Existing indicators of drug-related policies and harms - official statistics, treatment data, and so on - are difficult to interpret and to compare across jurisdictions. We propose a conceptual framework including key dimensions of policy and harm. The framework is designed to guide creation of a new dataset in which existing empirical data can be used to assign city-specific scores on policy and harm dimensions. By an interpretive process involving local experts and a coordinating body, we believe that existing data can be evaluated and synthesized to derive these scores. We also propose a cumulative case-study method of querying the dataset to test hypotheses regarding relationships between policies and harms.