Microeconometrics Summer School

Course Overview

This course gives the basics for good empirical analysis of panel data. It starts by emphasizing the benefits and limitations of panel data over time series or cross-section data. Basic estimation and testing methods for random and fixed effects models are reviewed and illustrated with published empirical applications using Stata and EViews. 

Next, problems of endogeneity in panel models are studied and panel instrumental variable estimation methods as well as Hausman type tests are reviewed and applied using many empirical applications. 

The course also covers an introduction to dynamic panel data models, limited dependent variables and non-stationary panels.

Prerequisites

A solid course in econometrics. Helpful knowledge of software like STATA and/or EViews.

Course Outline

1. Benefits and Limitations of Panel Data
2. Basic Review of Panel Data Methods: Estimation and Test of Hypotheses
3. Simultaneous Equations and Endogeneity in Panel Data Models
4. Dynamic Panel Data: introduction
5. Limited Dependent Variables and Nonstationary Panels

References

Required Text (included in course fees):

Baltagi, Badi H., Econometric Analysis of Panel Data, 6th Edition, Springer, 2021.

https://doi.org/10.1007/978-3-030-53953-5 (23,442 citations by Google Scholar).

Recommended Texts:

Baltagi, Badi H., 2009, A Companion to Econometric Analysis of Panel Data, (Chichester, England: John Wiley & Sons).

Hsiao, Cheng, 2014, Analysis of Panel Data (Cambridge: Cambridge University Press, 3rd ed.).

Wooldridge, J.M., 2010, Econometric Analysis of Cross‑Section and Panel Data (Cambridge: MIT Press).

Arellano, M., 2003, Panel Data Econometrics (Oxford: Oxford University Press).

Baltagi, B.H., ed. 2015, Oxford Handbook of Panel Data, (Oxford: Oxford University Press).

Software / Hardware

STATA *LAPTOP REQUIRED: In order to participate in practical sessions, you must bring your own portable computer.

About the Instructor

Badi Baltagi is a distinguished Professor of Economics at Syracuse University. He holds a PhD in Economics from the University of Pennsylvania. He is Senior Research Associate at the Center for Policy Research at Syracuse University and is a Research Fellow at IZA, CESifo and several other institutions. He was editor of Economics Letters (2011-2021), Associate Editor, Journal of Econometrics, 1999-2018, Replication Editor, Journal of Applied Econometrics, 2003-2018, Editor, Empirical Economics, 1999-2018. He is currently on the Editorial Board of Econometric Reviews. He is a fellow of the Journal of Econometrics as well as Econometric Reviews and a recipient of the Multa and Plura Scripsit Awards from Econometric Theory. Also fellow of the Advances in Econometrics and recipient of the Distinguished Authors Award from the Journal of Applied Econometrics. He is also fellow, founding member and ex-director of the International Association for Applied Econometrics.

Course Overview

The use of survey data is becoming a common practice among economists and social scientists both at the academic and professional level. The main characteristic of this data is that it contains qualitative information, making the use of the regression model not suitable when we deal with models where the dependent variable is either a choice (having or not having private health insurance) or a status (being or not unemployed), or where the dependent variable only takes non-negative values and a significant percentage of the observations are zeroes (expenditure on some particular goods like tobacco). On the other hand, the use of information with a panel data structure is also becoming frequent because it allows to deal with some limitations of the estimation of models with cross-section data.

In this course we deal with estimation of these models, paying special attention to the interpretation of the estimates and the limitations of the different models in the literature.

The course is organized in four blocks: the first one (one and a half sessions) devoted to discrete choice models, including count data models; the second one (one and a half sessions) to limited-dependent variable models, including sample selection issues typical when dealing with cross-section data; the third one (one session) devoted to duration models (survival analysis) in which the dependent variable is the length of a spell in a particular status (e.g. unemployment) and we are interested in estimating how the duration in a particular status affects the probability of leaving that status; and the fourth block to an introduction to (static) panel data, paying attention to the econometric advantages of this type of information.

