The Road to Safety

2001-2005

Setting the scene: The Statistics, the Safety Problem, the Key Contributory Factors

Vital Statistics: South African Road Use

There are currently about 6 million licensed drivers and about 6,73 million licensed and registered vehicles on South Africa’s roads. This figure is made up of:

• 3,86 million sedan cars [57.36%].

• 252 000 minibuses - of which some 126 000 operate as minibus taxis [3.75%].

• 24 000 buses [0.36%].

• 227 000 motorised heavy freight vehicles [3.37%].

• 1,2 million light delivery vehicles (bakkies) [17.83%].

• 1,17 million other (trailers, including truck trailers, motorcycles, special vehicles, etc) [17.38%].

• The total road and street network is approximately 500 000 km in length.

• The annual vehicle/km travelled on SA roads is estimated at about 98 000 million - or 268m km a day - of which 153m is travelled daily on the rural road network (including inter-city national roads).

• Keeping all this traffic moving are fuel sales of about 9500 megalitres a year.

The Safety Problem: the Overall Picture

• There are currently about 512 000 traffic crashes a year [511 605 in 1998].

• Of these, about 28 000 are fatal or lead to serious injury [7 260 crashes involving fatalities, 21 265 involving serious injuries in 1998].

• In 1998, 9 086 people lost their lives on SA roads, while 36 246 were seriously and 84 358 slightly injured.

• The total cost of these crashes to the SA economy is currently estimated at around R13.8 billion a year.

The most recent finalised statistics on crash rates in South Africa (1998) show a figure of 776 crashes per every 10,000 vehicles per annum—or 400 crashes per every 100 million vehicle kms travelled (as per the two main types of International benchmark measurement).

In terms of the first benchmark, South Africa’s road fatality rate for 1998 was 13.73 per 10 000 vehicles.

Where does this put us in global terms? The statistics are sobering. In the chart below we give a representative sample of international fatality rates per 10 000 vehicles for the latest year (1996) when such global figures were available.

With the state of Victoria, Australia, registering 1.2 deaths per 10 000 vehicles and China registering 26, South Africa – despite the welcome drop from 15.13 deaths per 10 000 vehicles in 1995 to 13.73 in 1998 – still falls well to the "wrong side" of the mean. What emerges from this comparison is that our performance puts us firmly in the "developing world" category – ahead of China and India, but significantly behind Turkey and Brazil (and roughly on a par with Russia). The scale of the problem we face could not be clearer.

However, we need to keep a sense of historical proportion here. In the next chart we offer a long-range retrospective view of fluctuations in the South African road fatality rate (1935 – 1998).

From this chart, two facts immediately become evident. The first is that our record has always been bad, though with significant downward "blips" in the last years of World War Two (fuel shortages/less domestic travel) and in 1974 (compare with 1973) when, as a result of the global oil crisis, tight speed restrictions were enforced on SA roads).

The second is that in recent years – and particularly since the launch of Arrive Alive – we have witnessed what is starting to look like a significant downward trend. The aim of The Road to Safety, in combination with an all-year Arrive Alive programme, is to consolidate and accelerate this trend.

[Note: Arrive Alive is a multi-dimensional programme of enforcement, coordination and communication. But the emphasis it places on speed limit enforcement is set to intensify greatly over the life of this strategy. This approach is grounded not only in the lessons of 1974, but in the results of study after study conducted in developed, middle-income and developing countries, all of which prove an incontestable correlation between speed (both absolute and too high for the prevailing circumstances), insufficient reaction time, crashes and fatalities. (See "The Speed Debate" below)].

Driver/vehicle analysis

Having outlined the general picture, the next step is to focus on crash and fatality trends for the different types of vehicles (and by implication their drivers) using South African roads. This has for some time now given us the information we need to serve as a rough indicator of the areas to which remedial attention should urgently be directed. The next figure tells the basic story.

