Key messages

1. Development policies and programs increasingly operate in situations of high complexity and uncertainty.
2. There are growing efforts across the sector to design, plan and implement more adaptive responses that are more relevant and appropriate in such contexts
3. At the heart of effective adaptive programming is the capability to gather, interpret and use knowledge, information and data in real-time
4. Strengthening this capability requires a positive enabling environment—including tools and methods, structures and processes, leadership and management and culture and mindsets.
5. For adaptive development to succeed, there needs to be greater attention and investment in both knowledge, information and data systems and in appropriate enabling environments 


In October 2015, USAID, mSTAR, and the Institute of Development Studies (IDS) convened the Learning to Adapt workshop in London. The aim of the workshop was to convene a set of actors from diverse backgrounds and disciplines to share experiences and generate new ideas on the requirements for adaptive programs and policies in development, and how best to meet these. The workshop conveners had a particular interest in the often-neglected capability of gathering, interpreting and using knowledge, information, and data in real-time, and how this capability could be strengthened to enhance adaptive management efforts. The dialogue took place with a purposefully selected diverse group—which included development policy specialists, practitioners, software innovators, data scientists, innovation experts, participatory development researchers, and behavioural scientists—to ensure that the insights and ideas were as widely validated and applicable as possible. The workshop represented a first attempt to bring these actors together to establish common ground and shared dialogue. This briefing note draws out key reflections and lessons from the workshop for policy-makers, practitioners, and researchers with an interest in adaptive development.[1]

[1] This briefing note is accompanied by a longer Workshop Proceedings which presents the discussions from the Learning to Adapt workshop in more detail. 

The adaptation challenge (and why it matters)

The Learning to Adapt workshop began with the premise that complexity and uncertainty are becoming watchwords in international development and humanitarian work. To be effective, policies and programs need to be able to navigate and maintain their relevance in the face of a dizzying host of shocks, pressures, and stresses. There is growing emphasis on adjusting and adapting development efforts to a variety of rapidly changing contexts, needs, and interests. The key driver is for development to become more innovative and creative in the face of seeming insurmountable and interlocking challenges of poverty, deprivation, hunger, inequality, climate change, conflict—the list goes on.
While there are many different ideas, concepts and intellectual traditions within this loose movement, the attendees of the Learning to Adapt workshop identified a number of common principles in terms of how these ideas are being applied.
Underpinning much of these efforts is a common focus on knowledge, learning, information, and data. Specifically, being adaptive means:

  • Taking a more ‘wide angle’ lens to understand the issue or challenge faced
  • Understanding the wider context of the issue, drawing on insights of those closest to it
  • Specifying interventions that are relevant to that context, and working hard not to over-specify based on imperfect understanding
  • Supporting the ongoing and real-time sensing of information, insights and ideas through the implementation process
  • Evaluating or in other ways making sense of this information, insights and ideas in ways that are relevant for the program or policy
  • Ensuring more appropriate, contextually relevant decision making, and encouraging more appropriate changes and adjustments at a strategic and tactical level.

At the same time as this emphasis on learning for adaptation, pressures on aid budgets—both upward and downward—and the related need to demonstrate results and show accountability to donors, was seen as placing a greater emphasis on planning and control. The growing need to pre-specify interventions speaks volumes about the need for the development enterprise to demonstrate its value and worth. Although it has been argued (including by the present author[2]) that this needn’t be the case, in reality the emphasis and focus on result-based management can and often does run counter to the needs of understanding and navigating uncertainty and complexity. Instead people are asked to ignore the uncertainties and complexities and to plan and fix too much too early.
In many cases, the response to uncertainty has not been to invest in adaptation, but rather to reinforce a linear results agenda more firmly. This mentality is embedded in the practice, language, and culture of development. This has led to what might be described as the ‘adaptation gap’ in development and humanitarian work: the gulf between the need for more adaptive iterative responses and the cash, commitment and competencies to adequately meet this need. 
Development organizations do not only need to get better at adaptive management, but they also need to create space and support for such efforts. Adaptive management efforts have to thrive, and simultaneously create and foster the conditions in which they can thrive. This places major challenges on such interventions, because the failure may often be as much—if not more—attributable to the lack of an effective enabling environment as to the skills and abilities to be adaptive. For example, large-scale programs focused on adaptive responses for private sector development may be based on appropriate and relevant principles of dynamic decisions, but if the reporting and accountability requirements are not also adjusted, it becomes very hard to maintain the program as intended.
As one participant in the Learning to Adapt workshop put it, using terminology from the world of software development, this is akin to attempting agile techniques in a system that is built around waterfall approaches (see Box 1).

