I am an Economics and Data Science researcher primarily working in combining ML and Econometrics for policy evaluation and platform economics. Currently, I am a Postdoc at Stanford GSB and the Stanford Causal Science Center collaborating with Airbnb. Starting in September 2026, I will join NYU as an Assistant Professor of Economics and Data Science. In 2025, I received a PhD in Economics and Statistics from MIT. During the PhD I was a Meta Research PhD Fellow and a La Caixa Fellow. Previously, I obtained a BSc. in Econometrics and Mathematical Economics from the LSE (First Class Honours) and worked as a research professional at the University of Chicago Booth School of Business. I also have experience designing causal inference methods at Google, Quantco, and as a consultant for Ivalua (Catalan Government).

NYU Data Science PhD: For students interested in a Data Science PhD at the intersection of Economics, Econometrics and ML, apply to the NYU CDS PhD and mention me in the application. Happy to answer questions via email too.

Publications

Stretching the Net: Multidimensional Regularization (2023)
Econometric Theory
[Published Version]

Synthetic Controls in Action , with Alberto Abadie
Econometric Society Monographs (forthcoming)
[arXiv]

Synthetic IV Estimation in Panels, with Ahmet Gulek
Co-winner of the Best Student Paper Award of the International Association for Applied Econometrics (Thessaloniki, 2024)
Special Mention (runner-up) for the Best Job Market Paper Award (EEA and UniCredit, 2024)
Presentations: ASSA (2025), EEA-ESEM (2024), IAAE (by co-author, 2024), NASM Econometric Society (2024), BSE Summer Forum Micro-econometrics (2024), EWM Econometric Society (2023), American Causal Inference Conference (oral, 2023)

Abstract
We propose a Synthetic Instrumental Variables (SIV) estimator for panel data that combines the strengths of instrumental variables and synthetic controls to address unmeasured confounding. We derive conditions under which SIV is consistent and asymptotically normal, even when the standard IV estimator is not. Motivated by the finite sample properties of our estimator, we introduce an ensemble estimator that simultaneously addresses multiple sources of bias and provide a permutation-based inference procedure. We demonstrate the effectiveness of our methods through a calibrated simulation exercise, two shift-share empirical applications, and an application in digital economics that includes both observational data and data from a randomized control trial. In our primary empirical application, we examine the impact of the Syrian refugee crisis on Turkish labor markets. Here, the SIV estimator reveals significant effects that the standard IV does not capture. Similarly, in our digital economics application, the SIV estimator successfully recovers the experimental estimates, whereas the standard IV does not.
[Draft]

Predictor Selection for Synthetic Controls
Revise and resubmit at the Journal of Econometrics
Presented at: BSE ML and Energy Workshop (2023), Google (2022), Meta (2022), Two Sigma PhD Symposium (2022)

Abstract
Synthetic control methods often rely on matching pre-treatment characteristics (called predictors) of the treated unit. The choice of predictors and how they are weighted plays a key role in the performance and interpretability of synthetic control estimators. This paper proposes the use of a sparse synthetic control procedure that penalizes the number of predictors used in generating the counterfactual to select the most important predictors. We derive, in a linear factor model framework, a new model selection consistency result and show that the penalized procedure has a faster mean squared error convergence rate. Through a simulation study, we then show that the sparse synthetic control achieves lower bias and has better post-treatment performance than the un-penalized synthetic control. Finally, we apply the method to revisit the study of the passage of Proposition 99 in California in an augmented setting with a large number of predictors available.
[arXiv] [slides] (with a deforestation application)

Bayesian and Frequentist Inference for Synthetic Controls, with Ignacio Martinez
Presented at: EEA-ESEM (2023), DataX Workshop on Synthetic Control Methods (poster, 2022), American Causal Inference Conference (poster, 2022), RAND Causal Inference Symposium (2022)
Abstract
The synthetic control method has become a widely popular tool to estimate causal effects with observational data. Despite this, inference for synthetic control methods remains challenging. Often, inferential results rely on linear factor model data generating processes. In this paper, we characterize the conditions on the factor model primitives (the factor loadings) for which the statistical risk minimizers are synthetic controls (in the simplex). Then, we propose a Bayesian alternative to the synthetic control method that preserves the main features of the standard method and provides a new way of doing valid inference. We explore a Bernstein-von Mises style result to link our Bayesian inference to the frequentist inference. For linear factor model frameworks we show that a maximum likelihood estimator (MLE) of the synthetic control weights can consistently estimate the predictive function of the potential outcomes for the treated unit and that our Bayes estimator is asymptotically close to the MLE in the total variation sense. Through simulations, we show that there is convergence between the Bayes and frequentist approach even in sparse settings. Finally, we apply the method to re-visit the study of the economic costs of the German re-unification and the Catalan secession movement. The Bayesian synthetic control method is available in the bsynth R-package.
[arXiv] bsynth

