As usual they have a top group of presenters. They’re also offering a Zoom option if you can’t make it in person:
2023 Northwestern Main and Advanced Causal Inference Workshops
[please recirculate to others who might be interested]
We are excited to be holding our 12th annual workshop on Research Design for Causal Inference at Northwestern Law School in Chicago, IL. We invite you to attend. Our apologies for the length of this message.
Main Workshop: Monday – Friday, August 7-11, 2023
Advanced Workshop: Monday – Thursday (midday, August 14-17, 2023
Optional Machine Learning Primer: Sunday afternoon, Aug. 13, 2023
What’s special about these workshops are the speakers. The session will be taught by world-class causal inference researchers, who are experts in the topics they will discuss. See below for details. In person-registration is limited to 125 participants for each workshop.
There will also be a Zoom option, but please come in person if you can. The online experience is not the same.
For information and to register:
Bernie Black [Northwestern University, Pritzker Law School, Institute for Policy Research, and Kellogg School of Management, Department of Finance]
Scott Cunningham [Baylor University, Department of Economics]
Main Workshop Overview: We will cover the design of true randomized experiments and contrast them to natural or quasi experiments and to pure observational studies, where part of the sample is treated, the remainder is a control group, but the researcher controls neither which units are treated vs. control, nor administration of the treatment. We will assess the causal inferences one can draw from specific “causal” research designs, threats to valid causal inference, and research designs that can mitigate those threats.
Most empirical methods courses survey a variety of methods. We will begin instead with the goal of causal inference, and how to design a research plan to come closer to that goal, using messy, real-world datasets with limited sample sizes. The methods are often adapted to a particular study.
Advanced Workshop Overview: The advanced workshop provides in-depth discussion of selected topics that are beyond what we can cover in the main workshop. The principaltopics for 2023 are application of machine learning methods to causal inference; when and how to cluster standard errors, quantile and nonlinear difference-in-differences, doubly robust estimation of causal effects; difference-in-differences methods for staggered treatments (applied to different units at different times); and empirical Bayes approaches to estimating individual effects.
Target audience for main workshop: Quantitative empirical researchers (faculty and graduate students)in social science, including law, political science, economics, many business-school areas (finance, accounting, management, marketing, etc.), medicine, sociology, education, psychology, etc. –anywhere that causal inference is important.
We will assume knowledge, at the level of an upper-level undergraduate econometrics or similar course,of multivariate regression, including OLS, logit, and probit; basic probability and statistics including confidence intervals, t-statistics, and standard errors; and some understanding of instrumental variables. This course should be suitable both for researchers with recent PhD-level training in econometrics and for empirical scholars with reasonable but more limited training.
Target Audience for Advanced Workshop: Empirical researchers who are familiar with the basics of causal inference (from our main workshop or otherwise), and want to extend their knowledge. We will assume familiarity, but not expertise, with potential outcomes, difference-in-differences, and panel data methods.
Main Workshop Faculty (in order of appearance)
Donald B. Rubin (Harvard University)
Donald Rubin is John L. Loeb Professor of Statistics Emeritus, at Harvard. His work on the “Rubin Causal Model” is central to modern understanding of causal inference with observational data. Principal research interests: statistical methods for causal inference; Bayesian statistics; analysis of incomplete data. Website: https://statistics.fas.harvard.edu/people/donald-b-rubin. Wikipedia: http://en.wikipedia.org/wiki/Donald_Rubin
Jens Hainmueller (Stanford University)
Jens Hainmueller is Professor in the Stanford Political Science Department, and co-Director of the Stanford Immigration Policy Lab. He also holds a courtesy appointment in the Stanford Graduate School of Business. His research interests include statistical methods, political economy, and political behavior. Web page: http://www.stanford.edu/~jhain//. Papers on SSRN: https://ssrn.com/author=739013.
Matias Cattaneo is Professor in the Department of Operations Research and Financial Engineering at Princeton University, with positions in Princeton’s Department of Economics, Center for Statistics and Machine Learning, and Program in Latin American Studies. His research focus is on econometrics, statistics, data science and decision science, with particular interests in program evaluation and causal inference. Website: https://cattaneo.princeton.edu/.
