Zuzana Smidova
Lutecia Daniel
OECD Economic Surveys: Estonia 2026
Abstract
Public spending in Estonia is relatively efficient, and the budget framework and process are strong, with regular stocktaking of policy achievements. There is nevertheless scope to make improvements in how public services are delivered and adjust to policy priorities. Social spending has increased with generous supports for large families, while population ageing and rising demand for public goods will put pressure on pension spending, long-term care and healthcare. Moreover, Estonia faces wider fiscal pressures, such as higher defence spending, as well as costs associated with the climate transition, together with the need to ensure good quality public services and support growth. Mobilising additional revenues in a fair and growth-friendly way will necessitate important choices including a shift of taxation from labour to other revenue streams, such as corporates and property.
Estonia plans a fiscal consolidation to decrease the deficit and stabilise public debt (Chapter 1). In the longer term, it will also need to manage future spending pressures from ageing, climate and defence costs. Population ageing is likely to lead to a decline in the labour income share, as the number of those active in the labour market declines, leading to a lower revenue from labour taxation, which currently raises an important share of tax revenue. Current projections suggest that the working-age population could fall by 15-30% by 2060 (Figure 2.1), although increased participation of older people in the labour force may maintain the revenues from labour taxation at the current level (Crowe et al, 2022). At the same time, climate transition is likely to lead a loss of revenue from vehicles and energy taxes fall (Chapter 1). Estonia needs a long-term fiscal strategy to align spending and taxes in a growth-friendly and fair way. This chapter looks at how key public spending and revenues can be used more efficiently as the country faces population ageing and growing demand for public services.
Note: In Panel A, medium fertility variant of Statistics Estonia used. In Panel B, the medium scenario of the United Nation World Population Prospects 2024 used. OECD is weighted average of OECD countries.
Source: OECD Historical Population database and United Nations.
Estonia’s overall spending as a share of GDP is close to the OECD average, while several European peer countries with a similar level of incomes spend more than Estonia (Figure 2.2). This largely reflects relatively low spending on pensions and other social benefits, despite an ageing population. Spending on education is relatively high, given that it is a long-standing policy priority, while spending on healthcare has been low but is on an increasing trend. Defence spending, another key policy priority, was higher than in many other EU countries over the past decade and rose sharply from 2.1% of GDP in 2021 to 5% of GDP in 2026. Public investment is also strong, in part supported by various EU transfers. Public sector employment accounts for around a quarter of total employment with a public sector wage bill of 12% of GDP. Several studies indicate that Estonia’s public spending efficiency is generally above average (Box 2.1). This section reviews key spending areas that account for the bulk of public expenditure: pensions, long-term care, healthcare, family policy, education, public sector management and municipalities.
Note: In panel A and panel D, the value of the OECD and EU averages are weighted averages.
Source: OECD Economic Outlook 119 database and OECD Government at a Glance database.
At a macroeconomic level, the efficiency of the public sector has been estimated by relating inputs, such as public spending, to observable outputs (e.g. the number of students or hospital beds) and outcomes (e.g. healthy life expectancy, returns to education, poverty rates).
Broadly speaking, the academic literature uses two main quantitative approaches. The first is non-parametric and uses either the Free Disposal Hull method or Data Envelopment Analysis (DEA) to compare countries/regions and establish an “efficiency frontier” or best practice. This involves computing relative efficiency scores to show how far countries are from this frontier. Efficiency can be measured using either an input- or output-orientated perspective. Public sector performance is often estimated using a single composite indicator covering multiple dimensions such as education, healthcare, infrastructure, administration, which may rely heavily on data proxies or other composite indicators. Assumptions about returns to scale (constant, variable or non-increasing) also matter. This approach has been used for instance by Afonso and Kazemi (2016), Dutu and Sicari (2016), Sutherland et al (2007), Hauner and Kyobe (2008).
The second approach is parametric and uses a stochastic frontier analysis (SFA). This necessitates making assumptions about the specific production functions that generate the services and goods provided by the public sector, which can be challenging. However, it allows to take account of measurement errors as well as country-specific inefficiencies. Unlike the non-parametric approaches, standard statistical tests can be used to assess variables. This approach has been used for example by Grigoli and Kapsoli (2018) and Apeti et al (2023).
Recent estimates using the parametric approach place Estonia above world sample average (Apeti et al 2024), while non-parametric estimates show both input- and output-oriented efficiency close to the OECD average (Afonso et al, 2023). Together, these suggest a relatively efficient public sector.
Note: In Apeti et al (2023) higher score signals better efficiency. Estonia’s score of 0.67 suggests that on average it could increase its performance by 33% for the same level of resources, which is similar to the average country in the dataset. In Afonso et al (2024), input efficiency scores suggest by how much government expenditure could be lowered to obtain the same outcomes This was around 30% for an average OECD country and around 40% for Estonia in 2017. Output efficiency scores suggest by how much outcomes could be increased for the same level of inputs. For an average OECD country this was around 30% and 37% for Estonia in 2017.
Source: Afonso et al (2023), Apeti et al (2024).
2.2.1. Private savings should play a key role in future pensions provision
Estonia is transitioning from a pay-as-you-go social security-based pension system into one where private savings would play a larger role, reducing the cost to the public budget. However, today, private savings are voluntary and there is a risk that people will not save enough, leading eventually to pressures to spend more on public pensions than currently assumed. In 2024, 40% of those aged 65 and over were at risk of poverty, that is, living on household income below 60% of the national median, one of the highest shares in the OECD. Meanwhile, the net pension replacement rate of 38% for an average earner remains well below the OECD average and is among the lowest in the OECD (Figure 2.3). In 2025, the average pension was EUR 817 per month, 40% of the average wage. Unlike in most other OECD countries, the poverty rate among older people has increased compared to the early 2000s.
The three-tier pension system has a broad coverage, but the adequacy of benefits is low. The first tier, financed on a pay-as-you-go basis, is composed of a flat basic and a redistributive earnings-related component based on the number of contributory years. The current contribution rate is 20% (reduced to 16% for those also participating in the second tier). Upon reaching the statutory retirement age, a residence-based minimum pension of EUR 393 (48% of the average pension) becomes available. The retirement age is currently 65 years and is linked to developments in life expectancy, with a maximum annual increase capped at three months. Over time, this will result in one of the highest retirement ages in the OECD: a worker retiring after a full career in 2071 would be 71 years old (OECD, 2023a).
Early retirement is possible with up to 5 years prior to the statutory age, however with a penalty of around 30% and depends on the years of contributions (for example, to retire 5 years early, 40 years of contributions are needed). In 2024, around 20% of eligible people took an early retirement and this share has doubled since 2020. Postponement or drawing half a pension are also possible, but less popular, despite good incentives. Public pensions are indexed to developments in wages and inflation, weighted 80% to the increase in wages and 20% increase to the inflation. While price indexation is more common in OECD countries, given the low level of benefits and ongoing productivity catch up of the Estonian economy, reflecting partially wage growth is appropriate.
Note: In Panel A, OECD Pension models do not include private asset-backed pension savings. Panel B is based on Ageing Working Group’s long-term projections.
Source: OECD Pensions at a Glance 2025 database and European Commission – Ageing Working Group, OECD.
The second tier of the pension system, formed by asset-backed defined-contribution pension plans, was introduced in 2002. Around 62% of 18–60-year-olds hold such a plan. Initially, the second pillar was mandatory for new labour market entrants. Since 2021, that has no longer been the case, but new labour market entrants are enrolled automatically and can opt out. The minimum contribution rate is 6% (shared between the employee – 2% and the employer – 4%), and participating in the second pillar decreases contributions to the first pillar by 4 percentage points. Employees can opt to increase their contributions from 2% to 4 or 6% as of 2026. These contributions can be deducted from an individual’s taxable income. Pensioners can choose between a life-time annuity or pension over the average remaining life expectancy, neither of which is taxed. Other payout options are possible but subject to an income tax rate of 10%. Furthermore, a quarter of the working-age population holds a voluntary asset-based pension plan, which represents the third tier of the pension system. This third tier is used mainly by high-income individuals, who would benefit most from the possibility to withdraw lump-sum amounts at a preferential tax rate of 10%.
Despite recent improvements, contributions to the asset-backed plans remain below the level required to provide adequate future replacement rates (Figure 2.4). The average replacement rate from the first public tier of 49% is expected to fall to 31% in 2070, by when the asset-backed plans are expected to provide an average replacement rate of 13%, attaining an overall replacement rate of 44% (Ministry of Finance, 2024a) (Figure 2.3). Maintaining income level of an average wage earner at 70% of his/her pre-retirement earnings would require additional contributions to the third pillar of 5% of gross earnings (Foresight Centre, 2025a).
Investment restrictions in the second pillar were relaxed in 2019 to allow conservative funds to invest up to 10% in equity and other funds up to 100% which led to a shift of investment, with public and private equity representing over 60% of investment at the end of 2022. Management fees have come down considerably to around 0.5% of assets over the past five years. Both changes contribute to better rates of return in the pension funds, making the investments more attractive. An information campaign explaining the benefits of asset-based pension plans, including model saving scenarios, is under way, which is welcome. An online pension calculator has been available since the end of last year and includes a projection tool of future pensions.
Source: OECD Pensions at a Glance 2025 database.
