Bryan Caplan  

Do Liberals Use Five Foundations After All? A Question About Questions for Jonathan Haidt

Liberals and Markets... Your Public Education at Work...
Will's latest blogging on Jonathan Haidt's "five foundations" theory of morals inspires me to publicly ask Haidt a question that's been bugging me: How hard did you try to include items about Ingroup/Loyalty, Authority/Respect, and Purity/Sanctity that would specifically appeal to liberals?  Examples that occur to me:

Ingroup/Loyalty: How much would somone have to pay you to vote Republican in an election if you knew the Democrats would win for sure?  To cross a picket line?

Authority/Respect: How much would someone have to pay you to privately ask Clinton embarrassing questions about the Lewinsky affair?  To dance on Martin Luther King's grave when no one was looking?

Purity/Sanctity: How much would someone have to pay you to throw one McDonald's cup out of your car window in Yosemite?  To scream racial epithets in a sound-proof room?

I could be wrong, but when I read your research, it seemed like such questions were rarely if ever asked.  If you included them, I suspect that you would discover that liberals, like conservatives, use all five moral foundations in a serious way.  The liberal-conservative difference, in other words, is largely about which group they identify with, which leaders they respect, and what they consider sacred.  Am I mistaken?

Comments and Sharing

COMMENTS (22 to date)
Contemplationist writes:

Excellent questions, Bryan.

I think for lefties the most sacred concept is that of the 'Environment' untainted, pure, holy.

sampleminded writes:

Great Post! But I think you are proving too much. The five factor theory doesn't hold that liberals never use the other 3 factors, just that they use them much less, I imagine in survey data there is a continuum. The fact that you can design a situation more likely to evoke them doesn't mean that those situations are very common or relevant to topical issues.

Also, maybe I'm an outlier as liberal/libertarian I really would require very little money for any of them. I think the five foundations theory mostly stands. The hardest for me would be polluting Yosemite, but I don't think one person doing this would result in much harm so I'm not going to stress to much about it. Look here I am thinking about the harm it will do, not the purity aspect. Those five foundations are still standing as far as I'm concerned. I also personally have no sense respect for Authority, I know I'm not normal in that regard.

dWj writes:

Intuitively, I think liberals think in terms of group loyalty far more than conservatives do. The people who obsess over racial, gender, and other identities seem reliably to be on the left. Maybe I completely misunderstand what he means by this factor.

Les Cargill writes:


First, only harm/care and fairness reciprocity ( is there a difference here ? ) make any evolutionary sense at all. The others are essentially cognitive biases (which are evolutionarily sound, but are made obsolete by modernity, and are therefore liabilities) dressed up like "morals". When I read the word "purity", I thought of "purity of essence" as used in "Dr. Strangelove". Not that this does not happen, understand, but it's very creepy when it does. The last three items *literally* form the basis of forms of "kink". And seems like twice a year, some famous Republican ends up falling out of the closet in a humiliating manner. At what point to cognitive biases turn into cognitive dissonance?

If there is such a thing as a conservatism, it is as Burke laid it out. This includes skepticism about progressivism and respect for tradition. Well, we're past hoping to have real conservatism these days. There is an argument that says that we live in something post-Western Civilization - we no longer have religion as the core of our society, and that makes this quite different from 50 or 70 years ago. Please note that I am not advocating religion, just observing a dramatic change ( actually repeating the observations of somebody else, whose name escapes me).

I may well be wrong, but the most excellent "From Poverty to Prosperity" by the esteemed Dr. Kling and "We Are Doomed" by J. Derbyshire touch on the apparent doom of our system - it works too well. We can't manage to distribute wealth well enough to keep from having negative feedback loops that essentially silence it. So we try to redistribute, which I would actually support if it was to work, and this not only fails, but *apparently* makes things *worse*. It may well exacerbate income inequality.

It is increasingly clear to me that we are complete and total prisoners of our cognitive biases, The unconscious of man will never be dominated by the conscious mind.

As I also understand it, there is a hypothesis that 1) banking panics are a product of pretty much the Bank of England and the US 2) these banking panics led to apocalypticism and what we call "fundamentalist" religionism in politics, and 3) that fundie-ism now stands athwart our political system with a rope around its neck. Since the fundies were having so much fun getting their cognitive-bias freak on, the Progressives adopted the same hyper-surreal tactics. The resulting cognitive dissonance is literally like a drug. Since we have in effect banished any negative feedback to these insanities through the rapidity of economic growth, only something like the Confucian Revolution in China could hope to interrupt the flow. And this may well be the future - who is the rising power, again?