Practical sessions will be devoted to how to estimate these models using STATA and how to interpret the results. This will be done using real data for different fields: demand analysis, health economics, labour economics, among others. Empirical papers will also be described and discussed in the lectures.

Course Outline

1. Discrete Choice Models (I)
a) Binary choice models
b) Multinomial Models
c) Ordered Models
d) Count Data Models​
2. Limited Dependent Variables Models (II)
a) Tobit Model
b) Sample Selection Model
c) Double-hurdle Models
​3. Duration Models
a) Basic Concepts
b) Continuous Time Models
c) Types of Data
4. Panel Data (Static Models)
a) Panel Data versus Cross-Section
b) Linear Model: Basic Estimators
c) Correlation between Individual Effects and Regressors

References

• Cameron, A.C. and Trivedi, P.K., Microeconometrics. Methods and Applications, Cambridge University Press, 2005
• Cameron, A.C. and Trivedi, P.K., Microeconometrics using STATA, STATA Press, 2010
• Wooldridge, J.M., Econometric Analysis of Cross-Section and Panel Data, MIT Press, 2010

Software / Hardware

STATA * LAPTOP REQUIRED: In order to participate in the practical sessions, you must bring your own portable computer. 

About the Instructor

Jaume García-Villar is Professor of Economics at Universitat Pompeu Fabra and Affiliated Professor of the BSE. He received his PhD from the London School of Economics and Political Science (1985). He was President of the Spanish National Statistics Institute from 2008 until 2011.

Course Overview

The main challenge for policy evaluation is to establish a causal link between interventions and outcomes. The objective of this course is to introduce the main econometric approaches used in the evaluation of public policies: randomized evaluations, natural experiments, regression discontinuity design, selection of observables, difference-in-differences, and synthetic control methods. The course presents the strengths and weaknesses of each approach in terms of internal and external validity. During the morning sessions, each approach will be presented and illustrated with specific examples in the areas of labor economics, health economics, and the economics of education. In the afternoon sessions, we will replicate the results of a prominently published study for each evaluation approach in Stata. Students are provided the corresponding data and code in advance so they can prepare.

Course Outline

• Randomized Evaluations (Experiments)
• Natural Experiments
• Regression Discontinuity Designs
• Selection on Observables (Linear Regression, Matching)
• Difference-in-Differences, Event Studies, Synthetic Control Methods

References

• Abadie, A. and M. D. Cattaneo, 2018, “Econometric Methods for Program Evaluation,” Annual Review of Economics, 10: 465-503.
• Imbens, G. W. and J. M. Wooldridge, 2009, “Recent Developments in the Econometrics of Program Evaluation,” Journal of Economic Literature, 47 (1): 5-86.

Software / Hardware

STATA *LAPTOP REQUIRED: In order to participate in practical sessions, you must bring your own portable computer.

About the Instructor

Albrecht Glitz is an Associate Professor at Universitat Pompeu Fabra, an Affiliated Professor of the Barcelona School of Economics, and a Researcher at IPEG. He received his PhD from University College London in 2008. His research interests include labour economics, the economics of migration, and microeconometrics. His work has been published in the American Economic Review, the Review of Economic Studies, and the Journal of Labor Economics, among others.

Course Overview

This course deals with methods and applications of dynamic discrete choice structural models in Economics. The methods that we will discuss are often used by researchers with interests in Labor Economics, Industrial Organization, Quantitative Macro, Development Economics and many other fields in Applied Micro. The course focuses on the estimation of dynamic discrete choice structural models that allow for modelling the behavior of forward looking agents making discrete decisions. These models constitute a useful tool for policy evaluation and an interesting complement to reduced form approaches. In particular, they have the advantage of allowing for ex-ante policy evaluation, and of providing external validity for the inference, subject to the model assumptions. They also have the benefit of providing a close link between economic theory and empirics, and the possibility of making inference about the predictions of a model.

