Evidence about park and ride services derived from a number of cities is cited, but inevitably most relates to Oxford. This is the case because Oxford has the largest and oldest system. It is the system which has been the most extensively studied, and is the example known best by the author . However, the intention has been to use this experience to provide a commentary that is sufficiently general and robust to be relevant to other cities.
What problems have park and ride services sought to solve?
Over the last three decades, short-range bus-based park and ride schemes have been implemented with various intentions and justifications. The present section identifies eleven variations on six main aims.
1 To provide an alternative to car use and public transport.
a) To reduce the amount of car use within the host city by people who live beyond reasonable walking distance of public transport services, by providing for an alternative means of accessing the bus stop.
b) To reduce the amount of car use within the host city by people who live within reasonable walking distance of public transport services, but are nonetheless unwilling to use them. This is achieved by providing an appreciably better service for park and ride passengers, either in absolute terms, or relative to the alternatives.
2 To intensify vehicle occupancy in the city.
a) To permit a greater number of trips to access the city for a similar number of vehicle movements and similar amount of road space.
b) To reduce urban congestion by increasing average vehicle occupancy while holding the number of trips to (or through) the city centre constant, and so reducing the flow of vehicles ‘downstream’ of the park and ride site.
3 To allow for more economically efficient provision of parking capacity.
a) To increase total effective parking provision for the host city without using scarce, valuable, city centre land.
b) To transfer parking capacity from the central area to the periphery so as to make city centre land available for other purposes.
4 To improve journey quality for the motorist.
a) To save motorists and their vehicles the stresses of congestion, and/or locating a parking space in an urban area which may be unfamiliar.
b) To lower the perceived generalised cost of travelling into the host city centre so as to attract more trips to the city. This may be achieved by reducing either:
· the actual or perceived relative money costs for park and ride compared with one or more of the alternatives, or
· the perceived effort costs relative to the other modes available to the individual traveller.
5 To make overall transport policies politically acceptable.
Park and ride is popular with both motorists and traders and can be used as a bargaining device to gain support during the political negotiation process for a traffic restraint strategy.
6 To contribute to environmental objectives.
a) To assist in meeting local air pollution targets or addressing hotspots by allowing either a reduction of traffic levels in the central area or a restructuring of traffic flows away from the centre. Buildings in the protected area can be expected to benefit from less stonework damage and vibration damage. Quality of life should be improved in terms of greater freedom of movement and reduced visual intrusion by traffic.
b) To contribute to national and international goals for global environmental management, notably targets for the control of carbon dioxide emissions, by reducing emissions due to car use.
How far has park and ride been successful in achieving these aims?
1a) In terms of attracting the classic car-dependent motorist from a very rural location, park and ride can be regarded, with some reservations, as successful. Taking the example of Oxford (Table 1, after Parkhurst & Stokes, 1994), data showed that at least 40% of users do live in rural areas, defined as settlements with an area smaller than 1km2. Another 12% do not live in classically rural contexts, but only have direct access to ‘poor’ public transport.
Table 1: Types of Origin of Oxford Park and Ride Users Related to Public Transport Provision Origin % of Sample Urban with ‘very good’ public transport provision 26.2 Urban with ‘good’ public transport provision 9.9 Urban with ‘poor’ public transport provision 12.2 Rural (assumed to have poor provision) 39.8 Beyond 45 km (origin types not analysed) 11.9
1b) At the same time, though, at least 25% of users live within 1000 m of a bus service with a departure at least every 10 minutes in the a.m. peak. Another 10% live within similar range of a bus or rail service that is not quite so frequent or attractive. Such services still offer a reasonable, although not necessarily very attractive alternative.
2a) Park and ride has been very successful in congested cities in providing a practical and acceptable means of concentrating vehicle occupancy. In Oxford, for example, traffic levels in the city centre have fluctuated since 1974, but not increased. Over this period, car ownership and use have increased in the City’s suburbs and elsewhere in Oxfordshire. A gradual increase in park and ride capacity over the same period has offered a means of providing for traffic growth without constructing new roads or increasing parking capacity in the city centre. From the perspective of the retail trade, the alternative would probably have been for some shoppers, deterred by parking charges, to have switched their allegiances to more car-friendly centres, whilst Oxford city centre retailers would have further specialised on their university and tourism markets. In practice, the extra flow of people has resulted in a total expansion in retail activity, with the traditional retailing sectors largely maintained in a context of greater worldwide travel for tourism.
