What 2,487 Pins Say.

The civic tech map began as a list of links we kept sending to people. Someone would ask who works on freedom of information in Hungary, or whether there's a FixMyStreet for their country, and we'd dig through bookmarks. At some point it was easier to put the bookmarks on a map.

The map holds 2,487 entries. We wrote 307 of them ourselves and went through each one against the project's own site or official records. The other 2,180 come from the Civic Tech Field Guide, whose crowdsourced registry is the largest list of its kind we know of. Those entries keep the registry's own descriptions and link to each project's website.

Checking took longer than we expected, and it changed how we see the thing. When you read hundreds of summaries in a row, patterns jump out that you'd never notice clicking pins one at a time.

The map in October 2026

initiatives on the map

2,487

initiatives on the map

countries

85

countries

cities

352

cities

Entries by region

• North America 1243
• Europe 657
• Asia-Pacific and Middle East 374
• Latin America 141
• Africa 72

Top five cities

• London 174
• Washington, D.C. 165
• New York 158
• San Francisco 128
• Berlin 103

The map leans towards the United States, and that's partly on us

The United States holds 1,196 entries, almost half the map. Europe has 657. London, Washington, D.C., New York, San Francisco and Berlin hold 728 between them. Africa has 72 in total, fewer than Berlin on its own.

Some of that is real. Foundations, digital government teams and the funders behind them do cluster in a few capitals. But a lot of it is about who does the listing. Our own 307 entries are mostly European, because we read mostly in English and a handful of European languages and follow the newsletters our friends write. The Field Guide's registry is mostly American, because that's where its community grew up. Merging two lists doesn't cancel out their biases, it stacks them. A Kenyan or Indonesian lab would draw a different map, and it would be no less accurate than ours.

We'd rather say that up front than pretend the map is a census.

A pin shows where an organisation is registered, not where its software runs

The health category is where this is most obvious. DHIS2 is pinned in Oslo, because that's where the university that builds it is, but ministries of health in dozens of countries use it. OpenMRS sits in Indianapolis. The Community Health Toolkit sits in San Francisco and is used across Africa and Asia. OpenLMIS and OpenELIS Global are in Seattle, a long way from the supply chains and public health labs they run.

So the map undercounts where civic tech actually reaches people. We haven't found a good way to show deployments without turning every pin into a cloud.

Two economies: money or code

Every entry we wrote ourselves ends with one thing a reader can do. We picked whatever fit the project best, so the action field turns out to be a decent proxy for what each project needs most. The numbers here cover those 307 entries; the imported ones simply link to the project's website.

What each kind of project asks you to do

Category	Entries	Asks for money	Points to code
Digital rights	43	31	6
Accountability	45	20	2
Transparency	38	10	7
Participation	24	2	9
Health	17	0	14
Public services	30	0	26

Covers the 307 entries we wrote ourselves; entries imported from the Civic Tech Field Guide link to the project's website. Each of ours ends with one suggested action. "Asks for money" counts entries whose action is to support or join the organisation; "points to code" counts entries that link to a repository.

The split surprised us. In digital rights, 31 of 43 entries ask for money or membership. Only six point to a repository, among them the Hermes Center, where the GlobaLeaks whistleblowing software began, and HURIDOCS's Uwazi, a database for human rights documentation. In public services it's the reverse: 26 of 30 point to code and none ask for donations. Health looks like services, with 14 of 17 pointing to code.

It makes sense once you see it. What noyb, La Quadrature du Net or the Electronic Frontier Foundation produce is litigation, policy work and campaigns. You can't fork a lawsuit, so the way to help is to pay the people filing it. What GOV.UK Notify or DHIS2 produce is software, and the way to help is to use it, report bugs and send patches.

We call all of this civic tech, and the label hides the difference. A developer who wants to help digital rights should probably reach for their wallet, not their editor. A donor who wants to help public service software might do more good paying a maintainer than giving to a campaign.

