DRAFT: A Small, Intensive, Data-Sciencey Seminar in Long-Run Economic History
I have decided on my teaching next semester: two things that are half-courses (half of grad student intro to economic history, and the economic-history outside-speakers seminar), and this 25-student seminar to quantify the grand arc of economic history—and to make tools that will be usable by everyone, STEM majors or not. We’ll estimate, simulate, and argue about the quantitative shape of human economic history all the way from evolving foragers to the attention economy…
Well, I have decided what the rest of my teaching load is going to be this forthcoming spring…
For quite some time now, I have been worried as I see CP Snow’s “Two Cultures” problem—the division between those for whom math is a way of thinking, and those for whom it is an obstacle to thought—growing bigger and bigger, as it steadily has since he laid it out nearly a century ago. And for some time now, I have been thinking that getting people comfortable with what are now called the “Data Science” tools—statistics, quantitative optimization, decision-making under uncertainty, basic operations research, and so on—are going to be for our students the equivalent of learning for their day what learning to write a fine chancery hand was for students at the mediæval university. And for some time now, I have been thinking we commit educational malpractice every time we allow a student to leave Berkeley without acquiring basic competence and comfort with numeroliteracy.
Back when I was in the rotation teaching the “applied math” version of undergraduate, macro economics, I had what I at least regarded as considerable success in moving my problem sets off of paper and into Python Jupyter Notebooks, and in the process and the process making them superior educational materials as engines of estimation and counterfactual simulation. But whenever I tried to move these things to my economic history courses, I found that the exercises I produced were (a) so simple as to be terminally boring for 1/3 of the students while also being (b) so complex that they were completely impenetrable to 2/3 of the students. I could not find the sweet spot. For 2/3 of the students who show up in my economic history courses, numbers, algebra, calculus, statistics, and so on are not tools for thought but rather arcane ritual stumbling blocks, where sometimes they can learn to mouth the proper responses, but still find them large obstacles to understanding.
So I am going to make one more attempt to shift this—this time by hand-picking twenty-five students for a seminar, meeting four hours a week, and watching them very closely to see what they find straightforward and what they find impenetrable:
DRAFT: ECON 196: SEMINAR: Quantitative Long-Run Global Economic History
Mostly behind the paywall as it is only an early draft…