Jakarta went from thirty-third to first among the world's largest cities without one person moving house.
Anyone who follows this subject will get to the bottom of the opening screen without having seen a building. What occupies the frame instead is an instrument mid-calibration — a grid laid over ground, a threshold, a cluster of cells that has just been classified as a single settlement — held at reading distance, unhurried, closer in register to a page than to a place. Administrative and cartographic, not urban. The method of measurement arrives first; the thing measured arrives second.
What decides whether a city works is written down somewhere unglamorous — in a delineation, a land register, and a budget line that has to be voted every year — and none of the three is visible in a rendering of the place.
The cities that get drawn and the cities that get built are, at this point in the century, almost disjoint sets. Most of the additional urban population to 2050 is projected into seven countries; the projects that dominate the image of urban futures are in none of them. The correction that matters is not a moral about where attention goes — it is a specific, checkable claim about what the constraint actually is. Recorded tenure, collectable municipal revenue and the length of network per household can each be measured, they are what the evidence in this Journey keeps landing on, and an architectural image is structurally incapable of showing any of the three.
For every claim in this Journey, find the document behind it first: the line that defines the city, the register that records who lives inside it, the appropriation that keeps it running, and the invoice that arrives later for whatever was never built.
This subject runs on more than one time axis. The chapters are not one sequence.
The reader stops treating the urban share as a measurement of the world and starts treating it as an output of a convention that can be, and was, replaced.
Degree of Urbanisation · Twenty-first-century urbanisation · Jakarta
The growth acquires an address, and a second property the address does not advertise: it is arriving at falling density, which lengthens every metre of network per person served.
Sub-Saharan African urban growth · Urban dedensification · Informal settlement · Urban housing supply · Agglomeration · Land tenure security
A flat coverage line stops looking like failure and starts looking like enormous effort exactly cancelled by arrivals — and the reader can name the two things underneath it, neither of which is construction.
Urban water supply and sanitation · Land tenure security · Municipal fiscal capacity · Urban housing supply · Informal settlement
The received story about the empty smart city is corrected in front of the reader, and the correction is sharper than the myth: the housing worked, and the thing the housing was justified by did not form.
Two projects in different decades, on different continents, meet the same obstacle — not engineering, not demand, but the second and third years of funding — and the repetition is what makes it a finding rather than an anecdote.
NEOM · Nusantara · Jakarta · Brasília · Municipal fiscal capacity
The chapter before it stops being readable as "planning fails". Both of these were drawn whole by open competition in the same century; the records attribute the difference between them to a hundred years of amendment rather than to the quality of either drawing.
Runs alongside 2 other chapters, not after
The reader gets the inverse profile of everything in the two preceding chapters — short, aimed at neighbourhoods that already exist, cheap to build and expensive to keep running — and learns that the expensive half is where these succeed or fail.
Medellín · Metrocable Line K, Medellín · Curitiba · Bus rapid transit · Urban mass transit · Lagos Blue Line · Lagos · Municipal fiscal capacity
Unbuilt infrastructure stops being a gap and becomes a debt with interest: the reader can follow one chain from a household borehole to a sea wall, and a second from a hot street to a national grid.
Jakarta · Urban land subsidence · Urban water supply and sanitation · Urban cooling demand · Heat action plan · Ahmedabad · NEWater · Medellín
Tenure stops being an abstraction from chapter three and becomes the thing that decided which of two waterfronts in one city was permanent — and the reader is left holding the fact that the population in question has not been counted in twenty years.
Eko Atlantic · Makoko · Lagos · Land tenure security · Informal settlement · Twenty-first-century urbanisation
30 records in this Journey.
Showing 8 of 21 featured records. Atlas does not choose which of the rest matter.
One Lagos waterfront is being added to the map and one is being removed from it, in overlapping years, a few kilometres apart. Nothing about construction quality separates them: the manufactured ground comes with recorded title and the nineteenth-century settlement does not. Read together they are the cleanest natural pairing in the corpus for the claim that tenure, not building, decides permanence.
The graph already holds this as a chain, and it is worth walking end to end: a piped network that did not reach everyone, private abstraction where it did not, sediment compacting, ground falling faster in the north than the sea is rising — and, among the stated reasons for legislating a new capital on another island, that same subsidence. A water main is not usually thought of as a thing that can move a capital city.
