Five lessons for supporting resilience in conflicts and recurring crises: evidence from SPARC’s retrospective studies of livelihood support

Few published studies assess resilience-building interventions several years after completion. Since resilience can be observed only over time, almost nothing is known about what really helps people cope with shocks. Retrospective learning is even more important in conflicts and recurrent crises, where life is less predictable, and where informal rules and informal power relations are more important. Without such learning, investments in resilience continue to be based on untested assumptions.

SPARC published five retrospective studies of projects in Ethiopia, Kenya and Chad, implemented by governments, United Nations (UN) organisations, international Non-Governmental Organisations (NGOs) and local NGOs. They covered water development in the drylands, climate-smart agriculture, public works programming and market-based pastoral development.

Common lessons, apparent across these diverse studies, combine to explain why so many investments fall short of their expectations and show how better to support resilience in places threatened by crises.

This policy brief summarises the lessons from the five retrospective case studies, revisiting projects some three to five years after closure. All were selected because the implementing agencies indicated they were successful.

Key message:

• Support for resilience should focus on evolutionary change, rather than on seeking transformational leaps. It is more important that changes are easy for people to adopt than that they lead to more ideal outcomes.
• Approval of any intervention should depend on evidence that it is based on an understanding of the informal ways in which people currently live. The impacts of the intervention on informal institutions – and vice versa – must be planned for. Projects should first look to improve what already exists.
• No single package of ideas can match the varied needs and priorities of everyone, even in a single community. It is better to provide a range of ideas that people can adopt and adapt.
• The social processes by which change is likely to happen must be set out explicitly, including, for example, how attempts at elite capture of the benefits of any intervention will be mitigated. Theories of change need to be taken seriously. They should not be reduced to diagrams. Setting out the assumptions necessary for interventions to work makes it possible for those assumptions to be monitored – and for management to be adaptive.
• There are no excuses for continuing to rely on the same mistaken assumptions in project and intervention design, simply because they are never checked. It should be standard practice to revisit a sample of interventions a few years after they have closed, to learn what really happened.