1001 Evidence Brief
💰 Which smart-city projects publish what they cost?
Few of them, and not the ones you would expect. 236 of 1167 entries state a cost at all — 20.2%. A public-private partnership publishes a price 38% of the time, a publicly funded project 19%, and a privately funded one least of all at 10%. Disclosure follows the procurement announcement, not the public purse.
✔ Every published entry is checked against the public sources it cites before publication.
Who publishes a price, by who is paying
States a costStates none
| Funding model | Entries | Cost disclosed | Share | Entries |
|---|---|---|---|---|
| All published entries | 1167 | 236 states a cost, 931 states none | 20% | 236 |
| Public-private partnership | 79 | 30 states a cost, 49 states none | 38% | 30 |
| Mixed / blended | 103 | 32 states a cost, 71 states none | 31% | 32 |
| Grant-funded | 77 | 18 states a cost, 59 states none | 23% | 18 |
| Utility-funded | 74 | 14 states a cost, 60 states none | 19% | 14 |
| Public budget | 647 | 122 states a cost, 525 states none | 19% | 122 |
| Not stated | 145 | 16 states a cost, 129 states none | 11% | 16 |
| Private | 42 | 4 states a cost, 38 states none | 10% | 4 |
The ordering is the finding. A partnership or a mixed-funding project is announced through a procurement event — a contract award, a concession signing, a bond issue — and that event has a number in it. A line in a municipal budget has no press release, and a company spending its own money has no reason to publish one at all, which is why privately funded entries sit at the bottom at 10%.
Representative entries
- Nairobi Expressway Electronic Tolling A price that exists and is still not a fact: the reported cost runs between KSh 51bn and KSh 88bn, and this atlas records the range rather than picking from it. A disclosed figure and a checkable one are different things.
- Porto Maravilha: CEPAC development-rights financing The partnership pattern in its clearest form. Both the R$3.5bn bond issue and the R$7.609bn contract are public because each had to be announced to be executed — the money left a paper trail because of how it was raised, not because anyone chose transparency.
- NYC311 Service Requests & Open Data One of the most-linked entries in this atlas, running for two decades on a municipal budget, and there is no figure anywhere in its sources for what it costs to operate. Nothing is hidden; nobody was ever required to publish it in one place.
Method
An entry counts as stating a cost when its record carries a cost figure at all — any currency, any precision, capital or programme. That is 236 of 1167 published entries.
The funding model is the atlas's own classification of who pays, from a fixed vocabulary: public budget, public-private partnership, grant, utility, private, mixed, or not stated. Entries whose funding this atlas could not establish sit in the last group and are shown rather than dropped.
The category breakdown is computed the same way; the domain that discloses most often is Energy at 32%. Every figure is regenerated at build time from the published data.
What this cannot tell you
- A stated cost is not a comparable cost. The figures here are in different currencies, from different years, and cover different things — one entry's capital budget against another's whole programme. This brief counts whether a number exists, and nothing about its quality.
- Absence may be our reach rather than the city's silence. The atlas records what its sources state; a cost published in a budget document nobody indexed is invisible here and would still be public.
- Some disclosed figures are ranges or disputed, which the entries say so where it is the case. Counting them as disclosed is the honest choice — a contested number can at least be argued with.
- The funding model is an editorial classification, not a legal one, and a large project often has several sources of money. Entries with more than one are grouped as mixed.
- Nothing here is an accusation. Most of these projects are small enough that no disclosure rule applies to them, and the finding is about where price information happens to survive, not about who is hiding what.
The slice behind every number
Every figure above is computed at build time from the published data, and the rows it counts are downloadable as one file: which-smart-city-projects-publish-their-cost.json. Recomputing the numbers is a matter of counting that file. The full datasets are on the open data page, under CC BY 4.0.
Where to read next
The costs page lists the largest disclosed figures and what they bought. Who publishes the evidence asks the same kind of question about sources rather than money. And the brief on who evaluates smart-city projects finds that an audit office — the institution most likely to publish a cost — reaches almost none of them.
All briefs · Who publishes the evidence · How the record ages · The state of the evidence