Behavior Based Traffic Safety

Note: This article originally appeared in Vol. 1(3) of Behavior Analysis Quarterly.

By Todd A. Ward, PhD, BCBA-D

According to the National Conference of State Legislatures (NCSL), traffic fatalities are the leading cause of death in the U.S. for those aged 3-34 (Teigen, Shinkle, & Essex, February, 2005). Speeding accounts for approx. 10,000 annual fatalities, while running red lights accounts for 750 deaths and 260,000 injuries (NCSL, February 27, 2015; Teigen, Shinkle, & Essex, February, 2005).

In an effort to make our roads safer, an increasing number of law enforcement agencies have embraced the technology known as automated enforcement. From a law enforcement perspective, the attractiveness of the program is based on efficiency of resources—automation implies an agency can do more with less (NCSL, February 27, 2015). In the U.S., more than 400 communities use some form of automated enforcement to enforce red light laws, while over 40 use the system to enforce speeding laws (NCSL, February 27, 2015).

The largest study to-date comes from a multi-year project in Washington D.C., funded by the Insurance Institute for Highway Safety (IIHS News, September 1, 2015). The IIHS President was quoted by IIHS News as saying “We’re all accustomed to seeing posted limits ignored, but it’s a mistake to think nothing can be done about it. Automated enforcement is one of the tools we have at our disposal.” If every county in the U.S. adopted automated enforcement, the IIHS estimated “more than 21,000 fatal or incapacitating injuries would have been prevented in 2013.”

The study began in Montgomery County in 2007, with the introduction of speed-detecting cameras in select locations. After only six months, the county witnessed a significant reduction in speeding and, after seven years, individuals speeding by more than 10 mph decreased by 59% in comparison to control roads with no cameras. During the same seven year period, the probability that a crash would produce a fatality or incapacitating injury fell by 19%.

In 2012, the IIHS took their program a step further with the introduction of speed corridors, or long stretches of road targeted with cameras rather than isolated sections. The IIHS noted “the cameras are regularly moved to different locations on those roads so drivers don’t become familiar with their exact locations.” With the introduction of corridors, the probability of fatality or serious injury dropped 30%, in addition to the 19% reported previously. What’s more are IIHS reports of a “spillover effect” wherein significant speeding reductions are observed on non-targeted roadways.

State law varies widely on how municipalities can use traffic cameras, the key component of automated enforcement. For example, the District of Columbia authorized the use of cameras for the enforcement of “all moving infractions” (Teigen, Shinkle, & Essex, February, 2005, p. 52), while Texas permits cameras for the enforcement of red light laws only, and West Virginia prohibits any use of photo enforcement (Teigen, Shinkle, & Essex, February, 2005).

However, Ohio presents an interesting case that taps into the incentives to use or discard automated enforcement. According to The Vindicator, the state recently passed a law that permits camera-based enforcement as long as an officer is physically present, similar to a more traditional speed gun (Kovac, September 9, 2015). Law enforcement agencies oppose the law for its requirement that an officer be present, which they say removes any incentive to use the cameras in the first place. As mentioned before, the primary incentive from an agency perspective is that automated systems allow for the more efficient use of organizational resources. As a result, many Ohio communities have ceased camera-based enforcement all together (Kovac, September 9, 2015).

Automated enforcement programs also have ancillary benefits for the communities in which they reside. For one, the potential for a significant revenue boost to the city is very likely. For example, Edmonton Canada anticipates an increase in photo-radar enforcement revenue from $30 million in 2014 to $47.8 million in 2015, all from traffic violations (Jones, September 7, 2015). The Washington Post reports that Chicago tops the list of U.S. cities in terms of revenue gained from enforcement cameras at over $90 million per year, followed by New York with $41 million per year (Crunched, May 1, 2015).

