Zoning envelope architecture

Why do so many buildings in one city taper in the same peculiar way?

Stand in midtown Manhattan and look up at a dozen towers from the 1920s and 1930s. Different architects, different clients, different budgets, different decorative styles — and all of them stepping inward as they rise, in the same tiered, ziggurat profile. Then walk the City of London and notice how several recent towers are sliced off along one sloping face. Then look at photographs of Tokyo, where whole streets of small buildings have their upper corners shaved at the same angle.

Fashion is the obvious explanation, and it is a bad one. Fashion does not hold steady for forty years, and it does not stop at a municipal boundary. Yet in each of these cities the shared shape stops, cleanly, at the city line. What sort of cause behaves like that?

Start with the New York case, because the record is unusually clear. In 1915 the Equitable Building went up at 120 Broadway: some forty storeys rising sheer from the property line with no setback at all, and a shadow that darkened its neighbours for much of the day. The following year New York adopted the first comprehensive zoning resolution in the United States, and the reaction to that building is written all over it.

Now the part that matters. The 1916 resolution did not say "buildings shall be stepped", or prescribe a shape at all. It defined an invisible sloping surface — a sky exposure plane — rising from the far side of the street at an angle fixed by how wide that street was, and required the building to stay beneath it. Above a certain point, a tower could rise without limit, but only on a quarter of the lot.

So the code specified a volume you may not exceed, and said nothing whatever about what to build inside it. Ask yourself what happens next. Who fills that volume, and with what aim?

A developer has bought the land, and what is sold afterwards is floor area. So the objective is straightforward: maximise rentable floor area, subject to staying under the plane. And an optimum against a binding constraint has an unhelpfully predictable form — you press against the constraint everywhere it will let you. Which means the building's silhouette becomes a tracing of the invisible surface: sheer to the height where the plane bites, then stepping back, again and again, then a slender tower on the permitted quarter.

Do you see why the buildings match? Not because the architects agreed, and not because anyone drew that shape into law. Every one of them solved the same maximisation against the same constraint, and the constraint had essentially one good answer. The shared profile is the shadow of the rule, not a style.

There is a vivid piece of evidence for this. In 1922 the delineator Hugh Ferriss, working with the architect Harvey Wiley Corbett, drew a sequence of charcoal studies showing simply the maximum mass the new law permitted on a large lot — no client, no program, no windows, no purpose. They are not designs. Yet they look uncannily like the buildings that were actually put up over the following fifteen years, because the envelope was already doing most of the designing.

Which suggests a test. If the shape follows the constraint surface, then changing the surface should change the shape. New York rewrote its resolution in 1961, dropping the setback plane as the governing device in favour of floor area ratio — a cap on total floor area expressed as a multiple of the lot's area — plus a bonus: build a public plaza at ground level and you may have extra floor area above. New objective, new optimum. Instead of a wedding cake filling its lot, the profitable answer became a plain slab set well back behind an open forecourt, and that is precisely what got built through the 1960s and 1970s.

Notice too what the bonus rewarded, and what it did not. It paid for the provision of plaza area, not for the plaza being a good place to be, so the city acquired a stock of windswept granite aprons nobody wanted to sit in. Later amendments added design standards, which is an admission of the same principle from the other direction: the shape you get is the shape the incentive priced.

The natural misreading is that the code dictates the architecture. It does not, and the difference is the whole point. A code that specified forms would have to anticipate every site; instead these codes specify a surface and a price, and let a thousand private maximisers discover the form. That is why the results are so uniform without anyone intending uniformity, and why they change abruptly and citywide when the arithmetic changes.

Two honest qualifications. First, the envelope is not the only convergent pressure. Elevator cores, standard structural bays, and the depth of floorplate at which daylight still reaches a desk all push towards similar plans anywhere in the world; the envelope explains the silhouette far better than it explains the plan. Second, the arrow sometimes runs the other way. The Seagram Building of 1958 stood back voluntarily behind a plaza, giving up buildable area for an effect its owner wanted — and is usually credited with inspiring the plaza bonus that later made the gesture profitable for everyone else. A famous exception can become the next rule.

The same reasoning reads other cities on sight. London protects sightlines of St Paul's Cathedral from designated viewpoints, and a tower that intrudes on one must be cut away along that plane, which is why the Leadenhall Building leans back along a single sloping face. Japanese rules limiting how far a building may cast shade and how it must taper from the road produce the shaved upper corners visible on ordinary streets everywhere. In each case the peculiar shape is a rule made visible.

• How transferable development rights let an unbuilt envelope be sold to another site, and what that does to the resulting form.
• What incentive zoning has bought elsewhere — arcades, theatres, affordable housing — and how often it bought the letter rather than the spirit.

Every term the collection defines is gathered in the glossary.