What is Education | Education as a Public Good and Social Contract — Why Civilisations Teach Beyond Private Benefit

Education as a public good, education as a common good, the right to education, public education, the social contract for education and the wider social benefits of education all point to a question larger than schooling: why does a civilisation organise, finance, regulate and protect learning when much of the immediate benefit appears to belong to the individual learner? A person who becomes literate can earn, communicate and navigate institutions more effectively. A graduate may gain access to occupations and income. Those are private returns. Yet education also changes the quality of work, parenting, public reasoning, professional standards, health decisions, scientific capability, cultural memory and the competence of institutions. Those effects do not stop at the learner’s front door.

That is why searches for why education is a public good, education as a common good, public versus private education, benefits of public education, social returns to education, right to education, social contract in education, education policy, education and democracy, and collective benefits of schooling cannot be answered by saying simply that education is “good for society.” The serious work is to identify which benefits are private, which spill over to others, which require collective coordination, which can be provided through markets, which require public guarantees, and which choices remain legitimately contested. Education is not one economic object, one political institution or one moral claim. It is an ecosystem in which private choice and public consequence are tightly entangled.

The central proposition of this article is therefore narrower than a slogan and broader than a funding argument: civilisations teach beyond private benefit because education produces capabilities, relationships and institutional conditions whose value is shared, cumulative and often impossible for one buyer or provider to capture alone. This does not imply that every school must be state-operated, that every educational decision should be centralised, or that private investment in learning is unimportant. It means that a civilisation has reasons to care about who learns, what can be learned, whether access is meaningful, how quality is judged, and whether the knowledge system remains trustworthy across generations.

---

The first distinction: private benefit is real

Any account of education as a public or common good becomes weak if it denies the obvious: individuals often benefit directly from education. Literacy allows a person to read contracts, instructions and information. Numeracy allows them to compare prices, understand interest, estimate quantities and interpret evidence. Technical education can open occupational pathways. Higher education can provide specialised knowledge, professional identity and networks. Language learning can widen mobility. These benefits can be experienced personally, and many can be reflected in income, employment, autonomy and choice.

Private benefit matters because it shapes motivation. Families invest time and money in education partly because they hope a child will have wider future options. Adults undertake demanding study because they expect a transition, promotion, profession or intellectual gain. Employers fund training because they need capability. Pretending these motives are somehow impure would misunderstand ordinary human agency. Education systems work partly because private incentives and public purposes can reinforce one another.

The analytical problem begins only when private benefit is treated as the whole story. A literate nurse does not keep the entire value of literacy. Patients benefit. Colleagues benefit. Records become safer. A competent electrician creates value for every building user who never meets the electrician. A teacher’s knowledge changes children who later become adults in other institutions. Education generates effects that move through networks. Those spillovers are where the civilisational argument begins.

Public good is an economic term before it is a compliment

In economics, a pure public good is usually described through two properties: non-rivalry and non-excludability. One person’s use does not substantially reduce another person’s use, and it is difficult to prevent non-payers from benefiting. National defence is a familiar textbook example. Education does not fit this pure category neatly. A classroom seat can be rival when capacity is limited. A course can be excludable through fees, admissions or geography. A tutor’s hour cannot be consumed by unlimited people at the same level of attention.

So when people say “education is a public good,” they may mean different things. Some are making a technical claim about positive externalities: education produces benefits beyond the learner. Some are making an institutional claim: public authorities should guarantee access or finance provision. Some are making a moral claim: education should not depend entirely on purchasing power. Others use “public good” in the everyday sense of something beneficial to society. These meanings overlap, but they should not be collapsed.

This distinction helps keep the debate honest. If education were a pure public good in the strict economic sense, policy design would be simpler. It is not. Real education uses scarce teachers, buildings, technology, time and attention. The public question is therefore not whether scarcity exists. It is how a society decides to allocate scarce educational resources when the consequences of allocation extend far beyond the immediate transaction.

Externalities: when the learner is not the only beneficiary

An externality occurs when an activity affects people who are not fully represented in the transaction. Education can create positive externalities when one person’s increased capability improves outcomes for others. A better-trained worker may reduce accidents. A health-literate parent may make decisions that benefit children. A scientifically literate population may evaluate risk more competently. A skilled technician can improve the reliability of infrastructure used by thousands. A literate voter can read primary documents rather than depend entirely on intermediaries. None of these effects belongs neatly to one buyer.

Externalities help explain why purely private purchasing can underprovide some forms of learning. If I pay for a course, I consider the benefits I expect to receive. I may not be willing or able to pay for the benefits my future competence creates for strangers. An employer may invest in firm-specific training but hesitate to finance broad transferable learning if the employee can later leave. A family may value education highly yet be unable to finance the level of learning that would also create broader public benefits.

