Urban Design in New Cities

Order that grows

Jane Jacobs opened the modern argument in 1961 by pointing out that the qualities planners were busy demolishing, density, mixed use, short blocks and old buildings standing next to new ones, were exactly the ones that made a street safe and interesting. Christopher Alexander gave the insight a structure four years later with the observation that a city is not a tree: real settlements are overlapping webs of relationships, and a plan that sorts them into a tidy hierarchy destroys what it set out to organise. His later distinction between generative order, which unfolds step by step under rules, and imposed order, fixed as an image in advance, remains the sharpest framing of the whole debate.

James C. Scott extended it in Seeing Like a State, reading Brasília and Chandigarh as attempts to make a city legible to the authorities at the cost of the local knowledge people use to live in one. Brasília is the case that will not go away. Its pilot plan was sized for around half a million residents and has never exceeded that, while the surrounding Federal District passed two and a half million. The growth went into unplanned settlements the plan had no room for, one of which was known as Cidade Livre, the Free City.

Somebody still has to lay out the streets

Not everything can be left to emerge, though. Streets are the layer that cannot be changed afterwards: buildings turn over every few decades, but the right-of-way and the pattern of plots decide what can be built there for a century. Drainage, sewerage, road hierarchy and public transport are the goods a collection of individual developers reliably undersupplies, because no one of them captures the benefit of getting them right.

The distinction that resolves this is between regulating form and regulating use. The New Urbanists turned Jacobs’ observations into an instrument, the form-based code, which governs the shape of the street and the building line and leaves what happens inside the building alone. That is a far lighter rule than the use-based zoning standard in America since 1926, and it is the version of planning most compatible with a market. Alain Bertaud’s Order Without Design supplies the economist’s case: a city is first of all a labour market, and mobility and affordability are the two things a plan must not break. Tokyo is the working example, where a national framework permits housing almost everywhere and municipalities cannot write themselves out of it.

What the new cities are testing

The Free Cities projects are running two experiments worth watching. Próspera lets a builder choose among recognised international building codes instead of waiting for a local one to be drafted, and issues approvals in weeks. That turns the building code from something imposed into something selected, which is the same move the movement makes with law in general.

Ciudad Morazán near Choloma is the opposite of a showpiece, and the more instructive for it. A compact walkable district of some forty-five hectares built for Honduran factory workers, it sets street widths against how many people there actually own cars rather than against an imported standard, keeps plots small enough for a small trader to take on, and maximises the leasable land so the cost of infrastructure per household stays low. Its designers describe the public role as supplying the initial public goods and then leaving room for the place to grow.

Underneath both lies the same economics. Land value is the scoreboard, since good urbanism shows up as a price premium, and that gives a developer who owns the whole site an incentive an ordinary builder never has. Beauty behaves the same way: no single builder can capture the return on a coherent street, which is why supplying it takes either a covenant, a code or a single landowner. That is the real subject of this topic. Not what to build, but who is entitled to decide.