Ministry of Transport and Public Works

Evaluation of the Delft Bicycle Network Plan

Homepage Manual Contents Navigation

RESULTS AND CONCLUSIONS OF THE STUDY PARTS - Travel behaviour and choice constraints

This study part evaluates the Delft bicycle network plan from the perspective of actual and potential use of the bicycle. By investigating actual bicycle use the achieved increase in cycling will be determined in a quantitative sense, also in relation to other modes of transport. Encouragement of bicycle use is also assessed from a more qualitative point of view, e.g. removing choice constraints, improving the conditions for cycling as perceived by people.

In keeping with the distinction made above, the study has been divided into two parts:

• a descriptive study of the population's work-day travel behaviour; emphasis is on the proportion of trips made by bicycle, before and after implementation of the bicycle network;
• a subsequent in-depth study, offering an explanation for the level of bicycle use found, and furthermore providing insight into changes which occurred in the perception of conditions for cycling.

Data for the descriptive study were collected by means of a postal home interview survey. The more qualitative information was gathered by in-depth interviews.
Both approaches were used both before (1985/1986) and after (1985/1986) the actual implementation of the infrastructural measures. The study investigates the population of Delft. The survey was limited to two study areas and one control area. The main study area was North-West, where a large number of measures varying in scale and type were made at numerous locations. The minor study area was Tanthof, a new peripheral residential district which was linked to the city centre by major new bicycle links.
The Wippolder district is a control area where no measures were implemented during the study period.

The in-depth interviews and related analyses were done according to the situational approach model, developed by the German search institute Sozialforschung Broeg. This method especially provides good insight into existing choice constraint. Gaining experience with this approach was one of the objectives of this study.

• Mobility in general

  The residents of Delft showed a stable level of mobility between 1982 and 1985. The number of mobile people remained the same (88 per cent of the total), as did the number of trips made by the mobile group. This stability can be explained in view of the already high level of mobility in Delft; data from the National Mobility Survey shows that it is higher than in other comparable medium-sized towns in the Netherlands.

• Modal split

  The modal split has changed between the two measurement. The change in the control area is different both from that in both 'experimental' districts.
  In the control area an increase in car mobility terms of absolute numbers of car trips per person well as in terms of the proportion of trips made by car per day, as well as in terms of the proportion of trips made by car, car mobility increased by more than 10 per cent. The increase refers to both driver and passenger trips. This coincided with a reduction of public transport mobility. This reduction is an expression of the falling trend in public transport mobility since 1979, observed in all medium-sized Dutch towns.
  The model choice of the residents of Tanthof changed in a completely different sense. Although they also made less use of public transport, though to a limited degree, no increase was found in car mobility. On the contrary, car mobility clearly dropped and was compensated by an increase in bicycle mobility. The proportion of bicycles in the total number of trips grew from 38 per cent to 39 per cent.
  In North-West a similar shift was observed, but somewhat less pronounced with respect to cycling. There the proportion of bicycle trips grew from 40 per cent to 43 per cent. In contrast to the control area, car use did not increase.

  The increase found in bicycle mobility cannot be ascribed to a growth of travel mobility: there was no change in total mobility between 1982 and 1985. The increase observed is consistent with comparisons based on volume counts: in the study area and even more so in Tanthof bicycle volumes increased.

• Sources of increased bicycle use

  The growth of the bicycle share does not originate from one other mode of transport in particular, but from former trips on foot as well as former car- trips and, though to a lesser extent, former public-transport-trips. The contribution of former trips-on foot and former motorized private-trips is more or less equal. In general, a rather limited degree of mode change was found, insofar as 8 person's activity pattern remained the same. For both bicycle and car trips, a stability level of approximately 95 per cent was found.

  The increase of bicycle mobility cannot be ascribed to a larger number of people using the bicycle, but is the result of a larger number of bicycle trips per cyclist per day. This finding is consistent with the findings from the before- study, based on the situational approach, that many non-cyclists are captive users of other modes (car, public transport) and do not have freedom of choice.
  Bicycle use by men increased more than that by women. This finding is also consistent with that of the other studies in Delft. However, the increase in the test areas is concentrated especially in work and school trips.