[2] Barder, O. and Ramalingam, B. (2012) Complexity, Adaptation and Results, blogpost,

Thornton, D (2015) The Government should give a realistic commitment to Agile, blogpost,

As a result of attempting to meet these conflicting needs, many adaptive interventions to date have ended up as curious hybrids: trying to meet the needs of adaptation on the ground, while also meeting demands for uniformity and results from on-high. This pressure frequently leads to both being done imperfectly. However, the adaptation challenge is not going unacknowledged or unaddressed.
The workshop participants identified a number of key lessons from adaptive responses that have been successfully undertaken in development. These include:

  • Managing tensions: Adaptive management in development contexts is not straightforward, but needs to continually manage a number of common tensions, including between decentralized decisions and information flows, and between existing organizational systems and processes (such as planning, contracts, and procurement) and more innovative programs.
  • Data and power: the use of data in adaptive management is fundamentally shaped by political dynamics: issues of power and control fundamentally shape what data is permissible, works, and is useful. In many adaptive challenges, better data can highlight the problem and the need for adaptation, but is not enough to motivate appropriate and adequate decisions and solutions.
  • Networks as safe spaces for learning: for many organizations, the inter-organizational collaborations were essential for creating space, time, and justification for adaptive management. It was through the development of communities of interest that many organizations were able to convince senior managers of the need for adaptive programming. Better collaborative learning would make more ‘natural experiments’ to try out a few different projects and see which work best where.
  • Adaptation favors the bold: The importance of leadership was emphasized, in particular where current generations of leaders need to be courageous enough to change the systems and processes that underpin their own sources of power and legitimacy. Leadership should motivate a process of continuously challenging how organizations operate and finding ways of changing the rhetoric of success and failure.
  • Having the right people:  The provocative point was made that development organizations may not have the right personality profiles to do adaptive work: “Do we fire everyone? Are we wishing for something that is inherently impossible?”
  • Making it work: There was general agreement that the justification for adaptive management was clear, but within a very specific group of individuals and networks. The key was to see how adaptive management could be incorporated into the mainstream of organizations, and used to change everything from decision-making to resource allocation. A fundamental issue was how to demonstrate the benefits of adaptive management, in terms of program effectiveness, organizational benefits, and development results.


A diverse and disconnected response (and how to join it up)

It was also apparent from the workshop participants that this adaptive challenge was not going unaddressed. Rather, there is a clear sense of a number of different alliances and networks with a central focus on strengthening adaptive responses in development. For the champions of these ‘adaptive movements,’ there is an urgent awareness of the need for development and humanitarian organizations to address the adaptation gap in a more concerted and strategic fashion.
Their efforts might be led by networks of researchers, practitioners, policymakers, or collaborations across all three. They often have good connections into specific policy and operational spheres, such as governance, health, or financial services.
For example:

  • The call to ‘do development differently’ through better understanding of the political economy of change and reforms;
  • A move towards better delivery in donors and implementing organizations, accompanied with a greater focus on improving the theories of change underpinning programs and policies;  
  • The rise in systems and complexity approaches to development as a way of better understanding and navigating wicked problems;
  • The focus on agile approaches as a core element of a move towards principled digital development efforts;
  • The growing interest in behavioral and ‘nudge’ based approaches and the need to iterate programs and policies to fully realize their promise; and
  • A renewed focus on participatory methods to engage communities and primary stakeholders, both to generate feedback for adaptive decision making within agencies, and to support communities’ own adaptive development.

