90350 - CAUSAL INFERENCE AND PROGRAM EVALUATION

Conoscenze e abilità da conseguire

The course illustrates the most recent identification strategies for the quantitative assessment of causal effects using observational data by referring to micro-econometric applications. It will cover matching and difference-in-differences strategies, and quasi-experimental approaches to identification. At the end of the class, student will be able: - to critically understand the application of these tools in the recent empirical economic literature; - to apply these approaches to design his/her own program evaluation.

Contenuti

Il corso è offerto solo in lingua inglese.

The course covers empirical strategies for applied (mainly micro-economics) research questions. The main goal of the course is to provide an overview of the statistical tools for counterfactual analysis and a deeper discussion of a selection of these tools (matching; synthetic control methods; IV-LATE and regression discontinuity design), from the theoretical and empirical point of view.

The course illustrates the identification strategies, estimation and other related issues (eg internal and external validity) that are relevant for the assessment of causal effects (or equivalently treatment effects) using observational data.

We will discuss Randomized Trials because they represent the benchmark of non-experimental methods. Experimental methods are illustrated and discussed in detail in other courses at graduate and undergraduate level.

Fixed effects and random effects model are covered in other courses offered within the program and will not be discussed in detail. We will touch on fixed effects models, because they are related to the Difference-in-Differences approach to identification and this is a pre-requisite to discuss Synthetic Control approach. The emphasis of the course will be on Synthetic control and recent advances in the literature on Difference-in-Differences.

We touch on Instrumental Variable Strategies as they are pre-requisite to discuss Fuzzy-Regression Discontinuity. Compared to standard econometrics courses, we will discuss advanced issues related to IV (LATE interpretation and external validity).

Features of each particular econometric tool will be illustrated from both the theoretical and practical point of view, often through the discussion of empirical applications.

The emphasis will be on the practical implementation of each approach.

Topics

1. Fundamentals of Impact Evaluation (approximately 6 hours, including practice sessions; week 1 classes)
   • The Fundamental Problem of Causal Inference
   • Potential Outcomes Framework
   • Basic Approaches to Identification:Randomized Trials
   • Basic Approaches to Identification:Selection on Observables (Propensity Score Matching)
     
     
2. Quasi-Experiments: IV (approximately 6 hours; week 2 classes)
   • Instrumental Variables: Identification, Estimation, Falsification Checks (Placebo), Interpretation (LATE), External Validity; Weak Instruments
     
     
3. Strategies exploting the structure of the data (approximately 12 hours, including practice sessions; week 3 and week 4 classes) 
   
   • Difference-in-Differences and recent advances on two-way fixed effects and difference-in-differences with heterogeneous treatment effects
   • Non-linear difference-in-differences
   • Synthetic Control Methods 
     
     
4. Quasi-Experiments: Jumps (approximately 6 hours,  week 5 classes)

• Regression Discontinuity Design: Sharp and Fuzzy Designs, Identification, Estimation, Falsification Checks
• Multiple cutoffs, Multiple Running Variables (eg Geographic RDD): insights
• External Validity: Extrapolating Away from the Cutoff
  
  

Testi/Bibliografia

Il corso è offerto solo in lingua inglese.

 

Lectures will be based on the following books and articles. We will rely on many empirical applications, referring to the recent literature. The list of papers with empirical applications will be distributed at the beginning of the class.

BOOKS

Mostly Harmless Econometrics, Angrist and Pischke

Mastering 'Metrics: The Path from Cause to Effect, Angrist and Pischke

Causal inference: the mixtape

ARTICLES

WEEK 1 reading list

Holland, Paul W (1986) Statistics and Causal Inference, Journal of the American Statistical Association 81 (396): pp. 945-970, with discussion

Abadie, Cattaneo (2018) Econometric Methods for Program Evaluation Annual Review of Economics 10:465–503

WEEK 2 reading list

Angrist, J., Imbens, G. and Rubin, D. (1996) Identification of Causal Effects Using Instrumental Variables, Journal of the American Statistical Association, 91 (434) pp. 444-455, with discussion

Angrist, J. (2004) Treatment Effect Heterogeneity in Theory and Practice, The Economic Journal, 114 (494) p.C52-C83

WEEK 3 and 4 reading list

Muralidharan, Karthik, and Nishith Prakash. 2017. "Cycling to School: Increasing Secondary School Enrollment for Girls in India." American Economic Journal: Applied Economics, 9 (3): 321-50.

