Bryan Caplan  

Biased Behavioral Econ

PRINT
Barbara Anderson, RIP... Milton Friedman on Segregation...
Behavioral economists often use their findings to argue for government intervention.  Critics of behavioral economics often accuse its practitioners of a tacit double standard: Human irrationality is a poor argument for government action because officials are human.  Question: Can behavioral economists really neglect such a basic critique of their position?

Yes.  In "Time for a Behavioral Political Economy" (Review of Austrian Economics 2012), Niclas Berggren classifies over 300 papers in behavioral economics.  Does the paper offer any policy recommendation?  If so, does the paper acknowledge that policymakers are subject to the same foibles as everyone else?  Berggren:
The study has two main purposes. The first is to document the prevalence of policy recommendations of a paternalist kind in leading research in behavioral economics. To what extent do researchers draw normative conclusions from the insight that economic actors often behave irrationally? The second is to investigate to what extent those behavioral economists that do offer policy recommendations analyze policymakers in the same way as they analyze economic decision-makers. Are the former also seen as suffering from cognitive imperfections and irrationality, or is it simply assumed that they are without such problems? To the extent that researchers do not apply assumptions about cognitive limitations and biases to policymakers, or motivate why such assumptions are superfluous, it could be argued that policy recommendations are based on an incomplete analysis. If policymakers are irrational just like others, the chances of success for the paternalist project can be put into question.
Results:
Our main findings are that 20.7% of all articles in behavioral economics in the ten journals contain a policy recommendation and that 95.5% of these do not contain any analysis at all of the rationality or cognitive ability of policymakers. In fact, only two of the 67 articles in behavioral economics with a policy recommendation contain an assumption or analysis of policymakers of the same kind as that applied to economic decision-makers. In the remaining 65 articles, policy recommendations are proffered anyway.
What difference does it make?  Berggren's handy flowchart elegantly shows what's at stake:

berggren.jpg

[Bigger version here.]

My one quibble: What Berggren calls the "behavioral political economy insight" is overstated.  The question is not, "Do politicians and bureaucrats have problems with their decision-making in the form of irrationality, poor self-control, cognitive limitations, etc.?"  The question is, "Are politicians and bureaucrat's problems with their decision-making in the form of irrationality, poor self-control, cognitive limitations, etc. serious enough to make the policies they choose a net harm?"




COMMENTS (18 to date)
Scott writes:

Couldn't a paternalist just respond to the first question by saying that private solutions exist but public solutions are more efficient, effective, etc.?

jc writes:

Could their counter be that behaviorally trained experts (who they will rely upon) are relatively immune to making these mistakes?

Many traits, e.g., IQ, being an expert in one's own domain, even believing that dispassionate analysis of hard evidence is the way to make choices, etc., may be less helpful than we like, or even orthogonal, with respect to how prone one is to cognitively biased thinking.

But is training people to look out for behavioral flaws fairly effective (at least when they're consciously looking)?

While there may also be private applications, and it does nothing to address public choice issues, has anyone pushing for government oversight (to rectify poor outcomes due to cognitive bias) made explicit arguments along these lines?

(My assumption is that arguments like this are barely even tacit, i.e., it's simply automatically accepted that potential market failure justifies government intervention...especially to folks who are either: (a) biased in this direction or, (b) simply not specialists in the analysis of links between government policies and real-world performance, and are thus illustrating the dangers of authoritatively stepping into someone else's field...though I also wonder how many simply add on "policy" arguments as interesting "discussion fodder" speculation, or to help make their work seem more important by offering real-world implications and solutions. In other words, they're married to their primary findings, while the rest is just kind of tacked on for whatever reason.)

Austin writes:

A quibble quibble.

"My one quibble..."

If you're going to make that quibble, it should be extended to every step.

--Do economic actors have sufficiently large problems...
--Do sufficiently satisfactory private solutions exist?
--Do politicians have sufficient incentive to...
--[Caplan's change] Do politicians have sufficiently large problems...

Perhaps more elegant to state, "By 'problem' or 'solution', we imply economic significance." than to insert it everywhere.

austrartsua writes:

This is all a nice utilitiarian analysis but it suffers from its own cognitive biases. In particular it places no utility on freedom itself. (Isn't that one of the main problems with utilitarianism, the very definition of utility is subjective?).

My point is that many people like myself see freedom as an end in itself. Let's say for arguments sake that a credible case has been made for government intervention in a particular area. The problem is that this analysis does not consider reduced freedom as an adverse impact. However this should always act as an impediment to intervention: intervening in people's lives removes some of their freedom, treating them like children. In fact, at this point, plain old utilitarian arguments can come back in. If you infantilize people, they will behave like infants. They will lose their autonomy. They will not be fully alive, fully conscious. In effect you have murdered them.

