The Mixtape with Scott

In this week’s podcast episode, The Mixtape with Scott, I am interviewing Paul Oyer. Paul Oyer is a labor economist at Stanford University and author of several books, including "Everything I Needed to Know about Economics I learned from Online Dating", which is one of my favorite "popular general interest books explaining what economics is", as well as "An Economist Goes to the Game" which is about sports and economics. Links below for both. He's a fun, funny and interesting guy whose work in labor economics and personnel economics follows many of my own interests -- how firms hire, what they pay, discrimination, and platforms, just to name a few. I had a lot of fun interviewing Paul and hope you like it too. Thanks for your support, but I welcome even more of it by becoming a subscriber! And of course share this with you friends, family, loved ones, and especially those people you hate. Really rub it in their face with how good your taste in podcasts is.

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

In this week's episode of the Mixtape with Scott, I interviewed Mike Luca, the Lee J. Styslinger III Associate Professor of Business Administration at Harvard Business School. Mike studies a variety of topics of high relevance to the underlying economic organization of the online sector, such as the design of online platforms, applied causal inference and data science. More to the point, his research helps organizations consume data or all types and produce insights that help them become more informed and about managerial obstacles and solutions. He is also the coauthor of "The Power of Experiments: Decision-Making in a Data Driven World" with Max Bazerman.

This is part of my longer series on what I’m calling "economists in tech", which includes my interviews with:

* Susan Athey (first chief economist at Microsoft, John Bates Clark award winner, and much more)

* Michael Schwarz (current Microsoft chief economist),

* John List (former chief economists at Uber and Lyft),

* Chris Nosko (VP Head of Science and Analytics at Uber),

* Kyle Kretschman (Head of Economics at Spotify),

* Ronnie Kohavi (computer scientist with a long career in tech spanning decades, early promoter of A/B testing)

This series is, I hope, particularly relevant to those PhD economists and adjacent workers, like data scientists and machine learners, within tech, but also those outside of tech wanting to learn more about the long story of the demand for and supply of PhD economists in tech — which I consider to be the result of a very disruptive, particularly important, technological shift that increased the value of the work that PhD economists do and can do, causing a long march of PhD economists into industry (which is likely to continue growing for a while). Which is itself part of my long term project to collect interviews that when pieced together in the longrun help tell at least a small sliver of the oral history of the last 50 years of economists across many disciplines, many parts of the world, many types of work, many departments, and beneath many influential watershed people and movements that I consider to be particularly interesting (to me).

But in many ways, the oral history element, while meaningful to me, is also a convenient hook for me to listen to and help broadcast the personal stories of real people. Real people matter, and their stories matter, not because of how they connect to some larger thing, but as people in and of themselves. And as I say often at the start of each podcast, I am firmly in the camp of those who believe that we navigate our own historical lives through stories, and for many of us, that includes the personal stories of others. These stories, as I call them, function as models that help us understand ourselves as well as compasses as we try to plot out where we are in our journeys, and inform our decision making under uncertainty as we try to navigate our lives in a way that is consistent with our values and help us find our place in society.

So thanks for tuning in! I hope you find this interview with Mike illuminating.

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Opening music by Wes Cunningham.

What a pleasure it is this week to introduce my guest on the Mixtape with Scott, Dr. Pedro Sant’Anna. Had you asked me a few years ago the likelihood I’d make such a good new friend this late in life, I would not have guessed it, but from countless conversations on social media, and even more in DM on our Slack channel with two other close friends, Pedro Sant’Anna has become one of my favorite people in life. A constantly upbeat, friendly, energetic man, patient to a fault to explain every single detail of econometrics, and enjoying himself as does so, he is one of the best in the profession. He is as many of you know one of the half dozen important young econometricians that have made major contributions to the difference-in-differences research design. His productivity is intense so I can’t name them all, but the two I know best, almost by heart, are:

* Callaway and Sant’Anna (2021), “Difference-in-differences with multiple time periods” Journal of Econometrics

* Sant’Anna and Zhou (2020), “Doubly-robust difference-in-differences estimators”, Journal of Econometrics

The first one has over 2000 cites and it was only published a little over a year ago. He also has an Econometrica with Jon Roth on issues related to functional form and parallel trends in diff-in-diff and a review article (also with Jon Roth, but also with John Poe and Alyssa Bilinski) for anyone who wants to in one stop learn everything you need to know about diff-in-diff.

