Car Clubs as Networks: Linking Interchangeability and Lot-Level Utilisation
Year: 2026
Type: article
Abstract: Car clubs (car sharing) are promoted as a way to reduce private car ownership and transport emissions, yet evidence on what shapes their use rests largely on self-reported member surveys. This study conceptualises a station-based car club as a network derived from revealed patterns of vehicle interchangeability. Using Enterprise Car Club hire records for Edinburgh (50,257 individual hires in 2019), bipartite member-lot networks are projected into unimodal networks of lots linked when hired by the same members, and standard centrality measures are calculated for each lot. Ordinary least squares models relate these measures, alongside distance to the nearest tram stop, to lot-level hire counts. In the analysed data, Strength (weighted degree) alone explains 67% of the variation in hires, compared with 7.1% for tram proximity, and all centrality measures are positively associated with use. Strength derived from 2018 hires is also associated with 2019 hires beyond each lot’s 2018 hire count, and the results are robust to count-data, robust-standard-error, and spatial-error specifications. Strength is closely tied to each lot’s number of distinct users, and an association independent of user-base size was not established. Four spatially coherent communities of lots are identified from member co-use alone. Network-based analysis of hire records offers operators and planners a practical and repeatable way to monitor car club provision and to inform its expansion.
Source: Sustainability
DOI: 10.3390/su182211682
Authors: Eran Livne, Ian Philips, Greg Marsden
Institution: University of Leeds
Language: English
Subfield: Transportation
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $2,896
OpenAlex ID: W7235125142
