Theory of change is the most misused phrase in the nonprofit world.

A theory is a falsifiable claim about cause and effect. 'We run after-school tutoring three days a week' is not a theory. It is a schedule.

This distinction sounds academic until you watch it decide an award. We have read a large number of theory-of-change sections over the past two years, and the majority of them are activity descriptions wearing a logic model's clothing: inputs, activities, outputs, outcomes, impact, arranged left to right with arrows that assert causation rather than argue for it.

The theory is the sentence underneath the arrows. Something like: we believe sustained small-group instruction at this dosage raises ninth-grade math proficiency, because the binding constraint for our students is practice time with feedback, not motivation.

Read that again and notice what it does. It names a mechanism. It identifies what it thinks the real constraint is. And it implicitly rejects a competing explanation — motivation — which means someone could disagree with it.

That sentence can be wrong. That is the point. A funder can evaluate it, ask what evidence supports the causal claim, and check whether your measurement plan would actually detect failure if the mechanism turned out to be something else.

Activity descriptions ask the funder to trust you. Theories ask them to evaluate you. The second is a stronger position, and it is stronger in a specific, practical way: it survives a change in program officer. Trust is held by a person. An argument is held by the file.

A practical test we use with teams: if your theory of change would still read as true after swapping in a completely different organization's programs, it is not a theory yet. It is a values statement. Values statements belong in the application — just not in that section.

A second test: can you name the condition under which you would conclude the theory is wrong? Not 'the program failed' — the specific observation that would tell you the mechanism was misidentified. If tutoring dosage rises, attendance holds, and proficiency does not move, what would you then believe? Teams that can answer this write measurement plans that are worth reading.

The most common objection is that funders do not actually want this — that a crisp falsifiable claim invites scrutiny an activity description avoids. In our experience the opposite is true at any funder sophisticated enough to matter. Program officers read dozens of logic models per cycle and they blur together completely. A stated mechanism is memorable, and it gives the officer something to argue with internally, which is what advocacy inside a foundation actually looks like.

There is also an internal benefit that has nothing to do with fundraising. An organization with a written, falsifiable theory can make program decisions against it. When a new opportunity arrives, the question stops being 'can we do this?' — the answer is almost always yes, and that is how mission drift happens — and becomes 'does this operate through the mechanism we believe in?' That question has a no in it.

Where do theories come from? Usually from the front line, not from the strategy retreat. The tutoring coordinator who says the kids are not unmotivated, they just have nowhere quiet to work and no one to check their answers — that is a causal hypothesis, stated in plain language by someone with direct evidence. Most organizations already have their theory. It is sitting in a program staff member's head, unwritten, because nobody asked them in those terms.

Getting it written takes about ninety minutes and three questions. What is the constraint we think is binding? Why that one and not the obvious alternative? What would we see if we were wrong?

Write the answers as prose, not as a diagram. Diagrams hide the verbs, and the verbs are where the theory lives.