.T25 Road Accident Prevention: the work of the Overseas Unit, TRRL G.D. Jacobs.BSc PhD MCIT MIHT Transport and Road Research Laboratory After taking a science degree at Brunel University. Dr. Jacobs joined the TIMR in 1961 and spent seven years working on a range of traffic and safety problems including the design of pedestrian facilities and work on drink/drive legislation. He joined the Overseas Unit of TMR in 1968 and is now the Head of the Unit's Urban Traffic and Road Safety Group. In 1976 he was awarded a PhD fram Surrey University for his work on road accidents in, developing countries. For the last ten years he has acted as the transport adviser to the British Overseas Development Administration. He has worked in over 25 countries in Asia, Africa andi the Middle East. Surprising though it may seem, road safety is a very old prcblem. In Great Britain for example, there were over' 1,000 deaths on the roads each year even before the advent of the motor car. In 1980, in those European countries supplying road accident statistics to the United Nations, about 100,000 persons were killed and over two million injured by motor vehicles. Over the last thirty years or so in Western Europe and North America much has been done to deal with this growing problem with substantial sums of money spent on a wide range of safety measures. -In a number of these countries, Great Britain, for example, the successful application of appropriate countermeasures can be judged by the fact that not only has there been a decrease on the accident rate (in terms of vehicle kilametres travelled) but the actual number of people killed and injured has also decreased. 2. By the early 1970's countries of the non-industrialised world were becoming increasingly aware that they too faced a growing road safety problem. This situation was also recognised by international organisations such as the World Health Organisation and the World Bank. 3. In 1972, following numerous requests made by African and Asian countries for guidance in the road safety field, a small research team was formed within the Overseas Unit of the TRRIM. The aim. of this team was to undertake research in the countries of Africa, Asia and Latin America with a view to establishing the nature and extent of their traffic accident problaem and, in the longer term to assess the effectiveness of renedial measures. The author uses scire of the major findings of this research team to provide an overview of the safety problem in these countries and to put forward suggestions for appropriate low cost countermeasures. 4. A preliminary study (1) of accident costs carried out by the Overseas Unit indicated that in those non-Western countries for which data were available, the total cost of road accidents was al.mest 1% of their gross national products. In countries such as Indonesia and Nigeria this means that road accidents may well be costing over £500 million per annum. In a range of smaller countries such as Colombia, Pakistan and Egypt equivalent costs may exceed £200 million, Clearly these are sums of money that these countries can ill-afford to lose. 1% of the total GNIP' s of all countries below £1000 GNP/capita. per annum curbined is approximately £10,000 million - a very crude estimate of the total annual cost of accidents in these countries. 5. A~nother way of illustrating the extent of the road accident problem is to cozaipr road accident fatalities with the number of deaths resulting from diseases and other known causes in non-Western countries. Using statistics published by the World Health Organisation and the United Nations, data were obtained fran 20 African and Asian countries using the most up-to-date classification of causes of death used by these two organisations. It was found that road accidents accounted for alrost 2.5% of all deaths recorded in these countries, making road accidents the tenth most important cause of death. The analysis was repeated for the age groups 5-64 years thus removing the very young and the elderly and in this case, road accidents accounted for over 6% of all deaths, a value exceeded by only five other causes. Mien repeated for the 5-44 years age group road accidents accounted for almost 10% of the total number of deaths reported and ranked second to the "rmultiple" cause of "all other accidents, homicide and suicide".- Detailed studies carried out in selected countries over the last few years where reliable data were more readily available also indicate the growing importance of road accidents as a cause of death. In Jordan for example, in 1962 road accidents was the eleventh most important cause of death but by 1980 it was the fourth most important cause, representing alnnst 6% of all deaths recorded. Although the countries for which data 553 were available are unlikely to be representative of the entire Third World, it is clear that road accidents represent a growing social problem particularly for juveniles, young adults and those in early middle age. 6. Numerous studies (2,3,4) carried out by the Overseas Unit have attempted to identify the magnitude of the road accident prcblem in non-Western countries and to rank countries in order of 'seriousness'. Comparisons have also been made between Western and non-Western countries in order to show that the safety problem is particularly severe in the Third World. In order