Bagged Polynomial Regression and Neural Networks, with Sylvia Klosin

Abstract
Series and polynomial regression are able to approximate the same function classes as neural networks. However, these methods are rarely used in practice, although they offer more interpretability than neural networks. In this paper, we show that a potential reason for this is the slow convergence rate of polynomial regression estimators and propose the use of bagged polynomial regression (BPR) as an attractive alternative to neural networks. Theoretically, we derive new finite sample and asymptotic L2 convergence rates for series estimators. We show that the rates can be improved in smooth settings by splitting the feature space and generating polynomial features separately for each partition. Empirically, we show that our proposed estimator, the BPR, can perform as well as more complex models with more parameters. Our estimator also performs close to state-of-the-art prediction methods in the benchmark MNIST handwritten digit dataset.
[arXiv] [code]

The Effects of Regulating Food Delivery Platform Design, with Alejandro Sabal
Draft available upon request.
Abstract
There is rising interest amongst regulators in understanding how different platform design choices affect welfare. In this project we focus on two important mechanisms platforms have at their disposal: (1) offering preferential treatment to producers in consumer search and (2) determining producer payments (commission fees). We study the welfare implications of different platform choices in the empirical setting of a food delivery platform that bargains with producers to set commission fees and to adjudicate fixed ranking slots in the consumer search wall. Using transaction level data and click stream search data from a large food delivery platform, we show that both mechanisms are important in practice. Producers with low commission fees are preferred by consumers, and low commission fees are used to attract valuable “anchor” producers that drive consumers into the platform. Search preferencing is also important. Using an A/B test in which rank was randomized we show that search rank is a driver of consumption. To evaluate the impact of different platform designs on consumer and restaurant welfare, we develop a structural model featuring a consumer demand with search frictions, bargaining between restaurants and the platform over ranks and commission fees, and restaurant and consumer entry into the platform. Using the model, we compute counterfactual experiments to assess the impact of regulations forbidding platforms from providing preferential rank to larger restaurants and from setting differential commission fees across restaurants.
[slides]

Pushing Back Against Private Practice: the Spanish Physician Public Exclusivity Bonus, with Jon Gruber, Nuria Mas and Judit Vall
New draft coming soon!

Recommendation Systems and Product Diversity in Online Platforms, with Chinmay Lohani

Abstract
Recommendation systems play a key role in shaping consumer welfare and market concentration on digital platforms. Better recommendations lead to better matches and lower search costs for consumers, but can also increase the market power of producers and the ability of the platform to extract surplus. In this paper, we consider a related, and less studied, consequence of improving recommendation systems: the effect on product variety. We develop a simple model of a two-sided platform with consumer and producer matches mediated by a recommendation system of varying quality. We show that as the quality of the recommendation system improves it becomes optimal, in equilibrium, for producers to offer products close to the average consumer. As a consequence, product diversity decreases, market concentration increases and consumer welfare becomes more unequal, with central consumers better off, but tail consumers worse off. Using the universe of songs from a large music content platform, we provide evidence for the main prediction of the model; with the improvement of recommendations, we see a decrease in music content diversity and a strong correlation between popularity and centrality in the content space.
[short slides]

Teaching

Econometrics for Managers: Correlation & Causality in a Big Data World (MBA) - MIT Sloan 15.034.
Teaching Assistant for Professor Roberto Rigobon and Professor Joseph Doyle. Spring 2023.
Overall rating 6.8/7.0
Selected quote: "Excellent TA. Really appreciated his lecture notes, which were the highlight of the course for me and many of my classmates"
[report]
[lecture notes]

Non-linear Econometrics (Graduate) - MIT 14.385.
Teaching Assistant for Professor Whitney Newey and Professor Max Kasy. Fall 2022.
Overall rating 6.7/7.0
[report]