Private car use still dominates everyday travel behavior in many regions of the world [1
]. In view of persistent population growth and urbanization processes, urban areas in particular have to face the negative impacts of further increases in individual motorized transport. These include congestion and extensive land use as well as air pollution and traffic noise affecting health [2
]. Within many cities, the promotion of cycling as a sustainable transport alternative has become an important measure in order to meet these challenges [6
]. A wide variety of interventions related to infrastructure, regulations, and marketing is being implemented all over the world to increase cycling or reduce car use [1
]. The concept of cycle streets that emerged in several countries within recent years aims to facilitate bicycle use and to discourage motorized traffic at the same time [12
]. It is therefore an approach counteracting present car dependency by fostering non-motorized alternatives, similar to other environmental developments focusing on the improvement of bikeability or walkability in urban areas involving bicycle paths and footpaths, safe crossings, and aesthetic elements [15
Cycle streets can be defined as shared streets within cities or suburban areas prioritizing cycling by means of certain regulations and design features [16
]. Within a cycle street, the entire carriageway is designated for bicycle traffic, even though in many cases it is shared with motor vehicles. As no standardized guidelines for cycle streets exist, different implementations can be found even within the same country as well as different notions such as “bicycle boulevard”, “neighborhood greenway”, “bicycle priority street”, and “bicycle friendly street” [19
]. In Germany, “bicycle streets” are specified within its national traffic regulations (“Straßenverkehrsordnung”). According to these, bicycle traffic within cycle streets must not be obstructed or endangered by motorized traffic, which is requested to adapt to the cyclists’ speed. The maximum speed is limited to 30 km per hour. Cyclists are allowed to ride alongside each other. Motorized traffic is only permitted to pass through the street if there is an extra permission indicated by an additional traffic sign [21
Benefits related to the implementation of cycle streets include the establishment of bicycle routes in a cost-effective way without the need for additional space, the enhancement of safety for cyclists, and the potential of shifting from the car to the bicycle [18
]. Thus, the intervention of implementing a cycle street refers to a strategy of transportation demand management, which aims for a change in travel behavior in order to foster sustainable, efficient, and affordable mobility [22
The idea of cycle streets emerged in several countries in recent years. Examples can be found in particular within Dutch, Belgian, and German communities and occasionally in Denmark, Sweden, Switzerland, Austria, France, Spain, the USA, and Canada [14
]. Despite the increased implementation of cycle streets, until now, research on its impact has been limited to only a small number of cities or streets [25
] mainly focusing on traffic volume or safety issues. Most commonly, studies relate to the use of a cycle street by cyclists based on traffic volume counts. For instance, the implementation of the first cycle street in Belgium in 2011 resulted in an increase in cyclists within this street of more than 50% within the first few months [27
]. Similar evaluations indicating an increase in the street’s bike traffic can also be found regarding implementations in Germany [28
], Switzerland [30
], and the USA [31
]. By means of GPS monitoring and surveys, some authors point out cyclists’ route choice preferences for passing through cycle streets instead of common streets, as these are associated with a higher level of safety and comfort [32
]. Based on police-reported collision data and cyclist count data, Minikel verifies these associations by identifying lower collision rates for cyclists within cycle streets compared to parallel arterial routes [34
On the contrary, evaluations of the impacts on car use are even scarcer and provide ambiguous results, indicating no significant decline in car traffic after the implementation of a cycle street [28
]. Several studies observed a decrease in speed from motorized vehicles [27
], which is dependent on the specific regulations established within the respective street [30
]. Some papers on cycle streets highlight the importance of design measures in terms of reducing traffic speed and discouraging car driving as well as improving acceptability and identifiability, including pavement markings, signs, impediments, and narrowing [19
Further studies on cycle streets concern the acceptance and the awareness of those affected by the intervention. Corresponding studies emphasize that cyclists in particular consider the implementations as an improvement of the street’s quality. Moreover, local residents, pedestrians, and even car drivers support the idea of cycle streets in several cases [30
]. Evaluations of the knowledge of the concept of cycle streets indicate that many of those people passing through the street do not even know exactly which regulations have to be followed [14
The review of previous research on cycle streets shows that the intervention’s influences on individual processes of travel behavior have hardly been considered. Although findings thus far suggest that cycle street implementations result in an increase in bicycle traffic within the respective street, uncertainty persists as to whether the intervention actually causes more frequent use of a bicycle or if existing cyclists just change their routes in favor of the cycle street. Thus, there is a research gap regarding the effects on individual willingness and intentions to change previous bicycle use and car use. Furthermore, the impact of perceived characteristics of the implementation, such as identifiability and emerging traffic conflicts, needs further examination.
