Refining the Complexity aware Theory of Change

When I wrote my last post about experimenting with new structures for a complexity aware Theory of Change (ToC) in Myanmar, I had a few elements in place, but still some questions. Going further back to an earlier post, I was clear that differentiating between clear causal links for complicated issues and unpredictable causalities for complex ones is critical. I have been thinking about that a lot and last week I have taught a session on monitoring in complex contexts and I think I have found the final piece of the puzzle.

Going back to my last post, here is the structure I proposed:

  1. Intervention: This level broadly describes the interventions without going into detailed activities as they might change frequently. Intervention are seen as triggers for stakeholders to react to.
  2. Uptake: This is about the immediate change the project wants to trigger on the level of the partner the team engages with. Essentially, this is the justification of why we work with these partners. What is the change in behaviour that we want to see on that level?
  3. Interaction of interventions/with context: This is where the changes taken up by project partners start to interact with each other and are exposed to the context. For example, how would increased knowledge of farmers on cultivation and production technique play out in the reality of a very difficult market that drives down rather than demands quality.
  4. Systemic change: this is the layer where we look at wider patterns of persistent failure or underperformance and how they could be more beneficial.

What I think makes sense are the first and second points. We develop a portfolio of interventions and for each intervention we define what reaction we expect, i.e. we define an uptake. This can and should actually contain both signs for success and signs for failure as well as strategies to amplify positive patterns that emerge and dampen negative ones. What I changed now are the layers 3 and 4 in the list above. Let me start ‘on top’. Layer 4 should give us the direction of travel – where we want to go with our project or programme. If we are locked into a logframe environment, this is often reflected in outcome and impact objectives. Of course this is not ideal in a complex environment as we cannot predict what success looks like. Therefore, this ‘top layer’ should be our strategic intent as defined in Systemic Insight. The strategic intent is not a fixed objective but gives us a sense of direction and allows us to decide whether a new emergent pattern is ‘better’ than what was there before.

Between the uptake layer and the strategic intent layer, there was still a void. Back in Myanmar, it became clear to me that although I knew that this is about interactions among the intervention and with the context to achieve the emergent improvement we want to see, I was not clear exactly how to capture this. Over the last weeks, this has become clearer and I could include my new idea in the session on M&E I did in last week’s Mesopartner Summer Academy. The reaction of the participants has been very positive.

What I added in this layer between uptake and strategic intent is a layer that depicts our conceptual understanding of the change. This conceptualisation depends on the context we are in and should be drawn from a solid understanding on how change happens in the systems we work in. For example if we work in an economic system, such as territorial economic development or market systems development, we want to use a conceptual understanding on how change happens in an economy. This understanding should as far as possible draw from deep theory, from an established scientific understanding of that change. This is obviously not always possible, so alternatively we can also draw from well-established heuristic frameworks.

Ok, so let me bring these thoughts together before giving two examples.

Here the layers of the ToC that I suggest, from the bottom up (although direction does not really play a role, so if you prefer left to right, right to left or top down, I don’t mind):

  1. Intervention: This level broadly describes the interventions without going into detailed activities as they might change frequently. Intervention are seen as triggers for stakeholders to react to.
  2. Uptake: This is about the immediate change the intervention wants to trigger on the level of the partner the team engages with. Essentially, this is the justification of why we work with these partners. What is the change in behaviour that we want to see on that level?
  3. Conceptual understanding of systemic change: This layer shows our conceptual understanding on how change happens in the given system and how the interventions potentially contribute to achieving this change. This layer typically starts off as being very generic but then the lifetime of the project becomes more specific as the team starts to better understand what influence the interventions can have on the system.
  4. Strategic intent: A formulation of the direction the project wants to see change in. Instead of defining an ideal future state, It starts with a deep appreciation of the situation now and how we got here and indicates what we expect to be a ‘better’ situation.

The elements of the Complexity Aware Theory of Change as I used them for the session at the Mesopartner Summer Academy 2016 in Berlin.

Two example on how we can use this.