to caipare the safety problem in different countries it is obviously meaningless to use total nurrber of fatalities or casualties because of the vastly different population sizes and degrees of motorisation in the various countries. Ideally comparisons should be made as though both their human and vehicle populations were the same. In the past, fatality rates (defined as road accident deaths per 10,000 vehicles licenced) have been used in order to com-pare the accident situation in different countries. The nuntber of fatalities as opposed to casualties or injury accidents have been used because the poor accident recording systems in most noni-industrialised countries means that only fatalities are recorded to any reasonable degree of accuracy. In addition, numrbers of vehicles licenced have been used as opposed to millions of vehicle kilometres travelled per annum because very rarely are accurate n-point or trend censes carried out in developing countries to provide such data. 7. The use of fatalities per 10,000 vehicles as a measure of death rate is far from ideal and recent attempts (5) have been made to express road accident fatalities as a function of both vehicles licenced and persons resident in a country. Results indicate that whatever measure is used, road accident fatality rates in African and Asian countries are considerably greater than in European and North American countries. Studies have also shown that whereas in European countries fatality rates have decreased steadily over the last 20 years, those in a considerable number of African and Asian countries have increased. In other words, over the last 10-15 years or so, the road accident problem in these countries has worsened. TH~E NATURE OF THE PRXLEM 8. Broad classifications of accidents and casualty types occuring in a country are inportant indications of the general strategy required in dealing with the country's road problem. The pattern of accidents taking place will clearly vary from country to country and Table 1 shows that for classes of road users killed, these differences can be considerable. Thus in Hong Kong 70% of all persons killed are pedestriansrwhilst in Indonesia the proportion is only 20%. In Indonesia however 34% of those killed are riders of motor cycles and scooters. In the UK, only 5% of persorrs killed were pedal cyclists; in countries such as China where the bicycle is the most carrn form of transport the proportion is likely to be very much greater. Important differences also exist between the age groups of those killed and injured in road accidents. More often than not these reflect the age distribution of the population in the different countries. This in turn means that there are proportionately more children killed and injured in road accidents in non-Western countries than in the case in Europe and North Anerica, suggesting that the education of children on road safety matters is particularly important in these countries. TABLE 1 - Percentage of fatalities by road-user class. 9. Wide differences between countries can also exist in the urban/rural split: for example, whereas 45% of fatalities in Great Britain (6) occurred on roads in non built-up areas in 1980, the equivalent figure for West Malaysia was 61% with a further 26% occurring in villages and only 12% 554 Country Year Pedestrians Cyclists Motor 'cycists Drivers and Total and scooterists Passengers Ethiopia 1976 84 1 1 13 100 Guyana 1977 45 13 10 28 100 Hong Kong 1976 70 4 7 19 100 Indonesia 1977 20 2 34 44 100 Jamaica 1978 41 5 17 37 100 Jordan 1979 47 1 2 s0 100 Kenya 1972 45 9 2 40 100 Kuwait 1978 55 2 2 41 100 Nigeria - 35 3 20 42 100 Sri Lanka 1980 51 10 10 28 100 Swaziland 1978 55 5 0 40 100 West Malaysia 1979 22 13 33 32 100 Zamrbia 1977 40 8 3 49 100 Zirrbabwe 1979 36 9 2 47 100 WK 1980 32 5 19 44 100 T25 occurring in towns. With this situation, it would appear appropriate to devote nore resources to rural accidents in West Malaysia than would be the case in Great Britain. 10. An analysis of one camplete year's accident records fran Kenya (7) showed that 16% of all road casualties were occupants of commercial vehicles. The equivalent value of mest Western countries is under 5%. In many African and Asian countries, conuercial vehicles are used to transport people to and fram places of work and greater attention than in developed countries to accidents involving cammercial vehicles might be needed. 11. In non-1Western countries a major road safety problem may be present that does not exist at all in Western Europe and North ~~~ica - accidents involving para-transit forms of public transport. Thus in Surabaya, the second city of Indonesia, 17% of all casualties (8) were drivers or passengers of betjaks (cycle rickshaws). According to surveys carried out in Surabaya there were, in 1974 an estimated 70,000 of these vehicles operating in the city. The drivers and passengers of these vehicles are often placed in a vulnerable position, not only because the vehicles provide little protection but also because the drivers frequently ignore all traffic rules and regulations. Other types of public transport commorn in these countries are the shared taxi such as the Dolmnus of Turkey and the Service taxi of the Middle East. In Jordan, Service taxis are involved in alnost a quarter of all accidents yet represent only 10% of the total vehicles registered in that country. This type of prcblem is rarely encountered in the developed world and remedial measures adapted from Western countries may do little to deal with this situation. 