Theory of Change vs Logic Model: One Nonprofit Program Mapped Both Ways
Theory of Change vs Logic Model: What Differs?
Theory of change vs logic model comes down to why versus what. A theory of change explains why a program should change people’s lives: the pathway from activities to a long-term outcome, with the assumptions named. A logic model shows what the program uses, does and produces (inputs, activities, outputs, outcomes) so staff can run and measure it.
A grant application has a box labeled “theory of change” or “logic model,” and the deadline is close. The quick question is which one to write. The better question is which one your staff will still open after the grant goes in.
In 2004, Heléne Clark and Andrea Anderson told the American Evaluation Association that the two terms were often used interchangeably, yet funders “almost all demand one or the other” (Theories of Change and Logic Models: Telling Them Apart). One fictional program, drawn both ways, shows the difference and what a logic model can miss.
Key Takeaways
- A theory of change explains why a program should work, with its assumptions named. A logic model describes what the program uses, does and produces.
- A logic model can list every input and output and still miss the assumption most likely to sink the program, such as whether anyone nearby is hiring.
- New program: write the theory of change first. Running program with a form due: fill in the logic model, then write the theory behind it within the quarter.
- Funders use the two terms loosely. Fill in their format and keep your own one-page version for staff and the board.
How Do the Two Compare Side by Side?
The two tools start at opposite ends. A theory of change starts from the long-term change and works backward, naming the conditions and assumptions it depends on. A logic model starts from the program’s resources and lists its parts in order, so staff can run and report on it.
Clark and Anderson sum up the difference as descriptive versus explanatory. A common objection is that a logic model has arrows too.
The UNM Evaluation Lab answers that the arrows in many logic models show only sequence: having resources leads to using them. A theory of change usually spells that reasoning out in more depth, with a reason behind each arrow. The table compares the rest.
| Theory of change | Logic model | |
|---|---|---|
| Question it answers | Why should this program lead to the change we want? | What does the program use, do and produce? |
| Where it starts | The long-term outcome, working backward | Resources and activities, working forward |
| Usual shape | A pathway diagram with a short narrative | A one-page table or flow chart |
| Assumptions | Named at each step so they can be tested | Often a short list, or left out |
| Best used for | Designing a program, or explaining its results | Running and reporting on a program |
| Effort | More work: a full version can take months | Faster for a program that already runs |
What Is a Nonprofit Theory of Change?
A nonprofit theory of change is the reasoning behind a program, written down. The Center for Theory of Change describes starting from the long-term goal and working back to every condition that must be in place for it to happen. Those conditions are preconditions. The beliefs that link one to the next are the assumptions.
The Center for Theory of Change calls the gap between what a program does and its goal the “missing middle,” and a theory of change fills it with steps you can check.
It can start small. The Center’s FAQ mentions one-sentence theories while arguing for more detail. A full version “is not something that can be done in a few hours or a few days”: some groups meet regularly for a few months.
The assumptions deserve the most care. Every activity can run on schedule and the program can still miss its goal, because one belief underneath it was wrong.
The most useful line in a theory of change is the assumption you are least sure of.
What Is a Nonprofit Logic Model?
A nonprofit logic model is a one-page picture of how a program works. The W.K. Kellogg Foundation’s guide defines it as “a systematic and visual way to present and share your understanding” of the links among your resources, the activities you plan and the results you hope for.
The Kellogg Logic Model Development Guide uses five parts: resources (or inputs), activities, outputs, outcomes and impact. Outputs are the direct products of activities. Outcomes are changes in participants’ knowledge, skills, behavior or status.
That line is the one to watch. “Forty learners enrolled” is an output. “Learners read two grade levels higher” is an outcome.
CNPC draws the same line in its own coaching: a completed six-session engagement is an output, and the progress a leader reports on the goals set at the start is the outcome.
What Does One Program Look Like Both Ways?
Take a fictional adult literacy program at a small nonprofit: two part-time tutors, a donated church classroom and adults who read below a sixth-grade level. Its goal is jobs. The theory of change explains why tutoring should lead to work. The logic model lists what the program runs.
The theory of change
In one sentence: if adults who read below a sixth-grade level get tutoring twice a week for a year, more of them will find work. The bet is that being unable to complete a job application without help is what keeps them from local entry-level jobs. Mapped backward from the goal:
- Long-term outcome: learners are employed within 18 months.
- Learners get interviews.
- Learners complete job applications without help.
- Learners read at least two grade levels higher.
- Learners attend most sessions for a full year.
Two more assumptions hold it together. Jobs are available: employers nearby are hiring for roles these learners could fill. Childcare is not the bigger barrier: a parent with no evening childcare will not come twice a week.
