How Long-Term Care Works in Sweden: Municipalities, Regions and the Welfare State

For an older person in Sweden who begins to need help with washing, meals, mobility or managing everyday life, the route into long-term care does not normally begin with a national insurance authority or a single national care programme. It begins locally. The municipality assesses social-care needs, determines what assistance should be provided and organises services such as home help or special housing. At the same time, healthcare needs may involve municipal nurses, rehabilitation professionals, primary healthcare, specialist services or hospitals operating within a different sphere of responsibility.

Understanding this division is fundamental to understanding Swedish long-term care. The Sweden Ageing, Long-Term Care & Community Support Knowledge Hub examines the different parts of this system and how they interact as population ageing increases demand for coordinated support. Sweden's model is strongly publicly organised and substantially tax-funded, but it is also decentralised. National legislation establishes important principles and responsibilities while 290 municipalities and 21 regions translate them into local services.

The result is neither a single national long-term care service nor a fragmented collection of entirely independent local schemes. It is a multilevel welfare system in which national government, national agencies, municipalities, regions and service providers hold different responsibilities. Its effectiveness depends on whether those responsibilities combine around the individual rather than becoming organisational boundaries that the older person or family must navigate themselves.

Long-term care sits inside Sweden's wider welfare state

Swedish long-term care is best understood as part of a wider welfare model built around substantial public responsibility for health and social protection. Older people may draw on healthcare, pensions, housing support, municipal social services, rehabilitation, home adaptations and community infrastructure as their circumstances change.

Care for older people therefore cannot be reduced to a single programme. The principal social-care framework is the Social Services Act, while healthcare is governed through separate health legislation. Municipalities carry extensive responsibilities for social services and older people's care, while regions hold the main responsibility for providing healthcare to their populations.

This distinction also separates Sweden from long-term care models predominantly organised through a dedicated social insurance scheme. Municipal care is largely financed from taxation alongside central government grants, with people making regulated personal contributions for certain services. The system consequently combines collective financing with individual needs assessment rather than giving every older person an identical package of services at a particular age.

The practical pathway is determined by need. An older person who remains independent may receive no formal long-term care. Someone who begins to struggle with everyday activities may apply to the municipality for support. Another person may require extensive daily home help, municipal healthcare and frequent regional medical input. Someone with very substantial needs may move into special housing.

The defining principle is therefore universal public responsibility rather than universal receipt of the same service.

The municipality is the central long-term care actor

Sweden's municipalities occupy an unusually important position in the everyday organisation of long-term care. Their responsibilities include assessing individual need and arranging much of the support that enables older people to continue living at home or, where required, access special housing.

Municipal social services can provide different interventions depending on assessed circumstances. These may include home help, personal support, meals, safety alarms, day activities and forms of residential provision. Local organisation differs, which is an intended consequence of municipal self-government rather than an anomaly in the system.

For international readers, the municipality should not simply be translated into a UK local-authority commissioning model. Swedish municipalities may directly operate services, contract with external organisations or combine public and private delivery. Their role encompasses political responsibility, administration, needs assessment, funding decisions, service organisation and quality oversight within the legal framework.

This makes organisational structure and accountability particularly important. An older person may experience care through an individual worker or provider, but accountability extends through provider management, municipal administration, locally elected political structures and national oversight.

Access usually depends on an individual needs assessment

Public responsibility does not mean that someone automatically receives a defined amount of home help because they reach a certain age. Municipal social services consider the person's circumstances and determine what support is required under the applicable legal framework.

This creates an important operational gateway. Assessment needs to identify not only obvious deficits but the person's ability to manage everyday life, existing support, living circumstances, preferences and the outcomes that assistance is intended to achieve.

Older people can differ considerably in what constitutes an effective response. One person who struggles to prepare meals may require practical support. Another may have the physical ability to cook but no longer initiate meals consistently because of cognitive impairment. Someone recovering from illness may need rehabilitation rather than permanent substitution of tasks. A person whose mobility has declined may benefit from adaptation of the home alongside care.

The quality of assessment therefore influences subsequent resource use. If it focuses only on tasks, services can become a timetable of inputs. If it understands strengths, risks and desired outcomes, the response can be more proportionate and capable of supporting independence.

This wider principle is reflected in person-centred planning and strengths-based support for older people. Swedish assessment has its own legal and administrative context, but the operational question is universal: does the support decision reflect the individual life that the service is intended to sustain?