12. The above examples show that major differences exist in the accident patterns of countries in different regions of the world. These differences have been used to emphasise the point that the order of priorities in road safety prograrrues in African, Asian and Middle Eastern countries could, mere often than not, be very different from that in Western countries. IHE APPLICATION OF REMEDIAL MEASURES 13. Over the past 40 years, European and North American countries have built up considerable experience in road safety and practice, which includes substantial body of research data. Potentially, this experience should be of value to non-Webstern countries in assessing priorities in their own road safety progranues. However, before attempting to apply research findings fram, say, Great Britain or the US to less developed countries, a nu~re of general reservations should be made : (1) As shown earlier, the nature of the prcblemn in non-Webstern countries may be considerably different from that in Europe or North America. (2) Countermeasures that are effective in developed countries may be ineffective in the develop-- ing countries, (and possibly vice versa). (3) Although there has been extensive research into the effectiveness of counterireasures in developed countries, the results of this research may be less definitive than might be desired. (4) Countermeasures that are appropriate in developed countries may, for financial or other reasons, be in appropriate in less-developed countries. 14. The road safety problem in developing countries is often markedly different from that in Europe or North America. Therefore, although research findings from Western countries can provide sane guidance, the inevitable uncertainties surrounding their transfer to non-Western countries exrphasises the need for caution in their application. As a direct consequence, there is a need to evaluate any countermeasures that are undertaken, thus enptasising the value of nounting local or regional trials of any countenreasures and carefully monitoring their effectiveness before using them nationally. 15. Over the last few years the Overseas Unit has collected annual police reports fran a nurrber of different countries. Most of these reports provide a basic surrmary of the road accident situation in each country, together with a list of the major "causes`' of accidents.. These data have clearly not been collected as accurately as those in a comprehensive 'on-the-spot' study (9) carried out by the TRRL and present a completely different data base. In the records collected, the police have ascribed a single "main cause" to each accident as opposed to listing the "factors involved". Nevertheless, this information does provide some insight into what the police regard as the major factors involved in road accidents in non-Western countries. (Note: If the police have little road engineering expertise it is likely that they will underestimate the part played by the road environment). 16. The results from the police data of five of these countries (see Table 2) show that roaduser error was identified as the main cause in at least 70% of the road accidents. The percentage of accidents attributed to the three main causes varied considerably from one country to another. However it is dangerous to draw conclusions about variations between these countries as there are 5 55 likely to be differences in the types of accident reported to the police and in the way in which the police analyse the accidents for causes.. 17. Also it is likely that the percentages are underestimates of the true contribution of these factors because in many of the accidents there are probably several factors involved and not just one. Thus the percentage of accidents due to adverse road conditions and environnent may in real- ity be much higher because many of the road user errors could have been due to inadequate road signing or markings. Nevertheless the results do indicate the importance of road user error as a contributory factor in road accidents in non-Western countries. Studies of Behaviour 18. Two neasures commornly used for controlling road users and improving their safety are traffic signals and pedestrian crossings. Studies (10) ware made by the Overseas Unit, TRR of the behav- iour of drivers at traffic signals and pedestrian crossings in selected cities in non-Western countries and comparisons made with results from Great Britain. For example, driver behaviour at `Zebra-type` crossings was observed in five Third World cities and ccsipre with behaviour at selected Zebra crossings in Reading and London. It was mandatory for drivers to stop for pedes- trians on the crossing in all cities studied. It was found that the average proportion of drivers stopping in four of the Third World cities ranged from 10 to 17%, whilst in Surabaya the percentage was well under 1%. The equivalent values in Reading and London ware 72 and 400 respectively. 