What was apparent is that these approaches are being used across all the sectors of development and humanitarian work, from epidemic response to community driven development, from government reforms to private sector development, and across the longstanding humanitarian-development divides. At the workshop there was also good representation from across these different sub-sectors in international cooperation.
These efforts are all in different ways focused on building capabilities for adaptive management at the individual, team, and organizational levels. The Learning to Adapt workshop brought together experts from across these different fields, and led to a common view that there were considerable benefits from each approach, as well as potential benefits from their increased and enhanced cross-fertilization.
The discussions identified that given the multi-layered nature of development and humanitarian efforts, there is a need for multiple perspectives and framings of the adaptive challenge, with different actors coming together at various points to combine and re-combine their ideas in novel ways. Some may draw on political economy analysis, others use network science; some will employ behavioral insights, while others will use participatory methods; some will use user-centred design while others will employ theories of change.
Increasingly these approaches are coalescing, seeing some exciting cross-disciplinary applications: for example, the World Bank has been using agent-based simulations from complexity sciences, together with participatory approaches, to better understand the political economy of governance reforms in real-time. However, on the whole, this diversity has not been capitalized upon. Many different actors and networks have been pushing for very similar changes in their own particular furrows, discovering the same challenges, learning similar lessons, but not benefiting from cross-fertilization.
Workshop participants also worked to apply these combinations of methods to specific, real-world adaptation challenges being faced by selected participants, in the form of virtual client groups. These rich discussions showed that, in keeping with the philosophy of adaptive management itself, there is no single best method or approach for doing all of the above, and that the problems of adaptive programming are such that no one actor or set of ideas can resolve them alone. More positively, real potential was seen from actively working to combine different approaches.
In the workshop itself, participants saw particular value from the following combinations:

  • Participatory methods together with digital development methods could help to engage communities more effectively and at much larger scale;
  • Behavioral sciences together with political economy analysis can give better insights into the limits and possibilities of individual and collective decisions; and
  • System and complexity sciences could be drawn upon to help to make theories of change more diversified and relevant to a broader range of challenges. 

The common thread of knowledge, information and data (and how to enhance it)