Currently, people can withdraw all their pension savings under the second pillar without any conditions, although it is then subject to income tax and there is some delay in the funds becoming accessible. A policy change in 2021 led to around a third of contributors withdrawing their funds, equivalent to around 4.6% of pre-reform GDP, but the level of assets has recovered since. To counter withdrawals, those who withdraw funds are not allowed to rejoin the fund for 10 years. The authorities have submitted a legislative amendment of the rules that would shorten this period to 5 years, which is welcome. These changes also introduce an option for a partial rather than full withdrawal of savings. A well-designed framework that would include conditions of how to make up for the ‘missed’ contribution years, is needed. For example, Canada allows withdrawals for house purchase under some conditions and people are able to make up for the lost pension assets during a certain window. Switzerland also allows withdrawals for key events, such as setting up a business and emigration. Estonia should allow only withdrawals that are tied to key events and under specific conditions to achieve a better balance between liquidity needs and adequate pension savings.
2.2.2. Long-term care spending is likely to rise
Needs for long-term care are likely to increase as the population ages and conditions such as dementia become more prevalent. The uncertainty about the increase of expenditures on long term care is large, particularly when delivery of and eligibility for care evolve over time. Estonia spends around 0.9% of GDP on long term care, one of the lowest shares in the OECD, although comparisons across countries are fraught with differences in data reporting. Currently, one in four men and one in three women over the age of 65 are in need of long-term care in Estonia (OECD, 2024c). The share of costs covered by the public social protection is among the lowest in the OECD and much of long-term care is provided informally by families, which can negatively impact labour supply, notably of women (Figure 2.5) (OECD, 2024c). Long term projections that maintain the current status quo of public expenditure on long term care show almost a doubling by 2070 (EC, 2024). Should formal in-kind care expand to that of the EU average and care unit costs converge to the EU average, public expenditure on long term care could increase ten times, to 6.4% of GDP in 2070 (EC, 2024).
Municipalities are responsible for assessing a person’s needs and organising care, with an obligation to offer home care as a default. Benefits are means-tested but capped, and the current means-testing threshold is below median income. In 2023, public financing of nursing homes increased, reducing out-of-pocket payments but the early experience shows that the allocated financing is not sufficient, as nursing home care prices increased faster than estimated. The authorities recognise that the existing financing of long-term care is not sustainable and are reviewing policy options. As in other OECD countries, shifting to home-based care, better integration and coordination of health and social care services as well as increasing state funding are currently discussed. In 2025, White paper on the integration of health and social care was published, aiming to establish “welfare regions”, introducing cross-sectoral coordination of services, allowing for data exchange among professionals and the use of outcome-based financing. Constructing an integrated long-term care model can bring significant efficiency and quality gains. The experience of Sweden illustrates considerable scope for improving care outcomes and quality by managing the interactions between health and social sectors more effectively (EC, 2018). In the Netherlands, the nurse-led model of integrated care with considerable professional freedoms and responsibilities changed considerably community care (so-called Buurtzorg model), while some municipalities in Denmark created joint teams for interdisciplinary cooperation (OECD, 2025e). Possible reform options include regional one-stop ‘shops’ to coordinate and simplify access to long-term care services, creating joint clinical and care guidelines, single integrated information system to track beneficiaries and strengthening follow-up care for the elderly.
Funding models for long-term care vary across OECD countries, but government support tends to be a key component. Private insurance is available only in the United States and Japan, and even there it covers only a fraction of the overall cost. Reverse mortgages and home-equity schemes exist in a number of English-speaking countries. Given the uncertainty about long-term care costs, Estonia may consider setting aside funds for the future as done for instance in Germany, where a share of social long-term care contributions is reserved to pre-fund and smooth future care costs (OECD, 2024b).
Note: For Estonia, rules governing public social support in Tallinn use.
Source: OECD (2024) Health Policy Studies: Is Care Affordable for Older People?.
2.2.3. Spending pressures in healthcare are building
Estonia has historically spent a lower share of GDP than many other countries on health, but spending has been rising steadily since the early 2000s (Figure 2.6). Overall health spending as a share of GDP was 7.8% in 2024, with 6% of GDP of public spending and 1.8% of GDP of private expenditure, both below the OECD averages of 7.1% and 2.2% of GDP, respectively. In terms of outcomes, life-expectancy remains two years below the OECD average but increased by 8.5 years between 2000 and 2024, one of the largest increases in the OECD. Nevertheless, years spent in good health in older ages are among the lowest in the OECD, preventable mortality remains well above the EU average, particularly among men, and deaths due to alcohol abuse, drug overdose and suicide are double the OECD average.
The public healthcare system is well designed and efficient as discussed in detail in the 2024 Economic Survey of Estonia, with health outcomes consistent with the level of spending (Figure 2.6). However, medical staff shortages restrict access to care, in particular in rural areas, current revenues collected by the social tax are insufficient, inequalities in health outcomes remain substantial and out-of-pocket payments accounted for 22% total healthcare spending in 2023, higher than in most OECD countries and well above the national objective of 15% (OECD, 2024a).
Source: OECD Health Statistics database and Eurostat.
The biggest emphasis is on outpatient care, with almost 40% of public financing on primary and specialist outpatient care, slightly above the OECD average. This is followed by inpatient care, which accounts for 26% of public spending, below the OECD average of 28%. Care is organised around family doctors, who act as partial gatekeepers to specialist care, and provide services in a determined geographical area. A widespread network of public hospitals provides specialist and emergency care. Nursing care is available in hospitals, care institutions, and at home. While the Ministry of Social Affairs regulates and supervises the healthcare sector, the Estonian Health Insurance Fund is central to managing the financing and setting a minimum level of required service provision (OECD, 2024a).
The healthcare system has undergone numerous changes in recent years, lowering out-of-pocket payments, increasing supply of care and delivering greater value for money (Table 2.2). Scope for further improvement exists in terms of rationalising the hospital network. Implementing the 2024 Hospital Network Development Plan should help. The plan suggests merging of the two existing hospitals in the capital, which is currently under way, designating hospitals in Tallinn and Tartu as providers of advanced specialist care and integrating primary and social care in regional hospitals. Promoting wider use of generic drugs should be also pursued. The use of generics is below the OECD average, despite default prescriptions by the international non-proprietary name. Encouraging joint hospital purchasing and allowing purchases in other EU markets or directly from drug producers could help to decrease overall pharmaceutical costs.
Given the elevated mortality rates from treatable conditions, such as ischaemic heart disease, hypertension, stroke and cancer, prevention and treatment need strengthening. Monitoring of health conditions among older people could be improved, as well as participation in cancer screening programmes. Lifestyle factors also need addressing. Prevention and early treatment can bring significant efficiency savings, as treatment is often most costly at later stages of illnesses. It is therefore welcome that authorities are currently exploring potential for value-based payments based on pathways, which would bundle services across the continuum of care, target acute events or chronic conditions, thereby improving financial incentives for care providers. The healthcare system is digitalised, and Estonia performs well in terms on primary data use. However, data quality, interoperability and linking of datasets can be improved, which would allow faster and wider access to further leverage its use (OECD, 2024b).
Source: OECD.
Note: In panel B, only data for doctors on an employment contract are shown.
Source: OECD Health Statistics 2025 database.
A key challenge is to ensure a sufficient number of health workers. The number of health workers relative to the population is low compared to other OECD countries. The ratio of nurses relative to the population has been broadly the same over the past two decades, while it has risen in most OECD countries, while the ratio of doctors in Estonia has increased although this increase has been among the lowest in the OECD (OECD, 2024b). Moreover, almost a half of the doctors are older than 55 years, which is well above the EU average. A persistent shortage of health workers is a result of an ageing workforce, migration and an insufficient number of medical and nursing graduates. This has led to long waiting times and lower access to healthcare. The latest collective agreements increased salaries in healthcare by 20% in 2023, 10% in 2024 and 10% in 2025. This should bring the ratio of nursing salaries broadly in line with the OECD average (Figure 2.7). Nevertheless, in neighbouring Finland nurses can earn more than a quarter more, when adjusted for differences in purchasing power. For general practitioners, the overall compensation is already competitive but can be improved further for salaried doctors.
To address staff shortages, the authorities are training more medical staff, which is welcome. There is significant potential to attract more foreign doctors and nurses as the barriers to entry into the medical profession are high. Non-EU qualifications are not easily recognised. The government is developing legislation for recognising foreign non-EU medical qualifications. This should be accelerated while ensuring foreign qualifications meet minimum quality criteria. Estonia currently requires practicing healthcare professionals to possess a high-level proficiency of Estonian. While it is important for people to be able to access medical care in Estonian, there is scope to relax this requirement initially and allow for greater flexibility so that foreign nurses and doctors can start practising earlier, while further improving their knowledge of Estonian.
The current funding of the public system by health contributions levied on wages and tax revenues earmarked as a payment for pensioners is insufficient to meet rising demand and costs. The Health Insurance Fund, the paymaster, has in recent years drawn down reserves built up during the pandemic. The social security tax raises around 4.8% of GDP and the public transfer for non-working pensioners around 0.8% of GDP while the health spending of the fund reached a peak of 5.9% of GDP in 2024 and is projected to remain at that level in 2026. Numerous long-term projections confirm the need for additional financing with a financing gap between 1.7-2.6% in 2040 (Aaviksoo et al, 2011; Koppel et al, 2020; Piirits and Vork, 2015, 2016; Vork and Piirits, 2023). Projections by the EU’s Ageing Working Group, which serve also as the national long-term projections, show that public health expenditure, excluding long-term care, can increase by over 1% of GDP by 2070 (EC, 2024). Population ageing and increasing incomes will bring rising consumption of healthcare. Moreover, healthcare expenditure tends to grow faster than GDP due to non-demographic factors, such as technological change and Baumol’s cost disease, as illustrated by the experience of other high-income OECD (Hagist and Kotlikoff, 2009; IMF, 2010). Any systematic underfunding could result in rationing of care, stalling of the progress achieved so far.
