Abstract
Community‐based coalitions are popular structures for creating community benefits. But desired effects have been reported only for single cases, the overall documented evidence to date for positive coalition outcomes being weak. Methodological obstacles may account for much of the missing evidence, and research possibilities for addressing these obstacles are suggested. Alternative interpretations include the positions that coalitions in general are not effective intervention mechanisms, that traditional scientific methodology is poorly suited for capturing fine‐grained coalition outcomes, and that coalitions and similar collaborative organizations are too complex to be adequately evaluated by the methodology that is now available.