Estimating Difference in Differences (The Effect, Videos on Causality, Ep 53)

Please visit www.theeffectb... to read The Effect online for free, or find links to purchase a physical copy or ebook.
The Effect is a book about research design and causal inference. How can we use data to learn about the world? How can we answer questions about whether X causes Y even if we can't run a randomized experiment? The book covers these things and plenty more. These videos are meant to accompany the book, although they can also be viewed on their own.
This video relates to material found in Chapter 18 of the book.
A version of this video without background music can be found here: • Estimating Difference ...
Alright, so we've got the concept behind difference-in-differences down. But how can we actually estimate an effect using this design? This video goes through a few different ways.

[@juliuskasberger]

Hey! I really enjoy your videos, keep going! As a political scientists, however, I lack some understanding in econometrics and just discover the subject area. Could you say something about using DiD with a non-linear model? In my case, I use survey data and my dependent variable is on an ordinal scale. Would that at all be possible to implement given the case that the parallel trends assumption is supported?

[@NickHuntingtonKlein]

Nonlinear DID can be a bit of a headache, largely because what "parallel trends" even *means* can vary quite a bit from model to model, and it often represents an assumption that's more technical and less intuitive in nature than the linear version. Basically you're just sort of stuck hoping someone has hashed out the problem for your specific kind of model before. It does look like someone has done this for ordinal dependent variables, though, so you're in luck! I haven't read the paper but here it is. arxiv.org/abs/2009.13404