Ensuring long-term sustainability of healthcare financing in an ageing society will require a combination of efficiency savings and raising additional revenue. Private health insurance exists but is not widespread. It has gained in popularity since it became part of non-monetary employee benefits, as up to a certain threshold it is tax free. However, mandating private supplementary private health insurance is not an effective option because the market is small and out-of-pocket are payments already high, while the design of the public system is efficient but could be impacted by a growing private sector. To raise revenues and in line with many other OECD countries, making pensioners pay health contributions, other than those on low incomes could be one option. Nevertheless, increased financing from general taxation is likely to be needed and would be an effective solution with appropriate safeguards (OECD, 2024b).
2.2.4. Adjusting family policy from universality to targeting, from cash benefits to services
There has been a sharp increase in spending on family policies due to policy changes in recent years that expanded the generosity of universal benefits for families, pushing up spending in this area from 1.7% of GDP in 2005 to a peak of 3.14% of GDP in 2021, well above the OECD average of 2.35% of GDP. Estimates of 2023 show a similar level of spending. The aim has been to make Estonia a more attractive place for family formation and raising children, supporting the well-being of children, reconciling work and family lives and raising fertility (Ministry of Social Affairs, 2023). A number of other countries, including Denmark, Finland and Hungary, dedicate similar amounts of public resources to family policy, although their share of cash benefits is smaller. In Estonia, family policy outcomes are assessed based on indicators such as child poverty, social exclusion, fertility rate, female employment and the gender pay gap.
In recent years, two comprehensive assessments of family policy have been published, and ad hoc studies on various aspects of the family policy are regularly commissioned by the authorities (Ministry of Social Affairs, 2015, 2025). Child poverty rates have declined and are now below the EU average, at around half of the average poverty rate of the population (Figure 2.8). Nevertheless, similarly to other OECD countries, a declining trend in total fertility rate has continued, falling from 1.7 in 2019 to 1.2 in 2024. The experience of OECD countries points to complex dynamics between family policies, labour market participation and fertility (Fluchtmann et al, 2023). Decisions to have children are highly individual, but comprehensive public support, as well as employment of both partners, can be critical in deciding to have a(nother) child. A correlation between the decline in fertility and rising housing costs in many OECD countries point to another constraining factor (Fluchtmann et al, 2023). A recent analysis by the Ministry of Social Affairs echoes some of these findings also for Estonia (Ministry of Social Affairs, 2025b).
Source: OECD Social Expenditure database and Eurostat.
Source: OECD Tax-Benefit database.
The recent considerable expansion of family policy included an introduction of an allowance for large families (three and more children) in 2018, one month of father leave introduced in 2020, and overall benefit increases in 2023. Some of these have been clawed back since and further changes decreasing generosity took effect in 2026: the maximum level of parental benefit decreased from triple to double of the average wage (i.e. EUR 3 806 per month), a cap on maternity benefit at a similar level was also introduced. At the same time, a previous limit on parental earnings will be abolished. The elevated inflation of 2022-25 eroded some of the benefit generosity. While family benefits are set in nominal terms and thus subject to ad hoc changes, automatic indexation of child benefits had been discussed. In 2023, indexation of family benefits partially to price and wage developments was planned but abolished due to budgetary constraints. While several OECD countries link child benefits to price developments (e.g. Finland, Belgium, the United Kingdom), indexation to wage developments is less frequent (e.g. Denmark, the Netherlands).
Note: Panel B presents calculations using the EUROMOD J1.0+ model, EU-SILC 2022 and HBS 2015 data. Baseline for 2026. Household disposable income is equivalised using the OECD modified equivalence scale.
Source: Ministry of Labour and Social Affairs (EUROMOD 2024) and OECD (2026a, forthcoming).
Universality of both the child allowance and the allowance for large families warrants review (Figure 2.9). While a number of OECD countries provide universal child-benefits, others use income thresholds or tapering to avoid providing the full amount to high-income households. Simulations from the OECD Tax Benefit model show that the benefit for large families (with three and more children) are generous (Figure 2.10) (Table 2.3). The maternity benefit, available up to three months after the birth, provides a 100% replacement rate of the previous wage’ up to double of the average wage. Moreover, the parental benefit provides 17 months of the full-rate equivalent paid leave, one of the longest among OECD countries. In addition, the parent who is drawing the benefit is entitled to monthly contributions equal to 4% of the national average wage into their asset-backed pension plan for a maximum duration of three years per child, regardless of whether the parent returns to work. Withdrawing the child allowance gradually, for instance from an income threshold of the average wage by 20cents for each additional euro and adjusting it for household size would impact mostly the upper 7th and 8th income deciles, whose incomes would decrease by around 1.3% (Figure 2.9). It would lead to budgetary savings of around EUR 118 million or 0.3% of GDP. Poverty rates would remain unaffected, while inequality – measured by the GINI coefficient - would decrease (OECD, 2026a, forthcoming).
Currently, the emphasis is on cash benefits that represent two thirds of the spending, with in-kind services, such as childcare provision and tax breaks playing a smaller role (Figure 2.8). To manage the overall costs of family supports and make them more efficient, these benefits should be targeted. Targeting could take the form of caps or gradual phasing out of the benefits at higher income levels to avoid payments to those who do not need them, or a system more narrowly focussed on means-tested support for low-income households. Means testing of family benefits has been discussed but technical hurdles in determining household income limit immediate progress (ERR, 2025a). Furthermore, changing dynamics of family formation, dissolution and living arrangements are on the rise in OECD countries. As a result, an increasing number of children in shared custody divide their time between both parents’ residences. Around 7% of children alternate between two households on average across European OECD countries and in Estonia this share is estimated above 10% (OECD, 2025c). Accounting for the realities of diverse family and residential arrangements linked to evolving family life can be challenging. In principle, both parents should be able to receive a benefit, allowing children to be counted in two residences. In Norway, a number of family supports take into account such family arrangements (housing, social assistance, child support) (Hakovirta et al, 2024). While increasingly complex to apply, countries have nevertheless overcome these issues in order to improve the targeting of family benefits. Withdrawing them from higher income earners creates fewer risks than means-testing at lower levels of income.
Total family benefits for a two-parent, two-earner family with a youngest child aged 6, by number of children, as a % of average full-time earnings, 2023
Note: Estimates based on a two-parent, two-earner family with a youngest child aged 6, with one parent working full-time (40 hours per week) and one parent working part-time (20 hours per week), both on wages at the median of the full-time earnings distribution. Children are aged three years apart. Average full-time earnings/the average full-time wage refers to the average gross wage earnings paid to full-time, full-year workers, before deductions of any kind (e.g. withholding tax, income tax, private or social security contributions and union dues). See the OECD Tax and Benefit Systems website for more detail on the methods and assumptions used and information on the policies modelled for each country. Data for Chile refer to 2016.
Source: OECD estimates based on the OECD Tax-Benefit Models.
2.2.5. The education system delivers world class results but faces shortages and wage pressures
Estonia spends the equivalent of 5.7% of GDP on education, the same as the OECD average, while it achieves world class educational outcomes (Figure 2.11). Expenditure per student in primary to post- secondary education increased by 20% during 2015-22, more than in the average OECD country. In primary education, 15-year-old students ranked among the top OECD performers in reading, mathematics and science in the Programme for International Student Assessment (PISA) in 2022. Differences in learning outcomes between boys and girls, and between schools are modest (OECD, 2024c). These strong outcomes extend into adulthood. In 2023, the Estonian adult population scored highly in the Programme for International Assessment of Adult Competencies (PIAAC). While the level of adult skills in many OECD countries has remained stable or deteriorated over time, in Estonia proficiency in reading and maths between the two waves of PIAAC improved.
These very good results are supported by widespread attendance of pre-primary education, high levels of school autonomy and transparency of school results as well as highly educated teaching staff. While the compulsory education starts at the age of seven, 89% of children aged 3 to 7 attended early childhood education and care in 2024. Spending per student in pre-primary and primary education is at a similar level. Pre-primary staff are required to have tertiary education. Primary schools have a considerable autonomy in terms of curriculum and student assessment, and various school results and outcomes are made public (via an information website). There are around 30 private schools, enrolling 5-7% of students, but this is a growing sector.
Source: OECD Education at a Glance 2025 database.
In principle, children go to the closest school in the neighbourhood, but parents can apply to other schools. At primary level, around 45% of pupils do not attend their nearest school, at secondary level this share rises to 60% (ERR, 2025b). Due to the possibility of school choice, student composition differs across schools as informal mechanisms of stratification take place, based on ‘reputational sorting’ and oversubscription (Ferraro et al, 2026). This has led to an emergence of ‘elite schools’, which can select students due to high demand. These public elite schools, attended by around 5% of students, outperform the average schools and do better than the private ones. Given that ‘elite schools’ are associated with higher parental incomes, this trend can widen differences between schools, aggravating intergenerational income mobility. Such developments also undermine inclusion of special education needs students into mainstream classes, one of the policy goals since an inclusion reform was implemented in the mid-2010s (Ferraro et al, 2026). Assigning places to over-subscribed schools by lottery or applying quotas reserved for pupils from unfavourable socio-economic backgrounds and those with special needs should be considered.
