The Scottish Government

« Previous | Contents | Next »

Listen

MOTORCYCLE ACCIDENTS AND CASUALTIES IN SCOTLAND 1992 - 2002

CHAPTER FOUR: IN-DEPTH MOTORCYCLE ACCIDENT DATA ANALYSIS 1992-2002

4.1 The STATS19 data for accidents involving motorcycles was supplied by the Scottish Executive, Analytical Services Division (ASD) Transport Statistics. The eleven years of data from 1992 to 2002 were supplied and stored on a TRL bespoke Access database system. The advantage of this database system was that it had a built-in interface for analysis via SPSS (if required) and also to the TRL MAAP accident analysis system. In practice most of the analyses were conducted using MAAP with the tables and figures being stored in Excel spreadsheets.

4.2 The analysis has mainly focused on casualties. Annex 2 contains supporting tables of casualty data analysis, whereas Annex 1 contains tables of analyses of accident trends and numbers.

Supplied Data

4.3 TRL was supplied STATS19 data for all Personal Injury Accidents (PIAs) which involved a motorcycle in Scotland for the years 1992 to 2002 inclusive. This data was sent in a fixed column format with separate files for the casualty and vehicle information. Full information for the fields was provided by the SE for the data within these files.

4.4 A total of 11,339 accidents with 20,092 associated vehicle records and 13,140 casualty records were supplied. The database included accidents in which a motorcycle was involved but in which the rider or pillion were not themselves injured. The current project is very much focused on motorcycle rider and pillion casualties; hence the records where only non-motorcyclists were injured have been excluded from the following analyses.

4.5 Using the appropriate query in MAAP to isolate TWMV casualties only, the casualty numbers in the database supplied to TRL were virtually the same as those published by the Scottish Executive on motorcyclist casualties (Road Accidents Scotland, 2002). The data supplied to TRL are shown in Table 4.1.

Table 4.1 Casualty numbers in data received by TRL and used for the in-depth analysis

Casualties by year

4.6 The number and severity of casualties by year are shown in Figure 4.1. The figure clearly shows a steady decrease in casualties from 1992 to 1996 and an increase in casualties from 1996 to 2002. The rate of increase from 1996 to 2002 is approaching 9% per year for killed and serious (KSI) casualties and just over 6% per year for all casualties. This year on year increase in motorcycle user casualties is clearly concerning.

4.7 The average number of casualties between 1994 and 1998 are shown in dotted lines in Figure 4.1, this is a reference point for casualty figures on which to judge any increase or decrease in casualties after 1998 for the road accident casualty reduction targets set for 2010 and as discussed in the document "Tomorrow's roads - safer for everyone".

Figure 4.1. Casualties by year

4.8 The increase in motorcycle casualties can be referenced to the increase in the number of licensed motorcycles on the roads, Figure 4.2 shows the casualty rate per thousand licensed motorcycles for fatal and seriously injured casualties and for all casualties. It shows a steady decline in the casualty rate (apart from a minor increase in 1995), with 2002 having the lowest casualty rate in the previous 11 years.

4.9 The fact that the accident rate has decreased over the past 11 years is encouraging. This is not the same as saying that motorcyclists' accident risk is acceptable, or that no action is needed to improve motorcycle safety. First, any situation in which the absolute numbers of fatal and serious casualties are increasing merits attention from the road safety perspective. Secondly, motorcycle user accident risk is far higher than that of the car user. This is particularly true for accidents resulting in serious injury or death, where the casualty rate per 100 million vehicle kilometres is nearly 30 times higher for two wheeled vehicle users than for car users (Road Casualties GB, 2002).

Figure 4.2. Casualty rate per 1000 licensed motorcycles

4.10 In order to investigate where the increase in casualties is happening, the casualty (and accident) data are analysed in more detail in the following sections.

Casualties by built-up / non built-up ROAD

4.11 There are two primary uses of motorcycles, firstly they may be used for commuting purposes where the motorcyclist's journey will probably involve urban riding conditions, and secondly they are used for recreational purposes which is more likely to be taken in non built-up riding conditions. The casualty data have been examined by built-up and non built-up roads separately. The road where an accident occurred is defined as being built-up if the speed limit is 40mph or less in the accident's STATS19 details, and as non built-up if the speed limit is greater than 40mph.

