I posted a note on the other blog about the difference between internal and external coherence of political ideology. The basic idea is that, a particular person or small group can have an ideology (supporting positions A, B, C, and D, for example) that is perfectly internally coherent–that is, all these positions make sense given the underlying ideology–while being incoherent with other ideologies (for example, those people who support positions A, B, not-C, and not-D). What’s striking to me is how strongly people can feel that their beliefs on a particular issue flow from their being a liberal, or a conservative, or whatever, even though others with similar opinions will completely disagree with them on that issue.
I gave two examples: (1) positions on legalized gambling (where I’ve seen impassioned expressions pro and con by liberals and conservatives in both cases), and (2) the reaction of self-described “reactionary” blogger Dennis Mangan to a Swiss businessman who defied a government ban by erecting a makeshift minaret next to his shoe warehouse. I reacted that I was expecting he’d take the free-enterprise side and support the minaret, but he strongly supported the government ban in this case.
At his blog, Mangan replied that, as a conservative, he didn’t want to be coherent, that conservatism is not a particular ideology but rather its absence. That’s an interesting thought, although I’d think that the absence of ideology may be necessary for conservatism but can’t be sufficient. After all, you could take any bag of attitudes at random, with no ideology at all, but said basket of views would probably not happen to be particularly conservative. I do think Mangan has a good point, though, that ideological coherence is not necessarily a virtue. I wasn’t taking coherence as a virtue in my quoted blog entry either–I was being descriptive, not normative–but I could see how people could miss the point there.
I also took a look at Mangan’s commenters who, unsurprisingly, supported him against what they saw as a criticism from me. I just wanted clarify one point. In the comments, Mangan wrote that I “had assumed that because one holds certain political views, one must be a champion of free markets (to the exclusion of other values)” and his commenters wrote similar things, for example, that I “equate conservatism with a blend of neo-liberalism(cultural Marxism) and libertarianism.”
This isn’t quite right. I was indeed surprised by Mangan’s strong stance, but, no, I wasn’t saying that he was supposed to have a different position or that I “equate” conservatism with any particular configuration of beliefs. Quite the opposite. What I was saying was that I think Mangan’s other beliefs are consistent with different views on the minaret story (depending on various things such as his views about Switzerland, for example). On this issue which, I still think, a conservative (or, for that matter, a liberal) could go either way on, I was suprised that Mangan felt so strongly that he was able to simultaneously attribute “no loyalties beyond profits” and “liberal ideology” to the shoe store owner. These are two concepts that in another context he might very well place in opposition. Attitudes on issues depend on context, and, as Mangan wrote, “In another time, when we faced or will face different challenges, I might not have [been so critical of the Swiss businessman].”
So, to say it again: I am not making any statements about how conservatives (or liberals) are supposed to react to shoe-warehouse minarets, and I was not criticizing Mangan for lack of consistency. Rather, I was using Mangan to illustrate how a set of positions that can seem perfectly consistent to one group (Mangan and his blog readers), while a different set of views– agreeing with him on most issues but not on the minaret situation–can seem perfectly consistent from another perspective.
Stated that way, my position might appear true but obvious. The reason why it’s relevant, in my opinion, is that I think that politically-active people often feel their clusters of positions to be coherent, and sometimes it’s hard for a supporter of A and B and C and D to understand the logic of someone else who supports A and B and C and not-D.
P.S. Some other commenters point out that my remarks seem hardly enough to justify my appointment as a professor of political science. I’d agree with that one. My academic position comes from my academic work; this blog is something different (usually), it’s a place where, among other things, I can work out ideas and get comments on fragmentary thoughts. I often learn important things from blog commenters here and elsewhere.
About eighteen months ago, I heard someone on NPR — I would love to give credit, but can't recall who it was — ruminating about his own political beliefs. Paraphrasing, he said he felt two things very strongly: (1) he felt the Republicans had exactly the wrong stance on absolutely everything, and (2) he was upset at them for their lack of internal consistency:if they really believe X, then they have to agree with Y, but instead they argue for not-Y. He said it was somewhat disturbing when he realized that the combination of those two perceptions must mean that his own position is inconsistent.
I think it's obvious that, just as people are not rational economic actors (in the sense that there is no utility function that predicts their actions), they aren't rational political actors either. Unlike you, Andrew, I will make a normative statement and say that people "shouldn't" behave inconsistently — I think the world would be better if people thought things through, had an at least approximately coherent philosophy, and acted accordingly.
The problem is that because of US new speech the terms liberal and conservative have lost their meaning. It's not really possible to have a good conversation about political ideas in US English.
Most US citizens don't think that they are political ideologies that are about ideas but that they mean roughly the same thing as Republican and Democrat which are terms that have to do with political parties that consist of people.
It is not possible to be free of ideology; it is our ideology that determines our concept of the "good". The claim that "society is good if it enables the people in it to lead good lives," is just begging the question.
"and sometimes it's hard for a supporter of A and B and C and D to understand the logic of someone else who supports A and B and C and not-D."