The share of young adults with low educational attainment remains an issue. While there are considerable regional, nationality and gender disparities, according to the national data more than 15% of 18–26-year-olds did not continue education beyond lower secondary level in 2024 (Statistics Estonia, 2025). Although this share is somewhat smaller using internationally comparable OECD data it remains above the OECD average and although during 2005-2014 it declined, developments since have been mixed (Figure 2.12). This share has remained broadly constant in the past eight years despite the policy area receiving increasing attention. Drop-outs tend to be more common among men and those of Ukrainian and Russian nationality, as well as in regions such as Jarva, Valga and Laane, where drop-out rates reached above 20%, but also among pupils with disabilities (European Commission, 2025). The authorities have extended compulsory education by one year to 18 years of age in 2025 and a reform of the vocational education training system is under way.
Source: OECD.
Tertiary education attainment has increased steadily from 40% in 2015 to 43% in 2024 of the working-age population. The share of young attaining a master’s or equivalent degree has increased from 16 to 20% between 2019-2024. Young adults between the ages of 25-34 with a bachelor’s degree earn on average 23% more than those with only upper secondary education. The wage premiums increase to 50% for masters, doctoral and equivalent degrees, with the ICT and health study fields reaching the highest salaries (OECD, 2025). However, completion rates have been subpar. Only 40% of students enrolled in bachelor’s degrees complete within the theoretical duration of the programme, 57% with one extra year and 66% with three extra years.
Introducing tuition fees in tertiary education could lead to shorter completion times, as well as steer the choice of study fields to improve job matching, while providing some public expenditure savings. Estonia spends the equivalent of around 1.5% of GDP on tertiary education. This would have to be accompanied by means-tested scholarships or loans as educational inequalities persist across generations. While more than half of young adults with at least one tertiary-educated parent attains tertiary education, only 17% of those whose parents did not attain upper secondary education will go to higher education, well below the OECD average of 26%. Equitable access to tertiary education is key to social mobility, as educational achievement strongly influences success at the job market (OECD, 2025b).
The education sector has been marked by a shortage of teachers, similar to other OECD countries. The attractiveness of the profession has declined, with around a third of new teachers declaring teaching as their first choice, well below the OECD average of 58%, and the share of teachers considering leaving the profession over the medium term has increased (OECD, 2024h). Teaching staff in lower education is among the oldest in the OECD. The average age of teachers is around 50 years and the share of those over 60 years old has increased in recent years. Furthermore, the share of qualified teachers has decreased over the past decade. At primary level, 76% of teachers met the qualification criteria in natural sciences in 2023, falling short of a goal of 90% (National Audit Office, 2024). The situation is likely to worsen as compulsory education has just been extended by one year to the age of 18 and as teaching is gradually moving to Estonian-only instruction.
Low salaries, heavy workload and limited career prospects have been identified as major deterrents (Ministry of Education, 2025; ERR, 2025). Starting salaries in the education sector are indeed among the lowest in the OECD and the recent episode of fiscal consolidation limited salary increases, although they are being increased by 10% in 2026 (Figure 2.13). The average salary in the public education sector is at around 90% of the average wage, while one of the policy objectives is to raise teacher salaries to 120% of the average wage. A new career system is being rolled out in 2026, introducing four categories linked to qualifications and experience, which will partly be linked to pay progression. Other measures to improve teacher retention include increased availability of upskilling, leadership developments and guidelines on recommended workload and class sizes. Implementation of the new standards and career progression model lies largely with municipalities, who run primary schools.
Statutory salaries of teachers in public institutions with the most prevalent qualification, 2024
1. Weighted average of the statutory salaries across different subnational entities.
2. In practice, many teachers obtain higher tertiary degrees during their service and are placed in a higher salary range.
3. Combination of different salary scales for the same ISCED qualification requirement.
Source: OECD Education at a Glance 2025 database.
2.2.6. Public employment increased in recent years, but management follows sound practices
The public sector employs almost a quarter of all employees in Estonia, with an annual payroll equivalent to around 12% of GDP in 2025. This is on a par with the EU average, but above the OECD average and many other OECD countries (Figure 2.14). While public employment has remained relatively stable over time across the OECD, in Estonia it has increased by over 15% since 2010. Employment in state-owned enterprises is also among the highest in the OECD. The central and local governments employ around an equal share of public employees. Overall, public sector pay is competitive with salaries on average higher than in the private sector, but there are important differences between low and high-skilled workers. Wages of low income-employees in the public sector are often better than in the private sector, while this is reversed for high-income employees (Kong, 2025). This comparison may be distorted by large tax incentives for corporatisation of high-earning individuals (See Section 4.3.3), which depress the average private sector wage. In 2025, public sector compensation grew 5% and the growth is projected to slow to 3.8% annually in 2026-28, below the average nominal GDP growth during this period (Ministry of Finance, 2025a).
Source: OECD Economic Outlook database, OECD The Size and Sectoral Distribution of State-Owned Enterprises database and OECD Government at a Glance 2025 database.
Ministries have a relatively large autonomy in managing human resources, including around recruitment, remuneration, learning and development, and succession planning (Figure 2.15). While there is no single model of best practice across OECD countries, public management policies need to strike a right balance between recognising that each organisation and its leadership faces different workforce and talent development needs and maintaining some degree of central oversight and co-ordination to enable government-wide workforce transformations, address common challenges, ensure minimum standards across government, uphold common principles and statutory obligations (where applicable) of merit and fairness, and avoid uneven pay scales or inter-ministerial competition for employees (OECD, 2025c).
Index of delegation of management of human resources in central administration, 2024
Source: OECD Government at a Glance 2025 database and OECD (2024), Public Service Leadership and Capability Survey.
Pay scales and indexation of public sector wages are publicly available. Salary scales differ between civil servants and support staff, with salaries of civil servants based on job classification and the level of responsibility. Wages are indexed to wage developments in the economy (80%) and inflation (20%). Nevertheless, due to the recent fiscal consolidation, the allocation for public wages will be frozen in nominal terms during 2024-28. This does not apply to salaries of the security forces and some other public sector workers such as teachers. A proposal to reform the two-track system of civil servants and support staff has been put forward, which would ultimately unify employment conditions, as well as introduce a maximum employment for senior officials of 8 years and extend the managerial fixed-term contracts to lower levels.
Performance assessments happen on an annual basis across the administration. Performance-related pay, introduced in 2012, applies to discretionary bonuses that tend to be modest and constrained by the budgetary situation. It is available mainly to senior officials and specialists. Nevertheless, in local governments performance assessments are less common, and serve mainly to assess training needs, with only a small share being used for awarding bonuses or setting pay (World Bank, 2024). Reporting on the pay-gap is becoming increasingly more common. Senior managers and certain mid-level roles are employed on fixed term contracts. Notice periods are comparable to those in the private sector and depends on the length of employment, with 15 days for contracts employment under one year to 3 months for contracts over 10 years. Severance pay for redundancy – one month for employees with 5-10 years of service and two months for those over 10 years of service – is the same as in the private sector.
2.2.7. The budgeting framework is robust and there is a focus on policy goals
Estonia’s fiscal policy has been historically prudent, supporting low debt and modest spending and the fiscal framework provides a strong medium-term anchor. Fiscal policy is set out in a medium-term plan (State Budget Strategy), which covers four years on a rolling basis and is updated annually and includes both fiscal and sectoral policy objectives.
Line ministries are required to submit annual performance reports to the Ministry of Finance and National Audit Office by June, which then feed into the preparation of a new state budget. Ministries are also obliged to include main stakeholders in the preparation and updates of ‘sectoral plans’. Four-year reporting on fiscal policy and structural reforms is submitted to the European Commission based on the State Budget Strategy. The medium-term fiscal plan is based on legislated measures and the latest macroeconomic projection.
In the most recent consolidation effort during 2023-25, all line ministries were asked to prepare zero-based-nominal growth budgets in 2024, but this was an exceptional exercise with the aim of identifying savings of EUR 150 million (10% of their expenditure over three years). So far, three key ministries have prepared these, and others are ongoing. The zero-based budgeting (ZBB) framework is built around three complementary principles: (i) comprehensive reviews of ministerial budget and activities assessing the relevance and efficiency of baseline spending that takes place every two years; (ii) thematic cross-cutting reviews covering specific policy areas or functions across ministries; and (iii) performance assessments of national development plans and EU structural funds to assess policy effectiveness (OECD, 2026b, forthcoming). The process starts with mapping and screening of services, followed by detailed analysis, leading to political decision-making. It places a strong emphasis on financial diagnostics, including analysis of carry-overs, spending patterns and under-utilised funds.
Estonia has also used departmental spending reviews. Two sectoral pilot spending reviews were carried out during 2016-18, covering around 3% of the state budget. The experience of OECD countries shows that to be effective, spending reviews require strong political ownership and commitment, and should have clear objectives and scope, transparency and accountability for implementation in addition to being integrated in the budgetary process (Tyggvadottir, 2022). The framework for spending reviews that aligns well with OECD best practices has been in place since 2020: a timetable ensures a link to budgetary process, the process is led by the government that also takes a decision on the results, the reviews are connected to the medium-term budget framework and use available evaluations. Introducing more regular or annual spending reviews would provide a structured way to systematically assess baseline expenditure over time, rather than relying on periodic exercises. To do so, capacity to carry out spending reviews should be strengthened within the administration.
