Curb cut effect

The name comes from a piece of civil disobedience. In Berkeley in the early 1970s, disabled activists, many of them wheelchair users from the campus independent living movement, took sledgehammers and cement to the kerbs the city would not cut, and the ramps that followed became the first municipal programme of its kind in the United States. What happened next is the part the term is actually about: the ramps were used constantly by people who had never been part of the argument, pushing prams, dragging suitcases, wheeling delivery trolleys, riding skateboards, walking with a cane. The count was never close. On any given kerb the intended user is a rounding error against everyone else rolling something.

Digital interfaces repeat the pattern so reliably that it is worth treating as a prediction rather than a happy accident. Captions were built for deaf and hard of hearing viewers and are now switched on by people watching in bed, in a gym, in an open-plan office, or in their second language. Voice control was built for people who cannot use their hands and became how everyone talks to a phone in a car. Autocomplete reduces the number of keystrokes a person with limited fine motor control has to make, and it is also why nobody types a full address any more. High contrast modes were built for low vision and get used outdoors in the sun. The constraint that forces a clearer design is the same constraint for everybody; the disabled user just hits it first, and hardest.

This matters most as an argument, because accessibility work is usually killed by the phrase “edge case”. The framing smuggles in a false model of the market: that the population needing the feature is both tiny and separate from the paying one. Neither half survives contact with the numbers. Disability is not a fixed minority but a state most people pass through, by age, injury, illness, or a badly timed hand full of shopping, and the feature almost never stays in its lane. A team that ships captions to satisfy a regulator ends up shipping a searchable transcript, better SEO, and a video that works with the sound off in a feed, which is where most video is watched.

Two cautions keep the term honest. It is an observation, not a justification: the ramp would be worth building for the people it was built for even if nobody else ever used it, and an argument that only ever cites the mainstream benefit has quietly conceded that disabled users alone are not reason enough. And it does not run in reverse. Designing for the majority and hoping it happens to work for a wheelchair user produces a kerb, not a ramp. The effect only appears when the constraint drove the design.