib | Con | Dem | Repub | Abs Gap | Rel Gap | Abs Gap | Rel Gap |
| Favor law against racial intermarriage | ||||||||
| Yes | 7.3 | 13.6 | 12 | 10 | -6.3 | 0.5 | 2.0 | 1.2 |
| Black person over for dinner recently | ||||||||
| Yes | 46.7 | 39.2 | 36.5 | 41.5 | 7.5 | 1.2 | -5.0 | 0.9 |
| Would vote for black president | ||||||||
| Yes | 96.4 | 93.8 | 93.3 | 95.4 | 2.6 | 1.0 | -2.1 | 1.0 |
| Whites hurt by affirmative action? | ||||||||
| Very likely | 14.5 | 22 | 15.9 | 21.8 | -7.5 | 0.7 | -5.9 | 0.7 |
| Somewhat likely | 48.1 | 50 | 49.2 | 52.4 | -1.9 | 1.0 | -3.2 | 0.9 |
| Not very likely | 37.4 | 28 | 34.9 | 25.8 | 9.4 | 1.3 | 9.1 | 1.4 |
| Close relative marry black | ||||||||
| Strongly favor | 17.9 | 9.5 | 14.3 | 9.7 | 8.4 | 1.9 | 4.6 | 1.5 |
| Favor | 11.8 | 12.7 | 12.5 | 12 | -0.9 | 0.9 | 0.5 | 1.0 |
| Neither favor nor oppose | 50.3 | 37.4 | 43.8 | 41.7 | 12.9 | 1.3 | 2.1 | 1.1 |
| Oppose | 10.2 | 19.9 | 13.1 | 19.3 | -9.7 | 0.5 | -6.2 | 0.7 |
| Strongly oppose | 9.7 | 20.6 | 16.2 | 17.4 | -10.9 | 0.5 | -1.2 | 0.9 |
| Have conditions improved for blacks | ||||||||
| Improved | 60 | 72.7 | 64.6 | 73.9 | -12.7 | 0.8 | -9.3 | 0.9 |
| Gotten worse | 4.8 | 3.6 | 3.7 | 3.5 | 1.2 | 1.3 | 0.2 | 1.1 |
| About the same | 35.2 | 23.7 | 31.7 | 22.6 | 11.5 | 1.5 | 9.1 | 1.4 |
| Has most in common with | ||||||||
| Whites | 16 | 16.5 | 16.5 | 17.9 | -0.5 | 1.0 | -1.4 | 0.9 |
| Blacks | 20.7 | 11.9 | 18.5 | 15.4 | 8.8 | 1.7 | 3.1 | 1.2 |
| Jews | 22.5 | 20 | 22 | 19.9 | 2.5 | 1.1 | 2.1 | 1.1 |
| Hispanics | 10.1 | 17.7 | 9.6 | 17.4 | -7.6 | 0.6 | -7.8 | 0.6 |
| Asians | 5.4 | 12.4 | 7.3 | 9.8 | -7.0 | 0.4 | -2.5 | 0.7 |
| Equal in common to all | 21.2 | 13.2 | 19.9 | 13 | 8.0 | 1.6 | 6.9 | 1.5 |
| Nothing in common with any | 4.1 | 8.3 | 6.2 | 6.6 | -4.2 | 0.5 | -0.4 | 0.9 |
| Number of blacks one is acquainted with | ||||||||
| 0 | 19.4 | 26.8 | 24.3 | 25.4 | -7.4 | 0.7 | -1.1 | 1.0 |
| 1 | 7.2 | 9.3 | 13.8 | 6.9 | -2.1 | 0.8 | 6.9 | 2.0 |
| 2-5 | 37.1 | 35.5 | 33.1 | 37.2 | 1.6 | 1.0 | -4.1 | 0.9 |
| 6-10 | 11.8 | 11.8 | 11.6 | 13.3 | 0.0 | 1.0 | -1.7 | 0.9 |
| More than 10 | 24.5 | 16.6 | 17.1 | 17.2 | 7.9 | 1.5 | -0.1 | 1.0 |
| Number of blacks one trusts | ||||||||
| 0 | 36.3 | 47.7 | 40.9 | 46.6 | -11.4 | 0.8 | -5.7 | 0.9 |
| 1 | 16.9 | 16.4 | 21.3 | 14 | 0.5 | 1.0 | 7.3 | 1.5 |
| 2-5 | 32.3 | 25.9 | 24.4 | 28.2 | 6.4 | 1.2 | -3.8 | 0.9 |
| 6-10 | 9.4 | 4.9 | 8.1 | 5.5 | 4.5 | 1.9 | 2.6 | 1.5 |
| More than 10 | 5.1 | 5.1 | 5.3 | 5.7 | 0.0 | 1.0 | -0.4 | 0.9 |
| Number of blacks in neighborhood | ||||||||
| 0 | 52.4 | 61.5 | 55.4 | 59.3 | -9.1 | 0.9 | -3.9 | 0.9 |
