Should we fix wrong behaviour?

I have spent the last three days at the Market Systems Symposium 2019 in Cape Town. I really enjoyed the event. Besides meeting good friends, there was a great number of practitioners with astonishing accumulated experiences on how to implement Market Systems Development (MSD) projects. There was also a good number of donor staff, which provided their perspectives on the challenges of implementing adaptive and learning programmes – unfortunately the European donors were largely absent (with the exception of one participant from the Swiss Agency for Development and Cooperation), most of the donor staff were from USAID. The majority of the active participants were those that are working on further developing the approach,  innovate within their project and generally try to make market development more effective – it was really exciting to hear what they figured out and what they are struggling with. There was a good energy around during the three days. But I have also picked up some concerning trends, particularly around the growing intent of MSD projects to change market actors’ behaviours.

Bridging the conceptual and the practical

What I enjoyed the most about the Symposium was the ability of the event to bridge conceptual discussions with very practical experiences. One of the discussions throughout the Symposium was on how to use concepts and frameworks from systems thinking and complexity in practice. But then there were also many sessions where people talked about their experiences, for example when a bunch of project leaders shared the stage in a plenary session and talked about what they are excited about and what keeps them awake at night.

I also enjoyed the thematic track on Market Systems Resilience as there are some really interesting explorations going on in various projects. Also there, the participants were able to discuss conceptual questions around resilience with practical challenges projects face when working towards Market Systems Resilience and attempting to measure it. There were lots of other interesting discussions in and between the sessions, for example on the question how we become more effective in learning as a field of practice and avoid repeating the same failures over and over again.

I want to send out a appreciation to the organisers who have been able to pull together an amazing bunch of people and that at only the second instance of the Symposium, which premiered in 2018.

Mesopartner provided a one-day training on complexity in economic development the day before the Symposium

Designing better behaviours?

There was, however, a concerning trend that I picked up in various sessions. As MSD projects do not generally provide grants or build infrastructure but facilitate change by working with and through partners, behavioural change has grown in importance, also driven by insights from behavioural economics gaining more broad recognition. The questions asked are for example: How can projects shape behaviours of buyers of agricultural goods to change their business strategy from short-term extractive to more oriented towards building a sustainable business? Or how can projects influence market actors to show behaviours that would improve their resilience, e.g. by making them save more, buy insurance, or diversify their assets?

In one exercise, the participants were asked to come up with statements a market actor would make who shows ‘bad behaviours’ (which often reflects current behaviours) and statement one would make who shows ‘better behaviours’. Once we know the two apart, we can convince the market actors that it is in their own interest to take up these ‘better behaviours’ that are described to them by the projects.

We know that behaviours are shaped by many different factors, including but not limited to a person’s history, knowledge and experience, their social and commercial relationships, formal and informal norms and customs, values and beliefs, etc. All of these factors, as well as how they come together in the particular moment, influence how people make sense of what is going on and how they behave as a result. Recent research even shows that the current mood as well as what people eat for breakfast that day influences their sensemaking. In the logic of behaviour change in MSD project this is unfortunately often conflated into a simple ‘they don’t know better what behaviour would be good for them’. Hence, if a project works with the people and shows them the light of a better way to behave everything will work out. I’m using on purpose a stark formulation here to make a point, reality is most certainly more nuanced.

Besides the complexity on how behaviours are shaped, also the complexity of the context needs to be taken into account. We cannot really define what ‘better behaviour’ looks like, as we cannot predict what outcomes a specific behaviour would result in in a complex context. If we convince a number of actors to become more inclusive and move their business strategy away from short-term extractive behaviour this could well lead to the end for these companies in a very competitive environment which does not value these traits. From evolutionary economics we know that it is not only the things that we try that counts, but also the selection environment in which they are expressed. So instead of designing ideal behaviours in an unpredictable context, we should encourage diversity and variety in behaviours and at the same time we need to work on the selection environment on which behaviours are valued in a market. In addition, whatever I have heard as examples of ‘better behaviours’ during the Symposium is strongly influenced by Western normative views on how the world should work. It doesn’t help if we say that a trader should build up more trusting and long-term relationships with his or her suppliers if the whole incentive structure around this person shapes a behaviour that is short-term focused and extractive. This can be a very rational reaction of these people to their reality. So we should assume that given all the factors the people perceive they are making the right decisions for them at the moment.

There is also a moral reason why I think directly targeting people’s behaviour is problematic. Who are we to go to a country or region and tell the people how to behave? On what moral basis can we as development actors argue that our interpretation of reality and the required behaviour to respond to this reality is superior to the people who actually live in this reality? The same moral argument has been made against nudge-type interventions in general, but I think it becomes much more powerful in development with development workers coming from developed countries and/or totally different cultures (and the symposium was, probably representing a good sample, still dominated by mostly white, western development workers) – this is simply patronising.

But what is the alternative? 

