LGRAS (00) 5 - ROAD SAFETY TARGETS FOR 2010
1. At the beginning of March, the UK Government, the Scottish Executive, and the National Assembly for Wales announced the road safety targets for 2010. These are to achieve, compared with the average for 1994-98:
a. a 40% reduction in the number of people killed or seriously injured in road accidents;
b. a 50% reduction in the number of children killed or seriously injured; and
c. a 10% reduction in the slight casualty rate, expressed as the number of people slightly injured per 100 million vehicle kilometres.
2. No announcement has been made by central government about "local" targets (e.g. for individual Police Forces or Councils), or about how figures should be monitored locally. However, it seems to be generally expected that central government will make available relevant statistics for individual Police Forces and Councils.
The first two targets
3. Numbers for each local authority corresponding to the first two targets can be produced easily from "STATS 19" data. The fact that local government reorganisation took place in the middle of the 1994-98 baseline period is not a problem: "STATS 19" figures for the present Scottish local authorities are available back to 1981, because each accident record has been allocated retrospectively to the appropriate current local authority area. In addition, account can be taken of the effect of the road network changes caused by the "detrunking" of some trunk roads, and the "trunking" of some other roads (which coincided with local government reorganisation), because the Scottish "STATS 19" data back to the start of 1992 were retrospectively coded in terms of the new trunk road network. Annex E of "Road Accidents Scotland 1998" discusses these matters.
4. So, for these two targets, the question is: what kinds of figures should be included in "Road Accidents Scotland 1999", and in subsequent editions? Members of the Group will know that "Road Accidents Scotland 1998" included several pages of figures for individual Police Forces and Councils, showing how their figures for recent years compare with the averages for 1981-85 (the baseline for the targets for 2000): see Tables 3, 40, 41 and 42. As well as the total numbers for each severity, some of these tables provide the figures broken down by type of road, or for pedestrians only. Should "Road Accidents Scotland 1999" include similar tables showing how the numbers for each area are starting to move relative to the 1994-98 averages?
5. The figures could be broken down in a number of ways - e.g.:
- by type of road - because local authorities are not reponsible for trunk roads; and because of the differences between major roads and minor roads, and between built up roads and non-built up roads.
- by type of road user - because there may be particular interest in groups such as pedestrians, pedal cyclists and motorcyclists.
- for particular population sub-groups - because there may be interest in figures for (say) pensioners, in addition to the interest in the figures for children (for the second target).
6. The attached skeleton tables indicate the kinds of breakdown that might be provided. Members of the Group are invited to comment on these - for example:
- should such tables be included in "Road Accidents Scotland 1999"?
- should fewer details be provided? if so, which should they be?
- should more details be provided? if so, what other information should be added?
The "slight casualty rate" target
7. The "slight casualty rate" target presents greater difficulty, because to calculate slight casualty rates one needs both the numbers of "slight injury" casualties and estimates of the total volume of traffic on the roads. The numbers of "slight injury" casualties are available from "STATS 19", but, so far, estimates of the total volume of traffic on the roads have not been produced for individual local authorities. Therefore, this paper describes, first, the traffic estimates that have been produced to date, and then the new traffic estimates which the Department of the Environment, Transport and the Regions (DETR) is developing, before discussing the possible publication of slight casualty rates for individual local authorities.
The traffic estimates that DETR has produced to date
8. DETR has, for a number of years, estimated the volume of traffic on major roads (i.e. Motorways and A roads) for individual local authority areas. Such estimates have been published in "Scottish Transport Statistics". However, to date, it has not been possible to publish even an estimate of the total volume of traffic on minor roads (i.e. B, C and unclassified roads) for Scotland as a whole in "Scottish Transport Statistics", because DETR's method of estimation was designed only to produce estimates of minor road traffic volumes for Great Britain as a whole. DETR's estimates of minor road traffic in Scotland are based on the total length of each type of minor road, as reported by local authorities, and traffic counts taken at a sample of 200 or so minor road sites per year, which is far too small for producing estimates for individual local authorities (the number of traffic counts each year in Scotland depends upon the amount that the Scottish Executive can spend on such work).
9. To date, DETR's estimates of the volume of traffic on major roads for the present local authorities have only been available with effect from 1996: its estimates for earlier years relate to the former Regions and Island Areas. Therefore, one cannot use the currentlyavailable estimates to calculate 1994-98 average slight casualty rates for major roads for the area covered by each of the current local authorities. The most that one can produce using the currently available major road traffic estimates for each local authority would be 19 96-98 average slight casualty rates (overall averages, and separate averages for each type of major road - i.e. Motorways, built up Trunk A roads, non-built up Trunk A roads, built up non- Trunk A roads, non-built up non-Trunk A roads).
