Chapter 8 - Conclusions and Recommendations

The objective of this study was to analyse existing data and information on the road safety situation in Sub-Saharan Africa in order to identify the data gaps and priority needs. Based on the poor response received from an earlier survey distributed by the ECA, a shorter survey (restricted to crash data only) was sent to each of the 42 countries. Unfortunately, this too had a low response with less than 20 per cent providing the data requested. Accordingly, the data analysis was based on published sources such as the IRF statistics, and data identified in the literature review and project reports.

The calculations for the current estimate of the road fatalities in Sub-Saharan Africa are shown in Chapter 3. Between 68 and 82 thousand lives are estimated to be lost in road crashes in Africa in 2000. After adjusting for reporting definitions (+15%) and under-reporting (+25-+50%), this estimate is 40-65 per cent higher than that officially reported by the countries. The use of official road crash casualty statistics thus underestimates the true extent of the health burden, especially that of injuries, of which only a fraction are believed to be reported by the police.

Africa as a whole was found to have about 10 per cent of global road deaths in 1999 but contained only 4 per cent of global motor vehicles. Fatality rates (expressed as annual road deaths per 10,000 vehicles licensed) in a number of African countries are amongst the world's highest with some (Tanzania, Ethiopia, Uganda, Malawi and Ghana) in excess of 100. (In comparison, fatality rates in OECD countries are of the order of 1.5-4.5 road deaths per 10,000 licensed vehicles per annum).

The average growth in road deaths over the period 1985/86 to 1995/96 (excluding the two dominant countries of Nigeria and South Africa) was found to be over 40 per cent indicating the continued increase in road deaths over time. In comparison, road fatalities in Western Europe and North America fell over the same time period by about 20 per cent.

In the absence of a detailed sectoral review and few questionnaires completed by respondents, the amount of detailed information on road crash patterns available in the review was clearly limited. Nevertheless enough information was available to identify a number of key issues relating to different classes of road users.

An analysis of crash patterns showed that pedestrians are the largest category of road users involved in six out of the eight countries for which detailed information was available. In five of the countries, vulnerable road users (i.e. pedestrians, cyclists and motorcyclists) accounted for over 50 per cent of all persons killed (the equivalent value in USA, Germany and France for example is about 13 per cent). As might be expected, pedestrian involvement is even greater in urban areas. Thus in Addis Ababa for example over 85 per cent of both fatalities and casualties were pedestrians.

Results of the analysis of types of vehicles involved in crashes showed that the involvement rate of buses was, in some countries, almost four times that of their fleet share. Thus in Kenya, for example almost 30 per cent of all crashes involved a bus or minibus whilst the equivalent value in most Western countries is less than 2 per cent. Research on public transport safety undertaken by TRL has shown that public transport vehicles in Africa (and Asian) countries are frequently poorly maintained, often overloaded whilst the drivers themselves receive inadequate training. Public transport in many African cities is provided not only by the conventional bus but by 'paratransit' vehicles such as mammy wagons (converted trucks), and Matatu (converted vans etc). Such forms of public transport are poorly regulated and controlled with many operating illegally. These vehicles now have a reputation of being particularly dangerous.

It should be noted that the fatality range presented here (68-82 thousand) is less than half that previously predicted by the WHO (WHO, 1998). Even the higher WHO road fatality estimate was only one-tenth that predicted for HIV/AIDS and one-fifth that expected to be lost to malaria. Africa's road safety record is the worst in the world in terms of deaths on a per vehicle basis, but the region has other much greater causes of premature mortality. Road safety is unlikely to become a top medical or political priority in Africa and limited resources can be expected to be available for road safety improvements. Maximum use should be made of the limited resources and lessons should be shared. The recommendations listed below pertain to the perceived priority areas where needs are greatest and opportunities are being missed.

8.1 Sectoral Needs Analysis

As mentioned previously, this study was the first comprehensive review of road safety in Sub-Saharan Africa. It was primarily a desk exercise whereas other regional reviews of the Asia-pacific region, Eastern Europe and Latin America-Caribbean included country visits with local data collection a priority. In order to have a better understanding of the current constraints and capabilities within Africa, consideration should be given to the following:

1. A sectoral needs analysis should be undertaken by local counterparts in Sub-Saharan Africa. Standardised sector surveys should be developed by a coordinating organisation who will oversee the data collection and be responsible for the data analysis. This approach would provide a much better understanding of local needs and priorities which could be used for guiding future technical assistance. The recommended needs analysis should cover all key road safety related sectors including crash data systems, traffic law enforcement and regulations, road safety engineering, road safety coordination and funding, driver training and testing, vehicle testing, traffic safety education and publicity, motor insurance system, and medical services

8.2 Crash data

The situational analysis has been limited by the data available. The crash data survey response was very low and the summary presented here was collected from published reports and donor projects. This problem is not unique to Africa and plagued the previous regional reviews. The establishment of a regional centre for crash data is a frequent recommendation in cases such as this, where data is difficult to obtain. Intended to raise the status of crash data, this step is not believed to address the key problems, which include insufficient location details and lack of application. Priority instead should be given to the following:

1. Crash location details should be standardised and monitored. Road authorities should provide traffic police with road maps, including strip maps from rehabilitation projects, to help document locations. The coding of locations and subsequent identification of hazardous locations should be one of the first objectives of a crash data system.
2. Joint training of police and engineers should improve the understanding of their roles and the development of crash data transferral procedures (from police to road authorities) should be included in any crash data system upgrading. The road authorities should be involved in the improvement of crash data reporting as they are expected to be the main users. The end objective is the application of crash data and not just its computerisation.
3. Examples of best practice would help others understand what is involved and how it can be achieved. Case studies of successful systems should be documented and shared with other countries.
4. Regional databases should be considered after the crash data has received more priority at the national and local level and the data quality has improved from local usage.
5. Investment in hospital reporting systems would produce better estimates of the number being injured in road crashes.

8.3 Road Sector Development Programmes

The World Bank led Road Maintenance Initiative (RMI) in Africa resulted in countries adopting a coordinated approach to donor financing of road rehabilitation projects while increasing their self-financing capability through the use of a fuel levy. Road Fund Boards were established with private sector participation. Road safety appears to have been bypassed in this process with only one of the five Road Funds even mentioning safety measures (Ethiopia). While environmental impact assessments and relocation policies have been addressed, road safety concerns have too often been forgotten. This is partially due to a short term focus on maintenance needs by the highway authorities and a multi-sectoral approach to road safety which has focused on the linkages between the sectors instead of highlighting the individual sectoral responsibility. Accordingly, it is proposed that:

1. All Road Sector Development Programmes should reconsider the need for incorporating a traffic safety policy which would ensure a minimum standard of road safety in all road projects. An example of a model traffic safety policy is shown in Table 7.1.

8.4 Road safety engineering

As noted previously, many African countries are currently undertaking national road rehabilitation and maintenance programmes. This is a prime opportunity to review and improve the road safety situation of the national and regional highways. At present, road safety is assumed to be addressed through adherence to geometric design standards and few road safety audits are being conducted. Safety audits are needed as they consider the safety of all road users (including roadside dwellers).

1. Safety audits should be a basic requirement of all major road rehabilitation projects with donor-financed projects containing demonstration audits and training programme on road safety audits.
2. Lessons learned from safety audits and any resulting changes should be centrally coordinated and shared with other countries in the region. This should be initiated by the donors who should request copies of all references and materials produced be shared with neighbouring countries.
3. Road safety engineering recommendations from national road safety plans should be integrated into the existing road authorities' 5 year plan and work programmes. Road safety engineering funding requirements should be integrated into road authorities planning and budgeting process.
4. Road safety engineering unit guidelines (suggested organisational structure, work programme, job descriptions, training progammes) should be developed as a key reference for the Institutional Strengthening projects which have often neglected safety in the past.

8.5 Funding

Whereas road maintenance has benefited from the use of dedicated user charges, i.e. fuel levies, funding for road safety interventions has been limited primarily to general revenue and donor assistance. User charges for safety programmes need to also be considered. A safety surcharge on motor vehicle insurance premiums, has been considered by several countries and is believed to be in place in Mali. The advantages of this approach are limited by the low insurance compliance found in many countries. South Africa and a few other African countries have avoided this problem by collecting third party insurance through the fuel levy. A simple proposal - in terms of practical application but not necessarily political or public acceptability-would be to increase the fuel levy to provide a dedicated road safety source of funding. As the existing road funds have been designed with maintenance needs in mind and not road safety, this could involve an actual increase in the fuel levy, as opposed to a share of any existing road fund revenue.

1. More priority should be given to financing road safety measures from user charges and that the most practical and appropriate measure would be to apply, (if not increase) fuel levy to pay for road safety interventions.
2. User charges should also be considered as means of paying for safety measures in other sectors, such as a safety levy on a driving license to pay for more or better trained driving examiners.

8.6 Community Participation

Experience from both higher income countries as well as Africa has shown 'top-down' approaches to have limited effect and lack sustainability. Local participation is required to maintain political support as well as ensure the approach adopted, both in terms of problem analysis as well as remedial measure identification, is appropriate and compatible with local priorities.

1. Community participation should be a basic component of any road safety programme, with involvement of both service organisations, NGOs, CBOs and the business community a priority. Local road safety initiatives should be incorporated in national road safety programmes with bottom-up efforts co-inciding instead of following national level action.
2. Priority should be given to sharing the experience gained from community traffic safety programmes in motorised countries, including the USA, Australia, New Zealand and Norway.
3. Road rehabilitation projects should include publicity campaigns in villages adversely affected by road improvements with the local community involved in both the identification of the problem and the development of the remedial measures.