In this mixtape episode, though, we learn more than just his papers. Pedro shares his story with me. I hope you like it and I hope as always you come to value both his story, but also the contemporary ongoing series I’m doing on the many stories of economists. Because to quote, Sue Johnson:

“We use stories to make sense of our lives. And we use stories as models to guide us in the future. We shape stories, and then stories shape us.”

Consider subscribing, sharing and possibly even supporting the substack as I continue to try and accumulate enough stories of living economists that we have those stories to help us make sense of our lives, but also an oral history of the profession. Thank you again for your support! Youtube below!

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

This week’s episode of the Mixtape with Scott is a little out of order. Season two’s episodes are going to be a little out of order, based on what feels like the best next episode to present at that time. So I decided after doing my interview with University of Arizona professor of economics, Hide Ichimura, that I wanted to release it because I had such a delightful time talking with him. Dr. Ichimura is an econometrician whose work I’ve gotten to know more recently because it’s been experiencing a little bit of a revival (though it’s always remained very popular over the years) within the difference-in-differences literature thanks, in part, to the Sant’anna and Zhao (2020) Journal of Econometrics on robust diff-in-diff, Callaway and Sant’anna (2021) paper on differential timing, and in many ways, other papers that conduct certain kinds of imputations and estimations that are similar in spirit like Borusyak, et al’s (2022) robust efficient imputation estimator, and even Abadie and Imbens (2011) selection bias adjustment method if you squint your eyes.

I had a wonderful experience talking with Dr. Ichimura today. This is sort of part of my broader interest, as I say in the intro, in interviewing econometricians who were active in the 1990s working on topics in causal inference, and to that end, I had in mine two Restuds by Dr. Ichimura with Heckman and Todd (1997) and a 1998 one in Restud also by Heckman and Todd (and the identical title!!), both on program evaluation. But I also just in general wanted to hear his story, and I’m so glad I did and that he would share it.

At the end of the episode, I asked him to share with me a paper that, maybe isn’t his favorite, but that has always stuck in his mind. He shared with me Stephen Nickell’s 1979 article in Econometrica entitled “Estimating the Probability of Leaving Unemployment”.

As always, opening introduction music is by Wes Cunningham (no relation).

And don’t forget to subscribe, share and maybe even support this! This podcast is subsidized by your donations and my workshops!

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

In this week’s interview on the Mixtape with Scott, I had the pleasure of interviewing Clair Brown, a labor economist at the University of California - Berkeley. Dr. Brown’s career has spanned several topics like discrimination, industrial economics, and climate. Dr. Brown’s late career has made several turns into environmental economics, particularly climate, but also a re-envisioning of the field of economics with her book Buddhist Economics. Dr. Brown’s work has always focused on issues around welfare that are often massaged out of her models, like meaning, community and fairness in labor markets. I thoroughly enjoyed our time together, and I hope you find it interesting too. Please remember to subscribe and share!

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Time for a new podcast episode! In this week’s episode, I have the pleasure of sitting down with Dr. Beatrice Cherrier, a widely recognized expert in the history of economic thought, and one of my personal favorite economists around. Beatrice is an associate professor of economics at the Center for Research in Economics and Statistics in France. She is a font of wisdom and insight about more parts of the field of economics than is found in most people who specialize in even one of those areas. But she is also a very endearing person and a pleasure to talk to, and I hope you enjoy this interview. Please remember to share and subscribe!

In this week’s episode of The Mixtape with Scott, I introduce you to the one and only, Dan Hamermesh. Dan is professor emeritus now at the University of Texas where he spent the last part of his career. He’s part of the story of Princeton that I’ve been interested in telling, as he was there from 1969 to 1973 before heading to Michigan State where he was until 1993. He then went down the street from me to Austin at University of Texas until 2014.

