Last Two Mixtape Sessions Workshops for Spring 2023
Rocío Titiunik on RDD; Brigham Frandsen on estimating individual treatment effects with machine learning
Before I introduce the last two speakers of the spring 2023 season of Mixtape Sessions, I wanted to briefly summarize the year so far. All the workshops are up. The average econometrics workshops in 2022 was probably around 120 people, but this year they’re up to almost 220 people per workshop. Mark and Dan’s innovative hidden curriculum workshop on “doing applied research” averaged around 50 people last year and this year is also up to around 65. It’s exciting because we’ve been able to use Zoom, scale, Vimeo password protected videos for participants with steep cuts in price along demand using price discrimination to encourage the best and the brightest to come to the platform and teach. Our vision of being a bridge that takes econometrics off the top shelf, taught by the econometricians themselves, and delivering this material to people all over the world, from all walks of life, with needs and wants to learn these tools but who can’t come to MIT, Brown or Princeton otherwise seems to be working, and that’s super exciting. So let me introduce our last two speakers of the spring 2023 season: Rocío Titiunik (Princeton, Political Science Dept) and Brigham Frandsen (Brigham Young University, Economics Dept). These two workshops are covering new material so let me tell you a little about them.
Rocío’s RDD Workshop on May 17-19th
For those who have ever used RDD have heard of Rocío Titiunik. She along with several regular players, like Cattaneo, Calonico and Farrell, have been pushing not just the econometric theory of RDD forward, but also the practice with their wildly influential -rdrobust- package (and accompanying packages) available in Stata, R and now python. This is the workshop for those of you want to learn regression discontinuity from one of the pioneers. It’s very amazing we have a chance to host her on the platform, so I highly encourage you to take advantage of the opportunity. Remember — the lecture notes and code are all available as public goods, but the recordings are club goods. All recordings will be uploaded to Vimeo, password protected, available to you indefinitely for as long as Mixtape Sessions is a thing. Important information is below and remember if you are a resident of a low income country, middle income country, a student, a predoc, a postdoc, or in between jobs, you are eligible for a discount, so just email me at firstname.lastname@example.org and send me proof that you qualify and I’ll send you a discount. The link to the workshop on the other hand is here.
Brigham’s individual heterogeneity and machine learning workshop on May 15th
We define causal effects at the individual level, but we estimate them at the aggregate level in the form of such familiar parameters as the average treatment effect (ATE) or the average treated on the treats (ATT). But in reality what we really wish is that we could narrow down as close as possible to the individual’s treatment effect because if we know a person’s actual treatment effect, then we are in a better position to allocate resources to them that actually are relevant to them. Just because most people liked a movie doesn’t mean you will, so wouldn’t it be great if we could just figure out if you would like? Or surgery. Or therapies. You name it!
Brigham Frandsen is back with another machine learning and causal inference workshop that is precisely on this topic. It’s a short 3 hour workshop that we’re experimenting with to see about demand. The same discounted prices are the same ($1 for residents of low income countries, $50 for residents of middle income countries, students, predocs, postdocs and people in between jobs) but only $295 for everyone else. We’ll reassess this in the fall, but for now this is an experiment in pricing. Information about the workshop is below — check it out! Even if you can’t make the time, having access to the recordings may be something you don’t want to pass on.
I’ll just stop here and say thanks to everyone who has supported the platform. It’s been quite an exhilarating fun ride to be able to build this thing called Mixtape Sessions. My idea of “democratizing causal inference” seems to be working, and now it’s more than just me on here. I will continue to teach my “flagship” workshops, but that’s just a small share of the total of what we are offering. In the fall, we are hoping to have another great lineup of both old material and new material. I’ll be teaching a Causal Inference III workshop devoted only to synthetic control so that Causal Inference II is only about difference-in-differences and event studies. So that’ll be fun. And Peter Hull will be back doing something old and something new. And I’m on the hunt for another new speaker that I think will add value to people’s toolkits so stay tuned!
Remember, if you don’t want to receive the Mixtape Sessions newsletters, just go into your settings and unsubscribe. A lot of people have asked to be able to get regular updates, and until I have a better idea, the substack is the easiest way to do it. Have a great weekend and rest of the week! Peace!
Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.