Public funding is one possible response because it can pool resources across beneficiaries who cannot coordinate individually. It is not the only response. Professional associations, philanthropy, employer consortia, community institutions and mixed funding can also support learning with wider benefits. The principle is simply that education often creates value outside the purchaser-provider relationship, so the funding architecture may need to reach beyond that relationship too.

The common-good frame is related but different

UNESCO has used the language of education as a common good to push beyond a purely instrumental treatment of education as individual investment. In that framing, education is not only something people consume individually; it is also a shared social practice shaped by relationships, participation, culture and collective responsibility. UNESCO’s Futures of Education work explicitly distinguishes the common-good idea from a narrow transaction model and connects it with the right to quality education throughout life and the protection of shared knowledge resources.

The distinction is useful because some educational outcomes exist only relationally. A shared language is valuable partly because other people share it. Trust in a qualification depends on collective standards. A classroom culture depends on how participants treat one another. Civic norms are meaningful because they coordinate people. A body of scientific knowledge becomes more valuable when it is preserved, criticised and extended by a community. These are not simply private commodities lined up on a shelf.

At the same time, “common good” is a normative and governance frame, not a magic policy formula. People can disagree about curriculum, institutional roles, public funding levels, school autonomy, parental choice and the place of private providers while still accepting that education has shared consequences. The concept is most useful when it opens those questions rather than pretending it has already answered them.

What a social contract means in education

A social contract is not necessarily a literal document signed by every citizen. In education, the phrase describes the web of expectations, rights, obligations and institutional arrangements through which a society decides what education is for and who is responsible for sustaining it. Families expect schools to teach. Teachers expect professional support and legitimate authority. Students are expected to participate while also holding rights. Governments collect revenue and set rules. Employers rely on qualifications. Universities certify specialised knowledge. Communities expect children to emerge capable of living with others.

UNESCO’s 2021 Reimagining our futures together report uses the social-contract language to invite renewed debate about shared educational purposes under changing technological, ecological and social conditions. That is a proposal and framing, not a binding global blueprint. Its value for analysis is that it makes invisible assumptions visible. Every education system already contains an implicit contract, even when nobody uses the phrase.

When the implicit contract becomes unstable, familiar disputes intensify. Who decides curriculum? How much discretion should teachers have? What responsibilities belong to parents? What happens when labour markets change? How should digital platforms enter schools? What should be guaranteed as a right? Which costs should be borne collectively and which individually? These are not peripheral administrative questions. They are negotiations about the boundary between private choice and shared obligation.

Education is intergenerational by construction

One reason education resists purely private framing is that generations overlap. Adults finance schools they no longer attend. Teachers transmit knowledge they inherited from earlier teachers, researchers and cultures. Children learn before they can finance their own education. Universities preserve knowledge whose creators are dead and teach students whose future contributions cannot yet be priced. Education constantly moves resources, knowledge and responsibility across time.

This intergenerational structure creates a civilisational bargain. The present generation maintains institutions whose benefits will partly appear after its members have left office, retired or died. Future engineers will maintain bridges not yet built. Future doctors will use research conducted today. Future citizens will inherit archives, laws, languages and professional standards. Education is one mechanism by which the present invests in people it cannot fully know.

That long horizon is difficult for markets and politics alike because both can be pulled toward immediate returns. Strong education systems therefore need institutions capable of remembering long-term purposes while still responding to current evidence. Civilisation is partly the art of making promises across time. Education is one of the places those promises become concrete.

Children make the market analogy incomplete

Many markets assume a buyer who can judge preferences, compare offers and walk away. Children complicate that model. A six-year-old cannot reasonably evaluate the long-run consequences of remaining illiterate, compare teacher quality across institutions or finance a future-oriented investment from expected lifetime earnings. Parents act on behalf of children, but families differ in information, income, time, language and ability to navigate institutions.

This is one reason societies often use compulsory education laws, public financing and minimum standards. The justification is not that families have no legitimate role. Quite the opposite: families are central. The issue is that a child’s future capability has consequences for the child that are too large to leave entirely to short-term preference or unequal purchasing power. The child is also a future adult whose rights and opportunities extend beyond the household’s present circumstances.

Compulsion itself raises legitimate questions about curriculum, freedom, alternative forms of education and the scope of state authority. Those questions should remain discussable. The important analytical point is that education policy must account for a learner who is developing the very capacity for autonomous choice that adult market models often presume.

