Findings#

What 32 cities and 16 dimensions reveal about participation infrastructure.


No city has broken 50#

The highest CommonScore is 44 (Singapore). The “strong participation infrastructure” band starts at 40 — and only one city has reached it. Singapore did it through comprehensive government engineering: hawker centres, community centres in every constituency, SkillsFuture credits for every citizen, ActiveSG memberships for the entire population.

No city has achieved this through organic, bottom-up infrastructure alone. The closest are London and New York (both 38), which combine deep institutions (City Lit, CUNY, three library systems) with grassroots participation (Parkrun, off-off-Broadway, Borough Market).

The question for every city: what would it take to break 40?


Civic infrastructure is the strongest predictor#

In 28 of 32 cities, Civic & Community is the top or second dimension. Cities with strong public library systems lead the rankings. This is not a coincidence — libraries with meeting rooms, maker spaces, and event halls are the closest existing model to a commons.

Top civic scoresInfrastructure
Helsinki 7.7Oodi (purpose-built commons)
Singapore 5.9108 Community Clubs
New York 5.93 library systems, 200+ branches
Boston 5.9BPL system
Toronto 5.9TPL 100 branches + 152 rec centres
DC 5.9MLK Library + federal access

Implication: If you want to raise your city’s CommonScore, invest in libraries and community centres. They’re already doing the work.


Convention center size doesn’t predict score#

McCormick Place is 2.6 million sq ft — the largest convention center in North America. Chicago scores 26. Oodi is 185,000 sq ft — not even a convention center. Helsinki scores 30. The building’s programming matters more than its size.

Convention centerSizeCity score
McCormick Place2.6M sq ft26
Messe Berlin1.7M sq ft33
ExCeL London1M sq ft38
RAI Amsterdam915K sq ft32
Oodi (library)185K sq ft30

Three models of participation#

ModelCitiesHow it works
Government-engineeredSingapore, Seoul, Mexico CityState builds and operates infrastructure at national/city scale
Institutional depthNew York, London, Boston, TorontoLibraries, universities, and arts institutions accumulate over decades
Organic/distributedTaipei, Portland, Austin, MedellínMarkets, food carts, tool libraries, maker spaces built from the bottom up

No model is better — they produce comparable scores (35-44, 32-38, 24-28). But the government-engineered model reaches higher because it fills every dimension systematically, while organic models leave gaps.


The hawker centre is the highest-scoring single institution#

Singapore’s 122 hawker centres score 6.9 on Food — higher than any single institution in any dimension in any city. Pike Place Market scores 4.9. Oodi scores 7.7 on Civic but that’s for a purpose-built commons covering 7 dimensions.

Hawker centres are: government-built, vendor-operated, daily, universal access, UNESCO heritage, multigenerational. This is what “induced demand” looks like after 50 years of systematic investment.


Music cities aren’t participation cities#

Nashville’s “Music City” scores 14 — last in the dataset. Austin’s “Live Music Capital” scores 27, same as Seattle. The brand emphasizes watching bands (consumption). Participation infrastructure — rehearsal studios, open mics, affordable all-ages venues — is a different thing.

The highest Music dimension scores: Melbourne 3.4 (465 venues), Seoul 2.2 (Hongdae busking + noraebang), Nashville 2.8 (songwriter rounds — the one thing that IS participation).


Convention centers contribute 14-25% of city scores#

Every city follows the same pattern: the convention center contributes to Industry Networking and Mega-Events, and nothing else.

CityCC share of score
San Diego25% (highest — thinnest city-wide infrastructure)
Seattle16%
Chicago15%
Helsinki10%
Singapore9% (lowest — deepest city-wide infrastructure)

The policy question: What if the building served the other 14 dimensions too?


The tool library pattern#

Three cities have tool libraries scoring as maker participation infrastructure: Seattle (Capitol Hill Tool Library, since 2007), Toronto (Toronto Tool Library, 3 locations, 5,200 members), and Denver (Denver Tool Library). All are volunteer-run, community-funded, pay-what-you-can or low-cost.

This is the Phoenix Comics pattern applied to fabrication: proven daily demand at retail scale. The bottleneck is always space, never demand.


It doesn’t have to be a convention center#

The dataset is organized around convention centers, but the CommonScore framework applies to any large underutilized space. Former factories, vacant office towers, old department stores, decommissioned industrial sites — the participation question is the same: what if this building served the dimensions where the city scores lowest?

The precedents are already in the data:

ConversionOriginal useWhat it becameCity
798 Art DistrictElectronics factory400+ galleries and studiosBeijing
Songshan Cultural Creative ParkTobacco factoryStudios, workshops, performanceTaipei
De HallenTram depotFood hall, cinema, maker storeAmsterdam
NDSM WharfShipyard250+ artist workspacesAmsterdam
WestergasfabriekGas factory17-building creative complexBerlin
Kaapelitehdas (Cable Factory)Cable factoryLargest cultural center in FinlandHelsinki
The BreweryPabst brewery120+ artist live/work loftsLos Angeles
CarriageworksRailway workshopContemporary arts centreSydney
Michigan CentralTrain stationFord mobility campusDetroit

But scale matters. An art walk in a repurposed office building works at one scale — it’s a proof of concept, like Phoenix Comics for gaming or the Capitol Hill Tool Library for fabrication. Art Basel in a convention center works at another — 268 galleries, 83,000 attendees, $640M in renovated facility. Songshan Creative Park (former tobacco factory) is closer to the second than the first: institutional-scale repurposing with government backing, permanent studios, year-round programming.

The CommonScore doesn’t prescribe which buildings to repurpose. It identifies which dimensions have the largest gap between proven demand and available space — and the gap analysis works whether the candidate building is a convention center, an empty Macy’s, or a decommissioned power plant.


Geographic inequality shapes scores#

Cape Town (27) is the first city in the dataset where geographic inequality directly shapes the CommonScore. Participation infrastructure exists on both sides of the divide — Woodstock studios and township art projects, Sea Point pools and township sports clubs — but the distribution is uneven in ways other cities don’t face at the same scale.

The CommonScore measures what exists, not how equitably it’s distributed. A city where half the population can access participation space and half cannot scores the same as a city with universal access — as long as the total capacity exists. This is a known limitation.


CommonScore methodology | Leaderboard | All cities