2b) Despite park and ride increasing vehicle occupancy, and the fact that, on average, 65% of vehicles parked at surveyed sites have been intercepted before entering the urban network, there is little or no evidence in the UK of significant congestion reduction directly attributable to park and ride provision (Parkhurst, 1996). There are a number of contributory explanations as to why this is the case.
· The role of park and ride is generally not large enough. Against the fluctuations in traffic caused by the background context of the national and local economies, one or two small sites might not be of sufficient magnitude to make a countable difference. Even in the case of Oxford, with the largest provision, park and ride in 1991 accounted for just 5% of all mechanised trips to the city centre. For trips to the centre from outside the city limits this share rises to 11% (CBP, 1992).
· Use of park and ride tends to build up over time, perhaps following capacity expansion through site extensions or new openings. As noted above, background traffic growth in this period may be able to exploit the benefits of freed road capacity and/or greater parking capacity.
· If, instead, the context is one of congested equilibrium, rather than background traffic growth, it is likely that suppressed demand will speedily exploit freed capacity.
· In particular, it has to be remembered that not all traffic using a congested radial route has a destination in the city centre. Some trips have destinations just outside the centre, or are merely using the city centre as the shortest available route. Such trips are not deterred by central parking charges and, unless they are restricted by other means, are actually attracted by reduced congestion due to other travellers being deterred.
· 35% of trips to park and ride sites are not intercepted, so the decongestion effect of the site is not as great as the crude usage data would suggest.
3a) Park and ride has had a notable success in enabling some cities to address parking problems without building new car parks in the city centre. In some cases this has been by providing bus links to existing, poorly-located sites or by utilising temporary sites for periods of peak demand. Most usually, such periods are Christmas or the summer tourist season. Hence, land which would have been converted for the relatively inefficient activity of parking has been used instead for housing, retailing or public space.
3b) Park and ride has less often formed part of a strategy to transfer parking capacity from the central area to the periphery, so as to make city centre land available for other purposes without increasing total parking supply. This resistance is understandable. Due to the demand for parking, and the investment of capital in the infrastructure, it is politically difficult to close down established facilities, and if they are owned by the private sector, virtually impossible. However, where on-street facilities are withdrawn or major redevelopment allows a renegotiation of parking standards, a reduction in capacity may be possible.
4a) Surveys of park and ride users show consistently that the facility is appreciated as a service. Avoiding the stresses and strains of driving is the prime reason for use in a significant minority of cases (10-15%). It is a partial explanation for anecdotal evidence that park and ride is used by a small number of people with free, private parking. On the other hand, concerns about the security of cars parked at park and ride sites and of those people who need to return alone to their cars in relatively remote locations after dark for their personal safety have been long-standing causes of resistance to use.
4b) The increased relative ease and lower cost of travel are outcomes of park and ride implementation which have sometimes been intended, where a commercial centre has been experiencing economic decline, and sometimes unintended. Reduced travel costs are important explanations for the phenomena of trip generation/attraction (around 6-12% of weekday trips to a site) and mode-shifting away from public transport services (depending on the particular survey and city considered, ranging from 8-40% of weekday trips to park and ride sites) (Parkhurst, 1996).
Park and ride often proves the cheapest means of accessing the city. This is the case for a combination of two reasons. First, park and ride is subsidised. The cost of providing and operating park and ride sites (whether marketed in the form of free parking or free bus travel) is considerable, and not fully met by the users. Second, park and ride bus services are relatively efficient to operate. The motorist pays, through his or her investment in private transport, for the part of the journey for which the public transport operator would find it expensive to provide, leaving the latter to operate high-volume limited-stop services from a very concentrated point of demand. In a deregulated economy for public transport services, the result is a rate per unit distance that undercuts conventional public transport fares.
These effects arise in addition to the more conventional problem of motorists tending to perceive only their day-to-day running costs as the relevant outgoings when comparing the relative costs of the modes available.
