New Tip from the RCS Stats Team

March 13, 2019

RCS Statistician Andrew Marder has compiled some resources for researchers working with observational study data who are interested in identifying causal effects:

Paul Goldsmith-Pinkham has some very nice slides about using difference-in-differences models to identify causal effects:

One paper discussed in the slides, Difference-in-Differences with Variation in Treatment Timing by Andrew Goodman-Bacon (, is particularly interesting. Goodman-Bacon (2018) shows how the causal effect estimated by a model with firm and time fixed effects can be seen as a weighted average of all possible 2x2 difference-in-differences models. For a quick introduction to Goodman-Bacon (2018), see his Twitter thread here!

See also: Stats Tips