Thursday, April 23, 2009

The Paternalistic Bias of Behavioral Economics

I find I’ve been waking up each morning with the outline of an essay in my head. Each time it’s a different topic, but the routine is the same: I stare groggily at the alarm clock as one part of my mind proposes objections and the other answers them by drawing out finer distinctions. This is terrible for my sleep, and it hasn’t led to anything productive because there is no way to write a new, worked-out essay each day. I’ll use this blog as a place to file these ideas away.

Today’s subject is what behavioral economists study, what they don’t study, and what this tells us about their underlying prejudices. And, as I realized at 6 am, this all goes back to the exchange between Walter Lippman and John Dewey in the 1920s, although I may not get to that here.

The domain of behavioral economics is the failure of individuals to process information according to the dictates of rational goal achievement, the idealized maximization routines one learns in microeconomics. The list is long and includes such issues as probability bias, accounting bias, availability bias, hyperbolic discounting, loss aversion (status quo bias), etc. All of these cognitive traits have been documented beyond dispute by psychologists, and there is now a vast literature demonstrating that they have nontrivial effects on economic decision-making.

But those familiar with psychological research will know that the behavioral insights gleaned by economists are not exhaustive. In particular, it is interesting that one of the biggest research fields in psychology, cognitive dissonance, has barely made a dent in the BE agenda.

The basic idea of CD is that people experience discomfort from holding two contradictory thoughts at once. The main application has to do with the reception of information that violates an individual’s self-conception: if you have chosen to do something, and this choice plays even a modest role in how you think about yourself, you will have a tendency to ignore or reject information that goes against it. The vernacular version of this is “denial”, which we see every day in real life.

I first became aware of the importance of CD in my studies of occupational safety and health. Of all the psychological impediments to rational thinking about work and health, surely denial is the most significant. There was a flurry of interest among economists in this topic following the publication of “The Economic Consequences of Cognitive Dissonance” by Akerlof and Dickens in 1982; I presented a much simpler but also sharper model in my book Markets and Mortality (1996). Researchers in health behavior took note; they had long had CD on their minds and were predisposed to see it in economic dress.

But read the current literature on BE and you will be hard-pressed to find any mention of CD at all. There is no end to studies of defective information processing, but the initial acceptance of information doesn’t make the cut. We have nuanced policy advice in books like Nudge, but the much larger questions posed by CD are off the table. Am I mistaken in thinking that denial, the refusal to acknowledge information that calls into question our economic behavior, is central to the disconnect between environmental knowledge and anemic or nonexistent policy?

So how to explain this? To put it bluntly, I sense bad faith. The topics researched by BE all have this in common: they point to traits that are likely to disappear with enough exposure to decision theory. We don’t succumb to loss aversion, probability bias or these other flaws. This means we can come up with clever schemes to fix them. But CD is another kettle of fish altogether; there is no reason to suppose that the experts (us) are any less likely to be subject to denial than the lay folk (them). In other words, I think behavioral economists find the paternalistic stance of their discipline congenial; there is less interest in pursuing questions that apply equally to their own judgment.

Ah, but what has this to do with Lippman and Dewey? That will have to wait.


Robert D Feinman said...

There is quite a bit of psychological research which shows that there are two broad types of personality (with much overlap in the middle).

Psychologist Robert Altemeyer has defined a characteristic he calls "right wing authoritarianism" and has studied people who score high and low on this measure. Those who score high tend to be conservative and show belief system defenses such as cognitive dissonance and the avoidance of discrepant information.

Those who score low are more likely to examine all sides of an issue and to incorporate such information in decision making.

To over generalize (based upon your thesis), those economists who lean to the prescriptive are also the most ideological and conservative and this leads to avoiding psychological studies as well as economic facts which don't fit their prejudices.

I'm not sure this is a left-right issue there seem to be some "liberal" economists who are also pretty set in their opinions.

Altemeyer has a free, online, book detailing his psychological research, you can find it at

One of the more useful findings is that arguing with high RWA's is a waste of time, they will never change their minds. He estimates this personality type to be about 20% of the population.

Lord said...

It is not surprising that rational argument would not be effective, but one has to wonder whether conversion of their authority figures would be, assuming they might not be of the same type though that might be uncommon.

Anonymous said...

I don't believe that I or any of my friends or peers have ever experienced (suffered from?) CD.

Peter Hollin said...

Brilliant observation. It has the sort of conceptual elegance that I believe lends itself to eventual empirical proof. I have a sneaking suspicion that selection effects are also a factor.

You should certainly continue this line of thought. I doubt you need to be told that, but I feel compelled to offer encouragement.

"The most exciting phrase to hear in science, the one that heralds new discoveries, is not Eureka! (I found it!) but rather, "hmmm... that's funny... " Issac Asimov

chrismealy said...

Since I slogged my way through "The Public and Its Problems" last year I take this as a personal challenge. But I can't work out the Dewey/Lippman angle here. Are the behavioralist economists playing the role of Lippman's expert class, mediating relations between the government and an incompetent public?

I used to think the behavioral economics approach was great because it took the neoclassicals down a notch, but lately I've been worried that all that nudging is really just trying to trick people into being more like Homo economicus. Homo economicus is a total jerk, and more of him is the last thing we need.

I could get behind a behavioral economics that led to a society where people could trust their best human values to guide them, and ignore free-market triumphalist jerks who got As in econ 101.

Sandwichman said...

I'm not sure that you have to go as far as bad faith, Peter. Wouldn't cognitive dissonance be enough to explain the behavioralists failure to focus on cognitive dissonance?

reason said...

What makes you think behavioural economists are not subject to all the other common irrationalities. After all, there is plenty of evidence that decisions are mostly made emotionally and then subsequently rationalised. Selective review of the evidence is just as important as CD I would think.

kevin quinn said...

The problem I have with BE is the inability to see that some of what they consider "biases," while they are biases relative to the conception of rationality peddled by the mainstream of the professsion, may not be biases at all relative to a richer and more accurate conception of rationality, conceptions of rationality which make it findamentally expressive, as in Elizabeth Anderson's work or narrative, as in Alasdair McIntyre's work. Framing effects often reflect treating differently actions with identical welfare consequences but different meanings. For a trivial example we distinguish between consequences which we intend and unintended consequences. In making distinctions like this we "make sense of ourselves" in certain ways and this "making sense of ourselves," in ways in which we are accountable to our fellows, is, again on alternative conceptions of rationality, at the very heart of what it means to be rational.

Barkley Rosser said...


Well, I think the more sophisticated of the behavioral economists argue that these so-called biases in fact reflect impulses that were selected for in our evolutionary pasts, some of which may be appropriate to current society, and some that are not. So, hyperbolic discounting reflects impulses coming from "non-rational" parts of the brain, but some of those sudden impulses also involve things important to survival, such as the "fight or flight" decision, or to reproduction ("Try to sleep with this person now!").

Regarding the authoritarian potential in BE, indeed some libertarians have been openly upset with people like Sunstein and Thaler who push all sorts of "default" social policies (e.g. for pension plans) on the idea of taking advantage of peoples' biases in order to bring about these supposed social goods paternalistically. I am not sure I agree with this concern, but I can see that it could get carried too far, possibly.

Shag from Brookline said...

Do behavioral economists focus upon the behavior of others but not of themselves?