How to build a state that delivers

A welfare state organises protection against risks that a person cannot always bear alone: illness, unemployment, disability, old age and childhood deprivation. It also finances services that expand opportunities. In this study, education and infrastructure are treated as foundations of prosperity and protection, although they are not included in every statistical definition of “social expenditure”. We do not add indicators with different scopes as though they measured the same thing.

OpenGov3's chosen objective is to improve individuals' lives with finite resources. This requires examining real income, health, learning, safety, time and freedom of choice. The amount received from government is only one part. One family may receive little cash and gain substantial protection through services; another may receive a transfer and still lack adequate schooling, care or transport. 1

Four mechanisms, four accounts

Universal does not mean without criteria. A service can be available to all eligible residents and still require access rules. Targeting means directing resources towards specific circumstances; contributory insurance links entitlements to contributions; funded schemes accumulate assets. These criteria can coexist in the same country. The Swedish, Danish, Australian and New Zealand cases show different combinations, not a simple division between “state” and “market”. 2 3 4 5

Autonomy does not mean the absence of support

The proposed criterion is to expand people's ability to direct their own lives. For someone with a disability or an ongoing need for care, stable assistance can be a condition of autonomy, rather than its opposite. Likewise, accessible healthcare can allow someone to start a business without concentrating every risk in the household budget. These are mechanisms to assess, not outcomes guaranteed by any programme.

Not every right has to generate a positive fiscal return. Protecting someone from hunger may be justified by dignity and protection, even without recovering the cost through taxes. What needs to be financially sustainable is the state's set of commitments. Efficiency selects better means; it does not decide on its own which objectives society should value.

A larger productive base allows more revenue to be raised without relying solely on higher tax rates. Education, health and care can, in turn, expand capabilities and economic participation. The historical study of the Nordic model highlights precisely this combination of economic openness and risk-sharing. This does not demonstrate that every expenditure increases productivity, or that any form of taxation is neutral for growth. 6

Public finances require a distinction between flows and stocks. Annual wages, profits, taxes and benefits are flows. Funds, property and accumulated debt are stocks. Selling an asset can finance a transition; it does not create revenue that recurs indefinitely. A fund can generate income, but returns fluctuate and must sustain its purchasing power and purpose.

With compatible units and periods, this identity explains why the same tax-to-GDP ratio does not buy the same services in economies with different income levels. GDP per resident is not a wage, median income or individual wealth. Nor is the calculated revenue freely available: it already finances pensions, services, administration and interest.

The bill is not paid only by the very rich

The systems studied combine broad-based taxes and contributions. The OECD identifies labour as the largest source of personal income tax revenue in the 29 countries reporting this breakdown. The conclusion concerns this tax, not all public revenue. 7

In tax comparisons, mandatory contributions to private pension accounts are not the same as taxes paid to government. Looking only at the tax-to-GDP ratio can therefore underestimate the total compulsory burden of an arrangement such as Australia's. It is also necessary to consider what families still pay directly for healthcare, housing, transport or care. 8 9

What sustains a permanent commitment

Recurring revenue, delivery capacity and adjustments over time need to work together. A new school requires construction, teachers, maintenance and operation. A pension depends on the number of participants, contributions, rules and the duration of payments. A transfer requires registration, funding, review and access for those who need it.

Borrowing can distribute costs across generations or cushion a crisis. It does not remove the real cost. Cutting waste is desirable, but unmeasured savings cannot be booked as funding for benefits already promised. The debt chapter will connect this constraint to interest rates, growth and maturities.

Swedish history shows roots predating the large welfare state, expansion during the twentieth century and reforms after crises. Denmark expanded its universalist approach particularly between 1956 and 1970. Norway was already expanding protection after the war before accumulating its oil fund. There is no universal sequence of becoming fully wealthy, paying off debt and only then providing protection. 10 11 12

The United Kingdom offers a counterpoint: the NHS began operating in 1948, after the war, and its introduction did not wait for debt to be eliminated. The OBR shows that the subsequent fall in the debt-to-GDP ratio involved nominal growth, interest-rate conditions and budget balances. This experience does not justify financing any current expansion through debt or inflation. 13 14

The useful historical conclusion is less comfortable and more practical: economic capacity, political agreement and administrative design need to mature together. An isolated reform can help, but it does not replace the entire system. Each country is examined below through the same sequence: starting point, choices, implementation, constraints and lessons.

These cases are useful because they combine protection, economic capacity and institutions that endure across governments. ETLA's comparative book describes economies integrated into international markets, with risk sharing that helps sustain acceptance of change. This is an evidence-based economic interpretation, not proof that a single formula exists. 6

What the idealised image tends to leave out

The labour market is part of the welfare state. Wage bargaining, training, mobility and predictable rules connect protection to production. Sweden's 1938 agreement and Danish flexicurity illustrate institutions built to manage conflict, not societies in which different interests have disappeared. 15 3

Care and taxation affect who can work. The expansion of childcare and changes to the taxation of couples reduced barriers to economic participation. They are elements of social infrastructure, not mere accessories to a cash benefit. The combination and timing varied between countries. 16

The rules were revised. Sweden redesigned its fiscal framework after the crisis. Finland reorganised health and social services in 2023. Iceland had to confront a banking collapse. Being an international benchmark does not eliminate errors, political disagreements or the need for reform. 17 18 19

Scale matters, but it does not explain everything

At the beginning of 2025, the Nordic region had 28.3 million inhabitants, including autonomous territories. Around 21% were aged 65 or over. A small population does not mean an absence of ageing pressures. 20

Comparing size also requires looking at density, distance, age, employment, productivity and administrative organisation. A larger population increases both demand and the potential tax base. In an exercise where all conditions per person are identical, the per capita calculation does not change simply because the country becomes larger. In the real world, those conditions rarely remain the same.

What this selection allows us to conclude

The cases help identify plausible mechanisms and differences in design. On their own, they do not isolate the causal effect of being Nordic, having a small population or raising the tax-to-GDP ratio. A reputation for prosperity must be checked against actual access, learning, health, disposable income and sustainability. A country can be a benchmark and still fall short in important areas.

What it means to call something a determining factor

A reform may have been historically important without our being able to calculate how much of today's outcome it caused. Here we distinguish between an institutional fact (a rule was created), evidence of an effect (a study estimates a change) and an interpretation of the mechanism (how the pieces may reinforce one another). The strength of the conclusion must match the strength of the evidence.

The cases reveal decisions, dependencies and costs. They do not constitute an experiment proving a single model of prosperity. Comparing countries only in the present also mixes prior conditions, simultaneous policies and external shocks.

Trust is not a sufficient explanation

Saying that it works because people trust one another leaves the main question unanswered: why do they continue to trust? Enforceable contracts, understandable rights, predictable services and correction of errors can reinforce that trust. The Danish contract of 1882 and public employment records illustrate concrete practices; they do not measure an isolated cultural effect. 7880

Nor is there consensus without conflict. Labour agreements and tax decisions are ways of organising disagreements. What can be transferred is not a population without differing interests, but institutions capable of negotiating, implementing and revising commitments.

1. Before broad benefits, a productive base

Sweden did not begin its development with the fully formed welfare state often imagined today. Ari Kokko's historical study describes rapid industrialisation in the late nineteenth century and the gradual shift from an advantage based on raw materials to one supported by human capital and knowledge. Social coverage and benefits expanded progressively. This challenges both the idea that prosperity emerged from a cheque and the claim that public services were irrelevant to development. 76

The economic relationship: a raw material creates more possibilities when a country develops the techniques, firms, workers and markets capable of transforming it. Tax revenue captures part of that value. But the transition from natural resources to productivity is not automatic; it depends on investment and learning. This is an interpretation of the mechanism, not a statistical decomposition of the causes of Swedish growth.

