Where have we got to on adaptive learning, thinking and working politically, doing development differently etc? Getting beyond the People’s Front of Judea

alan hudsondave algosoProps to Dave Algoso (left) and Alan Hudson at Global Integrity for making the effort to compare and contrast 9 different initiatives that are all heading in roughly the right direction in reforming aid

Aid, development, and governance practitioners increasingly recognize that change happens through iterative processes (trying, learning, adapting the approach taken, and trying again) as opposed to the linear assumptions that underpin much of the sector (do more X, get more Y). Pre-planned, linear, blueprint approaches to change fail in the face of contextual variations and shifting political interests. Progress occurs when efforts are more adaptive.

Several recent initiatives have brought together practitioners, policymakers, researchers, and others. Having multiple efforts as part of a movement can be very powerful, but the lines between these initiatives can be blurry; their missions, approaches, and stakeholders all overlap. As a result, there is a risk that participants in these initiatives end up unnecessarily splintering in competition over jargon and brand position (think Monty Python’s

Are we DDD or TWP today?
Are we DDD or TWP today?

People’s Front of Judea).

To help clarify the overlaps and differences, we’ve compared nine different initiatives. Important caveats: This is non-rigorous and each initiative is evolving, with multiple stakeholders involved, such that pinning down the essence or core strategy of each is nearly impossible. But we are sharing the general gist of each based on our informal observations.

Each initiative in brief

Let’s tackle them in batches of three, moving from the general to the specific.

Across the whole development sector

Initiative What is it?
Doing Development Differently (DDD) Community of researchers and practitioners convened by the Overseas Development Institute (ODI) and Harvard’s Kennedy School. Has a core manifesto calling for development to focus on locally defined problems, tackled through iteration, learning, and adaptation.
Thinking and Working Politically (TWP) Semi-regular convening of representatives from various donors, think tanks, and international NGOs that discusses the use of politically aware approaches to aid and development.
Global Delivery Initiative (GDI) Cross-donor collaborative (spearheaded by the World Bank, which currently serves as the Secretariat) to deepen the know-how for effective operational delivery of aid and development.

Focused on governance and accountability

Initiative What is it?
Global Partnership for Social Accountability (GPSA) Funds and convenes CSOs and governments on social accountability initiatives. Established by the World Bank in 2012.
Making All Voices Count (MAVC) Five-year program (2013-2017) funded by multiple donors (DfID, USAID, Sida, and Omidyar) to find, fund, and learn from innovations that support accountable governance.
TA LEARN Community of practice composed of transparency and accountability practitioners from many countries.

Focused on targeted aspects of the aid/development sector (but not governance/accountability)

Initiative What is it?
ADAPT(analysis driven agile programming techniques) Collaboration of two major NGOs (Mercy Corps and International Rescue Committee) to identify, develop, and spread the use of adaptive management approaches in complex aid and development projects.
Smart Rules DfID’s internal operating framework for programs, emphasizing how the agency adapts to and influences local context. Rolled out in 2014.
Collaborating, Learning, and Adapting (CLA) USAID’s framework and internal change efforts for incorporating collaboration, learning, and adaptation at its missions and implementing partners.

What we left out: This list is hardly exhaustive: among others, we left out Feedback Labs (focused on the specific issue of citizen/beneficiary feedback), Principles for Digital Development (guidelines for integrating technology into development programs), and the Open Government Partnership (multi-stakeholder initiative to open government).

Actors: Who’s involved with each?

Many of the actors in these initiatives overlap, which has provided some cross-pollination of ideas. Researchers and think tanks (especially ODI) have been active in all of these initiatives. But it’s the mix of ownership by donors and practitioners that most shapes each initiative.

Mixed stakeholders: The governance/accountability initiatives (TA LEARN, GPSA, and MAVC) all have mixes of actors from multiple countries, though naturally most partners work in related issues such as open governance, transparency, citizen engagement, etc. The DDD community has a similar mix, with a slightly broader set of issues represented (though most of their work is still grounded in governance, institution-building, service delivery, or policy reform).

