Transport risks and uncertainties

We hear a lot about risks of transport. Hardly a week passes without a dramatic story about a plane or train crash, and many people change their travel arrangements based on stories read in the media. Why is this, and what are the true figures?

The perceived risk often differs from the actual risk of something. This can be caused by media reports - the media tends to focus on large, rare events, and therefore rail and air travel are reported on more. This is added to the fact that many passengers do not fully understand how an plane works (or, occasionally, doesn't work) and feel that they are out of control in one, and you find that many people are terrified of air travel, and use it only when absolutely necessary. In addition, accidents in other forms of transport are also sometimes greatly feared.

So, what are the real figures? How do they compare with people's conceptions? All figures are taken from the Department for Transport - Road Casualties Great Britain: 2007 and are in micromorts/ 1000 kilometres and kilometres/ micromort.

Note (HEJ): See Table 7a on page 82. However, only the figures for bus/coach, rail, water and air match this. Where has Owen taken the others from?

A similar table using imperial measures is below.

This table shows the risk of each mode of transport, in micromorts/ thousand kilometres and kilometres/ micromort

The risk of each mode of transport
Mode of transport Micromorts/ thousand kilometres Kilometres/ micromort
Air Close to 0 A lot
Rail 0.3 3,333
Water 0.2 5000
Bus/ Coach 0.3 3,333
Car 2.7 370
Van 0.9 1,111
Motorcycle 111 9
Pedal cycle 34 29
Pedestrian 44 23

So what does this table tell us about risk? The first, and most obvious thing, is that the vehicles which have rare, dramatic crashes - such as airplanes and trains - are not, in fact, very risky to be on. The other thing is that the risk of traveling, even on the most dangerous vehicles, isn't very high. For instance, to have a 1 in 2 chance of being killed in a motorcycle accident, you'd need to travel 9,000,000 kilometres by motorbike. Media reports can rarely be used as an accurate gauge of the real risk of an activity. It also seems to say that walking a few miles is more dangerous than the average trip in an airplane... but it would obviously depend on where you are walking. Rail fatality figures do not include those trespassing on the railway lines or those attacked on railway premises, so the danger they cause to other people (for instance on level crossings) is not taken into account. The same goes for car accidents, which can cause deaths to pedestrians and cyclists. So the true danger to society of such vehicles may be higher.

The DfT gets the figures for rail travel statistics from the Transport Statistics Great Britain 2007 edition. The specific data for rail travel statistics can be found on page 108-9 on this page.

The air travel information can be found in the Civil Aviation Authority's 2006 annual report.


Note (HEJ): I can't see air fatality numbers on that page? Perhaps they're taken from Table 91 (page 156) onwards of here somehow?


The water travel fatality information is taken from the Marine Accident Investigation Branch's accident reports.

The water travel exposure information is taken from the Department for Transport Sea Passenger's statistics page.

Below is the same table in miles.

This table shows the risk of each mode of transport, in micromorts/ thousand miles and miles/ micromort

The risk of each mode of transport
Mode of transport Micromorts/ thousand miles Miles/ micromort
Air Close to 0 A lot
Rail 0.5 2,000
Water 0.3 3,333
Bus/ Coach 0.5 2,000
Car 4.3 233
Van 1.4 714
Motorcycle 179 5.6
Pedal cycle 55 18
Pedestrian 71 14


If you wish to research more into these issues, the DfT, RSB, CAA and MAIB all have details on accident reports and tables in close to excessive detail. If you want to find out about all the train accidents with more than 25 fatalities since 1945, the exact definition of "Airprox" or the Male casualties: by built-up and non built-up roads, road class and severity 2007, then there really are no better places to look. It's certainly good to know that someone is working away meticulously recording these things, just in case someone feels like analyzing them.

Levels: