The Roof

Of all the prices this volume has examined, none is felt so keenly or so universally as the price of a place to live. The rent of land has been, until now, something of an abstraction — the landlord growing rich in his sleep, the differential between the good acre and the marginal, the increment the community confers on ground it never owned. This plate makes it concrete, and brings it home, for the rent of land is the largest single force in the cost of the one thing every household must somehow buy: a roof. In the cities where the work is, a home has become the defining expense and the defining anxiety of ordinary life — the thing a young family cannot afford, the reason a worker turns down the better job in the better place, the asset on which the fortunate float and from which the rest are shut out. The roof has grown dear, and the question this plate asks is why; and the answer, from first to last, is rent.¹

To see it, begin by taking a house apart into the two things its price is paid for: the building, and the ground it stands on. The building is a manufactured thing, made of materials that cost much the same everywhere — timber and concrete and glass and wiring, and the labour to assemble them — so that the cost of putting up a given house in a cheap city is not very different from the cost of putting up the same house in an expensive one. The ground is another matter entirely.

Across most of America, a home sells for something close to what it costs to build — the structure, plus a modest and ordinary price for the land beneath it — and where that is so, housing is not dear. But in a limited number of places, almost all of them on the coasts, a home sells for far more than it costs to build, two and three and four times more, and the whole of that difference is the price of the land. The expensive home is not expensive because it is a better building; it is, very often, a worse one. It is expensive because of the ground it sits on — because of where it is.² And this raises the question on which everything turns: land near the heart of a thriving city is genuinely scarce, for there is only so much of it; but is it as scarce as its price implies?

It is not — or it need not be. For the scarcity of housing is not, in the main, a scarcity of land; it is a scarcity of permission to build. A single acre near a city centre could hold one house, or a row of townhouses, or a tower of two hundred flats — the same ground, multiplied many times over as a place to live, if only the building were allowed to rise. What forbids it is zoning: the dense body of law that dictates what may be built where — the rules that reserve the land for detached single houses alone, that cap the height and the density, that mandate minimum lot sizes and setbacks and parking spaces, that turn each proposal to add homes into a permission to be fought for and, more often than not, denied. Across the United States, roughly three-quarters of the residential land is zoned for single-family houses only, which is to say that it is illegal, on most of the land in most American cities, to build the very thing those cities most lack. The scarcity that makes the land so dear is, to a great degree, manufactured — written into law — exactly as the scarcity behind the taxi medallion and the occupational licence was manufactured in an earlier plate. The roof is dear because the law has made the right to raise it scarce.³

How much of the land’s value is manufactured in this way, and how much is the irreducible premium of a genuinely desirable spot, is a real and contested question. When Glaeser and Gyourko measure the gap between price and building cost and call it a regulatory tax, critics answer that the gap is simply the location value of land — the natural premium of a good address — and the artefact of no regulation at all. There is justice on both sides, and the volume’s own categories dissolve the quarrel: the price of a central location is a rent either way, part of it the natural scarcity of a place there is only one of, part of it the manufactured scarcity of a density the law forbids. The natural part no reform can abolish, though a tax can capture it; the manufactured part a change in the law can dissolve. And the proof that the manufactured part is large is the simplest kind there is: where a city has lifted the limits and let the building rise, the homes have come.⁴

Whoever holds a scarcity collects its rent, and here the holder is the existing owner. When the law forbids new building, the homes that already stand are made scarcer and so dearer, and their owners grow richer — not by any act of their own, not by improving anything, but because a community grew up around them and the law forbade anyone from adding to the supply. It is the unearned increment of an earlier plate in its purest residential form: a windfall conferred on the incumbent by the growth of others and the prohibition of competitors. The right to build, where it is granted, becomes a valuable thing in itself, an entitlement bought and sold; and the value of the prohibition, everywhere it binds, is folded into the price of every existing home and held there as wealth. The homeowner — the most sympathetic and least suspected of figures — has become a rentier, and the home is the rent.⁵

The diagram holds the whole of it. Where the supply of housing can answer demand, a growing city is housed at a roughly steady price, the arrivals absorbed as new construction. Where the supply is fixed by law, that same growing demand has nowhere to go but into the price, which rises without natural limit while the number of homes stands still. The prosperity that should have drawn people in and built up to receive them is captured instead, in its entirety, by those who got there first.

