We know that current development challenges are complex. But not all elements of a development programme are necessarily complex. How does a theory of change look like that shows us which aspects of a programme are complex and which aren’t? How does this help managers to develop appropriate strategies for interventions and focus their attention?
I have been thinking quite a bit about monitoring and how to find a monitoring framework that works in programmes that are facing the complexities of the real world (and I blogged about it before here and here). More and more monitoring guides speak about the complexities programme managers and staff face ‘out there’ and some guides even venture as far as to say that change in the real world is not linear or predictable. The new BEAM Monitoring guidance, which I co-authored, for example, states that ‘a market systems programme is unlikely to achieve its goals in a simple, linear way. It may be difficult to fully understand (at least in advance) how causes and effects will work at a system-wide level. There may therefore be significant uncertainties about how the overall market system may be re-oriented to serve poor people better.’
But what these guides subsequently recommend is usually still ‘traditional’ results chains, which are inherently linear and a fairly detailed prediction of what will happen and what is needed to make this happen.
This does not mean that because complex situations are unpredictable we cannot develop a theory of change. On the contrary, we need to capture our hypotheses of how we think we can get to the change we want. Being aware of complexity does not mean that the only thing we can do is to venture out there and try all kinds of random things. Trial and error needs to be systematised.
So how should a ‘complexity aware’ theory of change look like? Some others have been writing about this (for example Vogel et al. (2013)) and we have condensed current thinking and practice in the BEAM Monitoring guidance:
- The theory of change needs to be presented as an overarching framework that explains how the programme intends to work, but without detailing the specific mechanisms of change (i.e. interventions). This will help ensure the theory of change remains valid even if individual interventions are adapted, closed down, or scaled up.
- In conditions of significant uncertainty, an adaptive, learn-as-you-go approach is essential. It makes sense for programmes to include a range of exploratory interventions that can be scaled up, or brought to an end. These projects may run independently of each other, and each should be thought through with its own mini theory of change.
- Due to the adaptive nature of market systems programmes, the theory of change should be reviewed and modified regularly to reflect emerging findings, changing hypotheses, and adjustments to programme strategy.
- In a complex system, different stakeholders will have different perspectives and interpretations about what makes things work, which may not be amenable to analysis with a single model. It is therefore important to record different viewpoints and assumptions, and review these critically through ongoing monitoring efforts.
- It is perfectly feasible for a programme team to develop a theory of change ‘in house’. However, in light of the likelihood of a diversity of views emerging, it may be helpful to commission an external facilitator to help develop a theory of change.
What many people who write about complexity and theory of change stress is that we need to become better in focusing on the links between the boxes in the results chain, i.e. our assumptions of why some activity or change leads to a subsequent change. But how can we do that? To come up with a theory of change for the overall programme or results chain for an individual intervention, people usually sit together in a workshop and try to develop incremental causal steps from an intervention to the end results of systemic change and poverty reduction.
The first version of our theory of change should still be fairly general since we cannot yet know much about the intricacies of how change happens. It essentially is the representation of our knowledge and hypotheses we start off with. Once we have completed this first version of the theory of change, we can take all the assumptions that underpin our hypotheses, i.e. all the links between the boxes and write them on cards or post-it notes. We then collect them on a board or wall and start a Cynefin exercise. Through the exercise two things happen. On the one hand, we become clearer which of the causal links in our theory of change are obvious, complicated, complex, or – God forbid – chaotic. This will help us manage the intervention design. But secondly, we also develop a common understanding in the team for complexity and the difference between the ordered and the unordered space. The team will have a common vocabulary to speak about these things and a collection of examples to point to. So when a new situation comes up, a new causal relation needs to be discussed, the team can relate them to the examples used in the exercise and ask for example whether the new assumption is more like one from the complex domain or one from the complicated domain.
Causal links that are in the ordered side, i.e. either obvious or complicated, can be left as direct arrows in the theory of change or results chain. We either have the knowledge and evidence that the causality is linear and predictable or we think with more analysis or involvement of an expert we can determine what we need to do.
In the complex domain, there are multiple competing hypotheses of how the intended change could be achieved or could look like and the available evidence supports different and even competing perspectives. So rather than to add a straight arrow between two boxes, we should add an ‘exploration box’ that contains both boxes. This means that the causality between the two boxes is unclear and we need to run a set of experiments to determine how we can achieve the intended change. We also need to be aware that the relationship might be non-linear and that the intended change has to emerge due to the interaction between different changes – or put simply there are different things that need to change in order for us to see the change we want to see. In some cases we might even not know how a good outcome would look like before we see it.
What we end up with is a theory of change with different types of assumptions or arrows. Some are linear and fixed – we know that if we do A, B will happen. These are the areas where we need less involvement of the management other than in an oversight function. Then we have the ‘exploration boxes’. This is where we need to develop a portfolio of experiments and be exploratory and adaptive. This is where management needs to focus on and be involved the most.
This type of ‘complexity aware’ theory of change helps us to figure out which things we are sure we know and we can predict and which are complex. We know that different strategies are needed for the different domains (also described in the Cynefin framework). Furthermore, the ‘complexity aware’ theory of change approach also helps managers to know where they should focus their attention and use tools like adaptive management or agile approaches.
Vogel, I., Fisher, G. and Ramalingam, B. (2013). Complexity-Informed Theory of Change for DFID’s Private Sector Development Programme in Democratic Republic of Congo: Pilot Report.