Social Policy Evaluation using Stata


Social policies are of the utmost importance in contemporaneous societies. They consist of legislation and activities concerning governmental responses to specific human and social needs in many areas of intervention, such as health care, human services, criminal justice, inequality, education, migration and labour.

However, only in recent years the evaluation of social policies’ effects on targeted individuals, and on larger sets of a given population has become central in the policymakers’ agenda. Indeed, an unprecedented availability of new information collected by specialized surveys and administrative registries at individual, regional, and national level, has expanded the opportunity for researchers and policymakers to investigate social-related phenomena and policies in an increasingly finer detail. Moreover, today information technology based on larger computing power and storage, big data management, along with recent developments in causal modelling, make the analysis and evaluation of social policies (at any level) easier to carry out than it has been previously.

Social policy evaluation can have either an ex-ante or an ex-post nature, and can be either qualitative, or quantitative. This course focuses on the ex-post and quantitative side of the coin. Indeed, by making use of the most recent counterfactual statistical and econometric techniques, participants will become knowledgeable of the essential tools, both theoretical and applied, for a proper application of modern micro-econometric methods for social policy evaluation using counterfactual modelling in Stata.

The course will cover various counterfactual methods, such as, Regression adjustment (parametric and nonparametric), Matching (on covariates and on propensity score), Selection models, and Difference-in-differences. Moreover, it will provide the participant with Stata solutions to estimate social policy in the presence of indirect effects, and more specifically, neighbourhood effects.

In line with the general philosophy of our training courses, the lessons will be very interactive and will have mostly applied content. They will include numerous empirical applications on real social policy datasets. Participants will be able to experiment with the techniques learned through exercises performed by their own calculation stations under the guidance of the instructor.

After attending the course, the participant will be able to set up and manage a correct social policy evaluation design under observable and unobservable selection on their own, thus mastering the identification of the policy framework, the collection and management of suitable datasets, the use of the appropriate econometric methods, and the interpretation of results. Finally, the instructor will assist participants in setting-up their own personal policy evaluation case study.



Day 1

Session 1: Introduction to social policy evaluation

  • Social policy: definitions, aims, problems
  • The social policy “implementation cycle”: an overview
  • Ex-ante and ex-post social policy evaluation
  • Social policy effects: direct, indirect, and total effects
  • Social policy evaluation in the presence of neighbourhood effects
  • Randomized vs. nonrandomized social experiments
  • Setting-up ex-post social policies evaluation: some guidelines

Session 2: Counterfactual social policy evaluation

  • An introduction to counterfactual causality
  • Statistical background and notation
  • The selection problem: observable and unobservable selection
  • Regression Adjustment and Matching
  • The Stata treatment-effects estimation package teffects
  • Application on real data of teffects subcommands

Day 2

Session 1: Social policy evaluation under unobservable selection

  • The problem of unobservable selection
  • Estimating selection models in Stata
  • Difference-in-differences (DID) with applications

Session 2: Estimating social policy “neighbourhood” effects

  • The problem of indirect effects in social policy evaluation
  • Neighbourhood and mediation effects
  • Solutions via the sem and ntreatreg Stata commands

Learning Ratio

This course has a learning ratio of approximately 30% Theory, 30% Demonstration and 40% Practical


Principal texts for pre-course reading:

  • Wooldridge, J.M. (2010). Econometric Analysis of cross section and panel data. Chapter 21. Cambridge: MIT Press.
  • Cameron, A.C., & Trivedi P.K. (2005). Microeconometrics: Methods and Applications. Chapter 25. Cambridge: Cambridge University Press.
  • Cerulli, G. (2012), An Assessment of the Econometric Methods for Program Evaluation and a Proposal to Extend the Difference-In-Differences estimator to dynamic treatment, in: Econometrics: New Developments, Nova Publishers, New York.

Principal texts for post-course reading:

  • Cerulli, G. (2015), Econometric Evaluation of Socio-Economic Programs: Theory and Applications, Springer.

Daily Timetable

(subject to minor changes)

Time    Session / Description

09:00 - 09:20 Registration

09:30 - 11:00 Session 1a

11:00 - 11:15 Tea/coffee break

11:15 - 12:45 Session 1b

12:45 - 14:00 Lunch

14:00 - 15:15 Session 2a

15:15 - 15:30 Tea/coffee break (Feedback Session)

15:30-17:00 Session 2b


More Information / Apply Now

Start Date

End Date


Professional training


Regular fees: 430 - 1295 GBP



  • Knowledge of basic econometrics: notion of conditional expectation and related properties; point and interval estimation; regression model and related properties; probit and logit regression.
  • Basic knowledge of the Stata software

Dublin , Ireland

Start Date:

End Date:





Professional training