Public expenditure on unemployment, sickness and disability benefits was nearly 8% of GDP across the OECD in 2023. Expenditure on unemployment is around 1.6% of GDP and is mainly influenced by short term economic fluctuations. Expenditure on disability is around 5.9% of GDP, and cross-country differences cannot be explained by health status alone. This chapter sets out some of the current measures countries and economies are implementing to foster savings in this area. These measures fall broadly into a number of categories including tightening eligibility; increasing oversight; reducing benefit generosity; and strengthening incentives to re-enter the workforce. Common measures include reducing benefit duration, altering benefit structures, and expanding monitoring and compliance mechanisms to limit misuse. Some governments are also using active labour market policies such as training and tying benefits to job-search or reskilling requirements.
Restoring Public Finances
Enabling Effective Government
Abstract
Reform and savings measures related to unemployment and disability expenditure relate mainly to tightening eligibility, reducing generosity of benefits, and increasing oversight to avoid misuse. In several cases, these changes seek to strengthen the incentives to look for work, or to reskill. A greater focus on return-to-work measures and preserving people in work enhances the effectiveness of public spending in this area. Survey respondents are combining unemployment benefits with active job search or training requirements. In the area of disability and sickness benefits, aside from tighter eligibility, respondents are also focusing on prevention and early intervention as better co-ordination and streamlining of various social programmes can help achieve more cost-effective spending with improved outcomes.
Reform initiatives and savings measures
1. Tightening eligibility and increasing oversight
Narrowing eligibility criteria through stricter assessments of who qualifies and under what conditions (including increased job-search and training conditions), minimum ages and expanded means-testing.
Strengthening anti-misuse mechanisms through more frequent verification of incapacity, and greater oversight of sickness and disability claims.
Phasing out or restricting programmes that may create perverse incentives such as discouraging work, prolonging dependency, or are not well focused.
2. Reducing benefit generosity
Lowering benefit levels through reductions in baseline amounts, lowering of income replacement rates, and introduction of caps on total amount of support available.
Limiting benefit duration by shortening maximum entitlement periods and introducing reapplication requirements.
Adjusting indexation rules to constrain benefit growth over time.
3. Other measures to strengthen work incentives
Making unemployment benefit increasingly conditional on job-search, training or reskilling requirements.
Increasing sanctions for non-compliance (such as if recipient does not comply with employment-seeking required tasks), through penalties, waiting periods, benefit reductions and temporary suspensions.
Simplifying systems to reduce fragmentation and allow for better targeting of benefits.
Adjusting benefit incentives to avoid welfare traps.
Expenditure on unemployment and disability is intended to support working age people who are not in gainful employment. It is important to note that support for disability is significantly higher than unemployment expenditure in most countries (Figure 3.1 below). Employment and disability need to be considered alongside each other, as large cross-country differences in disability expenditure cannot be explained by health impairments alone. For example, sickness and disability expenditure in Norway as a percentage of GDP is nearly six times higher than it is in Ireland. Health impairment status cannot explain such large differences in expenditures. Rather, this may be explained by disability benefits frequently being more generous than unemployment benefits, differences in eligibility rules, and the role of disability schemes as labour-market exit pathways (Figure 3.1).
Expenditure on unemployment and sickness and disability benefits combined, 2023
Note: Data not available for Canada, Chile, Costa Rica, Mexico, New Zealand, South Korea, Türkiye, and the United States. OECD average is unweighted.
Source: OECD Public finance by function - government at a glance indicators, yearly updates, https://data-explorer.oecd.org/s/4hq.
However, the two components are evolving with different dynamics. In many countries unemployment benefits were a major source of support during COVID-19, and hence most countries experience a significant spike in spending in 2021. However, over time expenditure on unemployment as a percentage of GDP has been reduced across OECD countries. Conversely, expenditure on disability in most countries did not increase significantly during COVID-19, with a few exceptions, but at the aggregate level such expenditure has remained constant or slightly increasing, with several countries experiencing significant increases.
Source: OECD Public finance by function - government at a glance indicators, yearly updates, https://data-explorer.oecd.org/s/4hq.
These trends in expenditure on unemployment and disability have to be considered in light of demographic trends as most countries are expected to see declines in their working age population (Figure 3.3 below). This provides an overall context which explains the importance and the urgency of measures in this area, which are intended to extend working lives, both at older ages (see Chapter 2 on “Old age pensions”), as well as for women and younger age groups, as discussed below. For example, while the difference in employment rates for men and women has fallen by 7.6 percentage points between 2000 and 2021 in the OECD, this fall is slowing. Eliminating this difference in employment rates would increase annual OECD GDP per capita growth by 0.2 percentage points (OECD, 2025[1]).
