Skip to main content

Verified by Psychology Today

Jonathan Haidt's Moral-Political Psychology

Helping us understand liberal and conservative value systems.

Key points

  • Jonathan Haidt argues that liberals and conservatives largely differ when it comes to five moral value systems.
  • Haidt warns that disciplines like psychology may have massive biases against conservative viewpoints.
  • While Haidt raises important questions, he misrepresents the moral values of liberals.

As evidenced by Congress's historically low approval rating and the most recent debacle with the payroll tax, we are witnessing a period of remarkable political dysfunction, characterized by bitter disagreement and polarization between the major parties.

Although there is plenty of blame to go around, along with commentators like Paul Krugman, I believe that the lion's share of the fault lies with the Republicans. As seen in the weak group of presidential candidates, the Republican Party does not have a compelling narrative for the future. Instead, I believe it has largely been hijacked by uncompromising, ideologically driven, ignorant extremists who are lacking in the capacity for sophisticated self-reflection.

But is my narrative simply a function of living within a liberal elite "moral matrix"? Am I blinded by in-group biases and is my frustration with Republicans simply a mirror image of Republican frustration with liberals?

As a marriage therapist, I am quite familiar with how competing justification systems arise and feedback on one another through vicious cycles. That is why it is imperative to be as reflective as possible when such destructive positions emerge. And it is in that spirit, and continuing with analyzing leading developments in the field, that in this post I am going to comment on Jonathan Haidt's (pronounced HITE) in moral-political psychology as articulated in this TEDTalk on the moral roots of liberal and conservative politics.

Haidt's key messages

Haidt's talk contained several key messages. First, citing Steve Pinker (another hugely influential psychologist), Haidt argues that the blank slate "was one of the worst ideas psychology ever had" and it is now clear that "the mind" comes with much organization prior to experience, although that organization is plastic and molded by experience. [I generally agree with this statement, although I might want clarification about what exactly is meant by experience (e.g., surely there is prenatal experience, and that might be quite crucial)].

Early in his talk, Haidt mentions another key point, which is that liberals and conservatives demonstrate large differences in Openness to Experience, which most personality researchers consider to be one of the Big Five personality traits. Although I was aware of this finding, as I listened to his talk I wondered about the relationship between the trait of openness and political values.

Here is the issue: Traits, at least as conceived of by the major trait researchers Costa and McCrae, are seen as almost exclusively determined by genes (although it would be the subject of another post, Costa and McCrae consistently argue this point; I, however, would challenge their characterization as over-exaggerating the genetic case for traits). Political values are transmitted largely via the family (I assume this is true but would have to explore the research on it). So I could imagine a very interesting adoption study where the trait of openness of the biological parents could be analyzed and compared to the political ideology of the adoptive family in predicting the adopted individual's political values. If anyone knows of research on this topic, I would love to hear it.

The third and primary empirical point of Haidt's talk is about his own research on the five moral value systems that he argues underlie the liberal-conservative political dimension: 1) Care for Others/Do No Harm; 2) Fairness/Justice/Equality; 3) In-Group Loyalty; 4) Respect for Authority; and 5) Purity. His research shows across large numbers of people and many different countries that there are very reliable differences in the degree to which liberals and conservatives differ in the extent to which they endorse these values. Conservatives tend to value the five domains relatively equally. Liberals, in contrast, value the first two domains much more than the latter three.

Haidt's final point is we need metacognitive awareness about "the Moral Matrix". Speaking to a group dominated by liberals, Haidt argues that we should be very aware that such a group (and disciplines like psychology) likely have massive biases against conservative viewpoints. Moreover, Haidt argues (or implies), that conservatives actually have a more complicated moral system, consisting of all of the five values, whereas liberals are dominated by just two. He asks that we step outside our systems and understand the other point of view.

Positives about Haidt's work

There is much about Haidt's work that I like, not the least of which is that it opens up large doors to ask important questions about the relationship between psychology and politics. Consider, for example, the question: Why are so many psychologists liberal? Does psychology have an anti-conservative bias? Or is conservatism defined against psychology in some way? (Note: If you are curious, Steve Quackenbush and I wrote an article on the interface between Clinical Psychology and Politics). If psychology is going to have a large-scale impact on how humans think and govern themselves, this is a crucial question the field ought to wrestle with.

Another reason I like Haidt's work is that it is built on a sophisticated social-cognitive conception of mind and morality that is very congruent with the unified theory (e.g., see this article on the Emotional dog and Its Rational Tail).

Another positive is that he really has done an excellent job documenting key value differences underlying liberal and conservative viewpoints, which is extremely useful and helpful.

