Schools, the state and equality in Sweden

Background report: a historical timeline of decentralisation, the current division of responsibilities and funding, quality and equality impacts, and proposals for a new national school system with the selective return of higher education.

responsible intelligence in politics

Schools, the state and equality in Sweden

Executive summary

The decentralisation of Sweden’s school system did not happen in a single step, but through a chain of reforms from the late 1980s and early 1990s: state-regulated teaching posts were dismantled, municipalities assumed overall operational and employer responsibility, government grants were restructured, management by objectives and results replaced parts of the older rule-based system, and free school choice was combined with public funding for independent schools. In higher education, decentralisation took a different form: not municipalisation but increased institutional autonomy, particularly through the 1992/93 reform “freedom for quality”, the establishment of foundations for Chalmers and Jönköping University, and the autonomy reform of 2010/2011. Schools thus became more municipal and market-driven; higher education remained largely state-run but became more self-governing.

The main political motives were local self-government, proximity to citizens, more efficient use of resources, greater professional discretion and increased freedom of choice. Over time, Swedish inquiries, the Swedish National Agency for Education, the OECD and several research studies have indicated that the reforms also coincided with or contributed to widening differences between schools, greater school segregation, a stronger influence of socio-economic background, and uncertain quality in governance and monitoring. At the same time, the research evidence is not clear-cut: some studies find positive or small average effects from the expansion of independent schools, while others show increased sorting, more generous grading and negative effects for certain groups of pupils or educational choices.

The key analytical conclusion is that the issue should not be framed as “centralisation or decentralisation” in general. Local autonomy works best when surrounded by strong, competent and legitimate national capacity: clear common rules, robust funding, equivalent admissions and assessment systems, professional skills provision and transparency in monitoring. Where such counterweights are weak, decentralisation and school choice tend to widen differences between providers, schools and groups of pupils.

The report’s main proposal is therefore a new national school system for publicly funded preschool class–upper secondary education and adapted forms of schooling, in which the state resumes employer responsibility, provider responsibility, national admissions and most of the funding. Municipalities retain responsibility for premises, school meals, pupil transport and coordination with social services and leisure activities. Independent schools should eventually either be converted into non-profit-distributing, state-contracted specialist schools or phased out of the publicly funded core service. In higher education, the state should not re-centralise the entire sector, but selectively return teacher training, special needs teacher and special educational needs training, headteacher training and practice-oriented professional provision to a clearer national service model.

Starting point and terminology

In Swedish education policy, decentralisation means different things in different sub-systems. In schools, it primarily refers to the transfer of operational responsibility, the employer role and resource decisions from the state to municipalities, combined with management by objectives and results and, later, market elements through school choice and independent schools. In higher education, decentralisation instead refers to higher education institutions deciding their own organisation, range of courses, internal allocation of resources, admissions numbers and research direction within a state framework. To speak of “returning” the university system to the state is therefore misleading if ownership is what is meant.

It is also important to distinguish between system responsibility, provider responsibility, employer responsibility and professional autonomy. In today’s schools, the state has system responsibility through legislation, curricula, government agencies and inspection; the municipality or individual provider carries provider responsibility; the headteacher is responsible for pedagogical leadership; and teachers’ professional autonomy is embedded in this multi-level system.

The report therefore follows a principle of selective re-centralisation:

things that require national equality, economies of scale, long-term skills provision and legitimate accountability should be brought closer to the state
things that require academic freedom, local relationships with pupils and families or professional judgement should not be micro-managed from the centre

Historical timeline of decentralisation

Sweden moved within a short period from a relatively centralised school system to one in which governance became both more local and more dependent on the market. In higher education, the movement was more gradual: first mass expansion and a more unified sector, then increased institutional freedom and later stronger university leadership.

1977 – Higher Education Reform: the higher education sector was expanded and systematised nationally. Motive: broader access and a more integrated sector.
1989 – Prop. 1989/90:41: state regulation of teaching and school leadership posts was dismantled. Motive: municipal employer responsibility and greater control over staffing.
1990 – Prop. 1990/91:18: the state sets objectives and frameworks, while municipalities are responsible for implementation; older state school bodies were dismantled. Motive: clearer responsibility and less micro-management.
1992 – Prop. 1991/92:95: independent schools were given more equivalent conditions. Motive: freedom of choice and more alternatives.
1993 – Prop. 1992/93:230: the freedom-of-choice reform was broadened.
1992/93 – Prop. 1992/93:1 and the Higher Education Act 1992:1434: increased freedom over educational provision and internal organisation within a state framework (“freedom for quality”).
1993 – Chalmers and Jönköping University were transferred to foundation status.
2010/2011 – Prop. 2009/10:149: extensive deregulation of internal organisation and academic employment in higher education.
2020–2022 – Dir. 2020:140 and SOU 2022:53: the state investigates state provider responsibility or strengthened state responsibility for schools.

