Arnold Kling  

The Rabbit Hole of Behavioral Economics

One Gritty Piece of Academic R... Further Adventures with Econ V...

Frances Woolley writes,

Once you fall down the rabbit hole, you just have to keep on going. If people's choices are not a reliable guide to their well-being, you have to turn to something else. Ask people how happy they are and measure well-being in terms of happiness. Evaluate health care spending by looking at objective measures of health, such as mortality, morbidity, or survival rates. Chuck out the entire elegant theoretical framework of welfare economics.

Read the whole essay.

If the whole purpose of doing economics is to make policy recommendations based on application of the "elegant theoretical framework of welfare economics," then findings from behavioral economics represent one challenge. Hayekian considerations of local knowledge are an alternative challenge. I think that even if one could produce the grand synthesis of behavioral economics and neoclassical economics for which Woolley might long, the Hayekian problem would remain.

Comments and Sharing

COMMENTS (9 to date)
David R. Henderson writes:

This is a really nice essay. I love the way Woolley soul-searches in trying to figure out the right answer. I think I might use this essay, especially because it's so short, in my class.

Tracy W writes:

I've made a comment on the paper, but there's one thing I want to repeat here more generally.
As Woolley says, once you fall down the rabbit hole, you just have to keep on going. And what I get to as I keep going down the rabbit hole, is, well, if research indicates that people are bad at making choices, that they are affected by things like framing, salience, other options, then presumably all this applies to the people doing the research. If I can't trust people's decisions about their own healthcare, how can I trust psychologists' decisions about research into how people make their own healthcare? Isn't that set of decisions as likely to be biased as the healthcare decisions?

The odd thing is that I'm inclined to place a lot of value on the reports that people make bad decisions, so my intuition runs with the psychologists, so I'm in this odd situation where the more I trust their results, the more I find myself thinking that I should on this basis alone be distrusting their results - that the more I believe them, the less I should believe them. (And also the less I should believe any other scientific paper, but that's not so paradoxical).

hamilton writes:

From the article:

"Economists like to begin with a general model, work out its implications, and then go to the data to test the theory. ...

Psychologists begin by observing people's behaviour, looking for generalizable patterns."

It seems to me that economists (a) believe you can develop a 'general' view of the world without first finding out how, in general, the world functions, and (b) think that assumptions should be made based on something other than empirically observed facts about how said world works. Does that fit with others' experience of economics, and if so, do you think that is a strength or a weakness of economics?

(If it doesn't square with your experience, I'd also be really interested in said experience.)

For the record, I am a student of economics who wishes he had not sunk so much time into it, and was instead a psychologist.

Tracy W writes:

It seems to me that economists (a) believe you can develop a 'general' view of the world without first finding out how, in general, the world functions, and (b) think that assumptions should be made based on something other than empirically observed facts about how said world works.

Yes, and I think that's because economists are trying to explain how the world works. If you want to explain empirically-observed facts, you can't start by assuming them (or to be pedantic you can, but it's boring).

I think a big chunk of the difference in focus between economics and psychology is that many of the facts that economics tries to explain are relatively obvious - eg that diamonds sell for more per gram than water even though water is essential for life, that some countries are richer than others, that sheep and cows are in no danger of dying out, but fish stocks are. Psychology though is working on trying to extract reliable facts from a mass of individual data where the details can be overwhelming. Of course, the two overlap, and we have things like the EMH, which was at first an observation stated on empirical work, and only a bit later Samuelson put the theoretical underpinnings in.

Alex J. writes:

Speaking of down the rabbit hole:

[i]"For example, opt-out policies have much higher take-up rates than opt-in policies. But in picking between and opt-out and opt-in policy, we still need to know in an ideal world, would we want a high take-up rate or a low-take-up rate?"[/i]

Who's the "we"? Who's picking? I'm better served by choosing my own experts than by hoping my experts do the right kind of soul searching. If Woolley comes to a "first, do no harm" conclusion, he, or those like him, will likely be passed over in the search for rule makers.

hamilton writes:

"Yes, and I think that's because economists are trying to explain how the world works."

This is my difficulty: I don't know what that means. And I mean that sincerely, not snidely!

The models economists build do not actually provide explanations of how the world works, in the sense of the actual causal mechanisms. Economics provides "as-if" stories that (sometimes) fit the data, even when the individual causal pieces of those stories are demonstrably wrong. For my thinking, I'd rather have a series of demonstrably true empirical claims that do not fit well into a generalized framework derived from (not empirically demonstrated and often empirically incorrect) axioms.

So for those who like the framework, what benefits do you get from it? (And again, this is not meant to be a sarcastic question, so I apologize if it sounds like that!)

