Alejandro Sanchez-Becerra

*New draft alert* #EconTwitter Experiments + machine learning are changing the way we measure effect heterogeneity in policy settings. arxiv.org/pdf/2306.03363… My paper proposes the first valid confidence intervals for the VCATE, the treatment effect explained by observables.

Calling all aspiring data scientists seeking impactful careers! Join our virtual information sessions to explore how Emory's 12-month STEM-designated Master's in Data Science equips you with advanced skills in AI, Machine Learning, and Statistical Modeling.

Scan the QR to join today's live info session, and see how the Data Science MS at Emory can enhance your education and propel your career in just 10 months!

@EmoryDataSci, @EmoryEconomics, and @emory_sociology proudly hosted an international & interdisciplinary workshop on Causal Network Analysis on August 14, 2025 @EmoryUniversity. With 15 presentations & a keynote talk by Prof. Tom Valente (USC), the day was a splendid success!

@nathanihoff @emory_sociology Welcome to Emory!

I am thrilled to share that I will be joining @emory_sociology as an Assistant Professor this fall! Excited for a new chapter in Atlanta 🔥🍑

@YuyaSasaki12 @yan7812 @MetricReviews Congratulations Mingfeng and Karen!

Congratulations to Mingfeng Zhan and Karen X. Yan (@yan7812) for winning the Best Paper Awards! Congratulations to Maurizio Daniele for winning the Emerging Scholar Award! We look forward to your continued contributions to ER (@MetricReviews). tandfonline.com/doi/full/10.10…

@JoergStoye @CornellEcon Congrats Joerg!

Professional update: I am thrilled and honored to announce that I am now appointed Ta-Chung Liu Professor at @CornellEcon. Many thanks to my fantastic colleagues for the support!

BREAKING NEWS The Royal Swedish Academy of Sciences has decided to award the 2024 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel to Daron Acemoglu, Simon Johnson and James A. Robinson “for studies of how institutions are formed and affect prosperity.” #NobelPrize

14 different literature review types categorized by method used

I've posted a review article on deep learning for economists arxiv.org/pdf/2407.15339. The DL literature is vast, and I initially found it pretty daunting to find the parts of it that are most relevant to economic applications. I hope people will find this useful! #econtwitter

@AngYu_soci Covariates may not necessarily be predictive, even if all other assumptions are correct. You can quantify the level of heterogeneity predicted by covariates more explicitly, as I discuss here arxiv.org/pdf/2306.03363. Code in beta version for experimental settings.

In causal inference, does it make sense to check if "the outcome models are fitting the data well"? Our estimator is ML- and EIF-based.

Excited by the work of @I4Replication regarding replications of existing articles. This type of exercise is important for enhancing the scientific process.

We're thrilled to share 4 new reproductions/replications!

I feel so fortunate to be working with so many amazing students and collaborators! Check out below the work we've output in the last two months!

@VivianoDavide @clintmharris1 Thanks Davide!

@AlejandroSBe @clintmharris1 Interestingly whether and how we want to do indexing vs betw/ outcomes multiple hypothesis testing corrections depends on how we interpret such outcomes from a decision theoretic perspective. We have some discussion in Sec 4 here in case of interest arxiv.org/abs/2104.13367

If this seems like mathy snake-oil, consult your intuition! It is *surprising* that your treatment has "insignificant" positive effects on many, many good outcomes. Every time another positive point estimate pops out on another good outcome, we are a little more surprised.

@clintmharris1 When you construct a weighted average you might reduce uncertainty and have lower standard errors. But it your standard error estimates are correct, your coverage should still be at the intended level. Now, I do agree that estimating weights using data could introduce issues.

@clintmharris1 Good points! It seems like you have some type of index strategy in mind. The multiple comparisons corrections that I’m familiar with try to correct the family wise error rate (FWER), which means that in practice they inflate the critical values, not decrease them.

Very intuitive explanation of CDFs.

People sometimes find CDFs unintuitive. Here's how to understand them intuitively. A CDF is just magnitudes plotted over ranks.

Interesting reanalysis of AER papers. A very important exercise regarding the robustness of published research in economics. Hope to see more of this work in the future!