Fayssal Ayad
@faylosophe.bsky.social
PhD Applied Economics & Statistics. Humble man in search of truth. Views expressed are my own at the time of posting -- after that, who knows.
Reposted by Fayssal Ayad
Me trying to keep up with the DiD literature.
May 3, 2025 at 10:22 PM
Me trying to keep up with the DiD literature.
Reposted by Fayssal Ayad
Just posted updated version of our DID textbook! We now have drafts of all chapters, including the one on general designs! Now you can tell your friends still on X that they are DID-outdated :-) Happy easter for those of you that celebrate it. papers.ssrn.com/sol3/papers....
Credible Answers to Hard Questions: Differences-in-Differences for Natural Experiments
This book introduces applied researchers to modern Differences-in-Differences (DID) methods, that they can use to obtain credible answers to hard causal inferen
papers.ssrn.com
April 18, 2025 at 2:28 PM
Just posted updated version of our DID textbook! We now have drafts of all chapters, including the one on general designs! Now you can tell your friends still on X that they are DID-outdated :-) Happy easter for those of you that celebrate it. papers.ssrn.com/sol3/papers....
Reposted by Fayssal Ayad
Kirill Borusyak, Mauricio Caceres Bravo, Peter Hull: Estimating Demand with Recentered Instruments https://arxiv.org/abs/2504.04056 https://arxiv.org/pdf/2504.04056 https://arxiv.org/html/2504.04056
April 8, 2025 at 6:00 AM
Kirill Borusyak, Mauricio Caceres Bravo, Peter Hull: Estimating Demand with Recentered Instruments https://arxiv.org/abs/2504.04056 https://arxiv.org/pdf/2504.04056 https://arxiv.org/html/2504.04056
Reposted by Fayssal Ayad
link 📈🤖
Distributional Instrumental Variable Method (Holovchak, Saengkyongam, Meinshausen et al) The instrumental variable (IV) approach is commonly used to infer causal effects in the presence of unmeasured confounding. Conventional IV models commonly make the additive noise assumption, which is
Distributional Instrumental Variable Method (Holovchak, Saengkyongam, Meinshausen et al) The instrumental variable (IV) approach is commonly used to infer causal effects in the presence of unmeasured confounding. Conventional IV models commonly make the additive noise assumption, which is
February 12, 2025 at 4:26 PM
link 📈🤖
Distributional Instrumental Variable Method (Holovchak, Saengkyongam, Meinshausen et al) The instrumental variable (IV) approach is commonly used to infer causal effects in the presence of unmeasured confounding. Conventional IV models commonly make the additive noise assumption, which is
Distributional Instrumental Variable Method (Holovchak, Saengkyongam, Meinshausen et al) The instrumental variable (IV) approach is commonly used to infer causal effects in the presence of unmeasured confounding. Conventional IV models commonly make the additive noise assumption, which is
Reposted by Fayssal Ayad
Okay, I made an updated version of the guide "Python Packages for Applied Economists" to reorganize a bit, incorporate suggestions, and put it on Github like a grownup: github.com/clibassi/pyt...
Comments welcome!
Comments welcome!
GitHub - clibassi/python-packages-for-applied-economists: A curated collection of Python packages for applied economists, organized by functionality to support econometric analysis, data management, v...
A curated collection of Python packages for applied economists, organized by functionality to support econometric analysis, data management, visualization, and specialized tasks. - clibassi/python-...
github.com
November 24, 2024 at 10:18 PM
Okay, I made an updated version of the guide "Python Packages for Applied Economists" to reorganize a bit, incorporate suggestions, and put it on Github like a grownup: github.com/clibassi/pyt...
Comments welcome!
Comments welcome!
Reposted by Fayssal Ayad
I cannot wait to dive into this new simpler approach for Diff in Diff even though I am no longer programming my own Stata code. :-) #EconSky @edwardnorton.bsky.social @edwardnorton.bsky.social @jmwooldridge.bsky.social
www.nber.org/papers/w33026
www.nber.org/papers/w33026
A Flexible, Heterogeneous Treatment Effects Difference-in-Differences Estimator for Repeated Cross-Sections
Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...
www.nber.org
October 8, 2024 at 1:02 PM
I cannot wait to dive into this new simpler approach for Diff in Diff even though I am no longer programming my own Stata code. :-) #EconSky @edwardnorton.bsky.social @edwardnorton.bsky.social @jmwooldridge.bsky.social
www.nber.org/papers/w33026
www.nber.org/papers/w33026
#EconSky This is a brand new book by Chernozhukov et al on state of the art causal machine learning methods.
causalml-book.org
causalml-book.org
CausalML
causal machine learning book
causalml-book.org
February 22, 2024 at 10:24 PM
#EconSky This is a brand new book by Chernozhukov et al on state of the art causal machine learning methods.
causalml-book.org
causalml-book.org
Hi #EconSky. This is a new cool WP by Ahrens et al on the benefit of combining DDML with stacking for causal inference.
