S4E15: Dmitry Arkhangelsky, Econometrics and Machine Learning, CEMFI
Feb 11, 2025
auto_awesome
Dmitry Arkhangelsky, an Associate Professor at CEMFI in Madrid, shares his remarkable journey from Moscow to Madrid, intertwining anecdotes of cycling adventures and a deep-seated love for literature. He discusses his early explorations in economics, revealing the chaotic state of the field at Moscow State University. Dmitry elaborates on his innovative work in econometrics, particularly the "Synthetic Difference-in-Differences" method, showcasing the collaborative dynamics of academic research and its practical implications in policy analysis.
Dmitry Arkhangelsky's unique journey into economics was influenced by his upbringing in Moscow and a strong academic environment from his mother.
His experiences at the New Economic School and Stanford GSB cultivated a rigorous understanding of econometrics and innovative research collaboration.
Dmitry's groundbreaking work on synthetic difference-in-difference methodologies enhances econometric analysis by merging flexible frameworks with traditional methods for more robust policy evaluations.
Deep dives
Dimitri Arkhangelsky's Journey into Economics
Dimitri Arkhangelsky grew up in Russia and had a unique journey into the field of economics, which initially was not his primary focus. His upbringing in Moscow and later in its suburbs exposed him to various influences, including a strong academic environment from his mother, a research fellow in economics. His love for literature paved the way for an interest in subjects that combined social aspects and math, leading him to explore economics during his undergraduate studies. After realizing the significant opportunities within the field, he pursued a master’s degree at the New Economic School, which further fueled his passion for economic research.
The Impact of Higher Education on Dimitri's Career
At the New Economic School, Dimitri found a rigorous academic environment that enhanced his understanding and skills in economics, particularly in microeconomics. It was here that he was exposed to a graduate-level understanding of econometrics, which sparked a newfound interest in the subject. His experiences studying under a demanding and encouraging tutor also solidified his determination to further his education. This commitment led him to apply for a PhD program, seeking a more profound understanding of economics and potential career opportunities in academia.
Stanford as a Catalyst for Research Ideas
Dimitri's transition to Stanford GSB marked a pivotal moment in his academic journey, where his involvement in rigorous coursework and vibrant academic discussions flourished. During his studies at Stanford, he engaged deeply with influential scholars in econometrics, including Hido Embans, which shaped his research perspective. The academic atmosphere at Stanford allowed him to explore innovative ideas, eventually leading to collaborative work that merged econometrics with causal inference. This collaboration culminated in the development of a groundbreaking paper on synthetic difference-in-difference methodologies with his peers.
The Significance of the Synthetic Difference-in-Difference Paper
The synthetic difference-in-difference paper, co-authored by Dimitri, marked a significant advancement in econometric techniques by combining synthetic control methods with traditional difference-in-difference analysis. This innovative approach allowed researchers to create a more flexible framework for analyzing treatment effects in observational data. By enhancing the statistical guarantees of the original synthetic control method, their work made it accessible for a wider range of applications in policy evaluation. Within a short timeframe, this paper gained traction in the research community, illustrating the demand for robust tools in econometric analysis.
Future Aspirations and the Integration of Structural and Reduced-Form Methods
Looking ahead, Dimitri expressed excitement about exploring the intersection of structural economic analysis and reduced-form methods, seeking to enhance empirical research in economics. He envisions future projects that deepen the understanding of causal relationships while applying empirical techniques to address real-world economic questions. By bridging these two approaches, he believes researchers can contribute valuable insights to the field and address complex economic problems effectively. In particular, Dimitri aims to continue pushing the boundaries of econometric research, driven by his intrinsic motivation and interdisciplinary interests.
Welcome to episode 15 of season 4’s The Mixtape with Scott! I am of, you guessed it, Scott. And this is my podcast which is a podcast where I interview economists and ask them about their personal story. If you were dying to know the games that economists played when they were kids, or what books they read in high school, then man are you ever in luck because that’s literally what we talk about on here!
This week’s guest is Dmitry Arkhangelsky, an associate professor at CEMFI in Madrid, Spain. Dmitry is known to many people because of his 2021 American Economic Review article with an Avengers like team of econometricians and statisticians — Susan Athey, Guido Imbens, David Hirshberg and Stefan Wager — entitled “Synthetic Difference-in-Differences”. Synth diff-in-diff is a well known contribution to the pantheon of new causal panel methods and is quite versatile and flexible. Dmitry is currently on leave from CEMFI and had just arrived to Harvard for a research sabbatical when we did this interview.
Dmitry is an econometrician and machine learning, and as he’s connected to this new diff-in-diff and synth literature that has been exploding and evolving over the last few years, his work on those topics are well known. But I think as he’s not on social media, he’s not someone people may know as much about. So I hope you that this is an interesting interview for those of you wanting to learn about his life growing up in the bustling city of Moscow, Russia. It’s a bit of a rags to riches story in some way as unlike many Russian economists who are dialed into the best schools as a young person, where they are exposed to intensive training in mathematics early on, Dmitry’s journey was different, and I don’t want to spoil it. But I think it’s one that many of us may identify with.
Thank you again for all your support of the podcast. It’s a labor of love to get to have a chance to just pause, look at another person, and listen. I continue to believe that it’s in the moments when we can look at a person that we know ourselves. And so I enjoy doing it and appreciate your support and hope it is the same for you on some level. And thank you to Dmitry for being generous with his time to share a little about his life. Consider becoming a paying subscriber where you get full access to all kinds of weird posts ranging from econometrics, practical opinions about work, discussion of my classes, and taking care of my ailing dad, as well as a fairly regular reflection on the economic implications of new technologies.
Scott's Mixtape Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.