Governments build more of it than the story suggests

The usual picture of civic tech is volunteers working around a slow government. The services category doesn't fit that picture at all. GOV.UK Notify, Login.gov, the U.S. Web Design System, NL Design System, France's La Suite numérique, Italy's IO app, Singapore's FormSG and Estonia's X-Road were all built by public bodies and published as open source.

Some of that code has already travelled. Canada's GC Notify was adapted from GOV.UK Notify, which is the argument from our piece on public money and public code working exactly as advertised. Decidim and CONSUL Democracy, two of the best-known participation platforms, started inside the city councils of Barcelona and Madrid. Helsinki runs its own consultation tool, Kerrokantasi.

None of this means governments are doing enough. It means the ones that publish their code give everyone else something to build on, and the ones that don't are leaving that on the table.

The volunteers come in families

The volunteer side is there too, and it's more organised than it looks from outside. The map has 48 groups called Code for something. Thirty-nine of them are local Code for America brigades, from Pittsburgh to Hackensack, and the rest are in Canada, Germany, Japan, Australia, Romania, Hungary, South Korea and Portugal. Open Knowledge has chapters or projects in Brazil, Belgium, Nepal, Finland and Germany, where FragDenStaat grew out of Open Knowledge Foundation Germany. mySociety's tools sit together in London, and its Alaveteli platform runs freedom-of-information sites well beyond the UK. Transparency International shows up directly and again through Integrity Watch EU and Ukraine's DOZORRO.

What these networks pass along isn't mainly code. It's a way of running a group: how to hold a hacknight, how to file a request, how to talk to a council. That's much harder to copy from a repository.

Many of these tools were born in a crisis

Ushahidi was first built to map reports of violence after Kenya's 2007 election. Safecast started measuring radiation after Fukushima. SORMAS, an outbreak management system, was first deployed during the Ebola epidemic. KawalPemilu let Indonesians crowdsource polling station results so they could check their own election count.

A crisis is a great reason to build something quickly. It's a terrible business model. Whether a crisis tool survives depends on what happens after the cameras leave, which is the question we looked at in civic tech that lasted.

Berlin pays for a lot of the plumbing

Berlin has 103 entries, 16 of them written by us, and those aren't just projects. Several of them are the money and hosting underneath other projects: the Sovereign Tech Agency, the Prototype Fund, the Free Software Foundation Europe, Codeberg and Wikimedia Deutschland. Of the 32 commons entries we wrote, 24 are in Europe, and most of those are German or Swiss. Across the whole map, 116 of the 215 commons entries are in the United States.

That's good news for anyone those institutions fund. It's also a lot of weight on a few budgets, and budgets change. We went into the details in who funds the digital commons.

Civic tech dies quietly

The last pattern came from the checking itself. We removed 11 entries that were inactive or compromised, and corrected 32 more that had moved, renamed themselves or changed what they do. Just before that, we had dropped DemocracyOS after it lost its domain.

Almost none of those projects announced anything. Their sites went stale, then their domains lapsed, and in a few cases the address started serving something unrelated. A directory nobody checks slowly turns into a list of dead links, and some of those links stop being harmless. That's why each entry now shows the date we last confirmed it, and why a script re-tests every link once a month whether anyone's watching or not.

Resources

If the map made you want to dig further, these are the places we keep going back to. All of them are on the map too.

• Participedia is a crowdsourced database of public participation cases and methods from around the world, and a good place to see what a tool looked like in actual use.
• The Standard for Public Code from the Foundation for Public Code is a practical checklist for public bodies that want their code to be reusable by others.
• Developers Italia, code.gouv.fr and openCode are catalogues of software published by Italian, French and German public bodies.
• The EU's Open Source Observatory collects case studies on how public administrations across Europe use open source.
• The Open Contracting Data Standard is what to point your procurement team at if you want public contracts published as usable data.
• NLnet and the Prototype Fund fund small teams and individual developers, and the Sovereign Tech Agency funds maintenance of the infrastructure underneath.
• Software Heritage and the Wayback Machine let you save a repository or a page before it disappears. Given the previous section, we'd encourage it.

The map is big, and it still leans. Every entry carries the date it was last checked, and the 307 we wrote ourselves each tell you one concrete thing you can do.