Throughout the workshop, participants noted that knowledge, information, and data were often seen as a necessary but not sufficient condition for adaptive management. While this was accepted, it was also noted that paying too much attention to the broader enabling environment can often mean not paying enough attention to the core principles of what adaptive management actually is, what it looks like, and how to do it in different contexts. In many cases, this has meant not focusing enough on the knowledge and learning processes that lie at the heart of adaptive management.
There are numerous reasons for this. First, many may see adaptive management as being about creativity and originality, and not steeped in ideas such as evidence and data from older eras of the aid effectiveness debate. Second, and relatedly, knowledge, information, and data may speak to notions of rigour, and for many, adaptive management is more about flexibility. This is based on a mistaken idea that ‘rigour’ is the same as ‘rigid’, and also the fact that there are few credible ways of demonstrating ‘flexible rigour’. Third, there has been insufficient investment in knowledge, learning, and data that is focused on adaptive decision-making. Much of the information-related investment in development is aligned with the existing program management approach, which gives prominence to up-front designs and end-of-program evaluation. For knowledge, learning, and data that supports adaptive management, there is a need for more operational, real-time investments. In fact, the larger portion of information and evidence-related investments may need to be in the process of development programs, rather than its design or evaluation. This is a major challenge to the mindset and culture of most aid organizations.
Fourth, and building on the previous point, the focus on knowledge, information, and data cannot be disconnected from the people who hold and use that knowledge. Specifically, the development sector may not have the right people in the right positions in operational settings to lead adaptive responses. In fact, operational leadership is poorly thought about in development, where the tendency has been to promote away from the field. In many other settings where there is a premium on adaptive behaviors, there is a strong tendency to promote towards the field, to push decision-making capability, creativity, and intelligence closer to where they can have the most effect on the ground. To adapt one observation, ‘the brilliant surgeon stays in the operating theatre, the awesome lawyer in the courtroom, the excellent engineer on the project. But development experts often work far from the problem.’
With this in mind, participants worked together to develop a framework for how to best meet the knowledge, information, and data needs of adaptive programs and policies, using consensus methods. This highlighted the following three areas as centrally important:
1. Understanding the context
The first and primary purpose of knowledge, information, and data in adaptive programs is to ensure decision-makers can keep in touch and up to date with realities of their programs and the changing contexts in which it operates. This can span multiple aspects, including the social, political, economic, environmental, and technological. Historical knowledge of the past is also vital, and a major challenge for development organizations and their typical ‘next 5 years’ orientation.
Knowledge is also vital to better understand the relevant ecosystem of
actors, the connections between them, and how these are changing over time. In particular, this ecosystem needs to be seen as spanning organizational and disciplinary boundaries. Of central importance here is the need to identify the primary stakeholders of the programs in question—and to engage with them throughout project to identify their motivations, problems, and possible solutions. It may well be that these people are in the development system, but it is far more likely that they will be national, regional, or local actors. Establishing trusted and credible information flows to learn from these ‘customers’ is of vital importance—requiring the combination of sound tools (e.g. surveys, focus groups, immersions) with trust and effective relationships.
2. Effective experimentation and learning
Knowledge and learning also plays a more directly operational function in adaptive management. Specifically, it helps to filter ideas and test them in ongoing iterative cycles of experimentation, learning, reflecting, and adaptation. This enables development actors to better understand and respond to the central problem or challenge they are trying to address. If this effort is not underpinned by knowledge, information, and data, then adaptive management becomes chaotic and undisciplined.
This operational learning is often alluded to as having a greater appetite for failure, and the idea of failing fast, but participants identified this as problematic. Instead of failing fast, there was a need to fail at appropriate rates for the program in question: ‘fail fast but not too fast’.  
In keeping with the contextual aspects mentioned above, there is a need to ensure that the learning from experiments is diverse in nature and source—otherwise there is a considerable risk of group-think and biases.
3. Optimized data, used well
Effective adaptive programs are those with a strong focus on data and information, and how these are used in ongoing cycles of decisions. For truly effective adaptive programs there is a need for openness in relation to data sources and types, discipline and clarity in their selection, and intelligence in their use. This raises issues of balancing data quality and speed, to ensure this matches the decision-making needs around the problem in question. One participant suggested that there was a need to identify the minimum viable quantity, quality, and frequency of data for effective use. This will of course vary from program to program: data collection, management, and use will need to be contextually relevant and fit for purpose. There is also a need to invest in resources to make sense of data as it emerges, and to empower different actors to make use of data in different ways. Ultimately, better data should lead to more timely, relevant and accurate decisions for the program as a whole. However, it is also important to acknowledge that data use behaviors are diverse, and different individuals and teams will engage in different ways. In particular, there are in most adaptive programs ‘super analyzers,’ ‘super forecasters,’ and ‘super users’ of data, and these individuals need to be positioned so that their skills are of benefit to the program as a whole.
In addition to these three central knowledge, information, and data aspects, participants also identified the need to have the right people and incentives to undertake adaptive interventions. This would not be easy or straightforward, given the fact that this has not been a criteria for recruitment, promotion, or retention in development organizations to date. There was also a vital need to create appropriate organizational environments that help to support the core knowledge processes that underpin adaptive responses. Finally, across organizations, there was an identified need for effort to build common language and frameworks for adaptive management that can be used to further collective learning. 


Challenges and opportunities (and possible ways forward)

Over the course of the Learning to Adapt workshop, a number of challenges and opportunities were identified (see table 1).

Table 1: Challenges and Opportunities

  • How to do adaptive management effectively;
  • How to enable agility in a linear systems;
  • How to change the practice, mindset and language of development;
  • How to bring different adaptive movements together;
  • How to focus on core competency of knowledge, information, and data; and
  • How to adapt the way in which knowledge, information, and data is used in development.
  • The growing acknowledgement of complexity and uncertainty and the awareness of the adaptation gap;
  • The widespread interest in adaptive responses;
  • The opportunity to link up different ‘adaptive movements’;
  • The power of combining different methods and approaches;
  • The potential to use knowledge, information, and data to ask new questions about adaptive development and how to achieve it; and    
  • The scope for common frameworks (akin to the logical framework or sustainable livelihoods), that can support the wider dissemination of adaptive management. 

A vital way forward is to acknowledge the importance of real-time, relevant, operationally focused knowledge, information, and data. An enhanced focus on knowledge, information, and data in adaptive management could lead to improvements in the way that development organizations deal with such challenges. But this cannot happen without closer investigation of the role that knowledge, information, and data play in successful adaptive efforts. Specifically, this area raises the need to better understand:

  • How to learn effectively from needs and contexts, and the forms of knowledge, information, and data that can be most useful;
  • The kinds of capabilities and tools needed to anticipate and interpret problems as they emerge, and to better deal with uncertainty and change;
  • How knowledge, data, and information can be employed to support better, more decentralised leadership and decisions in ways that are consistent with overall ambitions and values;
  • How knowledge, information, and data can contribute to new networks and partnerships; and
  • How knowledge, information and data can contribute to new business models for development interventions. 

