#251 2004 · New York City Department of Transportation · Urban infrastructure / public administration
New York quietly disconnected most of its crosswalk buttons decades ago and never told anyone, because the button was doing its job either way
the problem
Removing an obsolete control that people rely on feeling in charge of costs money and provokes anxiety, but leaving a working control in place costs more
background
New York City installed pedestrian push-buttons at crosswalks starting in the 1960s, when traffic was lighter and a major street's signal would stay green by default until a pedestrian pressed a button to request a walk phase — a genuinely functional control at the time. As the city's traffic grew denser through the following decades, the Department of Transportation shifted most intersections to centrally timed, automated signal patterns that no longer needed or responded to individual button presses, but physically removing thousands of now-nonfunctional buttons across the city was never budgeted as a priority.
By the early 2000s the city had quietly deactivated the large majority of its buttons without public announcement, leaving them mounted at crosswalks, still pressable, giving no indication they no longer did anything — a fact city transportation officials confirmed only when asked directly, rather than one the city had proactively disclosed.
what everyone would do
Either physically remove the now-disconnected buttons, or clearly relabel them as inactive. Both cost something the city didn't want to pay: removal ran an estimated $400-500 per button, well over a million dollars citywide, while openly disclosing the deactivation would have destroyed the one benefit the buttons were still quietly providing.
what they saw
The situation revealed that the button's real value had shifted away from its original mechanical function toward something else entirely — the psychological effect of giving pedestrians a sense of agency over their own wait. A person who believes they've done something to speed up a signal waits more patiently than someone who feels powerless, so the button kept doing useful work even after it stopped being connected to anything, just not the work it was originally installed to do.
the move
The New York Times reported in 2004 that more than 2,500 of the city's roughly 3,250 pedestrian crosswalk buttons functioned as pure mechanical placebos — pressable, but disconnected from the signal timing. Rather than remove the disconnected buttons, at an estimated $400–500 per button (well over a million dollars city-wide), the Department of Transportation left them in place: pedestrians who pressed them still felt they had done something to speed their crossing, a documented psychological effect — researcher Ellen Langer's 'illusion of control' — that keeps people engaged with the act of waiting rather than jaywalking out of impatience.
why it works
A pedestrian who presses a button believes they've influenced the signal, regardless of whether it's actually wired to anything, and that belief measurably changes behavior, making people more willing to stand and wait rather than cross against the signal out of impatience. Because this effect doesn't depend on the button actually functioning, leaving thousands of disconnected buttons in place kept producing the same calming, compliance-inducing effect as when they worked, at zero ongoing cost. Removing them would have eliminated that behavioral benefit while costing over a million dollars to execute, so leaving them silently in place was simultaneously the cheapest option and the one that preserved a real effect — doing nothing outperformed both of the obvious active choices.
the payoff
By the mid-2010s only around 100 of the roughly 1,000 remaining buttons citywide were still estimated to be functional, with the rest left permanently in place; the city avoided the multi-million-dollar removal cost while the placebo buttons continued to serve their behavioral function of keeping pedestrians focused on the crosswalk while they waited, regardless of whether a press changed the signal.
where it breaks
The effect depends on the illusion staying at least partially uncontested — once the public learns en masse that most buttons are disconnected, as happened when The New York Times reported it in 2004, some of the psychological benefit erodes, since believing in a control requires believing it actually works. It also depends on the placebo genuinely producing a measurable behavioral benefit that offsets the cost of leaving it in place; a control whose placebo effect is negligible, or one that misleads users about something safety-critical rather than merely psychological comfort, risks causing real harm rather than helpfully calming a crowd. And it requires that maintaining the illusion doesn't itself become a reputational liability once discovered — an organization caught deliberately leaving a known-broken control undisclosed risks a trust cost that can outweigh the original savings, which is exactly the tension this case's own press coverage captured.
what came after
NYC's crosswalk buttons became the most widely cited real-world example of a 'placebo button,' a term now used across design and behavioral-psychology writing (99% Invisible, CNN, Popular Science) for any control retained purely for its psychological effect after its mechanical function was removed — elevator door-close buttons and many office thermostats are cited alongside it as members of the same category.
references
- [1]User Illusion: Everyday 'Placebo Buttons' Create Semblance of Control99% Invisible, 202299percentinvisible.org
- [2]Cities 101: Do Crosswalk Buttons in NYC Actually Work?Untapped New York, 2013untappedcities.com