LinksGlossaryMessagesSitemapHelp


Home

Policy Instruments

Select
Search
Filter
Cycle routes
SummaryFirst principles assesmentEvidence on performancePolicy contributionComplementary instrumentsReferences

Evidence on performance
Case 1:  Bicycle route network in Odense, Denmark

Case 2: Two Norwegian cycle towns – changes in travel behaviour
Case 3: Cycle routes in three Norwegian towns – a benefit cost analysis
Case 4: Cycle routes and accidents – a meta analysis



Case 1:  Bicycle route network in Odense, Denmark

Context

Denmark has a long tradition for cycling. 50 % of the total population of 5 millions are referred to as cyclists. The number of bicycles are 4-5 millions and the number of cars 1,6 millions. In order to give the municipalities incentives the state has from the mid-80’s supported cycle routes with 50 per cent of the costs and a dedicated planning guide was issued. In 1998 – 89 experimental projects were carried out in four cities, Odense being one of them. (Road Directorate & Ministry of Transport 1989)

Odense is a medium sized city with 200,000 inhabitants, the third largest in Denmark. Since 1976 the city has built cycle tracks according to a coordinated traffic and recreation plan. The target has been to connect all town areas of the municipality with Odense centre and give access to schools and green spaces. In 1984 Odense received a Government subsidy for building two cycle routes in the town centre and extending the cycle tracks. The city centre is the destination of 12,500 cyclists each day. The traffic plan of 1984 is based on the concept that all general car traffic is channelled via a ring road system, traffic with origin/destination in the town centre is led to a number of car parks close to the pedestrian streets.

The total cycle network in Odense includes 250 km of tracks and is signposted. The network is shown on a free map dispersed to the inhabitants (ADONIS 1998), see figure 3.

Figure 3: Map of the bicycle route in Odense, Denmark. Source: ADONIS.

Impacts on demand The provision of cycle routes has improved access to the car-free centre for both shopping bound cyclists and those passing through. According to ADONIS (1998) cycling has risen by 40 per cent in the city centre as a result of the scheme. Bicycle traffic to and from the centre increased from 12,500 bicycles in 1982 to 15,400 in 1990. The impact on car drivers and the number of travellers shifting mode is not known.   Impacts on Supply

There are four types of streets in the city included in the city network:

  • Pedestrian streets
  • Shopping streets with a two-way cycle track of 3,5 m. These are located within the centre of the city, slightly below the level of footpaths on either side.
  • Combined cycle track and bus streets of 6,0 m width and traffic in both  directions
  • One way streets for cars with cycle lanes on both sides.
 The construction of theses streets has altered the use of existing road space in the city centre in favour of cyclists. No studies have been found which indicate how this may affect mobility for motor vehicles.

Tracks for walking and cycling are separated from the carriageway by distance and for instance constructed in green areas, increasing the area that is used for road purposes.

Top of the page

Other Impacts  - Traffic safety

According to ADONIS (1998) the total number of cycle accidents per year in the city centre decreased from 49,8 to 38,0 from 1980/83 to 1989/90. The corresponding figures for accidents involving personal injuries were from 15,8 to 12,0.


Figure 4: Examples of cycle facilities that can be included in a cycle network. Source: ADONIS.

Contribution to objectives

The specific objectives of Odenses’ cycle plan are outlined above. The impacts on other objectives are not documented, but are evaluated informally.


Objective

Comment

Efficiency

The benefits will exceed the costs, when accident reduction and increased feeling of safety are taken into account.

Liveable streets

The traffic plan at large has induced less car traffic in the city centre and the streets have become more cycle and pedestrian friendly.

Protection of the environment

More cycling and possible reductions in car use may contribute to a reduction in environmental impacts.

Equity and social inclusion

Evidently better accessibility to and in the city centre for people without a car is achieved.

Safety

Cycle routes in the city with safe crossings have significantly increased safety.

Economic growth

Better accessibility to and in the city centre may support economic growth (shopping)

Finance

Substantial public funding is used (The cycle routes in the centre cost 1 million EURO and the total traffic calming 4 million EURO- 1989 prizes).

