This is a previous project site under CGIAR Research Program on Water, Land and Ecosystems (WLE). For more information and current updates, visit the WLE website.

Home > Basin > Nile > Outcome Logic Models: What They Are and How to Use Them

For the past few weeks, some of us at the M&E Team and friends have been having a discussion on OLMs, and their use in CPWF M&E. Here are some of the ideas we have gathered. We hope that far from finalizing the discussion, this brief recount generates new issues, especially your feedback on using OLMs!

What is a theory of change (ToC), and what is an Outcome Logic Model (OLM)?

A theory of change (ToC) describes how a project is expected to work: it is a description of the project’s activities, outputs, outcomes and impact and their inter-relationships.  Simply put: who are we contributing to change, why, and how will we do this? In solid research- for – development projects and programmes, where all researchers (coming from multiple disciplines) and all stakeholders should have a ‘say’ in the interventions of the project and should also have a space to contribute with information (be it local, regional or global) to understand ‘what may work’ and ‘what may not work’ in certain contexts, having an explicit ToC is a must. There are many benefits in  having an explicit theory of change, the first one being that all participants of a project are then ‘on the same page’, with articulated assumptions of the project’s causal logic. An explicit ToC can also be used as a basis for adaptive management (that is, for a project to plan- act- review- adapt) on a common set of ‘pathways’ or assumptions of what the project ‘thinks’ it will do. After the project is done, an explicit ToC can be the basis for evaluation: what did the project set out to do? How much of that was accomplished? Finally, a well articulated ToC can help quickly and easily communicate the ‘core logic’ of a project to its constituents, end users and the world in general.

There are many ways in which a project can make their theory of change explicit, usually ‘models’  or ‘maps’ that show ‘chains’. Some of these models begin at the activity level: what will this project do, which then, added to other activities, will ‘cause’ a result? (what I call activities here, others can call interventions, actions, initiatives and many others, and sometimes includes outputs. What i here call results are also known as goals, outcomes, impact, etc.) A project’s ToC can be made explicit in an Outcomes Logic Model (OLM). The main difference between an OLM and other ways of making the outputs- to – outcomes logic (ToC) explicit, including the famous ‘logframes’,  is that the OLM is actor- based: it is based on a definition of outcomes as ‘changes in groups of people’.

The OLM has been chosen by the CPWF as the tool to discuss, make explicit and adapt project and basin- level theories of change. OLMs are connected in the program to other project management and monitoring tools, which can all be found in the Project Workbook.

Tips to fill out the OLM

1. Column 1- Actor(s) who will change in the same way

Start with this question: if this project is successful, WHO will change? What groups of people or organizations will be affected? Be realistic, and prioritize the most ‘affected’. Then, try to be as specific about the actor groups as possible- in some cases you have the names or exact location of the actor(s)- use them here.

2. Column 2- Change in Practice / behavior

A change in practice or behavior is a change in the way people (in this case, the ‘actors’ in the 1st column) DO things. So here, try to use action  words- such as ‘use’, ‘coordinate’, ‘plant’, ‘participate in’, ‘integrate’, etc.

3. Column 3- Change in Knowledge, Attitude or Skills (KAS):

To be able to ‘use’ something, people usually need to first know/ understand it, and its advantages, and/ or have developed the skills to use it,  or at least to believe or trust the benefits of using it.  Look at your practice change, and make explicit 2-4 KAS changes that are key to having  the actor groups change their practice.

4. Column 4- Project strategies to achieve the change in KAS and Practice

How will your project contribute to make these changes happen? What are your ideas on how to help people change? Here, it may help to think what other projects have done before that has NOT worked well- what will your project do that is different? Better? Strategies can be the way (for example, co-develop instead of impose) you do things. Strategies are also the timing, methods, partnerships, ‘language’, etc. you choose to use.

5. Column 5- Project Outputs

Your project outputs come straight out of your road map/ research questions and/ or your contracted deliverables. Which is (are) your project’s main output (s) related to this ‘line of change’?

