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.
Razib Khan’s degrees are in biochemistry and biology. He has blogged about genetics since 2002, previously worked in software development, is an Unz Foundation Junior Fellow and lives in the western US. He loves habaneros.
September 7th, 2010 at 12:57 pm
strangely, the first comment was angered that
1) i didn’t bring up the issue of black racism
2) accused me of censoring the comment (i.e., moderation for first timers) because i was a democrat
obviously the moron didn’t read the post, as i’m not a democrat, and in general the post is i think somewhat more sympathetic to white conservatives than not i think. but, i guess the commenter was angered that i agree that white conservatives seem to exhibit on average more racialist sentiment than white liberals.
anyway, again, even if i’ve approved you before if it’s obvious you haven’t read the post, or try and take the thread off topic, i’m going to delete you. i did some “leg work” putting the data together, and i’m more interested in non-moron interpretation, from various political perspectives. but i’m also totally opening to just closing the thread if the stupidity gets on my nerves.
September 7th, 2010 at 12:57 pm
The difference between a racist and social activist is trivial: A racist finances his snit from his own wallet. A social advocate pays for her snit by robbing your wallet.
“Stupid” does not require an agenda. “Stupid” is a self-priming pump.
September 7th, 2010 at 1:05 pm
[...] This post was mentioned on Twitter by razib khan and Ron Simon, Vijay Vikram. Vijay Vikram said: haha! This is brilliant @Razibkhan. How much more racist are white conservatives than white liberals? http://bit.ly/dzTwh5 [...]
September 7th, 2010 at 1:18 pm
ok, can someone offer observations on the patterns in the data before going off into an aside? or is it just too hard looking at data when you’ve got prefab opinions to give?
September 7th, 2010 at 1:28 pm
Liberals do slightly better than conservatives, but Republicans do slightly better than Democrats.
(At least for the first few questions, that I could easily interpret.)
September 7th, 2010 at 1:40 pm
Scanning the table, I don’t see major or even consistent differences along either the ideology or party axis. In particular, focusing on the signs of the absolute differences, I see a roughly equal number of plus and minus signs in both axes. Without running a stat test on the frequencies of plus and minus signs, I would conclude that racist and antiracist attitudes are roughly equally distributed.
September 7th, 2010 at 2:05 pm
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.
The composition of the group being out of the control of the group is equally applicable to conservative and tea party movements. For instance, if the composition of the tea party groups is mainly characterized as people who view themselves as funders, rather than recipients, of government programs, and they are proposing race-neutral policies and elimination of race specific policies then this political viewpoint might not appeal to voters who are recipients of government programs and who benefit from race-directed policies. If the Tea Parties want to be race neutral and that message doesn’t appeal to people who want the government to be race sensitive, then I find it ironic that there is an implied reference to racism directed at the Tea Party people rather than directing that charge to those who reject race-neutral policies.
September 7th, 2010 at 2:23 pm
As to the data y0u’ve assembled, it appears to me that there is a little variance in behavior. What I find interesting is the difference in self-image versus behavior. For instance, on the question of the “number of blacks in the neighborhood” those who identify as Democrats show that 55.4% live in neighborhoods with no blacks compared to 59.3% of Republicans. Now, if Republicans are tagged as being more race-conscious then we would expect them to live up to that reputation and seek out white neighborhoods, but what’s going on with Democrats? They should be seeking out multiracial neighborhoods in order to live true to their philosophy and yet they’re acting little different than Republicans. What this looks like is a trade on reputation – they talk the talk, and score reputation points for being “enlightened” and yet very many of them fail to walk the walk. I do note that at the other extreme, 10 or more Black people in the neighborhood is when we see a marked 2:1 difference in favor of Democrats, but we’re only dealing with 10% of the respondent pool.
It makes me wonder on what basis the perception of greater liberal race tolerance is founded? Do the outliers accurately describe the behavior of the entire group? Is it the difference in the expression of attitudes, even if behaviors don’t match expression, that forms the truism?
September 7th, 2010 at 2:28 pm
What I find interesting is the difference in self-image versus behavior.
yeah, i think this is an interesting angle to to take. revealed vs. avowed preferences. one would have to account for confounds. e.g., whites in vermont are very liberal, while those in mississippi are very conservative. but obviously there are different racial proportions, and outside of mississippi’s far north the whites in that state will interact with blacks on a day to day basis.
though i think it is better to focus on the ideological difference, not party id. there are as i noted above a lot of wealthy socially moderate/liberal republicans, and downscale socially moderate/conservative dems. that pushes the two groups further together than if you sort by ideology.
September 7th, 2010 at 2:32 pm
I understand the greater difference between ideology and party – it’s quite clear that liberals are less racist than conservatives – but why the flip-flop on party (if we have the prior assumption that liberal would ordinarily be Democrat and conservative would ordinarily be Republican – that’s not the best way to put it, but oh well)?
Can you run these numbers again with regard to other indicators of racist attitudes/party affiliation?
September 7th, 2010 at 2:36 pm
well, correlations aren’t transitive always. i.e., if A correlates with B and B correlates with C, it does not necessarily entail that A correlates with B.
as i said, a certain number of high income whites vot repub despite social attitudes more in line with liberals, while a converse number of low income whites vote dem despite more conservative attitudes.
