What Can Development Policy Learn from Evolution?
In this YouTube/Blog presentation, Barder argues that development problems are highly complex, and are not solved with detailed project designs by ‘development experts.' He describes development contexts as ‘wicked problems' [particularly characteristic of crisis and conflict-affected environments] where problems are:
- Difficult to define clearly
- Have many interdependencies, multi-causal effects
- Actions have unforeseen consequences
- Often not stable
- No clear solutions
- Socially complex
- Not the responsibility of any single organization
- Involve changing behavior
Barder makes a compelling case (with a brilliant short video) that development problems in these contexts should be solved by evolution not design. Evolution requires variation and selection. By this is meant that development processes should include multiple, plausible variations, and use a process of feedback and evidence to select those variations that are most effective. The presentation notes that development agencies distort this process by requiring upward accountability for performance, rather than incentivizing a focus on direct input and feedback from beneficiaries and front-line workers. The task of those working on development is not to attempt better designs to solve problems [such as increasing equitable access to safe education] but to make and use better feedback loops.