Summary: After a VoxDev webinar with Rachel Glennerster and Stefan Dercon on the future of aid, I reflect on whether to retreat to radically simple interventions. I argue for embracing the complexity of social change and nurturing local learning systems. It may interest people thinking about aid reform.
- I understand the appeal of simple, evidence-backed interventions, but effectiveness depends on aligning with how change actually happens.
- Embracing complexity raises hard challenges, from causality to power and governance, and much good work is addressing them.
- Local learning systems should strengthen relationships and sensemaking, led by local actors rather than donors.
About 5 minutes to read the full piece.
A recent webinar featuring Rachel Glennerster and Stefan Dercon hosted by VoxDev explored an important question about development assistance, which recent aid cuts have brought into sharper focus: to what extent should development assistance focus on radically simple interventions which evidence (perhaps particularly favoured sorts of evidence) suggests are effective? (See also, this piece by Rachel Glennester and Siddharta Haria)
Rachel made the case for a radically simple approach, while Stefan - who agreed with Rachel that aid-recipient governments should focus on simple well-evidenced interventions - emphasized the importance of donors paying attention to the political economy dynamics of development and working in ways that support and strengthen governments in aid-recipient countries, rather than inadvertently making donors seem indispensable. The video is here (and below), and a write-up is here.
Watch on YouTube: Stefan Dercon and Rachel Glennerster on the Future of Foreign Aid (VoxDev)
In this piece I share my reflections on their discussion. My reflections are also informed by insightful questions and analysis shared by Rachel Leeds and David Jacobstein in the Local Systems Community 2.0 Google Group, which are well worth a look. Established by folks from USAID’s Local Systems Community but open to all, it’s a great group to be part of. You can join the Local Systems Community here, or from the Google Group’s homepage, here. I felt that the discussion might be of interest to folks beyond the Local Systems Community, so decided to share my reflections more widely.
Retreating from reality, or embracing the complexity of social change?
I understand the appeal of focusing on simple actions that evidence suggests are effective, as my 2011 review of key books on this theme by Dean Karlan & Jacob Appel, and Abhijit Banerjee & Esther Duflo makes clear. And I can see that the context of the brutal destruction of USAID might seem to require arguments for simplicity and control. However, given the complex realities of social change, I believe the wise course of action and the path that leads to effectiveness is to embrace that reality and deal with the challenges that entails. This seems to me a better option than retreating to focus only on “interventions” that seem to offer control and predictability - issues that “we know how to fix” as Rachel and Siddharta put it - and a supposedly clear path to accountability to parliament/congress and sustainable support from taxpayers.
Embracing the complexity of social change is, I believe, the wise course of action for two reasons. First, because much of the world is complex—uncertain, non-linear, unpredictable, and not amenable to cookie cutter solutions— with change emerging from ongoing interactions amongst multiple actors on an evolving landscape. And second, because it seems to me that the effectiveness of an approach to supporting social change is shaped significantly by the extent to which it is aligned with how change actually happens. Greater alignment with how change happens, leads to more traction, and greater impact. (See page 5 of ecosystems, emergence and social change, for a fuller version of this argument).
So then the question is, what are the challenges that an approach which is about embracing complex realities throws up, and how might they be addressed? For me, the challenges are ones about causality, entry points for engagement, evidence and evaluation, accountabilities, power and governance. These are big things. Tricky things. Not simple things. But they are important things worth investing time in exploring and finding ways to address, and which might, in the medium term, enable donors - and the countries they are supporting - to be more effective than they would be by retreating to simple things that can more easily be controlled.
Addressing the challenges of complexity
The good news is that there’s a huge amount of useful work going on as regards these challenges: to understand complex contextually-embedded causalities; to use political economy analysis to identify entry points for engagement in complex systems; to develop complexity-aware approaches to evidence and evaluation; to consider how to balance multiple dynamic accountabilities; to understand how patterns of power and incentives shape actors’ behaviours and the dynamics of social systems; and exploring how to nurture governance arrangements that emerge from and support the networks of relationships - and particularly the local systems - that shape communities’ and countries’ development trajectories.
There’s lots of interesting work going on in relation to this cluster of challenges. As well as USAID’s legacy of pioneering work on local systems, and Yuen Yuen Ang’s inspiring thoughts about going local 2.0, other things that spring to mind include:
- the work of Tom Aston, Marina Apgar and others on inclusive rigour, and related work on causal pathways by Jewlya Lynn, Lynn Coffman and more;
- the work of SOAS-ACE, the TransCap Initiative, UNDP, Climate KIC, the School of System Change and others on identifying entry points and crafting effective approaches - including systemic portfolio learning approaches - for engagement in complex systems;
- conversations - including those hosted by the UNDP MEL Sandbox - around complexity aware monitoring, evaluation and learning, and impact measurement in the complex landscape of systemic investing; and
- the work of Indy Johar and Dark Matter Labs, the Human Learning Systems initiative and more on new models of governance that are about stewardship, learning, and emergence, rather than control.
Nurturing local learning systems
A common thread or question to consider across these interconnected challenges seems to me to be about the extent to which the approaches taken to handling them are ones that strengthen relationships, support collaborative learning and sensemaking, and foster emergence in local systems, in ways that - shifting away from donor control - are led by local actors. In a way, this builds on and speaks to Stefan’s emphasis in the discussion on doing things that make use of and strengthen local systems, but goes beyond this by embracing not only political economy considerations (as per the title of the event), but also complexity and emergence and the role that collaborative learning and sensemaking, particularly amongst networks of actors in local contexts, plays in social change. So: Local (meaning causally close to the issue or problem, rather than simply geographically nearby), Learning, Systems.
I’m glad to be part of a community - the local systems community - that is well-placed to contribute to efforts to embrace the complex nature of social change, to address the challenges that that sort of approach entails, and to share the insights from this work in ways that will inform discussions about the future of aid and global development, as well as social change more widely. Rather than retreating to the apparent safety of simple interventions which give donors a sense of being in control, we can choose to embrace complexity, the local leadership that that requires, and the challenges and emergent opportunities that such an approach - a local learning systems approach - will bring.
PS: The May 29th piece by **Ken Opalo** on how development practice should approach the coming reforms in the aid sector includes consideration of Rachel and Stefan’s discussion. I’d say it’s a must read.
First published on LinkedIn, 28 May 2025.