5) There seems little doubt that, if implemented along with the appropriate bus priorities, traders do view park and ride as an important contribution to the prosperity of commercial centres. Although they may initially view park and ride, and public transport priority, with suspicion, preferring instead to see conventional car park construction and lower parking charges, they later come to appreciate that it is the pedestrian shopper who is most critical for buoyant trade, and that high-efficiency modes can bring more people into the centre. In their submissions to the formal consultation into plans for pedestrianisation in Oxford, the park and ride strategy was cited by a number of retailers as an important element in the future viability of the centre, and a component which they believed ought to be expanded.
Satisfaction amongst the majority of traders, together with popularity amongst users, suggests that park and ride has a real influence on the acceptability of policy. However, it is the overall package that has to be considered, and its subsequent effects. Park and ride cannot be regarded as an absolutely essential element, as publicly-acceptable strategies without this approach can be envisaged, and indeed, do exist. Rather the policy makes an important de facto contribution in the cities in which it has been developed to a significant degree.
6a) Within package approaches, park and ride can be regarded as contributing to attempts to control pollution where it is most concentrated, in the city centre. However, current state-of-the-art strategies seek ‘demand management’ rather than an actual reduction in demand. They provide for most of the displaced traffic to be accommodated elsewhere on the road network. Such policies may result in some suppression of demand for car use, by deterring ‘marginal’ trips and by persuading travellers to switch mode, but the principal effect will be achieved by transferring trips from the most sensitive area (typically the city centre) into the suburbs, and perhaps from the suburbs to the road network beyond the urban area. In this context, the transfer of pollution is clearly not to be welcomed, but as most of the pollutants involved are regarded as having safe levels, dispersion can be regarded as a legitimate strategy.
6b) At the global level, the role of park and ride in environmental strategy is much less well-defined, and the impacts uncertain. Park and ride services intercept car users and place them on relatively efficient vehicles. However, they do so for only a small proportion of the total trip. Data from Oxford and York (Parkhurst & Stokes, 1994) suggest that, on average, a car will be intercepted for 10-25% of the total trip, depending on the distance of the site from the city centre. The buses provided for the park and ride shuttle are not themselves environmentally neutral, but, other things being equal, simple calculations suggest that a modest environmental benefit in terms of reduced energy consumption and less carbon dioxide production would be derived.
In practice, however, as noted above, some 35% of trips must be considered in a different way to the above description, as they are trips transferred from public transport, or trips that have either been generated or attracted.
· Generated trips are the most straightforward to consider as all the car-kilometres travelled are additional.
· Similarly, in most cases of abstraction from public transport the service previously used will still be operated in the same way, with the same environmental impact shared between fewer passengers, while the car use of the abstracted passengers is additional.
· Trips attracted from another retail facility are problematic, because they may be either longer (because park and ride is particularly attractive) or shorter (because prior to park and ride provision people travelled further to take advantage of lower parking charges elsewhere).
The net implications strongly suggest an environmental cost for these effects. Furthermore, given the short distance of the bus journey, additional car trips will be very significant in the overall audit of the impact of park and ride.
Besides these unintended effects, impacts on land-use, both positive and negative, can be hypothesised. As part of a traffic restraint exercise which allows better access to the city centre, park and ride assists in reducing the demand for the expansion of out-of-town facilities. However, as noted above, that strategy may itself push traffic onto longer routes with unclear energy consumption effects. Further, reducing the cost of travel into the city by providing park and ride may encourage people to live further from the city, and/or live in places which are remote from public transport services. The historical precedents are clear, first with rail commuter development in the first part of the century, and later, with the development of one-car households, ribbon development allowing car commuting by the head of household and public transport use by the other members.
The effect of more people living in more remote rural locations would be to alter wider travel patterns in a way which is likely to increase demand, and at the same time make that demand less easy to provide for by high efficiency modes. This would mean an increase in car dependence.
As a minimum, claims that park and ride can assist in achieving global environmental targets have to be regarded as unsubstantiated. Further, in the current operating context of park and ride, they are probably ill-founded.