Leaning toward the donor side: Three of these initiatives are heavily anchored in donor agencies: CLA at USAID, Smart Rules at DfID, and GDI at the World Bank.

Implementer-focused: The ADAPT initiative is the only one led by and focused on non-donor agencies, in particular the IRC and Mercy Corps, though they have also engaged donors and others through pilot projects and events.

Focus: How central is adaptive learning? And working politically?

These initiatives see adaptive learning in different ways. To some, it’s a central driver of how change happens and a core strategic pillar. Others use adaptive learning more tactically, as a way to improve traditional approaches on the margins.

We see this as closely related to another concept: the political nature of development. Especially in governance and accountability work, navigating the fog of political interests—both hard to discern and likely to shift—is one of the core reasons why adaptive learning is so critical.

However, these initiatives don’t promote both concepts equally. We can expand on a spectrum that Duncan previously shared for TWP (from “evolutionary” to “revolutionary” uptake) into two dimensions.

algoso and hudson fig

Activities: What does each initiative do?

These initiatives’ strategies—whether explicitly stated or inferred from their activities—fall into four general approaches:

Practices: Identifying, developing, and sharing cases, methods, and tools. Nearly all of these initiatives have developed sets of case studies, framed with necessary context (organizational, political, etc), as an antidote to the reductive tendency of “best practices”.

Principles: Thought leadership, conceptual framing, and evidence. Nearly all of these initiatives are also working to provide the terminology and frameworks for furthering ideas.

Community: Cross-organization network building. Conferences and workshops create space for sharing practices and testing frameworks. Only GPSA and CLA actively encourage virtual exchanges separate from physical meetings.

Reform: Internal change at lead partner(s). ADAPT, CLA, and GDI all pair internal change at their lead organizations with external changes in the sector. Smart Rules has an exclusively internal focus.

adaptive-planOne approach seems relatively underutilized: Only GPSA and MAVC make grants of any kind to support project implementers who are developing practices. (Though some of the other initiatives also fund research.)

Conclusions (?): Where do we go from here?

First, we need clearer examples of adaptive learning in practice. We have a growing set of case studies, but few include a compelling counterfactual for what would have happened in the absence of adaptation.

Second, making the case with counterfactuals requires a bit more conceptual clarity. It’s not clear that we’re always talking about the same thing when we discuss “adaptation”, “learning”, or “working politically”. We need to clarify how adaptive learning leads to change.

Relatedly, one of the key concepts requiring clarification involves ownership over learning and adaptation. Despite the mix of stakeholders involved in these initiatives, few meaningfully distinguish between learning by external actors and by local actors, and the ways one can support the another.

Finally, this mapping suggests a potential for greater connectivity across the initiatives. There are some actors complexity signoverlapping several of these, but there were significant gaps until literally just a few months ago. Unfortunately, this is tied up a bit in issues of organizational ownership: some initiatives are only weakly owned, with no steering hand from the lead organizations, while others are so identified with a single organization as to leave others with no entry point. Efforts pushing the sector in broadly the same direction should push together.

Please let us know what you think of this mapping. This was not designed as a rigorous final say on these initiatives, but rather as a conversation starter. What did we get wrong? What did we leave out? Is it worth building on this with a more rigorous and perhaps broader version? Would a one-stop shop, or regular updates on what’s going on in the various initiatives, be of use?


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.


15 Responses to “Where have we got to on adaptive learning, thinking and working politically, doing development differently etc? Getting beyond the People’s Front of Judea”
  1. Jake Allen

    This is useful, but i have some disquiet, partly about this, but perhaps more about this as representative of the bigger discussion.

    The first problem that this shows is that there are too many different things saying the same thing, or similar things branded differently. This is an inherent ingredient of imminent dysfunction, in my limited experience. Linked to this, though not explicit in this piece is that it’s clear that because there’s lots of ‘things’ that are focused on adaptive programming, and which therefore need to be followed, because they are ‘things’, that these are now becoming exactly the normative, bureacratic obstacles they are aiming to be working to challenge.