Change in the working age population (20-64), 2022-2060
Source: Koutsogeorgopoulou and Morgavi (2025[2]); Figure 6B, based on United Nations 2024 World Population Prospects data; (OECD, 2025[1]).
Figure 3.4 and Figure 3.5 provide an overview of recent savings measures by RPF Survey respondents intended to limit expenditure on unemployment, sickness and disability. Many responses initially categorised as “Other” fit into two categories: ensuring greater levels of compliance, which involves efforts to reduce fraud and misuse of benefits, and increasing focus on activation, which involves active efforts to increase participation in the workforce. These will be analysed as such in the following sections.
The measures reported in the RPF Survey focus on identifying fiscal savings. They do not provide a complete overview of recent policy trends in the area.
Measures approved or submitted to parliament for the fiscal years of 2025 and 2026
Note: Results based on 39 RPF Survey responses. Measures reported as "Other" in the RPF Survey have been split into the following three subcategories, based on the qualitative information provided by respondents: "Shorten benefit duration", "Increase focus on labour participation" and “Other”. Data is not available for France.
Source: 2026 OECD Survey on Restoring Public Finances, Question 2: Unemployment and Social Assistance Benefits.
Measures approved or submitted to parliament for the fiscal years of 2025 and 2026
Note: Results based on 39 RPF Survey responses. Measures reported as "other" in the RPF Survey have been split into the following two subcategories, based on the qualitative information provided by respondents: "Ensuring greater levels of compliance, including reducing fraud and abuse" and “Other”. Data is not available for France.
Source: 2026 OECD Survey on Restoring Public Finances, Question 3: Disability and Sickness Benefits.
3.2.1. Adjust conditionality and/or eligibility rules
Adjustments to conditionality and eligibility rules were one of the most common approaches reported to generate savings for unemployment programmes and were also significant for disability programmes. For unemployment, several changes to conditionality rules to incentivise job searching and training and have been identified throughout the RPF Survey:
Czechia’s new super allowance provides greater income replacement levels to those in reskilling programmes. In Finland, those applying for social assistance must register as unemployed jobseekers. They must then seek full-time work, not just part time. Furthermore, 17-year-olds who have suspended their compulsory education will no longer be entitled to unemployment benefit.
Korea now requires individuals with repeated benefit claims to report job-seeking activity every two weeks (as opposed to the previous four) for the first six weeks. It has also strengthened face-to-face assessments for individuals who have received unemployment benefits more than three times within five years and requires submission of a job plan.
Slovenia has implemented a measure incentivising work specifically amongst the older unemployed. Workers accumulate unemployment-insurance periods while they are employed. When they lose a job and start receiving benefits, they normally “use up” these accumulated insurance periods. Once used, these periods cannot normally be used again. The exception for this is for people close to retirement, allowing them to continue using (or re‑using) insurance periods even after they’ve been exhausted. Under the new scheme, Slovenia has tightened this exception. Previously, someone could benefit from the “re-use” exception if they were 57 or older or had 35 years or more of insurance. Now they must be 59 or older, with 37 or more years of insurance and 16 months of insurance in the last 2 years.
Disability programmes have also seen several tightening measures. In Finland, the minimum age for receiving sickness benefit has increased from 16 to 18 years. Furthermore, participation in youth vocational rehabilitation generally would no longer entitle that person to rehabilitation allowance. A Danish government agreement will see individuals convicted of serious crimes lose their right to several public benefits, including supplementary disability pension. Furthermore, the scheme is expanded to cover more types of crime. A measure in the Slovak Republic aims to prevent people taking excess advantage of sickness benefit by requiring the treating physician to request the consent of the social insurance company to establish a new benefit upon termination of the original. Some conditionality modifications impact both unemployment and disability benefits. For example, in New Zealand, single 18- and 19-year-olds are not eligible for certain unemployment and disability benefits from 2025, if it is determined that their parents or caregivers can support them.
Changes in conditionality can go hand in hand with greater orientation of job search assistance to those who benefit from it most. Such assistance has been effective in several cases, including in Korea, where young recipients of unemployment insurance and social workers co-developed programmes through counselling, leading to an increase in job seeking intention (OECD, 2024[3]). Attaching job-search requirements to unemployment benefits, may involve costs, such as greater monitoring of job seekers, as we as the cost of job assistance. Much of this can be mitigated through use of digital tools, such as online portal where jobseekers upload applications and receive automated reminders. In cases where there are large numbers of jobseekers, agencies can prioritise monitoring, such as for those with a history of non-compliance or those who have been unemployed for longer. Funding decisions by central government towards agencies can also be based on performance, with agencies with higher success rates receiving greater funding.