19. Cbservations ware also made at signal-controlled junctions in the sane cities and the proportions of drivers (presented with a free choice) stopping at the red signal ware recorded. Results are given in Table 3. It can be seen that the percentage of drivers choosing not to stop at the red signal in the non-webstern cities was greater than in Reading and London. Studies ware carried out in Nairobi in 1975 and 1977; the results for 1977 showed a marked improvement over the 1975 value. This surprising result may be due to the fact that the nuirber of signals (and having them set correctly) may have brought about the observed iirprovement in driver behaviour. It should be noted, hovnver, that in Bangkok, Ankara and Surabaya many junctions ware signal-controlled, but behaviour was still poor. Thus, in at least the cases of traffic signals and pedestrian crossings, .there is evidence of road safety countermeasures being less effective in African and Asian countries compared with LW (although increased enforcement could perhaps have improved the performnc~ie of these measures). 20. Detailed studies of driver behaviour ware also carried out in Pakistan (11). Observations of driver behaviour ware made at a number of sites often as part of a moenitoring progranmr to detnrmire the effectiveness of remedial measures. From a summary of the results of this work shown in Table 4 it is clear that a very high proportion of drivers ware ccumitting errors at junctions particular- ly when turning right. Also it would appear that stop signs ware frequently disregarded even when traffic on the major road was close to the junction and in addition many drivers ware taking risks at bends by crossing over the centre of the road. TABLE 3 - Non-observance of the red signal in selected cities. Numbier of drivers Nuirber of drivers Percentage of who had a free choosing not to drivers city choice of stopping stop at red choosing not at red signal signal to stop at red signal Ankara (2 sites) 1974 101 36 35.6 Bangkok (9 sites) 1975 754 391 52.0 Naircbi (2 sites) 1975 203 101 50.0 Nairobi (10 sites) 1977 3045 210 7.0 Surabaya (6 sites) 1975 253 92 36.0 Surabaya (6 sites) 1976 396 130 48.8 Central London (11 sites) 1977 364 22 6.0 Reading Area (19 sites) 1977 726 30 4.1 21. Again, results suggest that safety improvenents such as road signs and markings may be less effective in non-Western countries than in European countries. Changes in behaviour can hopefully be brought about (and the effectiveness of safety measures improved) by the introduction of education and training progranrres and also by improved enforcemrent techniques (see below). Traffic Law Enforcerrent 22. With the generally low standard of road user behaviour that exists in many African and Asian countries - which may in turn be due either to lack of awareness of traffic regulations or to a 556 T25 general `attitude" towards road safety - it is inportant that adequate traffic law enforcement is provided by the police. Because little research has been carried out in this field it is difficult to assess the potential of police enforcement for accident reduction in non-Western countries, for in many of them the traffic police are not well trained or equipped as they are in Western countries. Further, in many developing countries the police are obliged to spend much of their tine controlling traffic, with little time available for traffic law enforcement. TABLE 4 - Percentaqe of drivers makinq errors in Pakistan. ()numtber of sites 23. The most promising evidence for the road safety benefits of enforcement in developing countries comes from Singapore and Egypt. In Singapore, a coubined publicity and enforcement carrpaign appears to have lead to a drop in fatalities of 19% and 50% in serious injuries, although there was a rise of 20% in slight injuries. In Egypt a combined package of police enforcement measures including radar, increased patrols and heavier penalties for traffic offences has had a significant effect on accidents on two major inter-urban roads.. On one of these roads there has been an overall reduction in the nurbr of accidents of over 50%. (Carparing a six-month period after the introduction of the improvements with a corrparable six-month period before their introduction). Vehicle Safety 24. Vehicle safety measures can be both "primary" and "secondary" in nature; the former aim at preventing an accident occurring whilst the latter attempt to protect the road-user during the course of an accident. Comnercial and public service vehicles are involved in proportionately mere accidents in Asian countries than is the case in Europe and North America. The way in which these vehicles are used leads to potentially dangerous situations with open lorries often carrying large nuirbers of workers and buses carrying people hanging on the outside of the vehicles. Paratransit forms of public transport, cycle rickshaws, shared taxis etc - also have a reputation of being dangerous vehicles in which to travel. The accident records to these vehicles could be cons iderdbly improved by legislation prohibiting lorries, buses and minibuses fran carrying passengers in a dangerous manner. 