    One of the causes of this, i think, is that the crucial aspect of this area, which is to empower and devolve management and decision making, isn’t happening yet (it’s talked up laudably in the Smart Rules, at least). If that did happen, one of the main bellwethers of its functioning, would – irony klaxon – be that peple would be politely but firmly putting the doctrine to the side (after understanding its essence), and getting on with things.

    But to finish with some micro nit-picking (is there a macro version?), the chart has MAVC as revolutionary, and as a tactic to tweak current practices, which seems contradictory. And secondly, the straight line and the squiggly line end up in the same place which only happens in the magical land of the project completion report 🙂

  2. Varja Lipovsek

    Thanks Dave & Alan. This could be helpful, yes. But not for purposes of synchronizing, linking, mainstreaming, or god forbid coordinating. I think multitude of initiatives is fine, and there is strength in people trying out different angles, approaches, etc. We don’t need to pare it down. But, a one-stop-shop to just keep up with who’s doing what would indeed be useful — at least for my kind of audience — i.e. I’m an implementer organization which is interested in understanding how various others are doing adaptive learning, and possibly trying out some of the models for ourselves. So I would not worry about having a terribly rigorous exercise; you could even have a crowd-sourced exercise. It’s mostly about having an open conversation as to what is being tried out where and by whom, and how they are faring. If we could weave into this conversation some honesty about the pull & push factors (doing adaptive learning because you’ve been told to, vs. able to choose / shop around for an approach that fits), as well as some honesty about how it’s going (including how it’s not going), then I think this is a conversation that I (my organization) would definitely be interested in following, and contributing to.

  3. Nice overview – and good take away fom the side event you both ran at the GPSA meeting. And you managed to do this with nary a mention of human centred design, innovation, participatory approaches or complexity theory! (although these ideas underpin most of these initiatives in some way and it might be useful to see how they take on ideas from these disciplines)

    More seriously though, for me the different initiatives are all a sign that we realize that our current linear approaches to planning and implementing development are not delivering the results we want, and that there is now serious efforts to find a better way. At the same time the fact that there are so many initiatives is also a sign that 1) we don’t know what approches are going to work 2) the initiatives we do try are often driven by our own internal organizational cultures and individual worldviews such that people often want to create their own flavour of change. Having a diversity of similar initiatives is probably great at the “emergent” stage of thinking about different ways to manage aid projects but at some point there will need to be some convergence of ideas or rationalization of approaches if any of this is going to have a meaningful impact on how the majority of development projects are managed.

  4. Catherine Dom

    I like the idea of regular updates on all these i.e. a sort of meta-COP. Not the “one stop-shop”… Never been sure what this exactly meant, but I doubt it would do as well as keeping plurality but more cross-learning etc. And… YES, need to be clearer about whose learning. Since the start of DDD I’ve kept asking “who’s doing”. A bit disappointed that this is only one of the key concepts. For me, it’s THE key.

  5. Cathy Shutt

    There is a group focusing on democracy assistance that has been facilitated by International IDEA which is pursuing some relevant approaches as alternatives to rigid RBM tools. Some interesting experiences were shared at a workshop in Stockholm last week. Though International IDEA is no longer facilitating, the group is intending to continue. Contact me if you want contact details.

    I do quite a bit of work with MAVC and am curious about why you have placed it where you have. Revolutionary uptake and a shift away from traditional aid approaches? Can you elaborate?

    • In reply to Jake’s and Cathy’s comments about MAVC: The thinking is that MAVC is ahead of the curve on the political nature of development, given its focus on citizen voice, accountability, etc; that puts it on the right side of the horizontal axis. However, on the vertical axis: MAVC doesn’t have adaptive learning at the core of its overall approach, or at the core of the approaches it encourages among its grantees. It encourages learning, of course, but learning as a tactic for improving or transferring knowledge, rather than learning as core to way change happens. Hence, it’s lower on the vertical axis.

      Still, there was some informality to this mapping. Reasonable people could place it elsewhere. I don’t think what we’ve done is contradictory though—rather, the contradiction (as I see it) would be in MAVC’s approach: Can you think/work politically without putting adaptive learning at the center? I’m not so sure you can.

