Street lighting is aimed at the road when people need to see faces

What follows is an argument about street lighting, and about where the received version of it stops being true.

Horizontal and vertical light are different quantities

Road lighting is specified largely as illuminance or luminance on the carriageway, because that is what a driver needs to see obstacles. A pedestrian needs to recognise a face at a distance, which depends on light falling on a vertical plane at head height.

A street can fully satisfy carriageway standards and leave faces in shadow, and many do. Guidance for pedestrian environments increasingly specifies vertical illuminance separately for exactly this reason.

Uniformity beats brightness

A sequence of bright pools separated by dark gaps forces the eye to keep readapting, and dark adaptation takes far longer than light adaptation. The result is that the dark sections are perceived as darker than their measured level suggests. Lower average light levels with high uniformity generally produce better perceived safety and better recognition distance than high average levels with poor uniformity.

On the ground, this argues for more, lower, closer-spaced luminaires rather than fewer bright ones on tall columns.

LED conversion changed more than the bill

Replacing sodium lamps with LED cut energy use substantially and changed the light from orange to broadly white. White light improves colour rendering, which helps recognition of people, clothing and hazards.

Measured properly, it also has a higher share of short-wavelength output, which raised legitimate questions about ecological impact and sleep disruption. Warmer LED colour temperatures address much of that concern and are now common in residential specifications.

Glare undermines the whole exercise

A luminaire visible directly in the field of view creates disability glare, which reduces contrast and effectively blinds the viewer to what lies beyond it. Flat-glass optics and proper cut-off control the effect, and cheap retrofits frequently do not. Older residents and anyone with cataracts are affected far more strongly, which makes glare an accessibility issue rather than a comfort one.

A street that measures well and glares badly will still feel unsafe.

Lighting and crime is a qualified story

Reviews have generally found improved lighting associated with reductions in crime, including in daytime, which points at community confidence effects as well as visibility. The evidence base is uneven, effect sizes vary widely, and displacement is difficult to measure. What is better established is that lighting affects whether people choose to walk somewhere after dark.

That behavioural effect matters independently of whether it changes the crime statistics.

Local statute governs most of this, and it varies street by street.

Darkness has value too

Excess light spills into bedrooms, disrupts nocturnal wildlife and washes out the sky, and all three are avoidable through optics rather than through switching off. Part-night dimming and adaptive controls reduce output when streets are empty and raise it when movement is detected. Cutting lighting entirely to save money has been tried in several places and reversed under public pressure more often than not.

The design question is where light goes, not just how much of it there is.

The takeaway

Judge a lit street by whether you can recognise someone coming toward you, not by how bright the tarmac looks.

The design decision is visible long after the people who made it have gone.