Road Pricing Context
Implementation and Evaluation
Dutch National Case
Urban Road User Charging Online Knowledge Base
What Do We Know About The Theme?
The previous section discussed the importance of the environmental theme. In this section, focus is shifted to environmental impacts of the transport policy measure of urban road user charging.
Greenhouse gas reductions – Fossil carbon–dioxide.
Climate change is a global threat, and greenhouse gases are equally harmful irrespective of where they are emitted. Many (in particular economists) would therefore argue that local RUC has little relevance to CO2 policy (Johansson-Stenman, 2005). Nevertheless, there is reason to include CO2 in the discussion here since it is currently considered to be the most difficult and important future environmental threat from transport, and since general CO2 arguments (“car traffic has to be reduced”) are commonly used to support RUC policy.
For any given fossil fuel, carbon-dioxide emissions will be proportional to fuel consumption. Any urban RUC scheme designed to reduce vehicle mileage in the charged area, will reduce fuel consumption and, accordingly, CO2 emissions. As a consequence of lower traffic volumes and less disturbance, driving behaviour will also change. In most cases, the reduction of mileage will therefore contribute to a somewhat more-than-proportional effect in CO2 emissions, since driving in congested stop-and-go conditions consumes more fuel than driving at even, self-selected, speed. Generally, traffic conditions are important determinants of emissions and fuel consumption. For example, fuel consumption factors (fuel consumption per km driven) may vary by a factor 2 between different street types. (Ericsson and Brundell-Freij, 2001).
However, current empirical evidence seems to indicate that when aggregated, the effect RUC has on fuel consumption factors is almost negligible (compared to the first-order effect of reduced vehicle mileage). In the Stockholm trial, the overall reduction of vehicle mileage was estimated to be 14% within the charging cordon and 3% on a regional level (county). A model that took these reductions into account, but also accounted for changed speed profiles (and consequently changing emission factors), estimated the corresponding reduction in CO2 emissions to be 15% and 3%, respectively. The additional reduction due to more even speed profiles was thus not more than 1% (Carlsson et al, 2006).
Without doubt, any policy that should contribute significantly to the required reductions in CO2 emissions has to be applied on a national scale. Having said that, however, the reductions obtainable from local urban RUC are considerable compared to other potential local-regional policy measures. In the Stockholm case, the congestion charging contributed more to CO2 reductions in the county than the top listed measure on the regional authority’s action plan for reduction of CO2 emissions from transport.
Local air quality
There is a general consensus that local exhaust emissions from traffic are a serious threat to life and health of urban populations. This threat materialises in a three-step process, indicated in Figure 8 1.
Vehicles in traffic contribute to emissions, measured by weight (g). Amounts emitted depend on vehicle mileage (veh-km) and emission factors (g/vkm). Emission factors, in turn, are determined by vehicle (engine) characteristics and gear changing behaviour, but also by speed profiles, and thus by traffic characteristics.
Emissions are then distributed in surrounding air, resulting in a concentration of pollutants (g/m3). Concentrations are given by the amount of pollutants emitted, but also by weather conditions, topography, height of surrounding buildings and street layout, all factors that determine the volume of air into which emissions will be spread. The narrow and deep “street canyons” that are characteristic of many European city centres, will typically contribute to high concentrations.
Finally, concentrations of harmful substances will not harm unless taken up by those organisms for which they are harmful (humans, plants etc). High concentrations will therefore be less harmful in sparsely populated areas, while relatively lower concentrations can have large health effects if large populations are exposed to them.
Figure 8 1 Process by which vehicle emissions affect human health
As was shown in Section 8.2, concentrations of harmful substances (in particular particulate matter, PM) are often considerably higher than those maximum levels and targets for environmental quality that have been set by European law. As a consequence, local authorities in major cities are legally bound to take action to reduce these concentrations.
For health effects of local air quality, too, both reduction in vehicle mileage and reduction of emission factors typically contribute to an improved situation after RUC is applied . But in addition to these, a third factor contributes to an overall disproportionate positive effect: With Urban RUC, reductions in immissions are typically larger than reductions in emissions, since emission reductions appear where people are most exposed to them, and where concentrations are high – that is in densely populated (day and/or night), and densely built, city centres.
