Connectivity, not distance, decides whether people walk

The points below about street network connectivity are ordered by how much difference they make, not by how often they get repeated.

Straight-line distance hides the walk

Catchment maps drawn as circles assume a network that lets you walk toward your destination, which many suburban layouts do not. The useful measure is route directness, sometimes called the detour ratio: actual walking distance divided by straight-line distance. Values close to one describe a grid; values well above one and a half describe a layout where walking has been made irrational.

A shop visible across a fence and twelve minutes away by road will be driven to.

Cul-de-sac networks concentrate everything

Branching layouts with few connections force all trips onto a small number of distributor roads regardless of length or purpose. That produces distributor roads carrying local traffic at speed, which makes them unpleasant to walk along and hard to cross.

The residential streets are genuinely quiet, which is the design intention and the reason the pattern was popular. The trade is quiet streets purchased with a hostile network everywhere else.

Block size is the underlying variable

Smaller blocks give more route choices, shorter detours and more frontage per unit of area. They also produce more junctions, which is a cost in vehicle delay and a benefit in vehicle speed. Traditional gridded districts and medieval cores both deliver small blocks by different means and behave similarly for walking.

Large superblocks can work for pedestrians only if they are crossed by dedicated routes, which requires deliberate design and long-term maintenance.

Filtered permeability separates the networks

A modal filter — bollards, a planter, a gate — closes a street to through motor traffic while leaving it open to walking and cycling. This produces a dense pedestrian network overlaid on a sparse driving one, which is the pattern most Dutch and Danish cities use. It is cheap, reversible and reliably contentious, because it redistributes convenience between neighbours.

Over a decade, monitoring in schemes of this type has generally found traffic reductions on filtered streets with smaller and more variable effects on boundary roads.

Severance is the other half of the problem

Railways, rivers, motorways and large parking areas cut networks whatever the local street pattern looks like. A single crossing point can add a kilometre to trips for everyone on the wrong side of it, and the effect persists for the life of the barrier. Retrofitting crossings is expensive, which is why severance created decades ago still shapes travel patterns today.

Assessing new infrastructure for the walking routes it will sever is not standard practice in most appraisal systems.

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

Measuring it is straightforward and rarely done

Route directness can be calculated from open street data for any area with a few lines of code. So can intersection density, block perimeter and the number of routes crossing a given barrier.

These metrics predict walking rates better than land use variables alone in most studies that have compared them. The main barrier to using them is that no one owns responsibility for the walking network as a network.

Everything above, in order of what to do first

1. Straight-line distance hides the walk. Catchment maps drawn as circles assume a network that lets you walk toward your destination, which many suburban layouts do not.
2. Cul-de-sac networks concentrate everything. Branching layouts with few connections force all trips onto a small number of distributor roads regardless of length or purpose.
3. Block size is the underlying variable. Smaller blocks give more route choices, shorter detours and more frontage per unit of area.
4. Filtered permeability separates the networks. A modal filter — bollards, a planter, a gate — closes a street to through motor traffic while leaving it open to walking and cycling.
5. Severance is the other half of the problem. Railways, rivers, motorways and large parking areas cut networks whatever the local street pattern looks like.
6. Measuring it is straightforward and rarely done. Route directness can be calculated from open street data for any area with a few lines of code.

The takeaway

Measure the walk, not the distance. The gap between them is the design.

Cities are built by a thousand small permissions, not one big plan.