Therefore, the aim of this paper was to investigate whether the implementation of cycle streets triggers travel behavior change and thus contributes to the objective of fostering sustainable transport within urban areas. As the intervention involves the promotion of cycling and the discouragement of car use, we focused on both modes of transport, questioning the effects on the willingness to increase bicycle and reduce car use.
For this purpose, we conducted a written household survey (n = 701) in two neighborhoods within the city of Offenbach am Main situated in the Rhine-Main metropolitan region in Germany. In one of these neighborhoods (the “Senefelder” neighborhood), a cycle street was implemented six months earlier. There is no cycle street within the nearby reference neighborhood. To determine the effects of the cycle street, we compared the survey data of the two neighborhoods considering the awareness, the use, and the individual perceptions of the intervention as well as bicycle and car related behavior and attitudes.
In order to evaluate possible changes in travel behavior, we employed the stage model of self-regulated behavioral change (SSBC) [39
], which was developed as part of the intervention evaluation approach “MaxSUMO” [42
]. By including several factors influencing the intention and the adoption of a certain behavior, the SSBC facilitates assignment to a stage within the processes of voluntarily turning away from previous behavior and adopting an alternative. In recent years, the concept of stage models has been increasingly applied to evaluate the impact of interventions on behavior change (see the review in [44
]). With regard to mobility interventions, the reduction of car use in favor of more pro-environmental transport modes, such as cycling or public transport use, has been investigated in particular. Examples include the effects of a marketing campaign for new residents [45
], of car reduction policies [46
], and of a mobility self-assessment app [48
]. Yet, to the best of our knowledge, cycle street interventions have not been evaluated based on the SSBC model. As part of this study, we examined several indicators of the cycle street with regard to possible correlations with the SSBC assignments to a stage within the process of behavioral change providing evidence on the behavioral impact of this intervention.
The remainder of this paper is structured as follows. Section 2
addresses the theoretical background to the evaluation of travel behavior change. In Section 3
, we describe the case study and the data derived from the survey. Next, results concerning the perception and the use of the intervention, travel attitudes, and behavior as well as the process of behavioral change are presented in Section 4
, indicating differences between the two study sample areas, which are discussed in Section 5
. The paper ends with conclusions in Section 6
2. Theoretical Background
Reviews on studies concerning the impacts of bicycle-related interventions point out prevailing uncertainties about the actual effects of specific implementations and policies caused by ambiguities and limitations of the evaluation methods applied [6
]. In particular, missing control groups and the neglect of individual perceptions and processes of behavior have been identified as shortcomings of previous research. Thus, several authors insist on comprehensive analysis instruments in order to assess behavior change possibly suggested by interventions [11
One such evaluation approach considering the processes of behavioral change due to an intervention is “MaxSUMO” [42
]. It involves three main levels of intervention evaluation indicators concerning the perception and the use of the mobility management service provided (e.g., a cycle street), the acceptance and the use of the mobility options offered (e.g., mode of cycling), and the overall effects [42
]. The latter refers to the main outcomes of the mobility intervention characterized by the adaptation of new attitudes and behavior related to mode choice and travel.