2. Build bargaining institutions before expecting consensus

The Saltsjöbaden Agreement of 1938, between workers' and employers' organisations, established rules for managing labour relations and disputes. The innovation was not to abolish disagreements: it was to create a process that made negotiating them more predictable. The agreement's history is described by PTK, an organisation within the collective bargaining tradition. 15

The Rehn–Meidner model combined wage coordination, the containment of macroeconomic imbalances and employment policies. The idea of bringing wages closer together for comparable work put pressure on low-productivity firms: they should not survive indefinitely simply by paying far less. In return, training and mobility needed to enable workers to find better activities. Erixon describes a partial historical application that changed over time. 21

3. Remove practical barriers so that more people can work

Individual taxation of spouses, introduced in 1971, stopped treating a wife's income as an extension of her husband's. The expansion of childcare and parental insurance open to both parents in 1974 completed another part of the design. It was not enough to permit people to work: combining employment and care had to become possible. The Swedish government identifies these reforms as part of the development of the dual-earner family model. 22

The institutional insight: taxation, early childhood education and parental leave are not three independent agendas. A financial incentive can fail when nobody is available to look after the child. A childcare place may have little effect if taking a job barely increases net income. The potential gain depends on the combination and must be measured through participation, income and quality of care, not just places announced.

4. Decentralise delivery with responsibilities and revenue

Municipalities and regions have constitutionally protected self-government and taxation powers. Central government regulates their activities and is responsible for equitable conditions in the provision of welfare. Decentralisation therefore means neither leaving each territory to fend for itself nor taking every decision in a single national ministry. 77

To understand this arrangement, separate three tasks: establishing rights and standards; providing funding consistent with needs; and organising local delivery. Transferring only the third task, without revenue or staff, does not reproduce the mechanism. The question is not how many agencies exist, but who can solve the problem and whom citizens can hold accountable.

5. Repair the finances when the design becomes unaffordable

The crisis of the early 1990s led to a stronger fiscal framework. It combines an expenditure ceiling, medium-term planning, local rules and external monitoring. The anchor of 35% of GDP is a reference for debt, not a universal safety threshold. The budget balance target is assessed over the cycle; from 2027, it becomes a balanced-budget target, replacing a surplus of one third of a percentage point of GDP. 17

The pension system has different components. The income-related component is not an individual account with all the money invested in financial markets. The funded component is a separate layer. The pensions agency describes allocations of 16% and 2.5% of pensionable income, respectively, alongside protections established by law. Confusing a contribution-based promise with assets already accumulated distorts the interpretation of the system. 2

6. What was decisive in the design and what remains unresolved

The central interpretation is complementarity: firms capable of producing, bargaining institutions, services that make work possible and revisable financial commitments. None of these elements guarantees the results of the others on its own. The fiscal crisis also shows that the system was not always balanced.

The IMF's macroeconomic assessment cited in the study identifies difficulties involving labour market integration, productivity, housing and private indebtedness. Their existence prevents us from treating the country as a completed model. This chapter does not estimate each reform's causal contribution to today's standard of living. 23

For adaptation to Brazil: before copying a benefit or making a contract more flexible, identify who could be left without transition support, what training matches real vacancies and how the support will be funded. The measure of success is the person's trajectory after the change, not a reduction in expenditure taken in isolation or the announcement of a right.

1. A history of production before a history of benefits

One document helps move beyond abstract debate: the contract of the Hjedding dairy cooperative, dated 1882. It specified milk supply, sanitary conditions, accounting records, payments and oversight. It set out consequences for adulteration, annual reporting and the appointment of auditors. Cooperation did not rely on personal trust alone; it involved verifiable obligations. 78

The mechanism illustrated: small producers can build capacity by sharing an operation, but the quality of the whole becomes vulnerable to each participant's behaviour. Contracts, records and controls make collaboration more sustainable. This differs from simply sharing revenue. The case does not establish the origins of the entire Danish welfare state; it shows a concrete way to build economic and institutional capacity.

2. Turn conflicts into affordable agreements

The Kanslergade Agreement of 1933 brought economic and social responses together in a political negotiation. The expansion of universal provision gained momentum particularly between 1956 and 1970. Building rights depended on agreements and gradual change, not on a single decision that solved everything. 2411

The relevant lesson for Brazil concerns the process: the demands of workers, producers, taxpayers and government had to enter the same negotiation. An agreement does not eliminate costs or make everyone a winner at every moment. It can, however, establish more predictable commitments than a succession of incompatible promises. Assessing that result also requires looking at who was initially excluded and how coverage changed.

3. How the flexicurity triangle works

The public agency STAR describes three components: mobility between jobs, income protection and active labour market policies. Legislation and collective agreements allow firms to adapt, while still imposing rules. Unemployment insurance depends on participation and eligibility. People outside that insurance scheme may have access to other support, subject to financial criteria. It is not the same unconditional benefit for everyone. 79

Training and job-matching services matter because insurance alone does not create a vacancy. Nor is it enough to make dismissals easier and expect reallocation to happen without friction. Policy must address different barriers: skills, health, transport, family care and local demand for labour.

4. The unseen work of reorganising municipalities

The 2007 reform reduced the number of municipalities from 271 to 98 and reorganised responsibilities. The process began with a commission in 2002, recommendations in 2004 and negotiation of the new boundaries. Small municipalities could be required to enter binding cooperation arrangements with larger ones. It was not simply an order to cut government offices the next day. 25

The problem was one of capacity: certain responsibilities require specialist teams, procurement, systems and sufficient scale of provision. Increasing administrative size can help, but also creates distance and integration costs. The relevant unit is the area capable of delivering the service well, not necessarily the smallest or the largest.

For a Brazilian municipality, the equivalent question would be: does it make sense to maintain every laboratory, specialist or system on its own? Or would a regional arrangement, with defined funding and responsibilities, provide a better service? This is a hypothesis for adaptation, not a general recommendation to merge municipalities.

5. Give frontline services the tools to manage

Jobindsats.dk, launched in 2007, brings together data on benefits, population registers, businesses, taxes and income. It enables monitoring of employment policies and comparison of territories. STAR reports that updates are typically monthly, with a lag and retrospective revisions. The service supports management and external scrutiny; it is not an infallible real-time snapshot. 80

What distinguishes it from a decorative dashboard is the link between records, action and follow-up. Even so, comparing two municipalities requires considering whom they serve. Rewarding only shorter benefit duration can encourage exclusion; rewarding only course attendance can produce training without employment. Evaluation needs to consider income, continued employment and groups facing greater barriers.

6. The model is also reformed

STAR published guidance on an employment reform approved in December 2025, with stages in February and July 2026. The record shows changes to organisation and contact requirements. It does not yet demonstrate that the new arrangements have delivered better long-term outcomes. 89

The fiscal framework and the link between retirement and longevity also contribute to sustainability. The study brings together IMF and OECD references on these rules and on the productivity gap between large exporters and the domestic economy. The success of one sector must not be confused with uniform performance across all firms. 2627

7. The transferable lesson

Denmark's distinctive feature is the connection between economic flexibility and protection that works in practice, rather than only on paper. This requires recurring revenue, negotiation, local capacity and information. The existence of eligibility criteria does not justify treating those who need assistance as suspects; it calls for an administration capable of distinguishing circumstances and offering appropriate paths forward.