6. Below each Outcome pathway, its Outcome Narrative row

Now start backwards- what will you be doing, how (strategies) and/ or producing (outputs) that will contribute to modify actor’s knowledge, attitude and skills? In this space for the Outcome Narrative, you can further test your assumptions by telling them as a story: our project will be doing this, to help actors better be able to understand and/ or use this and this, and therefore they will be likely to adopt this or this behavior.

Does it sound right? Is it really a ‘solid’ way to try to affect this group´s behavior? When you do this for each outcome line, you ‘narratively’ present the logic of your thinking in that outcome pathway to others, and it should also help you “test the sound” of your story. It is by narrating the outcome pathway, or “connecting the boxes”  of that line that many times you find the gaps.

7. The last row in the model, the Impact Narrative

This is the space for a narrative description of how the project predicts that the outcomes it achieves are likely to reinforce each other and eventually contribute to longer-term impact: it is your opportunity to tell the world how these changes will add up to make a big difference!

Why use OLMs?

To encourage communication and discussions among and across project teams

When project teams develop their OLM as a part of their planning stage they share their assumed ToCs. Also, OLMs are useful for looking at the potential synergies across projects when targeting change/ influence within the same actor or group of actors.

To communicate the project logic to outsiders

We can use attractive/ graphic OLMs (not a table) (for example, the ´models{  that can result from graphic modeling efforts), ´stripped logic’ OLMs, which should ideally fit in 1 screen or slide, and an example of a derived ‘graphic’ OLM.

For reflection (here we mean the reflection process used in CPWF as Monitoring tool)

Before they can be used in reflection, OLMs need to be stripped right back to the core logic, that can serve as a trigger for discussion about whether their Project is doing the right thing.

OLMs can be used for reflection to discuss some of these questions, leading to the adjustment of the project work plans (in the learning cycle):

  • Are our assumptions still true in the light of what we learnt and changes that are going on in the broader context?
  • Are we still trying to get to the same actors, or have we found other ones?
  • Are some changes already happening or do we need to focus on new ones?
  • Do we still think they change behaviors and adopt new practices in the way we first thought?
  • Will these changes in KAS be achievable? or do we need new ones?
  • Are these strategies + outputs working so far?
  • Do we still think these outputs + strategies will lead to these KAS and P changes?

To articulate the BDC program logic, and find gaps

An overarching program made up of individual but interconnected projects, such as the CPWFs six BDCs, has a broad ToC, or program logic, that must incorporate and take into account the interactions between the individual (project) ToCs. Project OLMs can ´add- up´ to a program logic, and this program logic should also adapt and change to what
is happening at the project level.

To help with adaptive management

OLMs can help us get away from projects and programs that simply sign up to deliver outputs without ever asking if they are still needed. By first making explicit and then refining and revising our project logic, or theor(ies) of change, we ensure that the outputs we design will have a real use later on in generating outcomes in our target
user groups. We can also then make sure that if these groups change, so will our outputs, and vice versa.  

To help with early engagement of next users of our project outputs

The OLM identifies who will be the next users of research outputs and what they need to know and think differently before they use it.  This can guide the design of participatory research methods that bring in next users as co-innovators.  A premise that underpins OLM thinking is that early engagement with next users will help create a research output that is more appropriate to the needs of next users and which next users are more inclined to use because they were involved in its development.

What  an OLM is not: some commonly held mis-perceptions about what OLMs can do

Outcome Logic Models:

…may not help with understanding a project ‘at a glance’. OLMs in their initial, formal form of table are not easy to grasp in one glance. To achieve this, you may need to derive a different product. See ‘Stripped/ Graphic OLM’ for more on project-at- a- glance help.

are not supposed to be used as a standalone management or even planning tool.  OLMs are complemented by other management tools such as gantt charts, activity timelines, budget assignments and others that address activities, milestones, indicators, budget, and other necessary project tools. The OLM can (and should) be well connected to these (what activities are necessary to contribute to the achievement of the changes described?) but all the needed detailed plans of a project do not fit in the OLM! In the CPWF, projects have ‘workbooks’ to do this: excel books that contain the OLM and other project tools.