In 2020, Estonia introduced performance budgeting (called activity-based budgeting) that feeds into the discussions preceding drafting of the state budget on a more systematic basis. The Ministry of Finance produces thematic one-pagers that feed into the annual pre-budgetary discussions and are updated and presented as a part of the state budget documentation, as well as an interactive dashboard of information on the actual fiscal year but also the last and next four years by performance area as well as line ministry or state agency. A key motivation for adopting performance budgeting was to address fragmented cooperation between ministries, lack of links between the budget, planning and objectives, lack of performance evaluation and monitoring. Performance budgeting can strengthen accountability by providing clear information on progress toward performance targets and on the actual results achieved, thereby catering for the growing need of key stakeholders for data-driven decision making (OECD, 2025c, forthcoming). While a number of OECD countries have adopted performance-budgeting elements, implementation remains challenging, and this is also the case in Estonia. Line ministries produce the performance-based annual reporting but a substantive connection between the expenditures and performance targets is low, debate in government and the Parliament continues to focus on the traditional “expenditure lines” of the state budget and changing the mindset to performance budgeting is challenging (National Audit Office, 2024). Functional audits are performed by the National Audit Office and cover a wide range of topics from individual programmes to processes such as procurement or budget execution. The office is constitutionally independent and reports directly to the Parliament. It monitors the implementation of its recommendations via follow up audits and reviews.
The Fiscal Council, established in 2014, issues opinions on both the macroeconomic assumptions and the draft budget, but its resources are limited. The central bank is also an active participant in public debates on fiscal policy. The role of the Fiscal Council in the public debate could be enhanced. For instance, it could provide fact-based analysis of trade-offs between spending and revenue and policy choices that Estonia is currently facing. Some independent fiscal institutions across OECD countries engage in explaining fiscal implications of different policy options to the wider public, costing legislative policy proposals, or political manifestos during elections. Larger budgets and more staff in these mean they can dedicate more resources to this as well as to effective communication activities, however smaller institutions such as the Irish Fiscal Advisory Council or Portugal’s Public Finance Council illustrate that it can be done also in smaller countries (OECD, 2025c). Taking on such a larger role is likely to necessitate more resources, as currently the Estonian Fiscal Council is governed by a board of five members, but the analytical staff is limited.
Stocktaking of policy achievements happens regularly and via multiple channels. Several strategic policy documents contain specific indicators that allow monitoring of progress: Estonia 2035, the Welfare Development Plan 2023-30, and the Estonian National Strategy on Sustainable Development. Statistics Estonia, the national statistical agency, publishes a dashboard (‘The Tree of Truth’) that monitors progress on the policy goals across 15 areas including education, energy, research and development and health in a user-friendly way (Box 2.2). The state budget documentation includes both simplified overview for each policy area, as well as detailed information on achievements. Other scoreboards and efficiency assessments are available for gauging the efficiency in local government (Section 4.2.9). The Foresight centre, a think-tank of the Estonian Parliament, carries out numerous analyses of wider long-term developments in the society and economy, identifying potential impacts on the country, including on sustainability of public finances. It engages pro-actively in communication of these issues to the wider non-technical audience.
The Tree of Truth is a visual tool featured on the website of Statistics Estonia, the national statistical office. It presents the status of key performance indicators by comparing the actual outcomes with quantitative policy objectives set out in three strategic policy documents: the Estonian National Strategy on Sustainable Development (Sustainable Estonia 21), Estonia 2035 and the Government Action Plan. Each indicator is represented as a coloured leaf:
Green leaf – the policy goal has been met;
Yellow leaf – progress toward the policy goal has been achieved;
Red leaf – deterioration or no progress towards the policy goal.
Certain indicators can appear across multiple policy areas, and the assessment can differ, as the assessment criteria can differ in the policy documents, and can aggregate other sub-indicators.
The branches of the tree correspond to 15 key policy areas: national security and defence, culture and sports, energy, education, information society, health, environment, internal security, cohesive society, welfare, foreign policy, R&D and entrepreneurship, agriculture and fishing, transport, effective state and rule of law.
The Figure 2.16 illustrates the assessment of the Government Action Plan against 18 performance indicators, some of them aggregated. The other leaves - not highlighted in the Figure – correspond to goals set by the other two policy documents (Estonia 2035 and Sustainable Estonia 21).
By clicking on any leaf, a side menu explains the policy goal and provides a direct link to the data. For example, various strategies as well as the government action plans contains a goal of increasing healthy life years to 62.6 for men and 61.5 for women by 2027. According to the most recent data, healthy life expectancy reached 56.8 years for men and 60.6 years for women in 2024. As a result, this indicator is displayed in yellow, signalling progress but indicating that further improvements are required to meet the 2027 target.
Assessment of policy goals in the Action Plan of the Government (coalition agreement)
Source: Statistics Estonia.
2.2.8. Rising responsibilities of local governments call for a review of municipal financing
Although Estonia is small in terms of population, considerable and persistent regional disparities exist. The country experienced already a considerable population decline in rural areas, with many moving to the capital region. This trend is projected to continue, increasing the need for adequate inclusion of the regional dimension in policymaking (Figure 2.17). Reducing regional disparities is one of the government’s long-standing policy goals, yet the regional dimension seems to receive limited attention (OECD, 2025d). Close to half of the population lives in North Estonia, which includes the capital of Tallinn, the only major metropolitan area and main economic hub, accounting for 62% of GDP. Tartu, a mid-sized urban hub with around 8% of the population and where part of the central administration is located, is surrounded by a predominantly rural area and located in the south of the country. The Tartu region accounts for just under 20% of GDP. In the north-east there are several small cities including Narva. The rest of the country, with around 20% of the population, is rural and remote, with limited access to urban areas. Economic disparities are visible along this rural-urban divide. GDP per capita in the capital region is twice as much as that in the poorest region of Central Estonia (OECD, 2025d).
Source: Statistics Estonia, Eurostat and OECD Centre for Entrepreneurship, SMEs, Regions and Cities (CFE) database.
Municipalities have limited spending autonomy and are largely financed by transfers from the central government. Estonia is among the least decentralised OECD countries (OECD, 2022). It has one tier of subnational government, consisting of 79 municipalities. A reform in 2017 abolished the intermediate level of 15 counties and established a minimum population of 5 000 habitants per municipality and a threshold of 3 500 for sparsely populated areas. The number of municipalities was reduced from 213 to 79 and the median population increased from just under two thousand to almost eight thousand. The average population increased to 16 653 inhabitants. While in terms of average population size, Estonian municipalities are still below the OECD small country average (30 804 inhabitants), the number of municipalities is well below the OECD “small country” average (OECD, 2022). The 2017 reform led to a decrease of around 10-15% of employment, mainly in the administrative support staff.
In 2024, municipalities posted a deficit of 0.4% of GDP and debt of 2% of GDP. The authorities expect the deficit to decrease to 0.2% of GDP by 2028 (Ministry of Finance, 2025a). Municipalities are responsible for around 23% of public expenditure, which is close to the OECD average of unitary countries. Nevertheless, fiscal autonomy of Estonian municipalities is limited (OECD/UCLG, 2022). They function largely as ‘paymasters’ of the central government, as almost 90% of their financing is earmarked via grants and subsidies. Around 40% of their expenditures go into education for which a block grant from the central government is based on the number of pupils and several needs indicators and is largely used to finance teacher salaries. Another large share is dedicated to healthcare, housing and transport. Such a high level of earmarking could create disincentives for efficiency gains and given projected population changes over coming decades, spending priorities are likely to change (Figure 2.17). The equalisation mechanism in place consists of an assessment of expenditure needs based on the size and age structure of the population, as well as fiscal capacity based on the income and property tax revenues received. Municipalities raise 14% of their own revenue, with the land tax the main source generating revenue equivalent to 0.15% of GDP (Section 4.3.5).
Economies of scale in service provision should be explored further, in particular in education and infrastructure. Some inter-municipal cooperation exists already in transport and development planning, but regional development should systematically happen at a larger scale (OECD, 2025). A network of 15 development centres at the level of former counties exist, but these appear to have made little impact so far. The low share of municipal revenue from own sources could be limiting incentives for inter-municipal cooperation. The central authorities plan to offer (financial) support for municipal merges. While an audit of the provision of public services in remote areas found no worsening in access following the 2017 administrative reform, a more recent analysis of the efficiency of local governments shows some backsliding in governance (i.e. management) in more recent years (Foresight centre, 2025; National Audit Office, 2021).
In 2025, government revenues reached 43.4% of GDP, mostly coming from taxation. At 37% of GDP, Estonia’s tax-to-GDP ratio was just below the EU average of 39% and above the OECD average of 34% (Ministry of Finance, 2025b). Estonia’s tax revenues were above what was collected in the neighbouring Baltic economies but comparable to the Central European economies in 2024 (Figure 2.18). The overall tax mix is skewed towards labour taxes, notably social security contributions, and VAT revenues, while the role of both corporate and property taxation remains low (Figure 2.19). Estonia needs to mobilise additional revenue in a fair and growth-friendly way, adapt to population ageing and to new challenges such as shifting from fuel taxes to distance charging. This necessitates making important choices about the future design on the tax system. To promote public debate and understanding, an ad hoc tax commission or a council of experts, civil servants and social partners could review and lay out options, including taxation of wages and capital income. In recent years, such bodies were in operation for example in Ireland, Belgium or Norway. Publication of alternative approaches and their impact would help inform the debate and ensure it remains evidence-based.
