School Clusters, Federations & Shared Administration

HEW-NODE-0161 · How Education Works · school clusters, school federations, shared administration, rural schools, small schools, network governance, specialist teachers, shared services, school identity, resource pooling, leadership, transport, accessibility and school-network design

Sometimes the choice is presented too crudely.

A small school is expensive, so close it. A remote school is hard to staff, so merge it. A village school cannot offer enough specialist subjects, so move the students somewhere larger.

But school systems have another family of options. Several schools can remain physically open while sharing leadership, specialist teachers, procurement, administration, professional learning, transport coordination, data systems, extracurricular provision or even parts of the timetable.

A school network can gain some economies of scale without requiring every community to lose its local school.

This node sits beside the How Education Works hub, School Closure, Consolidation & Teach-Out, School Mapping & Capacity Planning, School Catchment Areas, Boundaries & Rezoning, Education Demographic & Enrolment Projections and Education Shared Services & Administrative Consolidation.

Those pages keep their jobs. School Closure owns physical closure and consolidation decisions. School Mapping owns where schools and capacity should exist. Catchment Areas owns geographic access rules. Shared Services owns system-wide administrative consolidation. This node owns the middle structure: formal networks in which multiple schools retain sites and learners while sharing selected functions as one operating group.

The 60-Second Read

A school cluster is a group of schools that cooperate around defined shared functions.
A federation usually implies more formal common governance or leadership across separate schools, though legal terminology varies by jurisdiction.
Clustering can create scale without eliminating every small site.
Shared administration is not the same as shared teaching.
One cluster can centralise payroll, procurement or reporting while schools retain local heads and timetables.
Another can share principals, counsellors, specialist teachers, laboratories or extracurricular programmes.
Geography matters: a cluster that looks efficient on a map can become unusable if travel time is too high.
Younger learners generally have stronger reasons for provision close to home.
Specialist secondary subjects may gain more from shared teaching than early primary provision.
Virtual or rotating specialist teaching can complement physical clustering, but should not be assumed to solve every staffing problem.
Cluster governance needs clear decision rights.
School identity should not be treated as a decorative concern; local trust can affect participation and implementation.
Shared budgets can create cross-subsidy disputes unless allocation rules are explicit.
Shared leadership can reduce duplication and also overload one leader if span of control is poorly designed.
Data systems need school-level visibility even when administration is shared.
Performance should be measured at both cluster and site level.
Transport cost is part of the network model, not an afterthought.
Cluster design should be tested against demographic decline and growth.
Some schools should remain small because geography or community need makes consolidation harmful.
The objective is not to maximise school size. It is to design the smallest network structure that can sustainably deliver the educational offer learners need.

One-Sentence Definition

A school cluster or federation is a formal arrangement in which multiple school sites share selected governance, leadership, staffing, services or educational provision while remaining distinguishable institutions or locations within a wider operating network.

The First Distinction: Cluster Is Not Closure

School closure removes a site or institution from the network and transfers learners elsewhere. A cluster can leave the site open while changing how the site is governed, staffed or supported.

This distinction matters because many of the disadvantages of small schools arise from scale — limited specialist staffing, leadership burden, procurement inefficiency, professional isolation — rather than from the existence of the local site itself.

The Second Distinction: Shared Administration Is Not Shared Pedagogy

Two schools might share finance staff and procurement while teaching independently. Another cluster might coordinate curriculum, rotate specialist teachers and run common assessment moderation.

Cluster design should specify which layer is shared instead of assuming “federation” means everything becomes common.

The Third Distinction: Network Efficiency Is Not Simply Fewer Principals

A cluster can save administrative cost and still fail educationally if one executive head is spread across too many sites, travel consumes leadership time, or local decision-making slows because every question moves upward.

Efficiency must be measured against service quality, accessibility and the time cost of coordination.

