تجزیه و تحلیل هزینه فایده از سیستم شارژ تراکم استکهلم

This paper presents a cost–benefit analysis of the Stockholm congestion charging system, based on the observed rather than on the model-forecasted data. The most important data sources are travel time and traffic flow measurements made in the year before the charges were introduced (during April 2005) and during the first spring with the charges (during April 2006, 4 months after the charges were introduced). Using matrix calibration, effects on the non-observed link flows and travel times are extrapolated, enabling us to calculate the social value of changes in travel times and travel costs. Impacts on traffic safety and emissions are calculated using standard Swedish CBA relationships. The system is shown to yield a significant social surplus, well enough to cover both investment and operating costs, provided that it is kept for a reasonable lifetime: investment and startup costs are “recovered” in terms of social benefits in around 4 years.

The so-called Stockholm trial consisted of two parts: a congestion charging scheme that was in place between 3 January and 31 July 2006, and an extension of the public transport supply that was in place between 31 August 2005 and 31 December 2006. The trial was followed by a referendum (the result of which is reported and described in Eliasson et al., 2009). The charges were then reintroduced as a permanent system on 1st August 2007. A general description of the charging system can be found in Eliasson et al. (2009). It consisted of a cordon around the centre of the city of Stockholm, with a charge imposed 6.30–18.30 weekdays. The charge was 20 SEK during peak hours and 10 SEK during off-peak, levied in both directions across the cordon. The maximum amount payable per vehicle and day was 60 SEK. Various exemptions (for e.g. taxis, buses, and alternative-fuel cars and for traffic between the island of Lidingö and the rest of the county) meant that about 30% of the passages were free of charge. The purpose of this paper is to present a cost–benefit analysis (CBA) for the congestion charging system. What separates this from most other transport investment CBAs is that it rests mainly upon measured data – that is, not on modelling results. Most of the data stem from extensive traffic measurements during April 2005 and April 2006. The first underlying assumption is that the changes in traffic between 2005 and 2006 were only due to the introduction of the congestion charges. These assumptions are discussed in Section 6, where we argue that even if there are other factors affecting the traffic between the two years, they are likely to be small in comparison. The second underlying assumption is that the effects that could be seen during the period when the charges were in place will remain also in the future. Obviously, the analysis presented here are only based on short-term effects. There are relevant reasons to argue both that long-term effects may be higher and that they may be lower. Long-term effects are also discussed in Section 6. A particular potential problem is separating the effects of the charges from the effects of the transit extension. The costs and benefits of the public transit extension as such are not analysed in this paper, although we take into account the increased transit crowding caused by the charges, and the extent to which this was ameliorated by the extended train services. The extension of train services was a comparably small part of the public transit extension (63 mSEK of over 1300 mSEK), while almost all the rest of the cost went to the purchases of new buses and operating costs for new bus lines. Since the congestion charges were postponed (due to legal complaints on the tendering process), it so happened that the public transit extension started earlier than the congestion charges, and its effects on car traffic could hence be separated out. It turned out that it had no measurable effect on car traffic at all, a finding which was corroborated by on-board surveys, where only 4% of the passengers stated that they were “former car drivers” (Brundell-Freij and Kottenhoff, 2009). Since the new bus lines had around 14 000 boardings per day during spring 2006, this would mean that the effect of the new bus lines on car traffic amounted to somewhere around 600 car drivers per day. Comparing this to the decrease in car traffic caused by the charges (around 100 000 less vehicles per day across the cordon during charged hours), it seems safe to assume that the traffic decrease was virtually exclusively caused by the charges, and the travel time gains in the cost–benefit analysis can hence be attributed to the charges. Section 2 discusses investment and operating costs, and Section 3 the marginal cost of public funds. Section 4 presents in some detail the main benefit, namely the time gains. Section 5 presents other benefits – effects on traffic safety, emissions, transit crowding, etc. Section 6 discusses short- vs. long-term effects, and whether traffic was affected by other external factors between 2005 and 2006. Section 7 summarises all costs and benefits, and conclusions are drawn in Section 8.

Even if it is well established that perfect congestion pricing will yield a social surplus, it is neither evident that it will be enough to cover investment and operating costs, nor that a real congestion pricing system, with all its practical and political limitations, will be socially beneficial. The present analysis demonstrates that the Stockholm system yields a large social surplus, well enough to cover both investment and operating costs. The value of the time gains compared to the paid charges is remarkably high compared to most theoretical examples. This seems to depend on that such examples neglect network effects – for example, that many travellers not paying the charge will benefit from decreased congestion, and that bottleneck effects may lead to reduced congestion far away from the charged area. It is difficult to predict the development of costs and benefits in the long run. It seems likely that operating costs would decrease over time, from the 220 mSEK/year currently forecasted by the National Road Administration perhaps down towards the 100–150 mSEK/year of the Oslo system. It also seems likely that the benefit of a congestion reduction would increase over time, since the level of congestion will almost certainly increase. In reality, this may make it prudent to change the charge levels with increased traffic volumes, but the analysis of that question is left outside this paper. Finally, it is difficult to predict or even define the lifetime of the system – depending on whether this means the hardware (which is a fairly small cost), the computer programs (a larger cost), the system architecture (an even larger cost, and more durable) or is seen as a socio-technical-political “construction” – the latter being the most difficult and expensive to obtain. But given that the political will to keep the system remains, it seems that a lifetime over at least 20 years is a conservative estimate. According to the analysis carried out here, the system will have recovered its investment cost in terms of social benefits in about 4 years. In financial terms, investment costs are recovered in around 3.5 years. Hence, it seems that the encounter between the “theoretically irrefutable” idea of road pricing [in the words of the Smeed report (Ministry of Transport, 1964)] and reality will end in road pricing proving its case – even in a case where investment and operating costs certainly could have been lower, had the political process leading up to the implementation worked more smoothly.