Evidence on Performance

In this section case studies are presented to demonstrate the empirical evidence of the use of Park and Ride schemes as policy instruments. A summary of an Atkins report containing analysis of several sites in the UK including is presented. A re-evaluation of the Atkins case studies by Parkhurst (2000) is then given. Also included is global information relating to worldwide Park and Ride schemes.

The Travel Effects of Park and Ride Atkins

A report by WS Atkins and the DETR published in September 1998 studied the effects of Park and Ride systems in Brighton, Cambridge Coventry, Norwich, Plymouth, Reading, Shrewsbury and York. Each site was visited and discussions were held with local authorities, car park and bus service operators and city centre managers as appropriate. Questionnaire surveys based upon the behaviour and views of Park and Ride users were carried out in each of the sites, with a second post back questionnaire used in six of the eight sites for non-users of the park and ride. This would help separate the views of these two groups. Overall 1,479 of the questionnaires were returned (25%).

Results found showed that:

• Non-users and users were of a similar age and trip purpose;
• Over two thirds of users were female, with non-users being an even split;
• Half of users use the Park and Ride at least once a week;
• 85% of people drive to the Park, 11% walk, 2% are dropped off and 1% cycle;
• 92% of non-users know about the alternate possibility of Park and Ride; and
• 21% of users drive less than 2km to the site, indicating the potential for access by cycling or walking);
• 16% of those questioned said that they would not have made the journey had the Park and Ride been unavailable;
• Trip generation is high, particularly in Brighton (18%), Coventry (21%) and Reading (18%).

The study found that private car mileage did experience a net decrease from the use of Park and Ride facilities and the value of the decrease varied from site to site. The reduction in traffic was more significant in cities that also employed transport strategies that aim to remove long stay car parking in the centre and move it to the Park and Ride site, this effect was noted in York and Cambridge. This policy in York and Cambridge is helping to make it viable to increase the number of park-and-ride sites so decreasing potential diversion distance over time.

Other conclusions were:

• Consideration should be given for the provision of enhanced facilities for pedestrians and cyclists to encourage a modal shift in the case of short access trips,
• Bus abstraction will be experienced in areas where Park and Ride competes against local bus services,
• Park and Ride sites situated further from the centre allow for greater savings of mileage, although suitable sites to counteract the need for diversion on radial routes are ideal,
• Diversionary trips can be reduced further by comprehensive policies which include the location of Park and Ride sites on all key routes.

Reasons for non-users to avoid Park and Ride were perceived speed, ease of driving directly into the centre, proposed length of stay and limited mobility. Whereas cost, convenience, reliability, frequency, difficulty in parking and faster journey time were the reasons given by users for why they switched to Park and Ride. Of these factors, the non-users and users pointed to cost and journey time being the most likely to influence their decision on whether to switch to Park and Ride or not.

Table 7: Contribution to objectives

Objective	Comment
	A reduction in traffic in urban areas during the peak in particular represents an efficiency gain. The benefits may be partially negated by the increase in traffic in the scheme's catchment area.
	A reduction in traffic in urban areas during the peak in particular is likely to reduce air and noise pollution and perceptions of danger. The benefits may be partially negated by the increase in traffic in the scheme's catchment area.
	The net reduction in overall vehicle kilometres (urban plus nonurban) will lead to a reduction in CO2 emissions as long as the freed road space is not simply filled with traffic transferring from other modes (induced) and newly generated trips.
	Uncertain.
	Reduced road traffic in urban areas is likely to reduce accidents affecting vulnerable road users (pedestrians and cyclists) whilst the extra traffic in higher speed nonurban areas is likely to increase accidents that injure vehicle occupants.
	Any reduced traffic congestion is likely to be economically beneficial as is the alternative means of accessing the city centre. Furthermore, the newly generated trips are likely to boost the city centre. On the other hand taxes necessary to support the majority of schemes may stifle growth.
	Operating subsidy required in every case but one.

Parkhurst 1999 and 2000

A study by Parkhurst (2000) reviewed the work of Atkins for the DETR and offered an alternative appraisal of the same eight park and ride sites by separating the analysis into urban and outer-urban components. Parkhurst considered three types of traffic increase: 1) drivers who are intercepted detouring to reach park and ride sites; 2) passengers switching from existing public transport services; and 3) drivers making additional trips.

