Contribution analysis has gained traction as a useful approach for critically assessing how an intervention contributes to complex change processes. A recently developed guidance note offers the evaluation community insights into what quality contribution analysis looks like in practice.
By:
As programs and strategies operate in situations where change depends on a constellation of actors and factors in dynamic systems, evaluators are often faced with a challenging question: How can we meaningfully and credibly assess an intervention’s contribution to complex change processes?
Contribution analysis (CA) has become a widely used approach to address this evaluative challenge, often lauded for its ability to unpack complexity and explain how and why change happens as well as the relative role an intervention plays. But greater adoption of the approach has resulted in a variety of practices, making it difficult for practitioners and commissioners to understand what to expect from using CA and what makes an application high quality.
In 2025 we convened a community of practice of experienced CA practitioners through a series of virtual and in-person workshops to contribute to the evolving debate around these important questions. Quality Guidance for Contribution Analysis in Practice is the culmination of that process and sets out the core features and conditions required for a high-quality application of CA.
What sets CA apart from other complexity-aware approaches to evaluation?
The guidance highlights several characteristics that distinguish CA from other approaches, as well as the core features that determine whether a CA application is high quality. We outline a few of the most distinctive features in this blog.
Like many theory-based evaluation approaches, CA has an analytic backbone based on a theory of change. As such, the robustness of the theory used for CA is a critical component to a high-quality application. A quality theory of change for CA includes two distinct and interrelated concepts: articulation of how change happened and an explanation of how the “intervention” (ranging from a specific activity to a bigger project, program, or strategy) contributes to this articulated change process. This sequencing within CA, starting with the process of change first and then developing the intervention’s theory of change, supports high-quality CA.
The way this theory is used in CA differs from other theory-based approaches. Instead of using the initial theory as a hypothesis to test against, CA uses an abductive process, moving back and forth between multiple sources of evidence, theory, and alternative explanations to develop and refine the theory over time. By the end of a CA application, the initial theory of change has become the “explanation of change” based on the evidence.
Quality within CA is also enhanced through the meaningful inclusion of diverse perspectives during each step of the process, particularly in defining the theory of change at the outset and refining contribution claims. This means that high-quality CA often incorporates participatory processes to evidence gathering and sensemaking, especially with those directly involved with or affected by the intervention or situation of interest.
The aim of CA is to generate a plausible and meaningful narrative of complex change processes that explains the role that an intervention played, often referred to as the contribution story. Arriving at this final narrative, however, is not a linear process. As such, another distinguishing characteristic of high-quality CA is that the steps of CA are applied flexibly and often repeated over time with feedback loops between steps to formulate a strong contribution story.
Closing the gap in IEG’s own practice and beyond
Building this kind of strong contribution story is at the core of IEG’s work. The interventions and programs IEG evaluates are cross-cutting, global challenges, implemented in environments where the World Bank Group is one actor among many. In these settings, understanding causal processes, unpacking what drove outcomes, and capturing meaningful contributions are among the hardest questions in development evaluation. Advancing IEG’s ability to answer them rigorously has been a strategic priority.
CA is therefore a particularly relevant approach for IEG evaluations. Yet, an internal assessment of IEG’s evaluations found CA to be underused or used inconsistently. IEG evaluators increasingly turn to CA in Country Program Evaluation, for example, in the case of Morocco and Egypt, where capturing the Bank Group’s contribution to country outcomes is at the crux of the exercise. Yet, a common understanding of what high-quality CA entails in practice has been missing. Closing this gap became a priority for IEG’s Methods Team. This guidance is a concrete step towards building the practical foundations for CA to be applied rigorously and consistently in IEG’s own evaluations.
Beyond IEG, the Bank Group is focused more and more on telling a clear, evidence-based story about results—showing how its support for policy reforms, stronger institutions, and better-functioning markets helps countries address non-financial barriers and achieve lasting development outcomes. The 2023 scorecard introduced Results Narratives to that effect, which are aiming to use CA. This guidance, thus, aims to influence practice beyond IEG, in the Bank Group and across the evaluation and international development fields.
How to use this guidance
The guidance note is organized around the six steps of CA as originally proposed by John Mayne and further refined in his subsequent work. The intention is not to set out a road map for how to apply each step in a quality manner, since in the true spirit of CA, each application must be adapted to the particular evaluation context. Rather, the hope is that by setting out what high-quality CA looks like at each step, evaluators may use this guidance note to reflect on their current practice and begin identifying and adopting measures to improve their application over time. In this context, the guidance is intended to promote greater consistency in CA application across the evaluation community and to strengthen the collective use of CA.
Drew Koleros is a Principal Researcher at Mathematica with over twenty years of experience in designing and delivering mixed-methods evaluations and program monitoring, evaluation and learning (MEL) systems for both small scale and large scale social and economic development projects. He brings particular experience and expertise in using theory-based approaches that integrate complexity concepts and systems thinking into program and evaluation design processes, particularly using Contribution Analysis. His most recent work focuses supporting foundation strategies and portfolio programs aimed at catalyzing systems change to address complex societal problems from safe and affordable housing to youth employment to advancing health equity.
Giel Ton is a Research Fellow at IDS, at the University of Sussex, and co-leads the Centre for Development Impact (CDI). He specialises in research and evaluation of smallholder inclusion in agricultural value chains. He has published widely about the methodological challenges of impact evaluation of development interventions.
Thomas Delahais has been an evaluator for the past 21 years, first at Euréval and then at Quadrant Conseil. Quadrant Conseil is a worker-owned cooperative company which Thomas co-founded in 2013. Most of his work involves managing or evaluating programmes or policies, as well as providing strategic advice to French, European or international organisations. He also trains civil servants in evaluation and policy design and provides long-term support to organisations in setting up their evaluation systems.
Marina Apgar is a Research Fellow at the Institute of Development Studies, at the University of Sussex, and is currently leading the Research Cluster on Participation, Inclusion and Social Change at IDS. She also co-leads the Centre for Development Impact (CDI) with Giel Ton. She specialises in combining theory-based evaluation with participatory action research to support evidence co-production processes. She has convened a number of professional development courses on: Contribution Analysis for Impact Evaluation; Assessing the Strength of Evidence; Participatory Monitoring and Evaluation for Learning, via the IDS and the UK Evaluation Society. She is a member of the Causal Pathways initiative and the Inclusive Rigour Co-Lab, both networks aiming to support the development and use of appropriate forms of evidence production for more impactful and equitable development programming.
Diana is a Data Scientist in IEG’s Methods Advisory Function and the managing editor of IEG’s Methods Paper Series.
Estelle Raimondo is Head of Evaluation Methods. She advises and trains teams on a wide range of evaluation methodologies, spearheads innovations and leads the adoption of data science and AI for evaluation practice.
