A month or so ago I posted a substack entry saying that I was going to be starting an online class to teach causal inference. It had been something I had been thinking about for a long time. I got the idea to do it because there has been demand to be perfectly frank. And the more time I spend trying to get better at “explaining econometrics” to myself, the more confident I felt that I could explain it to anyone.
But it wasn’t just that I felt I was getting better at it. I feel like one of the things I’m “supposed to” be doing is putting in my hours of service for the profession. But, I also feel like I’m supposed to be doing for anyone. There are a lot of people in the world who work with data now. And there are a lot of people who are interested in causality. I mean when you think about it — the world is so different than it was 50 years ago in terms of who can do what kinds of empirical research how and on what topics. So much data is available now. A ton of it you can just download off the Internet for Pete’s sake. And with programming languages like R and python, both of which are free, then the only barriers to entry to doing empirical work are 1) a computer and 2) human capital. I can help with the second. I’m certain of it.
I know it sounds cheesy, but I actually have a thought out mission. I almost hesitate to say it out loud because if I say aloud what the mission actually is, you’ll think I’m joking. But I’m not. The mission of this platform is to democratize causal inference through teaching and training so that no one who wants to know it gets left behind. It takes a lot of people to help one another and I don’t claim to be the smartest person, or the most accomplished, but I love teaching this stuff, and I love helping people, and those both are true statements. And so the mission of Mixtape Sessions is to democratize causal inference, and the platform will help me do that.
So I am doing this on the side, sort of like a consulting company, only instead of consulting, I’m democratizing causal inference. And it’ll cover the material from my book, Causal Inference: the Mixtape. For now, I just plan to teach two classes. One on causal inference, and one on advanced diff-in-diff and synthetic control. I’m doing the latter because I am enjoying learning more about those things. And I think I have enough for a real class. Sessions will be on Zoom and we will use the app (and desktop software) Discord to communicate by text during the lectures, as well as to keep getting to know one another later. Part of the purpose of the platform — call it a secondary goal if you will — is provide the platform for you to make friends and build community, perhaps even to start research projects with someone, or ideally, make a friend. After all, the psychological sense of community is practically guaranteed when groups of people meet together regularly with a common purpose. And maybe through the workshops, and with Discord, we can make that happen.
There will be coding labs every day of my Sessions because coding together is how you democratize causal inference! I mean, it really is. People learn in all kinds of ways, but a lot of people learn econometrics by looking at code, and watching it loop through a bunch of instructions as weird and as alien as that may sound to an econometrician. But it’s true — sometimes, we learn econometrics by writing code, and then the code becomes a ladder that we stand on to change the light bulbs in our brain. It doesn’t seem like that could possibly be true, but I’m convinced it is true. So we will be taking breaks, and coding together, going through examples, working with real data, doing our best to interpret what it all means.
There will still be “In Person” Workshops
Before I show you the website, let me just get this out of the way. I’ve been doing in-person workshops on causal inference for two years, and they are a lot of fun for me. I get to travel, I get to meet students, I get be in front, and just socialize and teach causal inference. I’m still going to do that. I started Sessions, though, because not everyone can do that. But I am going to do both as long as I have the time and keep my other commitments.
Website Navigation and Content Description
Now, with that out of the way, let me introduce you to my new website. It is frankly gorgeous. So check it out, and let me tell you a little bit about it.
Why the name "Mixtape Sessions" anyway? Well, because the book was called Mixtape, I love music, I’m using a concert registration website to register people. Felt like a good idea. So you'll notice the theme of music and stuff throughout the site.
I am anxious about the website, because I figure something will go wrong, but I hired a very smart person to make it, and so I think it's not going to break if you look at it.
If you click through that link, you'll find the two workshops plus a Coming Attractions description of an instrumental variables class taught by Brown University economics professor, Peter Hull. Peter is a phenomenal economist and applied econometrician, and actually is a former advisee of Josh Angrist’s at MIT. As part of my effort to democratize causal inference, I’m trying to get the best of the best to teach short classes. I'll skip his, though, because we've not finalized a date for him. But once we’ve settled on a date, I’ll do a substack interview with him and tell you more about him and the class, and maybe one of his papers.
But let's look at mine. The first is causal inference which will run mid to late January over three consecutive weekends (Fri-Sat, Fri-Sat, Fri). I wasn't really sure to be honest whether most people would prefer Friday to Sunday over just two consecutive weekends or this way, but I'd already said I was doing it this way, so I needed to stick to it. The link to register is here:
The second workshop is the advanced DiD/synth class. It's dates are mid February (Feb 18-19, and Feb 25). You can find that link here:
Now maybe you some questions. Well I have a FAQ page!
So how I am I going to achieve my mission to democratize causal inference? I'm an economist, so of course, and a local monopolist, and if I didn’t use price discrimination and bundling, they'd take my economist card away . Price discrimination usually works better with services with goods, and even better wth add-ons, so I've got both. And I'm doing bundling, because I suspect there are diminishing returns to workshops, which therefore means I should bundle. So, here's the pricing:
if you are enrolled as a student/predoc/postdoc or from a low to middle income country, please email me for a discount code. The price for you is $50 for one workshop and $90 for two workshops. All I require for “student” is you be enrolled. It can be high school, community college, night school, college, or whatever. It just has to be that you’re enrolled, because this doesn’t apply to a metaphorical student.
But maybe $50 is still too high, though. For this first time around, that's what I'm doing, but please note that I will be offering the causal inference workshop for free this summer through remote student exchange, which is a week long class I did this previous summer. It’s about 40 hours of material frankly, and I encourage you to attend if you can’t afford the Sessions. Please email Johannes Haushofer at that link to learn more about his platform.
For everyone else, it is $595 for one workshop and $995 for two plus transaction fees. So that is the bundling — you get to pay less for two workshops. You just have to request a discount code.
If you do pay for the full price, then there is an add-on which is office hours. You can have 30 min of office hours per paid workshop if you pay full price.
Anyway, I wanted to bring your attention to it. I'm looking forward to seeing everyone. If it ends up just being a handful of us, then it'll just be a handful of us! See you all soon and have a great weekend. And I really really really really really hope that it will work when you attempt to pay, if you attempt to pay. Feel free to forward this to anyone you know who may want to attend.