Bryan Caplan  

How Econ Melts Your Brain

PRINT
Using monetary policy to preve... Low interest rates do not call...
Learning economics changes your life and improves your mind.  As Dilbert creator Scott Adams insightfully observes, "When you have a working knowledge of economics, it's like having a mild super power."  Yet when grading economics exams, I often notice a serious downside of studying economics: Many econ students literally get detached from reality.

I don't mean something lame like, "Some econ students disagree with me.  How dare they?"  What I mean is: I ask a question about the real world.  The question even contains the phrase, "In the real world..."  Then instead of discussing the real world, many economics students tell me about a model they learned in class.  Sometimes their answers even contain the phrase, "Assume model X."

For example, an exam question might ask, "What determines the price of water when two individuals bump into each other in a remote desert?"  Many economics students will then start talking about supply and demand, or even state, "Assume perfect competition.  Then blah blah blah." 

This wouldn't be so bad if the students would at least argue that, contrary to appearances, the perfectly competitive model applies.  But when students take a model for granted despite the violation of its core assumptions - such as no individual has any noticeable effect on the market price - something has gone terribly wrong.

Learning how to logically analyze hypothetical social situations is a great skill.  Economists should proudly teach it.  But mastering this skill can and often does melt students' brains.  Before studying economics, no one would imagine that you can discuss models in lieu of discussing the real world.  After studying economics, I'm sorry to say, many students tacitly embrace this absurdity. 

As economic educators, we bear a lot of the blame for the misconceptions our students acquire on our watch.  How can we do better?




COMMENTS (35 to date)
Grant writes:

Hmmm. Maybe being more self-conscious about the epistemology that under-girds modelling as a method--a firmer understanding of the epistemology might help students grasp the limitations of the methodology? I dunno. Too much philosophy, maybe.

I think economists, professors especially, sometimes worry about undermining their authority if they're clear and emphatic about the limitations of the tools they use.

Navin Kumar writes:

I usually love your posts but this is just silly. Students assume that they have to apply economic principles in an economics exam? What is this world coming to?!

Russ Nelson writes:

It seems like a very non-straightforward answer to that question. You'd have to make a lot of assumptions without turning the answer into a book. Like, "Assume the individuals are not willing to use violence." or "Assume that both individuals' ability to use violence is identical."

If the individuals aren't interested in free trade, then you have to do some kind of power analysis. Do the individuals know each other? Do they have any reason to trust each other?

I could go on and on for an answer that would be correct, but *extremely* time-consuming to grade.

Enial Cattesi writes:

It must be related to signaling somehow ... Joke aside, when you have a hammer, everything looks like ... signaling.

Matt Moore writes:

This is likely due to an overly focused exam preparation strategy. The individual hasn't moved at bargaining theory and was hoping it wouldn't come up. When does, they shoehorn in some technical knowledge they did study in the hope of getting some credit

Solution is to have one question on each main aspect of the course.

david writes:

It's an exam. Two paragraphs arguing that a shadow price is an inappropriate methodological device for considering how two individuals stumbling into each other in the desert would interact would be a poor gamble on scoring well.

The students sense - correctly - that they are being asked to exposit on some form of constrained optimization of underlying preferences. They're not sure what the desired form is, so they guess.

john hare writes:

It sounds like a lack of communicating on their particular level. Often it might be a lack of personal relation to the issues. "You're going to die if you don't get some of the water." Transform from two strangers to the individual and a stranger.

I have to explain some concepts to employees that aren't, like myself, a few decades past college age. Not making my point gets expensive so I find a hook that works on that particular mindset when I can. "Watering down concrete (to make it easier to handle) is like someone watering down your drinks." "Short (ineffective)strokes with the tool says a lot about your tool." And so on for construction workers. I have little idea of the particular hooks that make it personal to college kids.

Yaakov writes:

It is not unique to economics. Richard Feynman wrote about this problem with physics students in Brazil, who memorized everything they learned but could not apply it even in the simplest context. Read his book: "surely your joking..."