Rocio Titiunik (Princeton University)
Rocío Titiunik is Professor of Politics at Princeton University. She specializes in quantitative methodology for the social sciences, with emphasis on quasi-experimental methods for causal inference. Website: https://scholar.princeton.edu/titiunik.
Advanced Workshop Faculty
Christian Hansen (University of Chicago)
Christian Hansen is Wallace W. Booth Professor of Econometrics and Statistics at the University of Chicago, Booth School of Business. His research includes use of machine learning methods in estimation of causal and policy effects, estimation of panel data models, inference using clustered standard errors, quantile regression, and weak instruments. Website: https://voices.uchicago.edu/christianhansen/.
Jeffrey Wooldridge (Michigan State University)
Jeffrey Wooldridge is University Distinguished Professor at Michigan State University and the author of leading undergraduate and graduate textbooks on econometrics. His research interests include causal inference and the econometrics of panel data, including nonlinear models in difference-in-differences and general policy analysis settings. Website: www.econ.msu.edu/faculty/wooldridge/ .
Yiqing Xu (Stanford University)
Yiqing Xu is Assistant Professor of Political Science at Stanford University.His main methods research involves causal inference with panel data. Website: https://yiqingxu.org/.
Christopher Walters (UC Berkeley)
Christopher Walters is Associate Professor of Economics at the University of California, Berkeley. His research focuses on the topics in labor economics and the economics of education, including early childhood programs, school effectiveness, and labor market discrimination.
Main Workshop Outline
Monday, August 7 (Donald Rubin)
Introduction to Modern Methods for Causal Inference
Overview of causal inference and the Rubin “potential outcomes” causal model. The “gold standard” of a randomized experiment. Treatment and control groups, and the core role of the assignment (to treatment) mechanism. Causal inference as a missing data problem, and imputation of missing potential outcomes. Experimental design and applications to observational studies. One-sided and two-sided noncompliance.
Tuesday, August 8 (Jens Hainmueller)
Matching and Reweighting Designs for “Pure” Observational Studies
The core, untestable requirement of selection [only] on observables. Ensuring covariate balance and common support. Matching, reweighting, and regression estimators of average treatment effects. Propensity score methods.
Wednesday, August 9 (Jens Hainmueller)
Panel Data and Difference-in-Differences
Panel data methods: pooled OLS, random effects, and fixed effects. Simple two-period DiD and panel data extensions. The core “parallel trends” assumption. Testing this assumption. Event study (leads and lags) and distributed lag models. Accommodating covariates. Triple differences. Robust and clustered standard errors.
Thursday, August 10 (Matias Cattaneo)
Regression discontinuity (RD) designs: sharp and fuzzy designs; continuity-based methods and bandwidth selection; local randomization methods and window selection; empirical falsification of RD assumptions; extensions and generalizations of canonical RD setup: discrete running variable, multi-cutoff, multi-score, and geographic designs. RD software website: “https://rdpackages.github.io/”.
Friday, August 11: Morning (either Matias Cattaneo or Rocio Titiunik)
Instrumental variable methods
Causal inference with instrumental variables (IV): the role of the exclusion restriction and first stage assumption; the monotonicity assumption and local average treatment effect (LATE) interpretation; applications to randomized experiments with imperfect compliance, including intent-to-treat designs and two-stage estimation. Connections between IV and fuzzy RD designs.
Friday, August 11: Afternoon – Feedback on your own research
Attendees will present their own research design questions from current work in breakout sessions and receive feedback on research design. Session leaders: Bernie Black, Scott Cunningham, Rocio Titiunik or Matias Cattaneo). Additional parallel sessions if needed to meet demand.
Advanced Workshop Outline
Sunday afternoon, August 13 (optional) (Christian Hansen)
Primer on machine learning approaches to prediction
Introduction to “machine-learning” approaches to prediction algorithms, aimed at attendees with limited knowledge of machine learning methods.. Shrinking a large set of potential predictors. Predicting without overpredicting: training and test sets; cross-validation. Lasso, regression trees, random forests, and deep nets. High-dimensional model selection (function classes, regularization, tuning). Combining models (ensemble models, bagging, boosting), model evaluation, and implementation.