  The average bicycle trip length has increased, in contrast to the average trip duration. This means that cycling has become faster compared to 1982. To illustrate: in 1985 bicycle trips by Tanthof residents were approx. 6 per cent lower, while the average duration of a trip even decreased. In general, this suggests a positive effect of the improved continuity of the bicycle network, which extends its influence throughout Delft.

• Choice constraints partly removed

  Various constraints may limit the use of the bicycle. Compared to 1962, some constraints appear to occur less often. Particularly, the amount of time needed for travelling by bicycle is raised as an objection less often; this reduction may have to do with the improved accessibility. In figures, in 1982 for 27 per cent of trips by car and public transport, too much time was argued to be an objection to using the bicycle; in 1985 this objection was reported for 18 per cent of the non-bicycle-trips. General considerations of safety and comfort also form less frequently a restriction: previously for 17 per cent, now 8 per cent of the non-bicycle-trips: Specific road and traffic situations do not prevent people from cycling, neither in 1965 nor in 1952. This in itself indicates that improving only individual junctions and links hardly encourages the use of the bicycle. In other words: isolated infrastructural measures clearly offer less prospects for increased bicycle use than systematic, mutually interdependent measures, as were implemented within the framework of the Delft bicycle network plan.

  The measures taken near Tanthof proved to be effective, as indicated by the growth in cycling observed for Tanthof residents; the increase here is 1arger than in the district North-West. The new bicycle tunnels and routes not only improved comfort and safety, the improved connection with the town centre also implies a shorter route and a reduction in travel time. The importance of these improvements in the corridor involved has been clearly demonstrated by the reported perceptions of Tanthof residents.

• The perception of cycling

  The bicycle climate has evidently improved in Delft. The conditions for cycling are now perceived clearly more positively in comparison with three years ago. In Tanthof this improvement was found to apply to slightly more than half of the respondents, with respect to comfort, continuity and safety. It applies to less people in North-West, and involved more frequently comfort and continuity (less detours, shorter waiting times) than safety.

  On balance, a small minority of less than 10 per cent still perceives the bicycle system in Delft as negative in terms of comfort and continuity. The conditions for driving a car are experienced as being notably worse.

  A wide range of effects can be ascribed to the bicycle plan: positive effects are noticeable for a considerable part of bicycle trips being made now. This is the case for over a quarter of these trips made by residents of the district North-West and furthermore for approx. 45 per cent of the bicycle trips by Tanthof residents.
  The positive effect of the bicycle plan can also be derived from the notable reduction of the number of bicycle trips which experience problems en route. The proportion of 'problem trips' fell from one half to one third. All this is the combined result of removed bottlenecks as well as route shifts which have become possible.

• Transfer potentials

  Besides the use of transport modes actually found, an assessment was made of the extent to which the market position of the bicycle, relative to the other modes, had improved. The position of the bicycle was not very strong in 1962; the potential losses were bigger than the potential increase.
  The study has determined changes in transfer potentials between different modes of transport. A transfer potential to a certain mode is assumed to exist if there are no constraints that prevent that mode from being chosen. After the measures were implemented, the potential use of the bicycle grew considerably. This increase is largest for people who travel by car or on foot; the increase is much smaller for travellers using public transport. Tanthof residents show the largest increase. Although causality cannot be demonstrated, the conclusion is that the measures for encouraging cycling to a certain degree have increased the cycling potential by removing constraints. Besides the measures taken, other factors related to objective constraints have also extended this potential.

  Besides the increased number of trips by car, public transport or on foot that could be made by bicycle without any difficulty, there is also a reversed transfer potential from the bicycle to the car, to public transport or to walking. The potential which could shift from the bicycle to the other modes has also grown since 1982 , but less than the possible inflowing potential. In other words: the possible gain of the bicycle has grown more than the potential loss.

  In absolute numbers of trips the mutual transfer potentials between the bicycle and the car are balanced. This is not the case between the bicycle and walking, since cycling may loose more than gain (see figure 2).