An application for home help becomes a wider independence decision

An 82-year-old man living alone contacts his municipality after his daughter becomes concerned that he is struggling with shopping, cleaning and showering. The immediate request appears relatively straightforward: regular home help.

Assessment identifies a more complicated situation. He remains capable of many daily activities but has become cautious after a recent fall. He has reduced his walking, stopped travelling to a nearby shop and now relies heavily on his daughter. His difficulty is partly practical and partly a loss of confidence.

A narrowly task-based response could simply transfer shopping, cleaning and personal-care activity to paid workers. A stronger response considers which tasks require continuing assistance, whether rehabilitation or mobility support could restore capability, whether the home environment increases fall risk and how family involvement can remain supportive without becoming indispensable.

The eventual package may still include home help, but the assessment determines whether that support maintains independence or unintentionally accelerates dependency. The case also demonstrates why municipal social care, rehabilitation and healthcare cannot always be planned independently when an older person's needs begin to change.

Home help is a major expression of ageing in place

Home help is central to the Swedish long-term care model because it enables many older people to receive assistance while continuing to live in ordinary housing. Support may range from relatively limited practical help to multiple daily visits for people with extensive personal-care needs.

The operating complexity increases substantially as care intensity rises. Several short visits each day require reliable scheduling, continuity, travel time, information exchange and contingency capacity. Where healthcare needs are also present, home-help workers may operate alongside nurses and rehabilitation professionals, creating additional coordination requirements.

Home-based care can therefore be less institutionally visible than residential care while being operationally highly complex. Each home becomes a separate work environment. Staff travel between locations. Families may be present or absent. Equipment varies. Risks can change between visits. Workers have less immediate access to colleagues than staff based within one residential setting.

The challenge resembles issues considered within homecare workforce and scheduling: route design, continuity and staff deployment are not administrative details. They directly influence whether a person receives timely and reliable care.

Special housing provides a different level of support

Ageing at home is an important policy direction, but Swedish long-term care does not assume that ordinary housing remains appropriate regardless of need. Municipalities are also responsible for special forms of housing for older people who require substantial service and care.

Special housing provides an environment in which accommodation and extensive support can be brought together. For people with significant frailty, dementia or multiple care requirements, this can provide greater continuity and access to staff than dispersed home-based provision.

The decision to move is significant. It affects home, identity, relationships, finances and everyday control as well as care. Strong practice therefore requires the person's needs and wishes to remain central rather than treating special housing simply as the end point of an escalating care pathway.

Private organisations may operate special housing, but private delivery does not remove the public responsibility surrounding access and quality. Private providers operating services that require authorisation are subject to the relevant national requirements and supervision. The mixed-provider landscape therefore sits within a publicly governed welfare framework rather than operating as a separate private long-term care system.

Regions remain essential because older people also need healthcare

Municipal responsibility becomes only part of the picture when an older person has medical needs. Sweden's regions carry the principal responsibility for healthcare, including primary and specialist medical services and hospitals.

This means that a person living in municipal special housing can simultaneously depend on systems operating under different organisational responsibilities. The municipality may organise daily care and elements of healthcare while regional services remain responsible for medical functions that sit outside municipal responsibility.

The distinction becomes increasingly consequential as people receiving long-term care live with multiple illnesses, frailty and complex medication. A service model designed primarily around help with daily living must therefore interact reliably with healthcare capable of diagnosis, prescribing, medical review and treatment.

This is why the boundary between care and healthcare is not simply an organisational chart. It is one of the major operational interfaces in Swedish long-term care.

Municipal healthcare adds another layer to the system

Municipalities do not only provide social-care services. They also hold significant healthcare responsibilities within parts of long-term care. Municipal healthcare can include nursing, rehabilitation and other healthcare delivered in settings where municipal responsibility applies. Across much of Sweden, agreements between regions and municipalities have also transferred responsibility for home healthcare in ordinary housing to municipalities, although arrangements are not identical everywhere.

Physician responsibility remains with the regional healthcare system. This distinction creates an especially important interface for older people with complex needs. A municipal nurse may know the person's daily condition closely, while medical diagnosis or prescribing may require regional input. Care workers may detect deterioration, but an effective escalation pathway is required to translate that observation into appropriate clinical action.

Good coordination therefore depends on role clarity as much as professional competence. Staff need to understand who is responsible for what, how information is transferred, which service responds when a person's condition changes and how urgent concerns are escalated.