Comments on the important issues in the future development of park and ride policy
The areas in which park and ride policy is in need of development are:
1. how to make an effective, direct contribution to traffic reduction,
2. how to better target use of the facilities, and
3. how to play a supporting role to increasing the mode-share of public transport.
1 How can a de-congestion benefit be realised?
Two essential conditions for reducing traffic by intercepting cars at park and ride sites are that:
· park and ride site provision should not lead to a net increase in parking supply, and
· vacated road space should not be attractive to other car users.
In the past, parking restraint measures equal to the impact of the park and ride site have often not been introduced, because the most important policy motivation has been increasing total accessibility and car access was regarded as an essential component of this policy demand.
For traffic restraint purposes, it may not be sufficient just to reduce one central space for each park and ride space provided. Introducing or increasing charges for central parking or moving to a short-stay only regime may deter commuters, but encourage more trips by shoppers and hence increase the average daily trip rate.
Private parking is a major contributor to parking stock in most cities and is largely beyond public control. The imminent Government White Paper on Transport Policy may introduce the possibility of achieving public control through taxation, which may, in turn, deter use. In the absence of legislation, local authorities will otherwise be able only to pursue policies ensuring minimum provision in new developments and schemes for pedestrianisation. The latter have the effect of making access to private spaces less attractive while not actually infringing legal access rights and still providing for operational and delivery needs.
Indeed, the importance of road restriction measures was introduced above. The Oxford example shows that this traffic can comprise nearly 25% of movements in an only partly restricted city centre (CBP, 1991), hence it is not sufficient to restrain car use solely through parking controls.
It is often argued that there is a need to put park and ride measures in place before other changes, which have identifiable ‘looser’-groups are implemented. This is an understandable reflection on the necessities of the political process. However, practitioners also have to maintain awareness of another pressure: the longer the gap between park and ride provision being made and real restraint being applied, the more likely it is that behaviour will adjust to take advantage of the road space and parking space made available by other travellers who switch mode. If the public is allowed to make these adjustments, it will then prove harder to introduce the less-palatable changes.
Looking to the future, the emphasis on car commuting as the major problem is likely to reduce somewhat. As traffic growth continues, trip rescheduling will further erode the conventional peak periods as people adjust their behaviour to counter congestion and the economy offers less ‘9 to 5’, Monday to Friday jobs. The need to work more flexible, perhaps unpredictable, antisocial hours, combining work with other activities, will lead to more people seeking to use cars in the interest of factors such as personal security. Similarly, an ageing population with an increasing number of driving licenses and cars will make a greater number of demands for preferential access by car on the grounds of limited personal mobility.
In the aftermath of the phase of out-of-town shopping centre construction, travel for shopping is presenting yet further challenges. Sunday trading and later opening hours have, in effect, resulted in a major increase in parking supply for the majority of motorists who are able and willing to reschedule their shopping activity to take advantage of slacker periods. It has probably both increased total shopping activity, as well as spread it over a longer week. In consequence, congestion ought to have reduced along with the average daily trip-rate, helping to disperse pollution over time. However, from the point of view of global environmental impacts, this suggests a further growth in emissions without any further infrastructure provision for the car having been made.
2 How can park and ride provision be better targeted?
The evidence that a minority, but significant, number of park and ride trips transfer from public transport services is of particular concern given the strong national policy consensus that conventional public transport should carry a greater share of trips. If trips are lost from public transport in the context of a growing market, the costs may not be obvious to either the operator or the transport planner. In a declining public transport market, the risk is that the loss of passengers could be the critical factor. Even if abstraction were as low as 10%, from a typical new site of 500 spaces, this would amount to 50 bus passengers. These passengers would probably not be attracted equally from across the network but lost from specific local bus routes. The threat is that this would lead to commercial services being de-registered, or the local authority deciding that it has insufficient funds to maintain the full extent of its supported network. While it has been argued above that park and ride can play a role in supporting the acceptability of the overall transport policy, which in turn can enable the development of public transport services, its contribution would be more effective if the losses to conventional public transport could be stemmed.
The principle reasons for people switching from public transport are perceived problems with the existing public transport services (push factors) and price (pull factors). In terms of push factors, the buses used may be less attractive, although those on park and ride services are not always superior. On average, there may be further to walk from the house to the conventional bus service than from the car to the park and ride terminal, although in some car park designs this can also be considerable. The waiting conditions are often more comfortable at the park and ride site (either onboard a waiting bus or in a superior shelter).
