Insights from factors influencing cycling

Identifying the factors affecting bicycle demand or cycling to work may give us some insights on investigating what factors affect the uses of bikesharing though literature are still quite few.

There are a wide range of factors may influence individual’s choice to cycling such as demographic and socio-economic, cultural and societal, environmental as well as policy-related determinants. Here provides an overview of these factors as discussed below.

Dill and Voros (2007) point out demographics and environmental factors which incorporate both objective-oriented (e.g., climate and topography, land use, and infrastructure) and subjective-oriented (such as attitude on travel, safety perceptions, convenience, cost, and time valuation) would have impacts on the level of cycling based on the survey conducted in Portland, USA. It reveals that demographic characteristics vary between types of adult cyclist and the desire to cycle more but the impact of income and vehicle ownership on cycling seems to be unclear. It is also found that built environment in terms of both objectively and subjectively has impact on cycling to some extent. Additionally, positive perceptions of more bike lanes is related to higher cycling level and the desire to cycling more; and higher level of street connectivity is related to more cycling for utilitarian trips as well. However, it should be noted that the influences of some socio-demographic characteristics on cycling are still uncertain. For instance, age and income are negatively related to cycle to work in some studies while having a positive or even no impact in others (Handy and Xing, 2011).

29

Several studies believe that providing bicycle infrastructure, particularly bicycle lanes and paths, can increase cycling. Dill and Carr (2003) highlight the importance of bicycle infrastructure, indicating that higher levels of bicycle infrastructure are positively and explicitly associated with higher rates of bicycle commuting. However, it does not indicate that there has a cause-effect relationship since people may cycle more for commuting because of more bicycle lanes or paths. Alternatively, it is the higher cycling level that leads to build more bicycle infrastructure. Buehler (2012) explores the role of bicycle parking, cycling showers, free car parking and transit benefits as determinants of cycling to work, which is based on the Washington, DC context. It indicates that bike parking and cycling showers at workplace are associated with higher bicycle commuting; additionally, if combined these two facilities together, it is significantly that it has greater influence on bicycle commuting compared with only bicycle parking provided. However, the results show that transit benefits provided by employer seems to be not related to bicycling commuting. Zhao (2013) examines the effects of the built environment on bicycle commuting in Beijing through using multinomial logit (MNL) model for estimating workers’ travel mode. It is surprised that residential density has no significant effects in Beijing for bicycle commuting in comparison to Europe and North America. A higher level of public transit service is likely to decrease bicycle use in this case due to more new metro lines strengthening the pull effect and the competitiveness of low price.

It is also interesting that even though higher level of exclusive bicycle lanes relates to more cycling to work, but the elasticity analysis shows the impacts are smaller than mixed environment. Thus it implies that integrated bicycle facilities with urban design would encourage cycling to work more effectively. Traffic safety and air pollution are also the major factors influencing bicycle commuting.

Climate (long-term) including the weather (short-term) conditions is widely confirmed that it would have influences on the individual choice to cycling. Koetse and Rietveld (2009) present an overview of empirical findings from various literature on the impact of climate change and weather on transport. It is found that changes in temperature, precipitation and wind have impacts on bicycle use in the view of utility; for instance, rainfall and both very hot and cold weather decrease cycling trips. This finding is also in line with the research by Dill and Voros (2007). Noted that temperature and precipitation are the most significant factor for those cycling to work in summer rather than winter. In addition, recreational cyclists are more likely to be affected by bad weather than utilitarian cyclists. Miranda-Moreno and Nosal (2011) investigate the use of urban bike facilities in Montreal to identify the impact of weather conditions on

30

cycling ridership and to identify temporal trends of cycling ridership up to hourly scale through both absolute and relative ridership models. It is found that temperature, humidity, and precipitation have impacts on cycling ridership as expected; nevertheless, the impacts vary significantly across facilities and different time periods. Moreover, while temperature has a positive impact on ridership in most of time (if less than 28℃), the effect of humidity is negative.

The combination of heat and humidity would decrease cycling ridership as well. Flynn et al.