OECD Evidence: Consolidation Has Alternatives

The OECD’s Responsive School Systems review notes that small schools can coordinate educational offers, share resources or consolidate administration to gain efficiencies without necessarily removing local schools. It discusses clusters and federations as ways to share equipment, courses, purchasing and teacher development while preserving access.

This is especially important in rural and remote systems. The same OECD work emphasises that geographic isolation can make full consolidation inappropriate and that reasonable proximity remains especially important for younger children.

Current School-Network Work Is Increasingly Geospatial

OECD work on geospatial modelling for Latvia’s school network reorganisation combines school cost and accessibility measures rather than treating cost alone as the decision rule. More recent OECD work on demographic change likewise recommends standardised impact assessment for closures and mergers, including access, transport, quality and participation.

The same logic is useful for clusters. A network arrangement should be tested spatially: how far must a specialist teacher travel, how long do learners spend on transport, and which shared service can realistically reach each site?

What Problems Are Clusters Trying to Solve?

very small enrolment;
high per-student fixed costs;
difficulty recruiting specialist teachers;
principal workload in tiny schools;
limited subject choice;
professional isolation;
duplicated administration;
underused specialist rooms;
weak procurement scale;
fragmented student support;
limited extracurricular opportunities;
unstable staffing caused by part-time fragments.

The design should begin with the actual problem. A cluster built to share physics teachers will look different from one built to reduce finance administration.

Cluster Models Form a Continuum

Informal collaboration: schools share occasional activities.
Service consortium: schools jointly purchase or contract defined support.
Specialist network: staff rotate or courses are shared across sites.
Administrative cluster: finance, HR, data or procurement are centralised.
Leadership federation: one executive leadership team spans several sites.
Governance federation: schools share a formal governing body.
Multi-site school: separate sites operate increasingly as one institution.

The terminology differs across countries. The functional question matters more: what is shared, what remains local, who decides and how is accountability preserved?

Start With a Function Map

List the functions currently performed separately at each school:

leadership;
finance;
procurement;
HR administration;
timetabling;
specialist teaching;
learning support;
counselling;
assessment moderation;
professional development;
ICT support;
laboratory management;
libraries;
transport coordination;
facility management;
family communication.

Then ask which functions gain from scale and which lose value if moved away from the local site.

Fixed Costs Drive Much of the Economics

A school with eighty students still needs a principal function, basic administration, safeguarding, utilities, records, maintenance and statutory reporting. Many costs do not fall proportionately with enrolment.

Clustering can spread some fixed costs across more learners without necessarily moving the learners themselves.

But Some “Savings” Are Only Moved Elsewhere

If the system reduces school leadership cost but increases transport, travel reimbursement, substitute coverage and coordination time, the net saving may be much smaller than the headline suggests.

Whole-system costing should include new costs created by sharing.

Leadership Span of Control Is a Design Variable

One executive principal can support several sites when local operational leadership remains strong and sites are close. The same arrangement can fail across distant locations with complex programmes and high support needs.

Useful questions include:

How many sites?
How many students?
How many staff?
How much travel?
How many distinct programmes?
How much local decision authority remains?
Which decisions require executive sign-off?

Local Leadership Still Needs a Name

Parents, students and staff need to know who is responsible at the site today. A federation can have elegant central governance and still feel leaderless locally if nobody has practical authority.

Define local site leader, executive leader, governing authority and escalation route explicitly.

Specialist Teacher Sharing Can Stabilise Timetables

A small secondary school may not generate enough timetable hours for a full-time physics, music or computing specialist. Across four nearby schools, the combined demand might support a stable full-time role.

The network can then employ one specialist with an intentional rotation instead of relying on several fragile part-time contracts.

Travel Time Is Instructional Capacity

If a teacher spends six hours each week moving between sites, those hours are part of the staffing model. They cannot be treated as free.

Cluster timetables should minimise dead travel through block teaching, predictable rotation days and geographic grouping.

Some Subjects Can Travel Better Than Others

A specialist can carry a laptop and lesson materials more easily than a chemistry laboratory. Practical subjects may require students to move to a hub site instead of staff moving to them.