Additionally, the slower one drives, the less fuel is used, which means less pollution and more money in citizens’ pockets. For instance, Halsey (September, 1, 2015) noted that a Toyota Camry gets 40 miles per gallon when traveling at 55 miles per hour. Increasing the speed to 60 mph decreases gas mileage to 35 mpg, while increasing to 75 mph drops the mileage even further, to 30 mpg.

But the program isn’t without critics. Twelve state legislatures have explicitly prohibited the use of enforcement cameras, and a recent Washington Post poll suggested 40% of respondents opposed the idea—more so for people who lived in regions with cameras already in place (Halsey, September 1, 2015). At the city level, Cleveland recently voted to discontinue the use of cameras, which prompted Xerox to file a lawsuit against the city for breach of contract. Xerox signed a three-year deal to supply the city with cameras, which was terminated early due to voter responses (FoxNews, September 10, 2015). Ohio citizens complained that camera-based enforcement “were little more than a money grab” given that approx. 35% of revenue from fines went to a company that helps run the program. The same citizens alleged that the camera systems are not without error, resulting in citations going to the wrong people, and a lack of appeals processes suitable for the program (Kovac, September 9, 2015).

The story of automated enforcement presents an interesting challenge to behavior analysts. Data suggests the systems are effective at increasing the safety of our roadways, and the behavioral processes involved seem fairly straightforward. Under the traditional methods of traffic enforcement, the probability that any given driver would receive a ticket was relatively low. A ticket was contingent upon first contacting a police officer on the roadway, and many times drivers can see the police vehicle in the distance or infer the presence of a speed trap when other drivers in the vicinity reduce their speed. Granted, most people have likely received a speeding ticket in the past—some more than others—but the tickets are typically few and far between. In fact, the high dollar amounts coming into city budgets as a result of automatic enforcement is a testament to how many people do not obey traffic rules.

I have received a ticket from an automated enforcement program for running a red light. The ticket itself even had a link to a video where I could see myself running the light in question. Even though the ticket didn’t come in the mail for a few weeks after the infraction, my driving behavior has sensitized to cameras mounted at intersections. If I were consulting to a municipality on how to further improve the effectiveness of automated enforcement technologies I would recommend that the systems embrace the modern era of smartphones. Instead of using outdated “snail mail” technology to deliver tickets, the delayed contingency between the traffic violation and the receipt of a citation could be reduced to near a near instantaneous consequence if systems were in place to email or text an “e-citation” moments after the infraction.

But the ends don’t always justify the means. For cities like Cleveland that experience a large backlash against automated enforcement policies, those in power might find comfort in knowing that Applied Behavior Analysis has a sizable literature on interventions to increase driver safety, for a fraction of the cost of camera systems. For example, Van Houten, Nau, and Marini (1981) installed highway signs that provided daily or weekly feedback on the percentage of drivers who were driving at appropriate speeds. The researchers found that daily and weekly postings were equally as effective at reducing speeds. However, the effects of the signs vanished when the signs showed no numerical feedback. In a similar study, Van Houten, and Nau (1981) found that feedback effectively reduced speeding on urban highways, and was ten times more effective at reducing speed than were police surveillance and ticketing.

Van Houten’s effects have been replicated by Iclanders Ragnarsson and Bjorgvinsson (1991). The latter group posted signs along roads entering residential areas. As in the previous studies, the signs contained feedback on the percentage of drivers traveling at appropriate speeds. The team also created a condition in which a sign containing an ideal speed followed the feedback sign. Results mirrored those of Van Houten, in that the feedback signs significantly reduced speeding, with the follow-up sign producing much smaller speed reductions.

But the studies on highway feedback only scratch the surface of how behavior analysts have successfully intervened on traffic safety as a whole. The point is that evidence-based solutions to traffic enforcement exist. Behavior analysts have been demonstrating their successes for over three decades. When automated enforcement systems evoke resistance from the public, other options are available that, on the surface at least, appear benign.


Email Subscription

Leave a comment

Your email address will not be published.