Well doesn't this all depend on observed conditional probablilities? If P(D|A,B,C) != P(D), e.g. if P(D|A,B,C) is very close to one then I don't think it's surprising. As George Lakoff observed in "Moral Politics", in the current political discourse belief in seemingly independent issues is anything but, no?
Robert: These probabilities are not far from .5, as discussed in my article with Baldassarri, cited in the linked post.
Thanks much. I had only clicked through one level!
Ideologies are attempts to come up with ex-post and seemingly internally consistent rationalizations of preexisting preferences. I guess that human animals have some need to not hold wildly inconsistent beliefs. (Isn’t this a classic result of experimental psychology? Festinger, etc.) The preferences on specific policies and decisions are determined first and foremost by one's objective position is social stratification, and then by one's group affiliations (e.g., ethnic/religious/geographic). Ideologies are used to defend such preferences and make them seem as if they were derived from some more abstract set of rules. Think about these abstract ideas as axioms. Within a given ideological camp, consistency among policy preferences on specific real world issues is not required. What is needed, it seems, is the existence of at least one logical path from the axioms to the preference. I’d conjecture that this is enough to make people forget about the fact that their beliefs regarding different issues are logically inconsistent with each other.
For instance, in my country of origin, owners of small and medium-sized retail businesses are very happy to endorse free market ideologies when they want to justify why they want to be able to fire their employees at will. But they suddenly become regulation-happy when they want to prevent immigrant-owned businesses (e.g., Chinese-owned retail, North-African fast food joints) to compete with them on timetables (e.g., opening stores late at night or on Sundays). The two inconsistent preferences are held together, so to speak, by the fact that they can be derived from the axioms of the specific brand of conservatism popular in the country, that marries some fetishism for an “orderly society” (that is, one in which people go to sleep early and don’t shop on Sundays) with a respect for “free enterprise”. Similar inconsistencies in preferences can be detected among other social groups too, obviously. For instance, most of those on the left are happy to allow the government to tightly regulate labor markets, but they are opposed to any regulation of internet content. These again can be justified in a framework that values “workers rights” as well as “free speech”.
As I see it, most of what goes under the name of "conservatism" in contemporary U.S. politics is used to rationalize the support for policies that benefit white, Christian, "respectable" people. Conservatism itself revolves around axioms that claim that “markets are good”, “religion is good”, and “tradition has value in itself”. These are not necessarily consistent with each other: for instance, markets lead to “creative destruction”, which is not a tradition-preserving phenomenon. But by presenting them as a single “conservative ideology”, the inconsistencies among the axioms, as well as the inconsistencies among the real-world policy logical implications, can be ignored: inconsistency among axioms is in some way assumed away. Inconsistency between policy preferences can similarly be assumed away by invoking consistency between a specific policy preference and one of the axioms.
So one can be conservative and expound free market theories if this benefits members of a given social group. But free markets are quickly forgotten if the rights and well-being of non-white, non-Christian people are at stake. Religious freedom is similarly something fundamentalist Christians are happy to endorse, if it justifies, say, home schooling or the right to believe that the world is 5000 years old. If, on the other hand, someone wants to build a minaret, free markets and religious freedom are quickly forgotten, and the principle/axiom used to justify which is the “conservative” side becomes something like “preservation of traditional values”.
Ideologies that allow more leeway in terms of *internal* inconsistencies (in the sense that they can make logically inconsistent policy preferences appear consistent in the light of the ideological framework) can be used to rationalize more preferences than ideologies that are based on logically consistent axioms. Ideologies based on few (or a single) axioms at times are unable to justify one’s objective interest. The size of the set of real-world preferences that can be rationalized in the frame of a given ideology probably affects the success (in terms of popularity) of a given ideology. Hence, for instance, conservatism (vaguely defined, and derived from a wider set of primitives) is way more popular than libertarianism or Marxism, bundles of smaller sets of axioms. Libertarianism yields, for every binary policy decision, a clear prediction regarding which is the libertarian side and which is the anti-libertarian side. The reason why this happens is that libertarianism is based on one or very few simple axioms. If you think that government limitations of individual autonomy are bad, then you must be anti-taxes, be ok with a free market for marijuana, and oppose military interventions. On the other hand, some brands of contemporary conservatism allow one to rationalize: being spending-happy if the money is spent to kill people in some remote country (or those on death row); being a fiscal miser when it comes to health insurance for low-income kids; support restrictions on the use of medical technology to terminate a fetus; support the use of medical technology to terminate a convicted adult; support of free markets when Walmart is involved; opposition to a free market for recreational drugs; etc.
It would be interesting to think about how to test a proposition like “popularity of an ideology is an increasing function of the number of primitive axioms on which it is based.”
If I understand the paper correctly, the clustering among people (i.e. conditional probability not close to 1/2) is only true among party elites and strongly party aligned voters.
So given that party elites are more likely to be explicitly expressing their preferences (that's what politicians do right – talk about their beliefs?), perhaps George Lakoff and I have been hoodwinked by selection bias?