Course description
A graduate course in econometric theory and applied methods. The sequence develops probability and asymptotic foundations, linear estimation and inference, identification, cross-sectional and panel methods, time-series models, cointegration, and causal research design.
Prerequisites
Calculus, linear algebra, probability/statistics, and introductory regression. Students should be able to use at least one of R, Python, or Stata.
Learning outcomes
Students completing the course should be able to:
- derive OLS and IV estimators and characterize their statistical properties;
- distinguish identification from estimation and inference;
- select covariance estimators consistent with the data dependence structure;
- formulate and estimate panel and time-series models;
- evaluate cointegration, dynamic systems, and causal designs;
- conduct Monte Carlo experiments and produce reproducible empirical research.
14-week teaching plan
| Week | Lecture | Main deliverable |
|---|---|---|
| 1 | Probability to linear regression | OLS derivation and projection geometry |
| 2 | OLS inference | t/F/Wald inference exercise |
| 3 | Functional form and OVB | Marginal-effect and bias analysis |
| 4 | Robust inference | Covariance estimator comparison |
| 5 | Endogeneity and IV | Instrument validity memo |
| 6 | Simultaneous equations | Identification conditions exercise |
| 7 | Panel data | FE/RE comparison |
| 8 | Dynamic panels | GMM moment-condition defense |
| 9 | Time-series foundations | Stationarity and ARMA lab |
| 10 | Unit roots and breaks | Deterministic specification report |
| 11 | VAR/SVAR | IRF and identification report |
| 12 | Cointegration and ARDL | VECM/ARDL comparison |
| 13 | Causal inference | DiD/RD design memo |
| 14 | Research replication | Reproducible empirical presentation |
Assessment
| Component | Weight |
|---|---|
| Theory problem sets | 25% |
| Computational labs | 20% |
| Midterm examination | 20% |
| Replication project | 15% |
| Final empirical paper and presentation | 20% |
Expectations for written work
Every empirical assignment must state the estimand, sample, model, identifying assumptions, covariance estimator, diagnostics, and interpretation. Code must run from a clean session and include a data dictionary and software/package versions.
Core textbooks
- Hayashi, Econometrics.
- Wooldridge, Econometric Analysis of Cross Section and Panel Data.
- Greene, Econometric Analysis.
- Hamilton, Time Series Analysis.
- Lütkepohl, New Introduction to Multiple Time Series Analysis.
- Angrist and Pischke, Mostly Harmless Econometrics.
Recommended research papers
- White (1980), heteroskedasticity-consistent covariance estimation.
- Arellano and Bond (1991), dynamic panel estimation.
- Blundell and Bond (1998), system GMM.
- Engle and Granger (1987), cointegration and error correction.
- Johansen (1991), likelihood-based cointegration analysis.
- Pesaran, Shin, and Smith (2001), bounds testing.
- Callaway and Sant’Anna (2021), difference-in-differences with multiple time periods.