The most important pointers here are the following:

The sedan car is overwhelmingly the type of vehicle most often involved in crashes. This is obviously partly due to the fact that these vehicles comprise 57% of the total vehicle population.

But, if we exclude the category of "other" vehicles from our calculations (on the grounds that many of them are not regular users of public roads) we arrive at a more realistic road usage share, which gives us a reconfigured % of the (active, daily) vehicle population for each vehicle type.

This then allows us to more accurately represent their respective levels of involvement in crashes, and so get a clearer idea of comparable driver/vehicle behaviour between categories.

The picture that emerges looks like this:

In other words, we see heavy trucks, minibuses (including taxis), and buses all scoring above their representation in the vehicle population – with trucks scoring around 1½ times their weighting, minibuses scoring around double their weighting and buses scoring at nearly 2¾ times their weighting.

Summing up:

1. The high absolute number of vehicles involved in crashes represented in the graphic at the foot of p.12 tells us that:

• A large number of all drivers are misbehaving and/or seriously under-qualified/under-skilled.

• Far too many vehicles are unfit for the road.

1. The discrepancies between the different vehicle types in terms of their relative rates of involvement in crashes indicates that the combination of driver and vehicle unfitness is particularly acute in the heavy freight, minibus taxi and bus sectors.

This drives us towards our strategic emphasis on the following key areas of intervention:

• Further tightening up on-road enforcement of speed, alcohol limits and vehicle overloading (Arrive Alive operational issues);

• Restructuring our inspectorates, cleaning up driving licence and vehicle testing, formalising and regulating the driving school sector, setting tougher professional driver qualification criteria and tightening regulation of freight and public passenger fleet operators (system & structure issues).

The need for these targeted interventions is further confirmed if we examine trend-based data on both crash and fatality rates per vehicle type per 100 million vehicle kilometres travelled over the period 1991-1998. The figures (see over) tell a mixed story.

LDVs’ rate of involvement in crashes is lower than the average for all vehicles, and has shown a slight downward trend over the period.

• The crash rate for the category "all vehicles" has shown a steady, if very gradual, downward trend over the period.

• The crash rates for heavy vehicles declined from 1991-93, rose from 1994-96 and began declining again in 1997-98.

• The crash rate for buses was constant at just below 600 crashes per 100 million vehicle kilometres travelled from 1991-93, and then began to show a sharp upward trend, apparently peaking at about 950 in 1997. It is too early to say whether the slight decline in 1998 represents the start of a reversal in this trend or not.

• The crash rate for minibuses has been fairly constant at a high level of about 900-1000 crashes per 100 million vehicle kilometres travelled (apart from a peak of over 1200 in 1996). There is as yet no sign of any downward trend emerging.

If we now turn to fatality rates for the same period, we see additional patterns of unevenness.

This graphic tells us that:

• The fatality rate for the category "all vehicles" has come down from 11.5 (1991) to 7 (1998) per 100 million vehicle kilometres. This represents a significant downward trend, probably associated with increasing seatbelt usage and gradual improvements in the safety construction of new vehicles.

• The fatality rates for heavy vehicles and LDVs have fluctuated slightly but are fairly constant with a very slight decline.

• [Note: the fatality rate for heavy vehicles indicated here is somewhat misleading, since fatalities registered in crashes are based on vehicle occupants. In crashes involving heavy freight vehicles, the rate of occupant deaths is much lower than the rate of collateral deaths caused (persons in other vehicles and pedestrians].

• The fatality rate for buses showed a constant and alarming upward trend from 1991 to 1997 (from under 5 to 13 fatalities per 100 million vehicle kilometres travelled), almost catching up with the fatality rate for minibus taxis. It is too early to say whether the drop to 11 fatalities/100 vehicle km registered in 1998 represents the start of a sustained decline or not.

• The fatality rate for minibuses has fluctuated at a consistently high level of between about 15 and 20 fatalities/100 million vehicle km travelled. There is no sign yet of any reversal in this trend.

