I.~~~~~~~~~~~ TRANSPORT RESEARCH LABORATORY IA ;? '. ? O 1 J i J .. TITLE Road safety in Africa - background:,and overtvi ew : J. S Y~erre11 andA J Downing,- Overseas Centre Transport Research Laboratory Crowthome Berkshire United Kingdom by I1.. t A.-I j. .-i ,.' 1,...I ') : .:
.7 1 i 1 1 . .1 .' .. .1 C, - ' ...: ' --- .1 .' 1 .1 ...t ' --1t
.- 1'
.: 1 1 -L- A 'I- . y ., 1 . . . .1 1, : t '
i ; : ..1 5. . 1 ' ' ' .1 ' 1 - i 1 'I 0 .; ., 1 .' ' -l l'
1 k . ... i 1 . ..1 1 i .1 1
' j l.' : . 1 - YERRELL, jS.,-1989.. Road..Safety in. Africa -Background and overview. ln: Second African Road Safetv Coneress, Addis Ababa 16-20 October 1989. k ROAD SAFETY IN AFRICA -BACKGROUND AND OVERVIEW J Stuart Yerrell, Head, Overseas Unit,TRRL and Andrew Downing, Head of Road Safety Section, Overseas Unit, TRRL INTRODUCTION Road safety cannot be considered in isolation. Its problems and possibilities are intimately linked to social and economic conditions, as is the transport system in which they arise. This background paper takes a broad look at the road accident situation in the African continent, particularly the changes which have taken place in the last decade. It then attempts to set these in the context of some of the other major social and economic changes taking place, and draw some general lessons for the future as well as provide a background against which the more specialised papers of this conference can be considered. A recurrent theme will be the scarcity of reliable data on which to make statistically-sound statements, and the dangers in looking for "global" trends in a continent as richly diverse as Africa. ROAD DEATHS IN AFRICA Table 1 records the 'change in absolute fatalities for certain countries between 1980 and 1986. The detailed statistics are given in an appendix to this paper, and have been drawn from a variety of sources, including the questionnaire compiled and distributed specifically for this meeting. The table covers 22 countries for which reliable data exist for the two years in question. Egypt, Nigeria and South Africa, are listed separately, because the size and nature of their contributions could distort the statistical picture). The most striking feature of the final line of table 1 is the relatively small increase in absolute fatality numbers over the six year period. The earlier dire warnings of impending crisis and of a new form of "epidemic" appear, with the benefit of hindsight, to have been misplaced. Whether they were mistaken or simply premature remains, of course, to be seen. The fall in the number of fatalities in Nigeria is to some extent an artifact of the two years chosen for comparison; recorded totals have been volatile for the past decade rising from around 8000 in 1977 to a peak of nearly 11,400 in 1982.and falling back to below 8000 in 1987. The decline is clearly associated with a steep decline in the amount of travel, but unfortunately reliable data on vehicle ownership and use are not available. Even a cursory. examination of the other national data reveal some striking differences. The average of 9.4% for the 19 nominated countries conceals national increases of greater than 25% for Botswana, Cameroon, Ethiopia, Kenya and Rwanda, and decreases of greater than 25% for Ghana, Liberia and Niger. Although absolute figures are of concern to national administrations and policy makers, they can, by themselves, give a misleading impression of the true situation and they are, of course, unsuitable for international comparisons. Two measures which allow for exposure are commonly used; fatalities per head of population (personal safety) and fatalities per vehicle or vehicle-km (system safety). The latter are more customary amongst road and transport authorities and will be used in the subsequent discussion. Ideally, we would consider fatality and casualty rates per vehicle kilometer, but in the absence of this data for most African countries the simpler statistic of fatalities per 