I want to ride my bicycle: delimiting cyclist typologies

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rael20 Applied Economics Letters ISSN: 1350-4851 (Print) 1466-4291 (Online) Journal homepage: https://www.tandfonline.com/loi/rael20 ‘I want to ride my bicycle’: delimiting cyclist typologies Raúl Brey, José I. Castillo-Manzano & Mercedes Castro-Nuño To cite this article: Raúl Brey, José I. Castillo-Manzano & Mercedes Castro-Nuño (2017) ‘I want to ride my bicycle’: delimiting cyclist typologies, Applied Economics Letters, 24:8, 549-552, DOI: 10.1080/13504851.2016.1210760 To link to this article: https://doi.org/10.1080/13504851.2016.1210760 Published online: 25 Jul 2016. Submit your article to this journal Article views: 406 View related articles View Crossmark data Citing articles: 3 View citing articles ‘I want to ride my bicycle’: delimiting cyclist typologies Raúl Brey a , José I. Castillo-Manzano b and Mercedes Castro-Nuño b a Pablo de Olavide University, Seville, Spain; b Applied Economics & Management Research Group, University of Seville, Seville, Spain ABSTRACT Based around fieldwork in Seville (Spain), our article provides an empirical analysis with the aim of determining whether different typologies of cyclists exist depending on the type of bicycle for urban commuting (public bicycle/private bicycle). Our findings show that users of public bicycles are predominantly male, young, with a high level of education, and basically use the public bicycle for subsistence trips due to its easy intermodality; while private bicycle riders are mainly females who regularly make nonsubsistence trips and prefer a more flexible bicycle for their daily needs. KEYWORDS Trip purpose; bicycle ownership; public bicycles; discrete choice models JEL CLASSIFICATION R40; R42 I. Introduction A number of studies address the influence of personal characteristics and perceptions on cycling decisions (Vogel et al. 2014). However, little distinction has been made between Public Bicycle Sharing Systems (PBSS) and private bicycle users. The literature mostly focuses on individual determinants of bicycle ownership (Handy, Van Wee, and Kroesen 2014)while,as far as PBSS are concerned, we find studies on the impact of subjective perceptions of their features and quality (Bordagaray et al. 2015)since,asHekmanand Deisenroth (2013) state, choice behaviour and demand is clearly influenced by product quality. Our article aims to identify cyclist profiles for two cycling systems (PBSS/private bicycle) in the Spanish city of Seville, and to provide an empirical analysis of the influence of personal and environmental elements on the choice of bicycle type. The article is organized as follows. After this introduction, Section II presents the database and methodology. Results are discussed in Section III. The conclusions are set out in Section IV. II. Data set and methodology Two waves of surveys were randomly conducted among PBSS (SEVici) and private bicycle users in Seville during March–April, 2014. We obtained 1395 surveys from SEVici users and 451 from private bicycle users who were asked three types of questions based on covariates from prior literature (Table 1): (a) Socio-demographic factors. (a1) age (Fernández-Heredia, Monzón, and Jara- Díaz 2014): using binary variables: Age1520, Age2135 and Age3649. A value of 1 is taken when the cyclist’s age corresponds to the figures in brackets [15, 20], [21, 35] and [36, 49]. (a2) gender (Dill and Voros 2007): binary variable for male (1)/female (0). (a3) level of education (Castañon, Castañon, and Santos 2012): with 5 categories: (1) no primary education, (2) primary education, (3) secondary education, (4) three-year undergraduate degree, (5) higher education. The variables study_medium and study_high take a value of 1 when the level of education variable is 3–4 or 5, respectively. (a4) occupation (Handy and Xing 2011): with 6 categories: (1) employed, (2) self-employed, (3) homemaker, (4) unemployed, (5) student, (6) retired. The occupation_employed and occupation_- self_emp variables take a value of 1 when the occupation variable has a value of 1 or 2, respectively. (b) Travel preferences variables linked to journey frequency (Buck et al. 2013), purpose of cycle usage/ time patterns (Corcoran et al. 2014). (b1) bicycle_public (y) variable: takes a value of 1 (journey using SEVici) or 0 (journey using private bicycle). CONTACT Raúl Brey [email protected] APPLIED ECONOMICS LETTERS, 2017 VOL. 24, NO. 8, 549–552 http://dx.doi.org/10.1080/13504851.2016.1210760 © 2016 Informa UK Limited, trading as Taylor & Francis Group (b2) hour at which the trip is made. Two binary variables depending on observation distribution: hour11_15; hour16_20. (b3) frequency_route per week. (b4) purpose, with 5 categories: (1) work, (2) studies, (3) leisure, (4) sport, (5) other. The purpose_- work and purpose_study variables take a value of 1 when the purpose variable has a value of 1 or 2, respectively. (c) PBSS characteristics: number of PBSS stations/ per capita at origin/destination, and intermodality with other means of urban transport (Bachand- Marleau, Lee, and El-Geneidy 2012). (c1) other_mode variable: takes a value of 1 if the cyclist makes intermodal changes during the trip, 0 otherwise. (c2) station_origin and station_destination: density of SEVici docking stations (racks/thousand inhabitants) in the city district where the journey originates/ends. We use binary logit models. The observed variable to explain (y) is the cyclist’s choice between the private bicycle (y= 0) and the public bicycle (y= 1). Using this formulation, the likelihood that an individual imight choose a public bicycle is expressed: Pðyi¼1jxiÞ¼ expðαþxi0βÞ 1þexpðαþxi0βÞ(2) where αis a coefficient, βa vector of coefficients and x i a vector of variables. In order to account for the choice-based sampling design, the unknown αand β parameters are estimated using the weighted endogenous sampling maximum likelihood estimator, which provides consistent estimates (Manski and Lerman 1977). III. Results and discussion Table 2 gives the estimated coefficients and the effects of unitary changes of variables on choosing the public bicycle. Regarding age, cyclists in the 15–20 age bracket are more likely to use the public bicycle, which is consistent with Fuller et al. (2011). Concerning gender, males are more likely to use PBSS. Cycle usage in Seville is a quite recent phenomenon, so this result corroborates Fernández- Heredia, Monzón, and Jara-Díaz (2014), who find a more balanced ratio between male and female cyclists in countries with a well-established cycling culture. Eyer and Ferreira (2015) explain that females tend to be responsible for transporting their children and household responsibilities, and consider that public bicycles do not suit their urban mobility needs. A higher level of education seems to determine the choice of PBSS (corroborating Fishman, Washington, and Haworth 2013), perhaps because, as Bachand-Marleau, Lee, and El-Geneidy (2012) state, it could be considered to be more fashionable. The significance of occupation_employed suggests that salaried employees are more likely to use PBSS, perhaps because the private bicycle is more inconvenient (e.g., lack of parking at the workplace) (Bordagaray et al. 2015). Two peaks were found related to the hour variable. Taking the 8–10 time bracket as the baseline, Table 1. Descriptive statistics. Variable Mean SD Minimum Maximum bicycle_public 0.756 0.430 0 1 hour 12.971 2.829 8 20 station_origin 1.026 0.694 5.65E-02 3.326 station_destination 1.079 0.687 5.65E-02 3.326 frequency_route 5.316 3.210 1 47 other_mode 0.174 0.379 0 1 age 29.530 11.551 15 74 gender 0.623 0.485 0 1 level of education 3.650 1.043 1 5 purpose 2.142 0.943 1 5 occupation 3.909 1.520 1 6 Table 2. Estimated parameters a. Variable Coefficient t-statistic Effects