At the recently concluded meeting of the American Evaluation Association, Michael Bamberger, Penny Hawkins, and I led a Think Tank session on unintended consequences of development programs. One premise that motivated the workshop was the belief that unintended consequences are ubiquitous in development, and that evaluation would be well served by developing better methods for predicting and reacting to those kinds of surprises. A second premise was that to develop those methods, it is necessary to develop a community of interest around that topic. The Think Tank was a first step. The use of this blog to help the community develop is another step.
As a beginning we want to start collecting cases. Our hope is that people will post cases and that others will respond with suggestions about what might have been done. We realize that posting cases can be a delicate matter because discussing unintended behaviors might reflect badly on programs, sponsors, and evaluators. But we hope that people will be able to sufficiently anonymize their posts as to cover identities, while retaining the information needed to allow the rest of us to comment in an intelligent manner. It’s an experiment. Let’s see what happens.
Evaluation Uncertainty 
Thanks for the posting as a tool for continuing our conversation. One aspect of the conversation at the conference that I found helpful was distinguishing (a) unintended consequences — positive or negative, and (b) unexpected consequences — positive or negative. For me this was a useful 2×2 framework to include when planning an evaluation.
I agree with the unintended/unexpected distinction. From a systems point of view, I’m not so sure about positive/negative. Good or bad, they spring from the same dynamics.
Cutting to the chase (since I think the first conclusion about this needing more attention is clearly right), what should we do to find more of the unintended/unexpected consequences? I currently tell my students there are three main paths: (i) literature search; (ii) extremely thorough observation and interviewing with this specific goal (this needs special training); (iii) use a new staff member of a new consultant to operate in goal-free mode. In all cases of (iii) that I know of, this has turned up one or more significant new consequence. Anything I should add?
Michael Scriven
I have always been a big fan of DTC’s idea of consulting with program critics. Also I think there are some systematic scoping methods I talk about in my book that will help.
During the conference I’ve presented some work using participatory video & most significant change (MSC) as part of a participatory evaluation of girl programming. Using the MSC question without a domain or framing it to include negative consequences helped us discover unintended positive & negative consequences of programming. Here are the results in case you are interested: http://www.videogirlsforchange.org