So no, the five platforms don't do much for me. Sorry.

Justin Martyr writes:

This quote from Arnold's post cracked me up:

And if research shows that liberals are relatively indifferent to hierarchy and authority, who am I to question that? After all, this is only a blog post, not a peer-reviewed journal. And I don't have tenure.

And it goes against consensus!

zeljka buturovic writes:

i think you are right when it comes to behavior, judgments and intuitions. in other words, i think revealed preferences for different foundations are pretty much the same for liberals and conservatives, though, as you point out, relevant aspects are quite different between them.

however, i think liberals are less likely to use authority, purity and loyalty to justify their judgment and behavior. they will make more effort to cloth their purity and authority intuitions into a theory of harm and care. and conservatives don't feel compulsion to that to the same extent. that is why they appear different to some people.

JPIrving writes:


I have puzzled of this myself. I am sure many readers of this blog have damaged friendships/their reputations with lefties by expressing even mild skepticism over global warming/obama/whatever. (I know not all libertarians are GW skeptics..)

In my experience the left is much less tolerant, I remember debating religion with religious conservative friends back in my military school days. It got heated but no one took it personally. Do that with a progressive on something sacred like GWarming and many will never completely forget your blasphemy.

Am I a victim of small sample bias? I have lived mostly in Vermont and Sweden so my associations are certainly more with progressives.

aretae writes:


I asked him similar questions in private email about 18 months ago. No response. Since your post, I blogged about it.

Mike Gibson writes:

Nice Bryan. I've had similar thoughts while watching eco-worshipers cry before redwood trees.

If you look at Haidt's book proposal for his work on political psychology, he says he'll offer a "field guide" of sub-groups within liberalism, including greens and populists. So perhaps he will answer your questions indirectly with the true scotsman defense.

The Righteous Mind book proposal:

eccdogg writes:

I think the surveys are not set up well to measure libertarians.

I took some of the questionaires and found many things hard to answer or likely to be misconstrued.

For instance the questions about how important fairness was to me. I said it was very important, but I was refering to fairness of process not fairness of outcome.

There were many quesions that did not allow any such distinction.

Adam Ozimek writes:

Bryan, are you going to provide us with your answers to these survey questions? Or are you worried they'll never let you near Yosemite or Martin Luther King Jr's grave once you reveal how low your prices are?

Justin Martyr writes:

I've been thinking some more on this. I believe that Bryan's criticisms are sound and that liberals do use more of the five foundations. But I still think there are differences and I am mostly channeling my own experience as a former liberal atheist who is now a conservative Christian.

When I first became a Christian I had a thoroughly secular worldview but at the same time, like many conservatives, began to strongly feel the need for all five foundations (although I did not put it that way). It was the first time that I started to consider the possibility that the changes to society might not be good and might even be quite harmful. But I did not have the vocabulary to describe it and would grow quite frustrated.

Now I do, but it is primarily in terms of externalities (harm to others, perhaps in the form of weaker institutions and eroded social norms) and hyperbolic discounting (harm to ones future self). In that sense I have to wonder if my childhood conditioning is still at work and I still only lean on the same two foundations as liberals.

Snorri Godhi writes:

Brilliant post: I too wondered about that.
And brilliant comments too, most of them.
In particular, I second zeljka buturovic:
i think liberals are less likely to use authority, purity and loyalty to justify their judgment and behavior. they will make more effort to cloth their purity and authority intuitions into a theory of harm and care.

I'd like also to express my support for JPIrving:
In my experience the left is much less tolerant [...] Am I a victim of small sample bias?

On the basis of my experience with Ivy League "liberals", Albertan rednecks, Italian fascists, and Italian communists, the "liberals" stand out as the most likely to use ad hominem arguments. That is based on a small sample, but I note that the Italians who voted for fascism or communism in the last century are so open-minded that they have accepted reality, and stopped voting for fascism or communism. Meanwhile, the "liberals" have become more extreme, and are still whitewashing communism (and themselves).

8 writes:

For politics, shouldn't the question really be "what would you pay to forbid someone from doing X?" In your town? Your state? Your country?

For conservatives: gay marriage, drugs, prostitution, pornography.

