Theory of change, simply explained

A theory of change is a map of how and why a programme is expected to bring about the change it wants. It sets out the steps from activities to long-term change and, crucially, the assumptions that must hold for each step to lead to the next.

It is usually built backwards. You start from the long-term change, ask what must be true for it to happen, then what must be true for that, until you arrive at the activities you will run. Every activity then has a reason, and every reason can be checked.

This guide is for NGO teams writing proposals and CSR teams reviewing them. It explains what a theory of change is, how it differs from a results chain and a logframe, how to build one, and how funders use it, with a worked example from girls’ secondary education.

What a theory of change is

The OECD’s glossary of evaluation terms describes a theory of change as the way an intervention is expected to achieve change: how people understand change to happen in a given context, including the assumptions about the causal links between inputs, activities and results, and often the evidence and risks for each link.

In practice, a theory of change has six parts:

• The problem: whose situation should change, and why it hasn’t yet. This comes from your needs assessment.
• The long-term change: what will be different for people if the programme succeeds, usually over five years or more.
• The pathway: the earlier changes, or preconditions, that must happen first, in order.
• The activities: what you will actually do to bring about the early changes.
• The assumptions: the conditions outside your control that must hold for each step to lead to the next.
• The evidence: why you believe each link will work, from research, past results or the community.

It is often drawn as a diagram, but the reasoning matters more than the drawing. A one-page diagram with a short written explanation beats a wall chart with forty boxes and arrows going everywhere.

How it differs from a results chain and a logframe

People mix these three up, and funders sometimes use the words loosely. They are related, but they answer different questions.

A useful way to see it: the theory of change is the argument, the results chain is the outline, and the logframe is the plan for checking the facts. Our guide to the results chain and logframe covers the other two in detail.

How to build one, step by step

1. Agree the long-term change. Write it as a change for named people in a named place: “girls in these 25 villages complete Class 10”, not “empowered girls”.
2. Map backwards. Ask “what must happen before this?” and write each answer as an outcome, a change rather than an activity. Keep asking until you reach changes your activities can bring about directly.
3. Check the pathway. Read it forwards, asking “and then what?” at each step. If a step doesn’t lead to the next without a leap of faith, something is missing.
4. Write down the assumptions behind each link: about people’s behaviour, institutions, the season, the law or other programmes.
5. Choose the activities that will produce the earliest changes. Look for what you can do well, and what others, including government, already do.
6. Add indicators for the main outcomes, so you can tell whether each step is happening. See choosing indicators.
7. Write the narrative: a few paragraphs that explain the logic in plain words, for people who won’t read the diagram.

Do this in a workshop, not at a desk. Bring in the people the programme serves, frontline workers such as teachers, ASHAs and anganwadi workers, local officials and your own field staff. They know which steps are realistic.

Assumptions, evidence and risks

An assumption is something that must be true for one step to lead to the next, but which the programme doesn’t control. Many programmes that disappoint do so not because their activities weren’t delivered, but because an assumption was wrong: trained teachers were transferred, families couldn’t spare the time, or a scheme the project relied on stopped paying.

For each important assumption, note three things:

• How confident you are, and why. Is it backed by your needs assessment, by research, by a pilot, or only by hope?
• What you will watch to know if it is failing.
• What you will do if it fails: change the activity, add one, or work with someone who can deal with it.

Then test the weakest link first. Ask your team which assumption, if wrong, would break the whole chain, and check that one early, often with a small pilot in a few villages before you scale up.

A worked example: girls’ secondary education

Two assumptions changed the design. Disha first checked whether a state scheme already gave bicycles to girls in Class 9; where it did, Disha organised the travel groups and safe routes rather than buying bicycles. And because early marriage was a reason in some families, it added sessions with the panchayat and self-help groups, rather than assuming safe travel alone would keep every girl in school.

How funders and CSR teams use a theory of change

A CSR team reading a proposal uses the theory of change to answer one question: does the plan make sense? Good reviewers ask:

• Is the problem clear, and based on evidence from this place?
• Does each step lead to the next? They ask “and then what?” after each activity.
• Are the assumptions named, and is someone watching them?
• Did the community shape it?
• Is the long-term change realistic for the money and the time, and does the proposal claim only its own contribution?

It also helps the company afterwards. The annual action plan must describe each project’s monitoring and reporting mechanism (Rule 5(2) of the CSR Rules), and the theory of change tells you what to monitor. If the company is large enough to need an independent impact assessment, a good assessor will test the links in the chain, so a clear theory of change at the start makes the assessment far more useful. For NGOs, it is the backbone of a strong CSR proposal.

Common mistakes

• Boxes of activities with arrows between them. “Training” leading to “awareness” leading to “empowerment” isn’t a theory; it is a list.
• Skipping assumptions, or writing only safe ones such as “funding continues”.
• Claiming the goal. A reading programme in 40 schools contributes to better learning in a district; it doesn’t deliver it alone.
• Built by a consultant, for the funder. If field staff and the community don’t recognise it, it won’t guide the work.
• Never updated. A theory of change is a hypothesis. When monitoring shows a step isn’t happening, change the theory, not just the report.
• Ignoring harm. Ask what could go wrong for people as well as for the project: a girl teased for cycling, or a family stretched by a new demand on its time.

A one-page template

Revisit it at least once a year, with partners and, where you can, the community. A theory of change that has been revised in the light of evidence is a sign of a programme that learns.