Chapter 6 - Economic Cost of Road Crashes in Africa

6.1 Introduction

As part of the recent project undertaken on behalf of the Global Road Safety Partnership (GRSP), mentioned earlier in this report, it was estimated that the global cost of road crashes in 1999 was in excess of US $500 billion and the cost in the developing world was estimated to be about US $65 billion. Road crashes therefore are costing developing and transitional countries huge sums each year that they can ill afford to lose.

Detailed studies undertaken in the UK and elsewhere in the developed world indicate the high rates of return that can be obtained from the application of low cost traffic management techniques at sites where significant numbers of crashes take place. Results from studies undertaken in the developing world are now showing similar results (Baguley 1995). Thus apart from the essential task of reducing road deaths and injuries in developing countries, a sound case can be made for reducing road crashes in African countries on economic grounds alone.

A recent (unpublished) report undertaken by TRL and Ross Silcock for DFID (UK) reviews how the methodology used for costing road crashes in developing countries can be improved. This makes the point that even within the transport sector alone, hard decisions have to be taken on the resources that a country can devote to road safety. In order that this decision-making process has a rational basis, it is important that a consistent and soundly based method be used to determine the cost of road crashes and the value of preventing them.

Thus the first need for cost figures in African countries, is at the level of national resource planning to ensure that adequate investment in road safety takes place in a given country. It helps at the national planning level therefore to provide decision makers with an estimate of the annual cost of road crashes. For example, in a study undertaken by TRL in the early 1990's (unpublished) the annual cost of crashes in Mauritius was calculated to be US$32 million. A series of recommendations were outlined at a total cost of US$800,000 spread over a five year period which could reduce crashes (and hence costs) by 5 per cent p.a. (i.e. saving US$1.6 million p.a.). Thus the average first year rate of return on investment was estimated to be 1000 per cent. High rates of returns such as these are in fact, fairly common in road safety appraisals, and indicate clearly the value to be obtained in crash-reducing measures.

Conversely a review of road safety in Ethiopia by a joint TRL - Ross Silcock team estimated that the annual cost of road crashes was about Birr 400 million p.a. Thus it is reasonable to suggest that at least 10 per cent of this sum be spent on a crash reduction programme each year. However a proposal that 2 per cent of a newly established Road Fund, (about Birr 8 million) be spent on a comprehensive crash reduction programme each year was actually rejected. Bearing in mind the annual cost, of about Birr 400 Million an investment of say Birr 40 million (ie 10 per cent of the annual cost) would not have been unreasonable.

One particular problem might be that road safety in developing countries is not seen as the key priority for any specific organisation. Thus, from a highway and transport point of view, the main priority is that of maintenance and for the health sector the main priorities in Africa are AIDS, malaria etc. Whilst there is a desire to see road deaths and injuries reduced, there is, in virtually all countries, a lack of investment in crash-reduction programmes.

A second need for crash cost figures is to ensure that the best use is made of any investment and that the most appropriate improvements are introduced in terms of their cost-effectiveness. If specific costs and benefits are not applied to crashes taking place then widely different criteria in the choice of measures, the assessment of projects and the allocation of resources will result. As a consequence there will be an imbalance in the ways in which funds available for safety projects are allocated and also an overall under-investment in road safety.

Details of how road crashes can be costed in developing countries (together with associated problems) are given in the TRL Report, Overseas Road Note 10. Basically it is important that crashes are costed by degree of severity (ie fatal, serious, slight and damage-only) and that the Human Capital approach is used with sums included to reflect the pain, grief and suffering of those involved and also their loved-ones. An unpublished report to DFID explains the difficulties associated with the costing procedure, particularly the difficulties in obtaining accurate information on lost output and property (mainly vehicle) damage and also assessing a realistic value for pain, grief and suffering. (Another critical factor, discussed earlier in this report, is the under-reporting of road crashes, particularly those not involving a fatality).

6.2 Crash costing in Africa

There are relatively few examples available of efforts made to cost road crashes in African countries. Those countries where results are available include Zambia, Botswana, Kenya, Tanzania, Ethiopia and South Africa. All countries used the Human Capital approach (also known as the Gross Output method), or a variation of this method, and some included sums to reflect the pain, grief and suffering involved (see above). More details of crash costing in Africa can be found in Appendix C: Literature Review.

One way of comparing results from different countries is to express total annual national costs derived, (where costs by category of fatal, serious, slight and damage-only are multiplied by the numbers of those classes of road crashes taking place) as a percentage of Gross National Product (GNP) per annum. Early work by Jacobs and Fouracre (TRL, 1978) showed that road crashes cost any country, be it developed or developing, about one percent of GNP per annum. For many years this value has been used to provide a crude estimate of road crashes either on a national, regional or global basis. Evidence is beginning to emerge, however, that the annual cost of road crashes, certainly in the more wealthy countries, may be closer to two percent of GNP than one percent

Estimated costs in those African countries above show that, as a percentage of national GNP, costs range from about 0.8% in Ethiopia and 1% in South Africa to 2.3% in Zambia and 2.7% in Botswana to almost 5% in Kenya and the Kwa Zulu region of South Africa.

Costing road crashes, deaths and injuries is a difficult and contentious topic and any value derived is better for investment in life-saving activities than no value at all. It should be appreciated, however, that any value is likely to be little more than an estimate of the 'real' value.

Even in the west there is no simple method consistently used by all countries and comparisons should be treated with caution. Results which range from 1991 to 1998 suggest that in developed countries costs range from a low of 1.1% of GNP (Denmark) to 4% (New Zealand). Costs derived in different countries are in fact higher if the "willingness to pay" approach is used as opposed to the "Human Capital" method.

From the above it can be seen that it is difficult to state with any confidence that road crashes cost African nations a specific and precise percentage of their annual GNP. In order to derive a crude estimate of costs therefore, a range of values has been used, based on an estimated sum of all GNPs of countries in Sub-Saharan Africa for the year 2000 (of US$ 310 billion) (Table 6.1).

A cautious estimate, based on the value used for many years of one percent of GNP, which is fairly close to the value derived recently in South Africa and Ethiopia, suggests a regional cost of about US$ 3 billion. Higher estimates based on 1.5 to 2.0 % of GNP suggest $4.65 and $6.2 billion respectively.

Clearly, these are sums of money that the nations of Africa can ill-afford to lose every year.