Policy and programme impact evaluation
We help policy and programme teams design evaluations before rollout and assess interventions using existing data. We start with the outcome, the intervention and what would make a credible comparison.
Before and after rollout
Before implementation
We define who the programme reaches, the outcomes it is meant to change and how those outcomes can be measured. A phased or randomised rollout may help if programme operations allow it. We agree a feasible comparison and data collection plan before the intervention starts.
After implementation
We review timing, eligibility, outcome records and how people or sites entered the programme. If the data support it, we choose a quasi-experimental comparison. For example, a difference-in-differences design may fit a staggered rollout with a credible comparison group and suitable pre-intervention data.
What you receive
An agreed evaluation scope can include a design and data requirements, an analysis plan, documented checks, estimates with uncertainty, and a report that explains what the findings can support. We discuss reproducible outputs and handover when scoping the work.
See a worked staggered rollout with synthetic data, a pre-trends chart and a cohort-by-cohort calculation. Causal Capybara shows how VCESR makes study design, diagnostics and analysis scripts inspectable.
What if there is no credible comparison group?
We can describe what changed, assess the implementation and identify evidence gaps. A trend before and after a programme rarely isolates its impact on its own. If a causal estimate is unsupported, the report should say so and explain what new data or a future rollout could resolve.
To start, share the intervention and its dates, the outcomes that matter and a description of the available data. A first message does not need individual records.
Discuss an evaluation