What can urban transport policy achieve? How transport-policy-invariant characteristics relate to city-level car ownership and car use.

Most case and comparative studies discuss urban characteristics that can hardly be influenced by urban transport policies. However, these studies do not quantify the impact of the so-called contextual factors on car ownership and car use. This paper focuses on these contextual factors and their associations with car ownership and car use. Regression models show that contextual factors - such as city size, the share of households with children and the jobs-housing balance - explain a substantial part of the variance in car ownership and car trips in 44 German cities. Residual analyses reveal that neglecting contextual factors leads to biases in comparative studies and consequently in the evaluation of urban transport policies. Thus, there is a risk that cities with favourable contextual factors may be mistakenly viewed as successful examples of transport policy, whereas successful policies in cities with less favourable contextual factors may be overlooked. • Contextual factors explain a large part of variance in car ownership and car trips. • Residual analyses show: Neglecting contextual factors leads to bias in benchmarking. • Transport policies should be evaluated while accounting for contextual factors. [ABSTRACT FROM AUTHOR]