All of this means that freedom must be given the benefit of the doubt. People must be allowed full control over their own choices. Any government intervention must be seen as a grave, terrible last resort.

Jim Dow writes:

The problem with the flow chart is that it assumes that the existence of irrationality/bias is a "yes or no" question. A better question is do decisions making skills vary across the population (for various reasons) and is there a sufficient gap between the decision making skills of the regulators and the regulated?

This blog, among others, has argued that the poor are poor in part because they make bad decisions. In other words, given the same circumstances, some people will make mostly good decisions and and some people will make mostly bad decisions. If so, that opens up the possibility for an understanding of cognitive biases to lead to an activist government policy. Whether it would work in practice is likely an empirical question tied to the particular issue and so the flow chart doesn't get us very far.

Miguel Madeira writes:

The "behavioral economics" could be used to defend government intervention in two ways:

a) "If people were perfectly rational, the market will work well; unfortunately, most people are not perfectly rational, and because that we need government intervention"

b) "If people were perfectly rational, our proposed policy will not work (because the agents will adjust their behavior according to new policy); fortunately, most people are not perfectly rational, and because that this government intervention will work"

In discussions about "rational expectations", "ricardian equivalence", etc, are common variants of the position b).

I think the objection "Human irrationality is a poor argument for government action because officials are human" only applies to the position a); if anything, in the case of position b), is even better if government officials are also irrational (they will put the policy in practice with more conviction, if they also don't understand that their policy will not work if people were rational).

Jayson writes:

There's another, related issue here that bears mentioning. Virtually all the behavioral economics papers that make a policy prescription do so not based on actual evidence the policy will work but based on a logical deduction that the bias observed in their particular experiment will be offset by the prescribed policy.

ThaomasH writes:

Isn't this just a case of comparative advantage? "Behavioral economists" look at the "demand side" of intervention and "public choice" economists look at the "supply side?"

Jesse C writes:

austrartsua - Great point!

I find myself more and more frustrated that freedom isn't seen as THE primary issue/hurdle when contemplating government action - ideally, it would usually carry such weight card as to make the suboptimality a "poor" decision immaterial.

I'm further frustrated that people appoint themselves as philosopher-kings and contemplate the utility of other peoples' freedoms. (Even this post makes me feel a bit 'too big for my britches.')

Aaron McNay writes:

I seem to recall Richard Thaler complaining in several articles that this critique of Behavioral Economics is brought up regularly. I remember him saying that he does not understand why it keeps being brought up, as has addressed it in the past. My problem with his saying this is that the only times I see him "address" this issue is by bringing it up and then quickly moving on without actually examining it.

For example, here is an article where Thaler brings up that the federal government has a spending problem that seems to follow the same irrationality problems that Behavioral Economics says individuals have. His solution? Elect different politicians. That does not seem to me to be actually addressing the issue.

http://www.nytimes.com/2011/08/28/business/economy/washington-should-try-a-little-prudent-self-restraint.html?_r=2&ref=business

Does anyone know of a section of one of his books, papers, or articles where he actually examines the issue of government official irrationality; at least beyond only bringing it up as a possible problem and then ignoring it?

Gabriel Weil writes:

The biases of government officials and other would-be nudgers are discussed extensitvely in Nudge, Thaler and Sunstein's book laying out the case for interventions based on behavioral econ: http://www.amazon.com/Nudge-Improving-Decisions-Health-Happiness/dp/014311526X

Ben H. writes:

Most of the recommendations that I see coming from behavioral economics (and I am not a behavioral economist, so perhaps I am wrong here) come in the form of "nudges" that alter a default choice. For example, the default choice is often to assume that people do not want to be organ donors (opt-in); if we shift to assuming that they do want to be organ donors (opt-out), organ donation will increase substantially with no loss of freedom. With policies like that, much of the proposed flowchart seems to me not to apply, because there is not a choice A that represents "no intervention" and a choice B that represents "intervention"; there is merely a choice as to which option ought to be the default, and the government has to choose either A or B (since, following this example, it needs to record whether a person is or is not an organ donor). Similarly with assuming that people do want to participate in a 401(k) retirement plan with a high donation level, versus making them opt in to that; and so forth.

Robert K. writes:

@Ben H.

"...With policies like that, much of the proposed flowchart seems to me not to apply, because there is not a choice A that represents "no intervention" and a choice B that represents "intervention"; there is merely a choice as to which option ought to be the default."