The course is organized in three blocks. In the first one (3 hours), we will introduce the basic framework, and will review standard estimation techniques based on Maximum Likelihood, which involve solving the dynamic programming problems numerically. The second one (5 hours) will cover an alternative set of estimation methods grouped under the label Conditional Choice Probability estimation methods, which avoid the need of solving for the value functions in estimation. Finally, the third one (2 hours) will review the additional complications involved by dynamic problems that involve game theoretical responses to other agent’s choices. Theory will be complemented with examples from published papers in the literature.

Practical sessions will be devoted to the numerical solution and estimation of these models using STATA (MATA), using simple empirically relevant applications and instructing how to extend these simple applications into fully-fledged rich settings like the ones estimated in the published articles described in the lectures.

Both theory and practical sessions are expected to be interactive. Depending on the class size, breakdown sessions will be considered to foster interaction, especially in the practical sessions.

Prerequisites

Knowledge of Econometrics (at the level of Wooldridge, 2002) and Dynamic Programming (at the level of Chapters 1-2 of Adda and Cooper, 2003) is strongly recommended.

Course Outline

1. Full solution Maximum Likelihood approaches
   a. Introduction
   b. Basic framework: conditional independence
   c. Motivational example: Rust’s engine replacement model
   d. Estimation using full solution techniques
   e. Extensions: unobserved heterogeneity and equilibrium
2. Conditional Choice Probability (CCP) estimation
   a. Conditional value function representation
   b. Finite dependence
   c. Estimation methods
   d. Aguirregabiria and Mira’s iterative approach
   e. Extensions: unobserved heterogeneity and equilibrium
3. An Introduction to Dynamic Discrete Games

References

1. Adda, J. and R. W. Cooper (2003), Dynamic Economics: Quantitative Methods and Applications. The MIT Press.
2. Altug, S. and R. A. Miller (1998), "The Effect of Work Experience on Female Wages and Labor Supply", Review of Economic Studies, 65, 45-85.
3. Aguirregabiria, V. and P. Mira (2002), "Swapping the Nested Fixed Point Algorithm: A Class of Estimators for Discrete Markov Decision Models", Econometrica, 70, 1519-1543.
4. Aguirregabiria, V. and P. Mira (2007), "Sequential Estimation of Dynamic Discrete Games", Econometrica, 75, 1-53.
5. Aguirregabiria, V. and P. Mira (2010), “Dynamic Discrete Choice Structural Models: A Survey”, Journal of Econometrics, 156: 38-67
6. Arcidiacono, P. and P. B. Ellickson (2011), "Practical Methods for Estimation of Dynamic Discrete Choice Models", Annual Review of Economics, 3, 363-394.
7. Arcidiacono, P. and R. A. Miller (2011), "Conditional Choice Probability Estimation of Dynamic Discrete Choice Models with Unobserved Heterogeneity", Econometrica, 79, 1823-1867.
8. Berndt, E. K., B. H. Hall, R. E. Hall, J. A. Hausman (1974). "Estimation and Inference in Nonlinear Structural Models", Annals of Economic and Social Measurement 3, 653-665.
9. Eckstein, Z. and K. Wolpin (1989), “The Specification and Estimation of Dynamic Stochastic Discrete Choice Models: A Survey”, Journal of Human Resources, 24: 562-598
10. Hong, H. and M. Shum (1998), "Structural Estimation of Auction Models", In: Patrone F., García-Jurado I., Tijs S. (eds) Game Practice: Contributions from Applied Game Theory. Theory and Decision Library (Series C: Game Theory, Mathematical Programming and Operations Research), vol 23. Springer, Boston, MA.
11. Hotz, V. J. and R. A. Miller (1993), "Conditional Choice Probabilities and the Estimation of Dynamic Structural Models", Review of Economic Studies, 60, 497-529.
12. Keane, M. P. and K. I. Wolpin (1997), "The Career Decisions of Young Men", Journal of Political Economy, 105, 473-522.
13. Lee, D. and K. I. Wolpin (2006), "Intersectoral Labor Mobility and the Growth of the Service Sector", Econometrica, 74, 1-46.
14. Llull, J. (2018), "Immigration, Wages, and Education: A Labor Market Equilibrium Structural Model", Review of Economic Studies, 85, 1852-1896.
15. Llull, J. (2018), "Selective Immigration Policies and the U.S. Labor Market”, mimeo, MOVE, Universitat Autònoma de Barcelona, and BSE.
16. Miller, R. A. (1997), “Estimating models of dynamic optimization with microeconomic data”, in M. Pesaran and P. Schmidt (eds.), Handbook of Applied Econometrics, Vol. 2, pp. 246-299
17. Todd, P. and K. Wolpin (2006), "Assessing the Impact of School Subsidy Program in Mexico: Using a Social Experiment to Validate a Dynamic Behavioral Model of Child Schooling and Fertility", American Economic Review, 96, 1384-1417.
18. Rust, J. (1987), "Optimal Replacement of GMC Bus Engines: An Empirical Model of Harold Zurcher", Econometrica, 55, 999-1033.
19. Rust, J. (1994), “Structural Estimation of Markov Decision Processes”, in R. E. Engle and D. L. McFadden (eds.), Handbook of Econometrics, Vol. 4, Ch. 51.

