Abstract Indicators of the public provision of old age social care are routinely recorded in England and have been used for diverse purposes including performance monitoring. Despite long‐term policy guidance promoting more homogeneous service provision, large variations can still be observed between local authorities (the providers of state‐funded social care). Our aim is to better understand such variations in a small selection of key aggregate indicators. Drawing on multiple data sources and pursuing a two‐step strategy, we first assess the explanatory power of a set of structural predictors and then add to the models a set of specific care management ‘process’ predictors. We find that structural factors beyond the control of local authorities explain a considerable share of the observable variation. The additional explanatory power of care management characteristics is small in comparison. Therefore, our findings suggest that caution must be taken when aggregate indicators of service provision are used for performance monitoring purposes, as a degree of autonomy over outcomes may be implied which in light of the empirical evidence is unrealistic. Past attempts to influence the aggregate pattern of service provision – apparently seeking greater ‘territorial justice’– are likely to have had adverse implications for service users and the uniformity of service delivery across England. Questions are raised about the adequate role of central government in a policy environment characterized by longstanding local government responsibility.