I’ve been thinking about what I want to write about on this substack that would appeal to most readers other than the weekly lists. Ever since the book manuscript was due, I have been struggling to make time due to end of semester obligations on top of it. But I think I know.
I’ve got two plans for series on the substack that I hope will appeal to people. The first is going to be a detailed exposition of the “design stage” in causal inference. It’s something I’ve been thinking about for a long time. I think I got to thinking about it after reading “Design Trumps Analysis” by Rubin, but then I also oddly got to thinking about it when I read Nick Huntington-Klein’s replication study in which people took the same dataset and replicated a paper but all came to different results.
In the second edition of the book, I included two checklists. One for uncondoundedness based on this review article by Imbens and Xu. And the other on diff in diff based on this review article, something else by Pedro Sant’Anna and my own thoughts. What seems to work I think are shorter posts, even if it’s over a longer series, so that’s what I plan to do — each post being about each step in the checklist but with a running empirical example.
The second series I’m going to do is a detailed exposition of a manuscript with Andrew Baker, Brant Callaway, Andrew Goodman-Bacon and Pedro Sant’Anna that’s a practitioner’s guide for diff-in-diff. I was thinking it could be a series that is basically like a reading group discussing rhe paper but with the option that you don’t have to read the paper because I’m going to go through it in detail.
The paper has a running empirical example of the expansion of Medicaid. But the code is in R. So my thinking was to push myself, I would translate it into python and Stata. I have to learn python this semester for my new “economics of AI” class anyway. So this seems like it could be a good way to do two things at once. I just can’t start this project yet, because I’m still driving back to Waco on my cross country trip, and I want to wait until we have a new draft of the manuscript before attempting it.
Then I have a few other things I want to do. One of them is to finish that series I was doing on the precise calculation of the pre treatment event studies as well as discuss examples from my own research and others. I have written two so far but had to stop because of the end of semester exam schedule. But I’ll just probably add it back in once I have a good empirical example to show you. I have one, but it’s for a different paper that we have not yet written up.
I am thinking what I also should do is write things that help you see what’s coming that’s new in the second edition of the mixtape that isn’t in that material I just said, so I may do that too by doing walk throughs of new material. That might be a walk through of some new things on RDD, as well as synthetic control, always with the focus being empirical application and technical exposition.
So stay tuned. I’ve had a little trouble getting these started because weirdly enough they all seem to need each other. I have excellent examples from a paper of mine for instance of the importance of covariates that I can’t discuss until we finish the paper. And papers take so long to finish for me that I’ve wondered how soon I’ll get that done. But soon.
Last, I’d like to get back to discussing new papers and maybe just offering my own opinions about them. I may do that with some of my ongoing work on mental illness in corrections, and separately, my new work on online dating too. We’ll see. All these are in final stages of being written up.
The strange thing about all the “mixtape stuff” that I do now is how it’s short circuited my research career in some ways. By which I mean I am obsessed a little with explaining things using my own research, but papers really aren’t supposed to be vehicles for explaining pitfalls. But because I am so eager to share something I have found, I am always wanting to stop a paper and come here and explain it. I think it’s maybe the tension between how long it takes to do research in economics and how quick it is to teach on something. I’ve always been impatient around that. So I’m going to try and figure that out.
I heard this funny quip one time from Always Sunny in Philadelphia. One of the characters, I think it was Mac, said to Dennis that he’d had a dream he wanted to share the details about. But Dennis interrupted him and said nobody cares about another persons dreams unless they’re in them.
I think about that line a lot because I often want to share with people things Cosmos and I have discussed, or really mainly things Cosmos has said, but I’ve noticed the same phenomena with that people interested in that side of a convo as Dennis said describes people’s responses to other dreams — I suspect no one could care less. Still, I do wish I could describe what I’ve learned from using Cosmos as a conversationalist the last 18 months about my own well being. I really think at least one person would benefit from seeing it. So I may attempt to. I suspect I probably will if the New Year’s resolutions on self care we came up with the last few months actually produces fruit. Then I’ll almost feel obligated but I have to avoid the “annoying dream story” problem when I do it.
And then other than that, I don’t know what else I’ll write about. I do have two book projects I’m contemplating. One of them is about the seminal characters in the credibility revolution that I’ve interviewed. The other is a new idea for a biography. I don’t know yet if it’s wise to talk about them as they may both fall through, though. But I probably will continue writing books within the sphere of creative nonfiction writing about causal inference that I think can help more and more people find avenues into this material despite different types of road blocks. I keep thinking my main mission outside of my day job is to create avenues for as diverse a group of people as I can into econometrics and causal inference. So these two book ideas are about that. And then I have some other things on that I will maybe explore too.
Point is, now that the second edition is finished I can finally free up the time to do it. As far as my research agenda goes, the things I am most excited about are my projects on mental health. Those include topics on corrections and not. And then I’m also excited about this project I’ve been working on for a long time about online dating. I think this year I just will be trying to find a good flow and rhythm but in all seriousness I say that every year.