CAUSAL INFERENCE
Syllabus
EN
IT
Learning Objectives
LEARNING OUTCOMES:
- Obtain a comprehensive understanding of econometric methodologies used in policy evaluations.
-Learn how to define, identify, and estimate causal effects in the context of policy evaluations.
-Explore the connections between econometrics and causal inference in statistics and understand how these concepts relate to each other.
-Develop practical skills in using the R software for econometric analysis of policy evaluations.
-Discuss the latest research developments in methodological causal inference.
KNOWLEDGE AND UNDERSTANDING:
Learning an advanced statistical and econometric approach.
APPLYING KNOWLEDGE AND UNDERSTANDING:
-Apply knowledge and understanding of the estimation techniques by utilizing microdata in policy evaluation scenarios.
-Applications will be based on widely recognized welfare programs.
-Develop skills in using the following software: R.
MAKING JUDGEMENTS:
-Develop microdata analysis skills useful for preparing research projects.
COMMUNICATION SKILLS:
-Learn to present facts, analyze data, and address economic problems rigorously for both specialist and non-specialist audiences.
LEARNING SKILLS:
-By the end of the course, students will be able to define, identify, and estimate causal estimands in the context of policy evaluations.
-Apply these research techniques to independently develop research projects and ideas.
- Obtain a comprehensive understanding of econometric methodologies used in policy evaluations.
-Learn how to define, identify, and estimate causal effects in the context of policy evaluations.
-Explore the connections between econometrics and causal inference in statistics and understand how these concepts relate to each other.
-Develop practical skills in using the R software for econometric analysis of policy evaluations.
-Discuss the latest research developments in methodological causal inference.
KNOWLEDGE AND UNDERSTANDING:
Learning an advanced statistical and econometric approach.
APPLYING KNOWLEDGE AND UNDERSTANDING:
-Apply knowledge and understanding of the estimation techniques by utilizing microdata in policy evaluation scenarios.
-Applications will be based on widely recognized welfare programs.
-Develop skills in using the following software: R.
MAKING JUDGEMENTS:
-Develop microdata analysis skills useful for preparing research projects.
COMMUNICATION SKILLS:
-Learn to present facts, analyze data, and address economic problems rigorously for both specialist and non-specialist audiences.
LEARNING SKILLS:
-By the end of the course, students will be able to define, identify, and estimate causal estimands in the context of policy evaluations.
-Apply these research techniques to independently develop research projects and ideas.
Prerequisites
Econometrics and Statistics.
Program
One of the central applications of economics is the evaluation of policies and interventions.
Causality is important in econometrics because it enables economists to go beyond mere
correlations, providing a deeper understanding of how economic variables interact and influence each other. This understanding is essential for making sound economic decisions, designing effective policies, and advancing economic theory. For example, it helps determine whether an increase in the minimum wage directly leads to changes in employment levels. This course delves into advanced topics in causal inference, particularly focusing on 'irregular design’ and observation studies. Irregular designs represent complex scenarios where the standard assumptions required for estimating causal effects do not apply. Examples of irregular designs include randomized experiments affected by non-compliance, as well as observational studies with unmeasured confounding variables. Additionally, we explore other challenging irregular designs, such as regression discontinuity designs, where treatment assignment relies on specific thresholds, leading to issues of overlap and other methods for panel data with parallel trends or similar assumptions. In the course's second segment, our exploration extends to the realm of spillover effects. Here, we transcend the common assumption of independence between units and investigate how the treatment of one unit may reverberate onto the outcomes of others. We introduce cutting-edge statistical methodologies tailored for estimating spillover or peerinfluence effects, particularly within clusters of units or social networks.
The initial part focuses on general concepts and context: “Introduction” (approximately 12 hours).
The second part covers the following topic: “Instrumental Variable, Treatment Endogeneity, and other irregular designs” (approximately 12 hours).
The third part (approximately 12 hours) delves into the subject of “Spillovers and Interference: Randomized Experiments and Observational Studies”.
