On the role of economic rationality in teaching undergradute economics

In graduate school I was trained quite intensively in the "neoclassical model," where individual actors in the economy optimize, their postulated behavior an expression of economic rationality.  It is natural to teach what you yourself were taught.  I have done so for much of my career, even as I changed the courses I taught and even after I had to learn additional economics in order to teach the new course; I nonetheless clung to the tenets of the neoclassical model in my teaching.  Of course, I was aware of alternative approaches.  There were institutionalists to be sure.  The one I knew most about was John Kenneth Galbraith.  I read The New Industrial State as pleasure reading while an undergraduate  (perhaps while in high school, I can't remember).  But as a style of doing economics, the institutionalist approach seemed too touchy-feely.  At Northwestern at the time the program was highly mathematical, at least among the group I became part of.  Then, too, early in the fall of my third year it was announced that Herbert Simon had won the Nobel prize in economics.  This popularized the notions of "bounded rationality" and "satisficing," though at the time I only learned those ideas the way a school child learns, "Columbus sailed the ocean blue..."  Definitions could be recited but there was no attempt to tie those ideas into other economics I thought I understood.

Since then a variety of attacks have been launched on the neoclassical model.  To me, those attacks have mainly been on the mark.   One has been given by Mark Blaug on why static efficiency concepts (Harberger deadweight loss triangles, Pareto Optimality) are essentially useless.  (I discuss some of these issues in my own framing in the beginning of this essay on assessing student learning.)   Instead one wants measures of economic growth, but that issue is hardly discussed in a microeconomics course.  Another critique, on the macroeconomics front, has as its basis the misapplication of the general equilibrium model to the macro-economy (there is no such thing as unemployment in a G.E. model) and the utter lack of forecasting of the recent financial crisis.  The third critique is the one that interests me here.  Let's call it the Kahneman-Tversky critique, named after the authors who have done the most to change the way economists think about decision making under uncertainty.   Kahneman won the Nobel prize in economics in 2002.  (Tversky had passed away by then).  Interestingly, they were psychologists by training.  They had a research interest in how people actually make decisions in the presence of statistical information.  Their work firmly shows that we must reject rationality as an explanation for human behavior in the presence of uncertainty.  Instead, the vast majority of us have something like an optical illusion when it comes to statistical information.  A lot of training (like what I got in graduate school) can encourage more rationality, the popular expression for this now is behaving like an Econ, but even the best trained economists stray from rationality now and then. 

I've recently read Kahneman's book Thinking Fast and Slow which provides an excellent overview of this research as well as providing a reasonable model for our actual behavior.  I wrote several posts that critiqued the early chapters of the book, where I struggled with the approach because I wanted to tie what Kahneman was saying to my views about teaching and learning.  I had no such difficulties with the latter two thirds of the book where the focus is statistical thinking.  There I was in agreement with what Kahneman had to say.  He asserts we have a variety of mental crutches that we use when making choices under uncertainty and in the process of using these crutches we make systematic mistakes.

One of those is to rely on "anchors" and make the choice by considering it relative to a reference point.  Here's a simple and humorous example from Daniel Ariely to illustrate.  If you and your buddies are headed out to a bar with the intent of meeting people of the opposite sex, you will be more successful if your friends are not as good looking as you are - you will benefit from the comparison.  So if you have very good looking friends, interact with them in other settings than dating and mating.  ;-)

I used to argue in my teaching that even if we non-experts weren't rational in our decision making, we could trust that professionals in the field are rational, which is one reason to value their judgment.  Unfortunately, recent research shows that doesn't always seem to be valid.  For example, several empirically minded economists have looked at the decision of coaches in the NFL to punt or go for it on fourth down.  The finding is that that there is systematic bias in favor of punting.  (Along these lines some time ago I did a theoretical analysis about shooting the 3 point shot in college basketball.  The argument I made is that underdogs are risk seeking while favorites are risk averse. So underdogs should go for the 3 pointer in the hope of getting lucky, and the should slow down the game, especially if they get a lead.)