It happens also in Jewish studies. The final exam in Jewish law in some schools in Israel included a story, in which the students were to discuss different points and how they match the Jewish law. It was considered the hardest chapter in the test.

The solution, as in all education, is simple: practice. If students practice answering real world questions, they will succeed in doing so.

In electrical engineering I instructed a lab and put a little emphasis on having the students understanding the graphs. At first, the students had a hard time at it. But once word got around that that was what I was asking everybody came prepared.

The only question that remains is: what do you give up in order to spend resources on practicing answering real world questions?

Andrew writes:

This is not a phenomenon unique to economics. It is well known in physics teaching that students will not apply what they learn to the real world. In fact if a test question appears like it could occur in real life they will often fail to apply the physics they have learned, even when it is applicable. See the "force concept inventory" test.

It is as if students learn two types of physics, real life physics they use to solve actual problems, and classroom physics used to solve test problems. There is a firewall between these and on any given problem they choose the most applicable method.

Based on a talk by Eric Mazur one solution is to have students grapple effectively in groups with real world problems during class, and discover/agree on the correct answers themselves, with minimal facilitation. He had a system for accomplishing this, using targeted questions, class polling and group discussion.

mico writes:

They view your questions as framing to provoke a response in terms if concepts discussed in class. They do this because many other subjects - e.g. maths - deliberately structure exams this way.

Tracy W writes:

This is inflexible knowledge. It's apparently an unavoidable part of the learning process (although some rare people either skip it all together or just get through it too fast for the onlooker to notice). Basically everyone, or at the least the vast majority of people always go through:
1. Complete ignorance.
2. Rote knowledge: they've heard the words, but don't know what it means and can't solve any problems.
3. Inflexible knowledge: they can use the knowledge to solve problems very very close to the format they were given in.
4. Flexible knowledge: they can apply this knowledge in differing situations.

Getting to the 4th takes time, like years. It took me about 3 years of calculus (in NZ studied as part of a more general maths course) before I really understood why setting the differenced equation to zero resulted in the right answer. My teachers had given us explanations, and I had memorised the explanations and reproduced on tests, but basically it was just magic until then. Apparently practice in a variety of different situations on a variety of different problems helps.
If anyone knows how to get to flexible knowledge directly, you've made an amazing scientific discovery.

foosion writes:

Worse, many policymakers do this. For example, they demand "the free market" proceed without regulation in situations involving major market imperfections, such as monopolies, restricted information, high bargaining costs, public goods, large externalities, etc.

Pat writes:

I doubt those students are that detached in the real world. I think they're probably answering it the way they think you want them to.

gamma writes:

As a graduating senior, I have two observations.

First, Yaakov's observations hold. If you do it in class, and tell the students they will need to apply this kind of thinking in your exams, they will incorporate real-world thinking into their studying.

Second, students learn quickly that a full time course load makes it impossible to extract all the knowledge from all the classes. So they have to maximize their utility when selecting the knowledge they consume. There have been times when I have been frustrated that I couldn't give each class the time it deserves, and have sat in class, staring at the teacher bitterly and thinking, "Just tell me what I have to do to get an A!"

Which takes us back to my first observation.

Tom West writes:

Just wanted to say that I really like Tracy W's rungs of knowledge. I hadn't heard of it before, but it nicely encapsulates what I've just sort of understood for years.

Thanks!

Fred Anderson writes:

In the early 1970's, I finally found a professor who could teach me statistics such that I understood and could apply it. (I wish I could remember his name to give him credit; It was an Applied Statistics course at UICC -- probably in 1971 or 2.)

His technique was to show you the history of the idea. What was the original problem, and how did the original investigator grope his way through it -- fits, starts, dead-ends, refinements -- to eventually get to a solution. And to eventually encapsulate that solution into a formula / a theory.

I wonder if our problem as teachers isn't that we know too much. And to master all of that knowledge with our limited resources of intellect and time, we've had to develop a shorthand of acronyms, jargon, abstractions and theories from which the confusing clutter of reality's nuances has been surgically ablated.