These push factors could be addressed. Better waiting areas and investment in priority for conventional buses would be a good place to start, but this will require investment. However, there are strong rumours that the coming White Paper will allow local authorities to hypothecate the proceeds of road charges for expenditure on public transport. This possibility is particularly relevant in the case of cities with a large demand for park and ride, which tend to be the historic ones, to have the most captive workforces and shoppers, to be the most attractive to tourists, and to suffer most from congestion due to historic infrastructure.
Turning to the pull factors, there is considerable scope to alter the balance of user charges for the competing modes. The ideal would be a hierarchy of charges that rendered public transport use without a car-leg the cheapest option, and a scale for parking costs which made them increasingly expensive towards the city centre, and included both conventional spaces and those at park and ride sites.
Park and ride bus services are efficient to provide, and in some cases are run on a commercial basis. This suggests conventional bus services are already at a disadvantage, and there should be particular concern about subsidies for park and ride. Local authority subsidies for transport facilities are sometimes necessary, to avoid the undesirable effects of the free market. Most obviously, public transport is heavily supported outside the major cities. Indeed, sometimes both conventional and park and ride services are both subsidised despite being in competition! However, car use itself is already heavily subsidised, whether through the provision of company cars, or because the full costs in terms of pollution, congestion and accidents are not internalised. Therefore, there should be great concern that, beyond the level needed for the initial promotion of a new facility, subsidies to park and ride do not reinforce the attractiveness of car use.
In any case, motorists derive benefits from park and ride, and to the extent that these are worthwhile to them, it should in principle be possible to capture that value through user-charges, as is routinely expected of public transport infrastructure projects. The most obvious source of subsidy is the fact that bus users pay only a limited contribution, if anything, towards the costs of site provision, which is instead funded by developer contributions and by local taxpayers from cross-subsidies within the local authority parking account, or from the wider budget. At Oxford, the principle of explicit user payments for park and ride site related services has recently been established with the introduction of a 50p parking charge linked to the introduction of security patrols.
In a context of public transport improvements and/or reduced subsidy to park and ride, public transport would provide an increasingly competitive alternative to park and ride expansion, especially as a certain proportion of current users might be attracted back to public transport.
There is a potential problem that a policy of recovering a greater proportion of costs from users could deter the small proportion of rural poor, i.e., those who are sufficiently well-off to own cars, but have no alternative transport because their origins are classically rural, and whose disposable income is sufficiently limited for the increased charges to be a genuine hardship. This issue must be treated as a real (if numerically small) problem, but the intention is only, at most, to cover the costs of site operation with a possible contribution for provision costs. Poverty is a consideration, but ultimately not the sole responsibility, of transport policy and such genuine hardship should perhaps be addressed through more basic means, such as mobility credits, or in general terms through income support.
3 Role of park and ride in future public transport development
Charging for the user-benefits of park and ride would tend to have the effect of discouraging a proportion of marginal use and reducing the extent of selection of park and ride instead of public transport services. Nonetheless, planners are likely to face growing demand for park and ride facilities. Providing short-range park and ride opportunities to all of the motorists living outside all but the largest urban areas and who would benefit from, or merely enjoy, using them must be regarded as an unrealistic strategy. This is especially true in the context of increasingly-vociferous campaigns, particularly those by local interest groups, about the development impact of providing sites within the rural-urban fringe. This begs the question of “how much capacity should be provided for which target users?”
Overall, research on the degree of car dependence suggests that perhaps 20% of trips can be regarded as inevitably car dependent and 20% are public transport dependent. However, the majority, 60%, are currently made by car, but could in principle be made by public transport, assuming provision were considerably enhanced (TSU/RDC Inc., 1995). The critical policy question is: “how far should the long-term role of park and ride be to provide only for those people who are classed as absolutely car dependent, and how far should it provide for those who could conceivably be provided with an alternative public transport service, but would rather they weren’t, because it would be marginally less convenient or comfortable?”.
