96562 - Planning and Evaluation of Health Policy

Learning outcomes

At the end of the course, the student has developed the skills supporting evidence-based decision making, and has adequate knowledge of the evaluation approaches and their applications to the different health promotion programs.

Course contents

The course provides an in-depth exploration of the planning and empirical evaluation of health policies. It combines evaluation frameworks and outcome selection with experimental, quasi-experimental and model-based approaches. The central emphasis is on the credibility of the counterfactual, the assumptions required for causal interpretation, the interpretation of empirical outputs, and the reconciliation of evidence obtained from different methods.

Evaluation frameworks and policy planning

• Ex-ante, in-itinere and ex-post evaluation; the evaluation pathway and theory of change.
• Defining outcomes in relation to policy objectives, time horizon, perspective and target population.
• Mortality, morbidity, incidence and prevalence; DALYs, QALYs and economic outcomes.
• Overview of cost-effectiveness, cost-utility and cost-benefit analysis; willingness to pay and the value of a statistical life.
• High-level introduction to microsimulation, uncertainty, equity and heterogeneity in policy impacts.

Health-policy rationale and instruments

• The economic rationale for intervention: externalities, imperfect information, behavioural limitations, market power and distributional concerns.
• Information and awareness campaigns, fiscal interventions, regulations and bans, product standards, restrictions on availability, and behavioural or nudge policies.
• Stakeholders, intended and unintended effects, spillovers, implementation and equity considerations.

Causal inference and empirical methods

• The counterfactual problem; ATE, ATT, ATU and LATE; internal and external validity; SUTVA and spillovers.
• Randomized policy experiments, non-compliance and intention-to-treat analysis.
• Propensity Score Matching and Instrumental Variables, including diagnostics and identifying assumptions.
• Difference-in-Differences, parallel trends, group and time fixed effects, staggered adoption and the limitations of conventional two-way fixed-effects estimators.
• Dynamic Difference-in-Differences and Event Study designs; leads, lags and pre-trend assessment.
• Regression Discontinuity Design and Triple Differences.
• Robustness, falsification, sensitivity analysis, treatment-effect heterogeneity, and the use of multiple methods to reconcile potentially conflicting estimates.

Empirical applications

• The UK 5-a-day campaign: model-based counterfactuals, market responses and distributional effects.
• The Oregon Health Insurance Experiment: randomization, selection, non-compliance, PSM, IV and LATE.
• The OFCOM restrictions on advertising of less-healthy foods to children: Difference-in-Differences, pre-trends and matching.
• Minimum Legal Drinking Age in the United States: staggered policy adoption, Event Studies and Regression Discontinuity at age 21.
• Minimum Unit Pricing for alcohol in Wales: dynamic Difference-in-Differences, spillovers and Triple Differences.
• PROGRESA in Mexico: cluster randomization, theory of change and comparison of experimental and quasi-experimental estimates.

The final part of the course is devoted to cross-case synthesis, guided interpretation of empirical outputs and examination preparation.

Readings/Bibliography

The course will be mostly based on lecture notes and chapters/papers provided through the e-learning platform 

However, some key and very useful books are:

Cameron, A. C. & Trivedi, P. K. (2022). Microeconometrics Using Stata. Volumes I and II. Second Edition. Stata Press.

Angrist, Joshua D., and Jörn-Steffen Pischke, 2014. Mastering'metrics: The path from cause to effect. Princeton University Press.

Journals

A selection of journal articles will be recommended and distributed during the course

Teaching methods

The course combines theoretical recaps, policy case studies, empirical demonstrations and guided interpretation exercises using Stata. Each case study is organised around five questions: the outcome variable(s), the treatment and assignment mechanism, the relevant counterfactual, the estimand, and the assumptions required for causal interpretation.

Complete working codes, logs or selected outputs will be provided through the e-learning platform. Class time will focus primarily on:

• understanding the logic of the empirical design rather than reproducing code line by line;
• interpreting coefficients, tables, graphs and diagnostic tests;
• assessing identification assumptions, robustness and alternative explanations;
• drawing policy conclusions when different methods produce different estimates.

Short small-group activities will be used as formative checkpoints. These activities will require students to interpret an output, identify the relevant assumption or propose an appropriate robustness check. A guided final review and a mock examination will support consolidation before the assessment.

Assessment methods

The final grade will be based on a written exam (1 hour) structured in two parts:

a) Multiple-choice questions on the empirical issues faced when implementing empirical evaluations of health policies (50% of final grade)
b) A section on the interpretation and critical evaluation of the outputs from a health policy evaluation (50% of final grade)

The assessment focuses on the interpretation and critical evaluation of empirical evidence; writing Stata syntax is not an examination requirement.

IMPORTANT: With regard to assessment, the use of AI is prohibited. Any use constitutes a violation of academic integrity.

 

Teaching tools

The e-learning platform will provide access to lecture slides, selected readings, real datasets, Stata codes and logs, case-study reports, output-interpretation worksheets, and a mock examination. Students may use their own laptops during empirical demonstrations. Stata is freely available to registered University of Bologna students through the University software service, at this link

Office hours

See the website of Mario Mazzocchi