Territorial impact assessment: MOSA, an analytical model for structuring evaluations and integrating territorial understandings

Public policies increasingly promote interventions grounded in territorial specificities, yet evaluation practices remain poorly equipped to take territory into account. While territorial impact assessments (TIAs) are intended to address this, they rarely do so in practice. We argue that a key reason for this lies in the predominantly quantitative and top-down way in which TIAs conceptualize territory, despite its inherently multidimensional nature. As a result, important territorial assumptions underlying place-based policies risk remaining invisible. In response, we developed MOSA, a multidimensional analytical framework, informed by regional studies literature, and tested this with four French non-profit organizations involved in TIAs. We used two methods: a monographic analysis of documents and qualitative data collection through observations and semi-structured interviews with evaluators. We found that by making territorial assumptions explicit, MOSA can help prevent misalignment between evaluation approaches and territorial foundations, while offering the possibility of developing more complexity-sensitive evaluation practices.