arxiv.org/abs/2401.01645
arxiv.org/abs/2401.01645
Model Averaging and Double Machine Learning
This paper discusses pairing double/debiased machine learning (DDML) with stacking, a model averaging method for combining multiple candidate learners, to estimate structural parameters. We...
arxiv.org
January 10, 2024 at 10:34 PM
Hi #EconSky. This is a new cool WP by Ahrens et al on the benefit of combining DDML with stacking for causal inference.
arxiv.org/abs/2401.01645
arxiv.org/abs/2401.01645
#EconSky: A hot new survey WP have been dropped by the masters Arkhangelsky & Imbens on causal models for longitudinal and panel data. A must read if you want to cover everything from DiD & TWFE estimators to nonlinear models, synthetic controls, & design-based inference.
arxiv.org/abs/2311.15458
arxiv.org/abs/2311.15458
Causal Models for Longitudinal and Panel Data: A Survey
This survey discusses the recent causal panel data literature. This recent literature has focused on credibly estimating causal effects of binary interventions in settings with longitudinal data,...
arxiv.org
November 29, 2023 at 10:01 PM
#EconSky: A hot new survey WP have been dropped by the masters Arkhangelsky & Imbens on causal models for longitudinal and panel data. A must read if you want to cover everything from DiD & TWFE estimators to nonlinear models, synthetic controls, & design-based inference.
arxiv.org/abs/2311.15458
arxiv.org/abs/2311.15458
#EconSky If you're still diving in the ocean of propensity score, this is a very practical new WP that establishes very useful equivalence results when using the inverse probability tilting and the covariate balance propensity score methods.
arxiv.org/abs/2310.18563
arxiv.org/abs/2310.18563
Covariate Balancing and the Equivalence of Weighting and Doubly...
We show that when the propensity score is estimated using a suitable covariate balancing procedure, the commonly used inverse probability weighting (IPW) estimator, augmented inverse probability...
arxiv.org
November 1, 2023 at 8:37 PM
#EconSky If you're still diving in the ocean of propensity score, this is a very practical new WP that establishes very useful equivalence results when using the inverse probability tilting and the covariate balance propensity score methods.
arxiv.org/abs/2310.18563
arxiv.org/abs/2310.18563
#EconSky: This is a very cool brand new WP by Spiess, Imbens and Venugopal on exploiting double and single-descent phenomenon in ML to deal with highly over parameterized models in causal inference, including synthetic control with many control units.
www.nber.org/papers/w31802
www.nber.org/papers/w31802
Double and Single Descent in Causal Inference with an Application to High-Dimensional Synthetic Cont...
Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...
www.nber.org
October 29, 2023 at 7:46 PM
#EconSky: This is a very cool brand new WP by Spiess, Imbens and Venugopal on exploiting double and single-descent phenomenon in ML to deal with highly over parameterized models in causal inference, including synthetic control with many control units.
www.nber.org/papers/w31802
www.nber.org/papers/w31802
#EconSky: FE-TE estimator is biased under correlated heterogeneity (CH). Pesaran & Yang propose in brand new WP a test of CH which works well even if the time dimension is VERY short. To avoid bias they recommend using a new trimmed mean group estimator.
arxiv.org/abs/2310.11680
arxiv.org/abs/2310.11680
Trimmed Mean Group Estimation of Average Treatment Effects in...
Under correlated heterogeneity, the commonly used two-way fixed effects
estimator is biased and can lead to misleading inference. This paper proposes a
new trimmed mean group (TMG) estimator which...
arxiv.org
October 29, 2023 at 5:44 AM
#EconSky: FE-TE estimator is biased under correlated heterogeneity (CH). Pesaran & Yang propose in brand new WP a test of CH which works well even if the time dimension is VERY short. To avoid bias they recommend using a new trimmed mean group estimator.
arxiv.org/abs/2310.11680
arxiv.org/abs/2310.11680
#EconSky: principal ignorability (PI) assumption is usually invoked to estimate causal effects of compliers vs. non compliers. This is a cool WP by the team of Nguyen et al., tailoring new sensitive techniques for several PI based methods.
arxiv.org/abs/2303.05032
arxiv.org/abs/2303.05032
Sensitivity analysis for principal ignorability violation in...
An important strategy for identifying principal causal effects, which are often used in settings with noncompliance, is to invoke the principal ignorability (PI) assumption. As PI is untestable,...
arxiv.org
October 24, 2023 at 8:04 PM
#EconSky: principal ignorability (PI) assumption is usually invoked to estimate causal effects of compliers vs. non compliers. This is a cool WP by the team of Nguyen et al., tailoring new sensitive techniques for several PI based methods.
arxiv.org/abs/2303.05032
arxiv.org/abs/2303.05032
#EconSky: if you don't have a control group don't worry. This is cool WP allowing forecasting of counterfactuals by ML in your causal panel analysis. A dedicated R package is accompanying this methodology.
papers.ssrn.com/sol3/papers....
papers.ssrn.com/sol3/papers....
