Get a Flight Credit worth up to $350
when you apply with code* by May 6, 2024
Goverment Policy Evaluation
OVERVIEW
CEA CAPA Partner Institution: Universidad Carlos III de Madrid
Location: Madrid, Spain
Primary Subject Area: Economics
Instruction in: English
Course Code: 13688
Transcript Source: Partner Institution
Course Details: Level 300, 400
Recommended Semester Credits: 3
Contact Hours: 42
Prerequisites: Core courses in Microeconomics, Macroeconomics and Econometrics, as part of a standard Economics degree, at an intermediate level.
- Since the course places much emphasis on the motivation for and applicability of quantitative methods it is required that students successfully completed courses in Econometrics and Applied Economics (or equivalent courses), which are compulsory in the second and third year of the Grado del Economía at the Universidad Carlos III de Madrid.
DESCRIPTION
1.- Introduction
Introduction and Motivation
Definitions: Economic policies and treatment; treatment effects and causality (causal parameters of interest); control and treatment groups; observed and potential or counterfactual outcomes. Notation.
Problems in the identification and estimation of treatment effects, and their relationship to traditional econometric techniques.
2.- Randomized Experiments in the Social Sciences
Definitions and conditions of a randomized experiment. The advantages of randomization and how it enables the estimation of treatment effects.
Information from other variables: the possibility to verify successful randomization and to study the existence of heterogeneous treatment effects.
Problems and limitations of randomized experiments.
Examples and Applications:
- The effect of class size on educational outcomes (Project STAR).
- The NSW training and subsidy program for unemployed workers (Ham y LaLonde, 1996). Effect on the probability of finding work? Effect on wages?
3.- Observational Studies and Matching Estimators
Exogeneity, matching and multiple regression. Extrapolation. Matching based on the probability to be treated (Propensity Score Matching). Assumptions; estimation of the propensity score; estimator and algorithms; testing for common support.
Examples and Applications:
- Job Training Partnership Act: A program that provides job training and assistance in finding jobs for people in poverty.
- The NSW training and subsidy program for unemployed workers (Dehejia and Wahba, 1999)
4.- Natural or Quasi-natural Experiments
Exploit natural events or policy changes to identify the effect of treatment on the treated. Differences over time. The Difference-in-Differences Estimator: a basic estimator for repeated cross-sections and panel data; common or varying trends; additional regressors. Event study and heterogenous effects in the DiD estimator.
Examples and Applications:
- The effect of immigration on labor markets: the Mariel Boatlift (Card, 1990).
- The effect of minimum wages on employment (Card and Krueger, 1994).
5.- Using Instrumental Variables to Estimate Treatment Effects
The instrumental variable (IV) estimator using data from experiments and quasi-experiments. Wald estimator. Two-stage least square estimator. Interpretation of the IV estimator with homogeneous or heterogeneous treatment effects; eligibility rule; the local average treatment effect (LATE); the monotonicity condition.
Limitations. Marginal Treatment Effects (if time allows).
Examples and Applications:
- The Vietnam Draft Lottery: The effect of military service during the Vietnam War on civilian wages (Angrist, 1990).
- The impact of a Child care program Hogares Comunitarios on nutrition and health (Attanasio, Di Maro, y Vera-Hernandez, 2010 and 2013).
6.- Regression Discontinuity Designs
Sharp and fuzzy regression discontinuity (RD) designs. Continuity in potential outcomes and testable implications. The interpretation and estimation of fuzzy regression discontinuity designs by IV estimator. Parametric and non-parametric implementation. Local linear regression.
Examples and Applications:
- The effect of class size on test scores in reading and maths (Angrist and Lavy, 1999).
7.- The Estimation of Structural Models
Advantages and Disadvantages of atheoretical vs. structural approaches.
The estimation of structural models.
The importance and justification for dynamic models.
General equilibrium effects and models.
Introduction and Motivation
Definitions: Economic policies and treatment; treatment effects and causality (causal parameters of interest); control and treatment groups; observed and potential or counterfactual outcomes. Notation.
Problems in the identification and estimation of treatment effects, and their relationship to traditional econometric techniques.
2.- Randomized Experiments in the Social Sciences
Definitions and conditions of a randomized experiment. The advantages of randomization and how it enables the estimation of treatment effects.
Information from other variables: the possibility to verify successful randomization and to study the existence of heterogeneous treatment effects.
Problems and limitations of randomized experiments.
Examples and Applications:
- The effect of class size on educational outcomes (Project STAR).
- The NSW training and subsidy program for unemployed workers (Ham y LaLonde, 1996). Effect on the probability of finding work? Effect on wages?
3.- Observational Studies and Matching Estimators
Exogeneity, matching and multiple regression. Extrapolation. Matching based on the probability to be treated (Propensity Score Matching). Assumptions; estimation of the propensity score; estimator and algorithms; testing for common support.
Examples and Applications:
- Job Training Partnership Act: A program that provides job training and assistance in finding jobs for people in poverty.
- The NSW training and subsidy program for unemployed workers (Dehejia and Wahba, 1999)
4.- Natural or Quasi-natural Experiments
Exploit natural events or policy changes to identify the effect of treatment on the treated. Differences over time. The Difference-in-Differences Estimator: a basic estimator for repeated cross-sections and panel data; common or varying trends; additional regressors. Event study and heterogenous effects in the DiD estimator.
Examples and Applications:
- The effect of immigration on labor markets: the Mariel Boatlift (Card, 1990).
- The effect of minimum wages on employment (Card and Krueger, 1994).
5.- Using Instrumental Variables to Estimate Treatment Effects
The instrumental variable (IV) estimator using data from experiments and quasi-experiments. Wald estimator. Two-stage least square estimator. Interpretation of the IV estimator with homogeneous or heterogeneous treatment effects; eligibility rule; the local average treatment effect (LATE); the monotonicity condition.
Limitations. Marginal Treatment Effects (if time allows).
Examples and Applications:
- The Vietnam Draft Lottery: The effect of military service during the Vietnam War on civilian wages (Angrist, 1990).
- The impact of a Child care program Hogares Comunitarios on nutrition and health (Attanasio, Di Maro, y Vera-Hernandez, 2010 and 2013).
6.- Regression Discontinuity Designs
Sharp and fuzzy regression discontinuity (RD) designs. Continuity in potential outcomes and testable implications. The interpretation and estimation of fuzzy regression discontinuity designs by IV estimator. Parametric and non-parametric implementation. Local linear regression.
Examples and Applications:
- The effect of class size on test scores in reading and maths (Angrist and Lavy, 1999).
7.- The Estimation of Structural Models
Advantages and Disadvantages of atheoretical vs. structural approaches.
The estimation of structural models.
The importance and justification for dynamic models.
General equilibrium effects and models.
Speak with an
Admissions Advisor
Schedule an appointment to speak with a study abroad expert.
Book Appointment