New York rates four common options.
New York experienced major traffic safety improvements during the Bloomberg era, largely as a result of new street designs that offered a model for other U.S. metros to follow. The city hopes to build on that success during the de Blasio era, with the new mayor pursuing a "Vision Zero" plan to eliminate traffic deaths. That plan relies heavily on better enforcement of traffic laws, and despite a shaky start, there's hope it can become a governance model, too.
With so much focus on large-scale design and enforcement measures, it's important to remember that often, the quickest and most direct way a city can improve safety at an intersection is with traffic signals. Traffic lights are neither as ubiquitous as they might seem (even in New York, only a quarter of all intersections have four-way signals) nor as simple. New research in the journal Transport Policy evaluates four common light structures found across the city: basic signal installation, increased pedestrian cycles, the Barnes dance, and split-phase timing.
A quick note on the scoring system. The researchers first identified intersections across New York using each type of signal. They then compiled statistics on pedestrian and vehicle crashes for five years before the signal had been installed, and two years afterward — giving them a percent change in crash occurrence. Last, they found comparable intersections where signals had not been installed and gathered the same stats over the same period, to serve as a measure of the status quo.
As usual, we've charted the big results to satisfy your infographic fix (all figures via Chen et al, Transport Policy 32, 2014, 69-78):
Increased pedestrian cycle. These signals extend the time a pedestrian has to cross the street. This might seem like a clear win, but it means a longer green signal for traffic in the other direction, and thus the potential for impatient pedestrians to cross against a signal or for impatient drivers to make up time by speeding. Here, compared to intersections without a change, the researchers found a clear improvement in pedestrian incidents (50 versus 4 percent decline) but only a modest change in vehicle crashes (45 versus 37 percent decline).
The Barnes Dance. A Barnes Dance (aka pedestrian scramble) stops all cars and lets people cross every which way — even diagonally. Its benefits are clear but there are downsides: people have to wait longer to cross and get less time to cross, and cars are backed up considerably, which can lead to increased congestion across a network. The findings here were also mixed; pedestrian incidents dropped big time (51 percent versus 9 percent) but vehicle crashes actually rose (10 percent) even as they went down (12 percent) at comparable intersections.
Split-phase timing. These signals involve a third phase of the traffic cycle that stops all turns so pedestrians can cross. Walkers are well-protected with these signals, but they have less time to cross in a given cycle, since they only get the extra protection during the turn phase. The researchers found split-phase timing quite effective: there were larger declines in pedestrian incidents (39 versus 8 percent) as well as multi-vehicle crashes (56 versus 44 percent).
Signal installation. This is your basic traffic light. It offers far more order than a basic four-way stop, but it can also invite faster traffic, with cars seeing a green from afar and hitting the gas. The pros clearly outweighed the cons in this study: compared to intersections without a change, pedestrian incidents increased at a much slower rate when lights were installed (12 versus 67 percent), and vehicle crashes decreased at a much faster rate (49 versus 14 percent).
So what do we learn here? Well, all the signal measures help pedestrians, but in two cases the overall safety record is mixed. Increased cycle length didn't do much for vehicle crashes; it might be best near senior centers or for older populations most in need of extra walking time. The Barnes Dance, meanwhile, showed the odd potential to increase car crashes and might be most effective if heavy traffic flows are diverted around it.
The best-performing traffic signals were the split-phase and simple signal installation. Split-phase lights require a protected turn lane (an extra infrastructure expense) and might be most effective on narrow streets that can be crossed quickly. Perhaps the most bang for your policy buck comes from the good-old four-way traffic light. As urban traffic safety plans rightly become more sophisticated, cities should still remember the basics.