The right to education changes the baseline

International human-rights frameworks treat education as a right, not merely a consumer preference. UNESCO’s 2025 review of the right to education emphasises both long-standing access gains and new pressures created by digitalisation, conflict, migration and demographic change. The rights frame asks a different first question from a market frame. Instead of beginning with “Who can buy what?”, it begins with “What educational access and protection should every person be able to claim?”

Rights do not automatically solve implementation. A declared right requires teachers, buildings, materials, transport, language access, disability accommodation, finance, governance and mechanisms for remedy when provision fails. It also raises questions about quality. A child can be physically enrolled while receiving little meaningful learning. Access without capability is an incomplete fulfilment of education’s purpose.

The rights frame therefore creates a floor rather than a complete system design. Above that floor, societies still make choices about curriculum, provider diversity, assessment, vocational pathways, higher education, adult learning and funding. The floor matters because it prevents education from being treated only as a luxury whose absence is unfortunate but acceptable.

Public financing and public provision are not the same decision

Debates about education often bundle several choices together. Who pays? Who owns the institution? Who employs the teachers? Who sets standards? Who chooses the school? Who assesses results? These can be separated. A publicly financed system can include non-state providers. A privately financed programme can operate under public regulation. A public school can purchase private technology. A private school can teach a public curriculum and participate in public examinations.

Separating the decisions allows trade-offs to be analysed more clearly. Public financing can widen access and pool risk. Provider diversity can allow experimentation or specialised missions. Public provision can support universal coverage where markets would be thin. Regulation can protect minimum standards but can also become burdensome if poorly designed. Choice can increase fit for some families while also interacting with information gaps, transport constraints and selection incentives.

No single organisational form guarantees quality. The civilisational question is whether the overall architecture produces reliable learning, fair access, trustworthy standards and enough diversity to serve real variation without fragmenting the common knowledge and institutions on which people depend.

Markets can contribute to education

Private markets can expand supply, create specialised services, respond quickly to demand and provide options that public systems do not offer. Tutoring, publishing, educational technology, professional training, language instruction and private schools all illustrate areas where families and organisations purchase education directly. These markets can generate innovation and convenience. They can also reveal needs that large public systems have overlooked.

The presence of market provision does not erase public questions. Consumers may struggle to judge quality before purchase. Learning outcomes can take years to appear. Providers may have incentives to market credentials more aggressively than capability. Profitable markets may cluster where purchasing power is high. Data collected by digital services can create privacy concerns. A society therefore needs ways to distinguish useful enterprise from exploitation without assuming that every private provider is either superior or suspect.

The most productive analysis asks what problem the market mechanism solves and what problems it creates. Price signals can coordinate supply, but they do not automatically represent public value. Competition can encourage responsiveness, but it can also encourage selection of easier-to-serve learners. Choice can be empowering, but only when information and access make the choice real.

Public systems can fail too

Calling education a public good does not guarantee that public institutions will deliver it well. Public systems can become bureaucratic, slow to correct, uneven across regions, vulnerable to political cycles or overly standardised. Funding can be allocated by historical habit rather than current need. Accountability can focus on paperwork rather than learning. Central rules can protect fairness in one context and create rigidity in another.

This matters because a serious public-good argument includes obligations on public institutions. If citizens are asked to finance education collectively and families are asked to trust it, the system owes them competence, transparency and mechanisms for correction. Public purpose is not a substitute for performance. A publicly funded failure still fails learners.

The right comparison is therefore not ideal public provision versus flawed markets, or ideal markets versus flawed public bureaucracy. Real systems combine institutions with different strengths and failure modes. Governance is the discipline of making those strengths complementary while containing predictable weaknesses.

Education creates a shared language for complex societies

Large societies require strangers to coordinate. They use common units, legal concepts, calendars, maps, scientific conventions, technical standards and languages of public administration. Education introduces people to these shared representations. A person does not need identical beliefs to every neighbour, but they need enough common symbolic infrastructure to sign a contract, follow a safety sign, interpret a timetable and understand the basic rules of institutions.

This common layer should not be confused with cultural uniformity. Multilingual and plural societies can maintain shared civic and technical frameworks while preserving multiple traditions, histories and identities. The design challenge is to build interoperability without demanding sameness. Education is one of the places that negotiation occurs because curriculum always answers, implicitly or explicitly, “What should people living together be able to understand in common?”

The answer can change over time. Digital literacy became more important as services moved online. Statistical reasoning became more visible during data-rich public debates. Environmental knowledge matters differently as climate risks change. A shared curriculum is therefore not a museum. It is a maintained interface between individual development and collective life.

Qualifications are public trust instruments

A qualification is valuable partly because people who did not witness the learning process trust what the qualification represents. An employer cannot sit through every class an applicant attended. A patient cannot personally test a clinician’s entire education. A regulator cannot observe every hour of an engineer’s formation. Credentials compress information so strangers can coordinate.