10,000 vehicles will be used. The relevance of this is immediately obvious fro m table 2, which shows vehicle numbers for 1980 and 1986. (The sample size for the group of countries has dropped from 19 to 16, with the omission of Liberia, Mozambique and Senegal. The original 19 countries covered some 35% of the African population, and inclusion of Egypt, Nigeria and South Africa boosted this to 68%). The vehicle figures for Egypt and Nigeria have been estimated assuming that vehicle growth is directly proportional to gross national product -an estimate which is likely to be on the conservative side. Table 2 suggests that the fatality rate with respect to vehicles has in fact improved in our selected group of countries, but worsened in South Africa. To examine the situation in more detail, we need to include population numbers, and examine the relationship between fatality rates and vehicle ownership levels. Plotting national data of fatalities per vehicle against vehicles per head of population produces what is sometimes called a Smeed diagram, after the English professor who first used it to demonstrate a seemingly global statistical relationship between these two quantities. Such a diagram is shown in figure 1 and it summarises information for several African countries for the years around 1980. The regression. line from that figure appears again in figure 2, which also now includes national data points for 1980 and 1986. A general drift in the direction suggested by the regression line as motorisation increases can be detected, but there.is clearly a wide range of movements which makes any generalisations suspect, if not meaningless. This point is brought home more forcibly in figure 3, where the 16 sub-Saharan countries have been aggregated to. give a single "African" data point. The movement of this point between 1980 and 1986 indicates a decrease in the fatality rate per vehicle as the number of vehicles has grown, and even a slight improvement in the fatality rate per head of population. However, if the data from South Africa is included, the situation becomes almost static, since that country experienced a slight worsening---of its fatality,,rate per vehicle over the six years in question which does not appear to have been compensated by an increase in the number of vehicles. This limited analysis suggests that:- (i) the data should be re-examined and individual countries asked to provide more verified national data (ii) when more countries are represented, a clearer picture may emerge at the sub-regional level. allowed to deteriorate to save the rest. Tables 4 and 5 summarise information coming from the World Bank's Road Maintenance Initiative -a component of its sub-Saharan African transport programme (SSATP). Table 4 shows that the countries of West, and East and Southern Africa, some 39 in total, have only about 30% of the main network paved, and a quarter of this is in poor condition. Only 50% is classified as good, and the situation with the unpaved network is even worse. Although table 5, which shows the figures for 1984 in parenthesis, suggests that the situation is relatively stable, the projections for the future are not optimistic and point at best to a barely adequate network. There are two immediate lessons for road safety:- (i) there will be a premium on small low cost improvements which could be integrated with minor upgrading and maintenance activities; (ii) the number of accidents which can be directly attributed to poor road condition (and related factors such as marking and signing) is likely to increase on parts of the network. ROAD SAFETY AND THE YOUNG National safety statistics from Africa show that a high proportion of the casualties are young people. To a considerable extent this is to be expected, being simply a reflection of the typical age distribution. Figure 4 shows the population pyramid for Africa as a whole (although the data is for the male population, the female numbers