First example: For the project in Myanmar, I suggest that as a conceptual understanding of change in the rubber market we take the heuristic framework of the ‘five functions a market needs to work’ that we use a lot in our teaching on markets and market failures in Mesopartner. It draws from an influential book by John McMillan (McMillan 2002) and other scholars. Here are the five functions:

  • Information that flows smoothly.
  • Property rights that are protected.
  • People must be able to be trusted to fulfil their promises.
  • Side-effects on third parties must be curtailed.
  • Competition in the market is fostered.

These five points would then be further described specifically for the project in order to make them relevant. In the beginning, this could be done in a rather generic way (e.g. information on market prices and knowledge on improved production techniques flows more smoothly through the sector and particularly to the smallholders). In a later stage, theses five functions could then be define more specifically once more is understood about the potential influence the project can have and the interconnections and dynamics in the sector.

BEAM ToC empty.png

The BEAM Exchange Theory of Change without detailed content. It shows the levels of interventions, uptake, systemic change (using the CB2 dimensions) and strategic intent.

Second example: For the BEAM Exchange, we used a slightly adapted version of the model called Capacity Building 2 (CB2), which describes capacity building through four dimensions (Punton, 2014):

  • Personal change
  • Network change
  • Organisational change
  • Change in attitudes and norms

The BEAM ToC is based on a Logframe that was agreed on with DFID. Hence, the ToC includes the defined Outputs. The portfolio of intervention is defined through a number of work streams. The Outcome and Impact objectives are used instead of a strategic intent. The four boxes on the systemic change level are further refined by a number of bullet points with changes that are seen as required in each dimension in order to achieve the outcome and impact objectives. While the bullet points were very generic in the beginning, we are currently working on refining them based on a better understanding of the dynamics in the sector.

It would be great to get some comments and thoughts on this suggested approach to a more complexity aware Theory of Change.

McMillan, J. (2002). Reinventing the bazaar : a natural history of markets. New York, Norton.
Punton (2014). Viewing Capacity Development through Four Dimensions of Change. Itad Blog. [Retrieved: 12/07/2016]

8 thoughts on “Refining the Complexity aware Theory of Change

  1. Mike Klassen (@Mike_Klassen)

    Hi Marcus,

    A really fascinating post and I appreciate the way you’ve ‘tracked’ the changes in your own thinking over time. I really appreciate the intervention and uptake framing – makes sense, and by saying ‘uptake’ it forces us to think about the fact that it may well not be taken up for a variety of factors. I believe I grasp your 3rd layer of ‘conceptual understanding of systemic change’ but I think it would be greatly helped by an example, and ideally multiple examples to show the implications when you have different conceptual understandings. In the context of market systems programming, what would be the pervailing/dominant conceptual understanding? Something around the presence of competitive norms that lead firms to mimic/replicate new business models that prove effective? (i.e. crowding in).

    In terms of the examples, I found the 2nd example a bit confusing. It feels like you are fitting a square peg in a round hole by layering the complexity-aware ToC onto the standard logframe. Also, what would the ‘conceptual understanding of systemic change’ be for this context? Maybe part of the challenge on my side is that I can’t see the detail around the ToC for BEAM.

    Thanks for continuing to push the field forward and to share your thinking,


    1. Marcus Jenal

      Hi Mike. Thanks much for your comment. This is very new thinking so there are not many examples I can give you to illustrate the concept. I have given two in the main article above, the one from Myanmar and the one from BEAM, the latter being more matured. Unfortunately, the BEAM ToC is not publicly available right now, so I cannot share it here. I hope that the next version, on which we are working at the moment, will be available. I can give you some examples for each of the four boxes, though:
      Individual: Improved knowledge to implement market systems approaches; Greater availability of qualified professionals at all levels
      Network: A growing, engaged and self-confident Community of Practice; Community comprises an increased diversity of sectoral professionals
      Organisation: Procedures, policies and guidelines in donor organisations are conducive to market systems approaches; Implementing organisation’s value learning and adaptation
      Norms and attitudes: Greater consistency in the way market systems approaches are described and applied; Market systems approaches are seen as valuable for achieving sustainable development at scale in diverse sectors

      Now, these are still very generic and we are currently working to get greater consistency with our interventions (workstreams) and thinking about uptake.

