Congestion is perceived as a problem primarily because of the broad rangeand scale of impacts it imposes on individual travellers, the economy andsociety including delays, frustration, pollution and reliability problemsamongst others. Unsurprisingly, reducing congestion forms a key element of transport policy at all levels of government from the EU to the regionallevel. However, despite the widespread use of the term congestion there is still some ambiguity regarding how this is defined and what constitutes a state of congestion in practice. There is also disagreement on what the actual cost of congestion is - for example estimates for the UK range from2 billion per year to the often quoted Confederation of British Industry estimate of 20 billion per year. A further policy conundrum centres on a dual-role that congestion has in the system: firstly in inhibiting economicgrowth and secondly in restricting the environmental burden that transport places on the environment. The challenge for government policy is therefore to reduce the cost of congestion to the economy whilst simultaneously achieving a sustainable transport system. The purpose of this paper is therefore twofold - to set out the analytical approaches in seeking to empirically measure congestion and in deriving a figure for the cost of congestion. Some explanation is also given for why such a range in estimates of the latter exists. From this, measures of congestion appropriate to policy formulation can be identified which then point towards policies that might address the tension between alleviating congestion and managing travel demand. Two fundamental approaches to interpreting congestion are discussed: a 'traffic engineering' perspective (essentially relating volume to capacity) and an economic view (reflecting the impedance vehicles impose on one another). The latter is associated with measuring the cost of congestion, with three economic terms emerging that could justifiably be called the cost of congestion. These are the Marginal External Cost of Congestion (MECC), the Total Cost of Congestion (TCC) and the Excess Burden of Congestion (EBC). Early estimates of the cost of congestion focused on the TCC term, however, a shift in policy and research interest to demand management (particularly the impacts of road pricing) has led to increased interest in the MECC and EBC. A fourth term also exists that measures the cost of congestion against a baseline of optimal capacity with prices for road users equal to marginal social costs, though evidence on this is very scarce. The paper will review the range of evidence for each of the four economic terms, explaining the appropriateness to policy of each. Reasons why there are large variations in the estimates will also be proposed, alongside indications on how each term can be measured. At the practical level of empirically measuring congestion, approaches can be generally classed as travel time (or speed) based measures, volume based measures, area based measures and summary indices (or more complex model outputs). There is, however, a very large range of potential measurement methods which itself reflects how difficult it is to measure congestion on the road network. The review found that in practice the simpler measures are more commonly applied than relatively complex measures. In terms of the relationship between the way in which congestion is empirically measured and how the costs of congestion can be calculated, empirical measures of congestion do not lend themselves to monitoring changes in MECC or EBC but do support calculation of the TCC. Furthermore (as illustrated with data for Scotland) analysis of such measures indicates that even where congestion is perceived by the travelling public to be a problem, the percentage of trips that experience delay or reliability problems can be surprisingly low. The analysis also describes how congestion may be in evidence throughout the road network, but large geographic, temporal and journey purpose variations exist. In determining the need for policy measures it may therefore be the case that perceived congestion is as important as more objective evidence. The paper concludes bysetting out the policy challenges associated with reducing the cost of congestion without encouraging travel demand to increase. Practical challenges are also outlined in developing methods to monitor progress in reducingeach of the different measures of the cost of congestion, as well as improving our estimates of the cost of congestion. Throughout the paper the arguments are illustrated through examples, principally though not exclusively, drawn from the UK. For the covering abstract see ITRD E137145.
Abstract