What are we talking about really?

Many claims about dangers are based on observations about "excessive and inappropriate speed". What are they? Are there other sorts of speed? Which sorts of speed are dangerous?

The Government says "30% of accidents have excessive speed as a cause or contributory factor".

But what exactly do they mean? Are they talking about maniac speeding drivers? Drivers going too fast for the local conditions? Are these drivers likely to be slowed by speed cameras? And are they even exceeding the speed limit?

Of course, the Government research fails to support the claim. (click here) But at the very least we should examine what is meant by the claim, and what different sorts of "speed" they might be referring to.


"Speed" is not usually a helpful term in these contexts. Very low speeds can kill, for example a pedestrian could be crushed to death at under 0.5mph. Very high speeds can be safe too, for example a lone racing car at 200 mph on a track. And "speed" is always a feature in collisions simply because you cannot collide if you're not moving.

Numerical speed

We'll use the term "numerical speed" to describe the number on the speedometer. It's never possible to imply the degree of danger from the numerical speed without reference to the conditions. Reversing at 3 mph might be extremely dangerous while cruising on the Autobahn at 120 mph might be completely safe (and completely legal!).

You might not even think that there's anything special about road travel. (I don't) 25,000mph on 1969 on the way to The Moon wasn't dangerous. 1,500 mph in Concorde isn't dangerous. 250 mph in a high speed train isn't dangerous.

So numerical speed fails to tell us immediately of the degree of danger. In more useful forms, speed is a relative thing. The effect of a given numerical speed depends entirely n the conditions. As a simple example, 55 mph in a busy high street is dangerously fast, but 55 mph on a UK motorway in good conditions is usually pretty slow. 

Higher numerical speeds can be extremely dangerous in "snap shot" conditions (see below). To take an obvious case imagine a car 0.1 seconds before it impacts a 1,000 ton concrete block. At 1 mph there will be a little bump. At 20 mph there will be a frightening crash. At 40 mph and above the crash would almost certainly be fatal to the occupants of the vehicle.

Snap shot conditions

With reference to the "concrete block" example above. Many argue that since a higher speed at impact is more dangerous, then it follows that higher speeds are always more dangerous. In a single snap shot of road conditions this would frequently be true.

But real road conditions are forever changing, and in most driving circumstances drivers are constantly adjusting their speed to match conditions or the speed of other traffic. We need drivers who reduce their speed particularly in the vicinity of road hazards. It would usually be a truly awful driver who drove at a steady 30 mph in a 30 mph limit.

Suppose we did dump our snap shot concrete block in the road somewhere. How many would crash into it at all? And how many would crash into it without first adjusting their speed? We bet the answer to both questions is "very few indeed".

There's a dynamic interaction between a driver and his surroundings. When circumstances change the driver must adjust his speed or position to take account of the changes. Even on the straightest road with no obstructions if a driver closes his eyes for just ten seconds an accident is likely. Imagine what effect that same 10 seconds would have in a busy high street.

But if conditions were like snap shot examples we wouldn't need to concentrate so hard on the task of driving. This shows why snap shot conditions are so unlike real driving; in the real version we always have a driver adapting to the circumstances. Failures in the adaptation process lead to accidents.

Looked at from another angle, it's psychological factors that lead to accidents, and no simple matter of physics. However, let's look at the physics.

Speed and kinetic energy

Kinetic energy (KE) = ½.m.v² 

Where m = mass, and v = velocity. (velocity in physics terms is speed with a direction component. In the cases we're discussing speed and velocity can be considered to be the same)

Many people feel that larger numerical speeds are intrinsically more dangerous. Indeed simple physics appears to confirm this. After all damage and injury results from the sudden loss of kinetic energy when vehicles crash, and kinetic energy rises with the square of speed. This means that kinetic energy rises dramatically with increases in speed.

But kinetic energy doesn't imply the safeness of any particular numerical speed because sometimes there's very little risk of a crash at a high speed.  Indeed the very simple proof of this comes from the clear and well established fact that faster roads are statistically safer. The safest roads in the UK are the Motorways, and they are also the places where the highest speeds are used. 

Perhaps we should limit speeds so that the kinetic energy released in accidents is always non-lethal? While the idea has obvious attractions very few would consider it to be even remotely practical. This is because even at 20 mph the amount of kinetic energy stored is sufficient to kill. Most would immediately agree that limiting our fastest roads to a generally non-lethal 15 mph would be ridiculous.

Most people consider the UK motorway speed limit to be too low, and yet the kinetic energy in a car at 70 mph is easily sufficient to kill many times over.

Excessive speed

Excessive speed is usually defined as "speed in excess of the speed limit". It is not in itself dangerous to exceed the speed limit, and the vast majority of the UK driving population do so every day with no consequence whatsoever.

