About

Many regressions and instrumental variable (IV) specifications can be understood as leveraging the “design” of observed shocks for credibly estimating causal effects or structural parameters. This three-day workshop will build up this design-based toolkit and illustrate some of its advantages over alternative identification strategies. Questions we will seek to answer include:

• "What controls do I need to include to avoid omitted variables bias?"
• "Do I need to worry about ’negative weighting’ of heterogeneous effects?"
• "How should I be clustering my standard errors?"
• "What’s the payoff to considering nonlinear/’structural’ analyses?"

The course will include two programming exercises, where different techniques will be illustrated in real-world applications.

Schedule

This is a three-day (9 hour) intensive workshop, with 6-7 hours of lectures and two 30-minute coding demonstrations. The remaining time will be given to breaks. The coding demonstrations will feature me going through a real-world application, which will be handed out in advance if you’d like to attempt it on your own or in small groups beforehand.

Lecture 1: Selection-on-Observables

Lecture 2: Design vs. Outcome Models

Lecture 3: Design-Based IV

Live-Coding Application 1

Lecture 4: Negative Weights

Lecture 5: Clustering

Live-Coding Application 2

Lecture 6: Recentering

Lecture 7: Nonlinear Models

Readings

Here are selected readings that accompany the course.

Abdulkadiroglu, Angrist, Hull, and Pathak (2016) Angrist and Krueger (1991) Dale and Krueger (2001) Finkelstein (2007)