  Figure 2: Choice potentials regarding the bike, and of cyclists regarding other modes (after situation)

  This clearly shows the competitiveness of the bicycle compared to the other modes of transport. It is interesting to conclude that its competitiveness has improved since 1962, especially because of the improved position relative to the car. Nevertheless the bicycle still remains somewhat at a disadvantage.

• Assessment of the results

  The foregoing indicates that quantitative as well as qualitative improvements were made to the bicycle mode between 1982 and 1965. In view of the transfers found in the use of modes of transport, it is likely that the implementation of the bicycle network has prevented a further growth of car use and has strengthened the position of the bicycle.
  The question is how to assess this result. It could be argued that the observed growth of bicycle use is only marginal. In evaluating these figures, the following should be taken into account:
  
  

  • The increase in bicycle use occurred in a situation in which the chances of a reduction were greater than the chances of growth. In view of the existing choice restrictions, in 1982 more trips by bicycle could have been made by any other mode without any difficulty than the number of non-bicycle trips which could have been made by bicycle without problems.
  • The bicycle is not a feasible transport mode for a large number of trips in view of existing constraints, either of a more objective nature, or of a more subjective one. The situational analysis has shown that only a limited proportion of people have real freedom of choice.
  • Generating new bicycle traffic is certainly not the only aim of bicycle stimulating measures. Another objective is to improve the conditions for cycling in terms of comfort, safety etc; furthermore it should be realized that such a policy is necessary in order to prevent a reduction of bicycle use in the lone run.
  • In 1982, the proportion of trips made by bicycle was already quite high in Delft (approx. 40 per cent), although in some Dutch towns it is even higher. Therefore, a higher proportion is difficult to obtain. In this respect it is relevant to note that an increase in car use -as was the case in the control area- did not occur in North-West an that car use was even reduced in Tanthof.
  • The increase in bicycle use can not be completely ascribed to the bicycle network. For years now, bicycle use is growing in most medium sized Dutch towns.

  The effect of the bicycle plan can be traced even more clearly in a qualitative sense than in a quantitative one. The nature of the qualitative changes found are directly linked to the implemented measures:
  
  

  • to the extent that the bicycle is not used (yet), certain types of constraints (trip time needed and considerations of comfort) apply less frequently than before; this partly results in more freedom of choice;
  • a considerable improvement in the conditions for cycling was reported; first, less concrete problems during a cycling trip were reported. Second, people who experience cycling as being positive have increased in number.

  Another qualitative result is a better balance between the car and the bicycle with respect to mutual transfer potentials.

  The results obtained lead to the conclusion that the bicycle network has had a positive influence on actual bicycle use as well as on the modal choice situation. Especially the last, more qualitative result is important in the long run.

From this study the following conclusions can be drawn:

1. The Delft bicycle plan has succeeded in realizing an increase, though be it limited, in bicycle use of the inhabitants of Delft. Even more important is the fact that the measures for stimulating bicycle use prevented a growth of car use. In measures even coincided with a reduction in car travel.
2. The bicycle plan also has merits in a qualitative sense in the form of clearly improved perceived conditions for cycling with respect to safety, comfort and continuity. This improves the stability of the bicycle system in the long term.
3. Mowover, a qualitative result was obtained by establishing a balance between the position of the bicycle and other modes of transport, although at the moment the chances of a reduction in the use of bicycle are still slightly larger than the chances of an increase. In particular, the position of the bicycle compared to that of the car has become stronger.
4. The design philosophy of the bicycle network proves to provide solutions to the problems perceived in reality by cyclists and to their preferences. Since bicycle trips have a widely distributed pattern in a spatial sense and since the problems encountered en route (detours, unsafe and uncomfortable locations) are scattered, a policy accounting for these characteristics has favourable prospects.
5. Measures aimed at improving continuity in a corridor have a positive effect if heavy flows are involved; an example is the relationship between Tanthof and the town centre of Delft.
6. The situational approach proved to be a valuable method to investigate the bicycle potential and mutual transfer likelihood of modes. The method takes objective and subjective choice constraints into account which results in a realistic estimate of the existing freedom of choice and probability of behavioral changes. Through this, the method possesses predictive value.