Driver response is key to the success of any crash avoidance feature. If drivers see and respond to features and don't compensate by taking more risks behind the wheel, crashes can be reduced. But if motorists respond by changing how they drive, potential benefits may not pan out.
Drivers don't change their behavior in response to many safety features but do respond to those that give direct and immediate feedback — for example, when acceleration is boosted or brakes are improved (see "Risk compensation theory pops up where it's completely irrelevant," Oct. 13, 2007). This could be the case with studded snow tires, which have delivered a less-than-expected benefit, possibly because drivers who think they won't skid go too fast for conditions.
"Some motorists will respond to emerging crash avoidance features the same way," says Anne McCartt, Institute senior vice president for research. "They'll drive faster or pay less attention to the possibility of a hazard ahead if they think a gadget will alert them if needed."
Previous experience indicates that more automatic features — those that apply brakes automatically to avoid a collision, for example — will be the most effective. Yet many of the features being introduced on current cars give motorists feedback by activating a warning light. Then drivers have to take appropriate action to avoid crashes. Such features might not live up to their hype.
Of course, benefits will be disappointing if drivers simply turn off crash avoidance features that are annoying or don't seem to help. This could be the case with blind spot detection systems, which could activate continuously in heavy urban traffic.
"It's hard to predict," McCartt points out. "Automakers are designing so as to limit how much a feature may annoy drivers, but people's reactions are hard to foresee."
When antilock brakes were introduced on passenger vehicles, expectations were high for crash reductions, based on how cars with antilocks performed in tests by the Institute and others. But the initial outcome on the road was to increase crashes, and antilock never have produced a large safety benefit (see "Antilocks may not make the difference that many expected," Jan. 29, 1994). The problem could be that the tests involved skilled drivers in controlled settings while real-world driving is fraught with unforeseen circumstances for motorists whose skills may be limited. Drivers also might compensate for antilocks. Assuming they can stop quicker if they need to, they drive faster and/or delay brake application.
Brake light example
Experimental fleets equipped with center high-mounted brake lights were associated with crash reductions in the 1980s (see "Extra brake light reduces rear-end collisions," May 13, 1981), presumably because drivers took note of the new feature. Yet studies conducted after the lights became ubiquitous showed lesser effects, possibly because the novelty wore off.
The best way to assess the effectiveness of any new feature is to equip lots of vehicles with it and then evaluate its on-the-road performance compared with similar vehicles without it. Even this can be confounded if vehicles are equipped with multiple features at the same time, as is the case with the crash avoidance features being introduced now. Concurrent introduction makes it hard to distinguish the individual effects of each feature.