25. A study (12) of accident records of a nuibr of transport undertakings in India showed that buses were involved in about five tines mere accidents than might be expected from. their nuries cn the road or the annual vehicle kilametrage travelled by different classes of vehicle. Fatality rates per million bus kilometres travelled were about six times greater than for public transport in London and over ten times greater than for other cities in Great Britain. This study examndxed one comlete year's accident records of the Delhi Transport Corporation (DITC). In addition, about 10% of all bus drivers (580) were interviewed to cbtain information on their background, experience, knowledge of the highway code and working conditions. Surveys were made of the condition of 237 buses operated by DXr from five of the main depots so that a general assessment of the safety of vehicles could be made. These surveys showed that drivers' received inadequate training and that the general condition of buses was poor. 26. Perhaps the two mest important measures that can be adopted to protect the road user during the course of an accident are the use of seat belts for vehicle occupants, and crash helmets for metor cyclists. There is now considerable evidence from a nuffier of Webstern countries, that compulsory wearing of seat belts results in a significant reduction in injuries, particularly thcse of a mere severe nature. The benefits of wearing a seat belt in any particular accident situation should be similar in both developed and developing countries.. In view of the often poorer medical facilities, the benefits could in fact be greater in non-Western countries in the case of the mere serious injuries. Regretably, few of these countries have, as yet, introduced compulsory wearing of seat belts. 5 57 Driver error Percentage of drivers Failed to stop at red signal 13 (6) Failed to give way when turning left on red signal 12 (6) Failed to stop at stop sign - traffic near 52 (6) Failed to stop at stop sign - no traffic near 99 (6) Cut corner on right turn 48 (3) Turned right from wrong lane 42 (3) Failed to give way when turning right 36 (5) Drove wrong way down dual carriageway 51 (4) Crossed double white lines 15 (8) Highway Engineering 27. There has been increasing evidence fram the tS and the USA that relatively detailed local accident investigation, cozabi~ned with low-oost engineering remedial measures, can be highly costeffective. The experience being gained fran following this approach in these two countries is of particular relevance elsewhere. In T~stern countries, a grow'ing emphasis has been placed in recent years on cbtaining value for noney fran noney spent on road safety. With their lower gross national products, this must also be an important consideration for African and Asian countries. 28.. Work by Jorgensen and T~stat (13) in the US indicated clearly the high benefit-cost ratios that could be cbtained fran "spot" improvemants as compared with continuous widening or overall nodernisation projects. The limited data available to them strongly suggested that low-cost projects yield the greatest safety benefit per dollar expended. In the UK Duff (14) showed a sinilar result and his analysis of 29 schemes confirired that small inexpensive schemes could have a very marked effect on road safety. As far as the author is aware, no studies similar to those mantioned above have been completed in a non-J~stern country. However, the Overseas Unit, TPRLI has started work on the effectiveness of low-cost highwy engineering counterreasures in Egypt and Pakistan. 29. In order to investigate the relationship between accident rates and geometric design standards, a different technique was adopted by the author (15). Using data collected in Kenya and Jamaica, personal injury accident rates on main inter-urban roads were correlated with certain geoartric design characteristics. Step-wise multiple regression analysis, in which the accident rates were expressed as a function of several independent variables was used to correlate the n~me of personal Injury accidents per million vehicle kilcomatres with gearetric design features. It was found that the accident rates fell as the standard of the road improved; in the equation derived for Jamaican roads, road with and junctions per kilometre were significantly related to the aocident rate whereas in Kenya, junctions per kilomretre, horizontal curvature and surface irregularity were found to affect the accident rate. Since this work was carried out by the author, research woiikers in a nuxrber of countries, including India and Chile have also managed to correlate accident rates with georretric design. DATA COLITECI(X AND ANALYSIS 30. Little progress can be made on improving the road accident situation in a country until the prcblem itself has been clearly defined. Accident statistics must be collected over a period of tine so that an understanding is cbtained of where accidents are occurring, to what classes of road user, at what tinre of day and in what type of accident. Most of the accident patterns described in the previous section could not have been identified without there being sone sort of accident data collection system in operation. 31. Road accident data need to be collected over a wide range of levels from the broad perspetive of the national scene to the detail of the individual accident. Although analysis of national accident statistics show the source of prcbleni (as seen above), they do not indicate specific rerredial neasures. In order to do this, data need to be collected at the "regional" level. In such studies neasurenents of vehicular and possibly pedestrian flows are cbtained as well as accident statistics. Examples of studies of this type given earlier would include work carried out by the Overseas Unit in Naircbi and Surabaya. Finally, a detailed understanding of factors involved in road accidents can be cbtained only by "local" in-depth studies of road layout, vehicle design and road-user behaviour. At the present tine the Overseas Unit safety team are carrying out such analyses on major roads in Egypt, in the Islamabad. region of Pakistan, and also the Addis Ababa-Asab road Ethiopia. 