While many of the measures discussed in this section are specific and focus on adjusting conditionality and eligibility, and some of the following sections focus on freezing, or reducing the levels of benefits, a key theme for many respondents has been to tackle the broader underlying conditions of labour markets, as well as long term unemployment (see Box 3.1 below). The broader challenges of labour markets have been analysed as part of the multidisciplinary ‘OECD Jobs Strategy’ (OECD, 2018[4]), which offers extensive insights on unemployment benefits and their interaction with labour market support. Adjusting benefits may involve trade-offs in terms of potential costs of job mismatch due to people taking whatever job they can, rather than taking time to find a job that matches their characteristics, with potential adverse consequences for productivity growth.
Unemployment has cyclical and long-term components. Cyclical unemployment, which arises from downturns in the economic cycle, is often temporary and can be mitigated through passive measures that stabilise income and sustain demand. These include unemployment benefits that help individuals smooth consumption until labour market conditions improve. In contrast, long‑term or structural unemployment reflects deeper mismatches between workers’ skills or geographic location and the needs of employers. This type of unemployment can benefit from active labour market policies, aimed at improving employability and facilitating reintegration into the labour force. This can include training, job search support, and even recruitment subsidies (Knapińska and Woźniak-Jasińska, 2024[5]).
Prior to the COVID-19 pandemic, Nordic countries had comparatively high spending on active labour market policies, with an average spend of 1.06% of GDP (excluding Norway) compared to the OECD average of 0.45%. Employment rates of Nordic countries significantly outperform the OECD average and have done so consistently (OECD, 2025[6]). In recent years, and following the COVID-19 pandemic, they reinforced further the attention on active labour market policies. In particular, governments worked more with employers, increased online outreach efforts and implemented online job-matching services. Governments created partnerships with schools and NGOs to increase outreach with young people, aiming to establish contact with those in need to make them aware of the resources available. This can be beneficial, as such partners often have strong understanding of labour market trends.
Many Nordic countries combined this with skill assessment and anticipation exercises on current and future labour market skill demand. Such approaches can be useful for reducing unemployment in the longer term, by ensuring that job matches are suitable for both employers and employees (OECD, 2022[7]; OECD, 2023[8]). Other countries have adopted similarly proactive outreach approaches. For example, Bulgaria use mobile labour offices and workplaces to reach people in small or remote settlements, and also deploy family labour consultants who support whole households, allowing them to reach people even when they are not registered as job searchers (OECD, 2022[9]).
Efforts to address long-term unemployment must address the motivation of those who have been out of the workforce for a long time. For example, Spain’s launching pad programme brought together those receiving unemployment benefits to engage Iin a variety of group activities to build up fundamental skills and job-related competencies, as well as interact with potential employers. This approach both improved participants’ skill levels, and built up their confidence in professional contexts, helping increase motivation (OECD, 2024[10]).
Active labour market policies can also be useful in improving the number of people with disabilities in work. For example, in 2016, Estonia implemented the “Work Ability Reform". With the aim of improving the labour market situation of those with reduced work capacity. The Public Employment Services had dedicated counsellors trained to provide services for returning those with disabilities to work. The use of these services was made a pre-condition for continuing to receive disability benefits (OECD, 2022[9]; OECD, 2022[11]).
3.2.2. Freeze or reduce the level of benefits
Freezing or reducing benefit levels was also a common savings approach reported for both unemployment and disability programmes. In fact, inadequate levels of unemployment benefits can make job searches more difficult or provide less time for the recipient to find appropriate, stable work. However, overly generous benefits can act as a disincentive to find work, lengthening unemployment time and resulting in unnecessary spending. For example, a 2026 meta-analysis of 57 studies looking at how unemployment insurance affects unemployment duration (Cohen and Ganong, 2026[12]) finds that more generous benefits prolong unemployment, although the impact is moderate, and are likely to be more significant in countries with already high benefit levels. In the case of disability, the benefits are often more generous than unemployment benefits. Furthermore, there is a well-established link between disability benefit generosity and employment levels, particularly when it is difficult or impossible to combine benefits with employment (MacDonald, Prinz and Immervoll, 2020[13]). For example, countries with higher employment among disability benefit recipients (such as Estonia and Slovenia) tend to have gradual benefit phase-outs rather than cut-offs and allow for combining work and benefits. Some countries, such as Austria, provide stronger incentives to work at low earnings, benefitting low-income claimants. This is because of the use of a clawback mechanism that reduces disability pension payments differently depending on the total income, with reduction rates ranging between 30% and 50% depending on income level (MacDonald, Prinz and Immervoll, 2020[13]; OECD, 2022[11]).
In terms of RPF Survey results for unemployment benefits, several respondents reported reductions in benefits structured to make work more appealing. In some cases, these take the form of benefits that diminish as the period of unemployment goes on. For example:
Belgium will modify its benefit system so that instead of recipients receiving a constant level of benefit, they will receive a decreasing amount of benefit the longer they remain on it.
