The last plate mapped the gradient of urban rent — the price sloping down from the center — but it took the center itself for granted, and never asked why there should be a center at all, why anyone should pay a premium to be near one particular spot of ground rather than another. This plate asks. The answer is agglomeration: the cluster of forces that make being near others valuable — and, stranger still, that make the value pool, and keep pooling, where it has already pooled. The plate takes up the economies of agglomeration through Marshall, who first named them; through the long quarrel of specialization against diversity; and through the new geography of Krugman and the cumulative causation of Myrdal, which showed how activity, once gathered, gathers further. And it finds that the locational rent of the last plate is, at the root, the capitalized value of the agglomeration — the price of admission to a productive crowd that everyone in it makes, and the owner of the ground alone collects upon.
Begin, once again, with Marshall, who in 1890 asked why the firms of a trade so often huddle together in a single district — the cutlers of Sheffield, the cotton-spinners of Lancashire — when they might have spread themselves freely across the land. He found three reasons, three economies that flow not from the size of any one firm but from the size of the cluster, and that reward each firm for sitting near the rest. The first is the pooled labour market: where many firms of a trade gather, a deep pool of skilled workers gathers with them, and each firm finds the hands it needs, and each worker the work he seeks, far more readily than either could in isolation. The second is the specialized supplier: a cluster grown large enough can support the makers of specialized tools and the providers of specialized services that no single firm could keep in business alone, and all of them share in the benefit. And the third, the subtlest and the most powerful, is the spillover of knowledge — for in a crowded district, Marshall wrote in a phrase that has never been improved upon, the mysteries of the trade cease to be mysteries and are “as it were in the air,” passing from firm to firm and hand to hand by proximity alone, so that ideas breed the faster where the men who carry them are packed the closer.
The mysteries of the trade become no mysteries, but are, as Marshall wrote, as it were in the air.
And here the agglomeration explains the rent. For if being near the cluster confers these advantages — the pool of labour, the web of suppliers, the knowledge in the air — then a place within the cluster is worth more than a place outside it, and worth paying for; and what one pays for it is rent. The locational rent of the last plate, the premium upon the central spot, is at bottom the capitalized value of the agglomeration: the price of admission to the labour pool, the supplier web, the air thick with the secrets of the trade. The downtown commands its rent not for any virtue of the ground itself but for the company the ground keeps — for the firms and workers and ideas already gathered upon it, access to which is precisely what the high rent buys. The preceding plate described the gradient of price; the agglomeration is why the gradient exists at all — why nearness should be worth a farthing. Rent, in the city, is the price of proximity to a productive crowd.
But what kind of nearness pays? Here a quarrel runs the length of the subject. Marshall's economies were economies of specialization — the gains of clustering with firms in the same trade, the cutlers all together in one town. But Jane Jacobs, gazing at the great cities a lifetime later, saw a different and perhaps a deeper source of the agglomeration's power: diversity. The city's genius, she argued, lay not in the massing of a single industry but in the collision of many — in the cross-fertilization by which an idea borrowed from one trade sparks an invention in another, in the ceaseless recombination of unlike knowledge that only a various city affords. The specialized district makes an existing industry efficient; the diverse city, Jacobs held, breeds wholly new ones, by mixing what had never been mixed before. The two are not flatly opposed — a great economy wants both, the deep specialized cluster and the various generative metropolis — but they name two different engines of the gathering, and which matters more, and for what end, is a question the subject has argued over ever since and not yet closed.
And then the deeper question still: not what nearness gives, but why activity should heap itself in one place to begin with, rather than spreading evenly across a uniform plain. The answer came, in formal dress, from Paul Krugman, whose new economic geography showed how concentration emerges from the interplay of two forces — the increasing returns that reward producing at scale in a single place, and the cost of carrying goods across distance. Firms enjoying increasing returns wish to concentrate their making; and they concentrate it where the market is largest — but the market is largest exactly where the firms, and so the workers who are the market, already are. Demand draws the firms, and the firms draw the workers who are the demand, in a circle that feeds upon itself. From this circular causation, Krugman showed, a featureless plain will spontaneously divide into a crowded core and an emptied periphery — the agglomeration arising not from any prior virtue of the favoured spot but from the self-reinforcing logic of increasing returns and distance alone. The cluster, once it has begun, calls more into itself, and so grows.
This circle has a momentous consequence: the value pools, and keeps pooling, where it has already pooled. For if concentration breeds concentration — if a place that pulls ahead thereby draws to itself the very activity that pulls it further ahead — then small early advantages compound into large and lasting ones, and where the great clusters sit comes to be settled, in no small part, by mere historical accident, frozen afterward into place by the increasing returns that followed. Why is the technology of a nation gathered into a single valley, its finance into a single square mile, its films into a single town? Not always because those spots were best, but because the activity began there, a little, and then could not stop gathering. Gunnar Myrdal called this cumulative causation, and saw in it the engine of regional divergence — the reason rich regions grow richer while poor ones are drained, the favoured place feeding upon the periphery it impoverishes. The agglomeration that makes the cluster productive also makes it self-perpetuating; and it makes the map of prosperity a map of where prosperity happened, long ago, to begin.
The agglomeration that makes the cluster productive also makes it self-perpetuating — and the map of prosperity a map of where prosperity happened, long ago, to begin.
Three things must be weighed before the agglomeration is let stand as an unmixed good. The first is that it is, in great part, a good: the cluster is genuinely more productive than the scattering, the city a real engine of growth and of invention, and the economies of agglomeration a positive externality that lifts every firm and worker within reach — this is why the modern world crowds into cities, and why cities, for all their squalor and their strain, remain the places where wealth and ideas are made. But, second, the gain is neither free nor without limit: density brings congestion and filth and the thousand frictions of the crowd — agglomeration diseconomies that mount as the cluster swells, and that set, in principle, a ceiling on how large it ought to grow, though where that ceiling lies is far from settled, and many a city is held well below it not by nature but by the laws against building that the last plate decried. And third — the volume's own sting — the surplus that the agglomeration creates is captured, in great measure, by the owners of the land beneath the cluster, as rent: for the gathering is made by the crowd, by every firm and every worker who joins it, yet the value of their gathering settles upon the ground they gather on, and is gathered in by whoever holds its title. The agglomeration is the purest case in all this volume of a value created by the many and harvested by the few — the crowd makes the cluster, and the landlord banks it.
Return, then, to the volume's question, which the agglomeration sharpens to a fine point. The last plate found that the rent of the city is a price of nearness; this plate finds why nearness is worth the paying — because the crowd is productive, because the cluster's pooled labour and shared suppliers and knowledge-in-the-air make every member of it more than he could ever be alone. The locational rent is the capitalized value of that productive crowd; and so the question of whether it is earned or extracted comes down to this: the crowd creates the value, by the bare act of gathering, and the owner of the ground collects it, by the bare fact of owning. Of all the rents this volume has weighed, the agglomeration rent is the one most plainly made by the many and taken by the few — the surplus of the gathering, settling on the soil, harvested by its holder. And yet the gathering is real, and its fruits are real, and a price that rations the scarce ground of the cluster to those who can turn it to best account is not nothing. Whether the rent of the crowd is the just reward for furnishing a scarce and valuable place, or the unearned harvest of a value the crowd itself created, is the volume's question, asked now of the pooling of the world's activity — and bound up, as the bridge to the volume of growth will show, with the deepest question of all: why some places, and some nations, race ahead, while others are left in the long shade of the periphery. As the table says of every price within it, the argument remains open.