Official URL: https://doi.org/10.1080/02723638.2026.2726467

Vacancy has long been considered a key housing and urban development issue and many states produce datasets and official statistics that detail its extent and form. Yet, there is no internationally agreed definition or means of measuring vacancy. Nor is there a comprehensive understanding of the data politics and praxes that shape how states generate vacancy data. We first provide a typology of six ways in which vacancy has been measured globally (field visits, surveys, administrative systems, proxy measures, crowdsourcing, combination and modelling). Informed by interviews with key actors, we then document vacancy in Ireland post the 2007 global financial crisis, in particular focusing on how vacancy has been measured, and how the Central Statistics Office clarified and extended the vacancy data generated as part of the 2016 census, altered its methodology for the 2022 census, and introduced a new vacancy measure in 2023 using electricity consumption data. Drawing on our analysis, we contend that rather than using data denotatively there is a need for a reflexive approach to housing data that extends beyond a concern for data quality to consider the data politics shaping data production, how this changes over time, and its implications for analysis, interpretation and use.