ABSTRACT
Development projects are-rather, should be-policy experiments. From experimentation and piloting through replication and dissemination, projects can serve as tools to explore, test, refine, and monitor the implementation of policies (Brinkerhoff and Crosby, 2002; Chambers, 1997; Rondinelli, 1993). There is a tendency in current development thinking, however, to overlook projects as being too micro level to achieve significant results. This is a misconception, because it is precisely the ability of projects to generate interactions and synergies among all levels of intervention-macro (policy) , meso (institution), and micro (community, enterprise, household, and individual)- that defines their potential to serve as effective policy experiments (Jackson, 2000).