For liberals: carbon emissions, guns, teach creationism in a public school, logging

Steve Roth writes:

Are you saying that the surveys consist primarilly of the opposite--questions that will revolt Republicans in particular? Examples?

Troy Camplin writes:

Sounds right to me.

Jason Malloy writes:

Haidt's survey website has just added a "liberal purity" test, and this does reveal a substantial increase in the importance liberals place on purity. But conservatives are also much more likely to agree with liberal purity ideas than liberals are to agree with conservative/general purity ideas. Suggesting conservative feelings about purity are much more generalizable.

The same probably applies to authority and loyalty. But it still seems dubious to claim that liberal morality is "almost exclusively" based on two of the five moral dimensions.

mulp writes:

Well, your form of query isn't like those he uses, would you include queries to separate out conservatives like these:

How much would someone need to pay you to abort your unborn child?

How much would someone need to pay you to have sex with your 16 year old child?

How much would someone need to pay your to welcome a Muslim or atheist into your family by marriage?

How much would someone need to pay you to convince your child to be a suicide bomber?

After all, conservatives seem far more interested in getting rich than liberals, and generally consider getting rich a virtue, so how important are other value relative to increased wealth?

Troy Camplin writes:

I think the primary difference between liberals and conservatives has to do with values ranking.

For example, suppose that you had one person with the following values, ranked in the following order:

economic growth
preventing crime
eliminating poverty
clean environment
animal welfare

Now, suppose that you had another person with those same values, ranked in the following order:

clean environment
eliminating poverty
animal welfare
economic growth
preventing crime

What is the political position of each person? How might they differ in policy on creating a clean environment? on green jobs? on how to eliminate poverty?

Steve Roth writes:

Jason Malloy: "Suggesting conservative feelings about purity are much more generalizable. "

I would say more generalized.

"Generalizable" is pretty normative; it suggests greater accuracy, applicability, etc.

Jonathan Haidt writes:

Dear Bryan, and other commentators:

The question you ask is one of the most important ones we've been trying to answer for the last 2 years. The original theory was not developed to understand politics; it was to understand cross cultural variation, while drawing on evolutionary psychology to help pick the best candidates for being true foundations upon which cultures can construct many contradictory moralities. That's how we came up with the first 5. But once we applied it to politics, it quickly became clear that we were missing something about liberty/autonomy, and that fairness was much more complex than concerns about justice and equality (which liberals score higher on). As one of your readers commented, we do poorly by libertarians. But we're about to fix that. If anyone wants to see the data as it grows, on the various kinds of liberty and various kinds of fairness, please go to and take the "MFQ-Part B"

NOw, as to whether liberals have the ingroup, authority, and purity foundations at all: As one of your readers said, it's a matter of degree. So I've always thought that they have them, but don't build nearly as much on them. But the story is different for each one:

INGROUP: yes, liberals can do ingroup, but mostly just contra conservatives and racists. And they don't do it terribly well. The Democratic Whip has a much harder job than the republican Whip. Social conservatives take to it so readily. Liberals and libertarians can do it, but not as readily or as reliably. Liberals in particular are universalists; they are morally opposed to tribalism, although they can kinda do liberal tribalism. So yes, liberals would consider voting for a republican as a kind of treason.

AUTHORITY: This is the one that I think really is different. Many liberals tell me that we have authorities, but our authorities have to earn our respect, like a scientific authority. But i see this as something of a pun. The ethology of authority is related to dominance and submission, most primates do it, but conservative primates do it much more readily than liberal primates, and on the far left anti-authoritarianism is such a strong value. Dancing on MLK's grave is extreme sacrlilege (see purity), it is not defying an order, defying the teacher, father, etc. I think this foundation might be one that some liberals lack completely, others have weakly.

PURITY: I have long thought that liberal purity exists and is best found in liberal attitudes about the environment. I have a short blog post titled 'in search of liberal purity' here:

So yes, your question about littering is a very good one, i might test it out if you give me permission. To see the current items that we are testing, please go to and take the MFQ-C, which has items we are using to explore liberal purity.

Bottom line: Moral Foundations Theory, in its first draft, has done a surprisingly good job of capturing the culture war, particularly the old one with the religious right. But it is incomplete, it is constantly being improved, and questions and criticisms such as yours are one of the most important ways that we improve it. We're likely to come out with a revision in late 2010, based in part on what we find on the MFQ-B and C.

Thanks for posing these questions, and inviting me to respond.


Comments for this entry have been closed
Return to top