Greg Rakozy,, CC0 Licensing
Reading Time: 6 minutes

In my previous post I suggested that it was Authority, and the exercising of Authority through demanding Obedience, primarily through punitive learning environments, that defined the difference between Libertarians, Liberals, and Conservatives, and particularly between naturally conservative individuals, and people nurtured (actually inculcated) into it. So let’s look at the Schwartz-Duval model with Haidt’s Five Flavours super-imposed over the top in the approximate location that my sample put them. Starting, necessarily, with…



Figure 1: The Schwartz-Duval Circumplex with Haidt’s Authority/Subversion super-imposed over the top of it in line with my sample’s definitions (although a little too far over Achievement, and not enough in Power, but you get the general idea).

 Authority is a value contained in the Power segment of Schwartz’s model, which follows on from Achievement (from which I contend that one’s power is derived). In my sample, 44% pointed to Power as being definitional of Authority (lending weight to Schwartz’s inclusion of Authority in the Power segment), 19% pointed to Conformity, 10% to Tradition, 9% to Self-Direction, and 8% to Security. So four of the five most definitional values for Authority (accounting for 81% of my responses) are on the right (the exception being Self-Direction – which is presumably when you are the authority in question – which may explain the appearance of Libertarians on the right). Indeed the only right-wing value that didn’t score highly was Benevolence (3%) – the value that turns out to be more important to (probably religious) liberals than conservatives.



Figure 2: The Schwartz-Duval Circumplex with Haidt’s Loyalty/Betrayal super-imposed over the top of it in line with my sample’s definitions.

 Here we find that Loyalty is seen as most strongly associated with Benevolence (26%), and then Security and Tradition (18% each), then Power (12%), and Self-Direction (10%). In total Loyalty is 33% Self-Transcendence (though barely any of the reason for this is due to Universalism), and 43% Conservation, making it a highly conservative value, and even more highly right wing in general.



Figure 3: The Schwartz-Duval Circumplex with Haidt’s Sanctity/Degradation super-imposed over the top of it in line with my sample’s definitions.

Sanctity (also called Purity on occasion) was mostly deemed by my sample to be about Tradition (29%), but then, interestingly, Universalism (18%), and then Hedonism (13%). Recall that I said previously, when discussing related results, that the Hedonism score was likely to be due to a pull on the Degradation aspect (i.e. a negative), rather than Sanctity (the positive); this is interesting in and of itself. Sanctity, then, is a Conservation value (47%), but Openness to Change and Self-Transcendence are about equal second (23% and 22%, respectively).



Figure 4: The Schwartz-Duval Circumplex with Haidt’s Harm/Care super-imposed over the top of it in line with my sample’s definitions.

Harm was considered strongly to be about Benevolence first (31%) and Universalism second (24%), but then Security third (22%). Harm, therefore, is a strongly Self-Transcendent value (56% – containing both Benevolence and Universalism, two values held strongly by liberals), followed distantly by Conservation (32%), via the value of Security.



Figure 5: The Schwartz-Duval Circumplex with Haidt’s Fairness/Cheating super-imposed over the top of it in line with my sample’s definitions.

Finally, and in some ways, most interestingly, Fairness turns out to be seen as roughly equal in all quadrants. The strongest values related to it were Universalism (20%), Achievement (14%), and Self-Direction (13%), all left wing values. That having been said, Achievement is strongly held to by Libertarians and Conservatives, slightly ahead of Liberals, as mentioned in my previous post. Self-Transcendence did have a slight advantage overall (31%), but less than 10% separated the most and least definitional quadrants (Openness to Change scored lowest at 22%).

Haidt’s Five Flavours


Figure 6: The Schwartz-Duval Circumplex with Haidt’s Five Flavours super-imposed over the top of it in line with my sample’s definitions.

As you can see, above, Stimulation did not feature in definitions of morality, as far as Haidt’s flavours go. This is odd when you consider that most people would agree that knowledge is power. Stimulation is the basis of knowledge (from experience), and no one is arguing that Power is not morally salient. Self-Direction did feature, a little, as the basis upon which one recognises whether something is degrading or not, and whether something is fair or not. Both of which seem reasonable, and both of which suggest that Haidt’s focus on binding values is incorrect.


Figure 7: In the left table we can see how descriptive of Haidt’s Flavours that each Schwartz value was seen to be, in order, from most descriptive to least descriptive.

The centre table presents the same information, but this time ordered from Stimulation through to Self-Direction, which I contend is the best way to understand the Schwartz circumplex (the top three values describing Haidt are in bold).

The right table shows the descriptiveness of each of Schwartz’s quadrants of all of Haidt’s five flavours.

Over all, the left wing values of Stimulation, Hedonism, and Self-Direction all scored very low across all five flavours (see the middle table in figure 7), as did Achievement. Achievement was the only value that all three political types seemed to agree upon in my previous post (based on data from Iyer, Haidt, and colleagues[1]).

Two of the three Schwartz values that my sample said best-defined Haidt’s flavours are highly liberal, whether religious (Benevolence, 15%), or secular (Universalism, 14%). These were separated by Power (14.5%), a right wing (but not specifically conservative) value. It should come as no surprise, given my complaints about Haidt, that his flavours were said to be more compatible with Schwartz’s Conservation values, and less compatible with what I call his Liberal values (Openness to Change). If Haidt’s flavours were indicative of the full scope of human morality these numbers would all be closer to 25%, instead they show a strong pull towards “issues related to binding groups together” (Graham, Haidt & Nosek, 2009, p. 1044), as can be seen in the right hand table of figure 7.

If we were to include Freedom, Haidt’s sixth flavour, this would improve things slightly, as it would bring up the Openness To Change quadrant, primarily at the expense of Conservation. But in doing so we would move away from Haidt’s “issues related to binding groups together” (ibid.), of course, ironically, we would also be moving closer to Schwartz’s model.


Contrary to the very basis of Haidt’s work, morality is not exclusively about binding groups together, because binding people within a group necessarily pulls them away from people in other groups, and this is the basis of almost all conflict, and conflict is seldom a context for strong moral behaviour. By contrast Independence, Empathy for Others, Tolerance, Curiosity, and Creativity are a basis for strong moral behaviour, in no small part because in combination they attempt to avoid conflict.



[1] Iyer R, Koleva S, Graham J, Ditto P, Haidt J (2012) Understanding Libertarian Morality: The Psychological Dispositions of Self-Identified Libertarians. PLoS ONE 7(8): e42366. doi:10.1371/journal.pone.0042366