Introduction
· Potential Outcomes Framework
· Fisher’s Randomization Tests
· Permutation Tests for Cluster Randomized Experiments
· Fisher Randomization Tests for Multiple Outcomes
· Neyman Estimator for Completely Randomized Experiments
· Horvitz-Thomson Estimator
· Imputation-Based Estimation
· Stratification and Regression adjustment
· Observational studies and Unconfoundedness
Instrumental Variable, Treatment Endogeneity, and other irregular designs
· Treatment non-compliance
· Methods of Moments and Likelihood-based Inference
· Instrumental Variable in Observational Studies for Treatment Endogeneity
· Regression Discontinuity Deign
· Difference-in-Differences
Spillovers and Interference: Randomized Experiments and Observational Studies
· Definition of Interference and Spillover Effects
· Causal Estimands under Partial and Network Interference
· Two-Stage Randomized Experiments
· IPW and Randomization-based Estimators for Randomized Experiments
· IPW Estimator in Observational Studies
Causality is important in econometrics because it enables economists to go beyond mere
correlations, providing a deeper understanding of how economic variables interact and influence each other. This understanding is essential for making sound economic decisions, designing effective policies, and advancing economic theory. For example, it helps determine whether an increase in the minimum wage directly leads to changes in employment levels. This course delves into advanced topics in causal inference, particularly focusing on 'irregular design’ and observation studies. Irregular designs represent complex scenarios where the standard assumptions required for estimating causal effects do not apply. Examples of irregular designs include randomized experiments affected by non-compliance, as well as observational studies with unmeasured confounding variables. Additionally, we explore other challenging irregular designs, such as regression discontinuity designs, where treatment assignment relies on specific thresholds, leading to issues of overlap and other methods for panel data with parallel trends or similar assumptions. In the course's second segment, our exploration extends to the realm of spillover effects. Here, we transcend the common assumption of independence between units and investigate how the treatment of one unit may reverberate onto the outcomes of others. We introduce cutting-edge statistical methodologies tailored for estimating spillover or peerinfluence effects, particularly within clusters of units or social networks.
The initial part focuses on general concepts and context: “Introduction” (approximately 12 hours).
The second part covers the following topic: “Instrumental Variable, Treatment Endogeneity, and other irregular designs” (approximately 12 hours).
The third part (approximately 12 hours) delves into the subject of “Spillovers and Interference: Randomized Experiments and Observational Studies”.
Introduction
· Potential Outcomes Framework
· Fisher’s Randomization Tests
· Permutation Tests for Cluster Randomized Experiments
· Fisher Randomization Tests for Multiple Outcomes
· Neyman Estimator for Completely Randomized Experiments
· Horvitz-Thomson Estimator
· Imputation-Based Estimation
· Stratification and Regression adjustment
· Observational studies and Unconfoundedness
Instrumental Variable, Treatment Endogeneity, and other irregular designs
· Treatment non-compliance
· Methods of Moments and Likelihood-based Inference
· Instrumental Variable in Observational Studies for Treatment Endogeneity
· Regression Discontinuity Deign
· Difference-in-Differences
Spillovers and Interference: Randomized Experiments and Observational Studies
· Definition of Interference and Spillover Effects
· Causal Estimands under Partial and Network Interference
· Two-Stage Randomized Experiments
· IPW and Randomization-based Estimators for Randomized Experiments
· IPW Estimator in Observational Studies
Books
• Imbens, G. W., & Rubin, D. B. (2015). Causal inference in statistics, social, and
biomedical sciences. Cambridge University Press.
Web page: https://www.cambridge.org/core/books/causal-inference-for-statistics-socialand-biomedical-sciences/71126BE90C58F1A431FE9B2DD07938AB
• Ding, P. (2023). A First Course in Causal Inference. arXiv preprint arXiv:2305.18793.
Web page: https://arxiv.org/abs/2305.18793
• Cattaneo, M. D., Idrobo, N., & Titiunik, R. (2023). A Practical Introduction to
Regression Discontinuity Designs: Extensions. ArXiv. /abs/2301.0895
• Roth, J., Sant’Anna, P. H., Bilinski, A., & Poe, J. (2023). What’s trending in difference-
in-differences? A synthesis of the recent econometrics literature. Journal of
Econometrics.
• Hudgens, M. G., & Halloran, M. E. (2008). Toward causal inference with interference.
Journal of the American Statistical Association, 103(482), 832-842.
• Aronow, P. M., & Samii, C. (2017). Estimating average causal effects under general
interference, with application to a social network experiment.
• Forastiere, L., Airoldi, E. M., & Mealli, F. (2021). Identification and estimation of
treatment and interference effects in observational studies on networks. Journal of the
American Statistical Association, 116(534), 901-918.
• Vazquez-Bare, G. (2022). Identification and estimation of spillover effects in randomized
experiments. Journal of Econometrics
biomedical sciences. Cambridge University Press.
