Several studies came out in August with evidence for what works and what doesn't work in civic tech, especially regarding anti-corruption programming, open data and institutionalization. Plus there are several important of research collections, how-to methods guides, and oodles of case studies.
Last week had interesting findings on political mobilization, now with brain scans. Lots of discussions about appropriate methods for measuring government performance, improving statistics and facilitating adaptive programming. Useful resources from the Engine Room and Beautiful Rising. Oh, and Disco!.