Interoperability and system integration become particularly important where responsibility crosses organisations. Technology can support the exchange of information, but integrated care is not created merely by connecting software. Shared processes, professional relationships, legal information-sharing arrangements and clear accountability remain necessary.

A discharge exposes the practical boundary between region and municipality

An older woman living alone is admitted to hospital after pneumonia and significant loss of mobility. Medically, she improves sufficiently to leave hospital, but returning to her previous level of independence is unrealistic without additional support.

The regional hospital needs to communicate the changed clinical position. Municipal services need to understand what assistance will be required at home. Rehabilitation may need to continue. Medication arrangements have changed. Her daughter wants to help but cannot provide daily care because she works full time.

If each organisation plans only its own component, the formal discharge can succeed while the transition fails. The woman may arrive home before equipment is available, staff may lack current information, or the family may assume responsibilities that have not been explicitly agreed.

A stronger transition brings the different responsibilities together around the planned outcome: a safe return home with realistic support and a clear route for review. Information about function, medication, rehabilitation, home-help requirements and escalation needs to travel with the person.

The example illustrates a fundamental feature of Swedish long-term care. A person's pathway can cross regional healthcare and municipal care repeatedly. System quality therefore depends not only on the performance of each organisation but on what happens between them.

Taxation provides the financial foundation

Swedish long-term care is largely financed through public resources, particularly municipal taxation and central government funding. This collective financing model reduces dependence on individuals purchasing the full economic cost of extensive care from their own income or assets.

Older people can nevertheless pay charges for services and accommodation-related costs. Municipal fees operate within national rules designed to limit the amount individuals are required to contribute and protect an amount for ordinary living costs.

The distinction is important internationally. Tax-funded does not mean cost-free at every point of use, and universal does not mean that municipalities have unlimited resources. Municipalities have to balance care for older people with other statutory and local responsibilities while managing differences in population structure and tax base.

Central government grants can support national priorities or strengthen municipal capacity, but the underlying operating model remains substantially local. Financial sustainability therefore depends partly on the ability of municipalities to convert available resources into effective services.

This creates pressure for informed decisions about:

• the balance between preventive support and intensive long-term care;
• home-based provision and special housing capacity;
• direct municipal provision and externally operated services;
• workforce skill mix and use of technology;
• rehabilitation and maintenance of independence; and
• how quality and continuity are protected as productivity expectations increase.

These are not simply budget decisions. Each changes the experience of older people and the future pattern of demand.

Public responsibility can coexist with private provision

The presence of private providers sometimes leads international discussion to portray Sweden's older people's care as either public or market-based. In reality, those categories overlap.

Municipalities retain public responsibility while service delivery may involve both municipal and private organisations. In some municipalities, people may have a choice between approved providers under local choice arrangements. Elsewhere, organisational arrangements differ.

The central governance question is therefore not simply ownership. It is whether the municipality can assure itself that people receive appropriate, safe and high-quality services regardless of who employs the frontline workforce.

External provision requires effective specifications, contractual arrangements, information flows, monitoring and escalation. Direct municipal provision requires equally robust internal governance. Public ownership does not itself demonstrate quality, just as private ownership does not determine poor quality.

For organisations considering comparable purchaser-provider relationships, the Commissioner Evidence Builder offers a general framework for structuring contract, evidence and assurance questions. It is designed for a UK operating context and is not a Swedish administrative tool, but the underlying discipline of connecting expectations, evidence and review remains relevant where public bodies purchase services from external organisations.

Regulation combines local responsibility with national oversight

Decentralisation does not mean that every municipality can define acceptable care independently. National legislation, regulations, professional requirements and oversight create a wider framework within which local services operate.

The National Board of Health and Welfare, Socialstyrelsen, has an important national role in knowledge, statistics, regulations and guidance across health and social care. National guidelines can support decisions about priorities and evidence-based practice, while national registers and statistics provide information about services and outcomes.

The Health and Social Care Inspectorate, Inspektionen för vård och omsorg or IVO, supervises health care and social services, including older people's services. It also has responsibilities relating to authorisation for relevant privately operated activities.

This creates several layers of assurance rather than one inspection mechanism determining quality by itself. Providers retain responsibility for their services. Municipalities need oversight of care they organise. Healthcare professionals operate within professional and patient-safety frameworks. National agencies generate guidance, information and supervision.

