Street geometry sets driving speed and the sign only asks

Everything here earned its place by changing an outcome. Nothing about design speed is included to round the number up.

Drivers read the road surface, not the signpost

Speed choice is largely automatic and is driven by how much visual information arrives per second and how much margin for error the road appears to offer. A wide lane with long forward visibility and nothing close to the kerb reads as safe at high speed, whatever the sign says. This is why rural-standard geometry inside a built-up area produces rural-standard speeds.

Engineers call the speed a road invites its design speed, and where design speed and posted speed diverge, drivers follow the design.

Lane width is the single strongest lever

Narrower running lanes reduce free-flow speeds because the driver has less lateral tolerance and compensates by slowing. Many urban carriageways were built to widths inherited from highway standards intended for uninterrupted rural running.

Reclaiming that width for footway, cycle track or planting slows traffic and buys space at the same time. The counterargument that narrow lanes cause sideswipes is weak at urban speeds, though it holds on high-speed roads.

Edge friction does the work that markings cannot

Objects near the carriageway edge — trees, parked cars, buildings at the back of footway, planters — compress the visual field and reduce speed. The same road with a wide grass verge and no vertical features will be driven considerably faster. This is why suburban distributor roads with generous verges consistently overrun their limits.

Over a decade, adding vertical elements at the edge is usually cheaper than reconstructing the carriageway.

Forward visibility is a double-edged variable

Long sight lines are treated as a safety asset because they give drivers time to react. They also raise the speed at which drivers feel comfortable, which can cancel the benefit at the junction where the sight line ends. Deliberately restricting forward visibility on approach to a hazard is an established technique and sits uncomfortably with older design guidance.

Measured properly, the resolution is to match visibility to the intended speed rather than maximising it everywhere.

Enforcement is the expensive substitute

Cameras and patrols reduce speeds where they operate and much less between locations, so coverage has to be continuous to be effective. That makes enforcement a recurring revenue cost against a one-off capital cost for geometry. Political attention concentrates on cameras because they are visible and controversial, while the geometry decision is invisible.

A self-enforcing street continues working when the enforcement budget is cut.

Comparisons across cities are difficult because nobody counts the same things.

Limit changes without redesign underperform

Where cities have lowered urban limits without altering the street, measured speed reductions have generally been modest. Even a modest reduction matters because collision severity rises steeply with impact speed, so the intervention is not worthless. The mistake is treating the sign as the intervention and then concluding that limits do not work when the road was never changed.

Over a decade, programmes that pair the new limit with narrowing, planting and junction tightening report substantially larger effects.

Everything above, in order of what to do first

1. Drivers read the road surface, not the signpost. Speed choice is largely automatic and is driven by how much visual information arrives per second and how much margin for error the road appears to offer.
2. Lane width is the single strongest lever. Narrower running lanes reduce free-flow speeds because the driver has less lateral tolerance and compensates by slowing.
3. Edge friction does the work that markings cannot. Objects near the carriageway edge — trees, parked cars, buildings at the back of footway, planters — compress the visual field and reduce speed.
4. Forward visibility is a double-edged variable. Long sight lines are treated as a safety asset because they give drivers time to react.
5. Enforcement is the expensive substitute. Cameras and patrols reduce speeds where they operate and much less between locations, so coverage has to be continuous to be effective.
6. Limit changes without redesign underperform. Where cities have lowered urban limits without altering the street, measured speed reductions have generally been modest.

The takeaway

If you want a slower street, change what the driver sees, not what the sign says.

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