About 10 years ago, a quaint little neighborhood in Auckland, New Zealand, had a serious problem: "rat-runners." These were drivers who, stymied by heavy traffic on main thoroughfares, were turning down residential roads—and egregiously ignoring the posted speed limits. As a result, this stretch of churches, homes and shops had almost twice the number of crashes as any other part of the city.

Instead of rolling out more police patrol cars or slapping down speed bumps, Auckland officials tried a different tack: Working alongside traffic psychologists from the University of Waikato in New Zealand, they set about changing the entire feel of the neighborhood. They narrowed the smallest local roads by adding parklike spaces that might include a basketball court, community garden or trees, effectively moving drivers’ focus closer to the front of their cars. Larger collector roads got more delineated pedestrian crossings and cycle lanes, while obstructions to drivers’ forward views were removed to indicate it was acceptable to move a little faster on those roads. Arterial roads, meant to carry heavy and swiftly moving traffic, stayed the same.

Traffic psychologyFour months after construction was completed, the number of crashes fell by 30 percent, vehicular traffic decreased by 30 percent on the modified roads and pedestrian traffic was up 17 percent. Five years later, crashes were down by half, says study leader Samuel Charlton, PhD, whose work at the University of Waikato applies cognitive psychology to driving behavior. And at no point did the researchers change the speed limit.

"We did that not by changing people’s perceptions of speed while driving and not by changing people’s attitudes toward the speeds they choose explicitly," Charlton says. "We did this on an implicit, unconscious level."

Changing roads and vehicles, not people, is a hot topic in traffic psychology these days.

Speeding, in particular, is a persistent problem for road safety. The Auckland experiment is one example of the creative, sometimes subtle, ways in which researchers are seeking to slow down drivers—and save lives. And because there is no one cause of speeding, these approaches go far beyond traffic cops to encompass everything from public service announcements to in-vehicle technology.

Speeding kills

According to the U.S. National Highway Traffic Safety Administration (NHTSA), speed has been a factor in about one-third of the nation’s motor vehicle fatalities over the past 20 years. And though driving has become safer overall in that period, the short-term trends are alarming. The NHTSA’s preliminary 2016 numbers suggest that traffic deaths rose in both 2015 and 2016. (See figures below.)

While some dangerous behaviors, including drunk driving or failure to wear seat belts, have declined precipitously in the face of public health campaigns in the United States and other developed nations, speeding has not.

"Many drivers consider low-level speeding to be safe and normal," says Bernice Plant, PhD, an assistant lecturer at the School of Psychological Sciences at Monash University in Melbourne, Australia. Cultural messages around speeding are generally positive, Plant adds—even the children’s film "Cars" glorifies the need for speed.

Speeding can also be chalked up to optimism bias: Drivers who speed are convinced that they won’t be the ones to spin out of control if they step on the gas pedal, Plant says.

Research suggests that men and young drivers are particularly prone to optimism bias when comparing their own driving to that of their peers. Consistently across studies, young drivers also speed more frequently than older drivers, though such factors as having peers in the car make speeding more likely for youth.

Research suggests that people determine how fast to drive by using cues from the road and neighborhood. Alter the environment by adding small parks and bike lanes, and people will drive more slowly.

SpeedingPart of what makes speeding complicated to understand and prevent, Charlton says, is that it’s a continuous decision, made moment to moment throughout a drive. It’s not always a conscious decision, either, he says. People drive by habit and on "autopilot." In one study, Charlton and his colleagues stopped drivers and asked them their speeds after clandestinely clocking them with a speed gun. They learned that people often had little idea of how fast they were going and even fainter notions of how fast they were supposed to be going (Transportation Research Part F: Traffic Psychology and Behavior, Vol. 30, 2015). Drivers’ reports of their usual speeds were the best predictors of how fast they were clocked driving on the road, the researchers found, even better than the drivers’ estimates of how risky the road was to drive.

Road work

Given that people aren’t assessing risk on the road logically, the newest anti-speeding research is focused on the subtle.

One widely used strategy is to put up electronic speed-feedback signs that alert drivers to how fast they’re going. While they could also glean that data by glancing at their own dashboards, the external feedback has been shown to jolt people out of their habitual patterns, Charlton says.

Engineering the environment is another way to nudge people toward slower speeds, as the Auckland experiment shows. The researchers’ studies in both driving simulators and on real roads reveal that people decide how fast to drive using cues from both the road itself and its surroundings. In a 2017 study on how drivers gauge the appropriate speed for urban roads, a research team found that roadside factors such as landscaping, houses and sidewalks cued drivers to speeds of around 30 mph (50 kph), while wider roads with painted lines prompted them to judge the appropriate speed at about 50 mph (80 kph) (Accident Analysis & Prevention, Vol. 108, 2017). In a driving simulator, people drove a little slower than what they thought the speed limit should be on the slow roads, but they tended to drive a little over the assumed speed limit on the fast-moving roads. On rural roads, one of Charlton’s studies found, people drive slower when lanes are narrower, center lines are double yellow or wide, or when traffic is heavy (Accident Analysis & Prevention, Vol. 95, Part A, 2016).

Just posting a speed limit won’t slow people down if the road "feels" fast, Charlton says. In fact, he says, mismatches between the actual speed limit and what people think the limit should be are a major barrier to getting people to drive in the right range.

"Particularly roads designed in the 1950s and ’60s and ’70s—they’re big, they’re wide…and it just feels like you should go fast," he says. In New Zealand, there has been some effort to consider the look and feel of roads during the design phase, Charlton says, but there is still a strong emphasis on just the number of cars a road is expected to support, not how those cars will use the road. "Breaking away from just thinking about road volume and thinking about road functions is something we’re coming to grips with," he says.