Software / Hardware

STATA *LAPTOP REQUIRED: In order to participate in the practical sessions, you must bring your own portable computer.

About the Instructor

Joan Llull is the Director of MOVE, Associate Professor of Economics at Universitat Autònoma de Barcelona, and Associate Research Professor of the Barcelona School of Economics. He is also an External Fellow at CReAM (UCL).

Professor Llull's research focuses on labor economics, and more specifically on immigration, internal migration, occupational mobility, inequality, human capital, family economics, and health. His main research typically estimates dynamic discrete choice models of equilibrium, but several of his papers also use more reduced form approaches. His work has been published in the Review of Economic Studies, Journal of Human Resources, and the European Economic Review, among others.

Professor Llull is Data Editor of the Economic Journal, a member of the editorial board of the Review of Economic Studies, guest editor for a special issue on "Economics of Migration: Labour Market Impacts and Migration Policies" of Labour Economics, the Journal of the European Association of Labour Economists, and associate editor of SERIEs --- The Journal of the Spanish Economic Association.

Course Overview

This course covers difference-in-differences (DiD) estimators for policy analysis with panel data. The focus is on using and combining built-in Stata commands, along with recent user-written commands, to allow simple estimation methods, robust inference, and flexibility in the pattern of treatment effects. We will see how the usual two-way fixed effects approach can be extended in the staggered intervention case to allow for heterogeneous treatment effects across cohort and time. 

In addition, we will see how to incorporate covariates into the analysis in all settings. Simple solutions to violations of the parallel trends assumption also will be covered. Problems and approaches with unbalanced panels and issues of statistical inference with few treated or control groups will also be covered. 

The course ends with a discussion of my recent research on staggered designs with nonlinear models, with focus on binary, fractional, and nonnegative responses.

Prerequisites

Participants should have a good working knowledge of ordinary least squares estimation, fixed effects estimation, and basic nonlinear models such as logit, probit, and exponential conditional means. Sufficient background is provided by my introductory econometrics book, Introductory Econometrics: A Modern Approach, 7e, Cengage, 2020. My book Econometric Analysis of Cross Section and Panel Data, 2e, MIT Press, 2010, covers the background material at a higher level. An optional video of approximately 90 minutes will provide background on potential outcomes and treatment effects estimators. This material will be applied in DiD contexts.

Course Outline

1. Introduction. Two-Period Difference-in-Differences. Parallel Trends. Controlling for Covariates via Regression Adjustment and Propensity Score Methods.
2. More than Two Time Periods. Two-Way Fixed Effects. Heterogeneous Treatment Effects. Controlling for Covariates via Regression Adjustment and Treatment Effects Estimators.
3. Staggered Interventions, I. Pooled OLS and Extended TWFE. Aggregation. Testing and Correcting for Violations of Parallel Trends. Unbalanced Panels.
4. Staggered Interventions, II. Rolling Methods. Long Differencing. Propensity Score and Doubly Robust Methods. Imputation Methods. Clustering. Inference with few Treated Units.
5. Nonlinear DiD. Binary, Fractional, and Nonnegative Responses. Pooled Quasi-Maximum Likelihood Estimation. Poisson Fixed Effects.

