Why data-driven action is key to putting the brakes on road carnage

Whenever I travelled by matatu, it took me time to settle on a seat for the journey. The big debate that raced in my mind was the safest seat in case of an accident.

I always calculated it from the type of vehicle it was (14-seater, later 18-seater vs 29-seater), the amount and type of exposed metal and the likelihood of being in a head-on collision, a tailgating accident or a rollover.

Another of my favourite activities was watching a programme named Air Crash Investigation on the National Geographic channel. The programme, which was fascinating in a slightly morbid sense, detailed how plane crashes — and even near-misses —were investigated until the cause of the accident was known. Many times, the cause of the accident was a tiny mechanical part that failed, or pilot error that stemmed from something as innocuous as arrogance or sleepiness.

All this came back to me last week, after the latest of the fatal Ntulele accidents. On Thursday, a matatu was involved in a collision with a bus in Narok which led to the deaths of nine people. This was just a few metres, and a few months, from where a bus crash had left more than 40 dead. A day earlier, the Cabinet Secretary in charge of Transport and Infrastructure, Michael Kamau, had launched an initiative that would require all serious accidents to be investigated.

The second Ntulele accident came too soon for the Ministerial directive to have been implemented, but it showed the potential, and limitations, of Mr Kamau’s well-meaning initiative.

Kenya is one of the deadliest countries on which to use a road. Motor vehicles are operated by people who have no business being behind the wheel. Obviously defective vehicles ply our roads, or are used for the wrong purposes (motorcycles used to ferry whole families, or cows; station wagons used as 14-seater passenger vehicles).

Pedestrians dash across busy highways expecting that drivers will brake on time. Traffic policemen are more concerned with bribes than with stopping carnage, and public service operators treat the profit motive as the ultimate motivation, all else be damned.

Every time there is a serious road crash, much hand-wringing ensues, followed by earnest, wrong-headed pledges to finally do something about them. And many of these directives — from presidents to government ministers and police, are often utterly inadequate. Even worse, certain areas are declared “black spots”.

Bumps are erected after public outcry, and, most ridiculous of all, prayers are held to “cleanse” stretches of roads of “demons baying for more blood which can only be sated by the sacrifice of a few more humans”.

All these initiatives and directives fail at one core level. There is very little reliable data on road use and misuse, in Kenya. We do not know, to any useful degree, what causes accidents. We cannot ascertain whether particular makes of vehicles are prone to accidents, and why. We do not know why people cross roads at certain places, and thus determine where to place footbridges.

We’re not sure why these footbridges, even when they’re built at great cost, are shunned (except for guesses about laziness and African peculiarities).

The new National Transport and Safety Authority will succeed, or fail, on that one strategy. Lee Kinyanjui and his team have a chance to change the face of road safety, and bring down the terrible death and injury toll, if they become obsessive about data-driven action.
This will mean getting away from press-conferences and proclamations, and investing in a wealth of computing power and data-mining experts. It will mean that every single traffic incident would have to be investigated.

This may be a controversial recommendation, but think: most traffic accidents in urban settings do not lead to death or serious injury. Typically, even traffic police recommend that watu wasikizane – the fender-benders should reach informal agreement. But this leads to very little actionable data on urban traffic, which leads to bad road design and control.

Were data to be the key determining force in traffic control in Kenya (Air Crash Investigation-style), we would understand, to an unimpeachable degree, what actually causes road accidents in Kenya, and what mitigating measures to take to actually reduce them and their severity (and not simply rely on gut-feel or intuition).

It would instantly become obvious which vehicles, roads, and types of road users are more prone to mishap, and what to actually do about it.
Computing power and cost is coming down to a level when we can actually install data collection devices (black boxes) on every single vehicle on the road in Kenya, and use the information generated to make the roads less lethal.

And I would then, finally get an answer to my question on which the statistically-safest seat is in a matatu.

Mr Kantai is the Business Editor - NTV. @wgkantai