Development practitioners need to learn about changes in their local context and adapt interventions accordingly. Seems like common sense, right? But much development thinking and practice is stuck in a linear planning model.
In 1998 David Mosse argued that projects are commonly seen as “closed, controllable and unchanging systems” — and today this problem remains pervasive. The messiness of social change is a seen as a risk in project documents, rather than the starting point for interventions. Too often, the systems, tools and mental models we employ block rather than encourage critical reflection and change in interventions.
There are growing calls for this to change as part of broader attempts to “do development differently.” In our new paper, we argue that learning and adaptation are two sides of the same coin. We show how learning can be integrated throughout development programs. This means moving away from upfront analytical papers gathering dust on shelves, and away from monitoring and evaluation focused purely on upwards accountability. It means getting serious about learning.
Yet the ever present danger with such a revitalized push is that it quickly becomes shortsighted. Based on increasing engagement with donors and practitioners on these issues, here’s what to avoid as the push for learning and adaptive programing gathers pace.
1. Not taking it seriously.
Prioritizing learning within development programing has serious implications for the cost and time of an intervention. Simply making the environment more permissive might not be enough; it will require active donor support and nongovernmental organizations taking responsibility too.
Read related stories:
In short, learning has to be in the budget, and time needs to be allocated to it. Equally, taking learning seriously means taking adaptation seriously. It’s incredibly frustrating to generate fresh knowledge and have no realistic way of using it. In that respect, learning practices in development organizations need to be purposeful and relevant to the learner.
Reflecting on learning and using it to inform decision-making can be uncomfortable, time-consuming, and may reveal problems or failure — so the benefits of learning and using learning have to be direct and clear.
We should strive for so-called double loop learning, in which learning leads to a shift — often significant — in individuals’ worldviews and organizational culture. Double loop learning would encourage the kind of reflection that could prevent the rest of the pitfalls below.
2. Ignoring history.
An increasing focus on “learning by doing” and experimentation should not encourage ignorance of a country’s history, nor similar past development interventions. Even if we accept that ongoing analysis is more valuable than upfront expert contributions, our starting point still is not zero. Building strong, locally embedded teams who understand and work within a country’s history is key.
Knowing what has been tried before, how it played out and why is also crucial. For example, in Malawi, there have been 79 unsuccessful public sector reform programs, which gives pause as to whether another will work. In day to day learning practices, this means focusing on retaining learning (what has happened before), as well as taking in new information (what’s going on now).
3. Thinking short term.
Big intractable development problems take a long time to resolve — far longer than a typical development program lasts. Progress is rarely linear and far from guaranteed. A key part of learning for adaptation is taking “small bets” in development programs. The strong and valid rationale here is it’s better to try a few things out, learn as you go, and decide what works best, rather than go all in with something that you don’t know whether it will work out.
However, the danger is that this encourages a short-term mindset of constant learning, piloting and trials, in which sufficient time is not given to knowing whether there is indeed traction on the big changes the program hopes to create. Learning strategies therefore need a balance between rapid front-line implementer learning that leads to small-scale adaptation, and broader strategic thinking that takes into account the long-term nature of development processes.
4. Firing up the anti-politics machine.
Adaptive programing has emerged alongside calls to think and work politically. However in the United Kingdom’s Department for International Development, one danger is that being ‘flexible and adaptive’ becomes the dominant narrative and the need to be ‘politically smart’ slips from view. Research conducted by Overseas Development Institute on a roads program in Uganda, for example, indicates that a program that is flexible and adaptive but fails to work with the grain of local politics will be limited in impact.
For learning, this means putting political relationships at the heart of most development learning strategies by encouraging everyday political analysis. One example is the State Accountability and Voice Initiative, which conducted political economy analysis with state teams, supported by national experts; their findings influenced program decision-making. These kinds of practices are the bedrock of politically smart, adaptive programing.
5. Creating elitist and purely managerial practice.
The focus on flexible and adaptive programing takes many development professionals, as well as researchers, into a world that is beginning to feel like management consultancy. Perhaps that’s a good thing: effective programs need strong structures and systems to encourage appropriate learning for adaptation. However, this can also feel elitist, as if the program and its learning centers on the world of donors and big corporate implementers. The latter are often more likely to be able to demonstrate organizational structures and processes that can convince a donor of their ability to learn and adapt. This can shut out the learning of local partners, even though they are most in touch with contextual realities — and what’s going on in the country matters much more to them. As with politics, there is a danger that being locally led slips off the agenda for donors. This is a question of aid effectiveness: as David Booth recently argued in relation to two DfID programs in Nepal and Nigeria, programs need to be locally led to the furthest degree that will support effectiveness.
6. Falling into the traps of the results agenda.
As I’ve argued on Devex previously, the results agenda can be unrealistic, misleading and time-wasting. In some ways the push for what we might call a “learning agenda” is a reaction to the (often unintended) negative consequences of it. In our paper, we argue that a degree of accountability for learning may be appropriate. Indeed, we offer an annexed log frame that embeds learning processes as the main outputs — which if taken up more broadly would offer a radical departure from counting things that often tell us little about whether change is happening. However we also need to be wary of stifling learning, forcing it to be all about accountability to donors. Encouraging learning is not about donors, but helping us understand how positive change happens and what we might do to support it.
Join the Devex community and access more in-depth analysis, breaking news and business advice — and a host of other services — on international development, humanitarian aid and global health.