I just taught @pedrohcgs and Callaway's excellent paper on nonparametric diff-in-diff. I think when you do a simulation showing the bias of TWFE and then show their estimator which isn't, students freak out. Here's the lecture. Warning: it's long-is. https://www.dropbox.com/s/iv0dsj9vlk1pqad/cs.mp4?dl=0

Here are my lecture slides. I'm going to now do a thread. If you read this thread, and then you read @Andrew___Baker 's short piece (which I'll link to next), and then you read an imminent thread by @pedrohcgs , and then you read the paper, and then you do the simulation... 2/n

Then if one of those doesn't do it for you, one of the others will. But here's the slides. 3/n https://www.dropbox.com/s/r90za3fukp1y4nl/cs.pdf?dl=0

And here's @Andrew___Baker 's excellent writeup of a few of these new DD papers. 4/n https://andrewcbaker.netlify.app/2019/09/25/difference-in-differences-methodology/

But let me start at the top. Everyone on #EconTwitter has been hearing "oh the new DD papers this and the new DD papers that and something is biased, and something isn't good anymore and I'm busy, and we're in a recession," and so on. How do we think about these? 5/n

Well, here's how I think about these new DD papers. Some of them seem focused on parallel trend type stuff. For instance @jondr44 has a paper called "Pre-test with Caution: Event-study Estimates after testing for Parallel Trends." It's in the title! 6/n

But some of them have been focused on two things: 1) differential timing; 2) heterogeneity over time. Heterogeneity over time is different from heterogeneity cross-sectionally. Heterogeneity "cross sectionally" would be like this: My ATT is 10, but @jmorenocruz is 5. 7/n

The average ATT for us 7.5. OLS handles that fine. You can see me run it in the simulation, for instance -- TWFE handles group heterogeneity just fine. What TWFE can't handle is heterogeneity *over time* (with differential timing). TWFE is a sin in such situations. 8/n

And I illustrate that in the video, and you can see for yourself if you run this simulation here. @Andrew___Baker has the R code if you want to bug him for that. Just run it to see for yourself. Here's the dropbox link to the half completed simulation 9n https://www.dropbox.com/s/nikinluypt4tf5m/baker.do?dl=0

So, TWFE is biased. Run line 110 to see for yourself and compare that negative number to what it should be using line 96. It should at least be positive!! So, @pedrohcgs and Brant, what's this all about. That only took 10 slides (which is sort of like the lecture I gave). 10n

@pedrohcgs and Brant's estimator is used for all the reasons you don't want to use TWFE: 1) differential timing, 2) heterogeneity over time, and 3) conditional parallel trends. In other words, the *modal situation* for applied microeconomists doing program evaluation. 11/n

The notation in this paper is DENSE. So I'm going to give the white belt version. Basically, there's a billion treatment effects you can calculate if you have differential timing with heterogeneity over time - like literally a unique treatment effect per group/year. 12/n

Well, in @pedrohcgs and Callaway that has a name -- the group/time ATT. It's notation is ATT(g,t). What is that? Well let's say I got my PhD in 2007 and it's now 2020. My ATT would be ATT(2007, 2020). This is also what they mean by "the long difference". 13/n

One of the things I really like about this paper is it was a reward for spending time learning about the history of thought on inverse probability weighting, the different weights, and even an older Abadie paper on semi parametric diff-in-diff from 2006. Everything comes around

Here's the thing you have to keep in mind -- DD and TWFE are not the same thing. DD is a design; TWFE is an estimator. Just like IV and 2SLS aren't the same thing. IV is a design; 2SLS is an estimator. @pedrohcgs and Callaway's is a slightly nonparametric DD estimator. 14/n

At least, that's what I am calling it because really all you're doing in this ("all you're doing" says the guy who can't do anything) is weighting the "long difference" by normalized weights based on the propensity score. It harkens back to inverse probability weighting! 15/n

Briefly, here's situations where you may want it -- all of which have to do with dynamic heterogeneity. Here's an example of the group-time ATT. Here's the four assumptions needed for identification. And here's the funky estimator. Let's dig into this estimator. 16/n

Notice how the estimator is at its core just a weighting of the "long difference". You can take the ATT at any point in time, but you weight the actual outcomes based on whether it's a treatment or control differently. The expectations are with respect to time for a group. 17/n

Look closely at the C in the estimator, though. That's basically the never treated units. But what if you have staggered universal adoption (e.g., minimum wage)? Well, look at remark 1: just use the "not yet treated" as controls. What's the lesson here? 18/n

The lesson here is that when Adam sinned, it was because he compared treated units to *already treated units*, and with heterogeneity over time, that leads to bias. *So don't use them as controls*. And @pedrohcgs and Callaway don't. But econometrics isn't *that* easy. 19/n

So the paper is a lot to digest, but strangely, the simulation isn't. So I encourage you to study it, then look at these notes, then watch the video, then read the paper, and all between. But now the fun part. This method identifies a billion little ATTs. But you can aggregate!

After all, a policymaker probably is more interested in a more global ATT than some snapshot ATT at year t for some isolated group, right? Well like I said, they provide a couple of ways to do it. In R it doesn't automatically, but I'm having to manually do it and haven't yet.

So, here's the thing. You have to just run the simulation yourself to see. Do this: using line 96, make a spreadsheet of each individual ATT for a group/year like this. You should get 10 in 1986. And they do! 22/n https://www.dropbox.com/s/nikinluypt4tf5m/baker.do?dl=0

Okay, but maybe they got lucky. What about 1987? You should get 20. And they do! Yet TWFE? A giant poopy face negative value for the static parameter that looks like crap in the event study too. As the Hebrew say, "lo tov" (not good). 23/n

Still more that has to be done in this simulation: still have to aggregate the individual ATTs into "interesting parameters", which I'm working on. And need to bootstrap, which I also need to do. But putting those aside: wow. What a paper and a gift to us. Great job! 24/n