Teaching

Courses Taught

  • Economics of Family (third year undergraduate)
  • Mathematical Methods for Economics (second year undergraduate)
  • Quantitative Methods in Economics  (second year undergraduate)

My own course notes for the courses are available upon request.

Teaching Statement

Following my research interests, my teaching interests are in health economics, labor economics and applied econometrics. For undergraduate teaching, I was an instructor of a reading course on family economics; for graduate teaching, I held special lectures for MA and Ph.D. students in Labor Economics.

At the undergraduate level, their content are well-known to most instructors, needing no repeats in this statement; rather, let me begin by a discussion on my teaching philosophy:

Economics is a systematic way of thinking on how our society – a myriad of interconnected events – works; thus, I believe the key to successful economics teaching is to deliver the methodology well. They can use the methodology to analyze any situation they will meet; details are countless.

To be precise, establishing causality between events is the key in economics; otherwise it would not be a separate subject from statistics. Nonetheless, statistics is often taught as correlations in the junior year courses. While the techniques taught in these courses are certainly useful, the concept of causality is sometimes left out in the syllabus. Something should be done.

Thus, when I instruct readings to students in my upper-year undergraduate course, I explicitly emphasize the importance of causality from the start, by giving the following counterexample from Wikipedia: “As ice-cream sales increase, the rate of drowning deaths increases sharply. Therefore, ice cream consumption causes drowning.” This claim does not make sense, but the student may ask, “I know neither of them is causing the other, but so?”

The answer is that the policy-maker, based on this relationship, manipulates the cause and expects a result of saving people from drowning; he puts a ban on the sales of ice-cream. As economists, we make recommendations to the policy maker; for that purpose, we look for causation rather than correlations. We must exercise great care in this endeavour, because in the real world, bad policies can cause much more harm than losing the fun of having ice-creams in a hot summer.

The lecture continues: I organize empirical readings by a central theme of “solving a causality problem”. First, I explain that correlations are not causation, because correlations involves a selection effect. Then I tell students that there are three major methods to solve the problem: 1) Randomized experiments, 2) Control, and 3) Instrumental variable.

All these methods are introduced by linking them to specific papers, sometimes abstracted to ease the level of entry. To highlight causation, I often frame these papers as a comparison between a treatment and control group; in case of an observational study, “What does the author do to mimic an experiment?” is always the central question to be asked.

At the graduate level, content means more. Nonetheless, methods should still rule over specialization: this is particularly true for graduate Health and Labor Economics classes. Both fields admit widely different research approaches, ranging from purely reduced-form to full structural estimation. Under the same philosophy, my selection of papers would cover most of the popular approaches in the spectrum. Graduate students can always pursue a topic in depth later either by themselves under supervision, but they need a toolkit to start their own journey.