Why, then, does the scarcity persist, when so many are harmed by it and the remedy is known? Because the people who profit from it govern the body that sets it. The economist William Fischel gave the answer its name: the homevoter. A house is, for most families who own one, by far their largest asset, and an undiversifiable one — their savings are not spread across a hundred holdings but sunk into a single property on a single street. That concentration makes the homeowner intensely, and quite rationally, protective of anything that might bear on the value of the home; and the thing that bears on it most — the building of new homes nearby, which would ease the very scarcity that holds the value up — is decided locally, in the zoning board and the city council and the neighbourhood hearing, the arenas the homeowner attends and effectively controls. So the homeowner votes, reliably and successfully, to keep new housing out: to protect the value, and to forestall the traffic and the change and the loss of “character” that the new building is felt to bring.⁶

The result is a closed loop of remarkable stability. At the hearing where the new homes are debated, the people with a direct and quantifiable stake — the owners whose property values turn on the outcome — are present, organised, and opposed; the people who would gain — the renters priced out, the young who would move in, the workers who would take the better jobs — are diffuse, unorganised, and mostly not in the room, many of them not yet even residents of the town whose law excludes them. It is the transitional gains trap once more, wearing its most sympathetic face. The rent has been capitalised into the wealth of ordinary families, who did not design the system, who bought in good faith, and who would be genuinely harmed by its sudden unwinding — and who therefore defend it with the full and understandable ferocity of people protecting their life’s savings. And so the scarcity renews itself, election after election and lawsuit after lawsuit, against nearly every attempt to relieve it.⁷

The cost of all this is far larger than dear rent, though dear rent is cost enough. The deepest harm is that the closed city closes off opportunity itself. The cities that restrict housing most fiercely are, and not by accident, the most productive — the places where wages are highest because the work is most valuable, where a worker of ordinary skill earns more than she could anywhere else, by virtue of who and what surrounds her. By making housing unaffordable, these cities turn away the very workers who would come to share in that productivity, who are left instead in the places they can afford, doing less than they might. Labour is misallocated across the whole country — held back from where it would be most productive — and the economy is smaller for it. How much smaller is genuinely uncertain, and honesty requires saying so: the most influential estimate, by Chang-Tai Hsieh and Enrico Moretti, put the loss very high, holding that housing constraints in a few productive cities had lowered American growth substantially across half a century — but their figures were later found to contain errors, were revised, and have been challenged by others who find the true effect far smaller. The defensible statement is that the direction is not in doubt and the magnitude is: the closed city costs the country output, perhaps a great deal of it, though no one can yet say how much with confidence.⁸

And the loss is not only in output. The scarcity divides a society into those who own and those who rent — the former growing wealthier in their sleep while the latter surrender an ever-larger share of their income for shelter — a steady transfer from the young to the old, from the poor to the propertied, from the newcomer to the long-established. It pushes those who cannot afford the centre out toward the edges and beyond, into long commutes and sprawling development, with all their costs in time and traffic and carbon. And because the single-family neighbourhoods the law most carefully protects are, on the whole, the wealthier and the whiter, the scarcity inscribes and perpetuates a segregation by class and by race that long outlives the explicit exclusions it replaced. The closed city is dearer, smaller, more unequal, more sprawling, and more divided than the open one would be.⁹

What, then, is to be done — and here the volume’s threads draw together, for the remedy is one it has already supplied, in two forms that turn out to be the same medicine. The first is the tax of the previous plate. A tax on the value of land captures the unearned rent the scarcity confers — takes the windfall back from the incumbent and returns it to the public that created it — and does so, as we saw, without distorting anything at all; and it does something more, peculiarly apt here, for by making it costly to hold valuable urban land in an idle or underused state, it presses the owner to build on it, or to sell to someone who will. The second is to dissolve the scarcity at its source: to liberalise the zoning, to let the building rise, to legalise the homes the city forbids. These are not rival remedies but complements — the tax captures the rent, the upzoning removes the scarcity that creates it — and each makes the other work the better, for a tax that penalises idle land and a law that permits dense building together turn the underused lot into the new apartments.¹⁰

The case for the cure must be made honestly, and its evidence is itself a caution. When the city of Auckland upzoned three-quarters of its residential land at a stroke, the homes came — construction surged, and rents rose markedly less than they otherwise would have — which is the clearest proof we possess that the manufactured scarcity is real and that lifting it works. But the American reforms of recent years tell a more sobering story: the cities and states that have pared back single-family zoning — Minneapolis, Oregon, much of California — have so far seen only modest new building, held down by high interest rates and construction costs, by reforms too narrowly targeted to bite, by local resistance and litigation, and by the plain fact that such changes take the better part of a decade to show their full effect. The lesson is that scale and patience are decisive: large and sweeping reform works where piecemeal tinkering does not, and even the large reform pays out slowly. The remedy is sound; it is neither quick nor easy.¹¹

The politics follow from the homevoter. Because the beneficiaries of the scarcity control the local government that sets it, the relief has increasingly had to come from above — from state legislatures preempting local zoning, overriding the city council in the name of the wider public that the council does not represent. It is an awkward and contested instrument, this overriding of local control, and it provokes the predictable resistance and the predictable lawsuits; but it follows from the structure of the trap, for a body captured by its incumbents will not free itself. There is even a paradox of scale in the politics, the mirror of the one in the economics: the sweeping reform may be easier to pass than the narrow one, because singling out a few neighbourhoods unites them in opposition, while changing everything at once spreads the burden and the benefit alike. The path is hard, and slow, and uncertain. But its direction is clear, and it is the volume’s entire: capture the rent, and dissolve the scarcity that breeds it.¹²