ABSTRACT
Many scholars have sought effective ways to encourage people to cycle more. A considerable amount of effort has focused on the role of dedicated cycling infrastructure. However, knowledge on the roles and interactions of other factors that are influential in addition to infrastructure in cities where the cycling network is well-developed remains incomplete. In this study, we examined how various individual-level attributes, namely psychological, habitual, and socio-demographic, in addition to the built environment characteristics relate to cycling behaviour of employees of the Erasmus University Rotterdam, in the Netherlands, where comprehensive cycling infrastructure is provided. Specifically, we investigated how these factors relate to being a cyclist or not, as well as how they are associated with regular and irregular cyclists. An online survey was conducted among employees of the university and logistic regression models were utilized for the analyses. Our results showed that the perception of behavioural control is consistently correlated with different cycling behaviour while controlling for socio-demographic and residential built environment factors. Also, we found evidence supporting a trade-off between attitudes and habit across different types of commuters. Socio-demographic factors such as gender and year of immigration to the Netherlands are only related to being engaged in cycling but not to increasing the level of engagement among employees who cycled already. The type of residential area and population/business density at destinations does not appear as a consistent covariate. We concluded that psychological and habitual factors play key roles in encouraging cycling in a city with an extensive cycling infrastructure network.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 For instance, aiming to attract 200,000 additional bicycle commuters, the Government of the Netherlands announced an investment of 345 million euros in 'ultrafast' cycling routes and additional parking facilities, while alternative strategies appear unaddressed (Government of the Netherlands, Citation2018).
2 Single reported working locations by participants that fell outside a 500 m buffer from the two selected university locations were disregarded, as assumed to be misreported.
3 The figures for gender were complemented by information provided by the Erasmus MC Intelligence Centre. Age comparisons exclude employees from the Erasmus MC, since this data could not be accessed.
4 The presence of a supermarket is limited to 500m due to the difficulty of carrying stuff home without using a car as reported in literature (Bostock, Citation2001; O’Dwyer & Coveney, Citation2006). A shopping place is assumed to be a delimited area offering diverse types of shops and services. The distance limit follows Cervero et al. (Citation2009).