are virtually id'entical until over the age of 60) Nearly half the population is under the age of 15. The situation is projected to be very similar in 15 years time -only the absolute numbers will have been increased by some 50%. By contrast, the population of western Europe (shown on the right of figure 4) is remarkably stable and the only new problems facing the road safety planners is a modest increase in the number of elderly people who', hopefully, will be increasingly active as medical and social care improves. The implications of these curves for planning and resources is clearly immense, but they immediately suggest that measures which concentrate on school children should have both priority and long term benefits. They also suggest that such measures will consume an expanding share of available resources -. and, if directed at training and behaviour, will need constant renewal as well as growth. Collective measures, as distinct from individual ones, will be preferable. Progress in the general area of education may well, be. limited by the channels of communication, as much as by the availability of suitable materials. A survey of a sample of 33 countries reveals that just under a half have more than.80% of eligible children in primary education, and that at secondary level only one country (Egypt) can claim more than 60% attendance. The majority are below 20%. (Table 6). The figures vary considerably from country to country and careful country-specific studies will be needed before efficient programmes can be drawn up and implemented. URBAN ROAD SAFETY Another factor which is having, and will continue to have, an important impact on road accidents and their distribution is the growing urbanisation of the continent.- This is already reflected in the national accident statistics from several countries, and is often. accompanied by a reduction in overall severity and an increase in the number of reported accidents. Increases in pedestrian death and injury are also reported. Urbanisation is a global phenomenon, but whereas the relatively low income countries of the sub-Saharan Africa have the lowest absolute figures they can boast the greatest rate of growth (Table 7). Once more, the situation is diverse, with a sample of 37 countries showing a quarter with less than 20% of their population in towns in 1987, but an equal number in the range 40-49%. (Table 8). The relative affluence of the urban areas means that the levels of motorisation are well above the national average, even if that motorisation is reflected in two-wheeler ownership or fleets of public service vehicles rather than the private motor car which dominates western transport planning. The growth of urban traffic presents both challenges and opportunities to the road safety community. The challenges are self-evident, but the opportunities lie in the greater accessibility. of the urban network both for remedial measures (whether they be in traffic engineering, enforcement or wherever) and for post-accident emergency services and medical care. Also there is a tendency for urban accidents to concentrate along busy city corridors and in commercial areas, enabling safety measures to be targetted at the most sensitive parts of the network. ROAD SAFETY PROGRAMMES Road safety programmes will be dealt with in some depth by other papers in this conference and two of the themes which are likely to emerge are: (i) the wide diversity of road safety actions that are possible, and the need for cost- effective programmes which are tailored to suit the national and local conditions and situations; (ii) the need for multi-disciplinary action in devising and running such programmes, and a corresponding need for close collaboration and inter-sectorial cooperation to make the best use of scarce resources. Road safety is not unique. in. its diversity of a ctions and its possible spheres of influence, but the number of different sectional interests it penetrates