Appendix A: Estimating Road Fatalities in Africa
1 The Current Situation
1.1 Estimating Road Deaths in Africa
Previous global or regional reviews of road deaths undertaken by TRL, the World Bank and others have acknowledged the problems associated with data reliability and under reporting. That said, traditional reliance has always been on the use of officially published statistics based on police reports. This study describes the results from these official statistics and then makes a "best" estimate of the real totals using the described correction techniques.
The officially reported number of persons killed in road crashes in the different countries (for the latest year available) are shown in Table 3.1. Data are available in most countries for years 1996 to 1998 but in a few such as Angola, Niger, Sudan and Liberia, the latest year was as far back as 1985. In some countries the reported deaths are very low, for example 22 in Chad, 43 in Gabon, 58 in Central African Republic. There may indeed be relatively few road deaths in these countries but it is much more likely that they are significantly under-reported.
Two countries alone account for almost fifty per cent of all reported deaths, namely South Africa and Nigeria. The South African value of over 9,000 appears to be consistent over time. For example, it was at about the same value in the mid 1980's. Nigeria on the other hand at 6,185 deaths is now showing a dramatic reduction from a high of over 9,200 just a few years ago. This recent figure must be treated with caution in that it is unlikely that a large amount of investment in road safety activities has taken place in recent years resulting in this dramatic reduction. Other countries also showing significant numbers of deaths include Ethiopia, Kenya, Uganda and Ghana. The 42 country total of 35,394 deaths and the individual national totals are significant underestimates of the true totals.
To improve the estimate of the current fatality situation in Africa, four additional correction steps need to be undertaken. These are:
- Updating the fatality figure from the latest year available (which ranged from 1985 to 1998) to the year 2000.
- Estimating for those countries where fatality data was not available at all.
- Adjusting for the under-recording of fatalities by the police when the official database does not include all fatalities notified to the police.
- Adjusting for the non-reporting of fatalities by the public, i.e. where the police are not notified of the incident.
This correction process was developed as part of a global review undertaken by TRL for the GRSP (Jacobs et al, 2000) and the results for each stage are described in turn below.
1.1.1 Updates to the Year 2000
Most countries had published road fatality data with the latest year available ranging from 1985 to 1998 and the first step involved updating this to the current year 2000. However after investigating time series data for seven of the major countries of Africa it was apparent that the use of an overall average growth rate would produce misleading predictions because the figures would be biased by some countries.
A more accurate model was developed by subdividing countries with time series data into three groups where group one contained those countries with large numbers of fatalities (Nigeria and South Africa), hence dominating the continent. Another group was formed from countries with fewer road fatalities which included Ethiopia, Malawi, Tanzania and Zambia. A third group included those countries such as Kenya where trends were significantly different.
Regression equations were fitted to the total number of road fatalities in each group with linear regression used unless a strongly non-linear relationship was in evidence. The analyses suggested that the reported number of deaths in Sub Saharan Africa of 35265 for the latest available year could, by the use of regression equations applied to trend data, be adjusted to 39,700 for the year 2000.
1.1.2 Non-reporting by countries
Despite a review that included many sources, road fatality data could not be found at all for a number of countries in Sub Saharan Africa. The percentage of countries with no fatality data was about 25 per cent of the total which in population terms was about 20 per cent of the total population of all countries included in the study.
An attempt to represent these countries by motor vehicles would have been preferable but a significant number had no vehicle statistics available either. Accordingly, the number of reported road fatalities was adjusted upwards according to the ratio of population in countries with and without fatality data as follows.
Population Adjustment = Reported Fatalities x Total Population / Population in countries with fatality data
This adjustment increased road fatalities from 39,700 to 47,650
1.1.3 Under-recording of fatalities
The problem of under-recording occurs when fatalities are reported to the police but are not included in the official database. Reported fatalities may be excluded from the official database because of reporting definitions and updating procedures.
As defined by the Convention of Road Traffic (Vienna, 1968), a road death is deemed to have occurred when a person injured dies within 30 days of the crash (and as a result of the crash). However, not all countries use a 30-day definition with some countries using 'on the spot', within 24 hours, 3 days, etc. Adjustment factors have been developed by various organisations to bring these countries not using the 30-day definition `into line'. If this is not done, then a significant level of 'under-recording' will occur.
Adjustment factors identified included those recommended by the Economic Commission for Europe and the European Conference of Ministers of Transport (ECMT) and the United Nations. The more recent ECMT values indicate that if the definition 'on the spot' is used, then numbers recorded should be multiplied by a factor of 1.3.
In the most recent UK Casualty Report (1998), the ECMT values were applied to Turkey (30% increase for a one day reporting definition) and Korea (15% increase for a 3 day reporting period) (DETR, 1999). However, it was assumed that the ECMT values would not apply for all less motorised countries (LMCs). In Africa, a larger percentage of road crash fatalities can, unfortunately, be expected to die within the first day with the lack of medical facilities as well as the higher percentage of vulnerable road users involved.
Given the lack of information on the timing of LMC road crash deaths, the assumption was made to use half the ECMT values in this study. Thus those countries reporting road crash fatalities occurring only within the first day of occurrence would have their figures increased by 15 per cent rather than 30 per cent.
A second assumption was made regarding the standardised death definitions in African countries; while many countries state the use of a 30 day definition, this could be interpreted to apply at the local level and for prosecution purposes. Road crash statistics are based on report forms that are often to be completed as soon as possible, i.e. '24 hour report form'. It can be very difficult to modify previously reported or submitted figures, especially where manual reporting systems are used which is the case in most African countries at the local level, where casualty reporting occurs. Accordingly, it was decided that regardless of the official definition, a one-day reporting time period would be assumed to apply for all of Africa and thus a 15% inflation factor was logically applied to all reported fatalities.
1.1.4 Non-reporting by general public
Almost all countries require road crashes to be reported to the police. However examples are shown in the report to the GRSP 'Estimating Global Fatalities' (Jacobs et al, 2000) of the levels of non-reporting identified in various countries throughout the developing world which can be considered relevant to Africa.
Evidence was found that of under-reporting of fatalities ranged from a minimum of 25 percent (Brazil) to as high as 350 per cent (Philippines) in LMCs. Accordingly, to adjust for the extent of non-reporting of fatalities, the conservative but realistic decision was made to use adjustment factors of25-50 per cent for the countries of Africa.
The factors indicate a probable range of fatalities. Given all the uncertainty in the estimation, a range is much more appropriate than a supposedly precise figure and estimates will be shown for both factors.
1.1.5 Current fatality estimate for Sub-Saharan Africa
Based on the methodology described above, a realistic estimate of total road deaths for the 42 countries is between 68,500 and 82,200 for the year 2000. The calculations and totals are presented below (Table A.1).