What is risk compensation?

What is risk compensation?

The dark side of safety

From the introduction of the modern-day, three-point seatbelt in the 60s, to the standardisation of airbags during the 80s – vehicle safety has drastically improved over the course of the past half-century. However, although falling steadily since 1966, the number of road accidents in the UK remain high. In 2014 alone, England was the scene of 131,6571 crashes – of which, tragically, 1,386 were fatal. If vehicles are now safer than ever why do we still witness such a high number of accidents?

We recently researched2 the behaviour of UK motorists’ which yielded some unsettling insights. While eight in ten UK motorists consider themselves a safe driver, just under half believe errors made by drivers are the leading cause of road accidents. What’s more, a quarter deem poor driver behaviour or inexperience to have the second largest impact on accident figures. On the flip side – only 2% of people blame vehicle defects. The results clearly suggest that UK motorists acknowledge the fact that most accident are caused by drivers – rather than their vehicle or the competence of its safety features.

In his review of one of the safest vehicles on the market – the Independent’s Jamie Merril wrote, “I did, though, discover one worrying fault: me.”3 Merril expanded on this point and concluded his article with the following statement, “[vehicles] are getting more clever with each new model. We are not.” Merril – similarly to other UK motorists – identifies himself as the main flaw in terms of road safety.

But state-of-the-art vehicles, and more specifically their safety features, can also play an unexpected role in road accidents.

What is risk compensation?

Let’s take American football and rugby as examples. While die-hard fans will probably tell you they’re completely different – these two sports share several, similar characteristics. They’re both extremely physical sports which involve bone-crunching ‘downs’ or ‘tackles’ and flying ‘touchdowns’ or ‘tries’. Essentially, they both involve significant potential for physical harm. Yet a crucial difference is the fact that in American football, players wear a considerable amount of protective gear, while rugby players only a mouth guard. Taking this into account, when asked, people would probably say that rugby is the more dangerous of the two.

However studies4 have shown that, as humans, the level of risk we take is proportionate to how safe we feel in that specific moment. Consequently, as a direct result of wearing protective gear, American football players actually suffer a higher number and more severe injuries than rugby players5. This is due to the fact that, not only do they have the means to be able to deal more damage to opposing players when tackling, but more importantly, they’re more likely to involve themselves in heavy, high speed collisions due to feeling more protected.

This is known as risk compensation.

Risky business

In order to better understand this phenomenon, we spoke to Dr. Gerald Wilde of Queens University, Ontario. In his paper, ‘Target Risk’, Dr. Wilde proposes a theory he names risk homeostasis. Boiled down, the theory suggests that in any given situation people have a level of risk with which they’re comfortable– called ‘target level of risk’.

target level of risk

According to Dr. Wilde, when people are confronted with a situation in which their target level of risk is exceeded, they will act more cautiously until the danger they perceive is once again within their comfort zone. On the other hand – when people find themselves in a situation where risk is below their threshold, they will again change their behaviour. However in this situation, they will act in a riskier way to once again balance the scales and reach their target level of risk.

“The target level of risk is that level of risk where the net benefit of action is perceived as maximal”

Dr. Wilde explains that “people arrive at that target level of risk not on the basis of explicit numerical calculation, but on the basis of general knowledge, intuition and affect.” It can be said then, that the level of risk people perceive is based on their past experience of that particular situation and the belief they have in their driving and decision-making skills. This is influenced by many factors, but none more than age. Our survey revealed three-quarters of UK motorists think that 17 to 24 is the age group most likely to be involved in accidents – a figure which only supports the risk compensation theory. Having driven for the fewest number of years, not only are 17 to 24 year olds likely to have the least amount of past experience and skill, but they’re also the age group most likely to assess a situation as less risky than it really is.

Ultimately, according to Dr. Wilde, “the actual actions taken determine the accident rate and the accident rate in turn determines the level of perceived risk in a closed-loop feedback process that we call homeostatic. Hence, the target level of risk is the ultimate controlling variable in accident causation.”

Although this notion – and the theory in general – may seem a little unorthodox, numerous experiment have been carried out and an enormous amount of data has been collected to support it.

An example of risk homeostasis is what happened in Iceland and Sweden in the late 60s, when the countries altered from left-hand drive to right-hand drive. Although many predicted catastrophic consequences and a dramatic rise in accident rates per head, these in fact promptly dropped following the switch. However, within two years in Sweden and around ten weeks in Iceland, the figures returned to previous levels. In relation to the risk homeostasis theory, this was due to the fact that at first motorists overestimated the level of risk the change-over would bring. This resulted in perceived levels of risk being extremely high and caused motorists to be particularly cautious. Over time, motorists from both countries realised the situation wasn’t as dangerous as they originally thought. As a result, the perceived level of risk declined and once again aligned with the target level of risk, resulting in accident figures returning to normal. 

Ignorance is bliss

Yet in some situations – risk homeostasis doesn’t seem to apply. Where safety implementations are “invisible” – such as reinforced doors for example – driver behaviour may not be affected. Most drivers would barely notice the increased safety these features provide and, consequently, this will have a minor effect, if any, on their driving behaviour. On the other hand, if the safety alteration is visible, such as transitioning to a new car which brags lane assist, people will feel safer and consequently tend to act in a riskier manner.

Not all safety improvements increase our safety

How effective a new safety feature will be is almost impossible to predict. Dr. Wilde proposes that “such developments may well bring about a change in the particular pathways in which accidents happen, but not on their overall frequency per head of population. This is because drivers will adapt to these technological changes. Study after study shows that individuals drive more daringly after they have been supplied with safety features.”

But this is not to say that safety features lose all their value. Dr. Wilde also mentions that “there can be no question that the technological improvements over more than a century have led to major reductions in the accident rate per unit distance driven.”

It could be that the driver is in fact safer, but the danger has merely shifted elsewhere. To truly understand the repercussions of risk compensation, the safety of other road users – such as pedestrians and cyclists – should be taken into account.

Data based on most recent report. Updated 2015 data due for release September 2016.
1www.gov.uk/government/statistics/reported-road-casualties-great-britain-annual-report-2014
2An exclusive survey of 1,094 British drivers conducted online by YouGov for Aviva in conjunction with the Telegraph on 7-9 December 2015
3www.independent.co.uk/life-style/motoring/road-tests/volvo-xc90-t6-motoring-review-theres-one-danger-even-this-safety-first-car-cant-remove-10462793.html 
4riskhomeostasis.org 
5Knowledge for healthy living – Keeping their heads out of the game 

Back to top