Evaluations of participatory development efforts improved somewhat between 2007 and 2012, generating some new evidence. However, the evidence base for most questions relevant to policy remains thin, and far too little attention is still paid to monitoring and evaluation. Project localizing development: does participation work? 14 design continues to show little appreciation of context, and inflexible institutional rules fail to internalize the complexity inherent in engag ing with civic-led development. Unless these problems are addressed, participatory development projects will continue to struggle to make a difference. Local participation tends to work well when it has teeth and when projects are based on well-thought-out and tested designs, facilitated by a responsive center, adequately and sustainably funded, and conditioned by a culture of learning by doing. To ensure that it supports projects with these characteristics, the World Bank and other donor agencies need to take several steps: • Project structures need to change to allow for flexible, long-term engagement. Patience is a virtue. • Project designs and impact evaluations need to be informed by political and social analyses, in addition to economic analysis. • Monitoring needs to be taken far more seriously. The use of new, more cost-effective tools, such as short message service (SMS)–based reporting, could help enormously. • Clear systems of facilitator feedback as well as participatory monitoring and redress systems need to be created. • Most important, there needs to room for honest feedback to facilitate learning, instead of a tendency to rush to judgment coupled with a pervasive fear of failure. The complexity of participatory development requires a high tolerance for failure and clear incentives for project managers to report evidence of it. Failure is sometimes the best way to learn about what works. Only in an environment in which failure is tolerated can innova tion take place and evidence-based policy decisions be made