on Pðy¼1Þof changes in x b constant −3.297 −8.398*** hour11_15 0.306 2.115** Δ0.076 hour16_20 −0.615 −3.502*** Ñ0.149 station_origin c 0.901 9.734*** Δ0.090 station_destination c 0.579 5.993*** Δ0.090 frequency_route −0.068 −2.969*** Ñ0.017 other_mode 0.678 4.312*** Δ0.163 age1520 0.787 2.904*** Δ0.187 age2135 0.111 0.494 Δ0.028 age3649 −0.139 −0.573 Ñ0.035 gender 0.392 3.287*** Δ0.097 study_medium 0.857 3.547*** Δ0.202 study_high 0.484 1.963** Δ0.119 purpose_work 0.627 2.788*** Δ0.152 purpose_study −0.120 −0.63 Ñ0.030 occupation_employed 0.545 1.658* Δ0.133 occupation_self_emp −0.124 −0.497 Ñ0.031 a *p< .10, ** p< .05, *** p< .01. No relevant multicollinearity was found. b The baseline individual (an individual that presents a 0 value in all dummy variables) was considered. c One unit increase in the number of docking stations available in the most frequent district. 550 R. BREY ET AL. the likelihood of using PBSS is greater in the 11–15 time bracket, and lower in 16–20 time bracket, probably, due to productive and educational activity (Corcoran et al. 2014). The frequency_route is significant and negative, i.e., the greater the frequency with which a cyclist uses a route, the less likely it is that he/she will use PBSS (following Buck et al. 2013). The positive significant purpose_work and purpose_study variables show that it is more likely that a cyclist will use PBSS for subsistence trips (as in Buck et al. 2013). Both the station_origin and station-destination variables are significant and positive. As Bachand- Marleau, Lee, and El-Geneidy (2012) state, the greater the density of PBSS stations at the origin/ destination, the ‘nearer’the system is perceived to be and, therefore, the more likely its use. Finally, the positive and significant other_mode variable shows that cyclists opt for public bicycles whenever they need to make intermodal changes (Chatterjee, Sherwin, and Jain 2013). IV. Conclusions We conclude that differences can be found between PBSS/private bicycle user profiles in Seville that depend on age, gender, level of education, occupation, hour/frequency of cycling, trip motivations and PBSS characteristics. Compared to private bicycle riders, cyclists who choose PBSS are predominantly very young males with a high level of education who mainly use the public bicycle for subsistence trips (work or study) during class time or trading hours and are motivated by intermodal needs and the existence of a high number of docking stations within close reach. On the other hand, private bicycle use is more widespread among females who are also quite young, but less so, who use the bicycles to make regular trips that are considered to be nonsubsistence and that do not require their use in combination with any other means of transport. From a practical point of view, these results could help to guide urban transportation policy by identifying the profiles of the two segments of cyclists which, as we have seen, present significant differences. In fact, without disputing the PBSS’unquestionable success, the conclusion can be drawn that urban policies that seek to promote and generalize bicycle use should also envisage certain specific elements to facilitate private bicycle use. For example, measures could be implemented to lessen the likelihood of private bicycles being stolen or to facilitate their intermodal use, especially in conjunction with public transport. This is an issue that has still not been addressed in cities like Seville, where bicycles have not been a traditional or widespread means of transport. These results would also justify any improvements made to the design of PBSS bicycles to make them lighter and more ergonomic; anything to bridge the gap between the usually lightweight and agile private bicycle and the heavy, difficult-to-handle public bicycle used in many cities, including Seville (the multinational JCDecaux Company’s standard model). Making any such