| 1 | 12.7 | 10.2 | 12 | 9.3 | 2.5 | 1.2 | 2.7 | 1.3 |
| 2-5 | 22.3 | 20 | 20.7 | 21.2 | 2.3 | 1.1 | -0.5 | 1.0 |
| 6-10 | 2.6 | 4.6 | 1.7 | 4.6 | -2.0 | 0.6 | -2.9 | 0.4 |
| More than 10 | 10 | 3.8 | 10.1 | 5.7 | 6.2 | 2.6 | 4.4 | 1.8 |
| Number of black family members | ||||||||
| 0 | 77.7 | 83.7 | 80.1 | 81.8 | -6.0 | 0.9 | -1.7 | 1.0 |
| 1 | 11.3 | 13.4 | 10.4 | 8.9 | -2.1 | 0.8 | 1.5 | 1.2 |
| 2-5 | 8.5 | 2.9 | 5.8 | 9 | 5.6 | 2.9 | -3.2 | 0.6 |
| 6-10 | 0.8 | 0 | 0.6 | 0 | 0.8 | 0.6 | ||
| More than 10 | 1.7 | 0 | 3 | 0.3 | 1.7 | 2.7 | 10.0 | |
| Number of blacks in voluntary associations one involved with | ||||||||
| 0 | 41.2 | 43.4 | 40.8 | 42.7 | -2.2 | 0.9 | -1.9 | 1.0 |
| 1 | 11.2 | 8.2 | 10.3 | 7.4 | 3.0 | 1.4 | 2.9 | 1.4 |
| 2-5 | 28.4 | 26.6 | 26.3 | 25.9 | 1.8 | 1.1 | 0.4 | 1.0 |
| 6-10 | 9.2 | 8.3 | 9.4 | 10.9 | 0.9 | 1.1 | -1.5 | 0.9 |
| More than 10 | 9.9 | 13.4 | 13.2 | 13.1 | -3.5 | 0.7 | 0.1 | 1.0 |
| Number of blacks in current or previous work | ||||||||
| 0 | 23.5 | 33.3 | 26.3 | 30.6 | -9.8 | 0.7 | -4.3 | 0.9 |
| 1 | 11 | 8.2 | 14.6 | 8.6 | 2.8 | 1.3 | 6.0 | 1.7 |
| 2-5 | 28.3 | 25.9 | 29.2 | 24.5 | 2.4 | 1.1 | 4.7 | 1.2 |
| 6-10 | 16.2 | 11.6 | 11 | 12.3 | 4.6 | 1.4 | -1.3 | 0.9 |
| More than 10 | 21 | 20.9 | 18.9 | 24.1 | 0.1 | 1.0 | -5.2 | 0.8 |
To replicate:
Row: RACMAR MARBLK ACQFMBLK RACHOME BLKSIMP TRTBLACK ACQBLACK DISCAFF ACQNHBLK RACPRES MOSTCOM ACQVABLK ACQWKBLK
Column: POLVIEWS(r:1-3″Liberal”;5-7″Conservative”) PARTYID(r:0-2″Democrat”;4-6″Republican”)
Selection Filters: RACE(1) HISPANIC(1)
For those who don’t know the GSS URL: http://sda.berkeley.edu/cgi-bin/hsda?harcsda+gss08
The question of Hispanic or non-Hispanic status was only asked starting in the year 2000, so the data are all constrained to the aughts. I know tables are kind of hard to read, but I wasn’t sure as to the best way to visualize the results. But if someone wants to try, or has some ideas, here’s the data as a csv.
I’ll let readers engage in interpretation, but be warned that if it’s obvious you didn’t read the table your comment may not be published, or, I’ll just delete it. The only thing I want to add is that it isn’t a surprise that the political party division is narrower than the ideological one. Republicans are the conservative party, but there are wealthy social liberals within the party, while Democrats have some downscale socially conservative types.
Note: Sample sizes are small in some of the cases above, so don’t necessarily draw too much of an inference if the absolute value difference is marginal. That’s why I looked at a lot of questions.