For adaptation to Brazil: assess the whole package. Can the worker take the job? How much will their net income change? Does the course meet a demand? Is childcare available? Can the municipality provide follow-up? Without these answers, copying a benefit amount or a labour rule does not amount to copying the model.

1. The country's history did not begin with oil

The historical study of Norway's expansion between 1945 and 1970 describes the growth of health, education and protection after the war. The large financial reserve came later. This matters because a discovery does not automatically create an administration capable of taxing, contracting, supervising and delivering services. It would also be wrong to go to the opposite extreme: oil wealth materially expanded the country's possibilities. 12

The discovery of Ekofisk in 1969 was followed by production in 1971 and the creation of Statoil in 1972. The official history shows participation by foreign companies and the progressive development of a state presence. It was not a binary choice between handing over the entire sector and trying to develop all the necessary technology alone from the first day. 28

2. Capturing resource income takes more than ownership

The revenue system combines taxation, direct ownership interests and dividends. The official reference consulted reports a combined rate of 78% on petroleum income and a state holding of 67% in Equinor. These are sector-specific components, not the tax-to-GDP ratio of the Norwegian economy as a whole. 29

The economic insight: part of the gain comes from a limited resource, not just a firm's current effort. Policy seeks to capture this income for society. But effectiveness depends on the tax base, the allocation of costs and risks, oversight and investment terms. A high rate written into law does not, by itself, reveal how much revenue will be collected.

Separating the functions helps frame better questions: who authorises extraction? Who supervises it? Who represents the public owner? Who operates? Who decides how the revenue is used? Analysing these responsibilities is more useful than attributing the entire outcome to the existence of a state-owned company.

3. The gap between creating a fund and putting money in it

The law establishing the fund dates from 1990. The first net transfer took place in 1996. The fund manager's own timeline shows that announcing a fund and being able to contribute to it are separate moments. The institution's name does not generate savings; resources have to be withheld from consumption in the present. 81

That interval offers a lesson for Brazilian proposals. A fund financed by revenue already used in the budget may require cutting other spending, raising other revenue or accepting more debt. If the state borrows to invest, it has not automatically created net wealth: it now holds assets and liabilities, with their respective risks. This is a balance-sheet relationship, not a criticism of every fund.

4. Why invest abroad

The manager, NBIM, operates under a Ministry of Finance mandate and an international investment strategy. Policy decisions and portfolio management are not treated as the same task. The fund holds a diversified portfolio of assets, rather than operating solely as a cash reserve to finance domestic projects selected during each political cycle. 82

Interpreting the mechanism: investing abroad helps separate two concentrations of risk. Revenue originates in Norwegian petroleum activity; the assets need not depend exclusively on that same activity and territory. It also limits the temptation to turn every windfall into an immediate expansion of domestic demand. This does not eliminate market losses or currency exposure.

5. The 3% rule is not a guaranteed return

Since 2001, the fiscal rule has linked the use of resources to the expected long-term real return. The reference was reduced from 4% to 3% in 2017. It guides the structural non-oil deficit and allows cyclical flexibility. It does not mean that the portfolio earns exactly that amount every year or that the budget can spend any apparent increase in value. 31

6. The fund does not replace teachers, doctors or productivity

Financial assets provide purchasing power, but do not eliminate shortages of professionals and services. If spending grows faster than supply can respond, some of the pressure will appear in costs. Nor can an economy plan indefinitely as though income from an exhaustible resource were permanent.

This study therefore separates petroleum activity, the mainland economy, the fund and the budget. An average indicator that includes oil does not directly measure productivity in the rest of the economy. Similarly, gross debt does not show the assets on the other side of the balance sheet, and the fund's wealth is not freely available money in each resident's account.

7. What is exceptional and what can be learned

Oil availability per inhabitant is a distinctive condition, not the result of a policy that any country can replicate. The institutional design offers more transferable lessons: set rules for windfall revenue, make risks explicit, professionalise asset management and make withdrawals from the fund part of the budget process.

For adaptation to Brazil: start with the source of revenue and the opportunity cost. Then define the purpose, contribution policy, governance, risk and withdrawals. Promising a benefit first and only then looking for assets to sustain it reverses the construction process. The desired result is continuity of services and opportunities, not simply a large figure announced for a fund.

1. It did not begin with a ready-made system

Kela began in 1937 as a pensions institution. The original design included savings accounts; in 1956, those accounts were abolished and the system moved to another arrangement. In 1963, personal identification and an insurance card simplified administrative access. In 1964, health coverage expanded the population served. The institution's own history shows successive revisions and an expansion of capabilities. 32

The distinctive feature is not just the number of benefits. It is the creation of a structure that can identify citizens and process their entitlements. A policy can be sound on paper and fail because people do not know where to apply, must repeatedly prove their circumstances or wait too long for a decision. Administrative costs and users' time are part of delivery.

2. Comprehensive schooling changed educational pathways

Between 1972 and 1977, Finland gradually introduced a nine-year comprehensive school system, replacing the previous separation into tracks. A study used that phased geographical introduction to compare groups exposed to the change at different times. In the sample of men born between 1960 and 1966, it found a reduction in the link between parents' and children's incomes. This is not simply a comparison between a rich country and a poor one. 33

How to interpret the evidence: the study provides stronger identification of a specific effect of the reform. It does not demonstrate that all of Finland's prosperity came from it, or that an effect of the same size would occur for another generation or in Brazil. The contribution investigated concerns intergenerational mobility, not an endorsement of every outcome of the school system.

3. Changing schools also required changes in teacher education

Historical research on Finland and Sweden describes the incorporation of teacher education into higher education in the 1970s. In Finland, university-based academic training developed more fully. Institutional conflicts and implementation differed between the countries, despite initially similar proposals. Announcing the same objective, in other words, does not mean being able to implement the same reform. 91

The management interpretation is that teacher autonomy requires professional capacity. Copying a school with fewer visible controls while leaving everything else unchanged is not enough. Teachers need preparation and support, the curriculum must be clear, and learning must be observed. Trust may be a consequence of trustworthy institutions, rather than a cultural resource that simply exists or is absent.

4. The Nokia test: knowledge needs a new use

The decline in Nokia's activity tested that connection. Research using records of more than 20 thousand former employees, who left between 2009 and 2014, found that more than three quarters had obtained new employment by 2017. The remainder also included retirees and people who had returned to education. Education, previous position and region were associated with different trajectories. 83

The result shows a capacity to redeploy skills, but not uniform success. It does not prove that a specific benefit caused re-employment. Nor does it represent all unemployed people in the country: these were people who had worked at a large technology company.

The useful question for Brazil: when an activity declines, which of its skills can move elsewhere? Training has more value when other firms recognise it and when demand can absorb it. Preserving a company's structure for ever and preserving its workers' human capital are different objectives.

5. The administration also had to learn to share capacity

Kela records that the digitisation of documents and decisions between 2000 and 2004 allowed workloads to be transferred between regions facing greater and lesser pressure. This is a specific operational change: the process became less dependent on the physical location where demand arose. 32

6. In health, the administrative scale was redesigned

Since 2023, responsibility for organisation has moved to 21 wellbeing services counties, with specific arrangements for Helsinki and the HUS group. Funding comes predominantly from the central budget through a formula; municipalities retain public health functions. The European Observatory also describes additional channels, such as national insurance, occupational care and household payments. A public system therefore does not mean a single funding pool or an absence of differences in access. 18

Consolidation can facilitate recruitment, technology and regional planning, but the result must show up in access and quality. The OECD's 2025 reference, retained from the previous study, records financial and staffing pressures. Reorganising responsibilities without resolving the supply of professionals does not eliminate waiting lists. 35

7. The lesson is not that education solves everything

The combination of education, service institutions and a productive base is more convincing than a single-factor explanation. Education provides capabilities, but opportunities also depend on investment, firms and demand. Pensions and care remain exposed to ageing and the cost of labour-intensive services.