  6. Heather Marquette

    Thanks, Dave and Alan! Can’t agree more that some sort of synthesis and more regular coming together would be useful. My worry, though, is that it masks what could be important differences, even within groups. I’m part of the TWP Community of Practice, but I’d never put myself on the revolutionary end of the spectrum. My own belief is that this ‘field’, insofar as it can be called that, is far too young to justify revolution, and we’ve not begun to understand things like unintended consequences, taking small-scale reform projects to scale (or if that’s even possible or desirable), or even where we may find small ‘tweaks’ to everyday practice that could overcome some of the barriers to better working (see http://oxfamblogs.org/fp2p/what-are-governance-advisers-missing-with-political-economy-analysis-how-can-they-do-better/ for a bit on this). Yet there are other members of the CoP who believe that the revolutionary end of the spectrum is exactly where we should be, and our focus should be on finding ways to redesign programmes to fit the model that seems to be emerging in some of the literature. The great thing is that we can come together and have big discussions and debates, share evidence and be on the same ‘learning journey’ together.
    I agree that there’s a need to better work together in this space, but I worry about what could be lost. Any ‘coming together’ would crowd out debates that are vital in any newly emerging field of study and could stifle innovation around the edges. In all sorts of areas, friendly, collaborative competition is no bad thing for generating creative ideas after all. I reckon a bigger problem, which you don’t mention, is that there’s a lot of commercial pressures around this work: too few funders (because outside aid, who else is interested?) and a lot of incentives to flag-plant. So how do we create better incentives for collaboration and what role do/should funders play in this?

  7. Thank you, Duncan, for synthesizing this information. I would also second the idea of having regular updates on what’s going on with different initiatives. As someone who works on CLA, it’s helpful to have an understanding of the adaptive learning landscape as well as where CLA falls and how it can improve relative to other approaches.

  8. Thanks for the comments.

    We’re glad to hear that people see some value in joining the dots amongst the various conversations (and note the caution about bringing things together too hastily). We’re happy to help in that regard as we also try to ensure – as per Catherine’s comments about “whose learning?” – that the conversations are based on and support practical experience and action. Somewhat relatedly, and in response to other feedback we’ve received, we’re in the process of doing a short note outlining where people can look for examples of adaptive learning in action. And, we’re having a little think about how we can help people to easily stay up to date with the various conversations.

    Couple of related things going on in Washington DC next week – the first meeting of a USAID-led “Practical Adaptation Network” and a workshop on “Mapping the landscape of adaptive practice”. https://www.eventbrite.com/e/operating-in-complexity-mapping-the-landscape-of-adaptive-practice-tickets-25765751055

    Oh, brief rejoinder to Dave on Making All Voices Count (MAVC); there is certainly an aspiration within MAVC to make adaptive learning more central to the initiative. See for instance the Research & Evidence Strategy from the team at IDS http://www.makingallvoicescount.org/news/unpacking-making-all-voices-counts-research-evidence-strategy/ Indeed, a project we’re doing with MAVC – with real-time support for adaptive learning – will put that into practice.

  9. Brandy Westerman

    It is helpful to have this overview of the inter-related initiatives as well as the discussion of how they each relate. I agree that the emergence of these independent efforts suggests a trend in the right direction, and yet, I wonder if the significant overlap of stakeholders engaged in these efforts is itself a concern. Inclusion of prominent voices contributes to cross learning, but is it at the exclusion of others, especially practitioners directly wrestling with the constraints of the industry? The complexity of the web of initiatives and associated jargon and acronyms has itself become a barrier to more widespread participation in the dialogue. As a result, despite the importance placed on evidence, each effort struggles to build the evidence and participation at a field level that is necessary for real change in practice and effectiveness.

