Theory of Change vs Logical Framework: Which Does Your Program Need?

If you write proposals or manage social programs, you have probably been asked for a Theory of Change, a Logical Framework (logframe), or both. They are easy to confuse: a 2012 review for the UK Department for International Development (DFID) found that many people struggle to tell them apart, partly because both belong to the program theory family

This article explains each tool, how they differ and fit together, and how to build both with the people your program serves

What Is a Theory of Change?

A Theory of Change describes how and why a program expects its work to lead to a long term change. UN Development Group guidance from 2017 defines it as a method explaining how an intervention is expected to lead to a specific development change, based on a causal analysis of the available evidence

The work starts at the end. The team names the long term goal and works backwards to the conditions that must be in place first, which the Center for Theory of Change calls preconditions, before choosing activities to bring them about. Each link rests on an assumption, which the UN guidance suggests stating as “if X, then Y, because Z” and checking against evidence

The result is usually a diagram of boxes and arrows with a short narrative. The DFID review by Isabel Vogel found broad agreement on the basic elements: the context, the long term change sought, the expected sequence of change, the assumptions, and a diagram with a narrative summary. Important assumptions often surface as the group draws the diagram together

The approach grew out of program theory in evaluation, and evaluator Carol Weiss popularized the term: in a 1995 book from a roundtable on community change convened by the Aspen Institute, she suggested that a key reason complex community initiatives are hard to evaluate is that the assumptions behind them are poorly articulated. Today, UN guidance asks that a Theory of Change be developed with stakeholders, grounded in evidence and revised as the program learns, while the DFID review calls it both a process and a product

What Is a Logical Framework?

The Logical Framework Approach (LFA) was developed for the United States Agency for International Development (USAID) at the end of the 1960s. According to the European Commission's Project Cycle Management Guidelines (2004), it addressed three problems: vague planning, unclear management responsibilities, and evaluations that turned adversarial because nobody agreed on what the project was trying to achieve. NORAD published its widely used LFA handbook in 1990, and the Commission adopted Project Cycle Management, which is based on the LFA, in 1992

The guidelines separate the LFA, an analytical process of stakeholder analysis, problem analysis, objective setting and strategy selection, from its product, the logframe matrix. The matrix usually has four rows, from the overall objective (goal) through the purpose and the results (outputs) to the activities, and four columns: the intervention logic, indicators, sources or means of verification, and assumptions. Labels differ between agencies, but the logic is the same

Read from the bottom up, the matrix is a chain of if then statements: if the activities are carried out, the results are produced; if the results are produced, the purpose is achieved, as long as the assumptions in the last column hold. Read across, each row shows how progress will be measured and where the data will come from. After design, it guides contracting, work planning and monitoring, and later serves as the record of what was planned for evaluation

Theory of Change vs Logical Framework: Key Differences

The Results Chain Both Tools Use

Both tools rest on the same results chain, which the OECD evaluation glossary defines as the causal sequence of an intervention:

1. Inputs
2. Activities
3. Outputs
4. Outcomes
5. Impact

• Inputs: the funds, people, materials and institutional resources a program uses
• Activities: the work that turns inputs into outputs, such as running a training course
• Outputs: the products and services delivered, such as 50 farmers trained, which the implementing team controls
• Outcomes: the short and medium term effects of the outputs, such as farmers adopting new practices
• Impact: the higher level, longer term change, intended or not, such as more stable household income

Side by Side

How the Two Tools Work Together

As Julian Barr of ITAD puts it in the DFID review, the logframe is “good for defining success but not for defining reality”. So develop the Theory of Change first, with people who know the context, to agree on the pathway and test the assumptions. Then decide which part of the pathway your program is responsible for and write it into the logframe: the long term change becomes the goal, the key preconditions become the purpose and outputs, and the assumptions that matter most move into the last column. UN guidance likewise asks that the Theory of Change shape the results matrix, its indicators, risks and assumptions

Both then feed your monitoring, evaluation and learning (MEL) system. Logframe indicators and their sources become the monitoring plan, and the assumptions in the Theory of Change point to the questions an evaluation should test. If you plan a Social Return on Investment (SROI) analysis, the same work pays off again: the SROI guide asks you to build, with stakeholders, an impact map or theory of change that shows how inputs lead to outputs and outcomes

Common Mistakes

• A Theory of Change that only lists activities. If the diagram jumps from “training held” to “poverty reduced”, the middle is missing. The Center for Theory of Change notes that some so called theories of change are in essence just logframes
• Assumptions nobody tests. The DFID review found that in weak logframes the assumptions column has become almost a “too hard” column, where an entry such as “policymakers will read the research” puts the problem out of scope. Check key assumptions against evidence and monitor them
• Logframes written after the fact to satisfy a donor. The EC guidelines warn against preparing the matrix separately from the other project documents, as an afterthought, which leaves it out of step with the plan
• Indicators that only count outputs. The number of people trained shows delivery, not change. The EC guidelines ask that indicators for results (outputs) describe the measurable consequence of activities rather than restate them, so add outcome indicators with a baseline

Building Both in a Participatory Workshop

NORAD's handbook recommends holding LFA workshops in the project area with all parties and an independent facilitator, involving intended beneficiaries directly or through simpler sessions. A practical sequence:

1. Prepare. Gather existing evidence (needs assessments, past evaluations, baseline data) and invite community members, local government, implementing staff and the funder
2. Analyze the problem. Build a problem tree of causes and effects, with one idea per card so everyone can see and move them
3. Map the change. Turn the problem tree into a solution tree, agree on the long term change and map backwards the preconditions needed to reach it
4. Surface the assumptions. For each arrow, write “if X, then Y, because Z” and mark the links that lack evidence
5. Choose your share. Decide which pathways your program will take on and which belong to partners
6. Write the logframe. Move the chosen pathway into the matrix and set indicators, baselines and data sources, including who collects what and how often
7. Plan reviews. Set dates to revisit both documents with monitoring data and adjust them

Conclusion

So which does your program need? Usually both. A Theory of Change explains why the program should work; a Logical Framework sets out what it will deliver and how progress will be checked. Built in that order with the people affected, they give funders a plan and teams a pathway to test and improve. As the DFID review advises, keep theories of change proportionate to the scale of the initiative

SIDIA's Impact Program Design service works with partners from a community needs assessment to a Theory of Change and Logical Framework, and on to partnership blueprints, implementation roadmaps and SROI measurement