…are not necessarily good at showing research outputs that are produced by one project for the use of another. When you have several separate projects making up a bigger program, OLMs may not be the best way to find the dependencies and interrelations between projects. Because OLMs are actor- based (they start by the WHO) and not activity-based, establishing responsibilities and interdependencies between projects may be difficult. However, an OLM can be used to start these discussions, and for different
project implementers to start seeing when 2 or more of them are working towards a shared result.

(related to the above) …are not helpful for ‘assigning tasks’ to partners or understanding partner responsibilities. This is a job for the gantt chart, milestone plan, budget by partners, or other management tools your project might use.

…are not a project management tool that describes the production of outputs– this is done with a tool such as the Gantt chart.
What the OLM does do is provide guidance on how research should be carried out to achieve outcomes.

Common problems and FAQs about using the OLM

“The need to be specific with regard to target actors and corresponding strategies, at the same time as the need to be brief and thus the need to summarize and generalize. Strategies (such as communication) are very actor- group dependent, and need to be more specific than the space of the OLM allows for”.

Discussion: When projects in the CPWF are beginning, it is still not clear specifically who all the groups of actors involved will be. We know we need to influence general groups, such as ‘local policy makers’, but we do not yet know clearly who the strategic people to target in these groups are, and therefore, what our best strategies could be. When starting, is it  acceptable, then, to have a ‘general’ OLM, that we improve as we learn more about who we can work with and for? Definitely yes! As things get more specific, it will become necessary to ‘re-define’ (add definition). For example, many projects are working in different project sites and, in some cases such as the Andean Basin projects, even many different basins!. It may thus become necessary, as they learn more about each area (the institutional make-up of the area, and particularities of actors and local conditions) to re- define the OLMs. In some cases, eventually having separate OLMs per basin may be recommended (kudos for Gemán Escobar, Project AN1, for bringing this up and discussing it!) Another way in which this can be addressed is that strategies should not provide spurious detail (such as “x number of posters”): this type of information should instead be detailed in a communications plan/ activities plan.

How to handle OLM outcome lines dealing with intermediate products when research outputs are largely aimed at another project rather than outside next users? In principle, these can be handled within the existing OLM structure, but in practice the logic is weak.

Discussion: As explained above, OLMs are not the best tool for dealing with research outputs that are produced by one project for the use of another, or for finding the
interrelations between projects. Delivery of outputs within BDCs can be established in a timeline, or the Basin Gantt chart. Outputs that will heavily influence an outcome line for another project can be mentioned in the last column (‘linkages’, added specifically for CPWF projects for this purpose) and then described in detail elsewhere. A related thing is when 2 or more projects are working towards a shared result: these can be highlighted in ‘linkages’, and will be of course included in the ‘overall BDC logic’ and OLM.

If OLMs become over-elaborate, the underlying logic they try to capture becomes obscured in detail. People can get caught up in ‘filling in the boxes’ and then they lose analytical and reflective powers.

Discussion: This is true. OLMs, as described above, should be used to make explicit and describe overall ToC of a project. As soon as there is too much detail, the OLM looses its
advantages in the presenting of this logic. If your aim is to use the OLM to present your logic to others, detail ‘obscures’. This does not mean, however, that it is still not useful to generate and help guide the discussions on what this logic is. In writing up the OLM, it is good that project teams go into details, and challenge (even minute!) assumptions of the logic. The trick here is a well facilitated discussion that can gently lead the discussion from and to detail, to use the OLMs to enhance, rather than stifle, reflective powers.

Some examples (including project level OLMs, BDC level OLMs and graphic OLM)

Project level OLM:

Examples based on Project N2: RMS for Landscapes and Project AN3: COMPANDES- On designing and implementing benefit-sharing mechanisms and Building Collective Action for Water Benefit Sharing Mechanisms in the Andes

Other reading about ToC:


For more information contact Michael Victor ([email protected]).