Monday, August 14: Christian Hansen
Applications of machine learning to causal inference
When and how can machine learning methods be applied to causal inference questions. Limitations (prediction vs estimation) and opportunities (data pre-processing, prediction as quantity of interest, high-dimensional nuisance parameters), with examples from an emerging empirical literature.
Tuesday, August 15: Jeffrey Wooldridge
Advanced matching and balancing methods
Choosing among the many available matching and balancing methods. Estimators that aim directly at covariate balance. Combining balancing with regression and doubly robust estimators in cross-sectional and panel data settings. Synthetic controls.
Wednesday, August 16: Yiqing Xu
Advanced panel data methods
Causal inference with panel data using parametric, semi-parametric, non-parametric methods for addressing imbalance between treated and control units. Bias in classic DiD models using two-way fixed effects. Topics include interactive fixed effects and matrix completion methods, as well as reweighting approaches such as panel matching, trajectory balancing and augmented synthetic control. Relative strengths and weaknesses of different methods will be discussed.
Thursday morning, August 17: Christopher Walters
Empirical Bayes Methods
Empirical Bayes methods for studying heterogeneity and estimating individual effects in settings with many unit-specific parameters (e.g., school, teacher, or physician quality; neighborhood effects on economic mobility; firm effects on wages; employer-specific labor market discrimination). Topics will include methods for quantifying variation in effects, empirical Bayes shrinkage for estimating individual effects, and connections to multiple testing and decision theory.
We plan informal, wine-and cheese receptions for all attendees following the main sessions on Monday August 7 and Thursday August 10 for the main workshop, and on Monday August 14 for the advanced workshop.
Registration and Workshop Cost
The workshop fee includes all materials, breakfast, lunch, snacks, and the receptions.
Main Workshop: tuition is $900 ($600 for post-docs and graduate students; $500 if you are Northwestern-affiliated). .
Advanced Workshop: tuition is $650 ($450 for post-docs and graduate students; $300 if you are Northwestern affiliated).
Discount for attending both workshops: There is a $250 discount for non-Northwestern persons attending both workshops, for combined cost of $1,300 ($800 for post-docs and graduate students ($150 additional discount for Northwestern affiliates).
Zoom option: We are charging the same amount for in-person and virtual attendance.Partly, we want to encourage in-person attendance. We also want to allow attendees to switch from one format to the other.
You can cancel either workshop five weeks in advance, for a 75% refund – July 3, 2023, for the Main Workshop and July 10, 2023, for the Advanced Workshop – or carry over your registration to next year for full credit. There is a 50% refund after these dates but before three weeks in advance, July 17, 2023, for the Main Workshop and July 24, 2023, for the Advanced Workshop. After these dates no refund will be given, because we can’t realistically replace you. But you can carry over the registration fee to a future workshop.
We know the workshop is not cheap. We use the funds to pay our speakers and expenses. Prof. Black does not pay himself.
You should plan on full days, roughly 9:00-4:30. Breakfast will be available at 8:30.
Bernard Black (Northwestern University)
Bernie Black is Nicholas J. Chabraja Professor at Northwestern University, with positions in the Pritzker School of Law, the Institute for Policy Research, and the Kellogg School of Management, Finance Department. Principal research interests: health law and policy; empirical legal studies, law and finance.Web page with link to CV:www.law.northwestern.edu/faculty/profiles/BernardBlack/.Papers on SSRN: http://ssrn.com/author=16042.
Scott Cunningham (Baylor University)
Scott Cunningham is Professor of Economics at Baylor University. Principal research interests: mental healthcare; suicide; corrections; sex work; abortion policy; drug policy. Web page with link to CV: https://www.scunning.com .
On selected days, we will run parallel Stata and R sessions, following the main lectures, to illustrate code for the research designs discussed in the lectures. Some speakers will also build Stata or R code into their lecture slides. Presenters: Scott Cunningham (Stata) and Joshua Lerner (R).
We will also provide a repository (likely on GitHub) of datasets and code to illustrate the methods presented in the workshop.
Questions about the workshops: Please email Bernie Black (firstname.lastname@example.org) or Scott Cunningham (email@example.com) for substantive questions or fee waiver requests, and Sebastian Bujak (Sebastian.firstname.lastname@example.org) for logistics and registration questions.