Losing Control (Group)? The Machine Learning Control Method for Counterfactual Forecasting
Without a control group, the most widespread counterfactual methodologies for causal panel analysis cannot be applied. We fill this gap with the Machine Learnin
papers.ssrn.com
October 15, 2023 at 10:15 PM
#EconSky: if you don't have a control group don't worry. This is cool WP allowing forecasting of counterfactuals by ML in your causal panel analysis. A dedicated R package is accompanying this methodology.
papers.ssrn.com/sol3/papers....
papers.ssrn.com/sol3/papers....
#EconSky: 📢 Not used to self promote but just to announce that my paper: Mapping the path forward: A prospective model of natural resource depletion and sustainable development, has been published since a while in Resources Policy (1/n).
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Mapping the path forward: A prospective model of natural resource depletion and sustainable developm...
This paper introduces a sustainable development model of natural resource depletion that quantifies the optimal relationship between the two. Adopting…
www.sciencedirect.com
October 8, 2023 at 9:50 PM
#EconSky: 📢 Not used to self promote but just to announce that my paper: Mapping the path forward: A prospective model of natural resource depletion and sustainable development, has been published since a while in Resources Policy (1/n).
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
#EconSky: If you are still worrying on instrument validity for heterogeneous causal effects, this a new paper that proposes a general framework on the basis of a non-parametric approach.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Instrument validity for heterogeneous causal effects
This paper provides a general framework for testing instrument validity in heterogeneous causal effect models. The generalization includes the cases w…
www.sciencedirect.com
October 8, 2023 at 6:54 PM
#EconSky: If you are still worrying on instrument validity for heterogeneous causal effects, this a new paper that proposes a general framework on the basis of a non-parametric approach.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
#EconSky: if you are using the DiD design then you should worry on the PT assumption. This is an interesting WP that looks deeper on the issue of selection and it's link with PT.
arxiv.org/abs/2203.09001
arxiv.org/abs/2203.09001
Selection and parallel trends
We study the connection between selection into treatment and the parallel trends assumptions underlying difference-in-differences (DiD) designs. Our framework accommodates general selection...
arxiv.org
October 7, 2023 at 7:10 PM
#EconSky: if you are using the DiD design then you should worry on the PT assumption. This is an interesting WP that looks deeper on the issue of selection and it's link with PT.
arxiv.org/abs/2203.09001
arxiv.org/abs/2203.09001
#EconSky: In economics, we have many impossibility theorems. This is another impossibility result in inference with large clusters dependence.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Some impossibility results for inference with cluster dependence with large clusters
This paper focuses on a setting with observations having a cluster dependence structure and presents two main impossibility results. First, we show th…
www.sciencedirect.com
October 5, 2023 at 11:05 PM
#EconSky: In economics, we have many impossibility theorems. This is another impossibility result in inference with large clusters dependence.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
#EconSky If you've already moved to this plateform, this is a very cool WP on movers design.
I have a new working paper w/Ivan Badinski, Amy Finkelstein, & Matt Gentzkow!
We use a "movers" design to study the role of physician practice intensity for widely-documented geographic variation in healthcare spending
The punchline? Doctor sorting matters a lot!
www.dropbox.com/scl/fi/s3tci...
We use a "movers" design to study the role of physician practice intensity for widely-documented geographic variation in healthcare spending
The punchline? Doctor sorting matters a lot!
www.dropbox.com/scl/fi/s3tci...
October 4, 2023 at 11:37 PM
#EconSky If you've already moved to this plateform, this is a very cool WP on movers design.
#EconSky What can time series regressions tell us about policy counterfactuals? A brand new cool open access Econometrica paper.
onlinelibrary.wiley.com/doi/10.3982/...
onlinelibrary.wiley.com/doi/10.3982/...
What Can Time‐Series Regressions Tell Us About Policy Counterfactuals?
We show that, in a general family of linearized structural macroeconomic models, knowledge of the empirically estimable causal effects of contemporaneous and news shocks to the prevailing policy rule....
onlinelibrary.wiley.com
October 3, 2023 at 8:23 PM
#EconSky What can time series regressions tell us about policy counterfactuals? A brand new cool open access Econometrica paper.
onlinelibrary.wiley.com/doi/10.3982/...
onlinelibrary.wiley.com/doi/10.3982/...
#EconSky A brand new interesting WP checking inference credibility for TWFE models under PT assumption and HTE. @jondr44.bsky.social
arxiv.org/abs/2309.15983
arxiv.org/abs/2309.15983
What To Do (and Not to Do) with Causal Panel Analysis under...
Two-way fixed effects (TWFE) models are ubiquitous in causal panel analysis in political science. However, recent methodological discussions challenge their validity in the presence of...
arxiv.org
September 30, 2023 at 11:19 AM
#EconSky A brand new interesting WP checking inference credibility for TWFE models under PT assumption and HTE. @jondr44.bsky.social
arxiv.org/abs/2309.15983
arxiv.org/abs/2309.15983