Marginalised youth say ‘Enough!’ Guest post by Alcinda Honwana

November 4, 2015

Is China’s rise relevant to today’s poorest states?

November 4, 2015

Doing Development Differently: a great discussion on Adaptive Management (no, really)

November 4, 2015
empty image
empty image

Went to a fascinating workshop last week on ‘adaptive management’ hosted and designed by USAID as part of their work on complexity signKnowledge, Information and Data (see final para for more links) and facilitated by Ben Ramalingam, who has just started at IDS as their new digital, technology and innovation czar. A whole load of participants are going to write posts for this blog, which will go up over the next few weeks, but I thought I’d kick off with some personal impressions, not least because I’m doing a ‘brown bag’ staff talk today at USAID in Washington (on How Change Happens, what else?).

Firstly, what is adaptive management? According to Ben ‘It’s all about interaction and change: change is emergent and contextual; it relies on organizations having the right capacities and processes to generate novelty in day to day performance. It means bosses trusting people closer to the problem. In contrast ‘traditional management’ assumes you have the answers and you need to get them into place at the right time. AM assumes you don’t have the answers, and need a system to generate them as and when they are needed in any given context.’

Second, USAID’s explanation for why they schlepped all the way to London to discuss this was that much of the interesting discussion on smart/adaptive development is happening in the UK – see my earlier musings on whether Britain has become some kind of silicon valley-type cluster on aid and development.

Third, pleasingly inter-disciplinary – tech geeks, development programme people, donors, M&E types, academics, behavioural psychs, private sector entrepreneurs – and a great mix of creative tools and methods employed by Ben, from story telling to participatory games to consensus methodologies. That’s particularly important when people come from different disciplines, each with their own jargon, assumptions etc.

Fourth, I’m not a big fan of diagrams, especially the overcomplicated theory of change spaghetti variety, but two emerged in the discussions that really got me thinking.

USAID fig 1

The first, in a conversation with Robert Chambers, was a ‘data, adaptive management, participation’ triptych. Promising because each of the three points of the triangle compensate for the weaknesses of the other – participatory approaches ensure data is bottom up and relevant to people on the ground, not an exercise in extraction; and that adaptive management approaches don’t degenerate into more effective manipulation (see my concerns on the Doing Development Differently agenda); good data enhances both management and participation; adaptive approaches prevent rigidity creeping into data systems or participatory methods.

The second diagram emerged from a competition to develop a shareable framework on AM, and manages to get USAID fig 2beyond ‘are you for adaptation or logframes?’ polarities, by seeking to locate different approaches according to whether we have high/low levels of knowledge of the context in which we are intervening, or of the causative impact of a given intervention. That generates a handy 2×2: if you have high knowledge of both causation and context, then you can stick with traditional programming approaches; if you have good knowledge of causation, but context is uncertain or rapidly changing, the key is to have fast feedback and response systems in place, eg when doing cash transfers in a fragile state. If you have good knowledge of the context but are unsure on the causation, try lots of things – experimentation, monitor their impact, adapt and iterate (Matt Andrews’ problem driven iterative adaptation). If you have neither, try and escape from the bottom left quadrant by either finding a simple intervention that works (cash transfers again) or studying the context to get you towards the bottom right.

What do you think?

My overriding impression, which I shared in my closing remarks for the event, was that all of us pushing for adaptive development approaches should be thinking harder and smarter about how to change the aid sector. When we try and change the world we use a whole range of advocacy tactics that we know have worked in the past (eg finding champions, seizing windows of opportunity). And yet all too often we assume changing our sector is more of a technical endeavour (‘just show ‘em the evidence’). In its cross-disciplinary, dialogue-focused approach, this workshop was a great step towards a more collective approach to such change (building coalitions of ‘unusual suspects’ is another successful advocacy tactic) – let’s see if we can capitalise on it.