Shared provision should therefore follow the physical requirements of the subject.

Virtual Sharing Is a Tool, Not a Free Substitute

Video teaching can extend subject access across distance, but requires connectivity, classroom supervision, local support and pedagogical design suited to remote participation.

It can reduce travel and still create new dependencies. The model should compare both.

Shared Professional Learning Can Be a Major Gain

A teacher who is the only specialist in a tiny school has few same-subject colleagues nearby. Clusters can create professional communities large enough for moderation, lesson study, subject planning and mentoring.

This can be one of the most important benefits even when direct financial savings are modest.

Shared Student Support Needs Case Ownership

A counsellor, psychologist, therapist or inclusion specialist may work across several schools. The cluster gains access to expertise, but only if referral ownership is clear.

Who holds the case? Who speaks to the family? Where are records stored? Who follows up when the specialist leaves the site?

Governance Must Match the Shared Function

If schools share only procurement, a light agreement may be enough. If they share leaders, staff and budgets, governance needs to be much more formal.

What is the legal entity?
Who employs staff?
Who owns the budget?
Who signs contracts?
Who sets the curriculum?
Who appoints the executive leader?
Who hears complaints?
Who can leave the federation?
What happens to assets if the arrangement ends?

Shared Budgets Need Allocation Rules

If one school has 600 students and another 70, equal division of shared costs may be unfair. Pure per-capita division may also be unfair because the smallest sites often have higher fixed needs.

Possible bases include enrolment, weighted need, service usage, fixed-plus-variable formulas or negotiated shares. The rule should be visible before conflict appears.

Keep Site-Level Accounts Visible

Centralising accounting should not make it impossible to see what resources each site receives. Site-level visibility helps detect unintended cross-subsidies, underinvestment and service differences.

Identity Has Operational Value

A school can be a major community institution. Names, traditions, parent networks and local relationships can affect attendance, volunteering and trust.

A cluster that preserves local identity while sharing back-office functions may face less resistance than one that treats identity as inefficiency.

But Identity Cannot Become a Veto on Every Change

If local attachment prevents any resource sharing even when learners are losing subject access, the system can protect institutional history at the expense of current students.

Consultation should make trade-offs visible rather than imply that every existing arrangement can remain unchanged.

Access Must Be Tested by Age

Older secondary students may reasonably travel to a cluster hub for advanced laboratories or specialist courses. The same travel pattern may be inappropriate for young primary children.

Network design should distinguish education levels rather than apply one distance rule to all learners.

Transport Reliability Becomes Part of the Academic Model

If a shared mathematics extension class runs at the hub school but buses arrive late twice a week, the cluster has created access on paper and lost instructional time in practice.

Transport timetables, weather risk and after-school return routes need to be designed with the educational timetable.

Cluster Performance Needs Two Levels of Evidence

Cluster-wide averages can improve while one satellite school deteriorates. Site-only metrics can miss benefits from shared staff and broader course access.

Measure both:

cluster-level staffing stability;
site-level attendance;
subject access;
student travel burden;
teacher travel burden;
leadership workload;
administrative cost;
professional-learning participation;
support-service response time;
family satisfaction;
learning outcomes with appropriate caution about causality.

Demographic Change Should Trigger Scenario Testing

A cluster that works with 450 students may fail when enrolment falls to 270. Another may become too centralised when a growth corridor adds 800 learners.

Cluster design should be tested against demographic scenarios rather than frozen around today’s enrolment.

Case Study: Four Small Secondary Schools and One Physics Teacher

Illustrative example: four rural schools each generate only a fraction of a full physics timetable. Recruitment repeatedly fails because every role is part-time.

The schools create a specialist cluster. One full-time physics teacher teaches two full days at the largest site, one day at each of two smaller sites, and one shared laboratory day at a hub. Travel time is built into the job rather than added informally.

