The City as a Network of Communities: Decentralisation Without Fleeing to the Countryside

A city is not one community — and that is precisely why it can contain many communities

A large city is too large for all residents to experience it as one direct community. Most people do not know most other residents, and daily movement connects them to different parts of the city. This is not a defect that can be solved by treating an urban population as one enormous village.

A more useful picture is multi-level. A building or street can form a very close social unit. A neighbourhood can connect a school, park, shops, associations, local services and daily routes. A district can bring several neighbourhoods together. The city provides functions that one district cannot efficiently carry. A metropolitan area can exceed even municipal boundaries because people may live in one municipality, work in another and use the same transport, housing, water or economic network.

The OECD therefore does not measure cities only by administrative boundaries. It also uses the concept of a functional urban area: an urban core and the areas functionally connected to it, especially through commuting flows. This matters because the real boundary of a problem is not always the municipal boundary.

If a city is a network of different scales, decentralisation is not a choice between “city or village”. It is a question of which function belongs to the building, neighbourhood, district, city and wider metropolitan network.

Density is not only a problem; it is also shared infrastructure for opportunity

Rural romanticism often correctly notices urban weaknesses: high housing costs, noise, traffic, anonymity, pollution, dependence on complex supply systems and distance from sources of food or energy. The mistake comes only when we infer that lower density is automatically freer, more resilient or more humane.

Density also has advantages. The OECD links more compact urban form with better access to services and jobs, higher productivity from proximity between people and activities, more effective public transport, and lower land and energy use per person. The same density can also create housing pressure, infrastructure strain, noise and quality-of-life problems if poorly governed.

R80 is therefore not a defence of every form of urbanisation. The goal is to separate the benefits of proximity from the problems of concentrated power and poor spatial governance. A city can retain density while bringing part of decision-making, care and material capacity closer to residents.

In practice, decentralisation begins not by dismantling the city but by asking whether more daily needs can be met near home: basic retail, green space, childcare, school, some social services, some work, community space, local logistics or safe public transport into the wider network.

First, map the scales

One of the most useful tools for urban decentralisation is a map of scales. Rather than automatically moving every function downward or upward, we ask at which level its benefits, costs and consequences actually occur.

Building or block: shared spaces, minor repairs, energy efficiency, tool sharing, local mutual help and management of common areas.

Neighbourhood: small public-space improvements, parks, community space, local events, some social support, safe routes to school, micro-logistics and part of a participatory budget.

District or borough: larger local services, coordination across several neighbourhoods, local centres, some cultural and sports infrastructure, and coordination among local initiatives.

City: major roads, public transport, larger health and educational institutions, housing policy, major infrastructure, system-wide standards and redistribution between areas.

Metropolitan or regional area: commuting flows, rail and major transport, the wider housing market, water and environmental systems, logistics, major waste treatment, and some energy and emergency systems.

This division is not a universal law. In a small town, the same organisation may function as both neighbourhood and municipality; in a multi-million-person metropolis, one district may be larger than an entire Slovenian city. The principle is what matters: the scale of governance should follow the scale of the problem.

This is where R80 reconnects with R49 and R50. Subsidiarity is not automatic downward transfer but an evidential question. If a problem can be solved closer to residents without exporting costs and harm to others, there is a case for local decision-making. If consequences cross boundaries, the scale of coordination must grow with them.

The neighbourhood can be a more politically and materially real unit

Much of everyday urban life does not happen at the scale of the whole city. People use the same streets and parks, encounter the same dangerous crossing, the same noise, the same trees, the same school and the same local junctions. For these questions, residents have close knowledge that a distant administration may struggle to see.

Neighbourhoods or districts can therefore have councils, assemblies, participatory budgets, local funds or other mechanisms through which residents propose and monitor smaller projects. Ljubljana, for example, has 17 district communities that connect residents with the city administration, collect initiatives, run local programmes and prepare plans for small works. This is not complete decentralisation of the city, but it is a concrete example of multi-level urban self-government.

Such structures are most meaningful when they receive real but bounded authority, not merely the opportunity to submit opinions. UN-Habitat guidance on participatory spatial planning warns that participation should not remain a symbolic gesture; residents need clear influence, sufficient information and a way for their input to connect to actual decisions.

R58 already showed that a community does not need a permanent ruler, but it does need a decision process. In a city, this means moving some questions into smaller units without dismantling the entire urban system.

Local decision-making is not automatically fairer

Proximity can make participation easier, but it does not abolish power differences. A local process can be captured by a small, vocal and well-organised group. Renters, young people, migrants, people with less free time, older people, disabled residents or poorer residents may face greater barriers to participation than property owners with more time and information.

UN-Habitat therefore emphasises the inclusion of groups that are typically under-represented and warns that simply creating a digital platform or holding a consultation does not create representative participation. A decentralised decision is not legitimate merely because it is local; we must also ask who could participate and who was left out.

A second danger is local veto power. A neighbourhood may rightly decide many nearby details, but it cannot automatically gain the right to block every housing project, transport connection or service needed by the wider city. If local autonomy allows wealthier areas to retain benefits while pushing costs elsewhere, we have created local concentration of power rather than subsidiarity.

Good urban decentralisation needs two safeguards at once: more nearby influence and shared rules for basic fairness, access and responsibility toward other parts of the city.