make the institutional challenges a significant part of the overall problem. It is therefore encouraging that, of the 23 countries which responded to an OECD questionnaire, 8 reported active national road safety committees, either governmental or non- governmental, and actions being taken on a significant number. of fronts. Table 9 summarises the reported activity in the 23 countries, although it must be stressed that the replies gave no details of extent or duration of the measures planned or implemented. There is no apparent correlation between the level of activity reported and the GNP of the country. Ji Clearly African countries and indeed most developing countries have made encouraging efforts to improve road safety in the last decade and in doing so they have displayed a good deal of technical competence, goodwill and enthusiasm. It is hoped that these efforts will continue throughout the next decade and more and more countries will adopt a scientific approach to the wide range of national road safety programmes which Africa needs, with an emphasis on data collection, national 'analysis and experimentation and evaluation. Selected sources 1. 'Reducing Traffic Injury -A Global Challenge' by W Trinca et al. Royal Australasian College of Surgeons, Melbourne, Australia 1988. 2. Road Deterioration in Developing Countries. World Bank Policy Study, Washington 1988. 3. World Bank Atlas 1988, World Bank, Washington (Also similar for earlier years). 4. 'Road Deterioration in Sub-Saharain Africa' M Mason and S Thriscutt. Road Maintenance Policy Seminar, Harare, June 1989. 5. World Development Report 1989, World Bank, Washington. 6. World Demographic Estimates and Projections 1950-2025, United Nations, New York 1988. 7. World Development Report 1988, World Bank, Washington. ACKNOWLEDGEMENTS The work described in this paper forms part of the research programme of the Overseas Unit of the Transport and Road Research Laboratory and the paper is published by permission of the Director. C Crown Copyright. The views expressed in this paper are not necessarily those of the Department of Transport or the Overseas Development Administration. Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged. APPENDIX 1 Road accident and related data in African countries Country Year Population Faas Csate eils (Veh~)X 41 (Fatal) 4 (Casualties) 4 3 (Pop) (Veh ) x 10Veh )X1 (10) Botswana 1980 .814 116 1118 26825 329.5 43.2 .416.8 1986 1149 192 2632 .55604 483.9 34.5 473.3 Burkina Faso 1980 6908 - 47128 68.2- (Upper Volta) 1986 8101----- Cameroon 1980 8503 671 6820. 98000 .115.3 , 68. 5 695.9 1986 10151 1117 7992 1834600 180.7 60.9' .438 Congo 1980--- 1986 2000 106 315 . Cote dlIvoire 1980 .8034 719 10241 244707 304.6 29.4 418.5 1986 10650 728 9137 324224 304.4 22.5 281.8 Ethiopia 1980 30017 96 7 1494 53295 17.7 181.4 280.3 1986 44743 1258 4385 64427 14.4 195.3 680.6 Egypt 1980 42201 4999 22519 724821 171.8 69.0 310.6 1986 49739 - - - - - Gabon 1980 548 196 1119 23824 434.7 82.3 49.9 Ghana 1980 11450 992 9843 76544 66.9 129.6 1285.9 1986 14045 705 6977 63306 45.1 111.4 1102.1 Kenya 1980 16402 1413 8900 168718 102.9 83.7 527.5 1986 20384 1832 15377 299000 146.7 61.3 514.3 Lesotho 1980 1339 184 925 14267 106.5 129.0 648.3 1986 1586 230 1383 22000 138.7 104.5 628.6 ILiberia 1980 1873 118 1842 2897 15.5 40.7 6418.1 1 986 2253 51 256 - - - Mlalawi 1980 5968 - -38510 64.5- 1986 7380 53 7 2571 - -- Mauritius 1980-- 957 124 19fg44- 69188- 723._0 17.9-' '28'f.0 -.- - 1986 1029 109 2834 78229 760.2 13.9 362.3 Morocco 1980 20242 2 256 3 289 7 642948 317.6 35.1 511.6 1986 22466 - - 774064 344.5- IMozambique 1980 10473 515 3204 - -- 1986 14336 626 2761 - - - Niger 1980 5533 219 1791 39448 71.2 55.5 454.0 1986 7250 73 742 58998 81.4 12.3 125.8 INigeria 1980 85630 9150 25904 640045 74.7 143.0 404 .7 1986 103147 - - - - - Rwanda 1980 5046 187 3120 16685 33.1 112.1 1870.0 1986 6236 264 2231 23783 38.1 111.0 938.1 Senegal 1980 5661' 429 5475 115155 203.4 3 7.3 475.4 1 986 6770 450 7853 - -- ISeychelles 1980 65----- 1 986 66 5 205 4834 732.4 10. 