You are:

+ Share | Print page

Chapter 17. The process of prediction and evaluation of road accidents and its role in Motorway safety

Volume 1 Part 6
Brief résumé of Chapter 17:
The process of prediction and evaluation of road accidents
and its role in Motorway safety

R.H. Bird CB
formerly of the Ministry of Transport
with additional later economic material supplied by the Highways Agency

As Chapter 1 of this Volume has shown, the advocates of creation of motorways in Britain in the first half of the twentieth century adduced as a feature of their case the promise of greater safety for most road users from among whom the faster traffic would be removed. Those for whom motorways would provide envisaged relatively unobstructed passage. The form of exclusive motorway which they outlined had obvious intrinsic safety features, in particular their reservation to fast motor traffic and grade separation where vehicle paths would cross. In the 1930s their advocates in Nazi Germany sought to obfuscate their military significance by currying international regard for the design of their autobahns while in Britain concern to curb growing road casualties led the Royal Commission under Alness towards similar conclusions. As we have seen, disciplined analysis of the potential safety gains of such proposals successively by the RRL, TRRL and TRL was very significant.

The aim of research in these areas was, and still is, to find ways in which the incidence of accidents can be reduced and the physical consequences of accidents alleviated for drivers and passengers.

There are three main components to the study of road safety. The first is the contribution the road makes through its geometry, layout, surface characteristics, signing, lighting and road furniture. The second is the contribution of the vehicle in the extent of its all-round visibility, braking and crash worthiness. The third is personality and its effect on accident propensity and the extent to which behaviour can be modified by training and by abstinence from drugs or alcohol while driving.

Historically the first two were tackled initially because they offered the more immediate gains from research and implementation of the findings. Indeed, the effort in all highway design is to reduce the conflict between traffic movements. On this account the difference in geometry between motorways and urban roads is plain to see This major contribution to motorway safety is treated in extenso in other chapters. The change in vehicle design over the years is manifest in the improved all round sightlines provided by modern vehicles, as is their crash worthiness. Change in these respects is more difficult to implement because, while improvement in highway characteristics is entirely in the hands of H. M. Government or local authorities, improved vehicle design, from a safety aspect, is an issue for negotiation with private companies and only partly for legislation. Negotiation leans on the threat of legislation, and the prior influence implicit in placing the results of crash worthiness tests in the public domain. Even so, there is a plane of prior decision where again numerate analysis has to be pervasive.

This is the plane on which it is decided whether it is in the public interest, given the available technologies of transport and the very heavy preponderance of the roads in facilitating the population's dependence upon movement, that the motorway concept should be applied in particular situations. These provisions include the econometric analysis of recorded movement of traffic in current conditions of the transport infrastructure in a particular geographical area and, for comparison, estimation of what it would become in that area with a specific element of motorway incorporated. While one of the main differences illuminated by this comparison is expressible in the times taken to make journeys, the other is in the relative safety of travel per vehicle mile on motorway compared with travel on other forms of road.

That the differences between contemporary accident rates upon different categories of road are indeed appreciable. However, without much of that evidence being record the question was duly raised in a Departmental minute written in August 1965 by Richard Bird which forms this chapter. He posed the question whether numerate expression should be given to these differences in a calculus of the consequences of superimposing a length of motorway upon the network of roads and referred back to academic work as ground for regarding this as feasible

Its roots can be found in Chapter 10 which has referred to calculations made of the value of a life lost in 1970 at £17,000. The sequence of subsequent research and recalculations of this figure is traced here as it rose to a contemporary figure of just over £1 million, a figure profoundly relevant to consideration of the merits of providing or not providing elements of motorway.

At the same time the process continued of establishing or refining by research the specification of features of roads, and motorways in particular, which would be conducive to greater safety. But, as the law of diminishing returns started to affect research work, attention turned to work concerned with the effects of personality on accidents and what this meant for improving training regimes and in the conclusions which fed back into design by way of changed perceptions. The change was reported in Ron Bridle's keynote address to the Australian Road Research Board at their twelfth biennial conference at Hobart, Tasmania in August 1984 after the major part of the British motorway building programme was finished. Subsequently emphasis shifted to work on the effects of drugs and alcohol which modify personality. This work was planned and led by Geoff Maycock and simulators were built to help with the work.

Generally, apart from the truly physical changes such as reduced skidding distances and improved crash worthiness, much turns on driver perception and behaviour. Although physical changes in layout are regarded as variables impinging on safety, they take their effect in driver behaviour. The notion is to present a consistent level of required decision making as compared with the greater randomness experienced in unplanned situations. A motorway environment presents a uniformity of decision making which cannot be achieved in the randomness of historic urban areas. For example, reading signs far enough away before a decision needs to be taken is an example of the principle described. But drivers on motorways at night still experience sudden changes; for example, from illuminated areas to darkness. Scope for further study therefore continues both within the TRL's study of the impact of layout and road characteristics – work which has made its greatest impact in the development of motorway.