Finally, I very much share and agree with his perspective that we must have the metacognitive awareness to step outside the moral matrix and observe ourselves in relationship to others. I believe the ToK System is the ultimate meta-perspective that allows us to do that. Of course, that doesn't mean we should stay "above the fray" because there are real issues that we need to be for and against, but it does mean we should be able to get perspective.


The positives acknowledged, I do have some criticisms. The first criticism, and I think a fairly significant one, is that I think Haidt misrepresents the moral values of liberals. He argues repeatedly that liberals only value "do no harm" and "fairness." However, that is not really what his data say. A closer examination reveals that it is not accurate to claim that liberals think that in-group loyalty, respect for authority, and purity have no relevance. That would be a score of a 0 on his scale. If you look at the graphs, they actually generally cluster around a score of a 2, which I believe corresponds to "somewhat relevant." So the correct interpretation is that liberals value do no harm and fairness more than the other three, but they value the others to some degree.

To me, this interpretation changes the feel of Haidt's conclusions and message. Shouldn't we vote for a government that emphasizes "do no harm" and "fairness" over "respect for authority, in-group loyalty, and purity?" Haidt wants the liberals in the audience to become reflective and wonder about the extent to which they are blinded by the Moral Matrix. He uses the interesting argument that conservatives have five moral value systems operating, whereas liberals only have two. This catches the attention of the liberals, which I think is a good thing because it suggests that liberals may actually be less complex in their thinking, which would come as a shock to their system).

However, what if we simply asked the question: What values should our government be operating from? And answer that by saying that in-group loyalty, purity, and respect for authority should equal fairness and do no harm, my guess is many in the audience would say that is a seriously flawed value system. In-group loyalty, purity, and respect for authority are somewhat relevant, but not nearly as relevant as the other two. From this framing, the question shifts from why do liberals value only two of the five to why do conservatives value each equally? And that changes much of the implication of what Haidt is arguing for.

(As an aside, there is some very interesting work done on the construct of Ego Development, and from that literature, there is good reason for believing that a liberal view represents a higher stage in ego development than a traditional social conservative view. This analysis offers another answer to why so many intellectuals are liberal relative to conservative, one that is quite different than Haidt's message!)

The second criticism I have is that many of the variables are confounded. First, he talks about traits, which he does so to set the stage that the mind has a foundational architecture. But he does not really connect the dots between openness to experience and the five moral systems (although I have not read all of his stuff, I have not seen this in his writing either). I do not believe a factor analysis of openness to experience would yield Haidt's moral systems. So the connection between trait openness and the five moral systems remains nebulous to me.

A related criticism is that by connecting the moral values to liberal and conservative viewpoints, he is confounding morality with beliefs about the proper role and function of government. He also walks a very fine line between morality as a scientific construct (which means it is a descriptive construct...we operationalize it and examine moral beliefs and how this impacts actions and policies) and morality as a prescriptive construct (we ought to be functioning via valuing the five moral systems equally).

This last point gives rise to a final criticism, which is that the list of moral systems or concerns does not seem complete. First, while these may be five dimensions that liberals and conservatives consider when voting for a government, it seems to me there are other dimensions. Think, for example, of libertarians. Haidt mentions briefly in the talk that he has some ideas about libertarians, but he does not come back to them. He is working on this, and I recommend interested readers check out a blog that deals with moral and political issues and specifically has a paper dealing with libertarians.

But my point would be that libertarians hold extremely dear to the value of human freedom. (Come to think of it, so do I!) But where does freedom fit into the scheme? Surely, freedom is a moral-political-governmental value? The paper acknowledges the value libertarians put on freedom, but after reading through it I remained unclear exactly how freedom fits into the five systems. A related issue is that Haidt was talking about social issues, but as the libertarian perspective suggests, social and economic views are somewhat intertwined. This leads to the question: Are Haidt's five systems moral systems or are they beliefs about the things governments should concern themselves with?

In sum, Haidt is blazing some very important trails. The five value systems approach is intriguing, his data sets are impressive, and clearly, he has added much to our understanding of the differences between liberal and conservative values. He is also doing a great service to psychology by raising the issue of politics and the relationship between the discipline and political viewpoints.

Like Haidt, I do believe there are very important questions to be asked regarding the relationship between politics and science in general and psychology in particular. Areas for future development include clarity about the nature of the "moral systems" relative to other constructs like traits and confounding issues like governmental philosophy, clarity about descriptive relative to prescriptive aspects of morality and related questions about how he is framing his results, and questions about the completeness of the five systems, at the very least addressing the question of individual freedoms.