---

name: theory-of-change
description: "Build a theory of change and measurement framework that proves your program causes the outcome — not just that you ran it."

/theory-of-change

Activity without theory is how nonprofits end up reporting "we served 847 participants" with no evidence anything changed for any of them. Funders have learned to be skeptical of headcounts. The question they're asking now — and increasingly requiring answers to — is: "What change in the world do your activities cause, for whom, and how do you know?" A theory of change is the explicit causal argument that connects your work to your impact. Without it, your program might be the best thing happening in the community — or it might be coinciding with change that was already coming. This skill forces you to build the causal argument, name what you'd measure to test it, and be honest about what you'd stop doing if the evidence went wrong.

The Specific Change: For Whom, By What Mechanism

• Name the change in the world you are trying to produce. Not "improve health outcomes" — "reduce emergency department readmissions within 30 days for adults with Type 2 diabetes who complete the program."
• For whom: specify the population with enough precision that you could screen for it. Not "vulnerable youth" — "youth aged 14-17 who are in their first year of juvenile justice system involvement and have no prior out-of-home placement."
• By what mechanism: what is the causal pathway? "We believe that [input] produces [behavioral change] because [evidence/theory] which leads to [outcome]." Name each link in the chain. If you can't, you don't have a theory yet.

Inputs and Activities Required

• Inputs: what resources must exist for the program to function? Staff FTEs, funding, facilities, partner relationships, referral pipeline. These are the preconditions — if any input fails, the program fails.
• Activities: what do you actually do? Write each activity as an action with a recipient: "Coach meets weekly with participant for 45 minutes," not "coaching."
• For each activity, ask: is this directly connected to the mechanism? Activities that don't connect to the causal pathway are cost without theory. Name them honestly — they may still be worth doing for non-impact reasons (relationship, trust, access), but they shouldn't be presented as the engine of change.

Short-Term Outcomes: Proof You're On Track

• Short-term outcomes are things that change within 6 months and that, if they're moving, give you confidence the long-term outcome will follow.
• Examples: "Participants can identify 3 personal financial triggers within 60 days of enrollment" as a short-term outcome on the path to "reduced high-cost debt at 18 months."
• For each short-term outcome: how do you measure it? What instrument, what timing, who collects it?
• If a short-term outcome is moving but the long-term outcome isn't, your theory has a broken link. Name where it might break.

Long-Term Outcomes: The Impact Claim

• Long-term outcomes are the changes in the world that matter. These are what funders are buying. They should be measurable, attributable to your program (or at least plausibly connected), and meaningful to the community you serve.
• Attribution is hard. Be honest about what you can and cannot claim. "Correlation between program completion and outcome X at 18 months" is defensible. "Our program caused outcome X" requires a comparison group.
• Name what a comparison group would look like if you had resources to build one.

What You Measure and How

• Build the measurement matrix: outcome, indicator, instrument, frequency, who collects, who analyzes.
• Separate program monitoring (are we doing what we said?) from impact measurement (are participants changing?). Both are necessary; they're not the same.
• Name the minimum viable measurement system you could implement with current resources. A data system you can't staff is a liability, not an asset.

What You'd Stop Doing If Metrics Are Wrong

• If your short-term outcomes aren't moving after 12 months, what specifically would you change or stop?
• This is the honesty test for your theory of change. If the answer is "nothing, we'd keep going," the metrics aren't connected to decisions. Real theories of change are falsifiable.
• Name the kill signal: "If fewer than 50% of participants show [short-term outcome] at 6 months, we will [investigate mechanism / modify dosage / pause enrollment and diagnose]."

Rules

1. The change must be specific enough to measure. If it can't be measured, it can't be claimed.
2. The mechanism is a causal argument, not a description of activities.
3. Short-term outcomes must be on the causal pathway to long-term outcomes. They are not proxies for activity.
4. The measurement plan must be staffable with current resources.
5. The theory must be falsifiable. Name the signal that would make you stop.
6. Attribution claims must be honest. "Associated with" is not the same as "caused."

The output is a funder-ready theory of change document that shows the causal argument, the measurement system, and the honest conditions under which you'd revise or stop the program.