Tuesday, March 29, 2016

Trump and Authoritarianism: An Attempt at a Rigorous Definition

There has been a lot of talk about Donald Trump's followers being "authoritarians."  Well before you start throwing words around too loosely, you had better start defining what authoritarianism means. And a lot of these definitions are not very satisfactory.  This one, for instance, defines authoritarians as people who "value conformity and order, protect social norms, and are wary of outsiders."  But that simply means people who value Jonathan Haidt's moral foundations of group loyalty, authority, and purity/sanctity.  In other words, conservatives.  And I am one who believes that it shows a serious tunnel vision to pathologize conservatism.

Yet Haidt himself is also calling Trump followers authoritarians.  And if Haidt, who has led the charge against equating conservatism with authoritarianism, is calling Trump's followers authoritarian maybe we should be listening.  And maybe he can offer some useful insight into a better working definition of authoritarian than simply a synonym for conservative.

Haidt wrote an interesting column on the primaries on February 5 -- an eternity ago the world of primary politics.  Jeb Bush, Ben Carson, Mike Huckabee, Rand Paul, and Marco Rubio were still running, though not doing very well.  He applies the moral foundations test (naturally) to the followers of the candidates, rather than to the candidates themselves.  Supporters of each candidate were ranked by how their emphasis on each value compared with the views of the general American public.  A positive score meant that followers of that candidate valued that particular foundation more than the general public; a negative score did not mean that followers of a candidate opposed that foundation, but simply that they valued it less than did the general public.

For the sake of the article, he lumps the conservative values, or as he also calls them, the binding values, of group loyalty, authority, and purity/sanctity into one cluster -- call it social cohesion.  That left the values of harm/care, justice, and liberty.  Furthermore, it is pointless asking people how much they value justice.  Everyone sets a high value on justice.  The difference is not that some people value it more and some value it less, but that people with different political values define it differently.  Liberals define justice in egalitarian or at least utilitarian terms -- harm avoidance and the greatest good for the greatest number.  Conservatives are more likely to define it a proportionality, or just deserts -- letting the undeserving fail.  In simple English, conservatives are more likely than liberals to define justice in punitive terms, although Haidt does not refer to it that way.  His results were interesting and sometimes surprising.

Unsurprisingly, there was a marked negative correlation between care/empathy and the tendency to view justice as proportionality, or in punitive terms.  Wanting people to suffer the consequences of their bad decisions and lack of empathy are traits one would expect to go together, and they do.  One might call the combination the very definition of economic conservatism.  It should also not be surprising that there was a negative correlation between the value place on individual liberty versus social cohesion, since there is, indeed, a tradeoff between these things.  Preferring social cohesion to individual liberty could be called the very definition of social conservatism. If anything is surprising, it is that the negative correlation was not stronger.

But while one would expect a tradeoff between empathy and proportionality (economic liberalism/ conservatism) and between individual liberty and social cohesion (social conservatism/liberalism) there is no reason whatever for there to be any correlation between where people stand on the economic spectrum and where they stand on the social spectrum.  And, indeed, there was not.  While supporters of both Democratic candidates scored high on empathy and low on proportionality, only Sanders supporters scored high on individual liberty and (very) low on social cohesion.  Hilary Clinton supporters scored a bit low on both.  In other words, both Democrats draw economic liberals, but Sanders supporters are the more socially liberal.  On the Republican side, followers of Christian candidates Mike Huckabee and Ben Carson scored average to above average on empathy and fairly close to the mean on proportionality.  In other words, Carson followers were only slightly economically conservative and Huckabee followers were slightly economically liberal.  But both placed a markedly low value on individual liberty and a high value on social cohesion.  In other words, they were socially conservative.  Rand Paul followers, by contrast, had the exact opposite profile.  They scored lower than any others on empathy and high on proportionality.  By contrast, they scored higher than anyone else on individual liberty and lower than anyone but Sanders supporters on social cohesion.  Rubio and Cruz supporters stand out as low on empathy (Cruz supporters were second lowest after Rand, and Rubio's were third), and much, much higher on proportionality, i.e., punitive justice than anyone else, with Cruz supporters especially high.  Haidt adds that this marks them as the Tea Party candidates -- proportional/punitive justice is the Tea Party's leading issue. Followers of both men rate somewhat above average on social cohesion, with Cruz followers more so, while Cruz followers are slightly above average on individual freedom and Rubio followers neutral.