A similar system exists within Czechia’s new super allowance, where benefits start above 50% but then drop down to below 50% beyond the third month.
A consolidation package introduced in the Slovak Republic scales unemployment benefit based on number of months of receipt. Instead of staying at 50% constantly, it will go down to 40% in the fourth month, 30% in the fifth and 20% in the sixth.
Several respondents reported limiting the maximum amount that can be claimed in disability benefits:
Estonia has placed a limit on the amount someone can receive when claiming incapacity for work. For 2026, this has been set at EUR 126.87 per day.
Finland has reduced its sickness benefit income replacement rate at the upper income level from 20% to 15%. It has also reduced the income threshold at which his level kicks in.
Sweden has implemented a benefit cap, which includes a modernised national norm in social assistance that more accurately correspond with consumption patterns and a limitation of social assistance for households with more than three children.
In the United Kingdom, the government’s 2025 Universal Credit Act will reduce the level of a key means-tested disability payment from the equivalent of EUR 112 a week to the equivalent of EUR 58 a week for new claimants.
In any case, the calibration of unemployment benefits is a complex task, as several design components need to be considered (see Box 3.2).
The RPF Survey also provides examples of savings in disability benefits, often in the form of reductions in benefit growth:
In Belgium, the so-called prosperity envelope, which reevaluates certain benefits, including those related to sickness and disability, to preserve purchasing power relative to wages, will not be applied in 2026.
In the Netherlands, large employers will no longer receive compensation for paying a transition payment to employees dismissed due to long-term sickness or occupational disability, which should incentivise them to keep these employees on where possible.
While the “right” amount of unemployment benefits will vary from country to country, some design components can be of use more consistently. One approach is to link benefits with job search or training requirements. This links to the concept of “flexisecurity”, first conceptualised in Denmark (Aghion, 2025[14]). This sees employers having greater freedom to hire and fire, while in return the government provides employees with additional unemployment protection, tied to them as individuals rather than their jobs. However, these benefits are not aimed to replace lost income, but to support individuals in returning to employment as quickly as possible and therefore should be accompanied by training programmes and incentives to return to work.
One such incentive can be the use of in-work benefits for those transitioning into work from unemployment, as they reduce participation tax rates by smoothing the withdrawal of unemployment benefits. While in-work benefits are an extra cost to government, they can result in net savings if they lead to a significant number of people moving off unemployment insurance (OECD, 2018[4]). Another approach is to reward contributions to the unemployment system by giving those with longer contribution periods more unemployment insurance, creating an incentive to build this up in employment. Finally, governments can save by linking benefit levels and durations to macroeconomic conditions, so that in periods of strong economic growth where jobs are more easily available, benefit levels are lower (OECD, 2014[15]; OECD, 2023[16]; Institute for Fiscal Studies, 2025[17]).
3.2.3. Consolidate programmes and overlapping benefits
Countries can achieve savings through consolidating programmes and streamlining overlapping benefits. In fact, fragmentation of benefit programmes can lead to inefficient and uncoordinated spending, as the generosity and eligibility strictness of some government programmes relative to others can increase or decrease uptake. For example, applications to disability programmes tend to increase during economic downturns. This is likely because, as wages fall and job availability declines, disability benefits become more attractive relative to work. If disability benefits are more generous and less conditional than unemployment benefits, then during downturns people may chose disability benefits. This shift towards disability benefits can also occur due to policy changes even outside an economic recession. Disability benefits cannot be analysed in isolation from other programmes. If these programmes are managed by the same organisation, or different organisations with data sharing frameworks in place, it can allow for more thorough understanding different claims interact.
Other benefits to consider as part of the co-ordination are social assistance programmes, including minimum-income benefits, which are means-tested safety nets. They are not linked to past employment records or contribution histories and are intended for households with no other income source. Research in the Netherlands found that, following a tightening of eligibility for disability benefits, one-third of people no longer eligible transitioned to social assistance rather than to work (Immervoll and Pasteiner, 2025[18]).
Within the RPF Survey, several respondents have reported such consolidation measures. For unemployment benefits:
Denmark aims to simplify the support system for unemployed persons. It abolishes job centres, giving more power to municipalities and caseworkers, leading to a more direct approach. Furthermore, a new cash benefit system reduces overlapping benefits for unemployment by introducing three standard benefit levels. This reform is part of a broader reform of Denmark’s unemployment support system (see Box 3.3).
Norway plans to phase out a transition benefit provided to single parents. This is intended particularly to help women enter the workforce at greater rates.
Thailand has invested in data and digital systems to improve accuracy of assistance delivery and streamline procedures, including within the unemployment system. This includes integrating labour-related operations with relevant agencies to reduce duplication and ensure efficient use of public funds.