8.7 Collaborations and Partnerships

As highlighted by the earlier review of WHO figures, there are many other causes of death (HIV/AIDS, malaria, etc) which will rightfully absorb the health sector's attention. The health impact of road crashes will be overshadowed by other causes of death, especially at the national level. Whether it is injury prevention or AIDS related, partnerships with similar causes which require changes in behaviour and social responsibility should be sought. This wider approach has already begun in South Africa where CSIR have a safety education unit (instead of traffic safety). The WHO's recent initiative applies to all trauma victims with the Addis Ababa Emergency Medical Services Committee campaigning for increased trauma management skills and resources not only for road casualties but also for other trauma victims, such as from natural disasters such as earthquakes or floods.

1. Collaborations should be promoted between road safety initiatives and other health organisations and advocacy groups, ie. HIV/AIDS, trauma services, AIDS, community policing, urban development. Traffic safety education should be designed to complement awareness programmes on HIV/AIDS, as the overwhelming scale of the latter must take priority.

8.8 Monitoring and Evaluation

With limited resources available, it is easy to appreciate the need for proof of effectiveness and efficiency. Yet few evaluations were identified and the focus was on casualty reduction. The lack of complete and accurate crash data will handicap the use of casualty/crash reduction targets. Casualty reduction targets are also criticised for not adequately reflecting risk, as roads which are perceived to be dangerous may discourage use but this will not be reflected in the crash record. At this stage, skills and systems need to be introduced and implemented before any impacts will be seen and post-evaluations, which should be standard features of any project, should measure the change in capability, if not the impact on casualty levels.

1. Performance targets should be identified and agreed within each sector as proxies for effectiveness In many cases, casualty reduction targets will be premature given the uncertainty of the casualty data.

8.9 Central Road Safety Reference System

The literature review confirmed two suspicions: information on road safety practices/situation could not be found for many African countries, and secondly, where information was found, it was often contained in unpublished project reports from donor assisted projects. The latter has a very limited circulation list and the information collected is not being used to its full potential (both inside and outside the country). Reports are often shared on an informal basis by interested parties trying to build on previous work, instead of "re-inventing the wheel". With the move towards a global approach to road safety as seen by the establishment of the GSRP, it is recommended that

1. All donors agree to share their work and make project documents available to other donors and those organisations concerned with the promotion of road safety. Priority should be given to publishing recent project reports to allow greater dissemination and use whereas future reports should expect to be on general release.
2. There should be a central coordinating referral body identified for each country and one worldwide, preferably the GRSP. Inside a country, one organisation should have a copy of all road safety reports and within GRSP, all donor financed projects should provide abstracts if not copies of their reports in electronic format. Full copies of reports should be able to be downloaded from either donor websites or the GRSP website.
3. Consideration be given to converting the literature review contained in Technical Appendix C into a database with a standardised format so that additional references can be added. The African road safety literature database should be accessible via the internet (possibly GRSP or VTI's RETSNET) and updated on a regular basis.

8.10 Vulnerable road user safety

As a general observation it can be said that pedestrians are the neglected road users throughout much of Africa. Relatively few pedestrian crossing facilities exist in most major cities such as foot bridges, under passes, signal controlled crossings etc. Features that assist pedestrian safety, for example, guard rails or central reservations are also infrequently used. Many African countries are also characterized by quite heavy pedestrian activity along rural roads as people move from village to village but most people are obliged to walk in the road itself as few sidewalks or even paths are made available for pedestrians.

It has been the general experience of high income countries to overlook the safety of vulnerable road users during motorisation. Greater priority has begun to be given to the vulnerable modes of cycling and walking in motorised countries in recent years but the focus at present in low income countries is on vehicle operating costs and the efficiency of motorised transport.

1. Vulnerable road user safety should be a priority area, and pedestrian safety in particular. Training programmes should be undertaken to raise the awareness with engineers and police. Greater use should be made of CSIR's experience in pedestrian safety and consideration given to conducting regional training programme of CSIR's pedestrian facilities management and traffic calming courses. Donor funded road projects should specify vulnerable road user safety to be addressed.
2. Programmes and measures intended to promote vulnerable road user safety in motorised countries should be publicised in low income countries, i.e. speed reduction zones, traffic calming, home zones.

8.11 Professional drivers

With such low motorisation in the region and the reliance of freight and passenger transport on road, the priority should be on reducing the crash risk of professional drivers and commercial vehicles. Results of the analysis of types of vehicles involved in crashes showed that the involvement rate of buses was, in some countries, almost four times that of their fleet share. Much needs to be done to improve public transport safety in most countries of Africa, including improved vehicle condition and maintenance, greater central or local government control and perhaps above all, improved driver behaviour. Research on public transport safety undertaken by TRL has shown that public transport vehicles in Africa (and Asian) countries are frequently poorly maintained, often overloaded whilst the drivers themselves receive inadequate training. Action has begun in several countries, including South Africa which has targeted minibus and taxi driving and Zimbabwe which is believed to require defensive driver training for all its public service vehicle drivers.

1. Driver training and testing improvements and initiatives should target professional drivers. These should include extended on road tests with hazard perception checks, and defensive driver training. Refresher training should also be promoted along with close monitoring of drivers working conditions and crash involvement.
2. Driver training and risk management programmes should be proposed as a priority area for the newly established GRSP. Good practice code of risk management for fleet operators should be published.