The intensity of land use is of vital importance from an agricultural policy perspective, because it determines food production, income opportunities and environmental impacts arising from agriculture. And: space matters, as the provision of ecosystem services and disservices is highly spatially dependent. For example, regional clusters of intensive land use can be a threat to unique local landscapes and ecosystems. Farmers’ decision-making regarding land use intensity and resulting spatial patterns are not well understood. This paper aims to uncover driving forces behind land use intensity strategies, especially exploiting spatial clustering and difference across space and time. We use spatially explicit census data on 2018 for Swiss agriculture and focus on extensification decisions in grassland production. This dataset allows us to account for neighbourhood effects, i.e. spillovers across farms. We use a set of variables controlling for common, measurable conditions, as well as an instrument that controls for regional habits and cultural backgrounds. Within this setup, we identify a significant neighbourhood effect among farmers. Our findings highlight the need for a socio-cultural dimension in agricultural policymaking.