The stage model of self-regulated behavior change (SSBC) was developed in order to measure behavioral changes [39
]. In accordance with its predecessor model—the “transtheoretical model” by Prochaska et al. [55
]—within the SSBC, the process of voluntary change is described by means of four stages representing different levels of openness and willingness to question current behavior and to adopt an alternative: The stages of (1) “predecision”, (2) “preaction”, (3) “action”, and (4) “postaction” (Figure 1
). The presence of certain individual perceptions concerning the previous as well as the new behavior initiate the transition from one stage to another [39
]. Thus, the model enables analysis related to the two processes of abandoning the old behavior and of turning towards another, pro-environmental one.
In the stage of “predecision”, individuals show no preparedness and intentions for change. The transition to the second stage is initiated by the development of an individual “goal intention” indicating the willingness to change previous behavior (e.g., the frequent use of a car). This willingness is induced by obligations to fulfill personal norms [58
], by positive associations with and the perceived feasibility of abandoning previous behavior [57
]. The second stage (“preaction”) involves the evaluation of alternative behavior by means of related perceived behavioral control over the alternative as well as attitudes towards this behavior [59
]. The latter refers to the expected outcomes evoked by the behavior, which—when considered positively—strongly encourage the “behavioral intention” [40
]. Next, individuals in the stage of “action” are characterized by the strong intention to replace previous behavior with the evaluated alternative. The prerequisites indicate the willingness and the preparedness for behavior change and that explicit plans or first attempts for implementation exist. The transition to the fourth stage (“postaction”) is marked by the regular implementation of the intended new behavior. It involves the maintenance of a new habit, resulting in permanent turning away from previous behavior and adopting the alternative. Although the model constitutes a sequence of stages, the process of behavior change is not necessarily linear. Individuals might remain in a certain stage or even return to a previous one [45
Within previous studies, researchers have applied various approaches for operationalizing the assignment of individuals to a certain stage of behavioral change, involving variables related to individual intentions, attitudes, and actual behavior. Within several studies, the frequent use of an alternative transport mode (e.g., at least once a week or daily) was considered as an indicator of assignment to the last stage [46
]. Stated intentions of using alternative transport modes or reducing car use as well as certain attitudes were taken into account to determine the transition to the middle stages of behavior change [41
4. Evaluation of the Cycle Street Intervention in Offenbach—Results of the Case Study
In accordance with the approach of MaxSUMO, our evaluation of the impact of the cycle street involves the following steps: an analysis of indicators concerning awareness, use, and perceptions of the cycle street (Section 4.1
), a discussion of the use of and the attitudes towards the mobility options of cycling and car use reduction (Section 4.2
), and the application of two models of behavioral change (Section 4.3
) comprising an analysis of related determinants by means of multivariate regressions (Section 4.4
4.1. Evaluation of Awareness, Use, and Perceptions of the Intervention
To evaluate the implementation of the cycle street in Offenbach am Main, we examined aspects of awareness, use, and perception (Table 6
). The results show that, within the Senefelder neighborhood, awareness (93%) and regular use (89%)—regardless of the means of transport used—of the cycle street are significantly higher than in the reference neighborhood.
With regard to the factors of the perception of the implementation in Senefelderstraße and the cycle street concept in general, a higher level of satisfaction related to quality and positive effects as well as improvement for cycling can surprisingly found among respondents living in the reference neighborhood. However, participants in the reference neighborhood are significantly more likely to avoid Senefelderstraße, while those living in the Senefelder neighborhood perceive emerging traffic conflicts and speeding more often. The factors of identifiability and hindering car traffic indicate no differences between the investigation areas.
4.2. Evaluation of Bike and Car Use and Attitudes
In order to evaluate the possible effects of the cycle street implementation on individual mobility, survey items referring to cycling and car use travel behavior and attitudes were taken into account (Table 7
). The results show that frequent bicycle use is significantly more common within the Senefelder neighborhood where the cycle street intervention is taking place. 62% of the respondents in the Senefelder neighborhood stated to ride a bike frequently in summer; in winter still a share of 35%. With a proportion of 48%, frequent car use is less common than frequent bicycle use in summer in the Senefelder neighborhood.
With regard to bicycle attitudes, there are high approval rates relating to fun, flexibility, and freedom of riding in both neighborhoods. The feeling of a lack of safety associated with cycling is, however, significantly higher in the reference neighborhood.
