January is almost over, which means conference season planning is in full swing! If causal inference is your thing—and it should be—join us at the ACIC The American Causal Inference Conference is coming to Salt Lake City, May 11-14, and it’d be great to see some of you there!

📝 Submit your paper to COunterfactual Methods for Policy Impact Evaluation (COMPIE 2026) conference Paris, June 24-26 Deadline: Feb 28 🎤 Keynotes: • @pedrosantanna.bsky.social (Emory) • Michela Giorcelli (UCLA) Info: knowledge4policy.ec.europa.eu/event/compie... #econsky

Paul is here making our lives 1000 times easier!!! Thanks for this!!!

Our workshop Econometrics at Emory: Causal Inference with Panel Data is happening May 2–3, 2025! 🎯 We are thrilled to welcome Guido Imbens as our keynote speaker. But that’s just the start. 🎙️ We have a stellar lineup of speakers from both academia and industry to talk about causal inference!

We from @emoryeconomics.bsky.social are excited to host the "Econometrics in Emory: Causal Inference with Panel Data." 📅 May 2–3, 2025 🔗 econometricsatemory.com We aim to strengthen the relationship between academia and industry researchers so we all learn from each other!

Seven econometricians, all four census regions represented. At Clemson U.

Rest in peace, Ed Leamer

One of the things that I love the most is to give seminars, present new papers/results, and get feedbacks. I haven’t been able to do that as much as before (for the sake of the kids), but I’m really excited to present tomorrow at Stanford. If you are around, come say hi!

The course outline for the 6th Summer School is now online! 🎓📖 ✨ Explore the details here: summer-school.edu.gr/program/ @pedrosantanna.bsky.social @jmwooldridge.bsky.social

Join us in Greece for the 6th Summer School, June 22-27! @jmwooldridge.bsky.social and I will cover a lot of Casual Inference procedures. And we promise to make it fun (or as fun as it gets) 🥳😄 The program outline is now online! 🎓📖 : summer-school.edu.gr/program/ Hope to see you there!

It just hit me that it has been 5 years since the tragic deaths of Kobe and Gigi Bryant and seven other victims. I remembered being devastated when I heard, and that was 1 year before my kids were born!

To continue the public good streak, here are some cool materials that our @emoryeconomics.bsky.social student, Marcelo Ortiz-Villavicencio, prepared for his independent studies. marcelortiz.com/ECON697R/ I am biased, but the slides are clear and relevant to many econometrics topics.

@maxkasy.bsky.social uses POs in his bandit and reinforcement learning slides here maxkasy.github.io/home/ML_Oxfo...

Plenty of time to enroll! A combination of self learning and online discussion/interaction. Plus, an optional, friendly competition estimating policy effects at the end of the course.

This is very good

The book by Bibhas Chakraborty and Erica Moodie on Dynamic Treatment Regimes/RL is a helpful text, takes more of a stats perspective and mostly offline learning but has all the fundamentals using PO

I have been trying to get more familiar with the basics and advances Reinforcement Learning. I wonder if anyone have seen a paperthat use potential outcome notation within the RL world. This would make the life of many economists much easier. Anyone? All help is appreciated!

Really great stuff

Was thrilled to learn that @larsvil.bsky.social has a tutorial on reproducibility! Check it out, #EconSky. larsvilhuber.github.io/tutorial-rep...

I've decided to collect my DiD materials in a single place. psantanna.com/did-resources There, you will find - 14 lectures of my comprehensive DiD course - Shorter lectures/talks I have given on DiD - My DiD R/Stata/Python packages - Some DiD checklists - DiD materials from my friends Enjoy!

For those attending #ASSA2025, come to our session on ML/AI + unstructured data (e.g., text and images). We'll have a couple of papers on the econometrics of it, and I'll talk about incorporating this type of data in demand estimation. It's Sunday Jan 5 at 10:15 AM. www.aeaweb.org/conference/2...

I have now posted 14 decks of slides covering a lot of DiD materials. I also want to add a few more decks: - Triple Differences - DiD with Continuous Treatment - Instrumented DiD - Synthetic DiD and related extensions - Heterogeneity analysis in DiD I also want to add more applications to these!

The last four decks (at least for now) are here: 11 Problems of TWFE: psantanna.com/DiD/11_Stagg... 12 Staggered DiD (mostly CS): psantanna.com/DiD/12_CS.pdf 13 Challenges when Treatment turns on and off: psantanna.com/DiD/13_On_of... 14 Random Treatment Timing: psantanna.com/DiD/14_Rando...

Five more decks: 06 Double Machine Learning: psantanna.com/DiD/06_DML.pdf 07 Repeated Cross Sections: psantanna.com/DiD/07_RCS.pdf 08 Event Study and Dynamics: psantanna.com/DiD/08_ES.pdf 09 TWFE with multiple periods: psantanna.com/DiD/09_Twfe.... 10 Pre-tests psantanna.com/DiD/10_Prete...