  10. Katie Wiseman

    Thanks Dave, Alan and Duncan – useful mapping and I like the conclusions.
    I’m not sure ‘different things saying the same thing’ is necessarily a problem though as long as there is connectivity and collaboration – and what they’re saying is meaningful rather than jargon. I think this is starting to happen though must admit to some ‘Life of Brian’ style meetings in DFID debating acronyms and language (and lots of ‘is this anything new’, ‘this is just good programme management’, ‘we do it anyway’).
    I think we’re getting somewhere on shared understanding and most important practical implications for what this means for how we work in DFID and our own political economy. Lots more to do, for sure but we’ve got a strong internal collaboration going led by our Better Delivery Department with the Governance Dept in Policy Division and Research and Evidence bringing in staff from Finance, Procurement, Heads of Profession, country teams etc. Interesting challenges coming up on how you work adaptively in long and complex delivery chains and programming agreements; pooled funding arrangements and within results-based payment frameworks.
    We’ve realised a need to go beyond the formal ‘smart’ rules to build up the evidence and ‘case law’ for what happens in practice during programme design, procurement, annual reviews and negotiations over results frameworks and log frames etc. Agree with the need for better evidence.

    We’re just entering into a new collaboration with USAID on Global Learning for Adaptive Management. From a recent internal blogpost:
    “Complex development problems are often best solved by programmes which systematically learn about their environment, and then adapt their work to deliver impact. However, adaptive programmes require new ways of working. Senior leaders must create an environment which permits and promotes complicated work and calculated risk taking. In return, senior responsible owners need to assure they are making effective, evidence-based decisions about how to generate maximum impact. But how can we do this? How can we generate robust information quickly enough to make the right decisions? What quality standards should we apply to the information we use?
    Improving evidence use for management decisions is seen as key to managing risks and increasing the value for money and impact of adaptive programmes. The partnership will aim to improve donor and partners’ ability to use monitoring and evaluation information to make timely and effective management decisions. The partnership will also focus on harnessing new digital technologies to improve timeliness and availability of evidence on programme operations. ”

    Yes to creating incentives for collaboration.
    But keen to hear from others on what you think DFID and other donors can do to help (further) incentivise.

  11. Hi Alan, Dave, Duncan,
    Thanks for sharing the post. An interesting exercise which is useful for sparking more conversation on some of these issues.

    I just wanted to share a few open and honest reflections from my experiences as programme manager of the Research, Evidence and Learning component of MAVC in relation to the challenges of trying to work, and supporting others to work, in a more adaptive learning manner.

    MAVC is effectively supporting a large number of relatively small innovation projects over short time frames testing the innovative use of technology in efforts to address complex dynamic accountability problems. MAVC has been trying to support adaptive learning approaches at different levels – the programme level, the country level, and project level – and making the interconnections amongst the various levels. Each has its own challenges.

    ● Capacities – programme staff and grantees who may be used to more traditional ways of working can find it hard to change. It is important to consider whether people have the capacity and skills to work in a way which demands more critical reflection on their own practices and the realistic assessments of outcomes they might see. Many individuals are trained and organisations structured to implement, not necessarily to question, reflect and learn. The MAVC has offered many different opportunities to support reflection and learning but take up has been very patchy. I think part of this relates to understanding the need and value of working in these ways when addressing complex dynamic problems.

    ● Time – “old habits die hard” – more traditional linear ways of working, behaviours and the systems and mechanisms that support them are so ingrained that it will take time to move to other ways of working. Mainstreaming adaptive learning approaches isn’t going to happen overnight – the path, ironically, to learn and adapt practice to work in more adaptive learning approaches will take time.

    ● Our donors have generally been supportive of adaptive learning but are inevitably constrained in terms of their own pressures to hit targets and financial spends.

    ● Power – the power relations between different actors – grantees, grantors, donors – make it difficult to move to more adaptive learning approaches despite assurances that they will be supported. When pressure applies there is a tendency by all to resort to sticking to and reporting to “the plan” – or business as usual.

    ● Time – most projects supported by MAVC in isolation are too short term for adaptive learning to be especially useful within the duration of the grant. However, these approaches could be far more valuable when an MAVC innovation project is part of a longer term stream of work over a longer time frame.

