Who runs the smart city
Every published entry records the kind of body that operates it. Across 869 of 871 projects, this is what that adds up to.
The atlas kept arriving at this question from other directions. The South America region turned on it — a metropolis with no city government of its own, a metro that belongs to a pension fund. Half of North America turned on it too: the bike-share series, where identical Lyft terms sit on top of three different owners, and the camera contracts, where a vendor's cut of each fine changes what the programme is for. Ownership decided more than technology did, over and over, and the record could not be counted that way.
It can now. The operator field on each entry still names the specific body, because that is where the precision is — "Johannesburg Water SOC Ltd (municipal entity of the City of Johannesburg)" says something a category cannot. Alongside it, actorTypes says what kind of body that is, so the question becomes countable.
One warning about reading these pages: a legal form tells you nothing about ownership. Shenzhen Bus Group, Incheon International Airport Corporation and Rome's AMA S.p.A. all carry a corporate suffix and are all public bodies. They are filed here by what they are, not by what they are called — which is the entire reason the axis is worth having.
✔ Every published entry is checked against the public sources it cites before publication.
Every kind of operator in the atlas
| Run by | Entries | Share | Failed or scaled back |
|---|---|---|---|
| 🏙️ City government | 442 | 50.7% | 8.6% |
| 🗺️ Regional government | 47 | 5.4% | 6.4% |
| 🏳️ National government | 77 | 8.8% | 13.0% |
| 🔌 Utility or public operator | 277 | 31.8% | 5.4% |
| 🏢 Private company | 175 | 20.1% | 17.1% |
| 🏫 Research institution | 41 | 4.7% | 7.3% |
| 🤲 Civil society | 23 | 2.6% | 17.4% |
💸 Paying for it and running it are different things
46 entries are funded from a public budget or a grant and operated by a private company. Copenhagen's street lighting runs on Citelum, Bogotá's fare card on Recaudo Bogotá, Boston's Bluebikes on Lyft — in each case the public body is the client, not the operator. Only 8 entries run the other way. Reading a project's funding model as a statement about who is in charge gets it wrong 46 times in this corpus alone.
🏛️ Most of the record is run by public bodies
642 entries are operated entirely by city, regional or national government or by a utility. 175 involve a private operator at all, and 107 are genuinely mixed — a public body and a private one both with an operating role. The smart city, as this atlas records it, is mostly run by the state; the interesting cases are the ones where it is not.
📉 A failure gap that is probably about paperwork
Entries with a private operator carry a failed or scaled-back outcome at 17.1%, against 5.4% for utilities and 8.6% for city governments. That is an association and almost certainly not a verdict on private operation. A concession has an end date and a press release when it ends; a utility programme simply continues. The atlas records a failing outcome only when a named source says so, and contracts produce named sources. The same selection effect runs through the evidence grades.
2 published entries appear on none of these pages. 2 of them record no operator at all and cannot be classified; the rest arrived after the classification pass and are waiting for the next one. Guessing would put entries into a public filter on the strength of nothing, so they stay out until somebody has read them.
Open a set
Where to read next
The same record sliced other ways: The State of the Evidence counts who examined these projects, failures groups the ones that went wrong, and Your Smart City Is Owned by Someone Else is the essay this axis grew out of.