Last but not least – some cracking one-liners:

‘There are three kinds of social scientists: the ones who can count, and the ones that can’t’ (Robert Chambers)

‘Big Data is like teenage sex – everyone is talking about it, no-one knows how to do it, everyone thinks everyone else is doing it, so everyone claims to be doing it themselves’ (Ben R, definitely going to nick that and replace big data with theories of change…)

If you want to know more about USAID’s thinking on this, try this paper on its Ebola response, its Learning Lab or throw your hat into the ring for its Request for Proposals (RFP) on “Adaptive Programming Using Real-Time Data“, the deadline for which has just been extended to 20th November.

10 comments

  1. Can you expand on the claim we might have good knowledge of causation but not context? I would have thought that rapidly changing context would impact what happens, so would be sceptical of claims to knowledge here… Sure, there is evidence that cash helps families in fragile states but still lots of unknowns about what will work best.

    Also, I dont really understand the distinction between fast feedback (and then learning, then revision) and PDIA. I thought PDIA was exactly that?

    I guess what I’m saying is I’m not sure about the quadrants. Shouldn’t it all be PDIA? When is PDIA unhelpful?

    1. Fair enough Alice, it is a bit cryptic. Context = understanding the social, political and economic setting for any intervention; causation = being reasonably sure of the efficacy of a given intervention. So you know that cash transfers or vaccinations have a good chance of making a positive contribution even in contexts that you don’t understand (although there are exceptions of course).
      Fast feedback refers to feedback about the system – eg the communities have all fled from X to Y because of the violence, so we have to move resources to their new site. PDIA refers more to feedback about the intervention.
      Does that make sense? Nothing hard and fast, and of course there are blurred boundaries between causation and context, but I still found this 2×2 helpful

      1. I see. Maybe such diagrams simplify things. I don’t know if that’s a good thing.

        Continuing the theme of diagrams and theories of change, I’m not sure I’ve ever seen a useful visual representation of political economy. They tend to impose fairly misleading boundaries around nebulous, multifaceted phenomena, thereby obsfucating complexity, all in a bid to draw something pretty.

        Whoever saw one and then understood the world better?

  2. ‘There are three kinds of social scientists: the ones who can count, and the ones that can’t’ — I am distracted by noticing that the ones who can’t count are also the ones who don’t know their “who” from their “that”; or perhaps those who can’t count are also those who objectify the one person. One of many iterations might be: There are four kinds of social scientists: those who can count, those that can count, and those who can’t. As always with such comments, I live in terror of making grammatical mistakes myself and, unjustifiably, in hope that mine will be treated more kindly.

  3. Interesting – any discussion on AM and ‘how to’ in any form of shape is useful. The 2 x 2 doesn’t work for me because 2 of the fields are so similar (identical?). I agree with Alice that causation and context aren’t necessarily differences that make the critical difference. Another issue is the ‘who has the knowledge’ question. By making ‘knowledge’ the unit of differentiation, it matters where knowledge is located. So I can have a very different analysis than you if my knowledge base is distinct. I find Cynefin or the Zimmerman frameworks more distinctive and therefore easier to use.

    Like the triangle by the way!

    1. Thanks Irene and others, the 2×2 got a bit of a battering at CGD yesterday as well, so I’m going to have to mull over these comments and either tweak or go back to other frameworks!

  4. “When we try and change the world we use a whole range of advocacy tactics that we know have worked in the past (eg finding champions, seizing windows of opportunity). And yet all too often we assume changing our sector is more of a technical endeavour (‘just show ‘em the evidence’). In its cross-disciplinary, dialogue-focused approach, this workshop was a great step towards a more collective approach to such change (building coalitions of ‘unusual suspects’ is another successful advocacy tactic) – let’s see if we can capitalise on it.”

    I agree, surely that was the main message of Irene’s closing chapter in The Politics of Evidence and Results in International Development:Playing the Game to Change the Rules? http://developmentbookshop.com/the-politics-of-evidence-and-results-in-international-development

  5. Hi Duncan – well done for coming with the 2×2 framework – don’t give up on it yet! It could be a useful thinking framework. I would try and integrate it with a learning framework such as the Johari Window as a mechanism for expanding knowledge (trying something when we are in an unknown zone).

Leave a comment

Translate »