References

For some of the course, I will rely on my recent working papers, “Two-Way Fixed Effects, the Two-Way Mundlak Regression, and Difference-in-Differences Estimators” and “Simple Approaches to Nonlinear Difference-in-Differences with Panel Data.” 

Software / Hardware

STATA *LAPTOP REQUIRED: In order to participate in practical sessions, you must bring your own portable computer.

About the Instructor

Jeffrey M. Wooldridge is University Distinguished Professor of Economics at Michigan State University, where he has taught since 1991. From 1986 to 1991, Dr. Wooldridge was an Assistant Professor of Economics at the Massachusetts Institute of Technology. He received his bachelor of arts, with majors in computer science and economics, from the University of California, Berkeley, in 1982, with high distinction in general scholarship. He received his doctorate in economics in 1986 from the University of California, San Diego.

Dr. Wooldridge has published a large number of articles in internationally recognized journals, as well as several book chapters, including articles in the Handbook of Econometrics and the Handbook of Applied Econometrics. He is the author of Introductory Econometrics: A Modern Approach (South-Western, 4e, 2009) and Econometric Analysis of Cross Section and Panel Data (MIT Press, 2002).

Dr. Wooldridge is a fellow of the Econometric Society and of the Journal of Econometrics. His other awards include the Plura Scripset award from Econometric Theory, the Sir Richard Stone prize from the Journal of Applied Econometrics, and three teacher-of-the-year awards from the graduate economics association at MIT. He has also been an Alfred P. Sloan Research Fellow.

Dr. Wooldridge currently serves on the editorial boards of the Stata Journal. He has served as editor of the Journal of Economic Literature, Journal of Business and Economic Statistics, co-editor of Econometric Theory, econometrics co-editor of Economics Letters, and associate editor for the Journal of Econometrics, and the Review of Economics and Statistics.

Course Overview

This course provides up-to-date coverage of dynamic panel data models, discrete choice panel data models as well as censored panel data models and the estimation of dynamic panel data models subject to selection. Apart from a review of relevant theory, the focus of the course is on the practical application of these models to various data contexts: small T – large N, unbalanced panels, rotating panels and pseudo panels constructed from cohort data. 

Practical sessions provide the audience with a guide about how to proceed with these models, in particular, in how to phrase the restrictions and contrasts for an adequate specification of the model to be estimated. All the practical sessions are resolved with STATA.

Prerequisites

Some background in Econometrics is strongly recommended.

Course Outline

1. Review of Linear Panel Data models & Introduction to dynamic panel data models
2. Dynamic Panel Data and extensions
3. Censored Panel Data Models
4. Sample selection Panel Data Models. New developments
5. Discrete choice Panel Data Models: static and dynamic