One approach to the problem is to promote better integration of the options and allow the charging structure to guide travellers’ choices. Indeed, it is disappointing that experiments in combining conventional bus services with park and ride opportunities in the same way as they are provided on the rail system have not, hitherto, emerged. The synergies are obvious: the potential to combine service frequencies to provide a higher quality service or the same service for lower operating costs. There is also the possibility that such enhancements could initiate a virtuous circle, or at least halt the spiral of decline.
Three reasons explain why integrated services have not emerged. First, there has been an attitude amongst transport planners that park and ride services should be differentiated as far as possible from conventional bus services. The logic that motorists should be encouraged to regard park and ride services as something different followed from the idea that majority of motorists regarded conventional services as stigmatised. This motivation was understandable, but in the future would be counter-productive in a context in which public transport itself must be made more attractive. Fortunately, the intention to differentiate the two kinds of service has seldom been rigorously considered or developed, as demonstrated by the contradictory hopes of some promoters of park and ride that it might be the means of re-introducing motorists to public transport in the wider sense.
Second, congestion has a greater impact on bus services than car travel. This problem is addressed inside the city with bus priority measures, but not currently addressed outside the city. This is an issue that has begun to tax some local authorities. Oxfordshire County Council has received agreement and funding for short sections of bus lane alongside trunk roads, in order to link two park and ride sites with the city. The same authority has also commissioned a feasibility study into the possibility of adding a 15-km bus lane to the A40 route to Witney, which would initially target conventional services, although there would also be scope of long-range park and ride facilities.
Finally, there is the problem of the current payment-on-entry ticketing arrangements, which account for significant delays to all types of bus service, and erode the expensively-procured bus priorities provided by local authorities. The possible solutions are various, such as more use of prepaid tickets, exclusive use of prepaid tickets, purchase from on-board machines or the return to the use of conductors, and need to be addressed.
If all, or perhaps even some, of these problems can be resolved, there is no reason why appropriate city routes should not terminate at park and ride sites and longer-range park and ride strategies be explored. The vision for the future is perhaps of low-floor buses with comfortable seats operating on headways of 10 minutes or less, provided with priority both inside and outside the urban area, and in some cases using guideways. All waiting areas would at least be protected from the rain and perhaps feature electronic display giving real-time information. A variety of access modes would be supported. The emphasis and priority would be placed on foot and cycle, but there would be small-scale provision of parking facilities in villages to extend effective penetration of bus routes to the level of hamlets.
Conclusions
Park and ride services have been successful in attracting the car dependent, and those with alternative public transport that they are unprepared to use. They have played a role in packages that have been successful in managing the effects of traffic growth, but these packages have not actually resulted in reduced traffic. Provision of parking capacity at the edge of the city has meant more central land available for other purposes. Park and ride provides a valuable service to motorists, in some instances actually reducing their travel costs, and is politically popular amongst motorists and the business community. Park and ride can in principle make a contribution to addressing local environmental problems, through the appropriate package, but, at least in the short-range guise, is not a policy tool relevant to global concerns.
In the future, park and ride policy will need to target better those people who will remain car dependent even with a substantial improvement in the extent and quality of public transport services. The urban transport packages including park and ride will need to produce actual, measurable reductions in traffic, at least in the centre and probably also in the wider urban area. To this end, really effective parking controls will be required, coupled with the closure of all attractive routes passing directly through the city centre. Park and ride user charges will need to be aligned so as to provide an alternative to expensive city centre parking, but at a premium over rejuvenated public transport alternatives, so as in turn to provide appropriate signals within the housing market.
References
CBP (Colin Buchanan & Partners) (1992) Oxford Transport Study, Stage 1. Report to Oxfordshire County Council.
Parkhurst, G P & Stokes, G (1994) Park and Ride in Oxford and York: Report of Surveys (1994). Working Paper 797, Transport Studies Unit, University of Oxford.
Parkhurst, G P (1996) The Economic and Modal-Split Impacts of Short-Range Park and Ride Schemes: Evidence from Nine UK Cities. Working Paper 96/29, ESRC Transport Studies Unit, University College London.
TSU/RDC Inc. (1995) Car Dependence. Report to RAC Foundation for Motoring and the Environment, London.