Tracy W writes:

Hamilton, I don't understand what your terminology means. What is an "as-if" story? What do you think is an actual causal mechanism? Can you give some examples of economic theories that you think meet these descriptions, or don't?

This question is meant sincerely.

And also, what do you think is the generalized framework in economics? Because I can't think of such a framework in macroeconomics, though perhaps I'm not seeing the wood for the trees.

hamilton writes:

Tracy W,

Sure thing. The rational model of addiction is a good example of an as-if story. People becoming addicted do not actually weigh present discounted value of the highs they will receive--taking into account the increase in the product they require to achieve said high--against the health, material, and relationship costs of their addiction which occur in the future. But modeling people as if they make this tradeoff does a so-so job of explaining the choice to smoke cigarettes. This does not, according to people who deal with addicts, do a very good job of describing either how addicts became addicts, nor how addicts amidst addiction describe it (the references on the wiki page for this topic, if not the page itself, do a good job of summarizing the debate on this model, btw).

In fact, all "max U subject to..." is an as-if story, though again it does a pretty good job in the broad sense of explaining some (though usually very little) of the variation in the data. But the specific assumptions underlying it aren't just "reasonable assumptions"--there is actual empirical evidence to suggest that they are straight up incorrect (e.g., completeness & transitivity).

Constrained utility maximization is, to me, the generalized framework of economics. You can do a lot with it. But developing from axioms is not, for me, a fraction as compelling as trying to find the actual decision problems that the individuals whose behavior is under examination are grappling with, as opposed to how I would externally frame it for them were I to start from core micro theory.

On macro: um... yeah. I'm with ya. Maybe dynamic stochastic general equilibrium with real and nominal frictions when it comes to business cycles? Perhaps a macroeconomist could tell us. Certainly, one can say that all economics does (or should) go back to micro-foundations, but it certainly doesn't always feel that way to me (nor, does it sound like, to you, either:-).

Tracy W writes:

Hamilton - thanks, now what do you think would be an example of an actual causal mechanism, which is not a "just-as" story?

Not knowing this, I'm going to attempt to answer your questoins anyway.

For macroeconomics, I think that most macro economists would agree that ideally their subject would go back to microfoundations, but no one has solved the aggregation problem, and perhaps it isn't solvable. So in other words, macro economists have an ideal that they can't work out how to achieve.

This sort of problem is not uncommon in other areas of science. For example, physicists have been living for decades with the problem that quantum mechanics and relativity theory are incompatible, but both appear to be true. Evolution apparently suffered for a few decades from the problem that scientists didn't understand how genes got passed on (Mendel's contributions being overlooked), and thus how the variation in offspring was preserved, rather than children becoming simply the average of their parents.

In the case of microeconomics, we know that people manage to allocate their time to the tasks of finding food, water, raising children (and in much of the world, providing themselves with shelter and clothing), because otherwise we would not be around to worry about how they do so. So we know that people are solving the problem of how to allocate their time amongst different objectives, the resource allocation question, at least well enough to survive. The question for economists is "how are they doing that?" What's taught in microeconomics is what we think is the best model for how people do that, around at the moment. That doesn't mean that it's perfect, but, well, whatever the empirical data about preference incompleteness, clearly people are managing to make decisions. Normally people don't starve to death because our preferences between different food stuffs are incomplete, nor non-transitive.

Of course it would be great to have a theory that explains the empirical fact that people do make decisions about resource allocation, with truly solid empirical foundations. Perhaps you'll be the person to come up with one.

But generally, not just in economics, but in all areas of human decision-making, you can only replace a flawed model with a better model. People to have to make choices about the future, if the best models they have for making those choices are flawed, they'll still use those models. In engineering school, when I was taught about the mathematics of infintestimals (treating a differential as a very small bit), I recall the lecturer saying that when physicists started using this it was valid mathematics, then mathematicians decided for about 100 years that it was flawed and all wrong, yet physicists and electrical engineers kept using infintestimals because they wanted to solve problems and this was the best way they had of doing it, then a mathematician went back and worked out the detailed foundation to the mathematical theory. In the field of economics, people are going to ask questions about "how will people react to a change in policy (such as a new tax, or government spending)?", and people are going to try to answer that question. Similarly psychologists get called in to answer questions like "is this person sane enough to stand trial?" or "is this person likely to reoffend if we release them from prison?", or "what should we educate people in?", and no matter how bad psychologists' models are for answering those questions, they're going to keep being asked those questions. Shifting to psychology won't avoid your problems "as if" stories, unless I've missed something massive in my reading about psychology.

Comments for this entry have been closed
Return to top