Jonathan Haidt on Psychology and Politics

A little while back I read Jonathan Haidt’s book The Righteous Mind: Why Good People are Divided by Politics and Religion and had the opportunity to meet him and hear him speak at a book party in NYC. I’ve been meaning to say a few things about it and the recent appearance of a piece by Haidt in Time magazine prompts it, an article and a quiz that illustrates his points. If you are not familiar with Haidt or his methodology, I encourage you to click over and do the quick 12 question quiz now before reading further; even if you are, you might go ahead and do it anyway because it is fun and will refresh your memory. Go ahead, I’ll wait.

So what’s Haidt’s argument? His basic idea is twofold. First, that people do not rationally choose their ideologies. You do not come into the political arena as a blank slate and then just examine all the moral and consequential arguments for different policies and pick the one that is most “correct.” Instead, you come into the political arena with subconscious, largely unexamined psychological beliefs. Initially for Haidt what he focused on was ideas of “disgust.” Over time that has broadened and he describes five key vectors or values of psychological morality: (1) care/harm, (2) fairness, (3) loyalty, (4) authority, and (5) sanctity. Haidt finds in his research that self-described “conservatives” tend to value all five vectors of morality (as he defines them). Liberals, by contrast, place a high value on “care” and “fairness” and a lower value on loyalty, authority, and sanctity. On the two values that conservatives and liberals both value (care and fairness) they do not define those terms the same way, although they both value them according to their different definitions.

The second part of Haidt’s argument is that once you have subconsciously chosen your ideology (you don’t rationally choose what the important factors are) you also do not rationally and objectively weigh the evidence as to whether your ideological views are “correct.” Instead, people tend to subconsciously sift the information that they take in: you tend to overvalue evidence that supports your predispositions and dismiss evidence that is inconsistent with it. As a result, “evidence” becomes self-justifying.

In the end, this all becomes very bad for democracy. If people are choosing their ideologies based on their subjective psychological views and if people are subconsciously overweighting confirming evidence and dismissing contrary evidence, then it is hard to see how compromise and persuasion can occur.

Haidt also concludes that his model implies that there is a “conservative advantage” in politics: because conservatives value all five moral vectors and liberals value only two, this means that the conservative worldview will naturally tend to be able to pull in more people than liberalism.

So what to make of all this?

First, on Haidt’s central propositions, my sense is that he is correct. That ideologies are driven primarily by some sort of unexamined subconscious underlying psychological predispositions rather than rational argumentation seems to me to be correct. (Many readers will recognize a strong similarity here to Thomas Sowell’s fabulous book, A Conflict of Visions–a similarity that Haidt himself has noted).

On the other hand, I am not fully persuaded that the precise list of five values that Haidt identifies are necessarily the right list. I’m also not persuaded that they aren’t, but they don’t all seem uniformly persuasive to me. And indeed, in the book Haidt identifies a sixth value: the Liberty/Oppression value. This is not part of the original six, but Haidt says that it provides an explanation for why libertarians and conservatives tend to affiliate with each other. The question I had after reading that, however, is whether that sixth value actually swallows the other five, and so there really is only one key vector. To Haidt, I would add an additional hypothesis, which is that while the Liberty/Oppression axis appears probative for the libertarian-conservative overlap, it plays out differently for libertarians. My sense is that while libertarians root liberty in the individual, conservatives implicitly see the family as the fundamental moral/analytical unit and so “liberty” essentially means more about family autonomy than individual autonomy.  The larger point, however, is that I think that the central thrust of the research program–that there are some sorts subconscious psychological assumptions underlying all this, seems right to me.

It also seems to me that his second proposition is correct as well: that given this, in politics people do tend to subconsciously screen their evidence in a manner that tends to confirm their preexisting views and dismissing contrary evidence. This is in a large sense consistent with Bryan Caplan’s work in rational irrationality, which is that in politics (unlike markets) there is essentially no cost to being wrong or holding “incorrect” views, so if you gain any utility from doing so then you will persist in holding those views.

One other point that I find really interesting and important about Haidt’s work is his findings on the ability of different groups to empathize across these ideological divides. So in his book (p. 287) Haidt reports on the following experiment: after determining whether someone is liberal or conservative, he then has each person answer the standard battery of questions as if he were the opposite ideology. So, he would ask a liberal to answer the questions as if he were a “typical conservative” and vice-versa. What he finds is quite striking: “The results were clear and consistent. Moderates and conservatives were most accurate in their predictions, whether they were pretending to be liberals or conservatives. Liberals were the least accurate, especially those who describe themselves as ‘very liberal.’ The biggest errors in the whole study came when liberals answered the Care and Fairness questions while pretending to be conservatives.” In other words, moderates and conservatives can understand the liberal worldview and liberals are unable to relate to the conservative worldview, especially when it comes to questions of care and fairness.

In short, Haidt’s research suggests that many liberals really do believe that conservatives are heartless bastards–or as a friend of mine once remarked, “Conservatives think that liberals are good people with bad ideas, whereas liberals think conservatives are bad people”–and very liberal people think that especially strongly. Haidt suggests that there is some truth to this.