In both neighborhoods, respondents show a high level of awareness of the environmental issues of car traffic and little agreement with car travel relating to fun and passion. Overall, the attitudes towards car use show no differences between the neighborhoods.
4.3. Evaluation of Behavioral Change
As described in Section 3.3.3
, we applied two stage models to evaluate the processes of travel behavior change. To determine the significance of the cycle street intervention for the position within this stage model, we analyzed possible correlations between the intervention factors discussed in Section 4.1
and the stage affiliation by means of Pearson’s chi-square test and Spearman’s rank correlation coefficient [74
]. Due to the relatively small number of cases assigned to the middle stages, these were combined into one single stage (“preaction/action”) indicating a status of transition between the stages of habitual behavior.
The analyses reveal that proximity, awareness, use, and positive evaluations of the cycle street have positive correlations with the progress within the process of frequent bicycle use (Table 8
). A higher share of respondents within the postaction stage lives in the Senefelder neighborhood. In addition, the proportion of awareness and the regular use regarding the cycle street are much higher within stages 2–3 and 4 compared to the predecision stage. Furthermore, perceptions of the cycle street’s high quality and positive effects as well as of the concept of improving cycling effectively are more positive within the phases of preaction/action and postaction. Remarkably, within stages 2–3, the identifiability of the cycle street is evaluated better than within the other stages. Avoidance of Senefelderstraße and the perception of cycle streets as being a hindrance to car traffic are significantly higher within stage 1 than in the other stages.
For the analyses of reduced car use, we also examined potential correlations with the factors of the cycle street intervention. As Table 9
shows, being regularly within Senefelderstraße differs considerably and reveals a high value within the middle stages. Furthermore, a lower perception of high quality and positive effects of the cycle street as well as traffic conflicts and speeding is in the stage of predecision. Similarly, the values for forced car traffic detours and avoidance as well as for hindering car traffic are the highest within the stage of predecision and thus among respondents who show no willingness to reduce their car use.
4.4. Regression Analyses of Factors Influencing Behavioral Change
To determine whether the factors of the cycle street intervention actually have an influence on bicycle and car use behavior, as the results of the previous sections suggest, we conducted binary logistic regressions for each of the stages within the two stage models. Binary logistic regression is a multivariate statistical method for analyzing correlations between various independent variables and a dependent variable with only two possible values. The influence of the independent variables on the probability of the dependent variable having the value of 1 is calculated. Besides the variables referring to the intervention, control variables concerning travel mode access and socio-demographics were added to the regression analyses as independent variables (Table 10
). The binary dependent variables used indicate the respective stage affiliation (1 = assigned to this stage; 0 = not assigned to this stage) of frequent bicycle use or reduced car use resulting in six regression models.
Due to the inclusion of the factors related to the perceptions of the concrete implementation not assessed by all participants, the regression models are based on just 579 cases. All of the calculated regressions—except of the model for stages 2–3 of frequent bicycle use—reveal an acceptable coefficient of determination based on Nagelkerke’s R2
]. The regressions of stage 1 and stage 4 within the bicycle stage model provide a high goodness of fit in particular. The influence of the independent variables is indicated by the odds ratios [Exp(β)] expressing the change in probability of stage affiliation in the case of an increase in the variable’s value.
The results show that the availability of the respective travel mode—bicycle or car—influences the related stage affiliation. More specifically, the availability of a bike positively affects belonging to stages 2–3 and 4 of the frequent cycling model, while no access to a bicycle increases the probability of being in stage 1. Having a car at one’s disposal positively affects stage 1 and stages 2–3 affiliation within the reduced car model, whereas no availability favors belonging to stage 4.
Aside from transport mode availability, several intervention variables show a distinct impact on behavioral change processes, in particular using Senefelderstraße regularly as well as the evaluation of the intervention’s quality. Accordingly, the implementation of frequent bicycle use is positively affected by regular use of the cycle street (irrespective of means of transport), awareness of the cycle street, and perception of high quality and positive effects of the intervention. Furthermore, using the cycle street regularly and identifying its positive characteristics encourage affiliation to the stage of preaction/action for car use reduction and counteract the stage of predecision.