    ● Different actors – different actors come with very different perspectives and ideologies and bring with them different understandings of learning and approaches to adaptation. Technologists will often equate agile approaches to software development with adaptive management of projects and programmes to address complex social issues. There is a tendency for what happens in “agile” tech development processes to be focused on single loop learning which might help you do what you’ve decided to do better, but doesn’t necessarily get at deeper double loop learning questions that might result in a better understanding of what the problem is that you’re looking to address.

    ● These different actors also have different exposure and experience of working politically. If you have spent much of your life addressing technological challenges, you’re going to find it challenging to suddenly need to work in politically savvy ways.

    ● The MAVC programme is complex with a consortium of very different donors, a consortium of very different implementing organisations, and a large and diverse range of different grantees all bringing their own perspectives and assumptions which can be exaggerated due to complex power relationships between them – making shared learning and adaptation difficult at different levels.

    I hope this doesn’t sound negative – I genuinely believe reflection, learning, adaptation and working in politically savvy ways is a promising route to more positive governance outcomes. For example, in relation to supporting innovation in this space – I don’t think anyone should pitch an innovation project in any way other than as a knowledge endeavour. Doing so allows one to frame an innovation as adaptive learning process, be more realistic about potential outcomes, and value knowledge outcomes as highly as other potential outcomes. However, I think we need to be alive to the multitude of challenges in implementing adaptive learning approaches in complex programmes and projects looking to address complex problems (see Craig Valters Six pitfalls to avoid in learning and adapting).

    Finally, some immediate feedback on your mapping and some of the comments – I am really curious as to where your data has come from for making the judgements and inferences about the different programmes and initiatives in the mapping and in your comments. Without understanding what evidence you base these claims and inferences on makes it difficult to know how to interpret it. For example, my knowledge of the internal workings of a MAVC would put it elsewhere on your map and I would certainly take issue with Dave’s comments as to the centrality (or not) of adaptive learning in different areas of MAVC’s approach – it’s totally unrealistic to expect the range and diversity of MAVC’s grantees to all to be working with learning at the core of their work – I would say grantees are across a broad spectrum depending on their own experience and capacities. MAVC has been actively trying to encourage and support them recognising where they are on this spectrum.

    Best wishes,

  12. Hi Alan, Dave, Duncan,

    Thanks for a very useful overview of some of the initiatives that are trying to bring about a change in how we go about complex reform programmes. I am a project manager working in international development, rather than a technical expert, and have been working on the DFID funded Legal Assistance for Economic Reform (LASER) programme. This programme has set out to test what ‘flexible and adaptive working’ and ‘problem driven, iterative adaptation’ has meant in practice – with some good results I will add! See our website http://www.laserdev.org for more details on what we have done and achieved.

    In the LASER programme we – in partnership with DFID – have tried to explore how a programme guided by what I will call adaptive programming / iterative adaptation principles looks and operates in practice, and how we need to adapt and change what we do operationally to react to real time learning. What I have found lacking in discussions in the area is the acknowledgement that the way we design, procure and manage programmes will also need to change fundamentally, if we are to succeed in implementing programmes that aim to achieve greater impact through increased flexibility and adaptation based on learning what works. Without key changes in the way programmes are designed and even tendered, the accountability requirements we have to adhere to and the result frameworks we abide by, I am afraid this will end up being just another short lived ‘flavour of the month’ initiative – which would be a great shame!

    What is perhaps needed at this stage is to better understand how the management and oversight, as well as M&E and assessment, of these more iterative programmes could work – and to do that, we need to take risks, and indeed, learn through trail and error. We – as donors and suppliers – need to agree to take some risk,and indeed to invest in our own learning process (without being scared of ‘failing’ and Daily Mail type headlines!), and to go about testing new, sometimes radical approaches to management and oversight enable us to understand what the optimal balance between accountability and flexibility is.

    Looking forward to future updates on developments in this fast moving field!

  13. sr.marquez

    Is there an original piece or citation where we may find Algoso and Hudson’s comparison of the 9 different initiatives and approaches? Would be much apreciated!