I think this is far rarer than you believe. Consider your very own example: organ donation. There *is* a no-intervention option. It's the current default of no donation. In that case, if you die, the government does nothing visa vis your organs. If you do donate, they scoop out some of your insides for others or use your body for science or whatnot. If someone were to dig up a corpse without permission it is a trespass. It seems odd to claim that the government assuming it can own your body when you're dead and do what it pleases with it is somehow a non-intervention that results in no freedom lost.

Likewise for a 401K. If I, personally, were to steal money from you and place is all faithfully in an account earmarked for your retirement, this would not be looked at as a morally or legally neutral action. Why the special categorizing for government agents?

Miguel Madeira writes:

The example of the organ donors can be considered a case of "fortunately, most people are not perfectly rational, and because that this government intervention will work".

This is a case where the point is not to correct the "errors" committed by the "irrational" individuals (and making them to take the "rational" decision); it is almost the opposite - to exploit the irrationality of the individuals, making them to take the decision that the government wants but with an illusion of freedom.

Like I said, this not really requires that the rulers should be more rational than the ordinary citizens.

James writes:

Robert K:

The situation is more subtle than you seem to recognize. A default to opt in to a 401(k) is different than someone forcing you to contribute. It's more akin to someone forcing you to contribute until you show the first sign of unwillingness and then letting you do what you like. It's not morally neutral but not as severe as the analogy you present.

It would be a step in the right direction if Thaler and people like him would start to advocate replacing force with nudges, e.g. making Social Security opt in by default but letting people opt out freely. So far, Thaler's policy recommendations have all been to replace freedom with nudges.

Maybe some behavioral economics scholar can point to a case where Thaler or some other behavioral economist explictly calls for replacing force with a nudge.

James writes:

Scott,

The response you propose contains implicit ansers to all of the other questions in that decision tree.

The real mystery is why none of the believers in interventionist policies show any interest in what they ought to regard as the biggest behavioral biasof all: People in the public sector are able to improve upon the decisions of the people in the private sector in one situation after another. People in the private sector are so unwilling to go along with these improvements that they need to be threatened with fines and incarceration to get them to comply.

If this is how the world really is, why do behavioral economists waste time on small potatoes like hyperbolic discounting or endowment effects when they could be studying "paternalism aversion" instead?

Noah writes:

But officials put their reasoning on paper where we can easily spot any potential irrationality, whereas people's irrationality in daily decisions is not thrust in front of them on paper where they can plainly see it.

Stuart Buck writes:

The original post's objection to behavioral economists is one that I don't understand -- it seems like an easy gotcha, but doesn't actually apply (or if so, doesn't apply all that obviously for most policy questions).

Take 401(k) donations. The claim here isn't that ordinary citizens are dumb while anyone who works for the government is smart. The claim is that we are all biased some of the time, and more rational some of the time. So when we are making long-term decisions about our own well-being, we should hopefully do so in our rational moments.

By analogy, my rational self might think, "I want to be healthy and non-obese, and to live a reasonably long life. Thus, I shouldn't eat a gallon of ice cream every night." But my biased self will respond, "Look at that gallon of ice cream sitting there. A little taste won't hurt. And a little more, how could that hurt either." Before you know it, I've eaten the whole thing. So in my more rational moments, I might decide not to buy ice cream very often, or to buy it only in small individual packages that make it more difficult to keep eating non-stop. It's not that I am smarter than myself -- I'm the same person the whole time -- but that I want to structure my life so that my rational side wins out over the "pig out" side.

Similarly, in my more rational moments, I think, "I should really be saving in my 401(k)." And if I get old and haven't saved, I will think, "Wow, I really should have saved in my 401(k)." But in my less rational moments, I think, "Yeah, I'll start saving someday, as soon as I get around to it." But then I don't get around to it, because it takes time and trouble to sign up, and I always seem to have better things to do.

Since there has to be some default rule as to 401(k) participation, why not make the default rule be the pattern of saving that our more rational selves would like, and that we ourselves will appreciate when we're older and have more money saved up? There's no option of not having a default rule, after all. Similarly, I might decide that my personal dietary default rule should be "No purchasing of ice cream at the grocery store except in special circumstances" rather than "Buy at least a gallon of ice cream every time I'm in the store," because I know for sure that the latter default will result in worse long-term outcomes.

Thus, it's not that government officials are any smarter or less biased than anyone else. They are just as biased as everyone. But in choosing a default rule here, they are making the rational decision that their own rational selves (and the rest of us in our rational moments) find to work best, rather than a decision that we know will result in outcomes that we will later regret ourselves.

Comments for this entry have been closed
Return to top