Dan did his undergraduate in economics from the University of Chicago where he got to see some of the big labor economists of that era up close, like H. Gregg Lewis, who was also Gary Becker’s adviser. It was really interesting to hear Dan’s impressions of Lewis; for the first time, I wondered if all of the ethos and worldview we associate with Becker (“thinking like an economist” where economics was “everywhere”) might in fact be attributable to Lewis, even though Lewis was in many ways a traditional labor economist, and you could say Becker was a bit more nontraditional since he took economics in an imperialistic fashion into areas like the family and discrimination, among others. Dan provides some interesting perspective.

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Dan has been a premier labor economist his entire life, and is an important part of the story of labor and empirical micro of the last quarter of the 20th century and the first quarter of the 21st century. An unusually productive economist who blended well the price theory of Chicago with contemporary empiricism that we closely associate with labor economics in general. A very funny and energetic man, he is also one of those you tend to say “they’ve forgotten more economics than I’ll ever know”. I thoroughly enjoyed talking with Dan, but I always do. So please welcome MC Hammer himself, Dan Hamermesh.

Opening music credits: Wes Cunningham

Season two of the Mixtape with Scott is up and boy do I have a dynamite first guest. None other than the man himself, Dr. Jeffrey Wooldridge! Jeff, as I say in the opening, is the author of two phenomenally popular books (here, here and here’s the solutions) in econometrics that has raised an entire generation of economists. We have a great conversation about his life and career and I hope you enjoy it!

Expect new episodes every Tuesday morning. Thanks to my good friend, Wes Cunningham (no relation), for the amazing opener music he made for me — it’s perfect. And thanks to Arslan Yaqoob who set the music to the awesome montage of last season’s guests and has been producing them for me this entire time. Check out the YouTube video below if you enjoy watching more than listening.

My guest this week on the podcast is Phillip Levine, the Katharine Coman and A. Barton Hepburn Professor of Economics at Wellesley College in Massachusetts. I’ve only personally met Phil once — at a conference on the family many years ago and just briefly. But I have been a huge admirer of him for many reasons for a long time, ever since graduate school, and I wanted to interview him for a lot of reasons. First, he attended Princeton in the 1980s at that heady time when Orley, Card, Krueger, Angrist and so many others were there. The birth place of the credibility revolution is arguably the Princeton’s Industrial Relations Section where a shift in empirical labor took place that eventually ran through the entire profession and placed it on a new equilibrium. Phil was there, colleagues and students with those people, and himself part of that “first generation” of labor economists who thought that way and did work that way and I wanted to hear about his life and how it passed through, like a river bending and turning, the Firestone library and beyond.

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

But I also have a special interest in Phil. I actually first learned difference-in-differences from a book that Phil wrote on abortion policy entitled Sex and Consequences (Princeton University Press). I graduated from the University of Georgia in 2007, but the job market had started in 2006, and around the spring when I had accepted my job at Baylor, I was finishing my dissertation. I had one chapter left and it was going to be an extension of Donohue and Levitt’s abortion-crime hypothesis to the study of gonorrhea. My reasoning was that if abortion legalization had so dramatically changed a cohort by selecting on individuals who would have grown up to commit crimes, then it should show up in other areas too. My argument was relatively straightforward and I’ll just quote it here from the article I later published with Chris Cornwell in the 2012 American Law and Economics Review.

“The characteristics of the marginal (unborn) child could explain risky sexual behavior that leads to disease transmission. For example, Gruber et al. (1999) show that the child who would have been born had abortion remained outlawed was 60% more likely to live in a single-parent household. Being raised by a single parent is a strong predictor of earlier sexual activity and unprotected sex, evidenced by the higher rates of teenage pregnancy among the poor.”

It’s funny the order in which things go. I think I somewhat understood what I was doing because I already had planned to do my study before reading Phil’s book. I was going to use the early repeal of abortion in 1969/1970 in five states (California and New York being two of them) followed by the 1973 Roe v. Wade as this staggered natural experiment to see whether abortion legalization led to a drop in gonorrhea a generation later. I had adapted a graph I’d seen by Bill Evans to illustrate how the staggering of the roll out would lead a visual “wave” of declines in gonorrhea in the repeal stages among an emerging cohort that would last briefly until the Roe cohort entered. Visually, I believed you should see a drop in gonorrhea for 15yo starting in 1986 that would get deeper until 1988, flatten, and then disappear completely by 1992.

