Skip to content

guide

Evaluating peacebuilding programmes: how to show what worked

Peace is hard to prove, because you are counting what did not happen. How to set measures, gather evidence, and report what a peace programme achieved.

Takwimu editorial team, Editorial team7 min read

Evaluating peacebuilding means showing what changed and what part your work played in it. It's hard, since you're often counting something that didn't happen. The GSDRC topic guide says that monitoring should be built in from the design stage, and should track unintended effects. A simple record of what was done, for whom, and what followed is better than a grand claim.

Key points:

  • Decide what success looks like before you start.
  • Use a baseline and repeat your measures.
  • Claim contribution, and be careful with attribution.
  • Report failures as well as gains.

Why is it hard?

Peace has many causes. If a county stayed calm, was it your dialogue, the rain, the security deployment or the candidates? Often you can't separate these.

Time is another problem. Trust changes slowly, and funding cycles are short. Local people may also tell visitors what they want to hear.

What should you measure first?

Start with activities: meetings, people trained, disputes handled. Those are easy, but they aren't results. Move on to outcomes: did behaviour change? Did a dispute stay settled? Did people feel safer?

Then look at wider change: fewer incidents, more trust, more people using shared markets.

How do you set a baseline?

Measure before you begin. Ask residents about trust and safety. Count incidents from the last year. Note what's already being done by others. Without a baseline, you can't show change.

If you're late, ask people to recall how things were, and say that you did so.

What measures are useful?

Choose a few that you can defend. Time from warning to response. Disputes settled, and whether settlements held after six months. Trust in local bodies. Fear of violence. Returnees. School days lost. Cross-community trade.

Add measures from the cohesion barometer if it fits.

What is attribution versus contribution?

Attribution says that your programme caused a change. Contribution says that it helped, along with other things. In peace work, contribution is usually the fair claim. Explain the steps by which you think your work helped, and show evidence for each.

How do you gather evidence?

Combine surveys, interviews, group discussions, incident logs and official records. Ask people who didn't take part, and those who disagree. Compare with similar areas where you didn't work, if you can.

What about unintended effects?

The guide recommends tracking them from the start. Did your project raise tension? Did it favour one group? Did allowances attract the wrong people? Peace Insight reports claims of peace committee members provoking conflict to get paid, which is why you need to look.

Who should do the evaluation?

Your own team can track progress. An outside reviewer adds credibility for big decisions. Communities should take part too, by saying what they think has changed.

How do you protect people?

Don't record who said what in a tense place. Share only totals. Ask for consent. Consider whether an evaluation itself could raise tension.

What is an initiative record?

It's a running log of each initiative: what was planned, what was done, by whom, when, and what followed. TI's cohesion pack has a report by this name. It suits commissions, which are asked what they've done with public money.

A clear record beats a glossy report. An auditor or a funder can follow it line by line.

How should you use the data while the work is running?

Review it each quarter. The guide says there must be flexibility to adapt the project in response to findings. If something isn't working, change it. Don't wait for the final report.

How do you report to funders?

Say what you did, what changed, and what you can't prove. Show numbers with their source. Include what failed. Good funders trust a team that admits its limits.

How do you report to the public?

Keep it short and clear. What was the problem? What was done? What's changed? What's next? Share results with the communities involved first.

What are common mistakes?

Counting meetings as success. Measuring only at the end. Using a measure that shifts with the news. Ignoring those who were left out. Claiming credit for peace that had many causes.

What can a small organisation do?

Keep a simple log. Ask ten people the same five questions every quarter. Note what you did and what happened. Take photographs with permission. That's enough to learn from, and to tell a true story.

How does this help the next project?

Evaluation is about learning. Share what you've found with others in the county, so that they don't repeat your mistakes. Keep your records, because a future team will need them.

What does a good indicator look like?

It's clear, it can be measured and it fits your work. "Improved cohesion" isn't an indicator. "Share of residents in three wards who would accept a neighbour from any community" is.

Pick one or two for each goal. Say where the data will come from, who collects it and how often. If nobody can say, drop it.

What does a worked example look like?

A peace committee wants to reduce grazing disputes in two wards. Its baseline is 14 disputes last dry season, of which five turned violent. It trains mediators and revives a grazing agreement.

The next dry season has 11 disputes, of which one turned violent, and settlements held after six months in nine of them. Rain was similar. A survey shows more trust in the committee. The team reports a likely contribution, and it notes that a new borehole helped too.

How do you handle a bad result?

Report it. If violence rose in an area where you worked, ask why. You may have been working in the hardest place, or your project may have made things worse. Both are worth knowing, and a funder who hears it from you will trust you more.

How do you compare areas?

If you can, pick similar wards where the programme didn't run. That won't prove cause, but it helps. If a whole county improved, your project may not explain it. ENACT's account of 2022 shows how many factors can lie behind a calm election.

Where TI fits

Takwimu Intelligence (TI) prepares The initiative record, a report that follows each initiative from plan to outcome, drawing on surveys and field reports.

To see it on your own work, book a walk-through.

Takwimu Intelligence (TI) is a conflict early warning system for Kenya.

Sources

  1. 1

    Conflict sensitivity: topic guide · Governance and Social Development Resource Centre, 2014Opened 5 November 2026.

  2. 2

    Committed to peace, or creating conflict? The case of Kenya's local peacebuilding committees · Peace Insight, December 2014Opened 5 November 2026.

  3. 3

    Muted violence in Kenya's 2022 elections masked seething dissent · ENACT Observer, 24 April 2023Opened 5 November 2026.

More from the library