Why actuated signals are bad for pedestrians
1. Actuated signals communicate that cars are more important than people.
By almost constantly giving vehicles in one direction or the other a green light but giving pedestrians a walk signal only after a button press and a delay, the actuated signal embodies an institutional belief that the passage of vehicles is more important than the passage of pedestrians. This makes pedestrians feel like intruders on the street and contributes to drivers resenting the presence of any pedestrians who might slow them down. Only streets that provide some other extremely strong attraction can survive this implicit message that people are unwelcome.
2. Actuated signals make walking much slower.
With simple timed signals, delay to pedestrians from signals is minimal. Consider a typical street system (such as the north-south streets of downtown San Francisco) with intersections every 320 feet. Walking a mile involves crossing 16 streets. With optimal timing (each light beginning its walk signal 64 seconds after the previous block's signal) there might be no delay for pedestrians from the signals, so the total time to go a mile would be the 17:36 of actual walking time at 5 feet per second. More typically (with each light on a total 60-second cycle and generous 8-second clearance interval, but not timed to match walking speed), pedestrians will experience no delay 37% of the time and an average of 19 seconds of delay the other 63% of the time, for an overall average delay of 12 seconds per intersection, 192 seconds in total, contributing to a walking time of 20:48 for the mile.
With actuated signals, on the other hand, pedestrians never have a walk signal waiting when they arrive at a corner. If their walking speed exactly matches the green lights, they actually get the worst possible delay, 64 seconds at each intersection, waiting the entire length of a cycle for their walk signal, 1024 seconds waiting, making the total mile take a horrifying 34:40 to walk. More typically, they are unsynchronized to the cycle length, so their delay is an average of half the 60-second cycle at each intersection, 480 seconds in all, making the mile take 25:36. This is a 23% time penalty over timed signals in the typical case and a 97% penalty in the worst case.
3. Actuated signals ignore human factors research for user interfaces.
Actuated traffic signals are a user interface and, if they are to provide a satisfying experience, must follow the same rules of human factors as any other user interface. The standard rule of thumb is that interface components must signal feedback within 100 msec, should complete operations within 1 sec, and must provide a progress indicator for anything that takes longer than 10 sec or the user will assume that the program has crashed and will never finish. The typical actuated signal installation provides no feedback whatsoever until the light changes 30 seconds or more later, with the entirely predictable result that people will bang on the button repeatedly trying to determine if it is broken, and then give up and cross against the light before it eventually changes. (And in many cases the sensor really is broken and the walk signal will never come.) Newer installations often do have an indicator light to provide the 100 msec feedback, but it is still extremely rare to provide a progress indicator, let alone one that gives a plausible explanation for the length of the delay.
The human factors for bicycle actuation are even worse, because these never give feedback that the signal has been actuated, and the bicycle sensors are even more likely to be missing or broken than pedestrian buttons are.
4. Actuated signals deprive pedestrians and drivers of important information.
Because the human factors of signals lead people to believe that the walk signal will never come and the delay from waiting is so great, people frequently give up and cross against the light, especially if there is a green light for vehicles in the same direction. Because the pedestrian signal still shows "don't walk" as people cross illegally, drivers are more likely to turn without looking and hit the pedestrians, and pedestrians are deprived of the knowledge of why their signal might legitimately and importantly say "don't walk," such as a left turn arrow for vehicles in the opposite direction.