There is an emerging field called Behavioral Economics that stems from the work of Kahneman and Tversky.  The biggest and probably most well known proponent of the field is Richard Thaler, an economist at the University of Chicago Business School.  Thaler along with Cass Sunstein, who has just stepped down as the head regulator in the Obama Administration to return to being a professor at Harvard Law School, are authors of the book, Nudge, which talks about various interventions that one might take to improve people's choices.  For example, it has been shown that in making choices many people accept the default and go with that, whatever it is.   One can increase the individual saving rate with 401Ks rather dramatically simply by making the default that the employee will contribute to the pension plan.  Not contributing then becomes "opt out" and most people don't exercise the option.

Seeing that a behavioral approach can have real economic consequence in areas where the assumption of rational decision making would predict no consequence whatsoever, there is a temptation to want to teach about behavioral economics.  Indeed I offered a course on that in spring 2011.  Unfortunately I stumbled into a couple of issues, one anticipated, the other not.  The anticipated one concerned senioritis.  Attendance dropped below 50% fairly early into the semester.  Many of the students didn't seem very earnest about what they were supposedly learning.  The unanticipated issue resulted from using the Nudge book as the core reading in the second half of the course.  The underlying philosophy to guide the nudges is called Libertarian Paternalism.  It is libertarian in the sense that choice by the individual is retained.  It is paternalistic in that the nudges are defined to achieve some social good defined by other than the individual.  In preparation for the students in encountering this philosophy, I had the students read this piece by Amy Gutmann on paternalism unmodified, with regard to the State's interest in the education of children, and then I had them read the Supreme Court case Wisconsin versus Yoder.  Among those students who were earnestly participating in the class, several reacted quite negatively to these pieces on the merits.  They wanted as little government interference in the ordinary activities of individuals as possible.  From there, they reacted negatively to much that is in Nudge.

So I opted not to teach that course again.  I didn't want to have to do battle with the students on their politics as we were learning the economics.  Instead, last spring I taught a course on the Economics of Organizations.  This semester I'm teaching on the same topic though in a different format.   The environment that organization members operate under is full of uncertainty.  Different members have different information and they may very well also have different agendas for what the organization should be trying to accomplish.   So it is in this setting where I pose the question - what should the role of economic rationality play in teaching such a course?  Below I will list several possible roles, but first I want to offer up this caveat.

I don't know how to teach this course taking a behavioral economics approach.  Therefore I will use behavioral economics only to critique the economic rationality approach, but not to consider good organizational design per se.

• When I was a campus leader for learning technology, I found that I would frequently build little economic models to provide some explanation for what was really going on with whatever issue I was then working on.  From that experience I conclude that building such economic models is a life skill that students who will be managers one day need to have.  So it is the model building per se and not the particular models we cover that is the real goal. 

• A good chunk of the course is about how to provide appropriate incentives for employees to put forth their best effort on behalf of the organization as well as to explain why that isn't to be assumed the outcome a priori.  The economic rationality approach to this issue, via the principal-agent model, provides a good first pass at these questions.   

• We use the economic rationality approach as a baseline from which to talk about reality, which doesn't conform to the theoretical predictions in a variety of ways. We introduce various complexities to make the results more realistic, but we don't model the complexities in a mathematical way, the way we do the basic model.  

• Students need to learn abstract thinking and model making simply to be able to read stories about economics in the newspaper or in news magazines.  So it is an important life skill for everyone, manager or not.  Economic rationality is at the heart of such economic models. 

• The ideas of economists have evolved on these issues and not all economists see the world in the same way.  Between the standard principal-agent model and the repeated prisoners' dilemma model, one can capture a good chunk of this varied thinking.  So one uses economic rationality in different frameworks as a way to provide labels/descriptors of the various approaches.  

• This one goes back to what I started with.  This is how I was trained, so this is how I will teach.  Supposedly I bring some expertise to the endeavor.  Part of that is my experience as an administrator.  Another part is my background as an economic theorist.

I will stick with these reasons for now as they seem sufficient for the time being.  Near the end of the course I want to reconsider them.  I should also add that the above says nothing on how much to emphasize the math.  I will be grappling with that issue too.  My intent now is to be more intensive on the math than I was last spring.  We'll see how that goes.