Is risk a known, calculable entity, or a subjective assessment? Reading Brian Hunt’s excellent post on calculating risk got me thinking more about this question—one that’s of central interest and importance to anyone who forges ahead into unknown territory. Entrepreneurs are risking status and cash; physical adventurers may be risking health or life. But each does some kind of risk-benefit analysis … and most would say that they’re not taking irresponsible risks, not matter where they land on the risk continuum.
A clear answer to the question probably doesn’t exist. But Risk, a book by John Adams of University College London, has some fascinating points and case studies on the subject. More on and from his writing later, but here are a few points he makes that help explain why risk management is so difficult.
First, Adams agrees with Brian (although he doesn’t put it that way) that risk assessment is highly subjective. We all have different thermostat-triggers of what we consider “too risky” and why. But is it possible for a third, objective party to assess whether an activity is risky or not? Insurance companies and public safety entities make those calculations every day—but even accident statistics aren’t straightforward, according to Adams.
One reason, Adams says, is that if we perceive an activity to have risk attached to it, we compensate with our behavior to keep ourselves safe. If, for example, a road is considered dangerous because of the traffic, we’ll be more careful crossing it, and keep our kids away from it. So accident statistic purveyors will look at that road and see a low number of accidents. The road is, statistically, a low risk. But in fact, it’s very high risk.
The same holds true in the other direction, when it comes to safety equipment. We tend to drive faster and take more risks when we have air bags and seat belts, because we feel as if they will keep us safe. Adams argues that while seat belts unquestionably help protect car occupants from injury if they are in a crash, they may increase the risk of accidents in general, because of compensatory changes in driving habits. He quotes reviews of the impact of seat belt laws, including one by the British Department of Transport in 1981 that noted “available data for eight western European countries which introduced a seat belt law between 1973 and 1976 suggests that it has not led to a detectable change in road deaths.”
Interestingly enough, after Britain passed a seat belt law in 1983, he says there was an 8-13 percent increase in injuries to cyclists and pedestrians. So the risk was transferred to those outside the cars, now driven faster … until those cyclists and pedestrians also compensated for the added risk by removing themselves from the now higher-risk places.
Which is all just to say that assessing risk, let along managing it, is a very complex task. More from Adams later. But it seems that risk, while calculated every day, may be as difficult to quantify as an “objective” true value by a third party as by an individual.
Food for thought.
Lane Wallace is the Editor and Founder of No Map. No Guide. No Limits.
Next post: Do People See Entrepreneurs as Cowboys?
Previous post: Calculating Risk
Essential reading on risk: Nassim Taleb’s books “Fooled by Randomness” and “The Black Swan.” If you’re only going to read one, the first one is IMNSHO the better one.
“Is risk a known, calculable entity, or a subjective assessment?” I could be flippant and answer yes. And, like quantum complementarity, there would be truth in this reply.
Actuaries, bookmakers, and futures traders all perform objective measurements of risk as part of their business models. As you note, individuals also calculate some risk-benefit analysis. My speculation is that, since the individual calculus is by nature subjective, it is subject to unconscious manipulation.
I think humans crave security, but temper the results with complacency. In the 1980’s, federal regulators concluded that drivers were tuning out the taillights of cars in front of them and triggering needless rear-end accidents. The result was the center high-mounted stop light, intended to combat that. Now, ask yourself how often you are conscious of the CHMSL-equipped car ahead of you?
Automotive technology has advanced to the point that backup sensors are common. But slavish reliance poses risks. What if you slip behind the wheel of a non-sensor fitted car and hear silence as you reverse. Were you aware that you needed to clear behind you before backing up, or did you listen for a train of beeps and, hearing none, move the car? And what of new automotive systems that will automatically apply braking or keep pace with traffic ahead? Do we become so accustomed to the new technology that we revise our risk assessment to become lulled into such a sense of security that we do not do not observe traffic around us and cannot react in time if the system fails?
Similarly, has commercial aviation become so automated that too many of the pilot’s responsibilities are delegated to the aircraft? Does the pilot remain on board primarily to handle takeoffs, landings, and emergencies? I know this question is an exaggeration, but Airbus aircraft, in particular, rely upon a hierarchy of fly-by-wire flight control “laws” under which software maintains control and surrenders it to the crew in stages as the situation changes. If the crew has been relieved of mundane responsibilities, must they be better prepared for extreme eventualities? Could flight conditions under which an A330 threw in the towel and left its fate to a startled and task-saturated crew have resulted in the loss of Air France 447 off the coast of Brazil?
It’s said that if you treat people like idiots, they will behave that way. Hans Monderman of the Netherlands credited the judgment of drivers and removed traffic signs, lights, and markings from selected urban streets. He thereby manipulated the typical risk calculation of drivers. With their comfort level jarred, drivers had to revise their risk assessment to confront a new driving environment and their place within it.
Is risk so hard for individuals to calculate or manage? I don’t think so. It just seems the results get corrupted along the way.
And I toghuht I was the sensible one. Thanks for setting me straight.