This gives qualification systems a public function even when the institution awarding the credential is private. Standards, assessment integrity, accreditation and professional regulation protect the meaning of the signal. If credentials can be purchased without competence, everyone who relies on the signal bears the risk. The value of a qualification therefore depends on a collective reputation that no single graduate owns.

The same logic explains concern about credential inflation. When employers demand more qualifications merely as filters, individuals pay for longer education without a corresponding increase in job capability. The public-good question is not simply how many credentials a system produces. It is whether credentials continue to represent useful, trustworthy distinctions.

Research and universities generate spillovers far beyond graduates

Universities educate students, but many also produce research, preserve archives, maintain specialist expertise, operate laboratories, support professional standards and contribute to public debate. These functions complicate simple tuition-based accounting. A discovery may benefit people who never enrolled. A historical archive may serve future researchers. A medical school may strengthen a health system through training and research simultaneously.

Basic research is a particularly clear case of uncertain return. Its applications may be distant, indirect or impossible to identify in advance. Private firms can and do conduct important research, especially where intellectual property or product advantage allows returns to be captured. Public and philanthropic funding often support areas where knowledge has broad potential value but uncertain near-term commercial payoff.

The education system thus overlaps with the knowledge-production system. Teaching transmits what is known; research challenges and extends what is known. Civilisations need both. A system that teaches brilliantly but stops producing new knowledge eventually becomes dependent on others. A system that produces research but cannot teach or translate it struggles to convert discovery into widespread capability.

Vocational education creates public capacity, not only private careers

A plumber, electrician, machinist, nursing assistant, laboratory technician, transport operator or construction supervisor earns a living from skill. That is private benefit. But the skill also supports systems on which strangers rely. Water remains safe, electrical faults are repaired, factories run, samples are handled correctly and buildings meet standards. Technical competence is part of civilisation’s operating capacity.

This is why the status hierarchy that places academic education above vocational learning can damage public capacity. If talented young people are discouraged from technical pathways because society treats them as second class, essential occupations become harder to staff and maintain. The public value of a skill is not measured by the social prestige of the classroom in which it was learned.

Good vocational systems also solve a coordination problem between employers and learners. Individual firms may be reluctant to train broadly if competitors can hire the worker afterward. Shared apprenticeships, sector standards and public co-funding can distribute that cost. The result is a pool of capability from which the whole economy benefits.

Teacher capability is a shared asset

A teacher’s skill affects many learners over a career. Professional development therefore creates a multiplying effect. Improving one teacher’s ability to explain fractions, diagnose reading difficulties, manage discussion or use formative assessment can influence hundreds or thousands of students. This is another example of why education investment cannot be evaluated only at the level of one transaction.

Teacher quality also depends on working conditions, curriculum coherence, leadership, class composition, materials and time. Treating teaching entirely as an individual attribute can obscure the system that enables or constrains performance. A highly capable teacher inside a chaotic environment may spend expertise compensating for preventable institutional failure.

The public-good view therefore extends from paying teachers to maintaining a profession: recruitment, preparation, induction, mentoring, career development, professional standards and enough institutional memory that good practice can survive turnover. The profession itself is part of the educational commons.

Equity is not the same as identical treatment

If education matters for future capability, unequal access can compound. But equal budgets or identical programmes do not necessarily create equal access. A rural student may face transport barriers. A student with a disability may require accessible materials. A recent migrant may need language support. A child from a low-income family may have less access to quiet study space, devices or enrichment outside school.

Equity therefore asks what different learners need to reach meaningful participation and capability. This can justify unequal resources in order to address unequal barriers. The design is difficult because additional support should not become lowered expectation. The aim is to change access conditions while preserving the seriousness of the learning goal.

There are legitimate debates about how far redistribution should go, how needs are measured and how targeted programmes interact with universal provision. A neutral systems analysis does not erase those disagreements. It clarifies the trade-off: a society must decide whether equal formal opportunity is enough when starting conditions are substantially unequal.

Geography can turn a formal right into a practical impossibility

A school place sixty kilometres away is not equivalent to a school place down the road. Distance creates transport cost, time cost, safety concerns and attendance risk. Thinly populated areas may struggle to sustain specialist teachers, advanced courses or vocational facilities. Urban areas may have many providers but severe differences between neighbourhoods.

Public systems often respond through school transport, boarding arrangements, regional hubs, digital delivery, teacher incentives or multi-grade models. Each solves one problem while potentially creating another. Online learning can widen subject access but depends on connectivity, devices and independent-learning capacity. Regional consolidation can improve course range but lengthen travel.