For disability benefits:
In Chile, a two-day unpaid waiting period for all medical leave is introduced and aligns remuneration levels between public and private workers.1
Estonia has discontinued payment of sickness benefits to recipients of unemployment insurance benefits, to prevent unnecessary overlap.
Iceland has introduced a new payment structure for disability and vocational rehabilitation benefits, which will involve an assessment of the individual’s ability to perform in the labour market. This involves the use of co-ordination teams across the country.
Mexico has merged several health and disability care programmes into a set of consolidated programmes. This aims to generate savings by eliminating administrative overlaps and intermediation costs.
Another interesting example of consolidation is “Czechia’s Super Allowance” which consolidates four benefits, as well as introduces stricter means testing criteria and is therefore presented in the following section, although it has multiple and cross cutting effects (Box 3.4).
In 2025, the government agreed on a comprehensive reform of the employment system. The aim is to make the individual citizen the focus, so that case handlers can help them find work more directly, which is expected to reduce expenditure by 2the equivalent of EUR 350 million in 2030 and EUR 360 million permanently. Key policies include:
Abolishing the job centre requirement, instead giving municipalities more power to work in accordance with local conditions. This also involves having one clear channel of access for businesses to engage with municipal employment services.
Reducing the number of unemployment categories (which receive different interventions) from 13 down to 5. This is intended to reduce the number of people switching between groups, which can be confusing for them and results in greater administrative burden.
Removing many procedural programmes (rehabilitation, personal job facilitators, reading/writing/math support) while allowing municipalities discretion where needed.
Reducing the number of mandatory meetings and activation offers for citizens closer to the labour market, and allowing more discretion on how to handle citizens further from the labour market.
Sources: (Ministry of Employment, 2025[19])
3.2.4. Introduce or apply stricter criteria for means-testing
Measures to introduce or apply stricter means-testing were another commonly reported reform:
Czechia’s new super allowance will introduce stricter means testing, including through assessments of assets as well as income, and eligibility based on an entire household’s situation, not just individual income.
In Austria, the possibility of receiving unemployment benefits while being marginally employed (i.e. earning EUR 550 a month or less) is being abolished. Some exceptions exist, including older workers and those with disabilities.
Latvia has raised the minimum income threshold to be eligible for the State Social Security disability benefit (for residents with disabilities who do not have access to pension at statutory retirement age) from 20% of median income to 22%.
In 2025, Czechia introduced the “super allowance” social benefit. This replaces four separate benefits – child allowance, housing allowance, housing top-up and the living allowance – for any new applicants, while for those already receiving benefits a transitional benefit applies where they will not receive the consolidation version until July 2026
The super allowance consists of four components:
Living allowance: this helps households whose income does not meet basic needs, meaning they are unable to earn a living. It helps cover daily essentials like food, clothing and hygiene. Households must have income lower than 1.43 times the household subsistence minimum, and the adult applicant must be working or actively looking for work.
Housing component: this is aimed at households with housing costs such as rent and energy bills. Housing costs must exceed 30% of household income, up to a certain threshold.
Child component: this is for households with dependent children. A household is eligible if household income does not exceed four times the household subsistence minimum. Adult members must work, and children must attend school regularly.
Work bonus component: This is for households whose members work, intended to be motivational. It is provided to people receiving one of the other three benefits who are working at least 30 hours a month, run a business or are registered as a job seeker.
The benefit is subject to a means test, which assesses not only household income but also savings and bank account balances, property ownership and vehicles owned. These eligibility conditions will be stricter than those for existing benefits Some vulnerable groups (seniors, people with disabilities, parents with small children) are exempted from certain conditions.
The aim is to decrease the administrative burden, as the application will be submitted for the entire household using only one form – a simpler system than current four separate benefit allowances. The benefit is also intended to help those who aim to earn money themselves rather than fully relying on government assistance.
Source: 2025 OECD Survey on Restoring Public Finances, Ministry of Finance, Czechia
3.2.5. Shorten benefit duration
Several respondents have shortened benefit duration:
In Belgium, there will be a limitation in the time allowed to be on unemployment benefits up to maximum period of two years (depending on career history).
In Iceland, the maximum duration for the payment of unemployment benefits will be reduced by twelve months, so that such benefits may be paid for a total of 18 months instead of 30 months.
In New Zealand, the duration of the Jobseeker Support grants was reduced from 52 weeks to 26 weeks, meaning recipients will need to reapply to continue receiving benefit payments.
3.2.6. Ensuring greater levels of compliance for disability, including reducing fraud and abuse
A large number of respondents report efforts to increase levels of compliance concerning disability benefits, including significant efforts to reduce fraud and abuse of disability benefits. For example:
Chile has reinforced the supervisory powers of the Preventive and Disability Medicine Commission, strengthening their ability to identify behaviour inconsistent with prescribed medical rest through requesting information from relevant agencies.