32. As stated, it is essential in dealing with road safety problema that a good accident data collection and analysis system be established. With this goal in mind, the Overseas Unit, TR1UL and the Egyptian Ministries of the Interior and Transport began a prograimme of cooperative road safety research involving the following catponents. (a) Police Accident Booklet Design. An experinent was conducted in which four different designs of police booklet were compared, including one heavily dependent upon symbols and pictograms. Fran the results, two compromise designs were drawn up and tested in field trials leading eventually to one preferred design. This final booklet has now been tested by the police throughout Egypt and its performance carefully mronitored. Since these initial trials took place the booklet has been introduced and further developed in Pakistan and Ethiopia. In this research, the aimn has been to optimise the ease, speed and accuracy with which the booklet can be ccanpleted, whilst at the sane tine ensuring that sufficient details are recorded for the purposes of accident analysis. (b) Microcomputer Analysis.. A low-cost microcomfputer system has been developed and tested, again with the errphasis on ease of use (16). The rapid developuents in m~icrocomputers in recent years is opening new possibilities in accident analysis. Their low cost and general 558 T25 robustness in difficult environments make them well suited for use in non-western countries. Micros also have the benefit of being readily accessible and available for iimmdiate use; since catputer facilities are fewer and more centralised in developing countries, this is probably a greater advantage than in more industralised countries. In view of these various potential benefits the development of an experimental low,-cost microcomputer system was included in the prograirne of cooperative research in Egypt. Early prototypes of the package where tested in Egypt and developrent of the system are currently under trial in Pakistan, Botswana, Ethiopia and Papua New Guinea. The system is designed for use at the "Local Authority" level (county large city, etc). Particular empasis has been placed on ease-of-use, with the operator merely having to select one of a nurber of options at most stages of the prograntes. Accident and casualty cross tabulations, accident record retrieval, histograms of accidents along a route and "stock diagrams" analyses are among the features under developed. In the field, the major hurdle to be overcome is that of inaintenance in the event of breakdown. (c) Accident Investigation. This work follows broadly the British "Local Authority" approach to accident investigation as described in the British Departxent of Transport Accident Investigation Manual and in the Guidelines for Accident Reduction and Prevention in Highway Engineering produced by the UK Institution of Highways and Transportation. Analysis has been carried out on roads in the countries mentioned above. Accidents at the most critical black spots have been analysed in detail using the "stick diagram" technique.. In this way coenen factors such as overtaking or nose-to-tail accidents have been identified. Not unexpectedly, marked differences in the nature of the accidents taking place were found with the different sites. Following these analyses appropriate low-cost remedial measures are being introduced and their effectiveness evaluated. SUMMARY 33. Since 1972 the Overseas Unit of the UK TMR has been engaged in a progranne of research on road accidents in non-Western countries. Results to date indicate that fatality rates (per licenced. vehicle) are high in comparison with those in developed countries, and whereas in Europe and North Arrerica the situation is generally ihmroving, many developing countries have experienced a worsening situation, Particularly over the last five years.. A preliminary study indicated that road accidents cost on average alhost 1% of these countries' annual gross national product and it is clear that road accidents are utilising scarce financial resources that the countries can ill afford. 34. Using statistics published by the World Health Organisation and United Nations, data was obtained fram 20 non-western countries using the most up-to-date classification of causes of death used by these two organisations. It was found that road accidents ranked highly as a cause of deathi in these countries. Although the countries for which data were available may not be representative of the entire Third World, it is clear that road accidents represent a growing social problem, particularly for juveniles, young adults and those in early middle age, and also a growing econanic problem for the countries as a whole. Almost all countries of Africa and Asia suffer fran a lack of financial resources and the sums of money available to spend on road safety improvreents, road rehabilitation and maintenance, police enforcement, etc. will be severely limited. Consequently it is particularly difficult for these countries to deal effectively with their road safety problems. 