Figure 4.3. Fatal and serious casualties by built-up /non built-up road

4.12 Figure 4.3 shows the fatal and serious casualty numbers by year and by built-up or non built-up road. It shows for both types that there has been an increase in the number of casualties since 1996. However the number of fatal or serious casualties on non built-up roads is, in 2002, much larger than for built-up roads. This is probably because non built-up accidents are more likely to be on faster roads (i.e. speed limit greater than 40mph), and the outcome of any accident is more likely to be a fatal or serious casualty.

Figure 4.4. All casualties by built-up /non built-up road

4.13 Figure 4.4 shows all casualties by built-up or non built-up road. It again shows an increase from 1996 to 2002 (with a slight reduction in 2002 for non built-up casualties). The numbers of accidents by road type are fairly similar in later years, and since we saw in Figure 4.3 that there are more fatal and serious accidents on non built-up roads there must be more slight accidents on built-up roads. This is consistent with the view that accidents on non built-up roads are likely to produce more fatal and serious casualties than those accidents occurring on built-up roads.

Casualties referenced to 1994-1998 average

4.14 Reference was made in an earlier section to the bench mark levels for the road accident casualty targets as discussed in "Tomorrow's roads - safer for everyone". Figure 4.5 shows the percentage of yearly casualty figures compared to the 1994-1998 baseline averages. These have been computed for built-up, non built-up roads and all fatal and serious accidents. Figure 4.5 shows the percentage increase by year, and shows that there has been a higher growth in non built-up road KSI casualties than in built-up road casualties. In 2002 the percentage growth over the 1994 to 1998 baseline for both built-up and for non built-up road fatal and serious casualties is 124% and 132% respectively.

Figure 4.5. Fatal and serious casualties by built-up /non built-up road referenced to the casualty target average of 94-98

4.15 Figure 4.6 shows the percentage increase by year for slight casualties, and shows that there has been similar growth in built-up road casualties as in non built-up road casualties. In 2002 the percentage growth over the 1994 to 1998 baseline for both built-up and for non built-up road slight casualties is 124% and 119% respectively.

Figure 4.6. Slight casualties by built-up /non built-up roads referenced to the casualty target average of 94-98

4.16 Generally there has been a higher percentage increase for fatal and serious casualties on non built-up roads when compared to slight injury or built-up road accident casualties. As stated above, this is probably due to higher speeds being ridden on non built-up roads resulting in more severe injuries (i.e. the severity of injury is associated with the speed of the road being ridden). It suggests that successful road safety campaigns focused on riding on non built-up roads would have a larger impact on the numbers of killed and serious injuries than campaigns focusing on built-up road riding.

Casualties by road class

4.17 The total number of casualties has been analysed by road class for built-up and non built-up roads separately. For built-up roads (i.e. a speed limit £ 40mph), there will be no motorways, hence the A-road plus motorway category shown in Figure 4.7 represents just A-roads. Figure 4.7 shows that, for built-up roads, most accidents occur on A-roads or unclassified roads. Further, that from 1998, it is the accident casualties on unclassified roads that have been increasing at the highest rate.

4.18 Figure 4.8 shows the number of casualties by road class for non built-up roads. It is apparent that most casualties occur on M+A-roads and that the numbers have been increasing since 1996. Given that motorways are generally 'safer' and that motorcyclists do not enjoy riding on motorways since they are not interesting, i.e. do not have bends, then it is not surprising that 96% of M+A-road accidents are on A-roads. The B-roads are the next highest class of road for motorcycle user casualties and these have also been showing a slight but steady increase from 1996.

4.19 The severity of the casualty is also of interest and has been analysed by major and minor road classes, hence the B-, C- and unclassified roads are combined in Figures 4.9 and 4.10.

Figure 4.7. All casualties on built-up roads by road class

Figure 4.8. All casualties on non built-up roads by road class

4.20 Combining B-, C- and unclassified roads defines the minor - and M+A-roads define major roads. The above analysis is repeated using these major and minor road definitions. Figure 4.9 shows the fatal and serious casualties by major and minor road class and if built-up or non built-up. It clearly shows that most fatal and serious casualties occur on non built-up major roads, and probably for the reason tendered earlier in that these are faster roads and hence any accident is more likely to result in fatal or severe injury. There has been a steady growth in the number of these accidents since 1996, albeit with a slight reduction in 2002. There has also been a slight growth in fatal and serious casualties occurring on minor roads and major roads in built-up areas.