Road crashes and deaths: the key contributory factors

Our research has regularly shown that the role played by each of the three major factors involved in crashes can broadly be broken down as follows:

We therefore have to ensure from the start that the attention we pay to each factor is proportionate to its significance for road safety and has the potential to achieve the greatest impact over the medium to long term. Many of the actions that appear in the list of short to medium term interventions (i.e. actions which will be unrolled during the life of the strategy to 2005) are "starter-actions" aimed at rebuilding eroded or historically inadequate institutional foundations. They will not solve all the problems instantly. But they will create a new basis for incremental strengthening of our regulatory and enforcement systems over time.

Key driver factors

Speed

• Excessive speed or speed too fast for circumstances plays a role in approximately 30% of all crashes and about 50% in the case of commercial freight and public passenger vehicles.

The "Speed Debate"

Arrive Alive continually drives home the simple message "Speed kills!" There are those who dispute this statement, drawing attention to issues such as the following:

• The general debate on speed does not properly distinguish between speed causing a crash, and speed influencing injury severity once a crash occurs.

• Most road crashes occur in urban or peri-urban areas (i.e. at speeds lower than average freeway speeds); and the crash rate on national roads is significantly lower than in these areas;

• The relative speed of moving vehicles (hence force of impact) is a more significant factor in the outcome of a crash than the absolute speed of a single vehicle;

• Moving violations (speed too slow for circumstances, improper lane usage, illegal turning, running red lights or stop signs etc.) are factors at least equally salient in crashes as is absolute speed.

• On-road enforcement has traditionally devoted disproportionate attention to "trapping" on open stretches of national roads and in small towns (allegedly for "purely revenue raising" purposes);

• Local road environment and human density factors are an integral component in the determination of what a safe speed is, and in the incidence of actual crashes.

We recognise the validity of these points; so long as it is recognised that they capture important aspects of the truth, not the whole truth. We stick to the emphasis on the negative role of speed for very good reasons—and with the following specific emphases:

• The most important issue is undoubtedly speed too high for prevailing circumstances. This applies to both national and local roads. It does not, however, invalidate the need to set speed limits. If you say that "speed is not the issue" - and one assumes that this means on national roads, not in built-up urban areas—where do you draw the line? 250 km/hr? An unrestricted autobahn approach to South African national roads? Our national road environment is characterised by sudden changes of width, surface quality, weather and visibility conditions, and is further complicated by the widespread presence of pedestrians and stray animals along both urban and rural roads. We must set limits that take proper account of this complexity, and of road-user fallibility (both drivers and pedestrians).

• International crash research over many years has led to a powerful consensus on the relationship between absolute speed (as well as speed too high for circumstances) and two other linked factors: driver information overload (leading to impaired decision-making capacity) and (compounding this) the reduction of available reaction time to allow for avoidance of a potential crash situation. Crashes almost always have to do with the unexpected (i.e. a sudden dramatic increase in the number of decision variables confronting the driver). In such cases (typically arising at intersections or along busy stretches of urban or national road) the driver’s "reasonable" safety calculations, based on "normal" driving conditions, cease to apply. The equation is simple: greater speed, exponentially less available reaction time, more certainty of a crash, more devastating impact. ("The faster you go, the harder you hit.")

• Absolute speed, being a critical factor determining the likelihood of death and the severity of injuries sustained by occupants in single vehicle or multiple vehicle collisions, is of major concern to us: for all road users, but in particular for passengers in public transport vehicles (buses, coaches and minibus taxis). The death and injury rates prevailing in these sectors are absolutely unacceptable; the passengers are entirely innocent parties; and the great majority of them are "captive" to these modes of transport, without access to alternative options. Their rights require special defence, in terms of speed limit enforcement, professional driver competence and vehicle fitness assurance.

