The Agenda

NRO’s domestic-policy blog, by Reihan Salam.

Tea Partiers and the Campbell-Putnam Feelings Thermometer


In my post on David Campbell and Robert Putnam, I asked the following question:

I’d love to see the survey questions and results that led Campbell and Putnam to conclude that Tea Partiers have “a low regard for immigrants and blacks.” 

This was a very lawyerly question, which suggested skepticism on my part. But as my friend Gabriel Rossman, a sociologist at UCLA, pointed out to me, Campbell and Putnam were actually quite careful about the claim in question:

I’m not positive, but Putnam and Campbell seem to be referring to this dataset, which I was not previously familiar with). 

Heere is the 2006 baseline survey instrument, which contains the racial feelings questions.

I’d like to get your feelings toward a number of well-known groups. I’ll read the name of a group and I’d like you to rate that group using something we call the feeling thermometer. Ratings between 50 degrees and 100 degrees mean that you feel favorable and warm toward the group. Ratings between 0 degrees and 50 degrees mean that you don’t feel favorable and don’t care too much for that group. You would rate the group at the 50-degree mark if you don’t feel particularly warm or cold toward the group.  Feel free to use the entire extent of the scale.

[Among the groups listed: gays and lesbians, blacks, whites, Asian Americans, Latinos, Jews, Catholics, Muslims, Evangelical Christians, etc.]

As you can see, they did not do the common thing of relying on questions about policies to ameliorate racial inequality (e.g., busing, “too much to support blacks,” etc), but rather asked a general “feeling thermometer” and a 4 point Likert scale on trust. These kinds of questions do not directly involve government action or policy and there’s no reason in principle that a devoted small government conservative couldn’t feel warmly towards and trust blacks. What’s more, these questions are vulnerable to social desirability bias and so probably represent a lower bound estimate of racial animus. Also note that they claim that the results are controlling for party ID and race, so it doesn’t seem like the analysis is skewed. On this limited evidence, I’m inclined to say that it’s a fair dataset and a fair interpretation, though i’d want a lot more details if I were a peer reviewer.

My initial skepticism was not warranted. Campbell and Putnam do appear to have fairly represented the views of survey respondents, which, as Gabriel observes, “are vulnerable to social desirability bias.” 

I would like to have a better sense of the sentiments behind the feeling thermometer result, and I am curious as to how this relatively low regard manifests itself in practice, e.g., does it reflect negative experiences with members of outgroups or does it reflect a lack of exposure of members of outgroups. In The Ordeal of Integration, now very dated, Orlando Patterson found that whites who held anti-black sentiments were often those who remained in cities with large concentrations of African Americans in the years after the white flight era. Yet we also have suggestive evidence that the Americans most likely to hold anti-Muslim sentiments are those who don’t know any Muslims.   

Campbell and Putnam’s finding will reinforce the belief that racial animus animates small government sentiments. This is tough to prove, and I’d want to know much more before I accepted the premise. But this is a data point that merits at least some attention. The most obvious and mercenary reason is that the demographic composition of the U.S. electorate is changing, and those of who care about the size and scope of government need to find allies from all American groups, including the overlapping groups I belong to personally. 


Sign up for free NRO e-mails today:

Subscribe to National Review