Can you run these numbers again with regard to other indicators of racist attitudes/party affiliation?
can you elaborate?
September 7th, 2010 at 2:36 pm
There is definitely something interesting going on, but mostly I think racism is not something that is particularly strongly embedded in most American’s personal self-reportable mindset, with some mostly geographical (and low population) exceptions…most of us are likely not good judges of our racial attitudes.
Tango man, less than 15% of the US is black, and it is not evenly distributed. I don’t think a survey like this can measure whether people “want” to live near black people based on if they do live near black people. But your knee-jerk perception of liberals as hypocrites based on this dubious data point is interesting in its own right.
Finally, it is a known Fact that republicans lie like the sociopathic regressives they are, especially on surveys, and especially about how many black people they know. Do I win the Most Hysterical Aside?
September 7th, 2010 at 2:36 pm
Also, slightly off-topic – Razib, did you ever do a post measuring the separately measured social and economic opinions of members of both major parties?
September 7th, 2010 at 2:38 pm
Can you run these numbers again with regard to other indicators of racist attitudes/party affiliation?
can you elaborate?
Socioeconomic status, etc.
September 7th, 2010 at 2:45 pm
did you ever do a post measuring the separately measured social and economic opinions of members of both major parties?
yeah, at some point i did. discover’s search engine kind of sucks, so i recommend google
most of us are likely not good judges of our racial attitudes.
there is “implicit attitudes” stuff which you can look up. that’s the main reason why i would be willing to accept that white conservatives have more racialist feeling than white liberals, though i have friends who are skeptical of the IAT method (liberal friends btw).
. I don’t think a survey like this can measure whether people “want” to live near black people based on if they do live near black people.
you’re right about the GSS not being a good measure, but as i’m sure you know there is an issue of residential race segregation in cities which have liberal and conservative white populations along with black populations. the point i’m getting at in my post is that in the case of segregation along racial lines where the white population is politically conservative i think the prima facie assumption is that it’s driven by racism. but in the case where the white population is liberal (e.g., berkeley) i think most readers of this weblog would point to various socioeconomic confounds.
September 7th, 2010 at 2:47 pm
katherine, so you’re curious about the numbers if i ran socioeconomic status, e.g., in the column?
September 7th, 2010 at 3:00 pm
I don’t think a survey like this can measure whether people “want” to live near black people based on if they do live near black people.
Razib provided many different questions which we can put on a scale of revealed versus avowed preferences. Some questions merely ask about attitudes (relatives marrying someone who is black), others ask about behaviors with no real immediate impact on one’s life (vote for a black president), one ask about the associations one freely enters into and one asks about one’s neighborhood.
If your objection to a non-uniform distribution does explain away the difference in behavior then what’s going on with freely chosen associations? Having friends and acquaintances of different races isn’t so dependent on all the factors that influence housing choice. Yet we see 41.2% of liberals associate with no one who is black compared to 43.4% of conservatives. Then in the 1-10 range of black associates, liberals and conservatives are nearly identical in representation and conservatives start to open up the gap in the category of 10+ black associates.
Really, what is the basis for the belief that liberals are more racially inclusive? It is tautological – liberals are less conscious therefore liberals are less race conscious, or is it that liberals talk a good game but don’t live true to their principles, or is the data in support of the proposition found in sources that Razib didn’t reference and if so how do we reconcile the discrepancy between GSS data and this other data set?
September 7th, 2010 at 3:01 pm
Interesting, but perhaps over complicating the issue. There has to be racism, too much funding and too many liberal arts degree jobs depend on it. It’s not like they can go work at the “Sociology Factory”. They need unsolvable problems to dedicate there lives too or they don’t have work. So if there has to be racism, there has to be racists. They are not going to call themselves racists, so it has to be the other “scary” white people. Remember, if there are no witches, there is no need for witch hunters.
“It is difficult to get a man to understand something, when his salary
depends upon his not understanding it!”—- Upton Sinclair
September 7th, 2010 at 3:09 pm
Really, what is the basis for the belief that liberals are more racially inclusive?
historically white liberals supported the rollback of jim crow legislation, while white conservatives did not. do you really not understand that that’s probably the big background variable which shapes perceptions of different behavior, or are you just trying to make a good argument for a maximalist position of no differences in race bias?
September 7th, 2010 at 3:12 pm
ok, i’m closing the thread to save everyone’s time. the comments are kind of boring retreads as people dig in and support their own side. the engagement with the data seems symbolic and cursory at best. some of the comments are coherent and valid, but they’re shit i’ve seen elsewhere, and don’t need any support from new data.
i’ll post an update on this post with other variables later katharine. hopefully this post will be useful on google queries for the original data.
peace.
update: a link from reddit. might have missed a bullet re: comments
update 2: i put up a post at secular right if people still want to discuss this topic. i won’t be moderating too closely so no skin off my back.
September 7th, 2010 at 4:00 pm
[...] analyzed some GSS data over at Discover. The commenters were only cursorily engaging the data, and I don’t have much patience for [...]
September 12th, 2010 at 3:55 am
[...] of the Living God“, “Genetic Differences Within European Populations“, “How Much More Racist are White Conservatives Than White Liberals?“, “Sexual Selection: Lowered Expectations [...]