• Speeding is in many instances also related to a mind-set and behaviour patterns that are selfish, aggressive and unconcerned with the safety of other road users. It is often closely connected with the other forms of moving violation that contribute to crashes. Changing these underlying psychological/cultural factors is therefore a critical long term need, particularly in the light of widespread negative driver attitudes that still reflect live elements of South Africa’s violent, divided, highly individualised past.

Speeding violations therefore must and will be enforced, with the following proviso. Speed enforcement will pay—indeed is already paying—much greater attention than in the past to data-driven targeting. This means concentrating enforcement on identified rural, urban and peri-urban hazardous locations and road stretches, at the times of week and hours of day and night during which our statistics show the highest concentration of crashes occurring (including, of course, pedestrian collisions).

We hope that the discussion above has made it clear that, while we will continue to treat speed as a key safety issue, this does not mean that we elevate it to a level of importance above all other factors. It should be evident from what we have said that we explicitly recognise the importance of all the other factors at work—driver, vehicle and environment-related—and you will see that the whole thrust of this strategy is to develop interventions in each key area of road safety that will be both locally effective and mutually reinforcing from an overall systems perspective.

Other driver factors, which to a greater or lesser degree contribute to crashes, include:

• Driving under the influence of intoxicating liquor (continuously targeted by Arrive Alive) and/or a drug having a narcotic effect. (Testing will be stepped up during the life of this strategy, as accurate evidentiary technologies are introduced—see box);

• Fatigue resulting from exceeding the maximum acceptable uninterrupted driving hours (regularly targeted by Arrive Alive communication campaigns, and currently under negotiation with the road freight and public passenger transport sectors in the context of regulation of professional driver working hours and conditions – see Operator Fitness below);

• Basic driver incompetence (driving without proper training, without a licence or with a forged or fraudulently issued licence).

Note: The issue of basic driver incompetence moves to a very high priority position in this strategy, since our research into recurring patterns of driver error in crashes points to a clear causal relation between such error and the low levels of skill and awareness prevailing amongst thousands of drivers on South Africa’s roads. This is in turn clearly linked to the prevalence of low standards of driver training—both informal and in many driving schools—and to widespread fraud and corruption in our driving licence testing and issuing system.

[The Driving Licence Testing Inspectorate currently estimates that up to 50% of all licences currently entered into the Driving Licence Register may be licences issued or obtained in an irregular manner].

The short-term strategy tackles all these issues head on.

Other human factors include:

Pedestrian road use

South Africa’s hybrid first world-third world economy and infrastructure are sharply reflected in an extremely high rate of vehicle-pedestrian collisions and fatalities, as is illustrated by the graphic below.

In crashes involving pedestrians, jaywalking and walking under the influence of alcohol or drugs are the major contributory factors, being present in 40 to 50% of vehicle-pedestrian collisions in urban areas and in 30 to 40% in rural areas.

Enforcement focused on pedestrian behaviour contributing to collisions will be increased during the life of this strategy. Such enforcement will be carefully targeted as part of our broader approach to pedestrian safety, which also highlights continuous audits of hazardous locations, low cost engineering and signage upgrades, improved land use planning and sustained school and community education and participation programmes.

Vehicle fitness factors include:

• Under-inflated tyres;

• Smooth or worn tyres or fitting the wrong types of tyres;

• Poor brakes;

• Faulty steering;

• Poor lights/vehicle visibility;

• Poor general maintenance of other vehicle parts such as the chassis, body, wheel alignment and shock absorbers.

Operator fitness factors include:

• (Over)loading of passengers and freight vehicles used for commercial purposes;

• Management of vehicles (maintenance, safety and quality assurance regimes) and drivers (qualifications, refresher training, medical fitness, fatigue etc.)

Road environment factors include:

• Poor maintenance of the road surface, resulting in potholes and poor ride quality;

• Poor maintenance of the road reserve and/or road verges - e.g. tall grass at junctions obscuring oncoming vehicles, broken fences allowing cattle and other domestic animals to stray onto roads, and poor maintenance of gravel shoulders;

• Inadequate signs and markings.