Athey and Imbens (2006) Identification and Estimation in Nonlinear Difference-in-Differences Models, Econometrica 74(2) pp.431-497

De Chaisemartin and D'Haultfoeuille (forthcoming) Two-Way Fixed Effects and Differences-in-Differences with Heterogeneous Treatment Effects: A Survey,  Econometrics Journal.

Abadie, A. and Diamond, A. and Hainmueller, J. (2010) Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program, Journal of the American Statistical Association, 105(490), pp. 493-505.

Abadie, Alberto. 2021. "Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects." Journal of Economic Literature, 59 (2): 391-425.

Arkhangelsky, Dmitry, Susan Athey, David A. Hirshberg, Guido W. Imbens, and Stefan Wager. 2021. "Synthetic Difference-in-Differences." American Economic Review, 111 (12): 4088-4118.

 

WEEK 5 reading list

 Hahn, J. and Todd, P. and Van der Klaauw, W. (2001) Identification and Estimation of Treatment Effects with a Regression Discontinuity Design, Econometrica 69 (1)

 Research work by Cattaneo and co-authors https://sites.google.com/site/matiasdcattaneo/

Metodi didattici

Il corso è offerto solo in lingua inglese.

Each topic will be covered in class and in a practice session.

The lessons will cover the presentation of theoretical and applied topics related to the various econometric tools. The applications will be illustrated in the classroom and then resumed in practical lessons in the computer lab using STATA.

All teaching materials will be distributed through the e-learning platform of the University of Bologna.

Research articles listed among the references can be downloaded from the web. You may use the search engine: http://acnp.unibo.it/cgi-ser/start/it/cnr/fp.html

Most books in the reference list are available at the University libraries. You may check availability through the search engine http://sol.unibo.it/SebinaOpac/Opac?sysb =

To address the need to offer the whole course online, to make students' experience more interactive, I will experimentally adopt innovative teaching tools (such as peer instruction; see link with reference) using adequate techical support during lectures relying for instance on free available software such as Pingo (https://pingo.coactum.de/). Students do not need to install the software ahead but need to have a device (mobile phone or laptop) who can access internet during the lectures in which this approach will be implemented.

In addition, peer education requires a great deal of investment from students as students have to read the textbook before coming to class.

Modalità di verifica e valutazione dell'apprendimento

Il corso è offerto solo in lingua inglese.

The assessment will be based on

- replication study (70%; max 23 out of 33 points)

Students will have to submit: a) a STATA code to replicate the results of a published paper that uses at least one of the tools illustrate in the course ; b) a Fact checking report, summarizing the results of the replciation study (guidelines to be illustrated in class); c) slides on an extension of the replicated study (eg : same analisi on more recent data/on data on another country; different analysis -with the tools illustrated in this class- on the same data)

- in class participation (20%; max 7 out of 33 points)

- questions on the course program during the oral exam/replication study audit (10% max 4 out of 33 points)

According to the indications of the council of the School of Economics and Management, the indications on the graduation of the grade are reported.

• <18: insufficient

• 18-23: sufficient

• 24-27: good

• 28-30: excellent

• 30 sum laude: excellent with praise/with honors

 

Strumenti a supporto della didattica

Il corso è offerto solo in lingua inglese.

Slides, teaching material , practice using STATA.

Self-evaluation on-line tests will be made available through the e-learning platform https://elearning-cds.unibo.it/

Lectures involve the presentation of theoretical and applied issues of the various econometric methods. Applications are discussed in class and replicated during the computer laboratory session using STATA.

Software STATA: available for students of the Department of Economics (CAMPUS license) and at the Computer Lab of the School of Economics and Management.

Link ad altre eventuali informazioni

https://www.kuleuven.be/english/education/teaching-tips/activating-students/peer-instruction)

Orario di ricevimento

Consulta il sito web di Margherita Fort