Basic References

• Arellano, M. (1992), “Discrete choice for Panel Data”, Investigaciones Economicas, 2003.
• Arellano, M. and O. Bover (1995), “Another look at the instrumental-variable estimation of error components models”, Journal of Econometrics 68, 29-51.
• Arellano, M. and R Carrasco (2003), ”Binary Choice models with predetermined variables”, Journal of Econometrics, 115, 155-165.
• Arellano, M. and B Honoré (2003), ”Panel Data Models, some recent developments”, Journal of Econometrics, 115, 155-165.
• Baltagi, B.H. (1995), Econometric Analysis of Panel Data, John Wiley. 4th edition 2008.
• Blundell, R. and S. Bond (1998), “Initial conditions and moment restrictions in dynamic panel data models”, Journal of Econometrics 87, 115-143.
• Steve Bond, 2002, ”Dynamic panel data models: a guide to microdata methods and practice,” CeMMAP working papers CWP09/02, Centre for Microdata Methods and Practice, Institute for Fiscal Studies.
• Collado, MD (1997), ”Estimating dynamic models from time series of independent cross-sections”, Journal of Econometrics, 82, 37–62.
• Deaton, A. 1985. Panel data from time series cross-sections. Journal of Econometrics 30, 109-126.
• Dustman, C. and Rochina-Barrachina, M. R. (2007) ”Selection Correction in Panel Data Models: An Application to Labour Supply and Wages” Econometrics Journal (2007), volume 10, pp. 263293. doi: 10.1111/j.1368-423X.2007.00208.x
• Jimenez-Martin, Labeaga and Rochina-Barrachina (2009), ”Comparison of estimators in dynamic panel data sample selection and switching models”, mimeo, presented in the Cambridge PD conference.
• Sergi Jiménez-Martín & José María Labeaga, 2016. "Monte Carlo evidence on the estimation of AR(1) panel data sample selection models, "Working Papers 2016-01, FEDEA.
• Hansen, L.P. (1982), “Large sample properties og gneralized method of moments estimators”, Econometrica 50, 1029-1054.
• Jiménez-Martín, S. (1999), ”Controlling the endogeneity of strike variables in the estimation of wage settlement equations”, Journal of Labor Economics, 17, 587-606
• Jiménez-Martín, S. & J.M. Labeaga & M. al Sadoon, 2020. "Consistent estimation of panel data sample selection models," Working Papers 2020-06, FEDEA.
• Jones, A.M. and J.M. Labeaga (2003), “Individual heterogeneity and censoring in panel data estimates of tobacco expenditures”, Journal of Applied Econometrics, 18, 157-177.
• Kyriazidou, E. (1997), ”Estimation of a panel data sample selection model”, Econometrica 65, 1335-1355
• Kyriazidou, E. (2001), ”Estimation of dynamic panel data sample selection models”, Review of Economic Studies 68, 543-572.
• Labeaga, JM. (2001), “A double-hurdle rational addiction model with heterogeneity: estimating the demand for tobacco”, Journal of Econometrics, 93, 49-72.
• Labeaga, JM. (2001), “Efficiency comparisons in dynamic panel data models with limited dependent variables”, WP, UNED, Madrid.
• Nickell, S. (1981), “Biases in dynamic models with fixed effects”, Econometrica 49, 1417-1426.
• Semykina A. and Jeffrey M. Wooldridge ”Estimation of Dynamic Panel Data Models with Sample Selection”, April 28, 2009
• Semykina, A. and Wooldrigde, J.M. (2005), ”Estimating Panel Data Models in the Presence of Endogeneity and Selection: Theory and Application,” mimeo.
• Windmeijer, Frank, 2005. ”A finite sample correction for the variance of linear efficient two-step GMM estimators,” Journal of Econometrics, Elsevier, vol. 126(1), pages 25-51, May.
• Wooldrigde, J.M. ”Selection Corrections for Panel Data under Conditional Mean Independence Assumptions,” Journal of Econometrics, 68 (1995): 115-132.
• Majid M. Al-Sadoon & Sergi Jiménez-Martín & José M Labeaga, 2019. "Simple Methods for Consistent Estimation of Dynamic Panel Data Sample Selection Models," Working Papers 1069, Barcelona School of Economics.

Software / Hardware

STATA *LAPTOP REQUIRED: In order to participate in practical sessions, you must bring your own portable computer.

About the Instructor

Sergi Jiménez-Martín is Professor of Economics at Universitat Pompeu Fabra and Affiliated Professor at the BSE. He received his PhD from the Universitat Pompeu Fabra in 1994. He is currently Chair of the FEDEA-La Caixa Economía de la Salud y Hábitos de Vida. He is also Associate Editor of Empirical Economics, a member of the Scientific Council of Applied Economic Perspectives and Policy as well as Cuadernos Económicos de ICE.