While perceived traffic conflicts and speeding within Senefelderstraße has no significant effect on one of the bicycle use stages, this factor is associated with the reduced car model, indicating that being one of those respondents who does not use a car frequently is related to experiencing traffic issues. The perceptions of the cycle street concept primarily affect stages 1 and 4 within both behavioral change models. Not being convinced of the concept’s benefits is positively associated with the stages of predecision, whereas positivity towards the concept encourages affiliation to both stages of postaction.
The socio-demographic variables also contribute to the regression models to some extent. Respondents who do not have a higher education or are not in employment or in education are more likely to be in stage 1 of the bicycle use model and show no willingness towards frequent cycling. Participants in employment or education, which indicates frequent activities and a higher need for traveling, are positively associated with frequent bicycle use. In contrast, data suggest that the higher the income is, the higher is the probability of maintaining frequent car use. This relation can be explained by the costs of affording one’s own car. Previous studies suggest that limited financial resources in particular inhibit access to a car [67
In order to face challenges related to motorized traffic, the promotion of bicycle use can be found within many urban areas around the world. By the implementation of cycle streets, an increasing number of cities aims to facilitate cycling while discouraging car use at the same time. Previous research on the impact of cycle streets is limited to a small number of studies indicating an increase in bicycle traffic within the respective streets [25
]. However, as these studies are primarily based on traffic volume monitoring, uncertainty persists about the actual effects of the intervention and its characteristics on the individual willingness to adopt sustainable travel behavior. Therefore, the objective of this paper was to evaluate whether cycle street interventions actually trigger travel behavior changes related to cycling and car use.
For this purpose, we conducted a written household survey in the city of Offenbach am Main involving residents directly affected by a cycle street implementation and residents of a reference neighborhood. The data used included perceptions of the cycle street as well as individual intentions and attitudes related to cycling and car use. In order to evaluate impacts on individual behavior, we applied two self-regulated stage models (SSBC) specifying the individual’s stage within the processes of behavioral change: one related to frequent bicycle use and the other to reduced car use.
The analyses indicate a positive impact of the cycle street implementation on both processes of adapting sustainable travel behavior, albeit in a different way. Foremost, the encounter with the cycle street seems to encourage the implementation of frequent cycling behavior in everyday life. Although no such effect on actual car use could be observed, the results show the intervention’s positive effect on openness and willingness to reduce car use. The individual perceptions of the cycle street’s characteristics are linked to the affiliations within both stage models. In particular, a positive evaluation of the implementation’s quality counteract low willingness of cycling and reducing car use. However, the participants’ perceptions also show that conflicts between cars and cyclists as well as car speeding pose a major problem, causing dissatisfaction with the cycle street.
Therefore, despite the positive impact of the cycle street intervention on bicycle use, the findings reveal the need for further research related to the design of the implementation, other types of interventions, and the evaluation of individual travel behavior. As the results indicate, the identification of design measures addressing the perceived traffic issues might result in an enhancement of the positive effects of the intervention. In particular, an improvement in the cycle street’s safety would meet the needs of inexperienced cyclists [81
]. Initial corresponding examples include speed management design elements, such as narrowing by means of curb extensions, and refuge islands [20
]. As the results show no impact on actual car use frequency, additional interventions as useful supplements of the cycle street should be implemented and evaluated. Previous studies claim that car use reduction can only be triggered effectively by extensive car deterrent strategies, such as parking or congestion pricing [10
]. Furthermore, the implementation of an entire network of cycle streets or other bicycle infrastructures, such as separated bike lanes [82
] and bike-sharing systems [83
], might reveal a stronger influence on individual cycling behavior. Moreover, further research regarding the evaluation methods of individual behavior and behavioral changes could provide additional insights into the promotion of sustainable travel. Although the suitability of the SSBC was analyzed within several previous studies [39
], additional studies verifying existing constructs or new variables could contribute to the model’s improvement.