Some functions are inherently city-wide or metropolitan

Transport is the clearest example. A bus or rail line is not useful only in the neighbourhood it passes through; it connects homes, jobs, schools, hospitals and other districts. If every municipality or district plans only its own segment without common coordination, the result can be disconnected timetables, duplicated investment and gaps at boundaries.

Recent OECD analysis of metropolitan governance therefore emphasises horizontal coordination across municipalities and policy sectors. When administrative boundaries do not match the functional city, transport, land use, housing, economic development and some social services are often better coordinated across the wider urban area.

The same applies to other systems. A major wastewater plant, waste treatment, specialist hospital, flood protection, water source, energy grid or emergency logistics system may require a scale beyond the neighbourhood. R77–R79 already showed that water, energy and infrastructure operate through physical networks that political boundaries do not stop.

This is not an argument for centralising everything. It is an argument for a nested system: the local level governs where proximity has an advantage; the wider level governs only where the shared system is genuinely wider.

A city can become polycentric without drawing new administrative borders

Polycentricity does not necessarily require creating new municipalities. It can also emerge through multiple nodes of daily life: local centres, markets, libraries, health points, coworking spaces, schools, parks, community kitchens, repair workshops, distribution points or energy communities.

UN-Habitat and the OECD increasingly emphasise accessibility and proximity in urban policy. Concepts such as the city of short distances or the 15-minute city are most useful as an accessibility test, not a rigid formula: how many basic needs can a person reach safely and reasonably without a long compulsory trip?

Such organisation can reduce daily dependence on one centre and one transport axis. At the same time, the city keeps universities, specialist hospitals, major cultural institutions, railway hubs, large labour markets and other functions whose value comes precisely from connecting many people.

A polycentric city is therefore less like a pyramid and more like a network of nodes. Each node holds some local capacity, while none is forced to reproduce every function of the whole city.

Urban resilience comes from redundancy between neighbourhoods, not their isolation

A large city can be vulnerable because its supply chains are complex and failure at a key node can affect many people. Yet the same scale can create reserves: multiple hospitals, suppliers, warehouses, service teams, transport routes and people with different skills.

UN-Habitat work on urban resilience emphasises a combination of locally led solutions, city-wide coordination, data and multi-level partnerships. For R80, the important conclusion is that resilience is not achieved by turning every neighbourhood into a closed unit. It is more useful for each neighbourhood to strengthen some basic capacities while maintaining agreed pathways of mutual aid and substitution between neighbourhoods.

One district may have a community centre, another a larger health facility, another service equipment and another a warehouse. If links between them are clear, the network can survive more disruptions than a system with everything in one centre — and more than a system in which every small unit must maintain every possible reserve on its own.

R81 will continue this logic under disasters and major disruptions. R80 only needs to establish the foundation: urban resilience is the capacity of local nodes connected through networks of mutual aid and shared systems.

A practical audit: decentralising the city without dismantling the city

For a building, neighbourhood or district, we can run a simple urban audit:

1. What do people use every day nearby? Park, school, childcare, shop, transit stop, shared space, health care, social support, repair, delivery?
2. What is currently decided too far away? Which smaller decision would be faster and better informed at building, street, neighbourhood or district level?
3. Which decision also affects other neighbourhoods? Transport, housing, water, flooding, waste, noise, major infrastructure or public institutions?
4. Who can participate and who is systematically absent? Does the process include renters, young people, older people, disabled residents, migrants and people with less time or digital access?
5. Which assets could have a local steward? Community space, small green area, part of a budget, shared tools, local energy asset, micro-logistics point?
6. What must remain a shared city-wide standard? Minimum service access, safety, rights, transparency, basic infrastructure quality and rules that prevent exporting harm to others?
7. Where do we need city or metropolitan coordination? Major transport, housing market, wider environment, water, major hospitals, logistics or emergency systems?
8. Do we have more than one node? If the centre fails, is there another point for services, information, supplies or support?
9. How do we measure results? Not only the number of meetings, but service access time, responsiveness, inclusion, quality of space, costs and initiatives actually implemented?
10. What is the smallest viable first step? One competence, a small fund, a local forum, a shared space, a resource register or a pilot project that can later be evaluated?

This audit does not require a major constitutional reform. It starts with the everyday city and searches for decisions that can move closer and systems that must remain connected.

Decentralisation without escape

Not everyone is willing or able to live in the countryside. Many people need proximity to jobs, health services, schools, public transport, cultural environments, specialist services or social networks that larger cities can provide more easily. If a political vision of local freedom depends on leaving the city, it becomes a vision mainly for those who have land, capital, transport and the ability to move.

A more universal question is: can a person gain more real local power where they already live? In a city, that means stronger neighbourhoods, transparent district structures, more locally controlled shared assets, shorter daily journeys, greater material capacity and clear channels through which local units participate in city-wide decisions.

At the same time, we should preserve what makes a city work: large networks of knowledge, employment, transport, specialist infrastructure and mutual support. A freer city is not a fragmented city. It is a city in which power and capacity are distributed across many nodes while shared functions remain shared only where the shared problem genuinely requires them.

The countryside is not the only place for local sovereignty. A city, too, can become a network of communities — if proximity becomes capacity and scale becomes cooperation rather than distant rule.