In Portugal, prior to 2024, the verification process for temporary incapacity for work was activated after an incapacity of over 30 days, but since 2024, it is activated for any incapacity of over three days. There is also a proposal to make it easier to fire people for fraudulent claims.
Slovenia has started withdrawing compensation in cases where the insured person was already medically unfit for work at the start of employment or activity. Furthermore, it has introduced a legal basis for determining the movement regime of the insured person during sick leave.
3.2.7. Increasing focus on labour force participation
In order to achieve savings, respondents are also increasing focus on labour force participation. The fiscal implications of population ageing are likely to be significant, both due to the increased pension cost of an ageing population and the reduced tax revenue from a shrinking labour force. Therefore, many countries are projected to see significant declines in their working-age population. As countries can no longer rely on continued demographic growth to support their labour market, they are increasing the focus on improving labour force participation. The section below focuses specifically on the RPF Survey results, while broader OECD work also discusses ethe value of specific efforts focusing on young people, women, and also maximising the effectiveness of such measures.
Most measures focus on activation of unemployment benefits as well as adjusting training requirements to ensure they are well directed and effective. For example:
In Austria, a reform requires longer employment periods for those with degrees than those without to be eligible for educational allowances, with the aim of focusing training resources on those who need it most. It also mandates labour-market counselling, increases attendance requirements, and increases required proof of labour-market relevance of courses.
Czechia now provides support for so-called chosen reskilling, which gives job seekers greater freedom to choose the profession in which they would like to retrain.
In Finland, the employment authority provides supplementary job counselling to jobseekers those receiving unemployment benefits to help them find opportunities and help them prepare their employment plan. They must apply for the jobs assigned to them by the employment authority.
New Zealand provides additional support to those receiving unemployment benefits, which will lower spending on JobSeeker Support and other benefits. This includes expanding community job coaching and increasing support to help jobseekers prepare for work.
Thailand now links benefit entitlements with registration and monitoring of unemployment status, with the aim of facilitating reintegration into the labour market. It combines this with promotion of workforce skill development, encouraging enterprises to invest in skills development.
Several measures also aim to smooth the transition from unemployment to work, to help ensure that work is appealing to those on benefits:
In the Netherlands, the amendment of the Participation Act2 contains over 20 measures to simplify and improve the law. The aim is to enable more people to find work more quickly. For example, a portion of earned income per month will now not be deducted from unemployment benefit, to incentivise people taking on even low-paid work.
Slovenia introduced incentives for the employment of older recipients of cash benefits. Those over 59 years, currently receiving unemployment cash benefit, who enter full-time employment with an employer who is not their last employer before unemployment, will be paid an 40% of the monthly net unemployment benefit for up to 12 months after starting a job.
The reform, which will be phased in from the beginning of 2026, contains more than 20 measures, with the aim, among others, of making work more attractive through:
Standardising the additional earnings threshold: Prior to the reform, different municipalities have different approaches to reducing social assistance alongside income, which can cause confusion and fluctuating benefits. The rules in this reform create a standard method to calculate how income reduces social assistance, making it more predictable.
Creating a buffer budget: This buffer budget means that short‑term fluctuations in earnings (e.g. variable hours) don’t immediately reduce benefits. This gives beneficiaries more certainty about their total monthly income when they work.
Increasing additional earnings limit: From 2027 (and from 2026 for people under 27), 15% of income from work in a month will be ignored when calculating social assistance, while the remaining 85% is offset against benefits. This applies for 12 months per employment period.
Source: (Stimulansz, 2025[20])
Refocusing support to help raise youth employment can help boost labour force participation. According to the OECD’s Youth Database, about 12.8% of young people between 15 and 29 years are not in employment, education or training (see (OECD[21])), and in ten OECD countries it is above 15%. While such measures were not reported on in the Survey, several OECD examples exist. For example, a Hungarian 90-day job trial available to young adults not in education, employment or training (NEET), and which covered a portion of their wages, saw positive results after 12 months, with the effects stronger for lower-educated participants (OECD, 2024[3]). Attaching job-search requirements to unemployment benefits can also be beneficial, with research by the Institute of Labor Economics finding that on average, one additional required application reduces the duration of unemployment by 3% (Institute of Labor Economics, 2018[22]).
Refocusing support to raise women’s employment can also be important in boosting overall labour force participation. Women participate in the labour market less than men throughout the OECD, and often take on more unpaid work such as family and carrying responsibilities. Effective mobilisation of the female workforce will require efforts to increase gender equality both in the labour market and increase equality of unpaid work. This can include facilitating female entrepreneurship and promoting family-friendly policies within firms. Such options are discussed in greater depth in the OECD Employment Outlook (2025[1]).
