What did I learn from a week discussing Adaptive Management and MEL?

June 20, 2018

     By Duncan Green     

Just got back from an extraordinarily intense week in Bologna, running (with Claire Hutchings and Irene Guijt) a course on ‘Adaptive Management: Working Effectively in the Complexity of International Development’.

The 30 participants mainly came from NGOs and non-profits, but with a smattering of government officials and consultants. What made the discussion different from previous AM chats is that they were largely involved in the nuts and bolts of Monitoring, Evaluation and Learning (MEL). Broadly, they fell into two camps: MEL people whose organizations find themselves increasingly being told to ‘be more adaptive’, and want to know what on earth that means, and others who have already bought into the importance of systems thinking and adaptive management, but now are wondering how to put it into practice in an aid sector that has lots of countervailing pressures – short project cycles, tangible and attributable results, the dreaded logframe and all the rest.

That meant that the week got much more into the practical side (here’s the outline – please steal Adaptive Management week plan). Students were asked to bring a ‘nut to crack’ from their day jobs, and Irene and Claire, who have decades of MEL experience between them, were able to provide lots of tools and ideas that can help navigate complex systems (while always eschewing a single AM recipe). I contributed my usual systems/ power/ how change happens waffle.

The main output was a set of ‘rules of thumb’ for putting AM into practice, generated by the participants. We’re still refining that – will post for comments once it’s ready. In the meantime, here are some random insights and observations from the week.

Important to think about when not to do AM – the Cynefin framework is really useful – trad linear approaches work in the simple and (with more thought) complicated quadrants. AM comes into its own in the complex one.

AM raises all sorts of problems and challenges for partnership, whether with organizations you fund, host governments that you are trying to work with, or partners in multi-stakeholder initiatives. Who decides when/what/how to adapt? What if they don’t want you to be adaptive, but would rather you just get on and deliver what you initially promised? Lots of host governments just want ‘training and trips’.

AM costs money: all this standing back, reflecting, real time evaluation, piloting, comms with donors and governments to keep everyone onside while the project evolves etc requires time and investment. Are donors and our own bosses ready for that?

The ‘toolkit temptation’ is alive and well: try as they might to stress that the key to AM is rising above the toolkit and ‘dancing with the system’, Irene and Claire were constantly being asked for them. That’s just how aid folk rock. I experienced the same feeling while writing How Change Happens and ended up compromising with a ‘power and systems approach’ that emphasizes ways of working, behaviour and questions, not processes or answers.

Do MEL teams need to be rethought, rebranded or even scrapped altogether and absorbed into project teams? Too often MEL is seen as low status bean-counting, internal police/executioner, or a combination of both. I used to be very rude about them til I realized that people like Claire and Irene are among the smartest people in Oxfam. In AM, the MEL function needs to emphasize the ‘L’ bit – how can MEListas become critical friends, incubators of new ideas and champions of AM, both within and beyond the organization?

Comms with donors, host governments and others are even more important in AM, eg during long inception and experimentation phases with few tangible results. But whose job is it? MEL teams are often happier with number crunching than telling inspirational stories. Should story telling become a recognized and separate function? Come to think of it, I guess it’s part of what I do at Oxfam, though we’ve never described it that way.

Trust is essential to AM. Donors need to trust that AM is not just bad management; host governments and communities need to trust that all this fiddling about is going to help them in the end. So how to build and maintain trust? There could be a role for ‘critical interlocutors’ – people at arm’s length from the project, who understand it and the issue, and can communicate with others with independence and authority.

But what happens if trust is abused? What happens when lazy or incompetent people do AM? It’s a bit like the US constitution – designed not for good governments but for bad ones (and getting a good stress-testing right now, as it happens). Is AM easier to game than traditional aid formats? On the margins of one recent discussion, I was told that unscrupulous management consultants now deliberately bid low to win contracts, then hide behind AM as an excuse to reduce their delivery targets.

AM in fragile/conflict settings. On the surface it looks like a no brainer – these are supremely complex places, so linear approaches are much less likely to work. But in practice, often aid workers can do little more than simply respond to events – riding the tiger/ ‘blaming ethe context’. That misses out on the other essential element of AM – reflecting on and revisiting assumptions and strategy.

How to spread the AM message? The aid sector typically propagates new orthodoxies through toolkits and protocols, but standardising those is anathema to AM. If not toolkits then what? Case studies? Charismatic champions (where’s the AM TED talk?). Exposing and ridiculing alternative approaches that fail? Other ideas?

Overall it felt last week like we have reached a crossroads: There’s now a ‘there’ there, a substantial case and growing body of evidence for when and how AM works, but numerous barriers and uncertainties remain. Maybe AM now needs its own more deliberate theory of influence – as one participant said: ‘what works best is waiting until we are 2 or 3 years into a project, when it hits a crisis and those in charge suddenly become interested in new approaches.’ Applying AM to the adoption of AM is a bit meta/turtles all the way down, but it may be what is needed.


June 20, 2018
Duncan Green