Here are the first five sets of slides: 01 Introduction: psantanna.com/DiD/01_Intro... 02 Classical 2x2 setup: psantanna.com/DiD/02_two_b... 03 Clustering issues: psantanna.com/DiD/03_Clust... 04 Functional form: psantanna.com/DiD/04_Funct... 05 Covariates: psantanna.com/DiD/05_Covar...

Lecture 02 is here: psantanna.com/DiD/02_two_b... This covers the 2x2 DiD setup without covariates + some clustering issues.

Lecture 02 is here: psantanna.com/DiD/02_two_b... This covers the 2x2 DiD setup without covariates + some clustering issues.

Here we go. Lecture 1: Introduction psantanna.com/DiD/01_Intro... I will add lectures in a proper GitHub repo later and polish them further. Hope these will find people in need!

I am thinking of making my DiD materials that I used for teaching in the last 5 years available. I am hoping these would find begineers in need of a step-by-step opinionated walkthrough of several developments Let’s see if I find the energy to do this!

Last Sunday of 2024!!! And I have been reflecting on how grateful I am for being blessed with my family! Taking this time to recharge as ny parental leave is now over. 2025, here we go!

Merry Christmas, everybody!!! From my family to yours!

So far, the hardest part with 4 kids under 4yo has been health. I think we have been with at least 1 kid sick for the last 45 days. The amount of mild fever, running noses, congestion, coughing, and the associated sleep deprivation is piling up. Boy, that is draining!

I am trying to work while these two keep knocking on my door! Hard to focus!

Just a reminder that @gmcd.bsky.social and @kylefbutts.bsky.social have made an incredible public good describing how to use data.table and fixest to encourage moving from Stata to R! stata2r.github.io

With @overleaf.com down, here's a quick tip to keep working. If you're already syncing overleaf with Dropbox, you can still find your files in Dropbox/Apps/Overleaf Install the latex extension in vscode or download an editor like texstudio. Should take 15 mins or less to get set up! #econsky

Applications are now open. Join is in Greece!

Big FOMO for the #SEA2024!! I was soo looking forward for the conference, but then, two sick kids force me to change plans in the last minute. For all going to the #SEA2024, enjoy! That is my favorite conference—it always feel like a reunion.

In order to add some clarity on why the answer is just "use robust SE", here's a simulations that follow Abadie et al.'s paper "When Should You Adjust Standard Errors for Clustering?" This is pure random assignment of treatment. Cluster SE are way too big, and robust SE are close.

Impressive! And this is why testing for cluster correlation in error terms can't be the right way to decide on the level of clustering. At least not in a design-based approach. Unobserved heterogeneity in the distribution of the treatment, or in the TEs, gets mistaken for "cluster correlation."

Thanks Pedro! Link to WP: arxiv.org/abs/2411.10959 and slides dviviano.github.io/projects/pre... We show how we can estimate treatment effects using remote sensed outcomes without having to correctly specify a model for remote sensed outcomes and leverage arbitrary ML methods. 1/n

A very interesting new working paper by the greats Ashesh Rambachan, Rahul Singh, and @vivianodavide.bsky.social: arxiv.org/pdf/2411.10959 It seems that this is another area where the empirical common practice was "too fast,'' and econometrics is catching up! Cool and empirically relevant stuff!

Sunday walk!

❗6th Summer School in Advanced Economics. ------------------------------------------------- ➡️Where: Spetses Island, Greece. ➡️When: 22-27 June 2025. ▶️ Instructors: Jeff Wooldridge @jmwooldridge.bsky.social and myself, @pedrosantanna.bsky.social We will cover a lot! For more info⤵️

A few weeks ago, I spoke on a panel about publishing at the CSWEP’s CeMENT workshop. I wanted to briefly share some of that advice here.

If you’ve been on bluesky for a while, had a twitter presence, and want to help thicken periphery of our network, a suggestion: Run sky-follower-bridge. (Or if you haven't recently, run it again!) Many new folks joining here recently that you want to follow and yr notification may be overflowing.

A #HealthPolicy & #EpiSky 🧵 (re-posting): 1. With renewed threats to the #ACA, my colleagues & I have a new paper out in #HealthAffairs #HAScholar on the effects of #Medicaid expansion under the ACA on #HIV care continuum outcomes across the US. academic.oup.com/healthaffair...

The super useful {afcharts} package applies colour palettes and a {ggplot2} theme which follows best practice set out in the UK gov’s Analysis Function guidance

The gist of presenting a new paper for the first time is really refreshing!!

If you're teaching R, or statistics, and you want your students to get more practice of data wrangling with "messy" data, you can use the {messy} package to turn the clean data you currently use for examples into more realistic looking data! #RStats #StatsEd

If you listened to Guido talk, he stated that he was using DiD and TWFE interchangeably. The problems he raised was about the TWFE estimator, and not DiD. But people now just say DDDiD 🤔 What else should they use that can cover the needs of empirical researchers? There is no silver bullet.