Math Camp and Graduate Economics

Gabriel writes that he has started attending "math camp" at Clemson this week. For those who are not familiar with the rituals of economics graduate programmes, "math camp" refers to a short course in advanced mathematics that all incoming graduate students are required to take, generally before the formal economics courses begin, at many top economics schools, including The University of Western Ontario and, apparently, Clemson University.

When I was a young turk in economics (50-odd years ago), I would have thought "math camp" was a great idea. I saw the increased use of mathematics in economics as a way to make the discipline more scientific and more rigourous.

I am pretty sure this is a mistake now. Here is why: 

Some years ago, I began using the Canadian version of Paul Heyne's marvelous text, The Economic Way of Thinking, a slightly watered-down version of the classic intro text by Alchian and Allen (University Economics). As part of my work with the graduate TAs (whom we could not require to attend our lectures), I asked them to take the exams I had compiled and provide feedback on the questions. To my dismay, they never scored higher than 70%, and their average on several of the exams was less than 50%. They may have known math pretty well, but they certainly knew very little economics.

The problem is that in graduate economics programmes, we have learned that the surest way to have students churn out dissertations that might lead to some piddly publications is to admit math and science majors who know little or no economics. These students excel at math and do not need a "math camp". But they don't know any economics.

If we are going to require incoming graduate students to take "math camp", then it is even more important that we require them to take "Alchian and Allen" camp; more than math camp, we should give them a two-week course "the economic way of thinking."

Comments

[KipEsquire]

My 1st-year econometrics professor in grad school once summarily declared: "All economics professors are merely failed mathematics professors."

[Mike Moffatt]

"The problem is that in graduate economics programmes, we have learned that the surest way to have students churn out dissertations that might lead to some piddly publications is to admit math and science majors who know little or no economics. These students excel at math and do not need a "math camp". But they don't know any economics.

If we are going to require incoming graduate students to take "math camp", then it is even more important that we require them to take "Alchian and Allen" camp; more than math camp, we should give them a two-week course "the economic way of thinking."

But don't economists teach that people respond to incentives?

If it is, then what incentive do programmes (or more specifically, the profs who run the programmes) have to ensure their students understand basic economics (rather than understanding basic math)?

Should profs put the needs of the field ahead of their own?

[Nick Rowe]

John: From my own experience, I have an awful feeling you may be right; that many graduate students in economics cannot handle basic first year economics questions. But I am wondering how we could better test this.

Your own example of having your TAs answer the first year exam provides one test, but without seeing the actual questions, and how they were graded, it's not 100% convincing. Maybe the questions were very specific to that particular text? In other words, your and my anecdotes won't change things much, but a more public test might.

Here is a proposal for you (I'm asking you to do all the work, but then you do have the blog!): write 10 multiple choice questions that any good student who has completed first year economics ought to be able to handle, regardless of which Intro text was used, or in which country it was taught. Post those 10 questions on your blog, then we (and anyone else) can try those questions on our TAs, on our first year students (before and after they take the course?), and compare the results. (Even better if you can figure out one of those web-based self-test programs.)

Really simple basic questions like:

Harry spends all his income on milk and pizza. If his income rises by 10%, the price of milk rises by 10%, and the price of pizza rises by 10%, then:
a) he buys more milk and less pizza
b) he buys less milk and more pizza
c) he buys more milk and more pizza
d) he buys less milk and less pizza
e) none of the above

By posting 10 such questions on your blog, the questions are open to scrutiny, the test is more credible, and we have a common standard so we can compare results.

[EclectEcon]

Nick:

The exams were multiple-choice questions. To see my old exams, go to this site:

http://home.cabletv.on.ca/~jpalmer/Eco020/Exams020/exam020.html

[Nick Rowe]

Thanks John.

I had a quick read through one of your exams. In general, these look like good challenging questions, but I am not altogether surprised that an economics graduate who had never taken your course might legitimately have difficulty with some of them. Some questions are Canada-specific (for example the ones on competition policy, or the effect of minimum wage on employment of auto workers where foreign students might not know that Canadian auto workers are paid high wages). And some questions are perhaps course-specific (for example it took me some time to figure out where you were going with the questions about sales taxes being regressive or not, and I'm still not 100% confident that I have interpreted those questions correctly). And some stuff we simply forget (for example, I had to try to remember what was on the axes of a Lorenz curve). I confess I would have got significantly less than 100%, without having taken your course.

I am talking about our grad students' background in much more basic economic reasoning. Here's another anecdote: a PhD thesis with a competitive (price-taking) general equilibrium where each firm had increasing returns and falling LRAC. That in itself is not shocking, because everybody makes mistakes, and sometimes misses seeing something in a complex mathematical model (or elsewhere). What was shocking was that he failed to see or understand the problem when I pointed out that he had a competitive equilibrium with IRS; and so did some of the younger profs! My drawing a downward-sloping LRAC and horizontal MR=AR curves was no help. Finally someone calculated the second-order conditions for a profit maximum, and figured out that something was indeed wrong. I asked a couple of political scientists, and even sociologists, and most of them at least vaguely understood that there was some sort of conflict between increasing returns to scale and perfect competition.

I don't blame the individuals; I blame all of us who teach and do economics. But we need a test of general economic literacy to make the case.