Training on Impact Evaluation using Difference-in-Differences and Regression Discontinuity Design

A policy launches.

A program rolls out.

A reform is announced.

Six months later, someone presents a graph showing improvement and proudly declares:

"The intervention worked."

Maybe.

Or maybe the economy improved.

Maybe seasonal factors changed.

Maybe people were already improving before the intervention started.

Maybe the comparison group was completely different from the beginning.

The uncomfortable truth?

Many impact evaluations answer the wrong question very confidently.

Because measuring change is easy.

Measuring causality is hard.

That's why serious evaluators rely on methods like Difference-in-Differences (DiD) and Regression Discontinuity Design (RDD).

These approaches help answer the question decision-makers actually care about:

"What impact did the intervention truly have?"

In this course, you'll learn how to:

• Design rigorous impact evaluations using Difference-in-Differences and Regression Discontinuity methods
• Estimate causal effects when randomized experiments are not feasible
• Identify valid comparison groups and counterfactual scenarios
• Detect biases that undermine evaluation credibility
• Generate evidence that policymakers, donors, investors, and stakeholders can trust

And yes, we'll discuss why some of the most celebrated program successes disappear once proper evaluation methods are applied.

Overview

Governments, development agencies, NGOs, international organizations, research institutions, and private-sector organizations increasingly require credible evidence to determine whether policies, programs, and investments are producing meaningful outcomes. Traditional before-and-after comparisons often fail to distinguish genuine program effects from broader social, economic, environmental, or institutional changes.

Impact evaluation seeks to identify causal effects by estimating what would have happened in the absence of an intervention. While Randomized Controlled Trials (RCTs) are often considered the gold standard, practical, ethical, political, financial, and operational constraints frequently make randomization impossible.

Difference-in-Differences (DiD) and Regression Discontinuity Design (RDD) are among the most widely used quasi-experimental methods for estimating causal impacts using observational data. These approaches provide robust frameworks for constructing credible counterfactuals and measuring intervention effects in real-world settings.

Difference-in-Differences leverages changes over time between treatment and comparison groups to isolate intervention effects, while Regression Discontinuity Design exploits eligibility thresholds and policy cutoffs to estimate causal impacts with high internal validity.

These methods are extensively applied in public policy, education, healthcare, labor economics, social protection, governance, infrastructure development, environmental programs, public finance, and international development.

This course equips participants with practical and strategic expertise in designing, implementing, analyzing, and interpreting impact evaluations using DiD and RDD approaches. Participants will learn how to assess assumptions, conduct statistical analysis, interpret findings, and communicate evidence effectively.

Through practical exercises, econometric workshops, policy simulations, software demonstrations, and real-world case studies, participants will develop the skills needed to conduct high-quality impact evaluations across diverse sectors.

Duration

10 Days

Who Should Attend

• Monitoring and Evaluation (M&E) professionals
• Economists and econometricians
• Policy analysts
• Researchers and academics
• Development practitioners
• Government evaluation officers
• Public health analysts
• Data scientists and statisticians
• Impact assessment specialists
• Donor agency staff
• Program managers
• Consultants and advisors

Individual Impact

• Strengthen impact evaluation expertise
• Improve econometric and analytical capabilities
• Enhance policy evaluation competencies
• Develop stronger causal reasoning skills
• Improve interpretation of evaluation findings
• Gain practical experience with leading evaluation techniques

Organizational Impact

• Improve the quality and credibility of impact evaluations
• Strengthen accountability and transparency
• Enhance evidence-based decision-making
• Improve resource allocation effectiveness
• Support policy learning and adaptive management
• Increase stakeholder confidence in evaluation findings
• Strengthen organizational research and analytical capacity

By the end of this course, participants will be able to:

• Understand the foundations of causal inference
• Design rigorous quasi-experimental evaluations
• Apply Difference-in-Differences methods correctly
• Implement Regression Discontinuity Designs effectively
• Assess identification assumptions and threats to validity
• Conduct impact estimation using statistical software
• Interpret causal effects accurately
• Evaluate policy and program interventions rigorously
• Communicate evaluation findings effectively
• Support evidence-based policymaking and decision-making

Module 1: Foundations of Impact Evaluation and Causal Inference

• Understanding causality and counterfactuals
• Correlation versus causation
• Experimental and quasi-experimental designs
• Sources of bias in evaluation
• Internal and external validity
• Exercise: Identifying causal evaluation questions
• Case Study: Common impact evaluation mistakes

Module 2: Introduction to Difference-in-Differences (DiD)

• Conceptual foundations of DiD
• Treatment and comparison groups
• Before-and-after frameworks
• Estimating average treatment effects
• Visualizing intervention impacts
• Practical: Building basic DiD models
• Case Study: Evaluating social protection programs

Module 3: Advanced Difference-in-Differences Applications

• Parallel trends assumption
• Testing identification assumptions
• Multiple treatment periods
• Staggered treatment adoption
• Event study approaches
• Robustness and sensitivity analysis
• Exercise: Advanced DiD modeling workshop
• Case Study: Public sector reform evaluation

Module 4: Introduction to Regression Discontinuity Design (RDD)

• Foundations of cutoff-based evaluation
• Sharp versus fuzzy RDD
• Assignment mechanisms and eligibility thresholds
• Identifying local treatment effects
• Visualization techniques for discontinuities
• Practical: Designing an RDD study
• Case Study: Scholarship and education policy evaluations

Module 5: Advanced Regression Discontinuity Analysis

• Bandwidth selection methods
• Functional form specification
• Manipulation and sorting tests
• Validity diagnostics
• Robustness testing procedures
• Exercise: RDD implementation workshop
• Case Study: Social welfare eligibility programs

Module 6: Statistical Software Implementation

• DiD analysis using R
• DiD analysis using Stata
• RDD analysis using R
• RDD analysis using Stata
• Visualization and reporting techniques
• Practical: End-to-end impact evaluation project
• Case Study: Real-world evaluation datasets

Module 7: Communicating Findings and Policy Applications

• Interpreting causal estimates
• Presenting evidence to policymakers
• Reporting standards for impact evaluations
• Ethical considerations in evaluation
• Emerging trends in quasi-experimental methods
• Capstone Exercise: Complete Impact Evaluation Design
• Case Study: Evidence influencing major policy decisions