Note: OECD and the EU averages are unweighted.
Source: OECD, Revenue Statistics database and OECD Economic Outlook database.
Statutory income tax rates on labour and capital are identical at 22% and tax filing and administration is highly digitalised, with 98% of tax returns filed electronically, well above the EU average of 75%, which helps reduce compliance costs. Tax expenditures from various exemptions are low, at around 3.5% of tax revenue, equivalent to around 1% of GDP (Ministry of Finance, 2025b). These are reported regularly in the documentation accompanying the budget, which includes a projection for the following year. A recent comparison of EU countries illustrates that over a half of Estonia’s tax expenditures in the PIT system went to family policies (child tax allowance) while tax expenditures from reduced VAT rates are relatively limited (Turrini et al, 2024). Given increasing spending pressures, it is important to maintain a broad-based tax system with limited exemptions.
Source: OECD, Revenue Statistics database.
2.3.1. Some scope remains to broaden the VAT base and raise environmental taxes
The standard VAT rate at 24% is well above the OECD average of 19%, and there are two reduced rates (13% and 9%). The standard VAT rate increased by 4 percentage points over the past two years, most recently in June 2025, and is now among the highest in the EU and the OECD and similar to that of the Nordic countries (OECD, 2024f). A reduced rate of 5% that applied to certain media publications was abolished in 2025. At the same time, accommodation services were moved from the 9% rate to 13%. The VAT registration is required for businesses with an annual turnover of EUR 40 000 or more, exempting roughly 60% of businesses. The threshold is higher than in neighbouring Finland or Sweden (EUR 20 000) but below that of other EU countries such as France or Italy.
One way of assessing the efficiency of the VAT tax is the VAT revenue ratio, which captures the difference between a theoretical revenue obtained by taxing all final consumption by the standard rate and the actual collected revenue. In 2022, this ratio was among the highest in the OECD, illustrating that the VAT system is efficient. This is also confirmed by the so-called VAT policy gap that captures foregone revenues due to reduced rates and exemptions, which is one of the lowest in the EU. Nevertheless, the compliance gap, i.e. difference due to fraud evasion errors and insolvency, almost doubled between 2019 and 2023 (European Commission, 2024). The existing exemptions and reduced rates could be reviewed and reduced further. Currently, exemptions include postal services, education, insurance, financial services, leasing and sale of immovable property and non-profit activities. Reduced rates apply to online education, vessels and aircraft for international transport including equipment and fuel as well as goods and services provided on board of these vessels and aircrafts. Compliance could be further strengthened by mandating declaration of all invoices in VAT reporting, removing the existing threshold of EUR 1 000 and expanding e-invoicing, as planned for 2027.
VAT can be regressive because those on lower incomes spend a higher share of their income on consumption. This is the case in Estonia, where regressivity of the VAT system increased over time (Figure 2.20). The bottom income quintile paid 13% of household income in VAT in 2010, while for the top income quintile this represented only around 7% of household income. In 2020, this difference increased: the bottom income quintile paid 14% while the share for the top income quintile had decreased to 5.5% of household income (Eurostat, 2024). A majority of OECD countries apply reduced rates on basic food items based on distributional concerns, although the scope of the reduced rates varies, they decrease the overall efficiency of the VAT system, yielding higher gains to higher incomes. Such measures have been discussed in Estonia and would lead to a considerable revenue shortfall, although the exact cost would depend on details of included goods and the choice of the reduced rate. Targeted social benefits provide a more effective way of addressing distributional concerns and helping those on low-incomes.
Distribution of VAT paid by households as a percentage of their gross income
Note: Panel B showing EU average. Experimental statistics.
Source: Eurostat.
Excise taxes and duties on alcohol, tobacco, fuel and packaging raise revenues of around 3% of GDP, above the OECD average of 1.9% of GDP. Fuel excise taxes represent the largest share of this revenue, followed by alcohol and tobacco. While the EU sets minimum excise tax rates for many of these products, Estonia applies rates above these minima, in particular for alcohol and energy. Due to cross-border leakage, Estonia has to take into account developments in neighbouring countries when adjusting rates, as illustrated by an experience with increasing excise taxes on alcohol in 2015 and their subsequent lowering in 2019. Estonia should continue regional cooperation on aligning excise tax rate changes, which would help mitigate such spillovers and support revenue stability. With more than half of the adult population overweight, the relative price of an unhealthy diet should be increased (OECD, 2024b). A tax on sugar in soft drinks was planned in 2017 but never implemented. OECD countries such as Finland or France tax sugary drinks, which has been found to reduce sugar consumption, and Mexico and Hungary tax unhealthy food. Estonia should implement the sugar tax and introduce taxes on unhealthy foods more generally.
Environmental taxation raises an equivalent of 2.4% of GDP, above the OECD average (1.3% of GDP). Most emissions are already subject to a net effective carbon tax rate and a new car tax and registration fee that partially reflect emissions intensity has been introduced this year (Chapter 3). As climate transition advances and emissions fall, revenues will decline (Chapter 1). Excise duties levied on fuels currently account for about 4% of Estonia’s total tax revenue, which will erode over time with a rising share of zero-emission vehicles. Estonia should prepare to introduce of distance-based charging over time to replace these revenues, drawing on the experience of New Zealand and Iceland. Moreover, Estonia should consider strengthening tax incentives for car fleet renewal, as discussed in Chapter 3, by phasing out current reductions for older cars.
2.3.2. The tax system is skewed to labour taxes, including social security contributions
Personal income tax (PIT), levied at a flat statutory rate of 22%, raises an equivalent of around 7% of GDP, less than in many other European countries but close to the OECD average of 8.2% (Figure 2.19). Estonia applies one statutory rate and a basic tax allowance to individual incomes of EUR 700 per month, around a third of the average wage (Figure 2.23). Child tax relief and a tax allowance for pensioners were abolished in 2024. Social security contributions, levied at 33% on wages and on income of the self-employed, are high and raise revenues equivalent to 12% of GDP, well above the OECD average of 8.8% in 2024. These contributions finance pensions and healthcare. Taken together with income taxes, the effective all-in rate (average tax wedge) is relatively high compared to the OECD average and varies little across income levels, from 35% for an employee earning 67% of the average wage to over 40% for an employee earning 133% of the average wage or more (Figure 2.21). The effective tax rate on lower income earners is high compared to many other OECD countries. Marginal tax rates create disincentives to work and potential incentives to informal work. In combination with the design of the corporate taxation regime, high income earners have strong incentives to incorporate, given that their tax wedge of over 40% is almost double the statutory CIT of 22% paid only on distributed profit (Figure 2.24)
Average tax wedge decomposition by level of gross earnings expressed as share of the average wage
Source: OECD Taxing Wages database.
Under-reporting of salaries, prevalent in construction and hospitality services, remains an issue and lowers revenues from personal income tax and social security contributions. According to a 2024 survey, a quarter of the population knows someone who received envelope wages and 4% of respondents admitted to receiving such wages directly, however such surveys are often affected by underreporting bias (Turu-uuringute, 2025). Among those who received envelope wages, in two-thirds of the cases, it was at the initiative of the employer. Undeclared income does not seem to be an issue among platform workers, who are estimated to account for around 3% of employment. The tax authorities estimate that around 0.2% of GDP are lost due to underreporting and undeclared wages. Moreover, a special tax regime available for the taxation of ship crews operating outside the European Economic Area introduced in 2020 should be reviewed. Estonia could consider introducing a cap on the income taxed at a zero rate under this regime, as is done for instance in Norway.
Population ageing is likely to lead to a decline in the labour income share, as the number of those active in the labour market declines, leading to a lower revenue from labour taxation. Current projections suggest that the working-age population could fall by 15-30% by 2060. On the other hand, increased participation of older people in the labour force may maintain the revenues from labour taxation at the current level (Crowe et al, 2022). Revenue from taxing pensions may increase, but this depends on how pension incomes evolve. The design of the social contribution and personal income tax system could be adjusted to reduce the burden on labour income, particularly of lower earners. First, the tax burden could be shifted towards the wider base of the personal income tax. This would allow the marginal rate to be lowered and bring more income into the net, including capital income and pensions. While the social security funds are largely self-financing, the role of general taxation in funding them has increased and ‘rates of return’ within the social funds vary across individuals and cohorts, so the insurance principle is only partially applied.
Note: Redistribution is defined as the difference between Gini coefficient of market income and disposable income inequality, expressed as a percentage of market income inequality. Inequality is measured using the Gini coefficient.
Source: OECD Income Distribution database.
