We clarify what changed, how it changed, and why it changed. This includes explicit contribution narratives linked to evidence, counterfactual estimation where feasible, triangulated causal stories, and displacement scanning so benefits are not simply shifting between groups or geographies.
Contribution Analysis
A causal story your board will sign — and your auditor will not challenge.
Causation is hard. Confusing contribution with attribution is easy. We make the difference visible so your results survive scrutiny.
What you get
- Contribution narratives linked to evidence, not assertion
- Counterfactual estimation where feasible (synthetic control, comparison groups, instrumental variables)
- Triangulated causal stories using qualitative, quantitative, and administrative data
- Displacement scanning so benefits are not simply shifting between groups or geographies
How we work
-
State the claim
What change do you claim? What change do you not claim? The boundary of the claim drives the design.
-
Triangulate
Three independent sources of evidence, each with explicit limitations.
-
Estimate the counterfactual
Whatever you call it — contribution, attribution, counterfactual — make the assumptions visible.
-
Scan for displacement
If X went up, did Y go down for someone else, somewhere else, or later?
Why Merstat
- Contribution narratives that survive technical audit
- Displacement checks documented, not skipped
- Limitations and assumptions in the executive summary, not buried in the appendix
Frequently asked questions
Is attribution analysis the same as attribution modelling?
No. Marketing-style multi-touch attribution modelling is not appropriate for development evaluation. We use contribution analysis in the Mayne sense — explicit causal narratives, plausible counterfactuals, and triangulated evidence.
Ready to talk?
Tell us about your context, your objectives, and your timeline. We will reply within one working day with a proposed approach and a partner call.