The design for this idea came from a paper I just linked to above — by Phil Levine. It was entitled “Abortion Legalization and Child Living Circumstances: Who is the “Marginal Child”?” coauthored with Doug Staiger and Jon Gruber, published in the 1999 QJE. It came out two years before Donohue and Levitt’s 2001 QJE on abortion and crime and arguably really set the stage for that paper. The two papers are very different — Phil, Staiger and Gruber are looking at who was aborted using instrumental variables with the five “repeal states” as the instrument. The abstract is worth reading:

“Cohorts born after legalized abortion experienced a significant reduction in a number of adverse outcomes. We find that the marginal child would have been 40–60 percent more likely to live in a single-parent family, to live in poverty, to receive welfare, and to die as an infant.”

They used, in other words, instrumental variables whereas Donohue and Levitt used a lagged abortion ratio measure, if I recall correctly. Phil’s paper really struck me as the more credible design at that time because the staggering of legalization gave such precise predictions — something about the timing, something about the location. It just really haunted me for a long time.

Well, while I was preparing for that project, reading the literature on the economics of abortion, continuing my ongoing interest in the economics of sexual behavior, Phil has a chapter where he sets up for the reader a table explaining something called “difference-in-differences”. While econometrics was my field, I couldn’t recall hearing what that was, because it wasn’t really best I could tell an estimator. Rather it was what we now call a research design. I don’t have the book here at the house, but the table made a huge impression on me because if you just walk through the before and after differencing, even without potential outcomes, you can see with your own eyes exactly why difference-in-differences identifies a causal effect. I have a version of the table in my book, which I’ll produce below.

Once I saw that, it was easy to understand triple differences — a design that many people find very confusing if they only think of it in terms of regression equations. Almost immediately after I understood Phil’s DiD table, I adapted it to my repeal versus Roe context and imagined “Well, what if there were other things happening in these repeal states later? Is there an untreated group I could imagine was affected by those unseen things but which wasn’t treated?” And I thought “Let me use a slightly older group of individuals in the same states as the within-state controls”. That approach — the triple difference — can be seen below in a table I mocked up for a lecture in which I teach triple difference using Guber’s 1994 paper that introduced the design for the first time.

And so I wrote the chapter, and of all my chapters, it was the only one I ever published.

Thank you for reading Scott's Substack. This post is public so feel free to share it.

Where am I going with this? I guess what I’m saying is that as luck would have it, I made a monumental jump in my understanding of this “way of thinking” about doing empirical work from a single table in a short little book on abortion policy by Phil Levine. That one table so completely captivated my mind that ever since I have only wanted to learn more about causal inference in fact. As odd as it may sound, something about difference-in-differences really unlocked for me what the whole empirical enterprise was about. As Imbens said, there is something about potential outcomes that just makes crystal clear what we mean by causality, and many of the research designs that have over time been fully mapped onto potential outcomes — difference-in-differences being one — extend that clarity for a lot of us. Phil’s work has consistently been part of the broader education of labor economists about what the Princeton tradition left us — make clear where the variation in the data is coming from, make clear who is and is not functioning as the counterfactual, “clean identification”, carefully collected data, on questions that matter.

Phil has had a very interesting life; I caught only a peek of it from this interview. He opened up and shared about being a young man growing up middle class where family experiences during difficult economic times appeared to cause inside him an interest in labor. He gravitated towards law but a chance research class in college placed him on a new trajectory. His professors encouraged him to go to Pr…

Chris Nosko is a PhD economist. He did his PhD in economics at Harvard in the mid 2010s before going to Chicago Booth take a job as an assistant professor. But for a year prior to taking that job, between Harvard and Chicago, he did a postdoc fellowship at eBay where he, Thomas Blake and Steve Tadelis met and worked together on a project involved a serendipitous event at the company in which eBay quit paying for branded key words (e.g., “eBay Volvo decals”, “eBay typewriters”) on some but not all search engine auctions. They asked for the data on traffic to the site before and after eBay quit paying for branded keywords for all search engines (both those they kept paying and those they didn’t), ran a simple event study diff-in-diff and found evidence that search engine marketing at eBay was perhaps not causing increased traffic to the site. They convinced management to field a large RCT which confirmed their diff-in-diff results, and that study was published in Econometrica. Not a shabby way to start a career as an economist.