improvements would be difficult in the short term, however, as they would have to be compatible with ‘tedious’measures to counter theft and vandalism that have to be part and parcel of public bicycles. Disclosure statement No potential conflict of interest was reported by the authors. Funding This work was supported by the Andalusian Regional Government and the European Union (ERDF) [grant number GGI3001IDIR]. References Bachand-Marleau,J.,B.H.Lee,andA.M.El-Geneidy.2012. “Better Understanding of Factors Influencing Likelihood of Using Shared Bicycle Systems and Frequency of Use.” Transportation Research Record: Journal of the Transportation Research Board 2314: 66–71. doi:10.3141/ 2314-09. Bordagaray, M., L. dell’olio, Á. Ibeas, R. Barreda, and B. Alonso. 2015.“Modeling the service Quality of Public Bicycle Schemes Considering User Heterogeneity.” International Journal of Sustainable Transportation 9: 580–591. doi:10.1080/15568318.2013.838722. Buck, D., R. Buehler, P. Happ, B. Rawls, P. Chung, and N. Borecki. 2013.“Are Bikeshare Users Different from Regular Cyclists?”Transportation Research Record: Journal of the Transportation Research Board 2387: 112– 119. doi:10.3141/2387-13. APPLIED ECONOMICS LETTERS 551 Castañon, U., J. Castañon, and M. Santos. 2012.“The Profile of Cyclists in the City of Juiz de fora.”Work: A Journal of Prevention, Assessment and Rehabilitation 41: 5866–5870. Chatterjee, K., H. Sherwin, and J. Jain. 2013.“Triggers for Changes in Cycling: The Role of Life Events and Modifications to the External Environment.”Journal of Transport Geography 30: 183–193. doi:10.1016/j. jtrangeo.2013.02.007. Corcoran, J., T. Li, D. Rohde, E. Charles-Edwards, and D. Mateo-Babiano. 2014.“Spatio-Temporal Patterns of a Public Bicycle Sharing Program: The Effect of Weather and Calendar Events.”Journal of Transport Geography 41: 292–305. doi:10.1016/j.jtrangeo.2014.09.003. Dill, J., and K. Voros. 2007.“Factors Affecting Bicycling Demand: Initial Survey Findings from the Portland, Oregon, Region.” Transportation Research Record: Journal of the Transportation Research Board 2031: 9–17. doi:10.3141/2031-02. Eyer, A., and A. Ferreira. 2015.“Taking the Tyke on a Bike: Mothers’and Childless Women’s Space–Time Geographies in Amsterdam Compared.”Environment and Planning A 47: 691–708. doi:10.1068/a140373p. Fernández-Heredia, Á., A. Monzón, and S. Jara-Díaz. 2014. “Understanding Cyclists’Perceptions, Keys for a Successful Bicycle Promotion.”Transportation Research Part A: Policy and Practice 63: 1–11. Fishman, E., S. Washington, and N. Haworth. 2013.“Bike Share: A Synthesis of the Literature.”Transport Reviews 33: 148–165. doi:10.1080/01441647.2013. 775612. Fuller, D., L. Gauvin, Y. Kestens, M. Daniel, M. Fournier, P. Morency, and L. Drouin. 2011.“Use of a New Public Bicycle Share Program in Montreal, Canada.”American Journal of Preventive Medicine 41: 80–83. doi:10.1016/j. amepre.2011.03.002. Handy, S., B. Van Wee, and M. Kroesen. 2014.“Promoting Cycling for Transport: Research Needs and Challenges.” Transport Reviews 34: 4–24. Handy, S. L., and Y. Xing. 2011.“FactorsCorrelatedwithBicycle Commuting: A Study in Six Small US Cities.”International Journal of Sustainable Transportation 5: 91–110. Hekman, A. J., and D. B. Deisenroth. 2013.“An Empirical Approach for Directly Accounting for Expectations in Behavioural Models.”Applied Economics Letters 20: 1318–1321. doi:10.1080/13504851.2013.802087. Manski, C. F., and S. R. Lerman. 1977.“The Estimation of Choice Probabilities from Choice Based Samples.” Econometrica 45: 1977–1988. doi:10.2307/1914121. Vogel, M., R. Hamon, G. Lozenguez, L. Merchez, P. Abry, J. Barnier, P. Borgnat, P. Flandrin, I. Mallon, and C. Robardet. 2014.“From Bicycle Sharing System Movements to Users: A Typology of Vélo’v Cyclists in Lyon Based on Large-Scale Behavioural Dataset.”Journal of Transport Geography 41: 280–291. doi:10.1016/j. jtrangeo.2014.07.005. 552 R. BREY ET AL.