For adaptation to Brazil: combine a precise educational goal with teacher training and support; measure trajectories, not just enrolments; connect skills to opportunities; and organise services at a scale that allows resources to be shared. Finland's experience shows both the benefits of patient institution-building and the limits of relying on an established reputation for excellence.

1. Protection was built through negotiation, not received ready-made

In the historical interview published by the Icelandic pension funds association, a signatory to the 1969 agreements describes the expansion of pension participation among unionised workers. Compulsory participation was extended by law in 1974, also reaching employees outside trade unions. Funds for some occupational groups already existed. The agreement expanded an earlier structure; it did not invent all protection at once. 84

The source is a participant's recollection and should be read as such. It is useful for reconstructing the mechanism and its conflicts, not for accepting the interviewee's praise of the system without assessment. The history matters because rights involved negotiation, contributions and financial design. They did not emerge simply because the population was small.

2. The problem of the first generation of retirees

The same interview describes a transitional arrangement for older workers who had not had time to accumulate entitlements in the new funds. Other funds and public resources contributed to that effort. 84

This is one of the least visible issues in any pension reform. If workers' contributions begin to build reserves for the future, someone still has to fund benefits already promised. The transition can be financed through taxes, debt, other contributions or rule changes. None of these options removes the cost; each distributes it differently among people and generations.

3. Funded pension schemes are not risk-free money

Pension savings can distribute resources over time and reduce exclusive reliance on future contributions. However, participants' entitlements still depend on the scheme's rules and assets. Markets, inflation, longevity and concentration can change its ability to pay.

This reasoning prevents two misunderstandings: pension assets are not freely available Treasury cash; and creating financial assets does not remove the need for an economy capable of producing the services retirees will consume. A portfolio can buy goods abroad, but it does not make the supply of local care unlimited.

4. In 2008, the problem did not start with social benefit payments

The cited IMF study describes a banking crisis that required financial reconstruction, stabilisation and controls on capital flows. Risk reached the state through the financial system and economic activity, not only through an expansion of benefits. This changes what putting one's house in order means: supervision, contingent liabilities and the currency of financing matter too. 19

The transmission chain: a credit crisis can reduce production and employment; tax revenue falls just as the need for protection grows. At the same time, guarantees and interventions may require public resources. A budget that looked comfortable before the shock can lose room for manoeuvre through several channels at once. This is an economic explanation of the risk, not a new estimate of Iceland's costs.

5. The sequence of the response matters

The IMF's 2011 account emphasises that automatic stabilisers were allowed to operate in 2009. When activity falls, some taxes decline and certain benefits increase without requiring a new programme each time. Fiscal adjustment was accompanied by financial reconstruction and attention to social protection. 36

This does not mean that all benefits were preserved unchanged. The 2010 report records both revenue and expenditure measures. It also addresses debt and liquidity. The experience cannot be reduced to claims that letting banks fail solved everything or that spending cuts produced the recovery. Several policies and external conditions interacted. 19

The practical question is the order of action: what must work immediately to prevent greater harm? What can be adjusted gradually? Which losses will be recognised? Who pays for them? An emergency programme without an explained exit carries risks; so does consolidation that destroys capacity before recovery.

6. A small population does not mean small risks

In a smaller economy, technical resources and revenue sources may be concentrated. The benefit of closer coordination coexists with less diversification. Population size alone does not reveal the relationship between financial liabilities and the economy's ability to absorb them.

The OECD's 2025 references retained from the previous study discuss problems in housing, services and basic skills. Recovery from a crisis does not automatically eliminate every subsequent bottleneck. Nor does an image of prosperity guarantee that young people can find affordable housing or that every group has the same opportunities. 3738

7. The lesson for Brazil

The central lesson is to plan the social system together with its risks. In pensions, show the costs of the first generation and of changing the scheme. In public finances, identify guarantees, companies and financial exposures that could reach the budget. In services, preserve delivery mechanisms during the shock.

The most reliable protection does not depend on pretending that crises will never happen. It depends on building tools to get through them: financing with appropriate maturities and currency, adequate reserves, supervision, emergency rules and explicit priorities. These are conditions for resilience, not a promise of immunity.

1. The social safety net predates market and management reforms

The Social Security Act of 1938 expanded protection and brought together benefits for circumstances such as old age, disability, unemployment and families. The official history records dedicated tax funding. New Zealand therefore did not establish its welfare state after the economic reforms of the 1980s and 1990s. It reformed institutions that already had commitments to the population. 39

This sequence allows the right question to be asked: how can a protection system adapt to a changing economy and a budget under pressure? Asking only whether the government spent more or less is insufficient. We need to distinguish the entitlement offered, the funding base, the management tools and the outcome actually achieved.

2. Make visible the costs absent from cash accounts

The Public Finance Act of 1989 established a new basis for financial information and agency responsibilities. The Treasury's guide describes accrual accounting and statements that go beyond cash inflows and outflows. The timing of payment is not the only relevant information for understanding a policy's cost. 40

3. Management autonomy in exchange for explicit accountability

John Janssen's retrospective describes how the State Sector Act of 1988 defined relations between chief executives and ministers, including renewable contracts and responsibility for staff. The new design sought to give managers room to organise resources within the budget while holding them accountable for specified outputs. Financial reform began to emphasise services and outputs rather than only authorised inputs. 85

The logic: demanding results from someone who cannot organise their operation creates fictitious accountability. Yet autonomy without targets, information and oversight can increase arbitrary decision-making. The challenge is to combine decision-making capacity, clear limits, evaluation and the possibility of correction. The experience does not justify replacing all public controls with a business contract; the state's rights and responsibilities are different.

4. The fiscal rule structures the future, not just the next financial year

The Fiscal Responsibility Act of 1994, later incorporated into the Public Finance Act, linked responsibility and transparency to a longer perspective. The government must make its strategy and objectives explicit, rather than presenting one year's budget as though subsequent years were independent. The framework rests on principles and disclosure obligations, not on a single magic percentage valid in every circumstance. 41

For Brazilian readers, this means asking: will commitments remain affordable beyond the government's term? Does a cash saving today create a larger obligation tomorrow? Which scenario is being used for revenue growth? What changes if interest rates or demographics differ from expectations? These are planning questions, not grounds for promising that a fiscal law will eliminate political conflicts.

5. Correcting course: efficient agencies can form a poor system

In an official response published in 2019, the government acknowledged that management tools had helped fiscal responsibility and accountability, but had also encouraged silos, short-term focus and risk aversion. The document proposed greater collaboration and attention to long-term public value. This is important self-criticism from a country often presented as an example of exemplary management. 86

Imagine a person who needs housing, healthcare and training. Each agency may meet its own service target while nobody resolves the problem as a whole. The answer is not to abandon targets. It is also to define the shared outcome, who coordinates it and what happens when a dependency prevents delivery.