35. Although research findings fran Western countries can provide same guidance, the inevitable uncertainties surrounding their transfer to other countries anpasise the need for caution in their application. As shown in the paper, the problem faced by many countries is often markedly different fran that in Western Europe and North America. This, coupled with major difficulties in road-user behaviour, knowledge and attitude introduce an element of uncertainty in the potential effectiveness of many countermeasures. Results however suggest that the introduction of improved education, training and enforcement could be highly beneficial in non-Western countries and the potential for improved road safety by these methods is greater than in the developed world. It is essential that scarce resources are not wasted and that any measures that are introduced are carefully appraised and an assessment made of their relative effectiveness. The careful monitoring of remedial nmeasu-es and an assessment of their cost-effectiveness is one of the goals of an ongoing prograxire of research within the Overseas Unit of TRRL. This points to the further and fundamental need for a good accident data collection and analysis system. This should be sufficient to produce essential information for accident investigation purposes but, at the same tine, it should not be too sophisticated either for the needs or capabilities of those who operate it or contribute to it. This again is part of the ongoing prograrrme of research in the Overseas Unit. 36. The work carried out by the Overseas Unit TRRL over the last ten years has done imuch to identify the magnitude and nature of the road safety problem in non-Western countries. Based on these research findings it is hoped that international organisations such as the World Bank, WHO etc, in collaboration with the various countries themselves, will be able to invest wisely and effectively in road safety programmre to contain this growing problem. 559 REFERENCES (1) Fouracre P.R., and Jacebs G.D. Caipaative accident costs in developing countries. TMR Report SR206tX, 1976. (2) Jaccds G.D., and Fouracre P..R. Further researcth on road accident rates in developing countries, TIEL Report SR207, 1977. (3) Jaccbs G.D., and Hards W.A. Further research on road accident rates in developing countries (s-econd report) TRRL Report SR434, 1978. (4) Jaocks G.D. The potential for road accident reduction in developing countries. Transport Reviews. 2,213-224, 1982. (5) Jaccbs G.D. Road accident fatality rates in developing countries - a reappraisal PTR1 Surre TVnnual Conference Sussex University July 1986. (6) Sabey B.E. Accident analysis in Great Britain. Paper presented to the 1st Int.Conf., International Driver Behaviour Research Association, Zurich, 1973. (7) Jacehs G.D., and Sayer I .A. An analysis of road accidents in Kenya in 1972. TPPL Report SR227, 1976. (8) Jacdbs G.D., and Sayer I .A. A study of road accidents in selected urban areas in developing countries. TRR Report LR775, 1977. (9) Sabey B.E., and Staughton G.G. Ineracting roles of road environmet, vehicle and road-user in accidents. 5th Int.Conf of Imt Ass for Accident and Traffic bledicine, London, September 1975. (10) Jaccts G.D., Sayer I .A., and Downing A.J.' A preliminary'study of road-user behaviour in dev-eloping~ countries. TRRL Supplenetary Report SR646, 1981. (11) Downing A.J. Road accidents in Pakistan and the need for iinprovements in driver training and traffic law enforccments. PTRC Stunner Annual Conference Sussex University, July 1985. (12) Jaccis G.D., and Downing A.J. A study of bus safety in Delhi, TMR Report SR758, 1982. (13) Jorgensen R., and Associates and T~state Research Inc. Evaluation of criteria for safety iniprovements on the highway. Report to Bureau of Public Roads, Washington, 1966. (14) Duff J.T. The effect of small road improvements on accidents. Traffic Engineering and Control 13 (6), 244-245, 1971. (15) Jaccbs G.D. A study of accident rates of rural roads in developing countries, TRR Report LR732, 1976. (16) Hills B.L., and Kassabqi DL A microccanputer road accident analysis package for developing countries. PTRiC Summer Annual Conference. University of Sussex July 1984. Czown Copyright. Any views expressed in this paper/article are not necessarily those of the Departbnent of Transport. Extracts fran the text may be reproduced, except for caffnercial purposes, provided the source is acknowledged. TABLE 2. The main causes of rood accidents as determined by police In different countries F ~~~~~JAMICA GHANA BOTSWANA MALAYSIA HONGa-XaNo 1977 1974 1976 1978 1977 Number of Number of Number of Number o Number of accidents accidents accidents acc ident accidents Main ue of (incl udin~g % (not kno-e (Including % (inc luding % (injury % accident damage If damage damage damage accidents only) onl only) only) only) ___________ - inoclulded) Road-user error 7,027 95 8.164 77 844 71I 41,997 87 3.309 92 Vehicle defect 108 Il 1.679 18 137 12j 656 11 Adverse rood conditions or 72 1 551 5 19 21 3,675 8 environment _____________ Other 225 3 262 2 176 15 1,963 4 303 8 TOTAL 7,432 100 10,456 100 1,176 1001 521191 100 3,612 100 .grouped aith 'other' 560