Top of the page


Case 2: Two Norwegian cycle towns – changes in travel behaviour

Context

The towns of Sandnes and Tønsberg/Nøtterøy in southern Norway were designated as "Bicycle Towns" in 1992.  The objectives were set to create:

  • A better environment in the urban areas
  • Better health and well-being in the population
  • Reductions in urban space consumption
  • Less traffic congestion, a more efficient service provision

Both towns have approximately 50,000 inhabitants. Sandnes invested 3,6 millions EURO and Tønsberg/Nøtterøy 5,9 millions EURO in cycle routes.The measures included 11/41 km cycle tracks, 4/4 grade-separated crossings and 300/400 parking places. In addition each town used half a million on campaigns and information (Kolbenstvedt et al 2000). 

Evaluating the effects of the set of measures carried out to stimulate cycling in the two towns was set as an important element of the cycle town concept. In practice, however, the evaluation that was carried out fell far short of the desired standard.

Impacts on Demand

Travel surveys were carried out in 1992 and 1994. Both years some 1,600 persons aged from 15 to 79 years in each town were interviewed by telephone concerning their bicycle use and cycling habits.

The survey responses indicate that people would like to cycle more. Many respondents claim to be cycling more now than previously. However, the actual amount of cycling has not changed significantly from 1992 to 1994, cf. table 4. Neither did one find a change in bicycle ownership. People may want to cycle more, but do not carry out this intention in practice. Data from the 1997 National Travel Behaviour Survey indicates an increase in weekly cycle trips. This share is higher than the average for Norway, which is natural, given the types of town.

Top of the page

Table 4: Cycle shares (adults > 15 years, daily and weekly trips) before and after the project period in two Norwegian “Bicycle Towns”. Sources: Borger & Frøysadal 1995, Kolbenstvedt et al 2000

 

Sandnes

Tønsberg/Nøtterøy

Cycle share

Daily

Weekly

Daily

Weekly

1992

21

43

29

40

1994

19

42

28

42

1997

   

26

58

The clearly preferred alternative to cycling is walking (48 – 58 per cent). Car was mentioned by 24 – 30 per cent and public transport by 6 – 13 per cent (Borger & Frøysadal 1995).

Impacts on Supply

The few new cycle routes constructed both increased road space and redistributed road space between cyclists and drivers (15 km in total equals approximately
30,000 m2.)

Contribution to objectives

The specific objectives of the “Bicycle town” action are mentioned above. The impacts on these objectives were not documented, only the cycle share. The evaluation below is primarily based on the fact that current cyclists have got better facilities and thus probably increased safety, jfr case 4.

Objective

Comment

Efficiency

The benefits exceed the costs, when accident reduction and increased feeling of safety are taken into account.

Liveable streets

The cycle routes have probably induced more liveable streets in adjacent areas.

Protection of the environment

No increase in cycling.

Equity and social inclusion

Better accessibility by cycle where the routes are located.

Safety

Cycle routes together with grade separated crossing facilities have probably increased safety.

Economic growth

No impact

Finance

Substantial public funding is needed.


Case 3: Cycle routes in three Norwegian towns – a benefit cost analysis

Context

In 2000 the Norwegian Parliament requested the Government to : “prepare a National Cycling Strategy where the main goal is to make it safer and more attractive to choose bicycle as means of transport. This Strategy must form a part of the National Transport Plan.”  As a part of the preparation a working group with members from the Norwegian Agency for Health and Social Welfare, and the Norvegian Public Roads Administration wished to carry out a comprehensive benefit cost analysis of cycle routes. Current plans for three Norwegian towns Hokksund, Hamar and Trondheim were chosen as cases for the study. Planned main network for cycling was 15, 33 and 220 km. The benefits and costs included are discussed below as potential impacts of the schemes. The study is reported in Sælensminde (2002).

Top of the page

Impacts on Demand

Development of walking and cycling networks may have a potential for increasing the amount of walking and cycling in Norwegian cities. However, since there is large uncertainty regarding substitution from car to walking and cycling, the analyses are presented as scenarios. It is assumed that additional measures (e.g. safer crossing facilities and safer parking facilities for bicycles) must be implemented to achieve high future shares of pedestrians and cyclists. Such measures are implicitly included in the scenarios, but not directly included in the benefit cost analyses of the cycle track networks.

The current and future distribution between different means of transport is estimated from the Norwegian nationwide travel survey from 1997/98. To estimate the average annual daily traffic (ADT) of pedestrians and cyclists on the walking- and cycling track networks the following assumptions were made:

  • Journeys less than 5 km by bicycle or as pedestrian have an average length of 3 and 1 km, respectively.
  • Journeys longer than 5 km by bicycle or as pedestrian have an average length of 6 km.
  • 70 and 20 % of kilometres cycling and walking respectively, takes place on the walking- and cycle track networks.