Figure 4.9. Fatal and serious casualties on built-up / non built-up roads by road class

Figure 4.10. All casualties on built-up / non built-up roads by road class

4.21 Figure 4.10, for all casualties, shows that most casualties occur on minor built-up roads as well as major non built-up roads (as was observed above). It also shows that there has been a general increase in all casualties on major and minor roads whether or not they are on built-up or non built-up roads since 1996.

4.22 The increase in motorcycle traffic by major or minor road type for 1992 to 2002 has been illustrated in Figure 3.2. Using this data a casualty rate per million kilometres was computed and is presented in Figure 4.11, it only shows rates from 1993 to 2002 because of the availability of comparable data for all years (as previously discussed). The plots by year suggest a gradual decrease in the casualty rate from 1992 to about 1996. The rate then fluctuates, with no overall pattern of increase or decrease for the next 6 years. This suggests that the casualty rate per million kilometres has not changed by much in the past 5-6 years.

Figure 4.11. Casualty rate per million km on major and minor roads

4.23 Figure 4.11 clearly illustrates that there is a much higher casualty rate for all casualties per million kilometres on minor roads than for major roads. It was seen in Figure 4.10 that there is a high number of casualties on minor roads in built-up areas, these are more likely to be slight accident casualties but these motorcyclists do not cover the greater distances of those riders on major roads. Hence it is built-up minor roads that carry the highest accident casualty risk per million kilometres, albeit these accidents are more likely to result in slight injuries.

Size of motorcycle

4.24 The size of the motorcycle is recorded on STATS19 as either less than 125cc or more than 125cc. The total number of accidents by size of bike for built-up and for non built-up areas is shown in Figure 4.12. There are far more large bikes licensed than small capacity bikes. The figure shows that there has been a decrease in the number of accidents on larger bikes from 1992 to 1999 on built-up roads, and then the number stays about the same. There has been a slight increase since 1996 in the number of accidents on large bikes on non built-up roads. The number of accidents on small bikes has increased since 1998 on both built-up and non built-up roads, although the increase is much bigger for built-up roads. This presumably reflects the increase in the number of scooter type bikes which are typically used for commuting purposes.

Figure 4.12. All accidents by size of motorcycle and built-up or non built-up roads

Motorcycle Manoeuvre prior to the accident

4.25 Analysis of the motorcycle manoeuvre prior to the accident shows that there are three main manoeuvres (excluding 'ahead only' which account for 58% of accidents). These are overtaking including overtaking moving and stationary vehicles (average of 9.2%), going ahead on a right-hand bend (9.0%) and going ahead on a left-hand bend (11.4%). These three manoeuvres thus account for just under 30% (on average) of all the manoeuvres recorded prior to an accident.

4.26 Figure 4.13 shows the percentage of these three manoeuvres from 1993 to 2002. Overtaking and going ahead on a right-hand bend follow very similar patterns with no particular trend. In 2001 and 2002 there appears to be a reduction in the percentages of these types of manoeuvre. However, going ahead on left-hand bends has a general upward trend from 1994 to 2002, this suggests that a higher percentage of motorcycle accidents are happening on left-hand bends than were seven or eight years ago.

4.27 The left-hand bend accident, when just going ahead, i.e. not turning or overtaking, is either caused by the rider going too wide and either hitting an oncoming vehicle or even not completing the corner and ending-up in the hedge or tree on the other side of the road, or by other vehicles cutting the corner and then hitting the oncoming motorcycle. It is not possible to tell which of these scenarios caused these left-bend accidents, but it does suggest a situation that could be addressed by a road safety strategy.

Figure 4.13. Manoeuvres made prior to motorcycle accident by year

4.28 A similar analysis as above, but for built-up and non built-up roads, found that the problem of increasing numbers of casualties on left-hand bends was mainly on non built-up A-roads. This is shown in Table 4.2 where 25% of casualties on non built-up roads happened on left hand bends in 2002, this is an increase of 6% on the 1992 figure. There was also a 3% increase between 1992 and 2002 for right-hand bend manoeuvres on non built-up roads. There is very little difference in percentages of each manoeuvre for built-up roads between 1992 and 2002, the most notable being that in 1992 11% of accidents occurred when overtaking compared to 7% in 2002.

Table 4.2. Main manoeuvres made prior to motorcycle accident on built-up and non built-up roads

* Only those data where a manoeuvre is coded are included

4.29 Figures 4.14a and 4.14b show the percentage distribution of manoeuvres in 2002. They clearly show which manoeuvres are the most risky in terms of having an accident.