New traffic estimates for local authorities, which DETR is developing
10. Because of the potential interest in estimates of the slight casualty rate for individual local authorities, DETR has developed a method of estimating the volume of traffic on minor roads for each local authority in 1998, which has been tested on data for English local authorities. The method uses information about (a) the length of each type of minor road in that authority, (b) the average traffic flows on non-trunk A roads (which are not "minor" roads) in that authority, and (c) the estimated total volume of traffic on each type of minor road in the country (or region of England) in which the local authority is located. The method's key assumption is that the average traffic flows on each type of minor road in an authority will be proportional to the average traffic flow on the non-trunk A roads in the same authority (in effect, it assumes that an authority's minor roads will be x% busier than the average for minor roads if its main roads are x% busier than the average for main roads). With this assumption, DETR can produce a proxy figure for the volume of traffic on each type of minor road for each local authority in Scotland, and scale these proxy figures so that they add up to the previously-estimated total volume of traffic on each type of minor road for Scotland as a whole.
11. Having produced estimates for 1998, DETR could estimate the volume of minor road traffic in each local authority for other years, by assuming that the rate of traffic growth for each type of minor road is the same for each local authority as it is for that type of road for the country as a whole.
12. DETR could then estimate the volume of traffic on major roads in 1994 and 1995 for the present local authorities, by using a similar process along with the existing estimates for the former Regions and Island Areas for 1994 and 1995 (and knowledge of how the current local authorities relate to these). Finally, DETR could combine the estimates of traffic on minor roads and on major roads for each local authority for 1994 onwards to estimate the total volume of traffic for each local authority for each year. One could then calculate the 1994-98 average slight casualty rate for each local authority, and the corresponding figures for subsequent years.
13. DETR recognises that the above estimation process may not give an accurate estimate of traffic levels for each local authority. However, it believes that it should permit a reasonable estimate of the change in the level of traffic in each local authority. This would, in turn, enable estimation of the change in the slight casualty rate. DETR is continuing to develop the method, and is consulting the Transport Statistics Users Group about it. However, it seems likely that a method like that described above is the best that can be developed using data that are available across Great Britain: central government does not have the detailed information about traffic at a local level that would be required to enable it to produce better estimates.
The possible production of slight casualty rates for individual local authorities
14. So, estimates of slight casualty rates could be published for each local authority, accompanied by notes on the assumptions that had to be made to produce the underlying estimates of traffic volumes. If this were done, the estimates might be used by the media to produce casualty rate "league tables". For example, suppose that the published estimates for two local authorities were as follows:
Slight casualty rate
Attention might focus on Y's rate appearing to be roughly double X's rate, rather than the improvement in Y's rate.
15. One could prevent direct comparisons of the estimated rates for different local authorities, by publishing only figures in index form, relative to the 1994-98 baseline average, for example:
Slight casualty rate: Index (1994-98 = 100)
This approach concentrates upon the apparent change from the baseline level. However, as noted above, the traffic figures for other years were estimated by assuming that the rate of traffic growth for each type of minor road is the same for each local authority as for the country as a whole, which may not be the case. For example, X's apparently poorer performance in 1999 could be due to a large increase in traffic in its area, which is not identifiable using the estimation method, causing an increase in the number of casualties recorded in "STATS 19". And, in some cases, because of year to year fluctuations in the numbers of accidents and casualties, this approach might show a sudden rise in the slight casualty rate for a local authority which had a low average for the period 1994-98: in such a case, this approach might give a misleading impression of the true position, because it would not show how low an authority's slight casualty rate is, and how a small change in the rate represents a large percentage increase.
16. Separate slight casualty rates could be published in "Road Accidents Scotland 1999" for each type of major road for individual local authorities (in the same way as accident rates are given for different types of major road for each Police Force in Table 5 of "Road Accidents Scotland 1998"). However, it is not proposed to publish separate major road slight casualty rates for different types of road user (eg pedestrians) or for different sub-groups of the population (eg children), because of a lack of reliable estimates of the "volume" of traffic on major roads for such groups.
17. In view of the assumptions that are required to produce minor road traffic estimates for individual local authorities, it is not proposed to disaggregate any slight casualty rates published for minor roads: it is envisaged that, at most, only the overall minor road slight casualty rate would be published for each local authority.
18. Members of the Group are asked for their views on these matters. For example:
- would they be content for slight casualty rates for individual local authorities to be estimated, using traffic volumes estimated by the method described above?
- what are their views on the publication of such estimates in "Road Accidents Scotland 1999"? For example, should they appear only in index form? Should separate slight casualty rates for different types of major road be published for each local authority? What are their views on other possible disaggregations of the slight casualty rates?