The estimated increase in km walked and cycled is shown in table 5. Both new cycle traffic and reduced motorized traffic are included. The “best estimate” were based on Lodden (2002) and the following assumptions:

  • Walking- and cycle track networks gives 20 % (5-40% in sensitivity analyses) induced walking and cycle journeys i.e. journeys that would not have taken place without the new networks.
  • 15 % (0-35 % in sensitivity analyses) of current car and public transport journeys less than 5 km, are substituted by walking or cycling. 1/3 is replaced by walking and 2/3 by cycling.

Table 5: Estimated total number of km walked and cycled in the different cities and the average daily traffic (ADT pr km) of pedestrians and cyclists at the walking- and cycling track networks in Hokksund, Hamar og Trondheim. Source: Sælensminde 2002

 

Current traffic

Future traffic (“best estimate”)

City/
Mode of transport

Number of km in city area

ADT at the w/c-track network

Number of km in city area

ADT at the w/c-track network

Hokksund

       

Walking

9777

129

12306

163

Cycle

6308

292

11012

510

Hamar

       

Walking

20671

126

26019

158

Cycle

13336

283

23282

495

Trondheim

       

Walking

143851

131

180369

164

Cycle

152058

484

228955

728

Top of the page

Impacts on Supply

The impacts on the supply of road space depend on where the elements of the cycle network are located. New routes on green spaces will increase road space while routes taken from the carriageway will redistribute road space. No figures on this is given.

Results of the cost-benefit analysis

Benefit components in the analysis include: traffic accidents, travel time, insecurity, school bus transport, reduction of less severe diseases and short time absence, reduction of severe diseases and long time absence, global air pollution, local air pollution, noise, congestion and parking costs. Best estimate of future walking and cycling traffic leave no doubt that building walking- and cycling track networks in tree Norwegian cities is cost effective. Net benefit cost ratios in these cities are between 3 and 14 c.f. table 6. 

It was assumed that travel costs and travel time for different modes of transport do not change because of a new walking and cycle track. These costs are therefore not changed in the generalised travel cost in the cost benefit analyses. Reduced subjective insecurity is the only change in the generalised travel costs. These benefits together with health benefits constitute a substantial part of the benefits calculated.

Table 6: Benefits and costs (based on best estimates of future walking and cycle traffic) of investments in walking- and cycling track networks in Hokksund, Hamar and Trondheim. Unit: NOK. Source: Sælensminde 2002. In EURO

Benefit- and cost components

Hokksund

Hamar

Trondheim

Benefits of walking- and cycle tracks (present value)

     

Accidents (assumed no change)

0

0

0

Travel time (assumed no change)

0

0

0

Reduced insecurity for current pedestrians

557357

360607

14313594

Reduced insecurity for current cyclists

1258548

814274

52955495

Reduced insecurity for new future pedestrians

72090

46642

1816818

Reduced insecurity for new future cyclists

469293

303630

13390175

Reduced costs for school children transport

342078

146918

480225

Reduced costs related to less severe diseases and short time absence 

2224862

4703994

35804136

Reduced costs related to severe diseases

12993194

27471324

209096153

Reduced external costs of motorised road transport

1256060

2655669

16549092

Reduced parking costs for employers

1261257

4594857

57627130

TOTAL BENEFIT

20434739

41097915

402032818

Costs of walking- and cycle tracks (present value)

0

0

0

Capital costs

3141622

2094415

79787234

Maintenance costs

206630

137753

5247745

Tax-cost factor, 20% of budget costs

669650

446434

17006996

TOTAL COSTS

4017903

2678602

102041975

Net benefit- cost ratio

4,09

14,34

2,94


Contribution to objectives

Objective

Comment

Efficiency

The benefits exceed the costs by far, when health benefits and increased feeling of security are taken into account.

Liveable streets

The streets planned are supposed to be cycle and pedestrian friendly.

Protection of the environment

Increased cycling and possible reductions in car use will contribute to a reduction in environmental impacts.

Equity and social inclusion

Better accessibility for people without a car will be achieved.

Safety

Cycle routes in the city with safe crossings will  increase safety and feeling of security.