Figure 4.14. Motorcycle manoeuvre prior having an accident in 2002

a. Built-up roads

b. Non built-up roads

Casualties by month

4.30 The month when the accident casualty occurred has been investigated. It is unlikely that the pattern of motorcycle accident by month changes much, and so the average over the period 1992 to 2002 was computed and is given in Figure 4.15. The figure shows the average number of casualties for each month by built-up and non built-up roads.

4.31 The figure shows less of a peak in the summer months for casualties on built-up roads than for casualties on non built-up roads. This is not surprising since many motorcyclists will only ride in the summer and they are more likely to be riding in non built-up areas. In contrast, those motorcyclists riding in built-up areas are more likely to use their bikes all year round, perhaps for commuting.

4.32 A good time to introduce a motorcycle road safety campaign would probably be in Spring, when many motorcyclists are returning to riding. These riders, in particular, may be 'rusty' from not having ridden during the Winter months and hence are a potential 'target group'.

Figure 4.15. Average (over 1992-2002) casualties by month

Casualties by time of day

4.33 The time of the day when the accident casualty occurred has been investigated. It is unlikely that the pattern of motorcycle accident by time of day changes much, and so the average over the period 1992 to 2002 was computed and is given in Figure 4.16. The figure shows the average number of casualties for each two hour time period by built-up and non built-up roads.

Figure 4.16. Average (over 1992-2002) casualties by time of day

4.34 The figures show for built-up roads a small peak in the morning and a much larger one in early evening, these peaks correspond to commuting periods. The average number of casualties on non built-up roads increases steadily through the day, from 7:00 in the morning, and peaks mid-evening. The evening peak may be as a result of tiredness, if riding all day or perhaps deteriorating riding conditions (sun angle, traffic volume etc.) during the summer, or as a result of riding on fast roads in the dark during the winter.

Casualties by day of week

4.35 The day of the week when the accident casualty occurred has been investigated. It is unlikely that the pattern of motorcycle accident by day of the week changes much, and so the average over the period 1992 to 2002 was computed and is given in Figure 4.17. The figure shows the average number of casualties for each day of the week for built-up and non built-up roads.

Figure 4.17. Average (over 1992-2002) casualties by day of week

4.36 The average number of casualties during week days is considerably lower for non built-up roads than for built-up roads. This pattern changes with the week-end, where the average number of built-up casualties decreases on Saturday and more so on Sunday but the number of non built-up casualties increases to a bit above the built-up number on Saturday and to considerably more above the built-up number on Sunday. This is a reflection of the fact that many motorcyclists only use their motorcycle over the weekend and when they do, they are far more likely to be riding on non built-up roads.

Accidents involving motorcycles by day of week and time of day

4.37 The pattern of motorcycle accidents for the years 1992 to 2002 by the day of the week and the time of the day (in 2 hours intervals), is shown in Figure 4.18. The pattern indicates that there is a small morning peak with a much larger peak in early evenings for weekdays. However, the weekend pattern only has the late afternoon peak, which is highest on Sunday.

Figure 4.18. Accidents from 1992 to 2002 by day of week and time of day

Age of motorcycle rider (driver) and year of accident

4.38 The number of accidents per year as distributed between riders of different ages has been investigated. The analysis suggests that the pattern of age distribution has been changing over the past 11 years. Figure 4.19 shows the numbers of accidents per year for each age group. It clearly shows that the number of accidents involving younger riders has fallen over the past 11 years, whereas the number in age groups 31-35, 36-40, 41-45 has shown an increase over the past four or five years.

4.39 It is possible that the increase in accidents for age groups from 31 to 45 may be due to riders returning to motorcycling - the 'born-again biker'. Those riders who have had very long breaks, and/or for riders with very little pre-break experience and training, the returning rider is in many respects a novice. He or she is likely to have an elevated accident liability as a result - at least in comparison with other riders of the same age. Returnees increase total motorcyclist numbers, and total mileage, and hence total motorcycle casualties. Returnees might well form a good target group for countermeasures such as training, since (a) they tend to be easily identifiable - for example, when they purchase a new motorcycle, (b) it is possible to deliver training literature (and even offers of free training courses) along with the new machine and (c) returnees may often be well motivated to improve their safe riding skills.

Figure 4.19. Number of accidents (1992 to 2002) by year and age of rider

« Previous | Contents | Next »