The 2019 Wellbeing Budget and the Living Standards Framework broadened the structure of debate about living conditions, distribution and resilience. They are tools to inform choices. The 2019 budget is a historical document, not the budget in force in 2026. A new name or dashboard does not, by itself, demonstrate a change in outcomes. 4243

6. Three retirement instruments that are not the same thing

NZ Super is a public benefit subject to age and residency criteria, without the same income and asset tests as Australia's Age Pension. KiwiSaver is individual saving with enrolment mechanisms and an option to opt out under the prescribed conditions. The NZ Super Fund consists of public assets intended to partially pre-fund future expenditure. They are not three names for a single account. 44445

That distinction changes who receives, who contributes, who bears risk and who has access to the assets. The public fund is not an individually redeemable balance. Individual savings do not automatically replace the public benefit. And a future public promise does not mean the country has accumulated the full amount needed in advance.

7. Yesterday's success does not settle tomorrow's bill

He Tirohanga Mokopuna 2025 examines long-term pressures, particularly ageing, health and pensions. Its scenarios maintain explicit assumptions to show how the trajectory could evolve. They are not a prediction that governments will remain inactive for decades. 87

The most useful reading is not the extreme number at the end of a projection. It is to identify which obligation grows, why it grows and what combination of productivity, revenue, benefit design and delivery cost could change the trajectory. Reform must consider who bears the adjustment, not just the aggregate balance.

8. What Brazil can learn

The most transferable mechanism is the connection between complete financial information, operational accountability and assessment of outcomes for the person. The first generation of reforms improved tools. The subsequent review showed that those tools could also fragment services.

For adaptation: clarify the manager's mandate, provide the capacity to fulfil it, record the full cost, require accountability and check dependencies between agencies. Then observe whether the citizen's experience improved. The state does not become efficient simply by adopting business language or putting wellbeing in the budget's title.

1. Early protection was targeted and included exclusions

Australia's federal law of 1908 established old-age and disability pensions subject to criteria. The National Museum of Australia's history also shows exclusions in the original design. The model admired today did not begin as universal or immediately include every group. Building protection involved expanding rights and correcting restrictions, not just organising the finances. 49

This point prevents an idealised history. When presenting a success story, we must ask whose success it was at each stage: formal workers, men, women, immigrants, Indigenous peoples or everyone? A contemporary national average neither erases historical exclusions nor demonstrates equal access today.

2. Public healthcare and old-age income follow different principles

Medicare, introduced in 1984, forms part of health protection. For retirement income, the Age Pension provides public support subject to income and asset criteria. Broad services do not require a country to use exactly the same eligibility criteria for every cash transfer. Nor does Medicare mean unlimited coverage of every procedure at no cost. 505

The design logic: preserve collective access to services while concentrating some cash support on those with fewer resources. This can reduce the benefit's cost compared with an equal payment to everyone, but requires reliable information, review and understandable benefit-withdrawal rules. Targeting trades some simplicity for more demanding administration.

3. How retirement provision entered the wage agreement

The Treasury records that the expansion of occupational pension contributions began in the prices and incomes accord of 1985. Part of remuneration was directed towards retirement instead of an immediate wage increase. The 1992 legislation known as the Superannuation Guarantee expanded and consolidated employers' obligations. The assets were therefore built through contributions over time. 51

The important point is not to memorise a rate. It is to recognise the distributional decision: forgo some possible consumption today to build resources for old age. Employers' legal obligation to make contributions does not demonstrate that profits alone bear the economic cost; wages, prices, hiring and margins may respond depending on the context.

The institutional reference cited records that the compulsory contribution reached 12% in July 2025. This chapter is not an eligibility or payroll guide: dates, the contribution base and operational exceptions must be checked against the tax authority's current guidance. 9

4. What funded pensions do and do not do

A pension account with assets differs from a promise funded solely by future contributors. However, creating compulsory savings does not mean the entire amount is additional saving in the economy. Households may reduce other reserves, and tax incentives have a fiscal cost. The Treasury explicitly discusses these effects without treating fund balances as free money. 51

Funded pensions also raise distributional questions: interrupted careers and lower pay tend to produce fewer contributions; fees reduce the balance; returns are uncertain; and longer retirements require resources for longer. The public income floor and eligibility rules must be analysed together with savings, not afterwards.

5. Compare services using a common language

The Productivity Commission's Report on Government Services organises assessment around efficiency, effectiveness and equity. It separates resources, outputs and outcomes and acknowledges limits to comparability between territories. This enables a better discussion than comparing total spending and declaring whoever spent least the winner. 52

A service may have a low cost per case because it excludes complex cases. Another may spend more because it serves remote regions. Evaluation must examine the population served, quality and access. In a federation, common information standards help distinguish legitimate differences in need from implementation failures.

The Closing the Gap agenda, with monitoring in 2026, keeps disparities affecting Aboriginal and Torres Strait Islander peoples visible. This source is used here for a methodological point: national averages are not enough to judge whether services reach everyone. No single score is assigned to the country. 90

6. Robodebt: automating a bad rule magnifies harm

The 2023 commission of inquiry into Robodebt issued recommendations on policy design, data, communication, vulnerability and oversight. Among them, it emphasises service to people, clear language and channels suited to different capacities for access. The existence of a digital system is not proof of legality, correctness or efficiency. 53

The management principle is simple: an apparent saving obtained by wrongly denying a right is not public-sector productivity. Before scaling automated recovery or benefit awards, the rule, data quality, exceptions and challenge process must be tested. The cost of errors to citizens must enter the assessment rather than disappear because it is absent from the agency's spreadsheet.

7. What was decisive in the design

This study's interpretation centres on the explicit allocation of functions: health services, an income floor, reserve accumulation and evaluation mechanisms do not try to solve exactly the same problem. Their results depend on coverage, oversight and the quality of their interactions.

For adaptation to Brazil: compare the total effort, including contributions and household spending; make the transition between schemes explicit; observe who accumulates few assets; and measure services through access and quality. The Australian lesson is not simply a smaller state, but the need to know where each obligation sits, who finances it and how citizens remain protected when one element fails.

Analytical synthesis of the preceding chapters. The selected features do not exhaust each country's experience. A funded pension does not replace primary care; a fiscal rule does not replace teachers; a universal service does not eliminate the need for productivity. The comparison is between complementary functions, not competing slogans.

United Kingdom: creating services without waiting for zero debt

The NHS began in 1948, bringing services together and redefining responsibilities. Britain's subsequent debt reduction involved nominal growth and financing conditions as well as budget balances. The case challenges a rigid historical sequence, but not a resource constraint. A useful question: what material capacity allows care to expand while debt is still high? 13 14

Netherlands: separating the floor, contributions and reserves

The architecture combines a basic public pension, occupational schemes and individual savings. Participation rules vary by sector and arrangement. The lesson is to distinguish benefit entitlements from ownership of assets, while making explicit who bears longevity and market risks. A useful question: is the pension promise financed by future contributors, accumulated assets or both? 56

Switzerland: discipline with room for the cycle

The federal debt brake, applied since 2003, adjusts the expenditure ceiling to the cyclical position and provides for extraordinary circumstances and offsets. It should not be confused with an absolute ban on deficits. Nor does the federal government represent the entire federation. A useful question: how can permanent expansion be contained without requiring mechanical cuts in every recession? 57

Canada: programme review before across-the-board cuts

In the parliamentary debate of March 1995, programme review was presented through six questions: public interest, the role of government, the responsible level of government, opportunities for partnership, efficiency and affordability. The source records the stated logic; it does not demonstrate all its subsequent social effects. A useful question: which programme should continue, change its design or change its provider, rather than cutting the same percentage from every programme? 58