The strongest governance model connects these layers. An incident discovered by frontline staff should be capable of informing provider learning. Recurring concerns should become visible to municipal leadership. Wider patterns should inform regulatory or national understanding where appropriate.

Organisations examining their own assurance architecture can use the Quality Dashboard Builder to consider how workforce, safety, experience and outcomes can be brought into one coherent view. It does not reproduce Swedish national quality systems, but it reflects the broader need to connect operational data with accountable decision-making.

Quality cannot be understood through compliance alone

Long-term care is delivered through repeated human interactions that are difficult to capture fully through structural measures. A service may complete scheduled visits and meet formal staffing requirements while still providing poor continuity or insufficient personal control.

Conversely, an isolated negative event does not necessarily reveal the quality of an entire municipal system. Governance requires patterns, context and proportionate interpretation.

Useful quality intelligence can combine service activity, workforce information, complaints, incidents, outcomes, user experience and evidence about continuity. Sweden's national information infrastructure and local datasets create opportunities for comparison and improvement, but data only become useful when decision-makers act on what they reveal.

This is closely related to quality data, KPIs and performance metrics. A mature system asks not only whether a measure can be collected, but whether it identifies something sufficiently important to influence action.

Completed visits conceal a declining service experience

A municipality reviewing home-help performance sees reassuring headline data. Most planned visits are delivered and serious incidents are uncommon. Yet satisfaction among people receiving frequent daily support has begun to decline.

Analysis shows that recruitment difficulties are producing constant changes in who attends. The scheduling system has successfully filled shifts, but some older people are seeing a large number of different workers over a short period.

For someone requiring straightforward practical help, this may be inconvenient. For a person living with dementia or needing intimate personal care, the effect can be much greater. Staff spend time re-establishing familiarity, preferences are less well known, relatives repeat information and subtle changes in health are easier to miss.

The municipality therefore adds continuity to its routine performance view rather than measuring delivery only through completed activity. Local teams reorganise deployment around smaller groups of workers for people with the highest continuity needs.

The example shows why long-term care governance needs measures that reflect how services are experienced. Efficiency and reliability remain important, but neither is fully represented by whether a scheduled task occurred.

The workforce is the infrastructure of the model

Sweden's institutional arrangements ultimately depend on a sufficiently capable workforce. Municipal responsibility cannot translate into support if there are too few workers to deliver it, and sophisticated care models cannot compensate for unstable staffing or insufficient competence.

Older people's care draws on care workers, assistant nurses, registered nurses, occupational therapists, physiotherapists, managers and other professionals. Regional healthcare adds doctors and wider clinical services. The workforce challenge therefore involves both the number of workers and the distribution of competence across organisational boundaries.

Population ageing increases demand at the same time that many municipalities face wider labour-supply pressures. Rural areas may experience particularly difficult recruitment conditions, while larger cities compete across extensive health and care labour markets.

A sustainable response involves workforce planning rather than recruitment campaigns alone. Services need to understand which tasks require which competence, where continuity has the greatest value, how career pathways can strengthen retention and where digital tools can remove avoidable administrative burden.

Technology is becoming part of care infrastructure

Swedish municipalities have substantial experience with welfare technology, including safety alarms, digital forms of supervision and other tools intended to support independence and service delivery. Technology is increasingly relevant because distributed home-based care creates both opportunities and operational pressures.

A digital solution can reduce unnecessary travel, give an older person greater privacy or enable staff to identify emerging risk earlier. Digital records can improve information availability. Scheduling systems can use workforce capacity more effectively. Remote contact can extend professional reach across large geographic areas.

Yet the effectiveness of technology depends on implementation. Introducing a device without redesigning the associated care process can add work rather than remove it. Staff need training. Contingencies are required for system failure. Information security and privacy matter. People who lack digital confidence need alternatives.

Most importantly, the decision to use technology needs to remain proportionate to the person. A digital night check may increase one individual's privacy while making another person feel less secure. A safety alarm can support independence but does not compensate for loneliness or unmet personal-care needs.

Organisations exploring comparable digital change can use the Digital Transformation Readiness Assessment to test whether governance, workforce adoption, resilience and strategy are sufficiently developed alongside technical implementation. It is not a Swedish regulatory assessment, but the readiness questions apply widely.

