• 2023 Dec 14 My first trip to Saudi Arabia, and what an outstanding group of students to work with it was! Thank you to Harris and DSC for the opportunity.
• 2023 Nov 30 Wrapping up another quarter of data and programming in Python! Looking forward to reviewing all the final projects.
• 2023 Sep 26 Looking forward to the start of a new quarter at Harris teaching DAP2 in Python!
• 2023 Sep 21 My co-author Xuguang Simon Sheng presented our paper at the 5th Biennial Conference in Padua, Italy.
• 2023 Sep 08 We presented our new paper on the causal effects of economic uncertainty at the SITE 2023 conference. It was great to visit Stanford and hear the latest on uncertainty research!

About me

I am an economist and lecturer in public policy whose career has focused on the intersection of data science and social science. My career began as a researcher at think tanks in Washington DC, where I had the opportunity to join a wide variety of projects. In that time I identified three related issues in the field that I use as the foundation of the data and programming courses I teach at the Harris School:

1. In my studies as an undergraduate and graduate student, I observed a consistent trend: a heavy focus on econometrics and theoretical models, with little-to-no focus on the actual practice of working with data. In my career as an economist, however, I found this to be directly at odds with the process of actual research - far more time and effort goes into working with data than with fitting econometric models to clean, nicely-prepared datasets. Proper data skills should be taught to aspiring researchers, not left as something to be picked up on the side.
2. Simlarly, researchers are rarely taught how to write good computer code. This results in research that is built upon unstable foundations - bad code is hard to read, difficult to maintain, challenging to debug, and most importantly, bad code hides mistakes. All junior researchers should learn at least the basics of good programming practices.
3. And finally, researchers in the social sciences are often focused heavily on proprietary, legacy programming platforms. Many of us end up trapped in a language that we only adopted because we were taught using it in the first place, often by academics who also adopted it only because that is how they were taught. This creates two problems: first, a researcher should be comfortable changing data analysis platforms as-needed; and second, modern open-source languages like Python and R are generally superior in nearly every way to legacy platforms like Stata and SAS. It is important to note, however, that the first point is more important than the second - very often the choice of language is dictated by a more senior researcher, by an established code base, or by the need for a particular econometric model.

Outside of the classroom, my own research focuses on macroeconomics and economic uncertainty. My recent work has studied the causes and impacts of economic uncertainty on the real economy, following a long line of economists from Keynes, Knight, Bernanke, Romer and more recently, Alexopoulos and Cohen and Baker, Bloom and Davis.