An important observation is that decentralisation coincided with a deep economic crisis in the early 1990s. The OECD’s review of Sweden emphasises that the general direction of the reforms did not necessarily lower results, but that shortcomings in their design and implementation, combined with the recession, probably contributed to weaker outcomes. Implementation can be just as decisive as the decision in principle.

Current division of responsibilities and funding

The Riksdag and the Government determine what pupils should learn through the Education Act, ordinances and curricula. The municipality is the provider for municipal schools, while the board is the provider for independent schools. The provider is responsible for resources, quality work and compliance with regulations; the headteacher leads the educational work. Alongside this, state agencies are responsible for standard-setting, inspection, special educational support, research and certain special forms of schooling.

Independent schools operate within a publicly funded system but are established through permit applications assessed by the Swedish Schools Inspectorate. In the 2025/26 academic year, 17 per cent of compulsory schools had an independent provider, and just over 179,000 compulsory-school pupils attended independent schools; 74 per cent of these pupils attended schools whose provider was a limited company.

Schools are still funded mainly from public sources, but the funding route has several strands. Municipalities carry the main cost through municipal taxation, while the state supplements this with general and targeted government grants. The Swedish National Agency for Education’s analysis of the grant for strengthened knowledge development shows that the state to some extent distributes more funding to providers with greater needs, but that needs-based allocation is often weakened when providers pass the money on to schools. This is a central argument for increased state financial power.

For compulsory schools, municipalities’ total costs amounted to approximately SEK 152 billion in 2023. Across the entire school system, including preschool, leisure-time centres, schools and municipal adult education, the cost amounted to SEK 368.3 billion in 2024; compulsory schools accounted for almost SEK 160 billion and preschools for just under SEK 94.4 billion. Returning responsibility to the state is therefore not primarily a question of finding new money, but of moving and reshaping very large existing public flows.

In higher education, the Riksdag and the Government have overall responsibility, but higher education institutions are described by the Swedish Higher Education Authority as largely autonomous. In 2024, direct government funding for first- and second-cycle education amounted to SEK 31.2 billion. In 2025, total research income was SEK 56.8 billion, of which SEK 23.8 billion consisted of direct government funding.

Quality and equality impacts

At system level, Sweden still has relatively strong average results internationally, but equality has weakened. In PISA 2022, Swedish pupils performed above the OECD average in mathematics, reading and science, but results fell back after 2018, socio-economic background has become more influential, and the differences between pupils with and without an immigrant background are among the larger ones in the OECD.

The Swedish National Agency for Education’s report on equality at the end of compulsory school shows that differences in school results and school segregation have stabilised – but at a higher level than twenty years ago. The agency judges that the increase is due both to residential segregation and to school choice. School choice does not operate in a vacuum; it reinforces or channels differences that already exist.

Research offers a more nuanced picture than the debate. Böhlmark and Lindahl find positive average effects of a growing independent-school sector on certain educational outcomes, while Holmlund, Böhlmark and Lindahl show that the increased freedom of choice after 1992 is associated with greater school segregation even when residential segregation is taken into account. The conclusion is that the reform improved certain outcomes while also carrying clear sorting costs.

At upper-secondary level, the research evidence is more critical. Tyrefors and Vlachos find that pupils at independent upper-secondary schools perform, on average, 0.06 standard deviations lower on externally marked tests, while internal assessments are 0.14 standard deviations more generous. IFAU’s 2025 report shows that greater exposure to independent schools is associated with a higher probability of choosing a vocational programme, a lower probability of choosing a natural sciences programme, and possible long-term negative effects on STEM orientation and employment at age 30.

Internationally, the picture is similarly mixed. Decentralisation is a high-capacity tool: it can be productive in systems with a strong profession, transparent monitoring and a legitimate chain of accountability, but risky where these conditions are unevenly distributed. The question is therefore less “should the state govern more?” than “which functions must the state carry so that the local level can function equitably at all?”.

The particular logic of higher education

Sweden already has a largely state-run university system. The central problem is therefore not ownership, but how autonomy, quality, regional accessibility and professional skills provision are balanced. Research on the autonomy reform shows that it contributed to stronger line management and weaker collegial bodies – a process of “decollegialisation”. A careless re-centralisation could risk solving the wrong problem.

The conclusion is that “returning parts of higher education to state service” should be interpreted functionally: teacher training, special needs teacher and special educational needs training, school leadership training, work-based learning infrastructure and practice-oriented school research should be given a clearer national service model, because these are precisely the parts that support equality and skills provision in schools.

Interpreting voices

What follows consists of source-based interpretations, not claims that Jan Stenbeck, Olof Palme or Ellen Key actually took a position on today’s institutional model.

Jan Stenbeck

Stenbeck is associated with challenging public monopolies through technology, competition and new business models. A reasonable interpretation is that he would have defended competition, a diversity of providers, digital transparency and performance management – but also demanded greater openness of information, simpler rules and clearer sanctions against poor performers than today’s system offers.

