quote:

Originally posted by Piano*Dad:
It's an intriguing idea, and Chetty is an engaging and brilliant expositor for it.

I have a number of reservations, however. Despite those qualms, the ultimate proof is in how well a different "introduction" prepares students for the rest of the existing curriculum, and for life after Harvard. That's an empirical question, and Harvard will be well-suited to test it since they will maintain two different versions of the introduction. I'm sure some clever thinker there will find a way to evade selection issues and craft craft a "natural experiment" to test which version works better, and for what purposes.

Some of "my" issues with Chetty's approach are well presented in the article. For instance, drawing conclusions from data without having a well-specified theory may be even more subject to bias than the traditional economics approach Chetty critiques. As I tell my students many times per semester, data never speaks for itself. It speaks through our theoretical scaffolding, or it speaks through the biases that you bring to the table. For people who have little or no training, relying on data alone to "prove" things is the hallmark of the person with a healthy dose of "my-side bias" or of the person who overrates their ability to know things. At some point, you must have a good theory that spits out potentially falsifiable hypotheses, or you are NOT doing science. I worry that many of the students in Chetty's class will confirm their biases rather than challenge their understanding.

Chetty's class has no prerequisites. I know Harvard students are so brilliant , but if they haven't had a rigorous introduction to statistics then they don't have the toolkit to make much sense out of the empirical world. They can be taught a few canned concepts so that they can digest Chetty's pre-cooked scenarios, but they're not really being educated to ask their own data questions, nor are they being given any theoretical framework for seeing logical patterns.

At W&M, we have "big ideas" classes for freshman that have no prerequisites. We could even offer a few in economics (if we had the resources). But our "big ideas" classes wouldn't count as major credit, and wouldn't prepare students for the next level of theory which is intermediate micro (where Chetty started his economics education). How will Chetty's students do in that subsequent higher-level class? Open question.

Many of the critiques of the "standard" class offered in this article seem to rely on straw man notions of standard principles courses. We already know the many ways that people are irrational, for instance. A standard principles class often mentions these limitations, but the meat of the class develops consumer theory and production theory since they are the basis of market analysis. We all agree that the "rational person" theory is incomplete. But it DOES have a lot of use. Presuming that people usually respond to incentives has given us a great set of predictive tools. There is a baby in the bathwater that Chetty is simply throwing out.

And we do teach about the many ways that markets fail to yield good outcomes (income inequality), and fail to yield efficient outcomes (externalities, public goods, etc.).

I could go on, but this is becoming an article ...