Inappropriate speed

Inappropriate speed is usually defined as "going to fast for the conditions". This could be over or under the speed limit. 69 mph on a motorway in thick fog could be a suicidal example. We abhor inappropriate speed because it is dangerous. Typically the opposite of inappropriate speed is safe speed.

Of course it is also possible for a speed to be inappropriately slow - Imagine the effect of a vehicle at 20mph in lane three of a busy and fast flowing motorway for example. 

Safe speed

A speed is called safe whenever the driver can stop safely within the distance that he knows to be clear. This is the Safe Speed Rule and is widely taught to drivers. It features in the Highway Code (Rule 105) and the Police Driver's Manual: Roadcraft. It is the most basic principle of safe driving. Our information page on the safe speed rule is here

So how are excessive and inappropriate speed linked?

One thing we can say right away is that the largest excesses of speed may often be both excessive and inappropriate. For an inexperienced driver, the speed limit might provide a useful cap to the degree of inappropriate speed.

Research in Canada found that in speed related accidents 2/3rds were directly attributable to inappropriate speed within the speed limit, while the remainder were in excess of the speed limit. Note that this also suggests that inappropriate speed above the speed limit was a cause or contributory factor in the remaining 1/3rd of speed related accidents. It is impossible to imagine any scenario where a speed could be excessive, dangerous and not inappropriate all at the same time. It's quite correct for appropriate speeds to be considered safe and inappropriate speeds to be considered dangerous.

Impact speed

So is it reasonable to infer that lower speeds across the entire road network would improve safety? It might be if we could simply turn down a magic knob a little, and have all speeds everywhere reduced by a percentage. But most crashes take place at relatively low speeds. We'd need to know that any realistic strategy we came up with would reduce crash speeds. Speed in non-crash circumstances clearly doesn't matter. So is there a general relationship between free travelling speed and crash speed? Impacts with pedestrians strongly indicate that there isn't, although if we never go faster than x mph it's clear that we'll never crash faster than x mph.

Factors influencing crash speeds are mostly driver behaviour. Poor observation and inattention lead to a lot of crashes.

We've examined the figures for accidents involving child pedestrians. It's quite obvious that free travelling speed plays little or no part in determining the accident outcome, by a factor of at least 500:1 actually. (click here)

To what extent might enforcement of a speed limit affect inappropriate speed?

Many dangerous behaviours on our roads are unlikely to be altered by speed enforcement. Indeed to the extent that speed cameras are tending to replace police patrols, some of these dangerous behaviours may well become more frequent. (click here)

Then according to Canadian research, 2/3rds of excessive speed accidents occur within the speed limit.

Then enforcement of the limits might bring their own dangers directly and indirectly. (click here)

There are cases where a speed limit could reduce the use of an inappropriate speed. We don't know of any research which put the number of cases at a significant level. We believe the true figure is around 2% of accidents. And some of those will be drivers unlikely to be affected by speed limits, for example escaping criminals and joyriders.

The real dangers of speed

We have seen how travelling at high speed may not cause danger, and we have seen that sometimes the places where lowest speeds are used are the most dangerous. (for example, in town) The general fact that relates these together in practice is that a good driver will vary his speed to control the degree of exposure to risk.

The real danger from speed can be usefully expressed as a driver failing to slow down when necessary. Obviously we should be going fastest when conditions are clearest. And the vast majority of drivers in the vast majority of circumstances manage this quite well the vast majority of the time.

So how might we seek to influence crash risk with speed policy? Obviously setting speed limits has two extremely beneficial effects:

  • Speed limits provide essential guidance to inexperienced drivers about local hazard density.
  • Speed limits often enable the police to prosecute people using speed dangerously, without the requirement for complex evidence about standards of driving.
But rigid enforcement will tend most strongly to affect free travelling speeds not crash speeds. We have a page with specific proposals about how to use the speed limits to improve road safety. The emphasis should always be on safe speeds, not speed limits.

We could express road danger from a typical accident as kinetic energy times crash risk. Where the crash risk is very low we can safely allow high speeds and hence large kinetic energies. Where the crash risk is highest we use the lowest speeds, sometimes slowing to a crawl. This is what good driving is all about.

We need drivers who slow down where the danger is greatest, not where the danger is smallest. This is where emphasis on speed limits misses the point, such emphasis tends to affect areas where the danger is least. On the other hand publicity about using speed wisely and safely would emphasis reducing speed where danger is greatest.


Comments on the above are welcome. If there is a demand we will create a comments page. We will be delighted to publish all suitable emails including those whose content we disagree with. Email comment.

We have a strict editorial policy regarding factual content. If any fact anywhere on this web site can be shown to be incorrect we promise to remove it or correct it as soon as possible.
You can't measure safe driving in miles per hour.