Doing Problem Driven Work, great new guide for governance reformers and activists

complexity signOne of the criticisms of the big picture discussion on governance  that’s been going on in networks such as Doing Development Differently and Thinking and Working Politically is that it’s all very helicopter-ish. ‘What do I do differently on Monday morning?’, comes the frustrated cry of the practitioner. Now some really useful answers are starting to come onstream, and I’ll review a few of them.

First up is ‘Doing Problem Driven Work’, a paper by Matt Andrews, Lant Pritchett and Michael Woolcock. It turns previous work on PDIA – ‘problem-driven iterative adaptation’ – into a toolkit, aimed primarily at those involved in reforming governance from the inside, whether government reformers, or big bilateral and World Bank donors with access to the corridors of power. However, there are clear parallels with and lessons for the work of more ‘outsider’ NGOs and campaigners.

It starts by noting that while many reform projects have failed in the past, those that succeeded often involved a ‘problem focus’. Problems ‘force policymakers and would-be reformers to ask questions about the incumbent ways of doing things’ and ‘provide a rallying point for coordinating distributed agents who might otherwise clash in the change process’. ‘Good’ problems are urgent and can be easily addressed by those in the room. They often spring from crises or other ‘critical junctures’.

The first step for a would-be reformer is ‘problem construction’, which ‘involves gathering key change agents to answer four questions: ‘What is the problem?’, ‘Why does it matter?’, ‘To whom does it matter?’, ‘Who needs to care more?’ and ‘How do we get them to give it more attention?’’ Defining the problem is key: it’s no good having a woffly ‘corruption is a problem’ type statement – you need ‘a real performance deficiency that cannot be ignored’, like ‘we can’t get education and health care to these communities, because the municipal officials keep nicking the money’.

Once you have a problem, you can get started, with your doughty band of reformers:

What is the problem? (and how would we measure it or tell stories about it?)

Why does it matter? (and how do we measure this or tell stories about it?) Ask this question until you are at the point where you can effectively answer the question below, with more names than just your own.

To whom does it matter? (In other words, ‘who cares? other than me?) Who needs to care more? How do we get them to give it more attention? What will the problem look like when it is solved? Can we think of what progress might look like in a year, or 6 months?

The authors stress the importance of ‘authority’. For insider reformers like them, the key is to get political backing for the reform, preferably from the president or similar, which opens doors and aligns incentives.

Authority forms part of a ‘triple A change space analysis’, together with ‘acceptance’ and ‘ability’:

We can do better.....
We can do better…..

Authority to engage

Who has the authority to engage: Legal? Procedural? Informal? Which of the authorizer(s) might support engagement now? Which probably would not support engagement now? Overall, how much acceptance do you think you have to engage, and where are the gaps?


Which agents (person/organization) has interest in this work?

  • For each agent, on a scale of 1-10, think about how much they are likely to support engagement?
  • On a scale of 1-10, think about how much influence each agent has over potential engagement?
  • What proportion of ‘strong acceptance’ agents do you have (with above 5 on both estimates)?
  • What proportion of ‘low acceptance’ agents do you have (with below 5 on both estimates)?

Overall, how much acceptance do you think you have to engage, and where are the gaps?


What is your personnel ability?

  • Who are the key (smallest group of) agents you need to ‘work’ on any opening engagement?
  • How much time would you need from these agents? What is your resource ability?
  • How much money would you need to engage?
  • What other resources do you need to engage? Overall, how much ability do you think you have to engage, and where are the gaps?

The questions also highlight some of the weaknesses of PDIA – there’s no power analysis here, for example, whether some people who are currently not involved in decision-making could become so, and what might enable them to do so; who are the blockers, and do they operate through the exercise of visible or hidden power? A good power analysis would definitely lead to more refined tactics. There’s also a lack of a real systems approach, for example looking for positive deviants that are already working, or emergent hybrid institutions that combine new and traditional approaches. It all feels quite top down and traditional.

On the other hand, I like the bottom-up construction of problem and solution, and the authors have been out there, doing this kind of work for years, so the paper is full of practical examples.

I’d be interested in what people think.


Subscribe to our Newsletter

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please see our .

We use MailChimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to MailChimp for processing. Learn more about MailChimp's privacy practices here.


2 Responses to “Doing Problem Driven Work, great new guide for governance reformers and activists”
  1. Jacob A

    I think this paper *is* useful…up to a point. The main challenge that keeps occurring to me is with the first step; the ‘P’ if you will. This is, explicitly and obviously the first step in this process, and the bit that if you don’t do it properly makes all the following bits somewhat moot, or at least likely to achieve lesser impact (or full impact for something that’s not relevant). What I’m not really seeing as yet is this happening. The ‘P’ is still to a large extent prescribed, and those of us who are then charged with implementing, and implementing using ‘an adaptive and flexible approach’ as we always are nowadays, are therefore somewhat cut off at the knees before we even start.

    Now, to argue against myself, I know that donor organisations are stuffed full of intelligent, able and willing people who understand all this. The trouble is they are often as hamstrung both by the strictures of what is politically possible and desirable – both to the mythical ‘Uk taxpayer’ who checks up on us, and the recipient government – and by the bureaucracy and small ‘p’ politics of design, like business cases, that the ‘P’ process, involving a beautiful collection of reform minded individuals (who obvoiusly always exist and are purely motivated) who can spend time really diagnosing the problem, can’t or doesn’t happen. And even if it did, and a problem was really well identified and split into nice doable chunks, I wonder if this would even be sellable to the political masters who need to sign off on things.

    I realise this is a bit of a harsh critique, and that there are *some* examples of good practice coming through. But two thoughts come out: firstly, if donors are serious about this, just saying they are isn’t enough, adaptive programming requires a political, cultural and practical shift; secondly, for those of us who are ‘doing’ we need to recognise that this is no easy task, and to help who do want this shift to happen by finding ways within a predesigned Problem to re-diagnose the Problem properly, and show how this works better. And i suppose thirdly (what have the romans ever done for us), it would be strategic, practically and psychologically, to accept that its likely that this kind of politically difficult work will always have to happen a bit under the radar, or within a nice framework that allows it to be accepted, but with enough wriggle room to make it work.

    I’m just worried that the new fad will arrive just as we’re starting to make this work, and we’ll all have to change hats again.

  2. This is a time warp, no?

    When Germany introduced the logical framework, the problem orietation was central.

    All stakeholders participated and a full day was used to analyse the problems. Debating why, for who, and how each problem was related. The group defened together the central problem, and how the problems were related.

    The Problem tree.

    On basis of this, how to move to solutions was discussed, the solutions tree, which became the objectives, activities, short, the logframe.

    This placed in the minds of all stakeholders a problem focus and amindset to keep the eye on it.

    I remember i started to use so often this mindset i had to stop using the phrase : the problem is…

    So, is the latest movement addressing the same central problem the logframe was supposed to address?