And now for the most surprising findings.  The candidates whose followers hewed closest to the mean were Jeb Bush -- and Donald Trump.  Bush followers were either at the mean or minimally above or below it on all categories.  So why couldn't the candidate who attracts supporters closest to ordinary American values gain more traction?  Haidt suggests that he is just too vanilla bland to generate much passion.  As for Trump supporters -- well, they are somewhat below average on empathy, but not spectacularly so.  They are also somewhat above average on proportionality and social cohesion, but nothing impressive.  And they are very minimally above average on individual freedom.  In short, they seem pretty vanilla bland, too.  On a left-to-right spectrum, Haidt places them smack-dab in the middle, as number five out of nine candidates.

Haidt then goes on to compare the candidates in other ways.  For instance, he compares each primary candidate with only members of their own party.  Here are the results for comparing Clinton and Sanders to your average Democrat:

They are both at the Democratic average on care/ empathy.  From a Democratic perspective, Clinton looks pretty bland, valuing liberty a little below average and authority a little above, but nothing remarkable.  Sanders, by contrast, ranks below average on proportionality and all measures of social cohesion, and above average on individual liberty.  Which is to say, he is a social liberal.  But you knew that.

Among Republicans, what stand out most is (1) Carson followers are high on care and low on proportionality (i.e., they are not economically conservative) and low on liberty.  The really outstanding supporters are Rubio and Cruz on proportionality.  They are exceptionally punitive, even among Republicans.  (Rubio supporters also seem to disvalue loyalty and sanctity, while Cruz and Carson supporters value it).  And here Trump supporters seem the most vanilla bland of all the Republican candidates -- their values differ only minimally from the average Republican.  Comments Haidt:
One of the biggest surprises in our dataset was that Donald Trump’s supporters did not appear particularly unusual. Part of this is due to the fact that his higher vote share means his supporters are closer to average.
But Haidt has one last trick up his sleeve.  He compares each candidate's supporters to other members of the same demographic.   In other words, they took the "age, sex, education level, income, race, and self-reported ideology" makeup of supporters of each candidate and compared their moral foundations to general members of the same "age, sex, education level, income, race, and self-reported ideology."  For instance, if Bernie Sanders followers were younger, whiter and more liberal than Clinton followers, he took a cross-section of the population of that was equally young, white and liberal and compared their moral foundations to specific foundations of Sanders supporters.*  His results were as follows:

  Thus Sanders followers did not score any higher on individual liberty than others of their demographic, but did score lower on proportionality and social cohesion.  Rand Paul followers, by contrast, scored overwhelmingly high on liberty (and low on social cohesion).  Cruz and Rubio followers scored high on proportionality, but while Cruz followers scored highest in absolute terms, Rubio followers scored even higher relative to their demographic.  Bush followers, though very close to the national average on proportionality, were much below it compared to their demographic.

And what of Trump followers?  The predictive effect is weaker than any of the other candidates, but compared to others of their demographic, Trump supporters tend to score higher on social cohesion and lower on empathy. Says Haidt, "These are the true authoritarians — they value obedience while scoring low on compassion."

It is significant what Haidt does not define as authoritarianism.  He does not define placing a high value on authority as authoritarianism.  He does not define it as value placing a low value on personal liberty.  Nor does he define it as the combination of putting a high value on authority (and social cohesion in general) and a low value on personal liberty.  These things do, indeed, tend to correlate and might seem an intuitively obvious definition of authoritarianism.  Haidt disagrees.  He sees these things together as simply a definition of social conservatism and equating the combination with authoritarianism is simply an attempt to pathologize social conservatism.

Similarly, Haidt does not define low compassion as authoritarianism.  He does not define a punitive (proportional) definition of justice as authoritarianism.**  Nor does he define the combination of low compassion, high punitive justice as authoritarianism.  These things, too, tend to correlated, but Haidt would presumably say that they are just economic conservatism and equating them with authoritarianism is simply an attempt to pathologize economic conservatism.

Instead, Haidt equates authoritarianism with two traits that do not normally correlate -- a high value on social cohesion and a low value on compassion.  Haidt has elsewhere said that he has nothing against libertarians who place a low value on compassion but a high value on personal liberty.  Here he indicates that he has nothing against social conservatives who place a high value on social cohesion but also on care.  Indeed, it is probably true that many social conservatives like Mike Huckabee value social cohesion in part as promoting care.  It is the high cohesion/low care outliers who Haidt sees as dangerous.

I intend to play with the subject (not necessarily in the context of Donald Trump) in my next few posts.

*I am rather skeptical of this approach myself.  It seems to me that if a certain demographic gravitates toward a certain ideology and a certain candidate, that is itself more useful than trying to see how supporters of that particular candidate differ from their general demographic.  But still, to the extent that it distinguishes Trump supporters from the general population, it is interesting.
**Although he has warned against the danger of assuming that karmic justice is not just an ideal, but an actual reality.

No comments:

Post a Comment