(2012) quantify the impact of weather conditions on individual decisions to cycle to work among a diverse panel of adult bicycle commuters for at least 2 miles. They find that temperature and precipitation are more likely to influence the likelihood of bicycle commuting, which is in line with other studies. In contrast, higher wind speed decrease the odds of cycling to work; moreover, snow depth seems to have a dampening effect that most respondents does not cycling to work in wither months. Noted that the survey focuses on morning commuting condition and neglects the likelihood of other work schedule. Saneinejad et al. (2012) explore the impact of weather conditions on active transport travel behaviour in the city of Toronto through using MNL model on five transport modes; and the interaction between weather and age as well as gender are also investigated through two sub models. It is found that the utility-based cycling decreases in the situation of temperature below 15 ℃ while cycling becomes insensitive if temperature higher than 15 ℃. Wind speed and precipitation are found that negatively influence cyclists more than pedestrians However, Nankervis (1999) suggest that neither weather nor climate would be a strong deterrent to bicycle commuting, and this is generally based on which the rider groups surveyed response. There are several more subtle factors in association with social or psychological act as deterrents to bicycle commuting. For example, the perception of the effect of weather conditions particularly in the traffic dominated by cars, or road conditions, traffic patterns, etc.

It is suggested and concluded that there are several main determinants of bicycle use based on the findings from various studies: demographic and socio-economic, cultural and societal, environmental including weather, urban spatial structure, and infrastructure, and policy-related determinants (Vandenbulcke et al., 2011). Similarly, Handy and Xing (2011) also propose a cross-sectional study of bicycle commuting in six small U.S. cities to explore the relationships between bicycle commuting and individual factors (e.g., socio-demographic characteristics and attitudes), physical and social environment of the work place. It is suggested that individual attitudes and constraints are the most important determinant of bicycle commuting while physical environment is likely to have only a marginal effect directly. It is also found that

31

females are substantially less likely to cycle to work than males, as consistent with previous researches. It is surprising that providing bicycle facilities such as racks, showers, etc. close to the workplace does not indicate a significant influence on bicycle commuting. It implies that the provision of facilities may be a welcome amenity, but it seems not to be a main deterrent for bicycle commuting which is in contrast with the findings by Buehler (2012).

Winters et al. (2013) build a spatial tool through mapping “bikeability” to identify areas that are more facilitative or less facilitative to cycling in terms of opinion survey, travel behaviour studies, and focus groups to forming the bikeability index and their relative importance. It is reported that bicycle facilities, aesthetics, topography, traffic and trip distance are the main built environment factors influencing cycling. While travel behaviour survey indicates that bicycle facilities, connectivity, topography, and land use are the domains for cycling, the focus groups provide useful information on the relative importance of built environment factors. Bicycle facility is the most important among built environment factors, which are about twice the score as traffic. Followed by street network, topography, environment, travel distance and neighbourhood land use. Population density is the least importance among these factors.

Rietveld and Daniel (2004) give a general framework of factors that have a potential impact on bicycle use, which includes: (1) individual features: income, gender, age, and activity patterns;

(2) generalised cost of bicycling: travel time, physical needs and comfort, traffic safety, risk of bicycle theft, monetary cost of bicycle use, and personal security; (3) generalised cost of other transport modes: such as parking cost, fuel tax; and (4) local authority initiatives: quality of capacity of bicycle dedicated infrastructure, spatial design of city and etc. However, the gap of the factors influencing the uses of bikesharing and cycling between bikesharing users and traditional cyclists does not addressed in this research.

Heinen et al. (2010) perhaps offer a comprehensive overview of academic literature on bicycle commuting so far to the author’s knowledge; and focus on empirical results in particular from various aspects such as travel behaviour, transport planning, psychology and health science, as shown in following Table 8. This study also examines the individual’s daily choice toward cycling or not in terms of frequency. It is found that using conventional mode choice models may be insufficient and not be able to address some determinants; hence, other kinds of knowledge should be introduced and help to investigate the gap between for those currently available for motorised forms of transport and active transport.