Economic growth

Better accessibility in central areas might support economic growth (shopping)

Finance

Substantial public funding is needed

Top of the page


Case 4: Cycle routes and accidents – a meta analysis

Context

While many cycle studies lack systematic data on demand impacts, there are  several studies on the effects on accidents. The Handbook of Road Safety Measures (Elvik et al 1997, Elvik & Vaa 2003), includes meta-analysis of 30 studies on the effects of different tracks for walking and cycling and cycle lanes. The studies analysed come from different European countries (28) Japan (1) and USA (1).

The meta-analysis assigns statistical weights to studies by sample size and sorts them by design quality and thus gives the most systematic overview of safety impacts.

This case presents some of these results others are given in Cycle lanes.

Impacts on Demand

A number of studies on traffic safety also shows that walking and cycling increases when tracks for walking and cycling and cycle lanes have been constructed.

Impacts on Supply

Not studied in this context.

Impacts on Ttraffic safety

Cycle routes and lanes will not give large reductions in cycle accidents, cf table 7. Pedestrians seems to profit more than cyclists from tracks for cycling and walking.

Cycle paths separated from motorised traffic lead to a statistically significant reduction of injury accidents. The mean estimate is 4%. For combined tracks for walking and cycling of the type most commonly found in Northern countries, cf. Definitions, the analysis do not indicate that the number of accidents is reduced. However, the risk per cycle km is not known and might have been reduced when the routes have induced more cycle trips.

Table 7: Effects of cycle routes on accident figures. Percentage change in the number of injury accidents. Source: Elvik & Vaa 2003

  Percentage change in the number of accidents
Measures Types of accident affected Best estimate 95% Confidence interval

Tracks for walking and cycling

Pedestrians along the road

-35

(-67; +29)

 

Pedestrians crossing roads

+1

(-32; +52)

 

All pedestrian accidents

-10

(-32; +21)

 

Cyclists along the road

+1

(-37; +62)

 

Cyclists crossing roads

+2

(-42; +78)

 

All bicycle accidents

+1

(-29; +45)

 

Accidents with vehicles

+1

(-10; +14)

 

All accidents

0

(-10; +11)

Cycle paths separated from motorised traffic

Pedestrian accidents

-5

(-12; +3)

 

Bicycle accidents

-2

(-7; +4)

 

Accidents with vehicles

-5

(-9; -2)

 

All accidents

-4

(-7; -1)

We know that walking and cycling increases when combined tracks for walking and cycling have been constructed. This can lead to an increase in traffic at crossings where safety measures may not have been adopted. It has also been documented that not all pedestrians and cyclists use tracks for walking and cycling. Those who continue to use traffic lanes meant for cars, can be exposed to increased risk. Finally, examples have been found where the road authorities in Norway have increased the speed limit from 60 km per hour to 70 km per hour on road sections where tracks for walking and cycling have been constructed. This must be assumed to lead to increased speed. Together, these conditions may contribute to a lack of reduction in the number of accidents.

Top of the page

Also, these measures can lead to cyclists increasing their speed and can sometimes also lead to increased car speed. These are factors that can contribute to less accident reductions than expected.

As a part of a comprehensive strategy or policy combining several cycle facilities however, cycle routes may reduce the risk of cycling. In the Netherlands a broad cycle policy has been implemented throughout the nineties. Results from this policy indicate that combined measures can induce a significant reduction of the number of cycle accidents even if the cycle share increases (CROW 1997).

Contribution to Objectives

From the data presented we can only conclude on safety impacts. Cycle routes separated from the carriageway can reduce accidents slightly, while the evidence on tracks for both cyclists and pedestrians are more uncertain.

Gaps and Weaknesses

Studies of urban cycling are focused on modal split in general, characteristics of cyclists and calculations of cycling potential. Case studies mostly present information of plans, strategies and practical solutions. Only rarely do they give scientific evidence of different transport or demand impacts. The ADONIS (1998) catalogue lies in this tradition and contains few data on impacts, only remarks of advantages and disadvantages.

The benefit cost analyses presented in case: “Cycle routes in three Norwegian towns” includes the benefits of both new cycle traffic and reduced motorized traffic. Estimates of future cycling traffic are uncertain. In addition, uncertainty is large in the benefit components included. Possible changes in cycling accident risk are not taken into account in the analysis of studies of accident effects.

 


Top of the page


Text edited at the Institute for Transport Studies, University of Leeds, Leeds LS2 9JT