Applied Econometrics and Policy Evaluation

Lesson 1 • Probability and Statistical Review

Covers random variables, distributions, and sampling theory essential for econometric modelling. Establishes the statistical language used throughout the course.

Lesson 2 • Causal Inference Framework

Presents the potential outcomes framework and the fundamental problem of causal inference. Grounds all subsequent identification strategies in a unified conceptual model.

Lesson 3 • Economic Models and Data

Introduces structural vs. reduced-form models and data types used in policy analysis. Connects economic theory to empirical specification choices.

Lesson 1 • Simple Linear Regression

Derives the OLS estimator and its geometric interpretation. Provides the algebraic foundation for all regression-based methods in later chapters.

Lesson 2 • Inference and Hypothesis Testing

Covers standard errors, t-tests, F-tests, and confidence intervals for regression coefficients. Connects statistical significance to policy-relevant effect sizes.

Lesson 3 • Multiple Regression and Controls

Extends OLS to multiple regressors and explains the role of control variables in reducing omitted variable bias. Directly supports causal interpretation of coefficients.

Lesson 4 • OLS Assumptions and Properties

Examines the Gauss-Markov conditions and their implications for unbiasedness and efficiency. Students learn which violations matter most in applied policy work.

Lesson 1 • Binary and Limited Dependent Variables

Introduces probit, logit, and Tobit models for non-continuous outcomes common in policy evaluation. Covers marginal effects and their correct interpretation.

Lesson 2 • Count and Duration Models

Covers Poisson regression, negative binomial models, and hazard models for count and time-to-event outcomes. Expands the toolkit for health, labour, and crime policy analysis.

Lesson 3 • Heteroskedasticity-Robust Inference

Derives heteroskedasticity-consistent standard errors and explains when they matter. Prepares students to report credible standard errors in applied work.

Lesson 1 • Sources of Endogeneity

Catalogues omitted variables, simultaneity, and measurement error as sources of endogeneity. Motivates the need for instruments before introducing the IV estimator.

Lesson 2 • Instrument Validity and Weak Instruments

Tests instrument relevance and exogeneity and diagnoses weak instrument problems. Students learn to defend instrument choices in policy research.

Lesson 3 • IV Estimator and Two-Stage Least Squares

Derives the IV and 2SLS estimators and their asymptotic properties. Connects instrument relevance and exogeneity to consistent estimation.

Lesson 4 • Local Average Treatment Effects

Introduces the LATE framework and heterogeneous treatment effects under IV. Clarifies what population IV estimates identify in policy contexts.

Lesson 1 • Random Effects and Hausman Test

Presents the random effects GLS estimator and the Hausman test for choosing between FE and RE. Connects model choice to assumptions about unobserved heterogeneity.

Lesson 2 • Heterogeneous Treatment in Panels

Examines staggered adoption designs and the problems with two-way FE under treatment effect heterogeneity. Introduces modern robust DiD estimators for panel data.

Lesson 3 • Panel Data Structure and Notation

Defines balanced and unbalanced panels, within and between variation, and the two-way error component model. Lays the notation used in all panel estimators.

Lesson 4 • Fixed Effects Estimation

Derives the within estimator and the first-difference estimator for eliminating unit-level unobservables. Explains when each approach is preferred in policy settings.

Lesson 1 • Experimental Design Fundamentals

Covers randomisation mechanics, stratification, and power calculations for policy experiments. Ensures students can design experiments that detect policy-relevant effect sizes.

Lesson 2 • Spillovers and Interference

Examines violations of the SUTVA assumption and methods for estimating spillover effects. Applies two-stage randomisation designs to capture network and market spillovers.

Lesson 3 • Estimating Treatment Effects

Derives ITT and TOT estimators and explains when each is the appropriate policy parameter. Connects experimental estimands to the IV framework from Chapter 4.

Lesson 4 • External Validity and Scaling

Evaluates site selection bias, LATE-to-ATE extrapolation, and structural approaches to scaling experimental findings. Prepares students to advise on policy scale-up decisions.

Lesson 1 • Regression Discontinuity Design

Introduces sharp and fuzzy RDD as a quasi-experimental strategy exploiting assignment cutoffs. Covers bandwidth selection, local polynomial estimation, and validity tests.

Lesson 2 • Parallel Trends Testing and Validation

Covers pre-trend tests, event-study plots, and falsification exercises to validate DiD designs. Teaches students to present credible evidence for the parallel trends assumption.

Lesson 3 • Synthetic Control Method

Constructs a data-driven counterfactual from a weighted combination of untreated units. Applies the method to aggregate-level policy interventions with few treated units.

Lesson 4 • Classic Difference-in-Differences

Derives the 2x2 DiD estimator and its identifying assumptions. Establishes the conceptual baseline for all extensions covered in the chapter.

Lesson 1 • Heterogeneous Treatment Effects

Estimates conditional average treatment effects using subgroup analysis, interaction models, and machine learning-assisted methods. Identifies which populations benefit most from a policy.

Lesson 2 • Regression Kink and Bunching Designs

Extends the RDD toolkit to kink designs and bunching estimators for tax and benefit schedules. Covers identification assumptions and elasticity estimation.

Lesson 3 • Communicating Results to Policymakers

Translates econometric findings into policy briefs, visualisations, and executive summaries for non-technical audiences. Builds the professional communication skills needed for policy impact.

Lesson 4 • Replication and Research Transparency

Covers open data practices, code documentation, and reproducibility standards in applied econometrics. Prepares students to produce and evaluate replicable policy research.