Family members operate alongside formal long-term care

Sweden's welfare model deliberately places substantial responsibility on public systems, but this does not make family support irrelevant. Spouses, adult children and other relatives may provide emotional support, transport, help with administration, communication with professionals and extensive practical assistance.

The important distinction is that family involvement should not become an invisible eligibility mechanism. Two older people with equivalent needs may have very different family circumstances. One may live near several relatives who actively want to help. Another may have no close family. A third may depend on an ageing spouse whose own health is deteriorating.

Municipal assessment therefore needs to understand family circumstances without assuming unlimited availability. Formal services and families can complement one another, but reliance on hidden unpaid labour can transfer system pressure into households.

This is particularly important when care becomes more complex. Relatives can become informal coordinators between home help, healthcare, pharmacy services and different professionals. Where they repeatedly have to repair communication between organisations, the system may appear coordinated administratively while remaining fragmented from the family's perspective.

Carer support and family partnership therefore need to include clarity about roles as well as involvement in decisions.

Local variation is both a design strength and an equity challenge

Sweden's 290 municipalities range from large urban administrations to small and sparsely populated communities. Their demographic profiles, geography, local labour markets, tax bases and existing service infrastructures differ considerably.

Expecting every municipality to organise long-term care identically would therefore be unrealistic. A dense city can organise home-help routes differently from a northern rural municipality where staff travel substantial distances. The viability of specialist teams, provider choice or certain technologies can also differ with population scale.

Decentralisation allows adaptation to these conditions. The challenge is determining when variation is justified and when it represents inequity.

An older person's access to good care should not depend arbitrarily on where they live. National standards, evidence, data and oversight therefore have an important equalising role even where service organisation remains local.

The relevant governance test is not whether municipalities produce identical inputs. It is whether differences in organisation still produce defensible access, quality and outcomes.

A rural municipality cannot reproduce an urban workforce model

A sparsely populated municipality finds that recruiting enough staff to maintain highly specialised separate teams is becoming increasingly difficult. Travel between homes consumes a significant proportion of working time, particularly for people living outside the main population centre.

Simply adopting the staffing model of a larger municipality would worsen the problem because workforce density and travel conditions are fundamentally different.

The municipality instead examines the whole operating model. It considers which expertise needs to be physically present, where professionals can work across traditional boundaries within appropriate competence, how remote specialist input can be used and whether scheduling can reduce unnecessary journeys. It also assesses which older people would be disadvantaged by digital contact and retains in-person provision accordingly.

Governance focuses on outcomes and safety rather than whether the rural service looks organisationally identical to an urban one. Continuity, response times, unmet need and workforce stability become especially important measures.

The scenario shows why decentralisation has continuing value. Standardising expectations for dignity, safety and access does not require every community to deliver them through the same operational mechanism.

The 2025 Social Services Act changes the direction of travel

Sweden's new Social Services Act came into force in July 2025 and strengthens the direction towards social services that are preventive, accessible and knowledge-based. Its detailed implications for older people's care warrant separate analysis, but the reform is relevant to the architecture of long-term care because it changes the expectations placed on municipal social services.

For long-term care, prevention means more than providing information before someone requires formal support. It can include recognising declining function earlier, connecting people with community resources, supporting rehabilitation, reducing isolation and making assistance accessible before difficulties develop into greater dependency.

Knowledge-based practice requires another shift. Decentralisation creates substantial scope for local decisions, but those decisions increasingly need to draw on evidence and systematic learning rather than local custom alone.

The implication is not that municipal discretion disappears. It is that the rationale for local choices should become stronger. Services need to understand what they are trying to achieve, what evidence supports the approach and whether local outcomes justify continuing it.

This makes continuous improvement part of system architecture. A decentralised welfare state needs mechanisms that allow local experimentation while spreading knowledge about what works.

Governance must join the parts the law separates

Sweden's institutional division of responsibilities is not inherently a defect. Distinct organisations can bring clear expertise and democratic accountability. Problems arise when the interfaces are treated as somebody else's responsibility.

The older person requiring long-term support experiences one life. Their needs do not divide neatly into municipal social care, municipal healthcare, regional medicine, housing and family activity.

Effective governance therefore needs to operate horizontally as well as vertically. Municipal leaders need assurance about their own services, but they also need visibility of recurring problems at regional interfaces. Regional healthcare needs to understand how its decisions affect home-based and residential care. Providers need routes for escalating patterns that cannot be solved operationally.