Then, on the model that education is just accelerated experience, we try to show naive students the desiccated conclusions without wasting the time to push them through the experiences that led to those conclusions. We hand them a raisin and ask them to imagine a sun-drenched vineyard.

I found in my own teaching that a little carefully chosen history helped.

William Bruntrager writes:

This complaint is absurd, for two separate and important reasons.

The first is that the whole point of using models and modelling is to simplify the real world and make it comprehensible. When you ask people to describe how something works in the "real world," they will use a model to develop their answer. The question is, will it be a reflexive, on-the-fly model of the type people tend to use before they study economics, or will it be an imperfect and inaccurate, yet tested and developed model of the kind students learn while studying economics? For all the validity of your protest that "the assumptions don't hold!" when trying to apply economic models to real-world situations, I often find that the logic itself doesn't work in models developed on-the-fly.

Secondly, as others have pointed out, you seem to be lamenting that students expect an economics exam to be testing their knowledge of economics, rather than inviting a wide-ranging discourse on fundamental human problems, drawing from all fields of human knowledge. By the time they get to your class, students have built up a lifetime of experience taking exams, trying to determine how the material they learned in the course can be best translated into an answer that will get a high mark from the teacher. And you expect them to discard all of this, just because your question contains the words, "in the real world"?

So in summary, you pose a series of extremely difficult questions about which intelligent people surely disagree, give students 10-15 minutes to answer each one, provide no guidelines as to how they should even approach answering the questions except to explicitly reject the approach endorsed by the course you teach... and you are consistently disappointed with how your students perform on these exams. I disagree with you that this experiment convincingly reveals a profound effect of learning economics on students' brains.

Jon Murphy writes:

I see that fairly frequently, especially when it comes to minimum wage. Someone always whips out the ol' monospony argument, not realizing (or not caring) it's completely inappropriate to the real-world situation.

As an aside, I work for an economics consulting company. We hired an economist a little while ago. The guy had 2 masters and could wax philosophical all day long about various models, but had no clue how they applied to the real world, or if they applied to various situations. He didn't last too long...

Bill writes:

The real world is a special case.

Tony Irons writes:

Make students do something in the real world before they are admitted to graduate level programs in Economics; just like MBA programs.

I began my graduate studies at 34 after having run my own business and lived in over a dozen states and countries (and having visited dozens more). It is unfathomable to me that I could have become a "low-level-calculus-&-algebra, model-dependent, detached-from-reality" "economist" (read: policy engineer) after having experienced the real world.

Every once in a while you might get lucky, but going from a socialist upbringing (public schools) into a quasi-wonderland (US university) and then straight to grad school gives all but the most intuitively economic-minded students almost no chance.

Doug writes:

You're asking for a novel. It's a trick question, Person A is a sociopath and immediately bites you!

Scenario is failing to take into account that the exam question is occurring in the real world, not a purely intellectual exercise. The students are providing what they think you want. Writing "in the real world" before a hypo is utterly meaningless.


Mason Econ '02

Tom West writes:

I see that fairly frequently, especially when it comes to minimum wage. Someone always whips out the ol' monospony argument, not realizing (or not caring) it's completely inappropriate to the real-world situation.

Actually, the minimum wage is my classic for "ECO 101 says this *must* be true, therefore any paper that finds differently must be disregarded."

My take is that the truth is so complicated and involves so many cultural/social/racial elements that for small raises, the straight economics aspects are almost (but not quite) noise compared to all the other elements.

And given these element change from location to location, everyone's idea of the "right" outcome can be proved correct somewhere! No wonder minimum wage is a good field to stake your claim. Everyone's right (if you ignore the non-conforming results)!

Jon Murphy writes:

@Tom West

And given these element change from location to location, everyone's idea of the "right" outcome can be proved correct somewhere!

Which is exactly why I oppose minimum wage, price controls, FDA, EPA, and many other forms of central planning. AS you correctly note, the elements of cultural/social/racial change so much from location to location (what Hayek referred to as the "particular knowledge of time and place" that any top-down procedure or program is doomed to fail!