In both London and Stockholm, a substantial reduction of those vehicle emissions with negative health impacts is estimated to have taken place as a consequence of RUC schemes. See Table 8 1 .
In both cities, NOx and PM emissions decreased less than vehicle mileage. That is primarily due to bus traffic increases as a consequence of the charging package. In comparison with private cars, buses have high emissions of NOx per vehicle kilometres. In Stockholm, there were some (few) streets on which the resulting impact of increased bus traffic and reduced car traffic was negative, so that concentrations of NOx increased after charging.
Table 8‑1 Effects from RUC schemes in London and
Stockholm on vehicle emissions within the charging zone/cordon.
Change in CO2 emissions
- 13% (inner city)
Change in NOX emissions
- 8 %
- 8,5 %(inner city)
Change in particulate emissions
- 7 %
- 13 % (inner city)
Change in number of vehicles entering the zone
- 17 %
- 17 % (April)
As was the case for CO2, the effect from reduced mileage on local air quality seems to outweigh the effect of reduced emission factors. However, the effect of “efficiently” located emission reductions may be considerable. In Stockholm County as a whole the average immissions of exhaust-related PM10 fell by 6% during the trial (that is, aggregating concentrations, and weighing them by the number of affected inhabitants). Thus, the fact that emissions were reduced where people were most exposed to them doubled the effect from what would have been achieved if the (3%) reduction in mileage had been evenly distributed in the county (SLB Analys, 2006).
It should be noted that for most locations where environmental quality targets are exceeded, RUC alone will seldom be able to solve the problem. Often, traffic volumes will have to be reduced by a factor of 2 or more, since background levels of PM from other sources than traffic is considerable. To acquire such dramatic changes in car volumes at spots with local air quality problems, very strong economic incentives would have to be used; which means implementing extreme charges in small areas. Such highly differentiated charging schemes are less likely at present, but may become feasible in some years when GPS based technology for charging is technically available.
However, already those reductions that may be achieved through RUC may imply substantial health effects. Estimates of reduced mortality and health effects are of course highly sensitive to dose-response relationships applied in modelling. Due to lack of experimental data and the many confounding factors involved, the exact nature of the relation is yet subject to scientific discussion. In Stockholm it was estimated - based on a recent Norwegian study of dose-response relationships (Naftstad et al, 2004) - that 25-30 premature deaths per year could be avoided if the lower concentrations of PM10 implied by congestion charging were maintained (SBL Analys, 2006). In contrast, Tonne et al (2006) (based on another dose-response relation) estimated that the mortality effects of the RUC scheme in London were “modest”.
In a system where some vehicle categories are exempt from charging, we should expect increased traffic by those categories that can enjoy lower levels of congestion without charges. As an example, Transport for London data suggests that motorcycle traffic within the charging zone has increased by 15%. As a consequence emissions from such categories will increase accordingly. The effect, so far, is however small. In London for the first year of charging (2003-2004) it was estimated that increased motorcycle traffic increased PM emissions by 1% (TfL, 2008). If motorcycles had not been exempt, we would have experienced a 13% reduction of emissions, rather than the actual realised reduction of 12%.
In Stockholm, too, the contribution from exempted powered two wheelers (which includes motorcycles and mopeds) on emissions are minor, since these powered two wheelers constitutes less than 3% of the total traffic over the cordon during charging hours.
For London, Beevers and Carslaw (2005) estimated total NOx emissions to have decreased by 12% within the zone, while they estimated a corresponding increase on the inner ring road of 1.5%. This is an example of an issue for concern (dating as long back as Buchanan (1963)): that second-best charging scheme designs (i.e. cordons or zone-based systems), may divert traffic, and thereby re-allocate air quality problems without necessarily reducing them. In the typical case, however, population density will be higher in the area, from which traffic is diverted, than around those ring roads where there will be an increase. This supports the hypothesis that air quality (as an average over the affected population) will improve.