Singapore: debt needs a complete balance sheet

The Ministry of Finance explains that assets exceed liabilities and that most issuance serves savings, investment or financial market purposes, rather than unrestricted financing of current spending. There are specific instruments for infrastructure. The budget also receives a contribution from returns under its own rule, which is not Norway's rule. A useful question: what assets, restrictions and flows lie behind gross debt? The city-state's scale limits the transplantation of its administrative arrangements to Brazil. 59 60

Greece: real constraints and inadequate design can coexist

The IMF's evaluation of the 2010 programme acknowledged problems with assumptions and sequencing, including a more severe recession and difficulties related to the treatment of debt. The case proves neither that deficits are irrelevant nor that any austerity programme works. A useful question: how can adjustment be distributed and liabilities addressed without destroying the economic capacity that should sustain recovery? 61

Estonia: reducing friction without pretending that technology eliminates oversight

The X tee infrastructure enables data exchange supported by agreements, authentication, authorisation, encryption and logs. Its value lies in avoiding unnecessary requirements and transfers while retaining accountability. It is not a single database that makes all information true. A useful question: what data does the state already hold and may legally reuse, with traceability, instead of asking citizens to provide it again? 62

These cases were added by function: services, pensions, fiscal rules, spending review, assets, crises and information infrastructure. To maintain focus, the extensive geopolitical discussion of France, China, the US and Russia in the study received was not reproduced in full. Its general risks reappear in the fiscal analysis; the original numerical table remains in the working material and is not reproduced on this page. 63

The government does not have to repay all its debt at once. Refinancing maturing bonds is a normal part of financial management. Difficulties arise when costs increase, maturities shorten, repayments become concentrated or confidence in future financing declines. IMF and World Bank guidelines treat debt as a portfolio with costs and risks, not just a number. 64

Five concepts that must remain separate

In the study received, the IMF primary balance excludes net interest. The difference between the primary and overall balances should therefore not automatically be treated as the gross cost of bonds. Interest revenue matters when assets are held. This convention belongs to the working material. Its table is not reproduced here because a complete re-extraction has not been completed. 63

The logic of the trajectory

In the illustrative model without financial assets, exchange-rate changes or other balance-sheet adjustments:

b is debt as a percentage of GDP. i is the portfolio's average effective nominal interest rate. g is nominal GDP growth. p is the primary surplus as a percentage of the following period's GDP; a deficit is negative. Rates i and g enter the calculation as fractions.

If nominal growth matches the effective interest rate and the primary balance is zero, the debt-to-GDP ratio remains stable in this exercise, even with interest to pay. If interest exceeds growth, a larger primary balance is needed to stabilise the same debt. Under these assumptions, the approximate stabilisation calculation is p = b × (i − g) ÷ (1 + g).

The comparison must use nominal growth, which combines real activity and GDP prices. It is incorrect to compare nominal interest with real growth and attribute the entire difference to imbalance. In reality, exchange rates, asset purchases and sales, recognition of liabilities and other changes also affect the stock.

Why the connection with the welfare state is direct

Interest and refinancing compete for resources and attention with other commitments. A predictable portfolio helps the government maintain services when revenue falls. But low debt does not solve shortages of staff, medicines or learning. Fiscal sustainability and delivery capacity are complementary conditions.

This study does not identify a universal debt percentage that defines a safe country or a default. Assessment must consider currency, revenue, assets, institutions, the maturity profile and future risks. The working material includes an annex covering 21 countries, not published on this page and containing no probabilities of default.

A good debt team can reduce vulnerabilities involving maturities, currency, concentration and operations. It cannot, on its own, make a policy of incompatible spending and revenue affordable. International guidelines aim to meet financing needs at an appropriate cost over the medium and long term, subject to prudent risk. Seeking only today's lowest rate can increase tomorrow's risk. 64

The proposed operational dashboard

This framework is a proposed operational synthesis, not a claim that every country uses the same dashboard. Iceland's experience illustrates the importance of maturities and liquidity during recovery. New Zealand's financial transparency shows why obligations and assets must be visible beyond cash accounts. 19 40

Having one's own currency helps, but does not erase the real cost

Foreign-currency debt can increase in domestic-currency terms when the currency depreciates. Debt in the country's own currency avoids this particular mismatch, but remains exposed to inflation, interest rates and confidence. The investor's nationality and the bond's currency are different things: a foreign investor can finance a domestic-currency bond.

In the euro area, a national government does not unilaterally control issuance of the common currency. In a country with its own currency, nominal payment capacity is different, but does not guarantee the payment's purchasing power. Nor does the central bank produce the professionals and inputs required by services. The next section separates monetary policy from fiscal transfers.

A shock does not reach the entire budget on the same day

An increase in rates on new issuance affects floating-rate debt or near-term maturities more quickly. A long-term, fixed-rate portfolio transmits the effect gradually. Extending maturities may cost more initially; accepting a lower short-term cost may increase risk. The choice must be made explicit, not hidden in a promise of low interest rates.

The relevant test is to simulate combined scenarios: lower growth, revenue shortfalls, higher interest rates, adverse exchange-rate movements and the need for financial support. Simply adding the projected deficit to current debt does not capture all this exposure.

The boundary between three different decisions

Social policy defines rights, eligibility and services. Fiscal policy makes the overall spending and revenue commitments consistent. Debt management chooses how to finance them and manage portfolio risks. Mixing the three enables misleading promises: changing a bond's maturity does not solve an unfunded benefit; cutting a service does not guarantee a safer portfolio.

The decision must withstand a less favourable scenario

Extending maturities may reduce refinancing pressure and increase the initial cost. Using short-term debt may look cheap while accelerating the pass-through of future interest rates. Foreign-currency bonds can change exchange-rate risk. There is no optimal composition independent of the country: scenarios and costs over time must be tested. These are the risks addressed by debt management guidelines. 64

The connection with services is about continuity: a liquidity crisis can interrupt payments even before a debate about long-term solvency. A cash reserve and a less concentrated portfolio can provide time to respond. They do not, however, replace sufficient revenue, priorities and implementation capacity.

Fiscal discipline is the architecture that keeps commitments affordable over time. Austerity is an adjustment that reduces the deficit through lower expenditure or higher revenue. Administrative frugality means avoiding luxury, waste and privilege. They may be related, but they are not synonyms.

A privilege can be eliminated without reducing a service. Useful investment can be cut in ways that weaken future revenue-raising capacity. Ineffective expenditure can also be preserved simply because it has been given a social label. The analysis must compare alternatives and make explicit who gains and who bears the cost.

The timing of adjustment matters

In a recession, lower tax revenue and certain increases in spending can cushion households' loss of income. Requiring an immediate offset for every decline can amplify the shock. On the other hand, maintaining permanent commitments without funding can make debt more expensive and reduce future response capacity.

During the crisis in advanced economies, Blanchard and Leigh found an association between more intensive planned fiscal consolidations and growth below forecasts. The evidence warns of underestimated multipliers; it does not eliminate the need for adjustment in every context. 65

The proposed approach is to examine composition, speed, credibility and operational capacity. Across-the-board cuts ignore differences in results; permanent postponements ignore financing. Neither shortcut replaces a concrete review of programmes.

Why printing money is not the same as creating prosperity

In the modern monetary system, commercial banks create deposits when they extend credit. Central-bank asset purchases change reserves and portfolios. This is not the same operation as a budget transfer to a household. The Bank of England distinguishes these mechanisms. 66

The material reasoning still applies: greater nominal purchasing capacity can stimulate production when resources are idle. When supply does not respond, part of the adjustment may appear in prices, imports and exchange rates. There is no automatic ratio that turns every newly created monetary unit immediately into inflation. Nor is there permanently free financing.