Urban water supply and sanitation · Urban land subsidence · Jakarta · Nusantara
Two federal capitals of the same century, both settled by open competition, both drawn whole before anyone lived in them, with markedly different occupation outcomes. The records do not attribute the difference to the drawings. One plan was implemented slowly and amended for a hundred years; the other was inaugurated after roughly four years of construction, dimensioned for a population it long ago exceeded, with the workers who built it already living outside it.
Set against a cooling load measured in thousands of terawatt-hours and growing, the intervention in this Journey with a published mortality evaluation behind it is a threshold, a warning and a set of instructions for who does what when the temperature crosses it. It has been replicated across Indian cities and states. The comparison is not that documents beat machines — it is that the two are financed, owned and copied in completely different ways.
The most widely copied urban transport idea of the last fifty years is also the Journey's clearest demonstration that a corridor is an operating expense rather than an asset: the design travelled to roughly two hundred cities, and in the city that invented it daily ridership fell to well under half its peak. What copies easily is the geometry. What does not copy is the annual subsidy.
How many people live in Lagos?
The question cannot be answered from the record. The most recent census was held in 2006, its Lagos figure was publicly rejected by the state government, a later attempt was halted, and no new date had been set as of 2026. The corpus deliberately keeps a range rather than tightening it, and the fact that it cannot be tightened is one of the Journey's findings rather than a gap in it.
Which urban share is the real one?
Both conventions are defensible and they measure different things. The grid method is more comparable between countries and less connected to how any country actually governs, because administration follows legal boundaries and not grid cells. The Journey carries both rather than choosing, which is why the first chapter exists at all.
Which boundary do the Songdo figures describe?
Retrieved figures attach to at least three different boundaries — the business district, the wider international city, and the free economic zone — and the sources do not consistently say which. The split between a residential half that filled and a business half that did not is robust; the specific numbers on either side of it are not, and a curator with access to Incheon statistics should pin both.
What is actually happening at NEOM?
Much of what circulates about the project traces to unnamed sources, and the corpus deliberately excludes the specific suspension dates, write-downs and contract values in wide circulation. What it carries is a commissioned feasibility review, reduced resident expectations and an omission from a pre-budget statement. Atlas also holds no coordinate for it, so it appears nowhere in this Journey's places — the most-rendered city here is the one canonical knowledge declines to put on a map.
Did the cable car reduce violence, or did something else?
A natural-experiment study found larger declines in served neighbourhoods than in comparable controls. But the programme bundled several investments at once against a changing national security backdrop, the city's overall decline has many causes, and the effect sizes quoted in circulation differ substantially between retellings. The direction is supported; the attribution is not clean and the Journey does not clean it.
Is the reported service coverage better or worse than the real thing?
Almost certainly worse. Coverage statistics rest on household surveys with real definitional slippage between improved, basic and safely managed service, and informal areas are systematically under-sampled — which means the parts of a city this Journey is most about are the parts least present in the figures describing it.
How many people were displaced at Makoko, and why?
The settlement's population is reported across a very wide range and the displacement figures for the recent clearance are contested. The state and the residents give incompatible accounts of the reason — proximity to a power line, or clearance for waterfront redevelopment. The corpus states both and adjudicates neither, and a curator should re-check whether the legislative halt still stands.
The two most-publicised cities in this Journey are both still proposals, and the one new city that actually got built succeeded at housing and failed at the thing it was justified by. Meanwhile the intervention with the best evidence behind it is about two kilometres long, and the cheapest one is a document that tells people what to do when it gets too hot. None of that is an argument against ambition. It is an argument about which ledger the ambition lands in: the capital budget everyone photographs, or the operating budget that has to be voted again next year by a government that did not announce anything. There is also a bill neither ledger carries. In Lagos, in overlapping years, one waterfront was added to the map with title attached and another was cleared for lacking it — so the cost of the century's urbanisation, at the one point in this Journey where it can be traced to a street, was paid by households with no recorded tenure, in a city that has not been counted since 2006.
This is unfinished in the world, not only in the telling.
Ask the world another question. Dynamic keeps this Journey as its context and never changes what is written above.
Open Dynamic Atlas →