Some of Parkhurst's key findings supported whilst others contradicted the Atkins report. These are presented below along with supporting tables.

a) Impact On Urban Areas

The reduction in car vehicle kms is greater than the additional car-equivalent distance run by park and ride bus services in seven of the eight case studies. Bus kilometres are weighted by 2.5 compared to car to reflect the buses’ increased impact per kilometre on congestion, road wear and the environment. In Coventry the additional bus traffic was 2.6 car kms greater per intercepted car, whilst in the remaining seven case studies the reduction in car kms per intercepted car ranged from 1.1 (in Brighton) to 6 in Shrewsbury. Whether this actually results in such a reduction is dependent on the extent to which freed up road space is filled by additional urban traffic.

Table 8: Changes in Traffic in urban area Per Intercepted Car (car kms)

b) Impact On Outer Urban Areas

Results presented in the table below (column 3) suggested an increase in traffic outside of the urban area in all eight case studies as the result of intercepted drivers detouring to reach park and ride sites, passengers switching from existing public transport services and drivers making additional trips. The first impact (detouring) can add between 1.5 car kms (Reading) to 6 car kms onto the outer urban journeys of intercepted car drivers. The second impact (modal switch) adds between 3.4 car kms (Coventry) to 20.5 car kms (Norwich) per intercepted car. The third impact (generated trips) increases car kms per intercepted car by between 4.1 (Shrewsbury) and 25.7 (Norwich).

c) Net Impact on Traffic Growth

The net impact on traffic growth is to increase traffic on the road network overall for each case study by between 0.9 (Shrewsbury) to 20.7 (Norwich) car kms per intercepted car.

Table 9: Changes in Traffic Per Intercepted Car (car kms)

Parkhurst concludes, “Urban-fringe bus-based park and ride provided with dedicated bus services is better described as a policy of car traffic redistribution than a policy of car traffic reduction”.

d) Benefits of Traffic Transfer From Urban to Outer Urban Areas

This will be beneficial where urban congestion and/or environmental externalities in the urban areas are more acute than in the outer urban areas. There might however be greater urbanisation around the park and ride catchment area.

e) Costs and Revenues

After comparing the annualised operating costs & capital costs with revenue from user charges only one scheme (Brighton) was estimated to cover its costs, whilst the others required a subsidy per car intercepted that ranged from £0.11 (York) to £5.87 (Coventry). In terms of subsidy cost per intercepted car km these ranged from £0.02 (York) to £0.53 (Cambridge).

Table 10: Capital and Operating Costs of Park and Ride Sites

Table 11: Cost of Urban Traffic Interception Due to Park and Ride Per Weekday

Table 12: Contribution to Objectives

Objective	Comment
	Overall vehicle kilometres increased in every case but were redistributed onto less congested roads that in the short term at least is probably more efficient. However, any efficiency gains will be negated if freed road space fills up in the urban area. It is notable that there was a net increase even in the towns that have a long-term policy to reduce parking in the town centre. With projected increases in traffic levels, the long-term effect may be simply to spread the congestion problem beyond the city boundary.
	Traffic reductions in all but one of the case studies in the urban area are likely to reduce noise, local air pollution and perceptions of danger. Problem has been moved to possibly less sensitive nonurban roads which may represent a net improvement.
	No direct evidence presented but it is possible that whilst overall vehicle kilometres have increased, because they are driven on less congested roads, in the short-term, emissions of CO2 and local pollutants may see a net decrease for some of the schemes studied.
	No evidence presented, impact difficult to judge.
	No direct evidence presented but likely reduction in accidents in urban area but an increase in nonurban area.
	Provides an attractive alternative for accessing shops and employment in the urban area and newly generated trips may boost the city centre. On the other hand the taxes required to support all but one of the schemes may stifle economic growth.
	All but one of the schemes run at a significant loss and are therefore a financial drain on the body that is responsible. There is also significant abstraction from existing public transport which may increase the subsidy requirement for those services.