A minimum wage of $X.XX may be too high for some places, too low for others, and exactly equal for a third. In the first case, the wage would have negative effects. In the second and the third, it would have no effect. This would allow an un-careful thinker to conclude minimum wage can have no effect, and therefore can be raised but that is not true. Thus, as Prof. Caplan says, "Learning how to logically analyze hypothetical social situations is a great skill."

Brad writes:

Since when do (today's) students know anything about the real world? They're short in the tooth and lack any real experience that might inform their thinking about a "real" world problem.

I "learned" econ in school but it wasn't until my mid 30s that I really got it. It just takes time.

Tom West writes:

In the second and the third, it would have no effect.

You mean no (or minimal) *negative* effect.

Let's not forget the main reason for minimum wage in the first place - to redistribute profits from the owner to the worker who may have a high marginal product, but doesn't have market power to demand anything above government mandated minimum wage.

(A goal you may not consider a positive, but is pretty much the whole reason for minimum wage and has wide spread social support.)

Thus the calculus is "does this positive outweigh the negatives of greater unemployment, etc."

As for central planning, well, every company does it as well, so I have no great horror of the idea. If business is to be favored over gov't, it has nothing to do with central planning.

Bill writes:

If a worker has high marginal product, why would it be necessary for the worker to have "market power" to be paid according to that marginal product?

Jon Murphy writes:

Let's not forget the main reason for minimum wage in the first place - to redistribute profits from the owner to the worker who may have a high marginal product, but doesn't have market power to demand anything above government mandated minimum wage.

Of course, that simply isn't true. I could give a multitude of reasons, but let's just say there is no decoupling of wages and productivity.

I guess one other thought on this: if a worker feels he has a higher marginal productivity than he is getting paid for (in other words, he is underpaid), then he is perfectly free to pursue opportunities elsewhere. If he is right in his assessment, he will achieve his higher wage. If he is wrong, he will not. If, indeed, he does not have market power, it is simply because his demands are too high; his perception of his wage is simply not accurate. I may think I am worth $1,000,000, but because the most someone is willing to pay me is $70,000 does not mean I lack market power.

As for central planning, well, every company does it as well, so I have no great horror of the idea. If business is to be favored over gov't, it has nothing to do with central planning.

Incorrect. Central planning refers to a top-down approach imposed by a central government authority upon everything within its jurisdiction, regardless of the "cultural/social/racial elements" (to use your words).

One final thought before I call it a night:

You mean no (or minimal) *negative* effect.

Incorrect. I mean no effect. If I meant no negative effect, I'd have said no negative effect. I do my best to be precise with my words for exactly this reason.

Simple analysis: if minimum wage is set above the prevailing wage in the area, it'll have a negative effect (a point which I assume you do not contend, as you did not object originally). If the minimum wage is set at the prevailing wage, it'll do nothing as everyone is already paid the wage. If it is set below the prevailing wage, then it'll have no effect as wages are already higher.

Pithlord writes:

if a worker feels he has a higher marginal productivity than he is getting paid for (in other words, he is underpaid), then he is perfectly free to pursue opportunities elsewhere.

This assumes that the worker would be just as productive elsewhere, and that there would be no transactional/frictional costs of quitting and getting another job. As Coase pointed out eighty years ago, on these assumptions, there wouldn't be firms at all.

Jon Murphy writes:

This assumes that the worker would be just as productive elsewhere, and that there would be no transactional/frictional costs of quitting and getting another job.

Oftentimes, these assumptions hold, especially the first one. A burger-flipper for McDonalds is just as productive at Burger King.

In fact, rival firms often do what's called "headhunting." They monitor candidates at rival firms and offer them more to come work for them, the idea being that said person is actually underpaid.

Really, the only major assumption I am making is that there is free movement of labor, which, in the United States, is not illegal (unless, of course, you are coming from another country).

As Coase pointed out eighty years ago, on these assumptions, there wouldn't be firms at all.