Second, the personal income tax system could be made more progressive to reduce the tax burden and marginal rates on lower income earners. PIT is one of the main channels of fiscal redistribution in OECD countries (Jalles and Karras, 2025). Moving away from a personal income tax with one flat rate towards a progressive rate schedule as used in most OECD countries would allow the tax system to raise revenue in a more equitable way. Several proposals for revenue neutral personal income tax reforms have been made in recent years. Such a reform can adjust the basic allowance and replace the flat statutory rate by a more progressive PIT schedule with additional rates (Table 2.4). Countries that have moved away from a flat PIT rate have often introduced additional tax brackets for higher income earners. Model-based simulations suggest that introducing a second tax bracket of 30% for incomes above three times the minimum wage would affect largely upper half of the income distribution. Estimates by the authorities using the EUROMOD model suggest that the largest effect would fall on the 7th to 10th income deciles, where incomes would fall on average by 1.4%. Among pensioners, this would impact only marginally the 9th and 10th income decile. Such a change would increase progressivity of the PIT and could yield additional revenue in the magnitude of 0.5% of GDP.
The initially flat income tax rate of 26%, introduced in 1994, fell steadily to 20% in 2015 and has been increasing since, most recently last year by 2 percentage points to 22% (Figure 2.23). The level of basic tax allowance increased significantly in 2019, was then eroded by inflation and stands under 35% of the average wage. During 2018-25, the basic allowance was phased out above a certain threshold. A further increase by 2 percentage points was cancelled last year against the backdrop of use of the national escape clause of the European fiscal framework for funding defence. Moreover, a withdrawal of the basic tax allowance at higher incomes was cancelled, effectively removing a so-called ‘tax hump’ in marginal effective tax rates. The tax allowance has increased from EUR 654/month to EUR 700/month in 2026.
Source: Ministry of Finance of Estonia.
Note: Espenberg et al 2023 calculation based on Euromod model and progressivity impact includes assessment of other taxes. * The impact is compared to the situation in 2022, when the personal income statutory tax rate stood at 20% and basic tax allowance was dependent on income. ** The impact is compared to a baseline where the statutory personal income tax was set to increase to 24% in 2026 and basic tax allowance was decreasing beyond a specific income level.
Source: Espenberg et al 2023, Bunda and Hanapppi (2025)
2.3.3. More revenue could be raised from capital and from firms
Estonia does not have a conventional corporate income tax and taxation of corporates yields 2.4% of GDP, below OECD and EU averages of 3.7% and 3.3% of GDP respectively (Figure 2.19). Since 2000, corporate taxes are levied only on distributed profits with undistributed profits exempt. The tax rate was initially set at 26%, decreased to 20% in 2007 before going up again to 22% in 2025 as part of a budgetary consolidation to fund rising defence expenditures (the “security tax”). There are nevertheless exceptions, under which a conventional CIT is applied: During 2015-24 a reduced rate of 14% applied to companies distributing profits regularly. Since 2025, banks (credit institutions) are subject to the regular annual corporate income tax of 18%. A special tax regime for small entrepreneurs exists, who pay an annual corporate income tax of 20%.
The corporate tax system should act as a backstop to the income tax system and ensure that capital and labour are taxed consistently, while avoiding that capital income, which mostly accrues to wealthier households, is ‘undertaxed’. Under the Estonian regime, distributed income and realised gains are taxed at a much lower rate than labour income given that they are not subject to social security contributions (Figure 2.24). This creates an incentive for company owners to retain their earnings. Estonia has no inheritance tax and estates are not liable for taxes on capital gains, but heirs would be required to pay taxes on the gains as ordinary income, making the system relatively favourable to capital and company owners. The authorities should continue efforts to ensure that manager-owners of companies to pay themselves a ‘reasonable compensation’ or a minimum wage, based on which they would pay personal income tax.
Note: In case of capital income, the effective tax rates (ETR) consider cumulative personal income tax and dividend income received by the individual and corporate income tax levied on the shareholder profits. The effective tax rates on labour also consider employers social security contributions. The cumulative individual and firm-level taxes are then expressed as a proportion of income received by the individual plus firm-level taxes. In the case of wage income, the ETR corresponds to the tax wedge. The difference is modelled for an individual earning five times the average wage.
Source: Hourani and Perret (2025) and OECD Tax Database.
Taxation of distributed profits was intended to favour investment, job creation as well as to attract foreign direct investment. However, a simple comparison of real business investment and FDI developments across the three Baltic countries does not show a marked change for Estonia following the reform, although the comparison is more complex and implicit tax rates were similar in the neighbouring countries (Masso et al, 2011). A firm-level assessment of the impact of the CIT change shows an improvement in liquidity and liabilities of Estonian firms following the reform, as well as the improvement in investment and productivity, although evaluating the macroeconomic impact of the reform would require a more recent assessment as past studies have looked at a relatively short time period (Box 2.4). Furthermore, these studies do not consider the wider impact, the fiscal impact or compare alternative policies.
Estonia adopted tax on distributed profits in 2000, with the aim of promoting investment, entrepreneurship and job creation (Masso et al, 2010). Model based studies showed positive impact on capital stock, output and consumption (Masso and Merikull, 2011; Funke and Strulik, 2003). Empirical studies assessing impact of the reform focused largely on firm-level perspective. They generally find an improvement in liquidity holdings and decrease of liabilities among firms, with some impact on macroeconomic variables such as investment and profitability, although most of them have covered only a short period after the reform and the macroeconomic effects have not been the main subject of interest (Table 2.5).
Using firm level data, Hazak (2007) finds that the reform increased the share of retained earnings and decreases companies’ liabilities. Masso et al (2011) confirm this finding using a difference-in-difference analysis and propensity matching scope of firm level data from the three Baltic economies during 1996-2003. They find a positive impact on investment and productivity, although measures of these two variables are limited due to lack of comparable data. Pikkanen and Vaino (2018) examine the impact of the reform until 2016 looking at a slightly larger country sample, going beyond the Baltics. Their findings corroborate those above, although they do not find an economically important impact on investment activity.
Source: Hazak (2009), Masso et al (2011), Pikkanen and Vaino (2018).
The taxation of distributed profits seems to have served well in gaining international recognition and building business capital buffers. Estonia scores well in terms of certain measures of tax complexity and compliance costs (Barrios et al, 2022). Nevertheless, in wider measures that encompass a comprehensive ranking of countries’ tax systems, including the quality of regulation, administration and practical implementation, countries such as Czechia, New Zealand or Sweden achieve better results (World Bank, 2025). The costs and benefits of a more conventional corporation tax regime along the lines of other OECD countries should be explored, as they have a potential to raise additional revenue while increasing efficiency. Since the Global Financial Crisis, the elasticity of investment to corporate tax rates has declined and other economic factors such as skilled labour, infrastructure, macroeconomic stability play an important role (Hanappi et al, 2023; Hanappi and Whyman, 2024; Feld and Heckemeyer, 2011). Although over the past decades statutory corporate tax rates declined across OECD countries, the tax base has been broadened, contributing to growing revenues in many countries (European Commission, 2025; OECD, 2025). This has not been the case in Estonia.
Banks and small entrepreneurs are already subject to conventional corporate taxation, i.e. taxing annual profits. Switching to such a regime for all businesses would allow taxing corporate profits at a moderate rate with a broad base. Smaller open economies often levy a relatively low tax rate to support their competitiveness, although evidence of the sensitivity of FDI to corporate tax rates is low. This regime could include favourable treatment for innovation expenditure and appropriate arrangements for expensing of capital. Special tax regimes for small firms can also help lower the tax burden, and Estonia already has such a regime in place. To ensure tax neutrality and support risk-taking, loss carryover provisions are important, especially for start-ups who take inherently more risk. Many OECD countries provide loss carry-over provisions (Australia, Denmark, New Zealand), although caps on how much tax deduction can be made annually also apply. The existing regime gives strong incentives to keep profits in companies, and creates a cushion for unfavourable economic conditions, but it can also ‘trap resources’ that are not invested productively or remain in small firms, which have generally lower productivity, affecting overall economic efficiency. Under the OECD Inclusive Framework on Base Erosion and Profit Shifting, Estonia will need to implement the EU’s directive on the global minimum tax agreement in 2030.
2.3.4. Property taxation should play a bigger role
Property taxation is minimal in Estonia, collecting revenue equivalent to only around 0.15% of GDP, one of the lowest shares in the OECD (Figure 2.25). The existing property tax takes the form of a land tax, and is currently being increased, following an update of land valuations. Beyond generating tax revenues, immovable property taxes are considered among the least damaging to growth and may also help to shift some investment out of housing into higher-return economic activity (Arnold et al., 2011). As noted earlier, revenues from the land tax are fully assigned to the municipalities, which can also set the rate within a given bracket (between 0.1% and 2% of taxable value, with the tax bracket for residential land 0.1-1%). A number of exceptions to the land tax are in place, such as for land plots where a principal residence is located (of up to 0.15 hectares in residential areas and up to 2 hectares elsewhere), and municipalities can grant a tax relief up to EUR 1 000. The authorities should direct the exemptions and any relief to low-income households only and consider a faster phase-in of the new valuation, at least for business property.
Once the current land tax revaluation has been fully phased in, an introduction of a tax on immovable property (buildings) should be considered. This will require building up a register of properties and their values. Such a tax should be carefully designed and implemented to avoid unfavourable impacts on lower-income households. To avoid the ‘home-rich, income-poor’ phenomenon and improve the political acceptability of such reforms, a deferral mechanism could be introduced, as done for instance in Denmark, or low value properties could be exempt from paying the tax. Increasing the revenue raising capacity of municipalities would also decrease the currently high level of earmarking, giving more leeway to decide on spending.
