Part of what this suggests is that the role of donors and others aiming to support change at scale may be more to build connections and linkages that allow key local actors to adjust broad efforts to the dynamic political realities of contexts. While we tend to think of this as "our project should be adaptive" it's much more "our programming needs to up the odds of adaptiveness", describing a world where even if we do great, our projects will not directly drive success, though they may play critical roles over time. This idea is simple to express (see the sidekick manifesto) but is so hard to really internalize, that the work of development projects is to change the odds and build the enabling conditions and infrastructure toward other folks' eventual successes not to make change directly.
David Jacobstein
DCHA/DRG Policy, Coordination and Integration Team
United States Agency for International Development