Evaluating spending towards labour market support can help increase the cost effectiveness of spending in this area. Prior research supports the link between better employment outcomes and active labour market policies (Hur, 2019[23]). However, the success of active labour market policies varies greatly. Meta-analyses of 100 randomised control trials indicate that only around a third of programmes produce significant improvements in employment outcomes. However, these programmes have much in common, allowing patterns to be identified. The best outcomes generally come from carefully focused programmes that are of adequate duration and are aligned with business cycle conditions. Furthermore, they tend to involve strong employer engagement. While activation measures without job-matching support can lead to short-term unemployment gains, these gains are more likely to be unsustainable (Shaikh and Ozturk, 2025[24]).
3.2.8. Ensuring greater levels of compliance
There are examples of efforts to increase compliance with unemployment benefit requirements in the RPF Survey, with a view to bringing people back to work. These generally take the form of sanctions. For example:
Finland’s social insurance reform allows the social insurance institution to reduce the basic amount of social assistance by 50% if the benefit recipients does not register with the employment services as an active jobseeker. It has also increased the speed at which a work requirement kicks in if the jobseeker does not carry out certain tasks.
Ireland has increased the deduction for benefit recipient who fail to engage in the mandatory activation process for EUR 44 to EUR 90 a week from January 2025, to encourage jobseekers to co-operate with efforts to assist them in securing employment.
3.2.9. Intervening early with sickness and incapacity benefits
Besides measures to tighten eligibility and freeze levels of benefits, there can also be significant gains from measures promoting early intervention and prevention. In fact, many disability reforms are correlated with a reduction in the number of people receiving disability benefits, but are not correlated with higher employment rates. This may be because reforms come too late for moves into work to be possible. Returns to work become unlikely after even five or six months of absence, and in many countries, sickness insurance programmes can last several years. Even if a benefit recipient does eventually get a job, that job may end up being low-quality or unstable, leading to employment gains fading over time (OECD, 2022[11]). Interventions can therefore benefit from focusing on earlier stages. One possibility is to require employers to continue to pay wages when workers fall ill. For example, Dutch employers must continue paying at least 70% of their employee’s wages during the first two years. This provides incentives for employers to support return to work. In the Netherlands, the employer must also use a doctor’s assessment (mandatory in the sixth week of sick leave) to draw up a return-to-work plan. Stricter monitoring of sickness leave is also important. In the case of Sweden, research finds that postponing the requirement for a doctor’s certificate increases the length of sickness absences, and that stricter monitoring increases the exit rate from sickness insurance (OECD, 2022[11]).
However, in the context of the RPF Survey, such measures were clearly identified by only one respondent Belgium. In 2026, Belgium has introduced several reforms to incentivise people to go back to work once they have taken sick leave, including:
Limiting the duration of sickness certification to a maximum of three months. While extension is possible, this is only the case three months at a time.
Developing a database for electronic sickness certificates. This database will allow the government to identify doctors who issue more, or longer periods of, incapacity as compared to other doctors.
Aligning parts of the funding health insurance institutions receive with success in getting people back to work. Metrics will include the number of initiated permitted work resumptions and the number of referrals, among others.
Employers with at least 50 employees will have to pay a solidarity contribution in the period of primary incapacity as of 2026, with the aim of encouraging employers to reintegrate their sick employees. Employers will also be obliged to have an assessment done of the work potential of their employee by an external service after eight weeks of incapacity.
Employers with at least 20 employees will have to initiate a reintegration trajectory no later than six months after the start of the work incapacity, and work rules must include a mandatory procedure for maintaining contact with employees with work incapacity.
3.2.10. Other savings measures
Several measures were reported which do not fit into the above categories:
Australia plans to more consistently determine participant budgets for its disability insurance scheme based on needs and clarify information participants need to provide for plan reassessments, as well as what type support can be funded. This aims to improve budget planning based on need, and provide support and information to participants to reduce spending.
In Canada’s Budget 2025, the Disability Pension for the Royal Canadian Mounted Police, the national police service, will be indexed to the Consumer Price Index, replacing the previous approach which used the greater of either CPI or the average increase in wages of certain categories of public sector employees. These changes will ensure benefits are indexed in a transparent manner consistent with approaches used for other benefits such as the Canada Child Benefit, Old Age Security, federal government pension plans, and the Canada Pension Plan.
References
[14] Aghion, P. (2025), “Only Creative Destruction will Boost European Competitiveness”, Project Syndicate, https://www.project-syndicate.org/commentary/innovation-policy-how-europe-can-return-to-technological-frontier-by-philippe-aghion-2025-12.
[12] Cohen, J. and P. Ganong (2026), “Disemployment Effects of Unemployment Insurance: A Meta-analysis”, American Economic Review: Insights, Vol. 8/1, pp. 1-18, https://doi.org/10.1257/aeri.20240441.