32

Table 8 Overview of determinants of cycling (to work)s

Determinants Influence mode choice Frequency Preference Comments

Built environment: (1) urban form

Trip distance increase results in less cycling;

(according to 27% of non-cyclists, compared with 2% of cyclists)

No studies found on how access and egress

distance affect cycling frequency • The built environment affects a person’s choice to commute to work by bicycle;

• Cycling share is influenced by the following factors: distance, function mixture, storage facilities, block size and density, the presence of bicycle infrastructure and its continuity, traffic lights and stop signs, land use, parking facilities and showers at workplace; and

• Of these factors, distance seems to be the most important factor; and

• The presence of infrastructure might not only result in more cycling, but a higher cycling frequency could also stimulate the construction of bicycle infrastructure.

Network layout No significant effect on cycling People living closer to city centre cycle more Density Higher density relates to more cycling people living close to city/town centre cycle

more (decrease from 56% to 46% of non-cyclists closer to the centre)

Function mixture Residential densities have no effect;

higher density increases bicycle share people living close to city/town centre cycle more

Built environment: (2) cycling infrastructure separate facilities (safety issue)

Adjacent to parking Roads with no parking perceived

to be safer Continuity of cycling

infrastructure Unclear Preference for continuous

facilities Number of bicycle paths More cycling infrastructure results in

more cycling (increase of 1-2%, but probably depending on location

No effect

Traffic lights More traffic lights in a city associates with

lower cycling levels Experienced cyclists perceive

them more negatively Built environment: (3) facility at workplace

Bicycle parking No significant effect on cycling important to cyclists

Shower at workplace If present more cyclists Seems not to result in higher cycling

frequency important to cyclists;

Bicycle locker mostly preferred

Locker at workplace No effect No effect important to cyclists

Natural environment

Hilliness and land scape Less cycling with hills • A climate with moderate temperatures and little

rain increase the share of bicycle commuting; and Season More cycling in summer and autumn

(differs between locations

33

Temperature Unpleasant temperature corresponds

with less cycling Unpleasant temperature corresponds with

less cycling Less influential for commuting;

more influence on women • Bad and uncertain weather negatively affects a person’s decision to cycle.

Precipitation Cold more unpleasant than heat;

negative effect on cycling May have effect Mentioned by cyclists as most negative weather aspect Socio-economic variables

Gender No Effect;

Men cycle more than women Men cycle more than women • The relationship between socio-economic factors

and cycling is unclear.

• In most countries, men cycling more than women;

in those countries where cycling is very common, such as Belgium and the Netherlands, women cycle more;

• Car ownership has a negative effect on cycling;

logically, bicycle ownership has a positive effect;

and

• Most research merely mentions or examines the relationship between socio-economic factors and cycling, but does not allow us to make any inferences about the causality of this relationship.

Age No effect;

Women cycle more than men;

Cycling declines with increase;

Age is not significant

Income Positive connection between income and cycling;

Negative connection

Employment status No significant connection Part-time workers commute more frequency by bicycle

Car ownership Car ownership decreases cycling

Car ownership has no effect Car ownership decreases cycling;

having few cars increases cycling frequency Psychological factors

Attitude Cyclists have a more positive attitude

towards cycling • There is a relationship between commuting by

bicycle and people’s attitude as well as perceived values. More cycling may result from positive perceptions of cycling or negative perceptions of car use. If people’s social surroundings have a positive opinion of cycling, then higher chance of cycling

Perceived social norm Cyclists have a higher perceived social norm;

No effect on being a cyclists

Habit A cycling habit increases the cycling share A cycling habit increase the cycling frequency

Cost, time, effort and safety Cost of other means of

transport It higher, more cycling • It is thought that people sometimes decide

whether cycling to work with other transport options in terms of cost, travel time and safety.

Negative factors relating to car use or public transport could lead them to develop a more favourable view of cycling; and

• Travel time and safety seem to be more important for cycling than for other modes of transport.

Travel time Experienced cyclists prefer short

travel time

Safety A reason not to cycle Subjective safety does not always

correspond with objective safety

Source: Heinen et al. (2010), Kim et al. (2012)

34