The public-good lens makes geography visible because the system’s obligation is not satisfied merely by announcing that education exists somewhere. Access is relational: a learner must be able to reach, participate and benefit. Infrastructure beyond the school gate becomes part of education policy.

Disability reveals the difference between provision and access

An education system can be open in principle and inaccessible in practice. Stairs, uncaptioned video, incompatible digital platforms, inflexible assessment, inaccessible transport or lack of specialist support can exclude learners without any rule explicitly saying they are barred. Disability therefore shows why rights need operational design.

Universal design can reduce some barriers before individual accommodation is required. Clear structure, captions, readable materials, multiple means of access and accessible physical environments benefit many learners. Individual supports remain necessary where needs are specific. The important distinction is between changing the route to participation and changing the intellectual standard without justification.

From a civilisational perspective, exclusion also wastes capability. When a person is prevented from learning because an avoidable barrier was built into the system, society loses potential contribution as well as violating the learner’s interest. Inclusion is therefore both a rights question and a capacity question.

Migration turns education into infrastructure for belonging

Migration places learners between systems. Records may be incomplete. Curricula differ. Qualifications may not map neatly. Language can block access. Children may have experienced interrupted schooling, while adults may possess skills that are difficult for employers to recognise. Education becomes a bridge between prior capability and a new institutional environment.

Language-of-instruction support is one obvious requirement, but belonging is wider than language. New learners need orientation to school norms, pathways, rights and expectations. Schools also need the capacity to understand previous learning rather than assuming that unfamiliar records equal absence of knowledge. Recognition and diagnostic assessment can prevent both unnecessary repetition and unsupported placement.

Here again, benefits extend beyond the individual student. Successful educational integration can improve communication between families and public services, reduce wasted skill and create more durable connections across communities. Poor integration can leave capability stranded and deepen social separation.

Education and social cohesion: shared institutions without forced sameness

Schools bring people into repeated contact under common rules. That can create shared experience, but contact alone does not guarantee cohesion. If schools are perceived as unfair, segregated or humiliating, they can reinforce distrust. If they create legitimate procedures, opportunities for cooperation and a sense that different learners belong, they can contribute to the social capital that lets strangers coordinate.

Social cohesion should not mean suppressing disagreement. A cohesive plural society can contain deep differences. What matters is whether disagreements occur inside institutions people can continue to share. Education contributes when students learn to distinguish evidence from identity, argument from attack, rule from arbitrary power and compromise from surrender.

These outcomes are difficult to purchase individually because they depend on interaction among people. My capacity to cooperate with strangers is partly shaped by whether institutions create situations in which cooperation is practised and reciprocated. That relational quality is one reason education fits the common-good frame.

Civic education is not partisan instruction

Civic education deals with institutions, rights, responsibilities, public reasoning and forms of participation. It becomes dangerous when confused with teaching students which political choices to make. The educational job is different: help learners understand how institutions work, evaluate claims, identify evidence, recognise legitimate disagreement and participate without treating opponents as enemies.

OECD work published in 2025 examines civic education as a pathway to civic and social engagement and notes persistent participation gaps associated with background characteristics. The descriptive lesson is not that one curriculum produces one political outcome. It is that participation itself depends on knowledge, confidence and opportunity, and those capabilities are distributed unequally.

A neutral civic curriculum can therefore be rigorous without being politically directive. Students can learn how laws are made, how budgets work, how statistics can be misread, how sources are verified, how public institutions can be challenged lawfully and how competing values create real trade-offs. Education supports agency by making the machinery legible, not by choosing a side for the learner.

Health is another route through which education creates shared benefit

Education can affect health through multiple pathways: literacy and numeracy shape the ability to understand instructions; qualifications can affect work and income; school environments can influence habits and social networks; scientific literacy can affect interpretation of health claims. Health, in turn, affects educational participation. The relationship is bidirectional and intertwined with wider social conditions.

From a public perspective, health literacy creates spillovers. A person who understands infection control, medication safety or emergency guidance can make decisions that affect others. A trained health workforce is an obvious public capacity. These examples do not mean education alone determines health. They show that the boundary between “education policy” and “health policy” is more porous than administrative departments suggest.

This cross-sector effect is one reason narrow return-on-investment calculations can miss part of education’s value. If the only counted outcome is wage gain, improvements in family health decisions, civic competence, safety or social trust disappear from the ledger even though civilisation experiences them.

Culture and memory make education more than workforce preparation

Education carries languages, histories, literature, scientific traditions, craft knowledge, music, rituals and ways of interpreting the world. Some of these have direct economic value. Others matter because they allow a community to remember itself. A civilisation that trained only for current vacancies would be productive in a narrow sense and dangerously forgetful in a broader one.