Olof Palme

Palme repeatedly returned to the school as a democratic institution and to education as a matter of real – not merely formal – equality. If municipalisation and the school market undermined equality, he would probably have favoured stronger state responsibility, but for him centralisation would have been legitimate only if it strengthened equality, participation and social mobility.

Ellen Key

Key criticised the harsh, standardised school and placed the child at the centre of pedagogy. She would probably have been critical both of a market-driven school that treats pupils as customer flows and of a state model that reduces teaching to standardised performance. Her voice argues for state-backed equality without soulless uniformity.

Proposal for a new national school system

The status quo is not enough to repair the structural problems. Merely strengthening state responsibility without changing the provider structure leaves the fundamental problem intact: quality, skills provision and resources are filtered through very different providers with very different capacities. The most robust solution is a national core structure for schools combined with the selective re-steering of higher education.

Objectives

equal quality – comparable minimum conditions for teacher density, teaching materials, pupil health, special educational needs provision and school leadership
clear responsibility – the state becomes directly responsible for publicly funded core services
professional stability – a unified pay and career structure for teachers and headteachers
full accessibility – universal design plus individual support

Governance

From 1 July 2028, the state becomes the provider and employer for all publicly funded preschool classes, compulsory schools, adapted compulsory schools, special schools, Sami schools and upper-secondary schools currently run by public providers. The organisation should consist of a new agency – the National School Agency – with regional school areas. The headteacher remains the pedagogical leader, but headteachers and teachers become state employees. Municipalities retain responsibility for premises, school meals, pupil transport, and leisure and social coordination through legally regulated service agreements.

Within a seven-year transition period, independent schools should either become state specialist schools without profit distribution and with the same admissions and transparency rules as other publicly funded schools, or become wholly privately funded operations outside the per-pupil funding system.

Funding

a national basic grant per pupil for teaching, pupil health, school leadership and teaching materials
a compensatory weighted supplement based on socio-economic indicators, language profile, sparsity costs and documented support needs
a small number of strategic development grants for STEM, vocational education, culture, languages and special educational needs provision

Staffing policy

When teachers become state employees, the state can build a unified pay and career structure, guarantee national professional development and direct reinforcements where recruitment problems are greatest: a national minimum salary, recruitment allowances for sparsely populated areas and disadvantaged schools, a strong professional programme and a clear career ladder.

Curriculum and assessment

The curriculum remains national, but micro-management that has not improved quality should be reduced. Assessment becomes more legally secure through external moderation of national tests, sample-based external marking in more subjects and a national data repository for comparing the relationship between tests and grades over time and between schools.

Accessibility for people with visual impairments and dyslexia

The system should include a national accessibility requirement for all publicly funded teaching materials, central licensing of accessible learning tools, guaranteed access to Braille and tactile resources, and compulsory further training for school leaders and teachers on visual impairment, reading and writing difficulties, and assistive technology.

Selective return of higher education

A national professional skills provision pathway should cover teacher training, special needs teacher and special educational needs training, headteacher training, school library and teaching-materials expertise, and practice-oriented school research – funded through multi-year national assignments with higher core resources, clear quality requirements and links to state teaching schools.

Implementation plan

Preparation 2027–2028: new provider legislation, new agency structure and transitional provisions; inventory of staff, agreements, data, premises and pension issues; national professional assignments for higher education institutions.
Transfer 2028–2029: the state takes over employer responsibility for teachers, headteachers and the core professions in pupil health services; regional school areas are established.
System change 2029–2031: new funding model, national admissions and placement system with compensatory rules, expanded external moderation of grades and tests.
Consolidation 2031–2033: evaluation of teacher supply, pupil outcomes, segregation and accessibility.

Cost estimate and assumptions

This calculation is an original policy estimate, not an official budget. The reform is estimated to require one-off costs of approximately SEK 8–12 billion over six years for the HR transfer, data and IT systems, a national admissions platform, harmonisation of agreements, accessible teaching materials and transition support. In addition, there would be a recurring net cost of approximately SEK 4–7 billion per year during the first decade. The reform is expensive but manageable in relation to today’s total education expenditure.

Risks and how they should be managed

The state takes over responsibility without sufficient regional and professional capacity – the solution is regional state school areas with real mandates, not a single super-agency in Stockholm.
Coordination around pupils with extensive support needs breaks down – premises, pupil transport and social coordination should therefore remain close to municipalities and regions through state-regulated agreements.
Re-centralisation damages academic freedom – the return of higher education functions must be limited to professional skills provision, regional presence and core resources.

The final assessment: Sweden should return schools to the state because schools are an equality institution whose responsibility is currently too fragmented. Sweden should not, however, return the whole of higher education to the state, because it is already largely state-run. The practically sustainable route is a national school system, municipal service links and a selectively nationally governed professional higher education sector.