Home/Guides/Theory of Change

Reference guide

Theory of Change: what it is, what it is for and how to build it

What is a Theory of Change?

A Theory of Change is the explanation of how and why an intervention produces the change it promises. It chains resources, activities, outputs, outcomes and impact, and makes explicit the assumptions that hold each link. It serves to design the program, to define what to measure and to know what failed when something does not work.

Its value is in the assumptions, not in the boxes. Anyone can draw a chain that goes from training to employment. The hard part is writing what has to be true for that leap to happen: that there are vacancies in the local market, that people can keep up their attendance, that the certification is recognized by those who hire.

When the result does not appear, the Theory of Change is what makes it possible to tell two very different diagnoses apart: the program was run badly, or the program was run well and the assumption was false. Without it, the conversation stays at something not having worked.

Theory of Change and logical framework are not the same

CriterionTheory of ChangeLogical framework
What it answersWhy this is expected to workWhat we are going to do and how we verify it
FormatNarrative plus causal diagramMatrix
Treats assumptionsAs the center of the analysisAs a column at the end
Allows alternative pathwaysYes, and makes them explicitNo, assumes a single chain
When it is usedDesign and strategic reviewProject management and reporting

They do not compete. The logical framework is derived from the Theory of Change: first the causal logic is reasoned out, then it is organized in the matrix the funder asks for. Building the logical framework without the theory behind it produces matrices that are filled in correctly and explain nothing.

How to build it in six steps

  1. Define the problem. In a single sentence, in terms of the affected population. "Young people in the area do not access formal employment" is a problem. "There is no training program" is not: it is the absence of your solution.
  2. Define the long-term impact. The ultimate change the program contributes to. Contributes, not produces: almost no social program is the sole cause of its impact.
  3. Work backwards to the intermediate outcomes. Ask what has to happen beforehand for that impact to be possible. Repeat backwards until you reach something the program does control. This reverse order is what avoids the most common mistake, which is starting from the activities that were already being carried out.
  4. Connect outputs and activities. Link each outcome with the outputs that make it possible and with the activities that generate them. If an activity does not connect to any outcome, it is surplus or a link is missing.
  5. Make the assumptions explicit. On each arrow, write what has to be true for that leap to work. Mark which ones are outside the program's control: those are the ones to monitor as risks.
  6. Validate with those who implement. Check the chain against the field team. If they do not recognize their work in the diagram, the theory describes a different program from the one being run.

Quick test. Ask someone who did not take part in the design to read the Theory of Change and explain what the program does. If they cannot, the problem is not with that person.

The mistakes we see most often

  • Confusing output with outcome. "500 people trained" is an output. The outcome is what changes in those people afterwards. When the theory stops at outputs, the MEL system measures activity and not change.
  • Assumptions written as good wishes. "There is political will" is not a monitorable assumption. "The municipality maintains the program's budget allocation in the following fiscal year" is.
  • Causal leaps without an intermediate link. From awareness-raising to behavior change there are several steps. When the chain omits them, the evaluation cannot explain where it broke.
  • A theory that no one on the team can explain. If it lives in an annex of the proposal and no one opens it again, it is not fulfilling its function. It is the clearest symptom that it was built for the funder and not for the program.
  • Never reviewing it. A Theory of Change that stays identical after three years of implementation probably means the evidence is not entering the decision cycle.

Frequently asked questions

Next steps. With the Theory of Change ready, the next step is to derive the indicators: how to design them so they can be measured. To see where it fits within the complete system, read what a MEL system is.

Does your Theory of Change guide decisions or fill an annex?

Redesigning it is one of the four-to-six-week scopes that changes fastest how a MEL system works.