Bryan Caplan  

Testing Unflattering Claims About Human Motivation

Resentment, Information, and U... The Merits of Cochrane's Case...
"Democrats resent the rich."  "Republicans disdain the poor."  When accusations like this fly, the accused tend to be coldly condescending.  "If you knew a single Democrat/Republican/whatever well, you'd know how ignorant your claim is."

Personally, I believe that 80% of everyone's accusations are true.  Of course Democrats resent the rich.  Of course Republicans disdain the poor.  But maybe I'm just being contrarian.  Is there any fair-minded way to advance the conversation?

I think so.  A vast number of surveys include Feeling Thermometers.  A typical wording from the General Social Survey:
I'd like to get your feelings toward groups that are in the news these days. I will use something we call the feeling thermometer, and here is how it works: I'll read the names of a group and I'd like you to rate that group using 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 toward the group and that you don't care too much for that group.
My prediction: Standard Feeling Thermometer questions will support at least 80% of common motivational accusations in relative terms.  Democrats will have noticeably colder views of the rich than Republicans.  Republicans will have noticeably colder views of the poor than Democrats.  And so on.

My word of honor: As I write, I have never played with this data.  I am making an honest-to-goodness Assert First, Look Afterward prediction.  My question for readers of all factions: Before you know the answers, are you willing to admit that Feeling Thermometers are an informative way to put unflattering claims about human motivation to the test?

If you have any intellectual reputation to bet, please "go on the record" in the comments.

COMMENTS (18 to date)
As written writes:

As written: "Democrats will have noticeably colder views of the rich than Republicans. Republicans will have noticeably colder views of the poor than Republicans. And so on."

No opinion on feeling thermometers, but I think I'll take the other side of your prediction because no amount of playing with the data is going to make "Republicans have noticeable colder views...than Republicans" ;)

David Jinkins writes:

What is the point of betting when the data is probably pretty easy to find? Is the idea to do an "out of sample" test for how accurate your predictions are, so that we can be more confident when your predictions cannot be tested?

Here is something related I whipped up from the world value survey website. It reports beliefs about the cause of success by political affiliation.

As you might expect, the point estimate is that republicans are more likely to attribute success to hard work rather than luck.

Pajser writes:

Yes, without looking at the data, I believe that Feeling Thermometers will turn to be informative.

Josiah writes:

I think it will be informative, though not definitively so.

ThomasH writes:

I hope your 80% guess is too high, but the point is, making policy directly on the basis of even 1% by either Democrats or Republicans is a mistake.

If one wants to tax the consumption of the rich it should not be from resentment, but rather to raise revenue or cut taxes on the consumption of the poor. If you want to reduce income transfers to the poor through program x it ought to be that the effects of program x on recipients is harmful.

The role for feelings would be in how much one is willing to reduce the consumption of the rich by in order to increase the consumption of the poor. This is my (probably right-wing) understanding of Catholic social teaching's "preferential option for the poor."

Emily writes:

I don't think the claim "Democrats resent the rich"="Democrats resent the rich more than Republicans," but rather something closer to "there is not much overlap between how much Democrats and Republicans resent the rich." Like, if we only knew how much someone resented the rich, that would be an extremely valuable piece of information in terms of being able to guess whether they were a Republican or Democrat.

I accept this as a way of testing those claims. And I don't they are likely to be accurate. I anticipate a great deal of overlap between groups.

JLV writes:

My feelings toward feeling thermometers are warm. But I doubt they will work as a test in the way you claim. Feeling therms don't match to English phrases closely enough to really settle debates.

Let's suppose that Republicans really do have colder views about the poor. What you won't hear, I think is anyone saying "OK, you got me, we really do hate the poor!" Instead, the response is likely to be more like, "we don't disdain the poor, we just believe that the welfare state has created a bad culture of dependence and poor family values. We really disapprove of the welfare state."

Alternately, the response could be essentially #notallmen.

MG writes:

I would be more interested in this: Observe the actual treatment of an actual person. Then repeat. Then aggregate. Bottoms up. That is data. In my own experience the results contradict the expectations.

Annabelle Smyth writes:

I agree with MG. I'd rather observe and report rather than have people self-report on their perceived actions.

Austin Middleton writes:

"Don't think about elephants."

And now you're thinking about them, aren't you.

Point being, the feeling thermometers might show a pattern in the prompted response, but their response might not have any relationship to motivations of actions made prior to the prompt in the same way you weren't thinking of elephants prior to reading this post. Which means all you might (are likely to) detect is a festival of social desirability bias.

The Original CC writes:

+1 for MG

Bryan wrote:

My prediction: Standard Feeling Thermometer questions will support at least 80% of common motivational accusations in relative terms.

I'm not sure how useful this is. Democrats accuse Republicans of just outright disdain for the poor; it's not like they're saying "you love the poor less than we do". Ditto for the opposite case.

Seb Nickel writes:

I second Bryan's prediction.

John Z writes:

I think Emily is exactly right. Put another way, how much of a difference is there between the two groups?

I think a chart showing the distributions of Rs and Ds is the right way to do it. I also think there will be substantial overlap

Cole writes:
"Democrats resent the rich." "Republicans disdain the poor." When accusations like this fly, the accused tend to be coldly condescending. "If you knew a single Democrat/Republican/whatever well, you'd know how ignorant your claim is."

I think the response to you is interesting. I don't often get that response. I've seen outright acceptance of the accusation, or just a slight 'narrowing' of who they hate. Like 'I only disdain the lazy poor who do nothing to get out of their situation'. Or 'I just resent the rich that got rich at other people's expense, or the rich that spend money to keep other people poor'. The criteria for narrowing down the hate often seems pretty arbitrary and doesn't actually seem to narrow down anything at all.

I wonder if people are just trying to use a 'you are in an ivory tower, so you dont understand the real world' argument on you. I think proving your observations about democrats and republicans will probably just get you to the step I'm at where they either justify the hatred, or pretend to narrow it to only the bad actors. Maybe the people you talk with are just at a better level where they have to tacitly agree that hating someone for their level of wealth is a bad thing. The people I talk to seem to think its fine to hate people as long as you hate the right people.

austrartsua writes:

feelings shmeelings. Republican's might feel colder towards the poor than democrats and vice versa for the rich, but that doesn't mean democrat policy is better for the poor or worse for the rich. Good intentions mean little.

Scott Alexander writes:

I doubt anyone denies that Democrats are colder toward the rich than Republicans; the "Democrats resent the rich" claim, if it's controversial, seems to be controversial as an *explanatory* claim.

Suppose some Democrat says she thinks rich people are parasites. You can have the causal arrow go one of two ways:

1. Democrat feels secret emotional resentment toward the rich, this comes out as a well-developed economic theory in which the rich are parasites.

2. Democrat uses pure reason to derive a well-developed economic theory in which the rich are parasites, knowing that they're parasites makes her very logically resent them.

I think claims like "Democrats resent the rich" are claims for (1) rather than (2); if that's true, the Feelings Thermometer isn't going to help much.

Miguelito writes:

I wonder how these thermometers reflect the views of people like me, for whom the whole notion of having a sentiment towards a group is repulsive -- the definition of bigotry. Unless it is a group defined by behavior. Are we really so few that surveys designed as though we don't exist are good public opinion measures?

LD Bottorff writes:

If Democrats resent the rich, why do they keep voting for them?. Check out Roll Call's list (
Democrats do not resent the rich. However, their policies make it difficult for those who want to become rich.
As a Republican I can tell you, I do disdain the poor; I wish there were fewer of them. I want them to graduate to the middle-class (or better).

Comments for this entry have been closed
Return to top