If it is the case that conservatives understand liberals better than liberals understand conservatives, why is that? Haidt’s hypothesis is that it is because conservative values are more overlapping than liberals–conservatives have a “thicker” moral worldview that includes all five values, whereas liberals have a “thinner” view that rests on only two variables. Thus, the liberal moral values are constituent part of the liberal views, but not vice-versa. So conservatives can process and affirm liberal moral views and liberals literally cannot understand how someone could be both moral and conservative–the moral values that might be animating a conservative (say, tradition or loyalty) are essentially seen by liberals as not being worth of moral weight. So conservatives who place weight on those values are literally seen as “immoral.”

As an aside, I think the “thinness” of the liberal moral worldview may explain a phenomenon that has puzzled me, which is the speed at which liberal views harden into orthodoxy and the willingness of liberals to use various forms of compulsion to enforce that orthodoxy. Consider same-sex marriage. For conservatives, this is actually quite a difficult topic and one sees a wide variety of opinion and discussion on the “conservative” side of the fence. “Conservative” opinion is not uniformly opposed to same-sex marriage and conservatives who support same-sex marriage are not ostracized or silenced for doing so. I think Haidt gives a sense why: same-sex marriage cuts across a lot of these moral dimensions in different ways–it simultaneously triggers sanctity (for religious conservatives) and authority (tradition), but it also triggers equality/fairness impulses and care/harm impulses for the individuals affected by it. So conservatives, I think, tend to see it as an issue on which reasonable minds can disagree and that those who hold contrasting views are not generally thought to be immoral or evil. I think this sense that there is room for legitimate disagreement is also consistent with the one near-consensus view of conservatives, which is that regardless of one’s position on the issue there is no constitutional right to same-sex marriage, as opposed to allowing the issue to evolve through democratic processes that permit disparate moral and other views to be heard and compromised.

Liberals, by contrast, appear to broach little disagreement from the orthodoxy on this issue (and others for that matter), and I think Haidt gives us a sense why. If they are processing this only through the care and fairness moral value frameworks, then that implies that only immoral people could be opposed to same-sex marriage. And if these people are immoral, then their opposition is hateful and unjustified. So a notion quickly hardens into an orthodoxy–no moral person could oppose same sex marriage. It is then a logical step to a willingness to demonize and try to silence opponents of same-sex marriage as holding not just wrong-headed but illegitimate views, much like the Inquisition, which was premised on the idea that there is potential harm and no value in tolerating “error.” (‘That’s an oversimplification of the Inquisition, of course.) Ditto for more petty forms of censorship and suppression of speech, such as university speech codes.

One thing about Haidt is that my general impression is that by and large libertarians are often particularly skeptical of Haidt’s methodology and conclusions. To some extent that skepticism is warranted: Haidt (by disposition a liberal) conflates an analytical distinction that is crucial for libertarians: the distinction between an act being immoral and illegal. Haidt implicitly assumes that because an act is thought of as moral (i.e., the desire for everyone to have access to health insurance or recycling) it should be mandated and if something is immoral (i.e., drugs or pornography) it should be prohibited. That obviously conflates two different categories. On the other hand, to some extent that skepticism is unwarranted: libertarians insist on that logical distinction, but the reality is that libertarians are a very small percentage of the population. So, in fact, even if Haidt’s conflation of the two categories is not logically defensible, it seems to me that within the scope of his project it is quite defensible–if he is seeking to simply describe and predict how people think, he seems on solid ground in working from the assumption that most people–liberals and conservatives alike–do not draw that distinction and implicitly do conflate morality and legality.

One final word on libertarians: Haidt has written a completely separate scholarly article analyzing the “Psychological Dispositions of Self-Described Libertarians.” While one can quibble with such things, his findings seem largely persuasive to me. In that article, Haidt applies the same tools to self-described libertarians and concludes that there are distinct psychological correlates to to libertarian morality that distinguished libertarians from both liberals and conservatives. Perhaps most striking is the libertarians emphasis on systematization. Now this, I think, is an important insight. For it explains a point that seems to be highly distinctive to libertarians: the recognition by libertarians, often with a high degree of pride, that libertarianism offers the only “consistent” ideology and that is one of the most compelling aspects of it. Well here’s Haidt’s point: Most people simply do not care whether their ideological views are consistent. For most people (liberals and conservatives), consistency is simply not a relevant variable or axis for determining what you believe or your ideological worldview. This explains, I think, the frequent bewilderment that libertarians face when they try to persuade someone to change their mind about, say, a social policy because it is “inconsistent” with their economic policy beliefs. It simply is not a relevant argument to them. This has obvious implications for communicating libertarian ideas to non-libertarians (i.e., the overwhelming number of people in America!).

Which raises a related point: Haidt finds that libertarians place a much higher emphasis on rationality and logical reasoning than do other ideologies. But that doesn’t mean that libertarian beliefs are less-motivated by unexamined psychological predispositions than other ideologies. Again, take the idea that libertarians believe that “consistency” is a relevant variable for measuring the moral worth or persuasiveness of an ideology. But that is not a self-justifying claim: one still must ask why “consistency” maters or should matter. So while libertarians may place a higher stated value on rational argumentation, that does not mean that libertarian premises are any less built upon subjective psychological foundations.

Also, I think it is possible that there are other measures of psychological attributes that might also be correlated with political ideologies and the like, so I don’t think that Haidt’s model is the only one. I think that more research along those lines would be useful.

Anyway, I’ve hardly done justice to Haidt’s argument. And while I’m not certain I agree with all of his conclusions, I think that his basic insights are more right than not and well worth thinking about.