Inflation caused by energy supply, food or other bottlenecks is not identical to inflation associated with excess demand. The appropriate response depends on the diagnosis. And even a government that pays in nominal terms can harm citizens if the currency loses purchasing power.

How to decide without falling into two extremes

The question is not whether to spend or cut as a matter of principle. It is: what combination preserves essential protection, increases productive capacity and keeps the trajectory financeable? This includes assessing subsidies, tax expenditures and tax cuts with the same rigour applied to social transfers. Distributional effects and future costs do not disappear when a policy is presented as reducing the size of the state.

1. Define the problem before the programme

The starting point should be an observable need: children not learning, waits for care, households without sufficient income or workers without access to opportunities. The programme is a means. Starting with the desired building, app or benefit can reverse the logic and leave the problem poorly defined.

Define the population, baseline, objective and alternatives. Separate what lies within the implementer's remit from what depends on other levels of government. The legitimacy to choose priorities must coexist with technical assessment of the means; experts do not replace debate about values.

2. Cost the full cycle

Record implementation, operation, maintenance, staffing, systems, updating and closure. A cheap building without maintenance can cost more over its useful life. A spending reduction can transfer costs to citizens, the courts or another agency without producing savings for society.

Distinguish projected, contractually secured and actually realised savings. Avoid financing a permanent commitment with temporary revenue or a reduction in waste that has not yet been demonstrated. Show which expenditure can no longer be accommodated when priorities change.

3. Organise accountability and capacity

Define who decides, implements, procures, verifies and is accountable for failure. Provide staff, data, systems and authority suited to the task. When a service requires regional scale, design cooperation and funding rather than merely transferring an obligation to a municipality without the means to fulfil it.

Technical career paths and administrative continuity are different from preserving every existing structure. The proposed objective is to retain knowledge and reduce dependence on individuals while allowing processes and providers to be reviewed.

4. Procure and deliver with evidence

Choose public provision, contracting or partnership according to the problem, risk and available capacity. The design should make price, scope, quality, responsibilities and the treatment of changes verifiable. The lowest initial price is not necessarily the lowest full cost.

Data and traceability need to reduce friction. Estonia offers an interoperability mechanism; the Australian case warns of the limits of automated decisions. Publishing a record proves that it was recorded, not that the original measurement was correct. 62 53

5. Measure access, quality and effects

Australia's Productivity Commission helps distinguish efficiency, effectiveness and equity. The framework below is a proposed adaptation for analysing services, not a list of those countries' official targets. 52

6. Review without erasing accountability

Combine audit, evaluation and channels for challenge. An indicator can be manipulated; an improvement can reflect the selection of easier cases or a change in definition. Comparisons should adjust for needs and local conditions without using those adjustments as an excuse to prevent any accountability.

Expand what has supporting evidence, redesign what fails and end what is unjustified, with a transition for those affected. This is a proposal for disciplined decision-making. It assumes neither that every service has a single financial indicator nor that citizens are merely customers of a company.

The practical sequence of building the system

The sequence below is a proposed method derived from the comparison. It is neither a universal chronology nor a ready-made programme for Brazil.

An example of connected policy: providing training does not solve a childcare barrier; building a facility does not automatically recruit its team; creating a savings account does not fund the generation already retired. Before approving the visible element, map the whole arrangement that makes it usable.

A rule for distinguishing efficiency from apparent savings

The proposal is to measure three dimensions together: what the state spends, what it manages to deliver and what happens to the person. This includes exclusion errors, access costs, waiting and continuity. New Zealand acknowledged the limits of isolated agency targets; the Robodebt inquiry emphasised the harm caused by processes that do not put users at the centre. 8653

More control also has a cost. The criterion is not to multiply requirements, but to choose checks capable of reducing material errors without making rights inaccessible. The more a decision affects a person's life, the more important explanation, review and accountability become.

Ageing is not just about paying pensions for longer

It changes demand for health, care and accessibility. In the Nordic region, 21% of the population was aged 65 or over at the beginning of 2025. New Zealand's scenario points to a declining ratio of working-age people to older people. The budgetary consequence also depends on labour force participation, productivity, health, rules and migration. 20 47

Raising the retirement age can improve financing, but distributes costs unevenly across occupations and health conditions. A complete analysis should make these differences explicit. Pension accounts can accumulate assets, but future care still requires professionals, facilities and real inputs.

Personal services do not scale like software

From a cost perspective, a digital register can serve more people at a small additional cost; a home visit or a lesson requires labour time. Technology can reduce repetitive tasks and improve organisation, but the same productivity should not be assumed across all services.

This distinction helps explain why efficiency does not simply mean reducing the number of public employees. We need to examine which activities can be automated and where fewer staff would compromise access, safety or quality.

Housing can absorb part of an income gain

Assessments of Sweden and Iceland include housing-related difficulties and risks. If housing costs grow faster than disposable income, individuals may experience little improvement despite higher wages or benefits. 23 37

The policy hypothesis is to compare transfers, housing supply, infrastructure and land-use rules. Supporting demand and increasing supply are different mechanisms. Their effects depend on local conditions; there is no automatic conclusion that every subsidy is fully absorbed into rents.

Economic participation and integration matter more than demographic labels

The financing effort depends on how many people work, how much they produce and how they contribute. Studying immigration requires distinguishing age, skills, employment, length of stay and the time horizon of analysis. Country of birth does not, by itself, determine fiscal effects, behaviour or religion.

The fact that 32% of Australian residents were born abroad proves neither that all immigration pays for itself nor that diversity makes a welfare state unviable. It shows why the argument needs specific mechanisms and data rather than cultural generalisation. 55

A national average does not reveal every inequality

The proposed evaluation should break down outcomes by income, age, territory and circumstances relevant to the service. Historically excluded groups are not well represented by a country's average. The same applies to people caring for relatives, those with interrupted careers and those living far from large cities.

Private risk must also be monitored. Iceland's crisis and Sweden's property-market warnings show that problems in banks and households can affect public capacity. The state's balance sheet must be read alongside the channels through which shocks spread, without indiscriminately adding all private debt to public debt. 19 23

Brazil does not have to choose between copying an entire country and ignoring all external experience. The OECD's 2023 structural assessment identifies problems with productivity, informality, investment and spending efficiency. It is a reference for development mechanisms, not a bulletin covering every legal and fiscal rule in force in 2026. 67

A proposed sequence of work

First, see the real commitment. Bring together the cost, population, funding and maintenance obligation. Analysis of each programme must connect with the full budget, debt and other priorities. Avoid making a promise look cheap because its cost has been transferred to another level of government or to the future.

In parallel, strengthen people and production. Basic education, health, infrastructure, security and predictable rules can expand opportunities. Each area requires its own implementation and evaluation. Raising the nominal amount received is not enough if access to work and services remains blocked.

Then expand what demonstrates capacity. Brazil's scale requires designs that can work in very different territories. Testing implementation under real conditions, documenting costs and outcomes and planning the transition differs from declaring an app to be a national solution.

Maintain continuous review. Fiscal consolidation should not happen only after financing becomes expensive. Nor should it be a pretext for removing resources from services without examining the effects. The review must address privileges, subsidies, tax expenditures and ineffective programmes with equal rigour.

