Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
Still Torturing My Students, Behavioral Economics Edition…

October 22nd, 2015 · 3 Comments
Behavioral Econ

The other day, my students complained right before an exam that they couldn’t see the clock. So they got this:

classroom clock

I’d like to think that my students now know to be careful what they wish for. Today my students get their exams back, and I’m taking a page from the Richard Thaler behavioral playbook. From Misbehaving: The Making of Behavioral Economics:

Finally, an idea occurred to me. On the next exam, I made the total number of points available 137 instead of 100. This exam turned out to be slightly harder than the first, with students getting only 70% of the answers right, but the average numerical score was a cheery 96 points. The students were delighted! No one’s actual grade was affected by this change, but everyone was happy. From that point on, whenever I was teaching this course, I always gave exams a point total of 137, a number I chose for two reasons. First, it produced an average score well into the 90s, with some students even getting scores above 100, generating a reaction approaching ecstasy. Second, because dividing one’s score by 137 was not easy to do in one’s head, most students did not seem to bother to convert their scores into percentages. Lest you think I was somehow deceiving the students, in subsequent years I included this statement, printed in bold type, in my course syllabus: “Exams will have a total of 137 points rather than the usual 100. This scoring system has no effect on the grade you get in the course, but it seems to make you happier.” And indeed, after I made that change, I never got a complaint that my exams were too hard. In the eyes of an economist, my students were “misbehaving.” By that I mean that their behavior was inconsistent with the idealized model of behavior that is at the heart of what we call economic theory. To an economist, no one should be happier about a score of 96 out of 137 (70%) than 72 out of 100, but my students were. And by realizing this, I was able to set the kind of exam I wanted but still keep the students from grumbling.

I teach behavioral economics, and I am generally anti-grumbling, so I rejiggered the points on the exam so that it’s out of 137 points. Their grades are still essentially determined by z-scores, but I’ll take even an illusory Pareto improvement where I can get one, especially when I’m dealing with this during the grading process:

Feel free to use my animals as examples of negative marginal product of labor.

Update: Economists are such sticklers for citations…

Tags: Behavioral Econ

3 responses so far ↓

  • 1 Michael Makovi // Oct 23, 2015 at 11:14 am

    This is indeed a very interesting phenomenon. I’m glad it worked for you, and yet I am disturbed that it did work. But I am disturbed not as an economist but as a subjective human being.

    If we defined “rationality” as simply “choosing what is preferred more over what is preferred less”, the economist qua economist would not be surprised by this result at all. It would simply be one more subjective preference among many. Some people prefer vanilla to chocolate, and some people prefer a score of 96 out of 137 over a score of 72 out of 100. There is no reason to consider these students irrational, any more than we could pronounce chocolate the objectively better flavor. These students demonstrate a preference and they make a choice by weighing alternatives against each other, and this ought to be sufficient to consider them rational.

    Now, the psychologist would be very surprised indeed. This is in fact a quite peculiar and interesting subjective preference. But the economist qua economist does not ask where subjective preferences originate; he takes them for granted as given data.

    There is no reason why “the idealized model of behavior [should be] at the heart of what we call economic theory”. This idealized model of behavior does not accurately depict how humans make choices, and there is no reason to believe that this is how humans ought to make choices. To assert that humans ought to choose in the manner predicted by idealized economic models, is to assert a normative theory of value. It is to assert that one set of subjective preferences is objectively superior to another set. But scientific, positive economists are to eschew any such assertion. Our task is to take subjective preferences as given data and to study the resultant calculus of choice. If people have psychological quirks, then so be it; this is simply one more datum for the economist to consider, but it does not make people any less rational. It merely means that their subjective preferences are not what the economist expected them to be.

    As Ronald Coase said in criticism of Milton Friedman, it is not enough for our models to yield empirically testable predictions. We need our models to realistically depict the actual world. The reason, Coase said, is that we do not primarily wish to predict, but to understand. We could predict the 1970’s energy crisis, Coase says, under the assumption that the US federal government wanted to impoverish American citizens. This would yield a testable – and ultimately verified – prediction about the future: that the American government would institute an energy policy that would make Americans poorer. But while the model’s prediction is accurate, Coase says, it tells us nothing about the actual real world. We want a model, Coase says, that does not merely make correct predictions, but which helps us understand the real world. The idealized models of economic theory are therefore next to useless, if Coase is correct in this criticism of Friedman. These idealized models may yield interesting predictions, but they tell us nothing about how actual humans choose. (These idealized models might be useful, however, as counterfactuals. We might model an idealized way in which humans do NOT choose in order to contrast it with a more realistic model of how humans DO choose.)

    Coase, Ronald. H. 1981. “How Should Economists Choose?,” in: No editor listed. Ideas, Their Origins, and Their Consequences: Lectures to Commemorate the Life and Work of G. Warren Nutter. Washington, DC: American Enterprise Institute: 57-79.

  • 2 Michael Makovi // Oct 23, 2015 at 11:22 am

    If I may illustrate Coase’s point using statistics: if a model makes correct predictions using unrealistic assumptions, then the model is misspecified and the correlation is spurious. You got the right result for the wrong reason. But since the reason is wrong, you are likely to get the result incorrect in the future.

    Suppose you’ve got the data points (0,0) and (1,1). You regress the data as a linear equation with the variable (x) and arrive at the equation y = x. But suppose the true model for the phenomenon is y = x^2. You should have regressed the data with the variable (x^2). So you have misspecified your model. The two models happen to produce the right result for x = 0 and x = 1. But as soon as you try x = 2, the two models diverge.

    So any model which produces correct results (for now) for the wrong reason, is a misspecified model.

  • 3 William Sears // Oct 24, 2015 at 1:05 pm

    Makes perfect sense to me. We are conditioned to think that a score in the 90s is good, while a score in the 70s is bad. By matching the test scores with expectations, people can continue to use the shorthand of “90s are good”, instead of calculating Z scores.

Leave a Comment