A theory is easier to test when it says how much is enough. In Clark and Anderson’s after-school example, students attend “at least 3 days per week for a minimum of 60 days” before test scores rise. For the literacy program, the threshold might be two-thirds of sessions.
The logic model
The same program as a logic model:
| Logic model part | Literacy program (fictional) |
|---|---|
| Inputs | Two part-time tutors, a donated classroom, workbooks, one small grant |
| Activities | Tutoring twice a week, a job-application workshop each quarter |
| Outputs | 40 learners enrolled, 96 sessions held |
| Short-term outcomes | Learners read two grade levels higher within a year |
| Long-term outcomes | Learners employed within 18 months |
What this logic model leaves out
Now look for the jobs. The logic model counts 40 learners and promises employment, yet nothing in this one asks whether anyone nearby is hiring, or who watches the learners’ children. Both questions live in the theory of change, where a board member or funder can test them.
If reading levels rise and hires do not, the logic model shows the gap but not the cause. The theory of change says where to look: the bet about applications, the jobs assumption or the childcare one. What that check finds decides where the next dollar goes.
Clark and Anderson’s conclusion fits: “A Logic Model that is a summary of an underlying theory is a much more powerful tool.”
A logic model can be complete and still be wrong.
Which Should You Write First?
For a new program, write the theory of change first, even as one sentence, and build the logic model from it. For a program that already runs, with a form due this week, fill in the logic model first and write the theory behind it within the quarter.
Sources split. Clark and Anderson say theories of change work best when they start from a goal, and a cluster evaluator quoted in the Kellogg guide says “plan backward, implement forward.” The UNM Evaluation Lab calls logic models “a more approachable first step.” One view suits a new program, the other a running one.
Writing assumptions down is slow thinking, and hard to do alone between board meetings. In CNPC’s executive coaching, a volunteer coach with no stake in the program asks the questions that test an executive director’s reasoning, such as which assumption they trust least.
Which One Do Funders Ask For?
It depends on the funder, and the words are used loosely. Some forms ask for a logic model, some for a theory of change, some for both. The Center for Theory of Change advises checking what a funder means by the term before you write.
The form usually tells you. Columns for inputs, activities, outputs and outcomes mean a logic model. A request to explain how activities lead to change means a theory of change. The Community Tool Box even uses “logic model” as a label that covers both.
Fill in their format, then keep your own one-page version in plain language for staff and the board.
How Do They Connect to Strategy and Evaluation?
Both tools feed planning and evaluation. The strategic plan names the priorities, and each program’s theory of change explains how it serves one of them. The logic model’s outcomes tell you what to track, with an indicator for each, and its outputs show whether the program ran as planned.
For the bigger picture, start with a nonprofit strategy overview and the steps of nonprofit strategic planning. A long-term outcome from a theory of change often becomes a goal in the plan.
A filled-in strategic plan template shows where program measures sit. A business plan is a third document, covering how one program will be paid for and run: see how business plans and strategic plans differ.
When Is This More Than You Need?
A small program with one funder may need only a logic model and a few sentences of reasoning. Neither tool proves a program works: that takes evaluation data, and sometimes an outside evaluator. A multi-site expansion or a rigorous outcome study needs evaluation expertise, a different job from coaching.
CNPC says so during matching: designing a program evaluation framework is technical work, and CNPC does not coach its way through a consulting problem. Coaching fits earlier: six sessions at $300 to $600 depending on operating expenses, because volunteer coaches donate their time. The theory of change stays yours to write.
A theory of change is a sentence about what happens to the people you serve, written carefully enough that you can check whether it came true. Quick win: this week, write that sentence for one program and add the assumption you are least sure of. For a thinking partner on the next step, Apply Now.
Frequently Asked Questions
Can a nonprofit have one theory of change for the whole organization?
Yes. An organization-level theory links the mission to a few long-term outcomes, and each program keeps its own logic model. The UNM Evaluation Lab notes that one theory of change can contain several logic models.
Does a theory of change need indicators?
Yes, once it moves past a first draft. Clark and Anderson list indicators as a requirement: each precondition gets a measure that shows whether it happened. In the literacy example, a reading test would show whether learners gained two grade levels.
Who should write the logic model?
The staff who run the program, with the executive director reviewing it. The Kellogg guide puts most of the value in the process of creating and revising the model, so build it with the people who do the work.
How often should you update a logic model?
Whenever the program changes: a new site, a lost grant, a different group of participants. Between changes, a practical habit is to review it each year with the program budget.
Test the Assumptions Behind Your Program
Testing a program’s assumptions takes a thinking partner. A coach works with you over six sessions, $300 to $600 by operating expenses, because volunteer coaches donate their time.
Apply for Coaching →
