Skip to content


Impact Evaluation of the SADA Millennium Villages Project in Northern Ghana Annex A: Preliminary Statistical Analysis

The independent evaluation of the Millennium Village Project (MVP) in northern Ghana is based on quantitative and qualitative data that was collected in three stages of the project: baseline, midline and endline

Edoardo Masset

This report provides the statistical analysis, based on the published Analysis Plan. Some of the analysis is also presented in the main report, whereas this annexes contains additional details. The report summarises:

  1. The evaluation design and the stratification of the comparison groups by distance to the MV villages – plus the statistical models used to assess impact.
  2. The datasets and an overall assessment of the quality of the data and their suitability for a difference-in-difference analysis.
  3. The take-up rates for most activities promoted by MVP, where we analyse the characteristics of households and individuals taking part in project activities and the beneficiaries.
  4. An assessment of the impact of MVP on its final outcomes, the Millennium Development Goals (MDGs), using a ‘dashboard approach’.
  5. An assessment of the impact of MVP on an aggregate measure of poverty using the multidimensional Oxford deprivation index.
  6. A discussion of the theory of change of MVP, where we illustrate the challenge of explaining the project impact in these circumstances, and a justification for the evaluation strategy adopted.
  7. The impact of MVP on expenditure, income and savings where we attempt to reconcile the impact observed on these aggregates within a unified conceptual framework.
  8. Detailed impacts by sector (in agriculture, health and education) on non-MDG outcomes and other intermediate indicators.
  9. A discussion of impacts not covered in other sections, such as migration, water access and vulnerability.

Read the full report, executive summary, briefing papers and other annexes on our project page.