Smart cars

The other half of the driving environment is the cars themselves. Vehicles have become smoother and quieter over the past two decades, says Emanuel Robinson, PhD, a human factors psychologist in transportation and safety at Westat, a research organization in Rockville, Maryland. Speeds that would have left passengers with their teeth rattling 20 years ago are comfortable today.

At the same time, the onboard technology of today’s vehicles can easily be used to stymie speeding, Robinson says. He and his colleague Doreen De Leonardis, PhD, along with other Westat researchers, tested voluntary, in-vehicle anti-speeding devices with a group of chronic speeders (Transportation Research Record: Journal of the Transportation Research Board, Vol. 2425, 2014). In tandem with Maryland’s Motor Vehicle Administration, the researchers identified drivers who had received at least three speeding tickets (issued by a police officer) in the past three years. From that pool, the researchers identified 101 people who agreed to have a unit installed in their cars that would alert them if they exceeded the posted speed limit by more than 8 mph (13 kph).

Participants were monitored without speed warnings for a two-week baseline period, then put through a four-week treatment phase in which speeding would trigger an electronic voice that warned, "speeding violation." After the treatment phase, notifications went silent and monitoring continued for two additional weeks.

That electronic voice had a significant effect, the researchers found. The frequency of speeding up to 8 mph over the posted limit (say, going 68 in a 60 mph zone) increased during the treatment phase compared with the baseline and follow-up, but the frequency of speeding over 8 mph (exceeding 68 in that 60 mph zone) decreased dramatically. The slow-down effect seemed to linger over the two-week follow-up, Robinson says, gradually weakening as drivers rediscovered their lead feet.

It would be technologically simple, he says, to integrate such technology into modern-day vehicles; indeed, many companies, including ORBCOMM and Lytx, offer devices to track truckers’ and other professional drivers’ speeds and other on-road activities, and some insurance providers use monitoring to reward drivers for safe practices. The question, Robinson says, is whether standard in-car speed alerts could be presented in a way that is appealing to consumers.

Public campaigns

"Appealing" is the watchword for one final thread of the anti-speeding effort: direct messaging. Research by Plant published in the journal Accident Analysis & Prevention found that ads viewed while off the road can make people slow down on the road, at least in the short term.

Plant and her colleagues compared the efficacy of two anti-speeding public ­service announcements (PSAs) from the Roads and Traffic Authority of New South Wales, Australia (Accident Analysis & Prevention, Vol. 100, 2017). The first, a 2004 ad called "Heaven and Hell," starts like a standard car commercial, all macho speed and quick cuts, and ends with the driver smashing into another vehicle. The second ad, called "Pinkie," aired in 2007 and focuses on the social side of speeding: As young men rev their engines, women trade scornful glances and outstretch their pinkie fingers, implying that the drivers are overcompensating for certain deficiencies.

After viewing one of these ads as well as an emotion-matched, non-speeding-related control ad, young drivers answered questions about how they thought the ads would influence their driving speeds, and then each took a spin in a driving simulator.

The death and destruction in "Heaven and Hell" didn’t turn out to be particularly motivational. Drivers were no slower after watching that ad than when they watched an emotion-matched PSA on another topic. But "Pinkie" was effective: Participants who had watched the sexually suggestive ad drove less speedily than those who’d viewed an unrelated, emotion-matched PSA. (Unfortunately, neither ad resulted in slower driving when participants were tested a week to 10 days after viewing.)

Interestingly, Plant says, the participants were terrible at predicting which kind of message would motivate them to slow down. Only 30 percent thought "Pinkie" would do the trick, while 70 percent predicted the carnage in "Heaven and Hell" would motivate them.

The mismatch between what drivers thought would motivate them and what actually did could be a quirk of how the researchers asked their study questions, Plant says, or it could mean people have poor insight or there is some missing link between their intention to change and the execution of said change. Either way, the finding suggests that the way researchers assess interventions might need some tweaking.

More immediately, Plant adds, the findings suggest that ads depicting social consequences for bad behavior might be more effective for young drivers than ads depicting death or injury. "Once we know the ‘how’ and the ‘why’ behind the persuasive effects of messages or interventions more generally, we can ensure that we develop effective interventions," she says.

One exciting development on the horizon in traffic research, Plant says, is similar to the technology that Robinson and his colleagues tested to slow people down. Onboard GPS devices that can transmit real-world information about speed are likely to return much more reliable data than people’s self-reports or even data from driving simulators. Studies using such devices could lead to the development of interventions that prompt people to drive more safely on real roads.

"Observing reductions to speeding in these studies would have the most promising implications for on-road driving speeds," Plant says.

Deadly toll
  • 9,723: Number of U.S. deaths in 2015 in which speed was directly responsible.
  • 70%: Percentage of drivers who admit to speeding at least some of the time.
  • 5.6%: Increase in crash fatalities from 2015 to 2016.
  • 4%: Increase in fatalities due to speed-related crashes from 2015 to 2016, from 9,723 to 10,111.

Source: U.S. National Highway Traffic Safety Administration www.nhtsa.gov/risky-driving/speeding

Further reading

Virtual Testing of Speed Reduction Schemes on Urban Collector Roads
Domenichini, L., et al.
Accident Analysis & Prevention, 2018

Using Endemic Road Features to Create Self-Explaining Roads and Reduce Vehicle SpeedsCharlton, S.G., et al.
Accident Analysis & Prevention, 2010

2011 National Survey of Speeding Attitudes and Behaviors
Schroeder, P., et al.
U.S. National Highway Traffic Safety Administration, 2013