The description of those wider economic considerations is the prime aim of this chapter to reveal together with the recognition of the changes adopted in respect of life and injury. These changes have had an important impact both on design and on cost benefit comparisons as between alternative road investments and their wider implications for reducing accidents. The research undertaken within the Hard Shoulder and Roadside Safety Group featured later in this chapter shows the value of work done co-operatively as between several interested bodies to curtail such accidents.

The value of the benefits of preventing road accidents and casualties is the subject of many Departmental publications listed in Highways Economic Note No 1 (2001). Here it is only necessary to distinguish between a fatal casualty which occurs when an individual is killed (e.g. a pedestrian knocked down with no collateral damage) and a fatal accident in which there is subsequent car damage and possibly one or more lives lost. Within the valuation of a fatality in a road traffic accident three components are distinguished:

*Human costs, which represent pain, grief and suffering to the relatives and friends (and deaths resulting from accidents up to thirty days afterwards , if relevant to the casualty) plus a representation of the loss of enjoyment of life over and above the consumption of goods and services;

*Loss of output - the value of the expected loss of earnings (plus any non-wage earnings, such as national insurance paid by the casualty's employer);

*Ambulance and hospital costs.

Human costs have been found to represent nearly 60% of the total valuation, loss of output amounts to about 34%, but ambulance and hospital costs are less than 9.1%. The value of preventing a fatal accident is thus likely to be greater than that of preventing a fatal casualty because a fatal accident may involve more than the death of one casualty, such as another casualty who dies or is injured, together with other costs such as those of the police, damage to property or administration of insurance.

The methodology applies, mutatis mutandis, to prevention of non-fatal casualties and to the prevention of accidents involving non-fatal casualties or only material damage. The following table derived from Highways Economic Note No 1 of January 2001 shows the values assigned to preventing several types of accident at June 1999 prices:


Average value of prevention of -












Damage only




Since the first version of Highway Economic Note No 1 was issued in the mid 1970s there have been twenty three versions issued, each necessitated by inflation. To the extent that real income may grow faster than retail prices, the values in the series have to be increased on that account also. The sequence has included a number of step changes. Without unchanged methodology, the figures in use respectively from the autumn of 1978 to the autumn of 1979 and from early 1981 rose from £56,700 by 110 % to £119,300, this being calculated with three years' difference (1977 to 1980) in the price levels. Similarly, in the autumn of 1985 the value stood at £161,170 at 1984 prices but by the autumn of 1987 it had risen by 58 % to £254,880 at 1986 prices.

In 1988, however, a different methodology was introduced to deduce the values of preventing the various contingencies by estimating the public's willingness to pay so as to reduce the risks of accidents arising. Society's willingness to pay to save a life or avoid accident costs can be inferred from the cost of measures taken to avoid accidents and the upshot in accident reduction. This approach used such empirical evidence and a consensus view of experts on the lives saved thereby. Such an approach produced for the prevention of a fatal casualty on the roads the broad figure of £500,000 at 1987 prices. By 1996, at current prices, the corresponding figure was estimated to stand at £848,000, an increase of 70 % over 9 years.

A further Departmental research review in 1997 made a fresh identification of a reasonable range as £1/4 million on either side of £1 million. With that formulation the figures quoted in the table above are notably consistent. Its reasonableness is reinforced by the fact that among western nations the cost of a fatality prevented on the roads gathered around a figure: about £17,000 early in the 1970s. It had reached about £1 million in 2001.

The consistency which this formula implies is, however, not matched by instances of the measures taken. Concern to avoid mass carnage implies that a greater value still is put on the measures, and so on lives lost in that context, than on correcting circumstances that might lead to many incidents involving a single loss of life over a longer time span. Serious accidents on motorways have tended to draw attention to causes like adverse weather and the value of crash barriers.

However, incidents involving large passenger-carrying aeroplanes or a railway's passengers tend to draw attention to the circumstances where a more systematic application of preventive action is feasible. So they are used to justify very expensive measures to avoid public transport accidents like railway re-signalling. In cases where accidents involve major structures too, there are four identifiable components in the cost of measures taken:

1. Loss of utility;
2. Cost of repairs;
3. Figures put on the loss of life, injury etc.;
4. The price of overcoming public over-reaction.

Bird’s study sets out the basis for the application of economics to this type of systematic thinking, being a transcript (less only the numbering of its paragraphs) taken from the PRO file MT 120/186, item 50A entitled "The Prediction and Evaluation of Road Accidents" as written by Richard Bird in August 1965.