I am not familiar with that statement by Coase, but if he did say that, I'd have to disagree. Firms would exist (somebody has to produce), and they'd be incentivized to pay the right wage.

Pithlord writes:

A burger flipper has relatively low firm-specific human capital. But entry-level jobs are not typical, and they are inevitably the lowest productivity jobs.

The problem of industry-specific-but-not-firm-specific human capital is a major one, because it can lead firms to underinvest in training, because they are worried their competitors will get the benefit of it.

I am not familiar with that statement by Coase, but if he did say that, I'd have to disagree. Firms would exist (somebody has to produce), and they'd be incentivized to pay the right wage.

I am just summarizing Coase's "Theory of the Firm". Firms wouldn't exist if there weren't firm-specific human capital and transaction costs because it would be more flexible for everyone to be an independent contractor. The very fact that there are hierarchical structures of employment shows that the returns from firm-specific human capital exceed these benefits.

Jon Murphy writes:

The problem of industry-specific-but-not-firm-specific human capital is a major one, because it can lead firms to underinvest in training, because they are worried their competitors will get the benefit of it.

Odd that a firm would willingly harm itself in order to prevent a competitor from potentially sniping a candidate.

But again, that actually goes to my point: if labor is allowed to move freely, then a competitor could potentially poach an employee. The way for him to do that, however, is to offer a better compensation package. If a firm is indeed underpaying a person, in other words paying him less than his MPL, then it makes the poaching easier (and economically efficient). That is why there hasn't been a decoupling of wages and productivity. That is why I reject the argument Tom West put forth and why minimum wage as a way of "correcting market power" is such a foolish argument.

But entry-level jobs are not typical, and they are inevitably the lowest productivity jobs.

Which is why they are minimum wage jobs.

Tom West writes:

Um, Jon, are you saying that supply has no effect on wages, only demand? I'd assume that like every other commodity, if there's more supply of unskilled or semi-skilled labor than demand, then the price goes down, even while the marginal product of that work remains the same.

Also, if everyone is earning their marginal product, then by definition, *everyone* who is earning minimum wage gets fired when minimum wage goes up, something I don't see matching observation.

Now, I agree that it seems highly likely that there is some job loss when minimum wage goes up. After all, *some* workers were earning their entire marginal product. But over all, the effect seems pretty small, leaving me to believe that the *vast* majority of workers were earning more than their marginal product.

(I am sympathetic to the idea that minimum wage deprives low marginal product workers of any job. Like every economic policy, there are trade-offs. And if I can't list the costs of my preferred policy, then it means I have little useful to say on the topic.)

Daublin writes:

I believe you about the symptoms, but I don't think you have quite identified the problem. Any discussion is going to have an implicit model it is working within.

To see this, consider that a given number of words can express only so many concepts. Any reasonably short conversation is just going to have to rely heavily on shared context, including a shared view of the world. What is a view of the world if not a model?

As another way to look at it, a test is fundamentally examining whether a student has a good model for the phenomenon being discussed. So I would say that you just don't like your students' models, rather than saying they are wrong to use models at all.


Tel writes:
As for central planning, well, every company does it as well, so I have no great horror of the idea. If business is to be favored over gov't, it has nothing to do with central planning.

Not every company does it well by any means. A great many companies go broke on a regular basis, or weaken themselves to the point where they become easy takeover targets. Thus, you may find if you only observe successful companies you could convenience yourself anyone can do it.

Some companies avoid internal central planning by allowing considerable autonomy to branches and departments, the degree to which they do this is also selected by success or failure in the market.

The real problem is when you get a central planner not subject to any market forces. Either a government, that can prop up failures with tax money and monopoly regulations, or a sufficiently large corporation that has indirect means to stifle competitors, or a large union with the power of "hold up" over capital.

Tel writes:

My best short and accurate "real world" answer about two individuals meeting in the desert is "depends on the individuals" but if I needed the grade I probably would write more than that in the exam.

Comments for this entry have been closed
Return to top