Nevertheless, there may be cases when the redistribution of traffic creates new environmental problems. Not least from an acceptability perspective, there may be reason to consider the potential regressive redistribution of welfare that may arise from relieving environmental pressure from (well-off) central districts, by moving those problems to less privileged suburban neighbourhoods (Jones (1998)). Such redistribution is a topic of concern for many types of environmental problems (not only those that are traffic related), and discussions have given rise to the concept Environmental Justice (see Dobson, 2003).
Negative effects on environmental justice may be compensated by a combination of measures in different package solutions. In total, Tonne et al (2006) estimated that the positive air quality effects in London were largest in more deprived areas, and that the RUC scheme therefore contributed to “reducing socioeconomic inequalities in air pollution impacts”.
In principle, traffic volume reductions are the most efficient way to reduce traffic noise disturbance, since such a reduction affects all noise everywhere simultaneously (for example both outdoors and indoors, in contrast to what can be achieved by double glazing). However, there is a problem in that very large reductions in traffic are necessary for the noise reduction to be noticeable, due to the non-linear sensitivity of the ear. A doubling of traffic is experienced as a just-about-noticeable increase in noise.
But the degree to which people are disturbed by traffic noise is highly individual and varies with emotional state and situation. Thus, reductions that are too small to be “observable” (on average) will cause some individuals to be less disturbed than they would otherwise have been.
In principle, it would also be possible to apply charging with the aim of reducing noise, for example by differentiating prices in sensitive areas and/or in sensitive periods of time such as night time. Another possibility would be to impose higher charges on noisier classes of vehicle. Conversely, exemption of motorcycles in London and Rome will have reduced the noise benefits of these schemes.
The traffic reductions achieved in the Stockholm trial would be expected to lead to an average reduction in noise of approximately 1 dB(A) (Miljöförvaltningen, 2006). This prediction was also confirmed by measurements.
Urban quality – Liveability and townscape
Urban streets are not only for (vehicle) travel but also for living and being, seeing and experiencing, walking and talking. Excessive car traffic is known to reduce the quality of all those activities.
In Stockholm, opinion polls show that inner city inhabitants were generally more positive to charging than people living elsewhere (Winslott Hiselius et al., 2009). (In Stockholm this was despite the fact that detailed analysis showed that inner city inhabitants paid more charges and gained less travel time than others). One possible explanation for the positive attitudes among inner city inhabitants was the extra benefits in liveability gained by those that spend more time (outside their car) in the neighbourhoods where car traffic is reduced.
However the attempts to measure and identify effects in these qualities were less successful in Stockholm. Neither “objective” criteria evaluated by expert architects (Gehl architects, 2006), nor subjective assessment as reported in a postal survey (Transek, 2006a), gave clear indications that liveability qualities had improved overall.
Neither are there clear indications in the corresponding studies relating to the London scheme, that the public have perceived an improvement of liveability aspects. Both citizens living inside, and outside, the charging zone refer predominantly to “reduction of congestion” as the main personal benefit of the scheme, and none of the five most frequently given (unprompted) responses referred explicitly to liveability (MORI, 2004). However, it is possible that there was nevertheless a (hidden) difference in the hypothesised direction. “Reduced congestion” may be used synonymously to “lesser travel time” by people going into the zone, while the same response may mean “emptier, nicer streets” when given by someone who lives there.
Where liveability and urban quality is regarded as a central objective for charging, it would often be possible to further enhance those effects through reallocation of “freed” road space to pedestrians and bicycles, and to use street furniture and landscape planning to change traffic oriented streets to streets oriented towards liveability. This could be achieved for example by pedestrianisation (concentrating vehicle traffic to some streets), reducing the number of lanes open to vehicle traffic, widen pavements and introduce bicycle lanes instead, or benefit from reduced demand for parking to allow more open space and green areas in cities. With such policies, travel speeds will not (or only to a lesser extent) be increased for vehicular traffic, despite a reduction in motorised volumes. Instead, that benefit will be exchanged for an improvement of the urban qualities for non-motorised road users (May, 1975).