3 424. 1 Population (Veh) 4 (Fatal) 4 (Casualties) 4 Country Year 3~~ Fatals Casualties Vehicles (Pop) X( 0 Veh) X( 0Veh )X10 South Africa 1980 29285 7572 81219 4118392 1406.3 18.4 197.2 1986 32436 9343 102867 4674784 1441.2 20.0 220.0 Sudan 1980 18691 592 5860 .130933 70.1 45.2 447.6 1986 22567 665 6055 1 67536 74 .2 39.7 361.4 Swaziland 1980 56181 937 34100 613.3 53.1 274.8 1986 689 157 677 43800 635 .7 35.8 154.6 Tanzania 1980 18263 954 6381 132000 72.2 72.3 483.4 1986 23049 1071 6542 151000 65.5 70.9 433.2 Togo 1980 2625 210 1539 - - - 1986 3144 275 2840 27200 86.5 101.1 1044.1 Tunisia 1980 6369 988 9716 273908 430.1 36.1 354 .7 1986 7311 987 10415 639600 874.8 15.4 162.8 Zimbabwe 1980 7360 1140 9538 281973 383.1 40.4 338.3 1986 8705 934 9508 347177 398.8 26.9 273.9 Table 1 ROAD ACCIDENT FATALITIES - AFRICA 1980 1986 GROWTH 19 COUNTRIES 10 719 11724 1005 9.4 EGYPT 5000 5200 200 4.0 NIGERIA 8736 8154 -582 -6.7 SOUTH AFRICA 7572 9343 1771 23.4 TOTAL (22 COUNTRIES) 32027 34421 2394 7.5 Table 2 VEHICLE NUMBERS - AFRICA (1013) 1980 1986 GROWTH 16 COUNTRIES 2304 3091 787 34 EGYPT 725 964* 239* 33* NIGERIA 640 567* 7 3 i -11.4* SOUTH AFRICA 4118 4674 556 14 TOTAL (22 COUNTRIES) ] 7787 9296 1509 19 -:ESTIMATED ASSUMING PROPORTIONAL TO TOTAL GNP Table 2 Table 3 AVERAGE ANNUAL GROWTH OF GNP PER CAPITA 1973- 1980- 1985 1986 1987 1980 1984 INDUSTRIALISED 2.1 1.3 2.4 2.0 2.2 COUNTRIES DEVELOPING COUNTRIES 3.1 0.7 3.3 3.1 1.8 SUB-SAHARAN AFRICA 0.7 -4.9 2.9 -0.2 -4.6 AV. 1980-87 -2.9% Table 3 - .- - 0~0 000 zb-4-'C --:c 0
4z E 0
4 E
(,
L=] CD Z - r.J1 '0(N 11li 1,111r-4i1 (N i i cc: i 11 k_4..4 1 i rx. Ln rn itn en : C4 i i fa 00 N- (Ni (N rn 17 .~~j~ 'n0~ al ' r- (Nr ( c en r- - -.T a.' o~ (N 0. ' - 0 -7 0 en- 0 (Ni Ln. -7 - C'CD 0' C' z ~ ~ ~ z~ CD P~ ~ 0 z rE z0b-4 z0U- 0 01' 00-C 0 z~-'C-, -'C W tn- CO- co 00-Cr 0 z -'C -'C 0 Qz Table 5, ROAD CONDITION SUB-SAHARAN AFRICA 1988 MAIN NETWORK CONDITION SOURCE: WORLD BANK ROAD DETERIORATION STUDY WEIGHTED AVERAGES ACCORDING TO'LENGTH PAVED UNPAVED %GOOD %FAIR %POOR %GOOD %FAIR %POOR EAST and 50 33 i 17 2830 42 SOUTHERN (20) ~(42) (32) (26) (42) I (30) (28) WEST (21) 54 1 7 29 1 9 34 47 (52) (23) (25) (20) (36) (44) TOTAL - ALL ~~52 25 23 29329 SUB-SAHARA (41) (47) (27) (26) (33) (32) (35) 1 ., Table 6 EDUCATION IN AFRICA - 1986 SAMPLE OF 33 COUNTRIES Table 7 URBANISATION - GLOBAL TRENDS 1 9 65 198 7 % RATE %RATE LOW INCOME 14 4.8 24 5.6 (less INDIA & CHINA) LOWER MIDDLE 3 9 3.8 51 3.5 INCOME UPPER MIDDLE 46 3.9 6,6 3.2 INCOME SUB-SAHARAN 1.4 5.5 2 7 6.9 AFRICA OECD i77 0.8 Table 8 URBANISATION IN AFRICA - 1987 SAMPLE OF 37 COUNTRIES *INCLUDES SOUTH AFRICA Table 9 RESPONSES FROM 23 COUNTRIES (SELF-COMPLETION QUESTIONNAIRE) ACTIVITY NO MEASURES MEASURES AREA REPLY PLANNED IMPLEMENTED Data collection 10 4 9 Data analysis 11 6 6 Traffic legislation 8 5 9 Law enforcement 6 4 13 Driver training/test 6 6 11 Vehicle testing 6 8 9 Engineering (road) 7 5 11 Education/publicity 6 8 9 Alcohol measures 12 7 4 Emergency services 14 4 5 Public -r-ansporc 13 5 5 Goods tr:ansport 14 4 5 TOTAL 113 66 97 1 .4%% .4% * 4% *- Cameroon 44 * Ma~~~lawi * *.% Ethopia Nigeria * * 4. 4*% S *",~~~ %.4 4. *4% 10 100 Vehicles per 1000 persons Fig 1. Developing countries 1978-80 Relationship between fatality rates and vehicle ownership levels Transport and Road Research LaboratoryM27 1000r- -100CU CD0 C, LL lob- .44% 1 1 A 1 M2076 South Africa: 1980 1986 F v 1o4V 103p 18.4 139 20.0 144 1978/80 Regression ln 1980 16 Sub-Saharan countries % 1986 16 Sub-Saharan countries plus South Africa 1 A~~~~~~~~~~~~~~~~~~~~~~ 100 10 Vehicles per 1000 population Fig 3. Aggregated data for changes in fatality rates and vehicle ownership 1980-86 Transport and Road Research LaboratoryM27 1- 1 1 1 11* 1 looor 000d 100 M2078 -n -0 0 0. 0~m (D 0. zJ, C- 0 N)000) 00 (01 a,C) 0a)0 m 0.m 0, Transoort and Road Research LaboratoryM27 M2077