Photo by: Kazi Ashraful Alam

Our four-pronged strategy

Photo by: NRSP

To provide evidence about the impact of development projects, the Center conducts rigorous, independent and neutral evaluations.

  • Experimental evaluation methods such as randomized control trials
  • Non experimental evaluation methods such as difference-in-differences, instrumental variables, matching and propensity score matching, regression discontinuity design
  • Mixed methods

Photo by: A. Avdeenko

In addition to scientific research and impact evaluations, the Center assists its partners in increasing the social impact of interventions, programs, policies and products.

C4ED provides a wide bundle of advisory and practical support areas, such as:

  • Monitoring and evaluation support
  • Survey methodology, questionnaire design, data collection, and information and communications technology
  • Statistical decision-making and big data analysis
  • Management of information systems, data and knowledge management
  • Process and implementation evaluation

Recycle!Up Ghana

We set up pilots and subsequently support the upscaling of successful innovations, thus directly impacting on the lives of the poor and vulnerable.

  • Exploration of effective solutions to fill evidence gaps and to foster sustainable development
  • Program design and product development
  • Set up of pilot and scaling-up support

Current pilots

  • Pre-school nutrition project, Egypt
  • Street-children program, Ecuador
  • Recycling and waste management program, Ghana

Photo by: E. Camargo

We leverage evidence about the impact of specific development interventions by disseminating research findings and recommendations to implementing organizations, policy-makers, civil society and the scientific community.

Our expertise

C4ED supports development projects in a broad range of sectors, drawing on solid expertise that encompasses:

  • Education access, quality and attainment
  • Child labor prevention, children and youth rights
  • Health, nutrition, food security
  • Community mobilization, peace-building, conflict and migration, empowerment
  • Labor market, decent work, entrepreneurship
  • Social protection, cash transfers, insurance, pensions
  • Microfinance, credit, saving schemes, financial sector development
  • Environment, climate change, disaster resilience
  • Rural development, agriculture, smallholder farmers
  • Experimental evaluation methods such as randomized control trials
  • Non-experimental evaluation methods such as difference-in- differences, instrumental variables, matching and propensity score matching, regression discontinuity design
  • Mixed methods and qualitative designs
  • Development of methods for imperfect situations: nonparametric IV identification with mediation, attrition, missing data, …
  • Heterogeneity analysis (quantile treatment effect) and mediation analysis
  • Monitoring and evaluation support
  • Survey methodology, questionnaire design, data collection, and information and communications technology
  • Statistical decision-making and big data analysis
  • Management of information systems, data and knowledge management
  • Process and implementation evaluation

To achieve excellence in impact evaluations, the Center uses the most advanced techniques and methods from social sciences, econometrics, statistics and survey methodology. In addition to evaluations and applied research the Center also has a strong track record in developing new methods and impact evaluation methodologies, particularly regarding quantitative and econometric techniques.

Impact evaluation methodology has seen a tremendous change in the last decade. With the advent of ever increasing computational performance, network computing and large data sets as well as new identification designs, new econometric methods have become available and implementable. Whereas quantitative impact evaluations in the past were based on simple linear regression models and least squares estimators, new nonparametric methods permit for a much wider analysis of treatment effect heterogeneity, advanced discontinuity designs, mediation analysis, imputations for survey non- response, missing data and attrition. As an example, mediation analysis helps discerning the causal chains of an intervention, i.e. the channels through which change might be brought about by the program or intervention. Disentangling these channels and measuring their magnitude is important in order to learn more about the intervention mechanisms, and derive suggestions and indications on how to improve overall and cost effectiveness of an intervention.

Besides statistical and econometric methodology, extensive research is being conducted in survey methodology and questionnaire design, embedding new Information and Communications Technology in data collection, field experiments and behavioral economics, mixed methods and qualitative approaches, analysis with big data sets and data driven decision making. These methodological contributions have been published in leading scientific journals in statistics and econometrics, including the Journal of the American Statistical Association, Journal of Business and Economic Statistics, Journal of Econometrics and the Journal of Royal Statistical Society.