The result is not a claim of automatic learning gain. The immediate system improvement is a stable qualified post and predictable subject access.

Case Study: Shared Principal, Lost Local Leadership

Illustrative example: three primary schools share one executive principal. Administrative cost falls, but parents report that no one appears authorised to resolve day-to-day issues at two sites.

The federation adds named site leaders with delegated authority for attendance, safeguarding coordination, family communication and routine operations. The central principal retains staffing, budget and strategic decisions.

Case Study: The Cluster That Saved Nothing

Illustrative example: a district centralises administration across five schools expecting major savings. It keeps all old local processes and adds a new central office on top.

Costs rise because the cluster added coordination without removing duplication. The repair is functional redesign, not merely a new organisational chart.

Case Study: The Shared Laboratory Hub

Illustrative example: three small schools cannot each justify advanced laboratory equipment. One central site becomes the specialist practical-science hub. Students travel for scheduled laboratory blocks while theory teaching remains local.

The design succeeds only after transport, accessibility, timetable blocks, equipment capacity and missed-session recovery are planned together.

Failure Modes and Repairs

Cluster without a problem: repair by defining the exact scale failure being solved.
Leadership overload: repair by setting span-of-control and local-authority rules.
Travel treated as free: repair by costing teacher and student movement.
Central office added on top: repair by removing duplicated processes.
Shared budget opacity: repair with explicit allocation formulas and site-level reporting.
Local identity ignored: repair through structured consultation and preserved site identity where useful.
Identity blocks every change: repair by making learner trade-offs explicit.
Cluster average hides weak sites: repair by reporting both cluster and school-level evidence.
Remote teaching assumed equivalent: repair by testing supervision, access and instructional quality.
Network frozen after launch: repair by re-testing against demographic and staffing change.

The School-Cluster Operating Chain

Map current schools and enrolment.
Identify the scale problem.
Map travel times and transport.
Map staffing and subject gaps.
Map duplicated administrative functions.
Identify functions that could be shared.
Identify functions that should remain local.
Choose a cluster model.
Define legal status.
Define governance.
Define executive and site leadership.
Define shared staffing rules.
Build travel time into workload.
Define shared-budget allocation.
Define asset and procurement arrangements.
Define student-support case ownership.
Define data and reporting architecture.
Consult communities and staff.
Model costs including transition costs.
Model access by age and learner group.
Pilot the operating model where feasible.
Measure site and cluster performance.
Correct coordination failures.
Re-test under demographic scenarios.
Decide whether the cluster remains preferable to independent schools, full consolidation or another network model.

A School-Cluster Dashboard

schools and sites in cluster;
students by site;
staff by site;
shared staff FTE;
specialist subject coverage;
teacher travel hours;
student travel time;
leadership span of control;
shared-service cost;
site-level operating cost;
administrative cost per student;
vacancy rate;
staff turnover;
professional-learning participation;
support-referral wait;
shared facility utilisation;
family complaints by site;
site attendance;
course access;
next demographic review date.

Canonical Owner Boundaries

School Closure, Consolidation & Teach-Out owns physical closure and transfer when a school site is removed.
School Mapping & Capacity Planning owns geographic location and capacity planning.
Education Shared Services & Administrative Consolidation owns broader system-level common back-office capacity.

This node owns the school-network operating arrangement in which several sites deliberately share selected governance, leadership, staffing, educational provision or administration while remaining present in the network.

Sources and Further Reading

The Return Path

Return to the small school that looked too expensive to keep and too important to close.

The answer does not have to be sentimental preservation or automatic consolidation. Sometimes the correct design is to keep the local learning site and change the machinery around it.

Share what benefits from scale. Keep local what loses value when centralised. Measure the travel, governance and coordination cost honestly. Then ask again whether the network is serving learners better.

Scale is a property of functions, not a moral requirement that every school become large.

Return to the How Education Works hub.

How Education Works | School Clusters, Federations & Shared Administration — How Small Schools Gain Scale Without Disappearing