Other forms of wealth-related taxation used in OECD countries include inheritance. Although they do not generate a large share of revenues because of the small size of the tax base, various exemptions and low tax rates, they can raise some additional revenue. In the OECD, inheritance taxes raise on average 0.2% of GDP, and up to 0.7% of GDP in Belgium, France or Korea. They can be used to address distributional concerns, improving progressivity of the tax system. Such a tax should be focused on the recipients of inheritance, take a lifetime accumulation approach and limit exemptions and reliefs such as those for business inheritance and transfers (OECD, 2021). Distortive effects of inheritance and gift taxes on investment behaviour and work efforts of wealthy taxpayers tend to be smaller compared to those of labour and capital income taxes, and the effect on the labour supply of their heirs is significantly positive (OECD, 2021; Guvenen et al., 2023).
2024
Source: OECD Tax Database.
Source: OECD.
References
Aaviksoo, A., Kruus, P., Leppik, L., Sikkut, R., Veldre, V., Võrk, A. et al. (2011). Sustainable financing options for Estonian social security. Praxis Centre for Policy Research, Tallinn.
Afonso A., J. Jalles, A. Venancio (eds) (2023), Handbook on Public Sector Efficiency, Edward Elgar Publishing, Cheltenham, UK, https://doi.org/10.4337/9781839109164
Akgun, O., B. Cournède and J. Fournier (2017), “The effects of the tax mix on inequality and growth”, OECD Economics Department Working Papers, No. 1447, OECD Publishing, Paris
Apeti A. E., B Bambe, A.A. B. Lompo (2024), Determinants of public sector efficiency: a panel database from a stochastic analysis, Oxford Economic Papers, 2024, 76, 741-758
Arnold, J. et al. (2011), “Tax Policy for Economic Recovery and Growth”, The Economic Journal, Vol. 121/550, pp. F59-F80, https://doi.org/10.1111/j.1468-0297.2010.02415.x.
Bunda I. and T. Hanappi (2025), Options to strengthen the tax system in Estonia, Selected issues paper, International Monetary Fund, Washington.
Crowe D., J. Haas, V. Millot, L. Rawdanowicz and S. Turban (2022), Population ageing and government revenues: Expected trends and policy considerations to boost revenue, OECD Economics Department Working Papers, No. 1737, OECD Publishing, Paris.
Dutu R.and P. Sicari (2016), Public spending efficiency in the OECD: Benchmarking health care, education and general administration, OECD Economics Department Working Papers, No. 1278, OECD Publishing, Paris.
ERR (2025a), Coalition says child benefit system cannot change to needs based in near future, News of the ERR, Accessed July 10, 2025.
ERR (2025b), Expert: Education and wealth gap increasingly clear in Estonian schools, News of the ERR, Accessed December 2, 2025.
ERR (2025c), Study: National-level management failures behind Estonia’s teacher shortage, News of the ERR, Accessed December 2, 2025.
Espenberg, S., M. Motte, A. Vork, K. Lees, J. Masso, U. Varblane, M. Hayford (2023), Distribution of tax burden of Estonian households and impact analysis of the tax-free income reform, University of Tartu, Centre for Applied Social Science research RAKE, Tartu.
European Commission (2018), Challenges in long-term care in Europe, European Commission.
Ferraro S., K. Poder, T. Lauri (2026), Inclusive education and parental choice: How student characteristics affect school efficiency, Socio-economic Planning Sciences, February 2026, Vol. 103
Fluchtmann, J., V. van Veen and W. Adema (2023), Fertility, employment and family policy: A cross -country panel analysis, OECD Social, Employment and Migration Working Papers, No. 299, OECD Publishing, Paris.
Guvennen F., G. Kambourov, B. Kuruscu, S. Ocampo, D. Chen (2023), Use It or Lose It : Efficiency and Redistributional Effects of Wealth Taxation, The Quarterly Journal of Economics, Vol.138, Issue 2, May 2023, p 835-894.
Hagist and Kotlikoff (2009), Who’s going broke? Comparing growth in public healthcare expenditure in ten OECD countries, Revista de Economica Publica, 188(1/2009), Instituto de Estudios Fiscales
Hakovirta M., D. R. Meyer, M. Haapanen (2024), Shared residence and social security policy: A comparative analysis from 13 countries, International Journal and Social Welfare, 2024; 33:956-980
Hazak A. (2009) Capital Structure and Dividend Decisions Under Distributed Profit Taxation, Tallinn University of technology, Tallinn.
Hourani D. and S. Peret (2025), Taxing capital gains: Country experiences and challenges, OECD Taxation Working Papers, No 72., OECD Publishing, Paris.
IMF (2010), Health care spending issues in advanced economies, Technical Notes and Manuals, IMF, Washington.
Jalles, J. and G. Karras (2025), “Rethinking tax progressivity: a new look at its role in OECD economic growth”, Int Tax Public Finance.
Masso Jaan, Jaanika Merikul, Priit Vahter (2011), Gross profit taxation versus distributed profit taxation and firm performance: Effects of Estonia’s Corporate Income Tax Reform, Working Paper Series, 2/2011, Bank of Estonia, Tallinn.
Ministry of Education and Research (2025), TALIS 2024 Results, Ministry of Education and Research, Tallinn, accessed, December 2, 2025.
Ministry of Finance (2024), Thematic one-pagers: Internal Security, Ministry of Finance of Estonia, Tallinn.
Ministry of Finance (2024a), 2024 Ageing report – Estonia Country Fiche, Economic Policy Committee – Ageing Working Group, Ministry of Finance, Tallinn.
Ministry of Finance (2025a), Summer 2025 Economic Forecast, Ministry of Finance, Tallinn.
Ministry of Finance (2025b), State Budget Strategy 2026-2029, Ministry of Finance of Estonia, Tallinn.
Ministry of Social Affairs (2023), Welfare Development Plan 2023-2030, Ministry of Social Affairs, Republic of Estonia, Tallinn.
Ministry of Social Affairs (2025b), Supporting Childbearing and Childrearing: Analysis and Proposals, Ministry of Social Affairs, Tallinn.
National Audit Office of Estonia (2024), Shortage of qualified teachers is greater than previously thought, National Audit Office of Estonia, Tallinn.
OECD (2016), Skills Matter: Further results from the Survey of adult skills, OECD Publishing, Paris.
OECD (2021), Inheritance Taxation in OECD Countries, OECD Tax Policy Studies, No. 28, OECD, https://doi.org/10.1787/e2879a7d-en.
OECD (2023a), Pensions at a Glance 2023, OECD Publishing, Paris.
OECD (2024a), Pension Markets in Focus, OECD Publishing, Paris.
OECD (2024b), Economic Surveys: Estonia 2024, OECD Publishing, Paris.
OECD (2024c), Is care affordable for older people? OECD Publishing, Paris.
OECD (2024d), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris.
OECD (2024e), Adult Skills and Productivity: New Evidence from PIAAC 2023, OECD Policy Briefs, OECD Publishing, Paris.
OECD (2024f), Consumption Tax Trends 2024, OECD Publishing, Paris.
OECD (2024g), Corporate Tax Statistics 2024, OECD Publishing, Paris.
OECD (2024h), Results from TALIS 2024 – Country Notes: Estonia, OECD Publishing, Paris, Accessed Dec 2, 2025
OECD (2025a), Tax Policy Reforms 2025, OECD Publishing, Paris.
OECD (2025b), Taxing wages, OECD Publishing, Paris.
OECD (2025c), The cost of raising children across evolving family structures – new estimates and policy challenges, Policy Insights, Centre on Well-being, Inclusion, Sustainability and Equal Opportunity (WISE), OECD Publishing, Paris.
OECD (2025d), Strengthening Place-Based Regional Development Policy in Estonia, OECD Multi-level Governance Studies, OECD Publishing, Paris.
OECD (2025e), Good practices in delivering integrated care – examples from the Netherlands, Denmark, France and Ontario, Canada, OECD Publishing, Paris.
OECD/UCLG (2022), 2022 Country Profiles of the World Observatory on Subnational Government Finance and Investment, accessed October 26, 2025.
OECD (2026a, forthcoming), Work incentives and distributional impacts of recent and ongoing tax-benefit reforms in selected EU countries – Assessing Estonia’s 2025/26 tax reform package, OECD Publishing.
OECD (2026b, forthcoming), Recent spending reviews in OECD countries, OECD publishing.
Pikkanen P. and K. Vaino (2018), Long-term Effects of Distributed Profit Taxation of Firms: Evidence from Estonia, Lund University, Lund.
Piirits, M. and Vork, A. (2015, 2016), Update of the health insurance sustainability forecast model for the Estonian Health Insurance Fund, Praxis Centre for Policy Research, Tallinn.
Republic of Estonia (2020), Estonia 2035 Development Strategy, Republic of Estonia, Tallinn, accessed October 10, 2025.
Tyggvadottir, Á. (2022), "OECD Best Practices for Spending Reviews", OECD Journal on Budgeting, vol. 22/1, OECD Publishing, Paris.
Turrini A. et al (2024), Tax expenditures in the EU: Recent Trends and New Policy Challenges, European Economy Discussion Paper 212, Publications Office on the European Union, Luxembourg.
Turu-uuringute (2025), Shadow economy Survey for 2024, Population survey, March 12-18, 2025.
Vork A. and M. Piirits (2023), Current research on Estonian healthcare financing and updated scenarios – policy analysis, Tallinn.
World Bank (2024), Estonia – State of the public service: Evidence from a survey of public officials, Key findings from an empirical study, World Bank.