Cultural transmission creates hard questions because no curriculum can include everything. Selection is unavoidable. Whose history appears? Which languages are taught? What balance should exist between inherited canon and new knowledge? These are legitimate public disagreements. The existence of disagreement is not evidence that cultural education has no public purpose; it is evidence that the purpose is important enough to contest.

Libraries, archives, museums and schools form a distributed memory system. Their value often appears decades later when a society needs records, historical comparison or language continuity. Markets can support many cultural goods, but preservation often requires institutions willing to maintain materials with uncertain immediate demand.

The knowledge commons is civilisation’s shared inheritance

No learner starts from zero. Mathematics, alphabets, scientific methods, maps, legal concepts and technical standards were accumulated across generations. Education gives individuals access to this inheritance. UNESCO’s Futures of Education vision explicitly links education with protecting and enriching a knowledge commons: shared knowledge resources created, preserved and transformed over time.

The commons idea does not mean every piece of information must be free of cost or ownership. Producing knowledge requires labour, and intellectual-property systems can create incentives for investment. The public question is whether enough knowledge remains accessible, preservable and contestable for civilisation to learn. A society that locks every foundation behind private gates can raise the cost of future discovery; a society that refuses any mechanism for creators to sustain work can weaken production. Balance matters.

Open educational resources, public libraries, publicly accessible research, archives and educational broadcasting are different tools for widening access to the commons. Their value increases when they are curated, searchable, translated and connected to teaching. Availability is not the same as usability.

Digital platforms complicate the social contract

When learning moves through digital platforms, new actors enter the educational contract. Software companies may host assignments, store student data, recommend content, score responses or provide generative tools. These services can increase access and reduce administrative friction. They can also create questions about privacy, commercial incentives, algorithmic bias, interoperability and dependence on proprietary infrastructure.

The public-good question is not whether technology is public or private. It is whether the system preserves educational interests when commercial and educational incentives differ. Who owns the data? Can a school leave the platform? Are accessibility standards met? Can teachers inspect or challenge automated decisions? What happens if the vendor disappears? These are governance questions created by infrastructure choices.

A mature social contract for digital education would make those responsibilities explicit. Convenience is valuable, but dependency should be visible. The more central a platform becomes to public education, the more its reliability and governance become public concerns even if the code is privately owned.

Artificial intelligence makes authorship and judgement public concerns

Generative artificial intelligence can lower barriers to explanation, translation and practice. It can also produce plausible errors and make it harder to infer what a student personally understands from submitted work. Assessment, authorship, privacy and academic integrity therefore become system-level problems rather than isolated classroom annoyances.

The response need not be a simple ban or embrace. Different learning jobs require different rules. Early practice may benefit from immediate AI feedback. High-stakes certification may require controlled conditions. Writing instruction may explicitly compare human drafts with machine suggestions. Professional programmes may need students to demonstrate both effective tool use and independent judgement when the tool is wrong.

Because qualifications operate as trust instruments, society has an interest in knowing what competence the credential still certifies. If tools change the production process, assessment must change too. That is the social contract adapting to a new capability environment.

Assessment is a public language of evidence

Assessment tells learners what they can do, helps teachers diagnose, supports progression decisions and provides evidence to institutions that cannot observe every classroom. Its power comes from shared interpretation. A score means little unless people understand what was measured, how reliably it was measured and what decisions the score can legitimately support.

High-stakes assessment creates predictable risks. Teaching can narrow to the test. Schools can become anxious about metrics. Students can learn strategic exam behaviour without durable understanding. But abandoning common assessment entirely can create other problems: hidden standards, inconsistent expectations and weak comparability. The design challenge is to use assessment without letting the proxy consume the purpose.

A strong system uses multiple forms of evidence where stakes are high and treats scores as indicators rather than identities. Public trust grows when the limits of measurement are visible. A transparent assessment system can say both what it knows and what it cannot know.

Accountability should make learning visible without creating theatre

Public funding creates a legitimate demand for accountability. Citizens want to know whether money reaches schools, whether children attend, whether learning occurs and whether gaps are narrowing. The danger is that accountability systems can become performative. When one indicator dominates, organisations learn to optimise the indicator rather than the underlying goal.

Good accountability combines financial transparency, learning evidence, professional judgement, inspection or peer review, and mechanisms for families and students to raise concerns. It also distinguishes support from sanction. A struggling school may need better leadership, staffing or specialist assistance, not merely another demand to improve its numbers.