Depending on the technology used (see Chapter 4), urban road user charging may lead to a requirement for new roadside signs and equipment. This can have an adverse impact on townscape, particularly in architecturally important streets.
Car fleet transition
According to EU policy a substantial part of the necessary reductions in greenhouse gas emissions are forecast to arise from a transition of the car fleet - from fossil fuel based to bio fuel based, and from cars with high energy consumption to more fuel efficient models. For this transition to take place, however, governments have to rely on the purchases of private car owners. Therefore, many countries are looking for incentives that may increase consumers’ willingness to buy “greener” cars. In such policies, RUC may play a role. Allowing green cars to be exempted from paying the charge may reduce life-cycle costs of owning such a car considerably for those affected by charging. In practical implementation green cars in Stockholm have so far been exempted from paying the charge all together. However, price differentiation might be another option where “green cars” pay the “correct” price with respect to the specific targets and all others pay even more.
Although the Stockholm charging trial lasted only seven months, there is anecdotal evidence that the fact that the scheme exempted green cars helped boost the market during that time. Also, this hypothesis is - if not confirmed - at least supported, by the fact that the share of “green” new cars on the market increased more in Stockholm than the Swedish average during the same period.
Modelled predictions of changes in the Swedish car fleet also indicate that other cost-reduction incentives of approximately the same total magnitude may have a significant effect on both the propensity to buy green cars, and car renewal rate for those affected (Naturvårdsverket, 2007). It has to be remembered, though, that it is typically only a small number of car drivers who will have much to gain by being exempt from charging. In a studied two-week period during the Stockholm trial, less than 5% of the private cars registered in the county paid more than the equivalent of €25 (Transek, 2006b).
The larger the part of the car fleet exempt from charges, the smaller effect will charging have on vehicle mileage. Therefore, before adopting an exemption policy for green cars, the balance between beneficial effects on car fleet composition on one hand, and detrimental effects on congestion relief, on the other, has to be carefully considered. The Swedish parliament’s decision to introduce congestion charges on a permanent basis in Stockholm included a time-limited exemption for green cars. For the first five years of charging (when the number of green cars in the fleet is forecast to be relatively low, and incentives at the same time can be expected to be necessary to stimulate the market) green cars will be exempt from charges. After this period (when a non-negligible part of congestion will be due to green cars), this exemption will be explicitly reconsidered.
In the intense debate on transport and the environment, one of the more overlooked negative effects is the intrusion into natural landscapes, recreational areas and ecosystems. EEA (2005) shows that from 1990 to 2000 “artificial land cover” – which is areas used for housing, industries, transport networks etc - increased by more than 10000 km2 in Europe. The major part of this uptake is due to housing, commerce and industry, but transport infrastructure adds to the picture.
Although there is now a broad consensus among researchers that metropolitan congestion can generally not be overcome by road building (Owens, 1995; Goodwin, 1997), neither political actors nor the public seem willing to accept this fact. If there is congestion, there will be public opinion pressure that new infrastructure is “needed”. Conversely, we would expect it to be simpler to protect natural landscapes from artificial intrusion if congestion is successfully tackled by other means, such as charging.
Theoretically, (from a normative transport economics perspective) it can be shown however that congestion charging may either increase or decrease the overall social benefit of infrastructure investment (WSP Analys & Strategi, 2007). The bottom line effect will depend on the detailed function of the networks and activity systems involved. However, when the effect of three real suggested investment programmes in Stockholm were simulated with and without congestion charging respectively, aggregate benefits decreased substantially if congestion charging was first introduced for all three investment programmes. Thereby social profitability decreased and (theoretically) the political desirability of building new infrastructure. Thus, for those three cases RUC should (again: in theory) reduce the public pressure for new infrastructure.
The argument to introduce congestion charging to save ecosystems from exploitation is not only theoretical. The Swedish Society for Nature Conservation has for many years been very active in the political debate for congestion charging. Their main motive for doing so is that they are convinced that only with RUC will it be possible to avoid new ring roads in the rural landscapes around Stockholm.