Integrated MEL Modules (Advanced Programme) – Evaluation of Public Policies: Empirical Analysis

On September 23, 2025, the first advanced module of the MEL Integrated Programme on Rule of Law Policies took place, bringing together participants from civil society organizations, academia, and the legal field. The session, led by Mimoza Agolli, focused on introducing participants to the main stages of public policy evaluation and to the analytical tools used to assess policy performance and impact.

The training opened with a comprehensive overview of the preparation phase of an evaluation assignment. Agolli explained how to construct a theory of change and a results chain as practical tools to illustrate the causal links between activities, outputs, outcomes, and long-term impacts. She emphasized that a clear and logical results chain helps to articulate how policies are expected to produce change, while also identifying potential risks that may affect the achievement of results. Participants then explored the formulation of evaluation questions and the development of measurable indicators. Agolli underlined that evaluation questions must be framed as testable hypotheses and that the corresponding indicators should follow the SMART approach — Specific, Measurable, Attributable, Realistic, and Targeted. Through guided discussion, participants examined how to select appropriate indicators, identify reliable data sources, and define responsibilities for data collection and analysis, while also considering ethical and resource-related implications.

The second part of the training focused on causal inference and counterfactual analysis in policy evaluation. Agolli introduced participants to the concept of the counterfactual — understanding what would have happened in the absence of a given policy or program — and explained how this principle underpins all impact evaluations. Using practical examples, she elaborated on the use of treatment and control groups, as well as the challenges in ensuring their validity and comparability. Agolli also presented the gold standard of randomized evaluations, along with alternative quasi-experimental methods used when randomization is not possible. These included Difference-in-Differences, Regression Discontinuity Design, and Propensity Score Matching. Participants learned how these approaches can isolate the effects of policies and provide evidence on whether observed outcomes are indeed caused by specific interventions.

The module concluded with group exercises, where participants developed indicators and hypotheses based on a World Bank project evaluation framework. They discussed potential data sources, identified key risks, and proposed analytical methods to measure policy impact.

***

The MEL Integrated Modules take place within the framework of the “Building Partnership on Fundamentals: Empowered CSOs in the EU  accession process” project, with the financial support of the European Union – IPA Civil Society Facility 2021, which is implemented by the European Movement in Albania and in cooperation with the Academy of European Integration and Negotiations (AIEN), Slovak Foreign Policy Association (SFPA) and the Center for Transparency and Freedom of Information (CTFI).