Abstract

The main cause of road traffic congestion is that the volume of traffic is too close to the maximum capacity of a road or network. Congestion in the UK is worse than many, perhaps most, other European countries. More important, it is getting worse, year by year. Current official forecasts imply that congestion will be substantially worse by the end of this decade, even on the very favourable assumption that all current Government projects and policies are implemented in full, successfully, and to time. This is because road traffic is growing faster than road capacity. This is not a temporary problem: it will continue to be the case, in the absence of measures to reduce traffic, because it is infeasible to match a road programme to unrestricted trends in traffic growth. The effect, using the current Government method of measuring congestion, and a long established method of valuing it, would be that the widely quoted figure of an annual cost of £20 billion, would increase to £30 billion by 2010. Under current social and economic frameworks, there are no feasible policies that could reduce congestion to zero in practice, or that would be worthwhile doing in theory. But savings worth £4b-£6b a year could in principle be made by congestion charging alone, over the whole network, of which (very approximately) half might be reflected in the prices of goods, and half in savings in individuals’ own time spent travelling. A good proportion of this could alternatively be secured by an appropriate package of alternative measures: priority lanes and signalling; switching to other modes including freight to rail and passenger movements to public transport, walking and cycling; ‘soft’ policies to encourage reduced travel by car; land-use patterns which reduce unnecessary travel; and associated measures to prevent benefits from being eroded by induced travel. The combined effects of road charging and a supportive set of complementary measures represent the best that could be reasonably achieved in the short to medium run. This could reduce congestion costs (as distinct from slowing down their increase) by 40%-50%. These broad-brush figures, though based on long-established methods, must be treated with great caution. The ‘cost of congestion’, as used for these calculations, is based on relationships which in reality are not exact, stable or even meaningful. The wrong indicator has been used, comparing average real speeds with average ideal speeds. But in the real world, speeds are different every day, and so is the level of congestion. For just-in-time operation, and for much personal and business travel, variability and reliability are much more important. The really costly effect of congestion is not the slightly increased average time, but the greater than average effect in particular locations and markets, and the greatly increased unreliability. During the near future, until road pricing is implemented, increases in road congestion can lead to some shift in the balance of attractiveness of rail freight, sufficient for a proportion of the freight market to transfer from road. This would in turn make a small but significant contribution to reducing congestion, especially in some specific important corridors. Even though rail freight is usually a small proportion of all freight, the annual economic saving in congestion cost, to road users generally, from transferring a 5-times a week, 200 mile round trip, mostly on congested motorways, from road to rail would be in the order of £40,000 to £80,000, to which should be added the commercial cost savings made by the freight operator who chooses to do so. It should be emphasised that sustaining this would require measures to prevent induced car traffic filling up the relieved road space. An example of the impact of factoring in unreliability is given by approximate calculations made for journeys such as Glasgow to Newcastle, Cardiff to Dover, or London to Manchester. In free-flow theory these could be 3-hour journeys, but moderate congestion requires adding an hour to the average time and another hour safety margin to ensure that a tight delivery slot is not missed too often. In congestion so severe as to double the average time, the extra safety margin for unreliability could be as much as 4 hours, which is simply not feasible in many cases. The ‘total cost of congestion’ is a large number, but it is practically meaningless and by ‘devaluing the currency’ it distracts attention from more important, achievable, objectives. It would be better not to use it as a target for policy. The two key important things to do are: · Strategic action to reduce traffic volume to a level where conditions do not vary too much from day to day. In some circumstances this will slightly increase average speed, though not always: in some road conditions a reduction of average speed can greatly improve the smoothness of traffic flow. But in both cases, it will greatly increase reliability, this being more important than the change in average speed; · Practical measures to provide good alternatives for freight and passenger movements which reduce the intensity of use of scarce road space in congested conditions. Even where this only applies to a minority of movements, significant effects are possible. The Government plans to ‘re-launch’ the Ten Year Plan for Transport this Summer or Autumn. It is not reasonable to expect that the re-launch will include congestion charging for cars within the decade, so it will need to plan for it as soon as possible after, and a short-term coping strategy of priority measures to protect the most important classes of movement (both passenger and freight) from congestion in the period before charging is implemented.