What to learn from each group

From the Nordic countries, the connection between production, protection, care and institutions. From New Zealand, transparency about assets, future costs and social objectives. From Australia, the distinction between a public income floor and private accumulation, and the method of comparing services. From crisis cases, the cost of reacting late or adjusting without assessing the sequence.

None of these mechanisms eliminates distributional conflict. Technical design must make choices explicit, not pretend that political priorities have disappeared.

The information sheet every proposal should provide

This information sheet translates the whitepaper's approach into a focus on scarce resources and delivery. It does not presuppose validation of its scores, Rebate Social, blockchain or prediction markets. These are distinct institutional proposals that would require their own tests and clarification of their design. The application here is methodological, without product implementation. 1

A Brazilian case helps avoid the idea that it only works abroad

Ceará: incentives accompanied by capacity. Research presented by the World Bank in 2020 describes a combination of technical support for municipalities, funding linked to results, evaluation, local autonomy and leadership continuity. Its focus included literacy and stronger teaching practice. The study examines historical development, particularly between 2005 and 2017; it is not an updated assessment of the entire network in 2026. 88

The relevant point is complementarity: demanding progress from a municipality without capacity can widen the gap with those better prepared. Providing support without monitoring results also weakens the ability to learn and correct course. The case offers a mechanism to investigate within Brazilian institutions, without requiring us to become Denmark first.

What to ask of a Brazilian welfare-state proposal

What concrete problem does it solve? Does the service exist for those who will have the income to access it? Who pays for maintenance? Which group is left out? What risk passes to the budget? Is there local implementation capacity? What outcome would justify expanding the programme, and what would require it to be corrected?

These questions apply equally to creating a benefit, cutting expenditure, granting a tax concession or privatising an operation. The focus is not on imitating the size of another state, but on building verifiable capacity to improve people's lives.

The research agenda most likely to change conclusions

This is a proposed agenda, not a claim that the answers are already available in this document. The priorities would be to deepen evaluations of specific reforms in the selected countries and reconcile the fiscal database with its original metadata before using the annex for any decision.

The counterpoint on transfers and work

Banerjee and co-authors reanalysed seven experiments in six developing countries and found no systematic evidence of reduced work in the programmes studied. This prevents us from assuming that every beneficiary stops working. It does not prove that benefits of any amount, duration or degree of universality have the same outcome. 68

Design tests should consider gradual benefit withdrawal, net income, care, transport, health conditions and job availability. Hypotheses about incentives must be measured; psychological or ideological observations do not replace that work.

The limit of descriptive comparison

A sequence in time is not enough to establish causation. A country may reform during a global recovery; a policy may work because another was already in place; an average gain may hide a loss for one group. Studies with causal identification, administrative data and independent evaluations can reduce this uncertainty, but they too have scope and limitations.

The study's conclusion: a lasting welfare state needs to build capacity, fund commitments, deliver services and correct failures. Debt must be managed to preserve that continuity. The ultimate aim is neither to maximise spending nor to minimise the state: it is to make protection effective and sustainable for people.

Population and taxation

1. Population: rounded values for 2025. Nordic countries, start of the year; New Zealand and Australia, June; Brazil and the US, July. These indicate demographic scale, not real-time counts for 2026. 69 46 54 70 71

2. Taxation: OECD definitions, including compulsory contributions to public social security. Values for 2024 in the 2025 edition, except for Australia, whose available figure in this table is for 2023. Brazil comes from the 2026 regional edition. Taxes are neither social spending nor all public revenue; compulsory private saving is not the same as taxation. 8 72

The table does not show revenue per resident in real resource terms because that would require combining GDP per capita, purchasing power and service costs over compatible periods. Nor does it measure median disposable income. The correct interpretation concerns scale and revenue-raising effort, not how much a citizen receives back.

Checking outcomes without inventing an overall index

Life expectancy is a broad population outcome. Avoidable mortality measures premature deaths from preventable or treatable causes and is age-standardised. Neither measure, on its own, captures the welfare state's contribution. 73

3. Source: OECD, Health at a Glance 2025, table 1.2. Life expectancy: 2023 in the selected countries. Avoidable mortality: 2023 for Sweden and Australia; data between 2020 and 2022 for Denmark, Finland and Iceland, as noted in the table. The differing dates prevent the column from being treated as a precise contemporary ranking. Not available does not mean zero. 73

The data show why maximum prosperity needs to be broken down into dimensions. They do not prove that a specific tax-to-GDP ratio caused the outcome. Assessing a system would require adding access, household costs, learning, income, economic security and distribution while preserving periods and definitions.

Enable JavaScript to change the parameters. Initial examples:

A. Financing capacity

With GDP per person of R$ 50,000, a tax-to-GDP ratio of 34% and a monthly benefit of R$ 600 for every resident, the benefit uses 42.4% of revenue. The balance of R$ 9,800 per person still has to fund all other expenditure and interest. Changing only the population does not change that proportion.

B. Debt trajectory

With initial debt of 100% of GDP, nominal interest of 4%, nominal growth of 4% and a zero primary balance, debt remains at 100% of GDP after ten years. Constant conditions, with no assets, exchange rates or other adjustments; this is not a forecast.

The exercises are not combined automatically.

The working material contained a transcription attributed to the IMF fiscal database. A new extraction allowing all its values and metadata to be checked has not been completed.

The table is not reproduced on this page. This avoids presenting a transcription as independent verification. The explanations of financing and debt and the illustrative exercises remain available, with their own limitations.

The source documents are identified in references 63 and 74.

One study, two connected problems

The study brings together the construction of social protection and the ability to finance it over time. The seven central cases combine institutional history, implementation mechanisms, distinctive circumstances and illustrative examples. Analysing what may have been decisive is not the same as estimating an aggregate causal effect. 75 63

Provenance of references: the labels distinguish consulted sources, revisited sources and background references. The latter have not been audited again in full. Their presence in the bibliography does not mean real-time verification. Each data point retains its period, definition and caveats.

Limits of the fiscal data: the table covering 21 countries in the working material is not reproduced on this page. A complete re-extraction of the database and verification of its metadata are still required before publication.

Four types of content

Background documents: the social protection and public debt studies and the OpenGov3 whitepaper. Attributing information to a source in those documents does not, by itself, amount to independent verification.

External research: history, rules and assessments from statistical agencies, governments, multilateral institutions and academic research identified in the references. A government page describing a policy does not, on its own, establish that the policy produced all the desired effects.

Synthesis and proposal: connections between mechanisms, lessons and OpenGov3's operational information sheet. These are arguments open to review, not a national architecture already validated. The project's objectives guide the question but do not replace evidence on outcomes.

Simulation: exercises whose numbers and relationships were defined to teach a principle. Assumptions and limitations appear alongside results. They are neither forecasts nor estimates of real countries.

Comparisons that require care

Historical sources and assessments from different years retain their own periods. A recent publication does not turn older data into present-day measurements. The different year for Australia's tax-to-GDP ratio is made explicit. New Zealand's 2019 Wellbeing Budget is treated as historical.

No causal effect of population size, the tax-to-GDP ratio, austerity or Nordic identity has been estimated. Nor has a legal analysis of implementation in Brazil been produced. Conclusions about efficiency are analytical syntheses, not estimates of causal effects.

The selection is deliberate: benchmark countries and cases that reveal mechanisms or failures. It is not a representative sample of the world. The analysis does not attribute personality, behaviour or fiscal effects to religious groups or national origins.

A policy only looks simple when we hide part of the cost. The purpose of the study is to make the relationships between people, production, services, financing and review visible without turning the explanation into a promise of an automatic solution.