Ideas, Policy, Truth, and Rationalization

From Russ Roberts.

Keynes saw economics ideas influencing policy. But maybe it is policy that influences economics. So as the world becomes more interventionist, the economists respond by finding arguments that rationalize that policy. (I am sure I’m not the first person to suggest this. Feel free to share references in the comments.)

…Obviously, this is not the whole story. Keynes was right–good ideas are powerful. Economists aren’t just affected by public opinion, they affect it in turn. But I do think our profession (like journalists) have a view of ourselves that is quite romantic–we see ourselves as truth-seekers. Well, yes, there is an element of truth-seeking in what we do. But it’s not the only factor.

I believe it helps to think in terms of two uses for reasoning. There is motivated reasoning, which is aimed at rationalizing one’s own actions and those of one’s favored group. And there is constructive reasoning, which is aimed at seeking the truth. The existence of motivated reasoning is well established in the literature on psychology and political beliefs. The existence of constructive reasoning is something that I take on faith.

It is tempting to say that I engage solely in constructive reasoning, while other people engage in motivated reasoning. Of course, the odds that this is the case are not very high.

In fact, I think that in contemporary America we are highly tribal in our political beliefs. Think of my three-axis model. If you make an argument that rationalizes the views of those who share your ideology and puts down the views of those with a different ideology, you raise your status within your tribe. If you do the opposite, you lower your status within your tribe. So once you become embedded in a tribe, your reasoning tends more and more toward motivated reasoning.

6 thoughts on “Ideas, Policy, Truth, and Rationalization

  1. One should add that there is nothing inherently disreputable about motivated reasoning–rhetoric has always and will always be the lifeblood of politics, salesmanship, and advocacy. These can all be honorable activities, and the thinking that they contain can often be innovative and sometimes sound. But Arnold is right that there is a big benefit in thinking of the “pursuit of truth” as a somewhat disconnected enterprise, and in trying to keep in mind the motivations of motivated reasoners.

  2. Motivated reasoning is always a danger as it can inure us to reality. If your goal is more than power but results, you cannot help but be motivated towards constructive reasoning, and in the end, results are the greatest power, though not always the most reassuring or satisfying. It is this discomfort and the belief, belief can be self fulfilling and we can make the world we want, that leads people away from it.

  3. I don’t agree that motivated reasoning, in the sense of “rationalization,” is well established in the literature.

    The first wave of the literature, based on Festinger’s notion of cognitive dissonance, was overturned by the cognitivist revolution in psychology, which showed that perfectly rational (not rationalizing) explanations of cognitive-dissonance results were available–but they depended on interpretive charity toward the subject who had previously been interpreted as rationalizing away “inconvenient truths.”

    The second wave of results, which has gone under the “motivated reasoning” rubric, is similarly subject to the interpretation that people direct more critical scrutiny at evidence that they have prior reason to think is false, or independent reason to hope is false (e.g., being diagnosed with a disease). Such behavior does not count as “rationalization” (in my view) unless one is *deliberately* overlooking uncongenial evidence.

    But the motivated-reasoning literature doesn’t show that. So-called motivated reasoning occurs only when the evidence people are interpreting is ambiguous enough to sustain a *logical* interpretation that is consistent with their hopes or their priors.

    Of course this poses a real problem in policy debates, where the evidence is always ambiguous. But still it seems to me that to be consistent with the charity-toward-all approach of this blog, we shouldn’t accuse people of being irrational when all they are doing is seeing what their prior experience and prior exposure to evidence has led them to expect is likely to be the case.

    Then the problem becomes not an attitude that, if we steel ourselves to be “really rational,” we can avoid–but a matter of the path-dependent information streams that lead each of us to see one or another interpretation of ambiguous evidence as more reasonable than the alternative interpretations.

    On motivated reasoning, see my intro and the Lee Ross article at http://www.tandfonline.com/toc/rcri20/current

  4. You wrote “The existence of motivated reasoning is well established in the literature on psychology and political beliefs. The existence of constructive reasoning is something that I take on faith.”

    I’m not sure how serious you are about the second sentence, but I’ve just seen someone reacting your post as though the claim is intended literally — and is literally correct. That seems mistaken to me: I don’t think the existence of reasoning in pursuit of truth needs to be taken on faith. To think it does need to be taken on faith seems similar to the fallacy Dawkins famously criticized with “show me a cultural relativist at thirty thousand feet and I’ll show you a hypocrite.”

    When people manage to solve technically demanding problems, it is strong evidence that reasoning and problem solving other than your motivated reasoning “aimed at rationalizing one’s own actions and those of one’s favored group” is taking place. Do you really think comm satellites are successfully developed, launched, and operated without a sizable amount of reasoning in pursuit of truth along the way? Is there really no difference in the amount of pursuit of truth (because it’s strictly zero in both cases!) between the most fanatically politicized organizations and the most successfully pragmatic ones?

    Now, I am cynical enough to grant that even in the most successfully pragmatic organizations there may be rather less reasoning in pursuit of truth than reasoning in pursuit of political rationalization. However, there is an important qualitative difference between “far less reasoning in pursuit of truth than we’d prefer,” which may indeed be what we observe, and “zero reasoning in pursuit of truth,” which doesn’t seem consistent with the facts.

  5. I’m always late with my comments; so this one might get missed again, but I want to repeat a complaint I have with the otherwise brilliant three-axis model.

    You say: “Think of my three-axis model. If you make an argument that rationalizes the views of those who share your ideology and puts down the views of those with a different ideology, you raise your status within your tribe. If you do the opposite, you lower your status within your tribe.”

    To square with this statement, shouldn’t the Progressive axis be positive-negative like the other two? I like Equality-Oppression, but there may be better constructions that fit my desire.