Leveraging Complexity Science to understand Human Behavior
Part 2 | Problem with Beginning with the end in mind where the end emerges in unpredictable ways
Few years ago, we started an engagement focused on promoting HIV prevention services and products amongst at-risk Adolescent Girls & Young Women (AGYW). This a very complex problem that is impacted by a host of societal and other structural factors as well as experiences of AGYW. Our project also was facing uncertainties around problem definitions (how do we define risk), research design (which regions should we focus), solution ideation (where should we focus our solutioning efforts), piloting (who will we partner with) and even evaluation (how might we measure success). Despite the inherent complexity of the task we were undertaking, we were still required and came up with our theory of change (ToC), a list of key deliverables, potential solution areas and potential evaluation metrics. We knew then and we know now that most of things we imagined at the start of program did not pan out as envisioned. We had to constantly adjust and adapt our outputs based on the learnings that emerged during the course of the project.
And this is by no means a unique experience. Most programs are required to build a theory of change that usually includes a logic model that is essentially a linear causation map and also a results framework that defines the program objective and intermediate steps / results necessary to achieve the objective.
ToCs are/may be determined or typically approved by the funding agency or the program sponsor. The overall intent is to have complete clarity on the causal direction of interventions, outcome / impact and clear indicators of success for the program. A ToC thus provides a definitive structure to improve program’s explanatory power. Multiple issues have been reported with this approach. The Tavistock Institute has published some interesting documents and talks that articulate the issues with current approaches to Theory of change and potential ways to address it.
In this post, we will focus on one key issue: The emergent nature of change which makes the outcome unpredictable. In the paper, “Emergence is coupled to scope, not level”, Alex Ryan provides a good foundation around the idea of Emergence in a non-linear world and the inherent limitations in our ability to predict it. Emergent properties arise due to the relationship between macro and micro states. A key property of emergent behavior is that if its present in macro-state, it will not be present in the micro-state and it may be only be exhibited in certain environments. A simple example is the interaction and relationship between a lock and key. A key in isolation does not have an emergent property. Its emergent property of being able to open or close a complementary lock is only available in an environment that includes both the lock and the key. Another common emergent phenomenon widely studied is the flocking of birds.
Going back to our HIV prevention project example, micro-state may refer to individual AGYW while macro-state could refer to the overall community including boys or sexual partners, teachers, family members, peers, health professionals etc. that she is part of. In this case, we could think of community norms and behaviors around HIV prevention as something that emerges from the interactions between all the members of the community. The norm is not something that is present with any individual in the community or even group of specific individuals (e.g., all AGYWs). Micro state involving individuals harbor actions and macro state then is the stage where norms germinate through the cumulative actions in the microstate. Any desire to affect these community norms means working with and influencing multiple interactions. To be able to predict specific changes to these norms would mean being able to model with a relatively high level of accuracy the change pathway at both micro and macro states.
Multiple theories of change have been developed to articulate this change pathway and the current evidence around their success isn’t very encouraging. Through the lens of complexity sciences, the problem here is that that we are aspiring to predict something that by definition is not available until it is near or has already occurred. In other words, we are aspiring to predict an end and we don’t really have the means to since we are working with non-linear change environment.
This brings us to the question, what can we do instead? Principally, we could change our approach in following ways:
- ToCs could be considered as a collective sense-making tool instead of top-down solution design. It could help different stakeholders such as donors, implementors, governments, and community actors understand on how they might work together with complexity. This reframe of ToC is becoming more acceptable in the development sector.
- Given that most systems like health or education are complex systems nested / inter-connected with other systems, we may want to develop multiple nested Theories of Change. The different TOCs could also be sequential in nature to account for the dynamic nature of change. Each ToC could have their own set of high-level logical maps. Perhaps, the one in the near future could be more specific in their actions and outcomes while those further way could be largely aspirational.
- The nested approach also suggests aiming for Transition Points which may be aspirational shifts in systems and behaviors that we are hoping to achieve over a period of time. The Transition Points could be potentially mapped as outcomes of the different ToCs. The key word here is aspirational and we need to be prepared for transitions that we haven’t imagined yet.
- Since we are working with an open system i.e., the outcome will be impacted by multiple environmental factors outside our intervention, we should assume that there are multiple pathways to change (positive and negative). This would mean learning and mapping the specific pathway of change during the course of implementation and updating our logical maps accordingly.
- The previous point also indicates that we may want to delay or abstain from setting baselines, indicators and targets until the change pathway (relationships between micro and macro) have been understood to a good extent. Setting them too early would again risk a linear approach guided by our limited knowledge of the system.
- The previous two points also suggests that we may want to set milestones to revise our ToC to ensure its adaptation based on learnings and reality that has emerged during the course of the project.
It’s evident that biggest change that we would need is a shift in our mindset in approaching these complex problems. A key shift being, that we may not be able to begin with the end in mind. The end will emerge and we will only know once it has arrived.
- Heino, M.T.J.; Knittle, K.Noone, C.; Hasselman, F.; Hankonen,N. Studying Behaviour Change Mechanisms under Complexity.Behav. Sci. 2021, 11, 77. https://doi.org/10.3390/bs11050077
- Ryan, A. J. (2007). Emergence is coupled to scope, not level. Complexity, 13(2), 67–77.
Final Mile brings unique and proven capabilities in addressing complex behavioral challenges. As one of the first Behavioral Science & Design consultancies, Final Mile has had the opportunity to bring these to practice in a wide variety of sectors and contexts. We have executed highly complex behavior change projects across a wide variety of areas covering Global Health (HIV, TB, Maternal Health, WASH), Financial Inclusion, Safety across Africa, Asia, Europe, and the US.
Final Mile is also building a pandemic playbook that can be used as a potential toolkit by policymakers and implementors in mitigating Covid19 and future such pandemics.
Reach out to us at firstname.lastname@example.org.