Contribution

“Contribution is not a softer word for impact; it is a more honest account of causality in a crowded system. ”

Contribution is sometimes used when attribution cannot be proved. The programme was present, results improved and the report says it contributed. Without further evidence, the word can become a polite escape from causal discipline. Contribution analysis, associated with John Mayne, offers something more demanding.

It builds and tests a contribution story: the intervention's theory of change is plausible, expected activities and mechanisms occurred, outcomes are observed, alternative explanations are examined and evidence supports a reasonable conclusion about the intervention's role. Complex sustainability change rarely has one cause.

Reduced child labour may depend on household income, school access, law enforcement, social norms, buyer practices and local monitoring. A single programme can be necessary, helpful, marginal or even counterproductive depending on context.

Contribution allows the causal role to be assessed without pretending the programme acted alone. The theory of change is the starting point. It should specify the sequence from action to outcome, the mechanisms expected to produce change, assumptions and external conditions. Training does not improve income by itself.

Farmers must understand, adopt and sustain practices; the practices must affect productivity or cost; markets must reward the result; risks and unintended effects must not erase the gain. Evidence should test each link. Administrative records show delivery. Observation or survey can show adoption. Outcome data show change. Interviews and process evidence help explain mechanism.

Comparison and contextual analysis examine whether other factors provide a stronger explanation. Alternative explanations should be treated as competitors rather than footnotes.

Prices may have risen. A government extension programme may have reached the same farmers. A drought may have reduced gains. Participants may have been better positioned before selection. A contribution claim is strengthened when it explains how these factors interacted with the intervention. The size of contribution can be difficult to quantify.

Some methods estimate shares or effects; others support qualitative judgements such as necessary, substantial, modest or enabling contribution. Precision should not exceed evidence. A well-supported qualitative conclusion can be more useful than an arbitrary percentage.

Contribution can occur through leverage rather than direct delivery. A buyer changes contract terms, enabling suppliers to pay on time. A civil-society organisation shifts local enforcement. A data system makes previously hidden deforestation visible. The causal pathway may operate through decisions made by others. Negative contribution matters too.

Purchasing pressure can contribute to excessive overtime even when the supplier directly schedules workers. A sustainability programme can contribute to exclusion if compliance costs fall on the smallest farms. Causal language should follow evidence consistently, not only when allocating credit. Contribution claims should also identify limits.

The intervention may have worked for farmers with land and labour but not for households below a viability threshold. It may have accelerated change without creating it.

It may depend on public infrastructure outside programme control. These qualifications make replication more realistic. Contribution analysis is strongest when it tests rather than merely narrates the theory of change. It should identify the expected causal chain, gather evidence for each link, examine competing explanations and assess whether the intervention's actions were necessary, supportive or incidental.

Agreement among implementers is not confirmation. Evidence from participants, non-participants, external trends and failed cases can reveal where the story breaks. The conclusion can be proportionate. An organisation may reasonably say it accelerated a policy change, helped remove a barrier or strengthened a capability without claiming sole credit for the final outcome.

This form of modesty does not weaken the achievement.

It separates a defensible role from reputational capture and leaves room for the agency of governments, communities and other actors. The discipline is to show the mechanism, not merely coexistence. What did the intervention change in the behaviour, capability, incentives or conditions of others? Which evidence supports that link?

Contribution earns its place when the causal story is exposed to alternative explanations and still holds.

Practical application

Develop a theory of change with mechanisms, assumptions and external influences. Gather evidence across the causal chain and identify plausible rival explanations. Use process tracing, comparative evidence and stakeholder testimony to test the story. State the strength and nature of contribution without false precision.

Include negative and differential effects, and update the causal account when implementation or context changes. Use an evidence matrix for the theory of change. For each causal link, record expected mechanism, supporting evidence, contradictory evidence, rival explanations and remaining uncertainty. Include cases where the intervention did not work.

The resulting contribution statement should name the organisation's role, the other necessary actors and the conditions under which the contribution appears to hold.

Why it matters

Most sustainability outcomes are co-produced by many actors. Contribution analysis supports credible learning and accountability without demanding exclusive causation or accepting vague association.

Common misconception

Contribution is often treated as a claim that requires less evidence than attribution. It requires different evidence: a tested causal mechanism, alternative explanations and a reasoned account of the intervention's role.

Connections

Theory of Change sets out the expected pathway. Attribution seeks a stronger causal allocation, while Counterfactual estimates the alternative without intervention. Impact describes higher-level effects; Contribution explains the intervention's role in producing them.

A question worth asking

Which link in your contribution story has the weakest evidence, and what alternative explanation is most capable of breaking it?

Selected references

Mayne, J. 2001. Addressing Attribution Through Contribution Analysis: Using Performance Measures Sensibly. Canadian Journal of Program Evaluation 16(1): 1-24. Mayne, J. 2012. Contribution Analysis: Coming of Age? Evaluation 18(3): 270-280. OECD. 2023. Glossary of Key Terms in Evaluation and Results-Based Management for Sustainable Development, Second Edition. Befani, B. and Mayne, J. 2014.

Process Tracing and Contribution Analysis: A Combined Approach to Generative Causal Inference for Impact Evaluation. White, H. and Phillips, D. 2012. Addressing Attribution of Cause and Effect in Small n Impact Evaluations.