The public-good principle therefore comes with reciprocal obligations. Schools and systems need resources and trust; they also need to show what they are doing with both. Accountability is not the opposite of professional autonomy. Done well, it protects autonomy by making competence visible.

Parents are participants, not customers in only one sense

Parents make choices, express preferences and sometimes purchase education directly, so the customer analogy can illuminate part of the relationship. But it is incomplete. A school serves a child, a class, future institutions and the broader community simultaneously. A parent’s preference is important, but one family’s preference can interact with the rights and learning conditions of other children.

This is why education systems need participation mechanisms more nuanced than “the customer is always right.” Parent councils, communication channels, transparent policies and appeal processes allow families to influence institutions while recognising that schools also have professional and legal obligations. The relationship is cooperative rather than purely transactional.

Schools should not use the public-good argument to silence parents. Nor should parent voice erase the expertise of teachers or the rights of other students. A workable social contract defines where voice, expertise, law and shared rules meet.

Students are not passive recipients

Education is co-produced. A teacher can design a strong lesson, but the learner still has to attend, think, practise, ask, attempt and revise. Students therefore carry responsibilities alongside rights. This does not mean blaming them for system failures. It means recognising that learning cannot be delivered like a parcel.

Student voice can improve the system because learners observe dimensions adults miss: whether instructions are comprehensible, whether rules feel consistent, whether bullying is hidden, whether digital tools create friction, whether feedback arrives in time to be useful. Voice becomes educational when students learn to turn experience into evidence and proposal rather than mere complaint.

Participation also prepares people for shared institutions. A classroom in which rules can be explained, challenged appropriately and revised when evidence warrants is a small model of legitimate governance. The lesson is not obedience. It is how authority can become accountable.

Education in emergencies exposes the public floor

War, disaster, epidemic, displacement and infrastructure failure reveal which parts of education a society treats as essential. When normal schooling is disrupted, the system must decide what to preserve first: safety, contact with learners, foundational instruction, examinations, meals, psychosocial support, documentation, digital access or teacher continuity. The priorities make the social contract visible.

Emergency education also shows why schools do more than teach subjects. They may distribute food, identify vulnerable children, provide routine, connect families to services and preserve records. When schools close, these hidden functions suddenly become visible. A narrow private-return model struggles to capture them because much of the value lies in social infrastructure.

Resilient systems therefore plan continuity before crisis. They maintain backups, alternative communication channels, flexible calendars and ways to recognise learning that occurred outside normal classrooms. Public value includes the capacity to keep education functioning when the market and the timetable are both disrupted.

The social contract extends beyond childhood

A civilisation that treats education as a shared project cannot end the contract at graduation. Adults face technological change, career transitions, migration, caregiving, health decisions and civic responsibilities that require new learning. Lifelong learning therefore extends the public-good question into adulthood: which kinds of re-entry should be publicly supported, which should employers finance, and which should individuals purchase?

There is no single answer because the balance depends on purpose. A firm-specific software course may reasonably be an employer responsibility. Foundational adult literacy has a broader public rationale. Professional licensing may combine individual fees with public regulation. Public libraries provide general access because the knowledge benefits are diffuse. The funding model should follow who benefits and who can realistically coordinate.

The key principle is re-entry. A person should not be permanently excluded from capability because an earlier educational pathway ended badly or because the economy changed after they graduated. The lifelong social contract keeps some doors open.

Why a civilisation cannot outsource all educational judgement

Societies can contract services, license providers and buy technology, but they cannot outsource the responsibility to decide what counts as adequate education. Someone still has to define minimum standards, protect children, recognise qualifications, decide what public money can fund and resolve disputes. Even a highly market-oriented system requires a public layer of judgement.

The opposite is also true: a central authority cannot possess all educational knowledge. Teachers know classrooms, communities know local conditions, employers know parts of work, families know children, universities know disciplines, and students know their own experience. Governance is therefore an information problem as much as an authority problem. Centralisation can coordinate; decentralisation can reveal local knowledge. Systems need both channels.

A mature social contract is thus neither total state control nor total market delegation. It is a negotiated architecture in which different actors hold different responsibilities and checks. The quality lies in the connections.

Failure mode: confusing public purpose with public monopoly

A public purpose can be served by different organisational forms. If the goal is universal literacy, a society might use public schools, regulated non-state schools, community programmes, libraries, digital resources and family support. Arguing that education has public value does not logically require a single provider model.

Monopoly can protect coherence and universal coverage in some circumstances, but it can also reduce responsiveness or experimentation. Provider diversity can create innovation and fit, but it can also fragment standards or stratify access. The relevant question is empirical and contextual: which architecture produces the desired public outcomes while protecting rights and quality?