Web page: https://www.cambridge.org/core/books/causal-inference-for-statistics-socialand-biomedical-sciences/71126BE90C58F1A431FE9B2DD07938AB
• Ding, P. (2023). A First Course in Causal Inference. arXiv preprint arXiv:2305.18793.
Web page: https://arxiv.org/abs/2305.18793
• Cattaneo, M. D., Idrobo, N., & Titiunik, R. (2023). A Practical Introduction to
Regression Discontinuity Designs: Extensions. ArXiv. /abs/2301.0895
• Roth, J., Sant’Anna, P. H., Bilinski, A., & Poe, J. (2023). What’s trending in difference-
in-differences? A synthesis of the recent econometrics literature. Journal of
Econometrics.
• Hudgens, M. G., & Halloran, M. E. (2008). Toward causal inference with interference.
Journal of the American Statistical Association, 103(482), 832-842.
• Aronow, P. M., & Samii, C. (2017). Estimating average causal effects under general
interference, with application to a social network experiment.
• Forastiere, L., Airoldi, E. M., & Mealli, F. (2021). Identification and estimation of
treatment and interference effects in observational studies on networks. Journal of the
American Statistical Association, 116(534), 901-918.
• Vazquez-Bare, G. (2022). Identification and estimation of spillover effects in randomized
experiments. Journal of Econometrics
Bibliography
SUGGESTED READINGS
• Hernán MA, Robins JM (2020). Causal Inference: What If. Boca Raton: Chapman &
Hall/CRC
Web page: https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/
• DiTraglia, Lecture Notes on Treatment Effects (or Completely Innocuous Econometrics)
Web page: https://www.treatment-effects.com/treatment-effects.pdf
• Hernán MA, Robins JM (2020). Causal Inference: What If. Boca Raton: Chapman &
Hall/CRC
Web page: https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/
• DiTraglia, Lecture Notes on Treatment Effects (or Completely Innocuous Econometrics)
Web page: https://www.treatment-effects.com/treatment-effects.pdf
Teaching methods
Classroom Lectures;
Laboratory sessions, Exercises
Workshops and students discussions on specific projects assigned
Lessons focused on problem-solving
Analysis of referred readings and textbooks
Suggested articles and readings
Laboratory sessions, Exercises
Workshops and students discussions on specific projects assigned
Lessons focused on problem-solving
Analysis of referred readings and textbooks
Suggested articles and readings
Exam Rules
The grade will be expressed in 30/30.
To acquire the knowledge of the methodology and correctly interpret the results: written exam on theory and applications of the methodology with R. weight: 70%;
To improve the presentation skills and the ability to conduct an independent research project: Group work for a research project, weight: 30%.
Evaluation criteria:
Failed: Significant deficiencies and/or inaccuracies in understanding the topics; limited abilities in analysis and synthesis, frequent generalizations.
18-20: Knowledge and understanding of the topics are barely sufficient with possible imperfections; adequate abilities in analysis, synthesis, and independent judgment.
21-23: Routine knowledge and understanding of the topics; correct abilities in analysis and synthesis with coherent logical reasoning.
24-26: Decent knowledge and understanding of the topics; good abilities in analysis and synthesis with rigorously expressed arguments.
27-29: Comprehensive knowledge and understanding of the topics; remarkable abilities in analysis, synthesis, and good independent judgment.
30-30L: Excellent level of knowledge and understanding of the topics. Outstanding abilities in analysis, synthesis, and independent judgment. Arguments expressed in an original manner.
To acquire the knowledge of the methodology and correctly interpret the results: written exam on theory and applications of the methodology with R. weight: 70%;
To improve the presentation skills and the ability to conduct an independent research project: Group work for a research project, weight: 30%.
Evaluation criteria:
Failed: Significant deficiencies and/or inaccuracies in understanding the topics; limited abilities in analysis and synthesis, frequent generalizations.
18-20: Knowledge and understanding of the topics are barely sufficient with possible imperfections; adequate abilities in analysis, synthesis, and independent judgment.
21-23: Routine knowledge and understanding of the topics; correct abilities in analysis and synthesis with coherent logical reasoning.
24-26: Decent knowledge and understanding of the topics; good abilities in analysis and synthesis with rigorously expressed arguments.
27-29: Comprehensive knowledge and understanding of the topics; remarkable abilities in analysis, synthesis, and good independent judgment.
30-30L: Excellent level of knowledge and understanding of the topics. Outstanding abilities in analysis, synthesis, and independent judgment. Arguments expressed in an original manner.