For many of us, a PhD in economics from Harvard, a successful partnership with eBay resulting in a study destined for a Top 5 and a tenure track job at Chicago Booth meant staying at Booth and having a career as an academic. No one outrightly says that the only meaningful life you can have as an economist is to be an academic, as it’s vulgar, opinionated and obviously false to talk that way about how someone else should live their life, but the norms are pretty powerful nonetheless. Well, starting around the time that Chris got his job at Booth, tech began experiencing a surge in hiring of PhD economists, largely driven by Amazon’s nearly insatiable appetite for them. Talking with people at Amazon, I have learned that behind this push was Pat Bajari, and behind Pat Bajari was Jeff Bezos who had long believed economics, and economists more specifically, had unique value. As Susan Athey said to me, though, in an interview earlier, Bajari though had to do pull a rabbit out of a hat. Whereas the first wave of economists to tech — people like Hal Varian, Susan Athey, Preston McAfee — had largely been micro theorists helping craft the foundations of a business model through auctions and advertising that would support search engines, arguably the core arteries of the internet itself — Bajari would have the task of bringing in young people, fresh out of grad school, and in Athey’s words, make them productive. And one of the people Bajari would ultimately tap do that was Chris Nosko, an assistant professor at Chicago Booth and someone trained in structural industrial organization, one of the economics’ more interesting experiments of fusing deep microeconomic theory with econometric estimation.

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Nosko was a Ariel Pakes student at Harvard and was well versed in so many different parts of economics and modern technology that it almost seems predestined that he would ultimately leave Chicago Booth permanently and go to Amazon when Bajari finally convinced him to, but that’s all selection on the dependent variable reasoning. When we look back in time at decisions we made, our mind tends to forget that there was a moment when we could’ve gone left instead of right. The same with Chris — there was a decision that had to be made to leave a career as an academic. The decision materialized into what it materialized, but to pretend it was easy, or that it didn’t have risk, or that Chris didn’t try to manage that risk in some ways is really unfair to our earlier selves or even our future selves who are in situations facing, not probabilistic risk but more like Knightian uncertainty in which no one truly has a clue what possibly could happen.

But Chris did leave. Sort of. He took “a leave of absence” from Booth in 2015 and took a job at Amazon, then permanently left Booth in 2016. He spent four years at Amazon before leaving for Uber, one of the more impressive firms to ever exist for creating an actual open marketplace solving two sided matching problems through algorithms and prices. Algorithms, prices and rules — three ways, no doubt there are others, in which modern economies coordinate productive activity. Is it really so surprising that economics might be valued by tech firms given the complex coordination they try to solve using all three?

Thank you for reading Scott's Substack. This post is public so feel free to share it.

Chris has been at Uber for four years. He is now Vice President and Head of Science and Analytics for Uber Product there.  Within tech, economists sort into tons of different jobs with titles that to an academic don’t make a ton of sense — just like so much of what academics’ lives takes place within administrative units that make little sense to anyone else. If Chris isn’t the chief economist, though, at Uber, I figure he’s probably up there. And he’s my guest this week on The Mixtape with Scott as part of my longer, unfolding series I call “Economists in tech”. Our conversation covered a lot of ground. We talked about growing up in rural Oregon, falling into programming early on and working a few years between high school and college during the early wave tech boom of the late 1990s and early 2000s as a programmer. It wasn’t exactly what he would do later, as that was more web design and less machine learning and statistics, but the aptitude of programming is very portable and his deep knowledge of tech sectors was anyway established or at least re-invested in while there. We talked about his love for his liberal arts education at the University of Chicago where he did his undergraduate degree, and his broad navigation of economics as a field and a career. 

All in all, it was a fun opportunity to talk to Chris, to learn more about his own path, about the world out there outside of academia, what economists do in tech, and how all of these things fit together for both economics but maybe more importantly just for Chris himself. I think a lot of people are going to find Chris’s story very interesting and personally intriguing as they may see him in themselves. You can read some of Chris’s work here. Thanks again for tuning in! I hope you enjoy this week’s interview as much as I did! If you are enjoying these, please consider supporting me by sharing the podcast and/or becoming a paying subscriber!

Scott's Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.