This distinction keeps the debate from turning into labels. “Public” and “private” describe ownership and funding arrangements, not automatic quality. Institutions must still be judged by what learners can access, learn and carry into the world.

Failure mode: confusing choice with equal power

Choice can be valuable. Families differ, learners differ and institutions differ. But having options on paper is not the same as having equal ability to use them. Transport, fees, admissions rules, language, information quality and time can make some choices realistic for one family and impossible for another.

This does not mean choice should be removed. It means systems should analyse the conditions around choice. If a policy relies on families comparing providers, information must be intelligible. If schools are far apart, transport matters. If admission is selective, the selection mechanism matters. If leaving one school harms another through funding instability, system effects matter too.

Public policy becomes stronger when it observes how choices interact rather than assuming isolated consumers. Education decisions are networked because students share institutions.

Failure mode: measuring only private economic return

Wage and employment outcomes are important. They help learners judge pathways and help governments understand labour-market alignment. But an education system optimised only for earnings can neglect capabilities whose value appears elsewhere: health literacy, caregiving, civic understanding, cultural knowledge, scientific reasoning, social trust and the capacity to learn again later.

Even economic value itself can be wider than wages. Reliable institutions, safer workplaces, innovation, lower transaction costs and easier coordination among strangers contribute to productivity without appearing as a simple graduate salary premium. Civilisations operate through systems whose quality depends on educated people in many roles.

The solution is not to stop measuring economic outcomes. It is to place them inside a broader dashboard and remain explicit about what each metric leaves out. What gets counted should not silently become the definition of education.

Failure mode: invoking the common good to silence disagreement

“The common good” sounds noble, which makes it rhetorically powerful. That power can be misused. Different groups may disagree sincerely about curriculum, assessment, language, school choice, religious education, technology or the balance between national coherence and local autonomy. Calling one side “the common good” does not resolve the underlying conflict.

A common-good framework is strongest when it creates procedures for legitimate disagreement: transparent evidence, public reasoning, representation, appeal, periodic review and protection for rights that should not depend on majority preference. The common good is then a governance discipline rather than a slogan.

Education itself can teach this distinction. Students can learn that shared institutions do not require unanimous beliefs. A civilisation becomes more capable when people know how to remain in a common system while arguing about how it should change.

A practical test for education as a public and common good

Ask a sequence of concrete questions. Can every child reach meaningful foundational learning? Are barriers visible before they become permanent? Can families understand quality? Can qualifications be trusted by strangers? Can teachers improve over a career? Can vocational capability receive serious investment? Can adults re-enter learning? Can public institutions correct failure? Can private providers innovate without escaping standards? Can knowledge remain accessible enough for future learners to build on it?

Then ask the distribution questions. Who pays? Who benefits? Who is excluded? Who carries risk? Who has voice? Who can appeal? Who controls data? Who bears the cost when a provider fails? Which benefits appear immediately and which appear a generation later? These questions turn abstract philosophy into system diagnosis.

If an education system can answer them coherently, the public-good claim has operational content. If it cannot, the phrase may be ceremonial.

Conclusion: civilisation teaches because civilisation inherits the result

Why does a civilisation teach beyond private benefit? Because the learner carries education into systems shared with other people. The child becomes a worker, parent, neighbour, professional, voter, patient, creator, carer and eventually an elder who transmits knowledge again. Capability travels. Error travels too. The quality of one person’s education can enter another person’s life through institutions they share.

This does not eliminate private choice, private provision or private return. It explains why they exist inside a wider architecture of rights, standards, public finance, shared knowledge and institutional trust. Education is simultaneously personal and social. The mistake is forcing it into only one category.

A mature social contract for education therefore asks more than who owns the school or who pays the bill. It asks what capabilities a civilisation must keep creating, what knowledge it must preserve, what floor it owes every learner, what diversity it can sustain, and how the system can remain trustworthy enough that people continue investing effort in it. Education becomes a common concern because civilisation itself receives the consequences.

---

Continue through the eduKateSG education and civilisation library

Use What Is Education? as the canonical definition, then continue through the Education Hub, How Education Works, How Social Institutions Work, How Social Cohesion Works and How Social Capital Works to see how learning becomes institutional capacity.

Evidence anchors and further reading

UNESCO’s Reimagining our futures together: A new social contract for education frames education as a public endeavour and common good while explicitly presenting its proposals as an invitation to debate rather than a universal blueprint. UNESCO also revisits the distinction between education as a public good and the complementary common-good frame, and its 2025 review of the right to education examines new pressures from digitalisation, conflict, migration and demographic change. These sources support the definitions and contemporary context used here; the system model and distinctions developed in this article are eduKateSG’s synthesis.