      And re your second comment: it is a compromise to fit this whole thinking around a logframe. It is not a new point that systemic change approaches are not really compatible with logframe thinking. But unfortunately for many donors log frames are still mandatory. The alternative is to have two separate logics. On the one hand the logframe which frames the contractual agreement with the donor and on the other hand the programme ToC, which is the basis for programme management, learning and adaptation. We have tried the former here. I agree with you, it is not totally coherent. In Myanmar, we are going the other way. We will see how this works.

      1. Mike Klassen

        Thanks Marcus! Helpful to get that clarification and context. You’ve certainly whetted my appetite to see the BEAM Theory of Change once it becomes public! I find the term ‘self confident’ attached to community of practice to be particularly interesting. It makes total sense to me, but had never heard that desired outcome described in that way before.

  2. Michaela

    Marcus, great on your effort to try to provide something that people could use and improve the current practice …but I am not really convinced about the need for a TOC in the context you are suggesting, where you deal with complex issues…I think you build it as you go and implement the project plan review and plan further, not develop it in such detailed form from the outset…What is the difference that your approach/toc is making to what already exists? It does not deal with the ‘heart’ of CA… I thought being “complex aware” is about dealing with unpredictability “ground yourself and your work in unpredictability” as someone has recently twitted; rebalance boundaries as you move… I am not sure that your approach and the TOC you are suggesting is really responding to these issues…

  3. Marcus Jenal

    Hi Michaela. Thanks for your comment. I don’t think you can do without a ToC, even in complex contexts. A ToC essentially describes why a project is set up and is doing what it is doing. You need to be able to describe why you do a certain intervention and not another one that would also be possible. We always have an implicit ToC, so it is better to make it explicit so we can discuss it and change it while we learn. So ToC yes in principle also in the complex domain. But it has to be adapted to complexity, I agree. It has to enable us to embrace uncertainty. If we are dealing with complex issues, a ToC is not about setting up an imaginary future state and describing the way of getting there. It is about documenting our understanding of the current, the direction of change we think is needed, and our strategy on how we get there – what systemic change we want to see, how we can nudge the system in the ‘right’ direction of change. This then builds the basis for us to learn. We adapt and replace interventions, reflect on the conceptual understanding of systemic change, etc. It is a living thing, not a lock-in for the next four years.

    ToC can help us to deal with unpredictability in showing us what we know and what we don’t know. Where are the known knowns, the known unknowns, the unknown unknowns? For these require different strategies. It is not about predicting the exact causal path, but an understanding of how change happens.

    I think this is quite a different approach to ToC than the one that is predominantly applied today and maybe I should not use the term ToC because it is not about naming blocks and connecting them with causal links. But this is what people are used to, so this is where I want to start off. It’s not about ‘if we do this, then that happens’, at least not in the complex domain. It is more about ‘this is our conceptual understanding of how change happens and these are all the things we are trying in order to see what happens’. Does that make sense?

    1. Michaela

      Thanks Marcus
      Probably we are talking about the same approach…What I am trying to say is that you build it as you go! You plan a bit, refine as you accomplish or not, and plan the next steps, refine plans build etc.. The learning I am getting from my research but also from practice working with the private sector for example is that this is the way you should go. I think the current frameworks used orb proposed to be used in development ( not in other sectors) are way behind this thinking for various reasons including pressure to predict everything…

  4. bhavfish

    Just found your blog, thanks for writing, really useful.

    You wrote – “We always have an implicit ToC, so it is better to make it explicit so we can discuss it and change it while we learn.” I generally agree with this but I am finding people get caught up in the words. Maybe sometimes a person can’t explain yet and so they need to say a little and then they need to do / test / experiment, and then they can start to understand their own implicit TOC.

    This is something I am working with / trying out, instead of trying to make everything explicit. My thinking behind it comes from the idea that we have a deep symbolic / metaphorical understanding of things that is held together abstractly and that we sometimes loose this when we are forced to “spit it out”!

    1. Marcus Jenal

      Hi bhavfish, thanks for your comment. I agree with you. Sometimes it is difficult to formulate something. What you are describing would be really great practice, but it needs a lot of trust within a team. If the manager is confident enough to give his/her team the ‘long leash’, this could work out really beautifully. It would also adhere to the principle in complex systems to decentralise the decision making.


Leave a Reply to bhavfish Cancel reply