[23] Hur, H. (2019), “Government expenditure on labour market policies in OECD countries: responding to the economic crisis”, Policy Studies, Vol. 40/6, pp. 585-608, https://doi.org/10.1080/01442872.2018.1533113.
[18] Immervoll, H. and F. Pasteiner (2025), “Weathering the storms? Minimum-income benefits as a crisis response”, OECD Social, Employment and Migration Working Papers, No. 323, OECD Publishing, Paris, https://doi.org/10.1787/d45ca27f-en.
[17] Institute for Fiscal Studies (2025), Options for unemployment insurance, https://ifs.org.uk/sites/default/files/2025-10/GB%20Options%20for%20Unemployment%20Insurance%20-%20PDF.pdf.
[22] Institute of Labor Economics (2018), Job Search Requirements, Effort Provision and Labor Market Outcomes, https://docs.iza.org/dp11765.pdf.
[5] Knapińska, M. and K. Woźniak-Jasińska (2024), “Labour market policy in OECD countries – selected aspects”, Argumenta Oeconomica, Vol. 2024/2(52), pp. 186-199, https://doi.org/10.15611/aoe.2024.2.13.
[2] Koutsogeorgopoulou, V. and H. Morgavi (2025), “Ageing populations, their fiscal implications and policy responses”, OECD Economics Department Working Papers, No. 1844, OECD Publishing, Paris, https://doi.org/10.1787/6aec03b3-en.
[13] MacDonald, D., C. Prinz and H. Immervoll (2020), “Can disability benefits promote (re)employment? : Considerations for effective disability benefit design”, OECD Social, Employment and Migration Working Papers, No. 253, OECD Publishing, Paris, https://doi.org/10.1787/227e7990-en.
[19] Ministry of Employment (2025), Aftale om reform af beskæftigelsesindsatsen [Agreement on the reform of the employment effort], https://bm.dk/media/ymybm2xe/aftale-om-reform-af-beskaeftigelsesindsatsen.pdf.
[1] OECD (2025), OECD Employment Outlook 2025: Can We Get Through the Demographic Crunch?, OECD Publishing, Paris, https://doi.org/10.1787/194a947b-en.
[6] OECD (2025), The Role of Public Employment Services in Promoting an Inclusive Nordic Common Labour Market, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/54cec4d6-en.
[3] OECD (2024), OECD Youth Policy Toolkit, OECD Publishing, Paris, https://doi.org/10.1787/74b6f8f3-en.
[10] OECD (2024), Strengthening Active Labour Market Policies in Korea, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/44cb97d7-en.
[16] OECD (2023), “Income support for jobseekers: Trade-offs and current reforms”, https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/02/income-support-for-jobseekers_a99cb15b/405ed506-en.pdf.
[8] OECD (2023), Nordic Lessons for an Inclusive Recovery? Responses to the Impact of COVID-19 on the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/2aa7bcc1-en.
[11] OECD (2022), Disability, Work and Inclusion: Mainstreaming in All Policies and Practices, OECD Publishing, Paris, https://doi.org/10.1787/1eaa5e9c-en.
[7] OECD (2022), OECD Employment Outlook 2022: Building Back More Inclusive Labour Markets, OECD Publishing, Paris, https://doi.org/10.1787/1bb305a6-en.
[9] OECD (2022), Reaching Out and Activating Inactive and Unemployed Persons in Bulgaria, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/7b91154a-en.
[4] OECD (2018), Good Jobs for All in a Changing World of Work: The OECD Jobs Strategy, OECD Publishing, Paris, https://doi.org/10.1787/9789264308817-en.
[15] OECD (2014), Society at a Glance 2014: OECD Social Indicators, OECD Publishing, Paris, https://doi.org/10.1787/soc_glance-2014-en.
[21] OECD (n.d.), Youth employment and social policies, https://www.oecd.org/en/topics/youth-employment-and-social-policies.html.
[24] Shaikh, O. and S. Ozturk (2025), “Assessing the impact of labour market spendings on unemployment dynamics across demographics in OECD countries”, Public Sector Economics, Vol. 49/4, pp. 527-562, https://doi.org/10.3326/pse.49.4.2.
[20] Stimulansz (2025), “Handreiking Participatiewet in balans [Participation Law Guidance]”, https://www.kennisbundel.nl/wp-content/uploads/2025/12/Handreiking-Participatiewet-in-balans-Stimulansz-17-november-2025.pdf.
← 1. A 3-day waiting period already existed for private sector workers on short term leave (up to 10 days). The reform reduces this waiting period to two days and extends it universally to all workers regardless of leave duration, aligning the treatment between public and private workers.
← 2. “Participatiewet in Balans”