Organisations examining these relationships can use the Governance Maturity Assessment to structure questions about responsibility, escalation and assurance. The framework does not replace Swedish legislation or public accountability, but it illustrates why mature governance asks not simply who owns a service, but who owns the outcome when several organisations contribute to it.

The system increasingly needs to manage complexity rather than just volume

Ageing will increase demand for Swedish long-term care, but aggregate demand tells only part of the story. The profile of people receiving support is also important.

As more people remain at home until needs become substantial, home-based services may work with greater frailty and complexity. People entering special housing may have extensive cognitive, physical and healthcare needs. Greater clinical complexity increases the importance of medication management, rehabilitation, observation, escalation and healthcare coordination.

This changes the operational unit of planning. Counting how many people receive home help is insufficient if each person's care requires more time, competence and coordination than previously. Similarly, counting beds in special housing does not reveal whether workforce skill mix is appropriate to the residents living there.

Municipalities therefore need capacity intelligence capable of connecting population projections with complexity, workforce and outcomes.

The Digital Twin Scenario Modeller can help organisations explore similar relationships between workforce, capacity and service stability. It does not model Swedish municipal entitlements or budgets specifically, but scenario-based planning is particularly relevant when demand is driven by interacting demographic and operational variables rather than one predictable growth rate.

Accountability ultimately needs to remain visible to the person

A sophisticated governance structure has limited value if an older person does not know whom to contact when support is wrong, late or no longer sufficient.

Accessible complaints processes, reassessment, opportunities to raise concerns and transparent decisions are therefore part of long-term care quality. People need to understand what has been decided, which organisation is responsible and how circumstances can be reviewed when needs change.

Person-centred care is not achieved simply because an assessment records preferences. It requires those preferences to influence delivery. Timing, continuity, communication, privacy, cultural identity and ordinary daily routines can determine whether support feels enabling or institutional even when it takes place at home.

At system level, service-user feedback and co-production can also reveal issues that administrative measures overlook. If people repeatedly describe uncertainty about who is responsible after hospital discharge, that is system intelligence rather than merely a collection of individual complaints.

What the Swedish model offers international systems

Sweden's structure cannot be separated from its taxation system, local-government tradition, welfare institutions and political history. Transferring the institutional model directly to another country would therefore make little sense.

Several underlying principles are nevertheless useful internationally.

The first is that public responsibility can coexist with decentralised delivery. National government does not need to operate every service directly in order to establish common rights, expectations and knowledge infrastructure.

The second is that decentralisation requires powerful feedback mechanisms. Local autonomy becomes safer when outcomes, variation and quality are sufficiently visible to identify where adaptation is productive and where it is creating inequity.

The third is that long-term care cannot be separated cleanly from healthcare once people have complex needs. Organisational boundaries may remain, but pathways need to bridge them.

The fourth is that mixed provision does not remove public accountability. Whether a service is municipally operated or privately operated, the public system still needs evidence that the intended outcome is being delivered.

Finally, universal welfare should be understood as a commitment to collective responsibility rather than uniform service design. The quality of universality is ultimately tested by whether people with different needs and circumstances can access appropriate support on equitable terms.

Conclusion

Sweden's long-term care system is built from several distinct layers: national legislation and policy, municipal social services, municipal healthcare, regional healthcare, public and private providers, professional responsibilities and national oversight. Its strength lies partly in distributing decisions close to communities while retaining a broad public commitment to supporting older people according to need.

That same architecture creates the system's central operational challenge. Older people increasingly live with needs that cross every institutional boundary. A municipal assessment may lead to home help, but effective support can also depend on rehabilitation, nursing, primary healthcare, hospital services, housing, technology and family involvement. Quality therefore emerges from relationships between organisations as much as from the performance of individual services.

Population ageing will make those relationships more consequential. Municipalities will need to manage greater complexity while sustaining workforce capacity and financial resilience. Regions will need to work effectively with municipal services around people whose health and care needs are inseparable. National agencies will need to convert data, evidence and supervision into learning that supports equitable local delivery.

Sweden's experience demonstrates that decentralisation and universal welfare are not opposites. They can reinforce one another when local flexibility is matched by clear responsibility, knowledge, transparent outcomes and effective oversight. The next stage of Sweden's long-term care development will therefore depend less on redrawing every institutional boundary than on ensuring that the boundaries already present do not become barriers in the life of the person who needs support.