We find an alternative approach in complexity thinking. We can catalyse attractors, change constraints or increase diversity. Catalysing attractors means that we can supply a place for new behaviours to crystallise. This can for example be through making existing divergent behaviour more visible, i.e. working with so called ‘positive deviance’, or by exposing market actors to different realities that might stimulate some exploration of different behaviours. Changing constraints would look at why certain behaviours are not taken up, e.g. because there are strong social norms that prevent them or because there might even be formal rules that do not allow them or severely hamper them (e.g. red tape). A project can work with such institutional or structural aspects like constraints, but the exact change in behaviour that will become visible on the surface will be an emergent result of this work – i.e. it cannot be predicted. This means we need to be cautious and tread lightly, ideally by using safe-to-fail probes. Increasing diversity and variety of things tried in the system opens up new options for different behaviours. This could be different business models, different processes, or different products. A project can stimulate this by working on and strengthening the local innovation system. It generally means working through meso organisations like technological or extension organisations, research stations or educational organisations like universities or vocational training institutions.

For me the fact that many are trying to design the ‘better behaviour’ is a symptom that we as a community often still default back to our traditional modes of working where we identify a problem and try to find a solution. So in that sense we haven’t made the switch to a way of working that is truly informed by complexity thinking. The types of solutions may have shifted from doing capacity building or building factories or cold storages to designing better behaviours. But it is still about fixing problems through designed solutions, rather than about attempting to work on the evolutionary potential and the development trajectory of the system. 

Another thing that I picked up is that people still ask for concrete tools and methods they can implement in their projects. This is also a symptom of the linear, reductionist way of thinking: “If I can find the right tool I can design the right solution”. In reality, complexity requires us to deeply explore a context by engaging with it. Pathways will emerge as we do that. There is heuristics guidance on how to do this (like our Systemic Insight approach), but here is no recipe as what needs to be done and how is unique in each context.

All in all, I think events like the Market Systems Symposium provide important spaces to grapple with these ideas, methods and concepts and learn together as a community. I’m already looking forward to the Market Systems Symposium 2020!

9 thoughts on “Should we fix wrong behaviour?

  1. John Turley

    Marcus – great blog again, on an important point. Do you think a propensity to view the world / market / company / whatever as a system, even a complex adaptive system, has a underlying assumption of control and therefore we would be better off thinking of human systems as something other than system?

    1. Marcus Post author

      Hi John. This is a great question. I think there is such a tendency, yes. And in a way it makes sense. The mechanical world view is so strongly engrained in our society that it is not easy to let go of it. Even I often catch myself trying to figure out causal chains even though there might be none. So what is needed in the end is a fundamental paradigm shift on the level of our mental models and deeply held beliefs of how the world works (is it a machine or is it an organic evolving system?). Nora Bateson actually argues that the term system is so strongly contaminated with a mechanistic world view that we need a new and different term. She suggests the term ‘symmathesy’. Have a look at this post and the further reading from there:

      1. John Turley

        Thanks Marcus. I’ll check out the link and keep reading. Thanks for the Nora Bateson tip.

  2. Sven Gehlhaar

    Thanks a lot Marcus, great thoughts. Lets see what my colleague Roberto will say about that.

  3. Rifaat Bashir

    thanks, what amazed me the statement(who are you to go to a country and imposed behavioural change over there. Really brilliant

  4. Sachin Gupta

    Great blog and I agree strongly. At a meso-level we need to appreciate complexity and design multiple interventions that can shift the system (not control it) – however at a micro-level it’s easy to forget that we need to think like market actors, and not development intervenors. This requires a much greater degree of empathy and humility than the concept of behaviour change implies. You didn’t mention market failure once – and I do feel too often practitioners over-complicate and forget simple economics .Maybe an entrepreneur doesn’t want to change their ”behaviour’ because the risk/return doesn’t stack up – our job is to understand and shift the underlying incentives in the system – not the individual actors. There is also a danger that behaviour change begins to sound like traditional direct interventions . However… To be fair I think ‘behaviour change’ advocates were primarily talking about meso/policy/system level disruptors….. Not market actors. The danger is if that crucial distinction is forgotten .

    1. Marcus Post author

      Hey Sachin, very important points you are making! The way I perceive it is that the behaviour change advocates often focus too much on the micro level, trying to identify the ‘wrong’ behaviours of enterprises and think what would be a ‘better’ behaviour (e.g. having an extractive business model vs. having one that nurtures collaboration and long-term relations). But for me the entry point to change this behaviour is the institutional incentive structure, which is often a result of the meso level arrangements – indeed, in Mesopartner we are strong advocates that the meso level is the main place to work when trying to overcome market failures. So if the institutional environment that produces the incentive structure is changed, then entrepreneurs will automatically change their behaviour. These incentive structures can be formal or informal – and most often are a mix of it. If we can change these institutional structures, I would call this transformational change (a.k.a. the elusive ‘systemic change’ we often look for).

      1. S Gupta

        Yes agreed. Meso level institutions such as rules or supporting institutions. However the other route to institutional change of course is some kind of disruptive business model innovation, either in the core transaction or supporting functions. So yes, I agree – too much focus on actor level behaviour change and not enough attention to how you identify and nurture genuine business model innovation with the potential to disrupt and bring about that systemic change.

  5. Anna Garloch

    thanks for this insightful post, Marcus. It prompted a lot of healthy internal discussion here at ACDI/VOCA among a number of colleagues. Look for an informal summary of that discussion in the coming weeks!


Leave a Reply