Preventing road accidents

The WHO's first ever Global Status Report on Road Safety (full pdf here) reveals that over 1.3 million people die every year due to road accidents, and that at 1,14,590 deaths in 2007 India topped the charts, followed by China with 89,455 deaths. The report, which describes road accidents an "epidemic", is based on the first broad assessment of the road safety situation in 178 countries, and uses data drawn from a standardized survey. It also points to between 20 and 50 million non-fatal injuries every year.

The report traces these accidents predominantly to poor enforcement of road safety laws on the five major risk-factors - over-speeding, drink-driving, not following safety precautions like helmets, seat-belts and child restraints - and finds that more than half the victims are pedestrians, cyclists and bikers. Expectedly, over 90% of world’s road fatalities are in low & middle-income countries, which have only 48% of world’s registered vehicles.

Within India too, Andhra Pradesh and Maharashtra tops with 1,37,000 deaths each, followed by UP with 1,25,000, and Tamil Nadu with 1,20,000 deaths.



In view of the relative ease of data collection and its availability, among the various policing subjects, road safety stands out as being especially amenable to scientific analysis. However, as the WHO report writes, the data collected on road accidents and injuries in many developing countries are both incomplete and of doubtful quality.

There have been a number of empirical studies in recent years which have explored various aspects of road safety - negative externalities generated by drunk drivers, effect of alcohol on driver risk, the effectiveness of safety devices like air bags and seat belts, children's car seats, effect of insurance, nudging drivers to drive safely, contribution of driver distraction to causing accidents etc.

Taking cue from these aforementioned empirical analysis of road accidents, here is an intutive illustration of how careful analysis of data can assist in minimizing accidents on highways. Take the example of the recent expansion of national highways (NH) across the length and breadth of the country. While this has had widely acknowledged economic and social benefits, it may also have imposed significant costs arising from increase in road accidents resulting from fast-moving vehicles. Here is one (and just one) simple illustration of how empirical analysis of data can be useful in preventing accidents along these highways.

A detailed analysis of time series data (of say, 12-18 months) on road accidents across a stretch of national highway could, for instance, help localize vulnerable areas (typically stretches at the entry to and exit from NH-side habitations), highlight accident-prone timings of the day (typically between two hours on either side of midnight), and identify commonest victims (say, pedestrians crossing the road or cyclists). Having identified the most likely locations, timings and causes, road patrolling can be made more focussed and prevention of accidents would suddently become a more hopeful task.

It therefore stands to reason that instead of spreading out their attention over the entire stretch of NH and for the entire duration of their shifts, and in the process diluting its effectiveness, each patrol team can become far more effective by concentrating their attention on the specific accident-prone areas, times and types of victims.