Conclusions and Recommendations

As the common experiences related by participants at the Learning to Adapt workshop identified, when change towards a more adaptive response happens, it is usually because of both a technical and a political awareness of the need for change. Such changes happen not in the mainstream of organizations, but rather at the margins, ‘under the wire.’ Delivering on the ambitions of adaptive programs requires safe space, honesty and trust, and at a more fundamental level, cannot happen without questioning and challenging assumptions.
Based on the discussions at the Learning to Adapt workshop, there was clearly a strong interest in, and clear awareness of the need for, more cross-disciplinary effort to rise to the challenges of adaptive development programs and policies. It was clear development and humanitarian organizations need to do a lot of things to become more adaptive, but important first steps are to let go of old, comfortable, but ultimately unhelpful linear models of change and development, and existing, silo-ed approaches to the problem. Instead, both a more dynamic mindset and a more collaborative approach are sorely needed.
The workshop provided a first step in this direction, through its cross-disciplinary focus, and grounding in novel ways of thinking about and understanding adaptation. The challenge ahead is about how to turn this into a serious coalition of change. One important suggestion was about using the tools of influencing and advocacy strategies, and for the adaptive management movement to practice what it preaches and take a more strategic and politically smart approach to changing the sector.
This means, at a minimum, reforming and retooling organizations, and striking a better balance between accountability and learning than is currently the case. This also means pushing for changes at a personal and a professional level. There needs to be positive ways of encouraging learning, experimentation, and risk-taking throughout the system. Adaptive management, in short, needs to be seen not as a one-off or an aberration but as a core competency in development practice.
The following five recommendations are a distillation of the next steps and future lessons from the workshop discussions.

  • Case studies, case studies, case studies. Although there is a wealth of research in development, it was noted that decent case studies of adaptive management are thin on the ground. This can and should be rectified, through more shared investment of time and resources to documentation and sharing. To be successful, these case studies will need to be able to capture the highly tacit nature of the adaptive development experience. Starting to share these in a small and low-cost way might be a useful initial step, using blogs and other forms of social media.
  • Peer-to-peer learning and co-working to enhance adaptation in action. The workshop identified the considerable potential for peers to support learning in this space, both in terms of providing fresh insights into ongoing challenges, working on new challenges together, and in terms of lending greater legitimacy to internal debates. Creating an action learning group from across the workshop participants was suggested as an immediate positive measure.
  • Networking across different movements to achieve synergies. There is a clear need for the different efforts represented at the workshop that are in different ways pushing for adaptive development to do more in common, share explorations, and to join forces where possible. At the present time, the whole of these efforts are not greater than the sum of the parts, but they could well be. In this regard, the diverse backgrounds and perspectives of the Learning to Adapt participants are a real potential strength—if they can be tapped and capitalized upon. A useful step might be to organize an adaptive development conference to convene these groups in critical mass.
  • Responding to windows of opportunity. There have been many different opportunities to frame the adaptation challenge as central for development and humanitarian practice—from unexpected regional crises such as Ebola, to delivering against ambitious new long-term global frameworks such as the SDGs. Each of these represents an opportunity for the adaptive management community to make its voice and perspective heard, and to convince a wider network of stakeholders. It may be worth initiating such an exercise, using advocacy tools, to see what the opportunities and messages might be.
  • Ask: whose adaptation counts? The final point is less a recommendation, and more a centrally important consideration. Instead of thinking of adaptation as focused on international agencies and their counterparts, it is important to question whether this is the adaptation that really counts for the changes that are sought by international development organizations. Even brief reflection should serve to highlight the central importance of enhancing, supporting and encouraging the adaptive capabilities of actors in developing countries: governments, civil society, private sector, and communities. Without such a focus, the adaptive management movement risks becoming another short-lived development fad. If these actors are placed in the foreground, with a focus on their adaptation, autonomy and creativity, then the adaptive development movement may have a chance to help transform how development works for the better.


This report is made possible by the generous support of the American people through the United States Agency for International Development (USAID). The contents are the responsibility of Institute of Development Studies and do not necessarily reflect the views of USAID or the United States Government.