This chapter focuses on the implementation of ILAs. It begins by examining governance and management structures, along with technical aspects such as user-friendly digital platforms and outreach strategies to effectively reach the target population. It then explores mechanisms to ensure high-quality training provision and discusses the broader enabling frameworks, including policies on training leave, career guidance, and the recognition of prior learning. The chapter concludes with key considerations for ongoing monitoring and evaluation, which are essential for assessing the impact of ILAs and informing their continuous improvement.
Advancing Adult Skills through Individual Learning Accounts
A Step-by-Step Guide for Policymakers
Abstract
The governance of ILAs is typically complex, involving multiple governmental bodies in both planning and implementation (Table 3.1). In the countries covered in this report, ILA schemes are supported by a legal framework and initiated by ministries responsible for labour, employment, or education. Public agencies collaborate with these ministries and play a key role in implementing ILAs.
Effective governance is crucial throughout the design and implementation stages. Engaging all relevant stakeholders from the outset – including in discussions on objectives, target groups, and eligible training programmes – is vital. During implementation, it is equally important that each stakeholder has clearly defined roles and responsibilities to ensure accountability and enable effective co‑ordination.
Note: *Education programmes must be aligned with the CROQF Register to be eligible for funding. While the Ministry of Science, Education and Youth is not directly involved in the voucher pilot, it oversees the register. Several institutions are involved in the development of qualification and occupational standards, including the Agency for Science and Higher Education, the Agency for Vocational Education and Training and Adult Education, and the Ministry of Labour and Pension System, Family and Social Policy.
Source: Information collected by the OECD project team.
Key actors and governance structures in ILA schemes
The design and implementation of ILA initiatives typically involve the competencies of multiple ministries – most commonly those responsible for employment, labour, and education, but also finance. While governance is generally led by a single ministry, it is carried out in consultation with others. Public agencies are usually responsible for the operational implementation of the schemes.
In Croatia, Czechia, France, and the Netherlands, the process of establishing ILAs was initiated by ministries responsible for labour and employment. In Croatia and Czechia, public employment services oversee ILA implementation, whereas in France and the Netherlands, implementation is divided between multiple agencies.
In France, the CPF scheme is managed by a public financial institution, the Caisse des Dépôts et Consignations, while France Compétences, the national authority for vocational training and apprenticeships, also plays an important role. The Caisse des Dépôts et Consignations administers the Mon Compte Formation digital platform, which hosts learner accounts, maintains a registry of training providers, publishes training offers, and supports platform development. France Compétences pools the Professional Training Funds and allocates approximately EUR 2 billion annually to the Caisse des Dépôts et Consignations to manage CPF funding. Additionally, France Compétences oversees the Répertoire RNCP and the RS, setting quality standards for training providers within the CPF framework.
In the Netherlands, the public employment service (Uitvoeringsinstituut Werknemersverzekeringen) managed the STAP portal, through which learners applied for courses, training providers reported participation data, and e‑invoices were submitted to the government. The Education Executive Agency (Dienst Uitvoering Onderwijs) was responsible for managing the STAP training register, which was connected to the portal.
In Singapore and Lithuania, ILA schemes were introduced under the mandate of ministries responsible for education. In Singapore, implementation is carried out by SkillsFuture Singapore, a statutory board under the Ministry of Education, which was established to drive and co‑ordinate the implementation of the national SkillsFuture movement and promote a culture and holistic system of lifelong learning. In Lithuania, where the ILA is funded through the EU’s temporary funding instrument, the Recovery and Resilience Facility, implementation is managed by the European Social Fund Agency. Programme evaluation for non-formal education and training providers is conducted by the Qualifications and Vocational Education and Training Development Centre (KPMPC).
In addition, labour market and social partners are key stakeholders in ILA initiatives (Sung, 2025[24]), engaging in various aspects of their design and implementation. Employers may play a crucial role in financing ILAs – through employer levies and other mechanisms – while also shaping the content of ILA-supported training by helping define skills needs. In France, for instance, the CPF is funded by employer levies, with employers also able to supplement CPF funds for specific training programmes through collective agreements negotiated via Skills Operators (opérateurs de compétences, OPCO). The 11 OPCOs, representing different sectors, previously managed CPF funds before this responsibility was transferred to the Caisse des Dépôts et Consignations in 2019. They also contribute to the development of professional certification standards under the RNCP, which determines CPF programme eligibility.
While involving employers in determining which training is eligible for funding can help align ILAs with labour market needs, it is important to ensure that this involvement does not compromise training quality or cater solely to the narrow interests of specific employers. For example, in the Netherlands, eligibility of training to be funded by the STAP budget could be demonstrated through recognition by industry and professional associations. However, this approach raised concerns about accountability – particularly in cases where training was later found not to meet regulatory and quality standards.
Beyond financing and programme design, labour market and social partners also play an advisory role in shaping ILA schemes. In Czechia, for example, the tripartite Upskilling and Reskilling Committee, established in 2021, brings together government ministries, employer organisations, and social partners at the national level. This platform facilitates expert discussions on the systemic development of retraining and further education in the country, including the ILA pilot. In Lithuania, a similar role is played by the Non-formal Adult Education Council, an expert advisory group that includes representatives of social partners and municipalities.
Legal frameworks supporting the implementation of ILAs
The introduction of ILA initiatives is typically supported by legal frameworks, even in the case of pilot programmes. Countries use a variety of legal instruments – ranging from laws to decrees, ministerial orders, and other regulatory measures – to establish and govern these schemes. In some cases, these instruments provide detailed provisions for scheme implementation. In others, they offer only the minimum legal foundation necessary to enable operations.
In France, the Labour Code provides a detailed legal basis for the CPF, outlining its objectives, eligibility criteria, scope of training, and co-financing mechanisms involving individuals, employers, and other stakeholders. The CPF and its subsequent reforms have been introduced and updated through dedicated legislation. Key laws include:
Law No. 2014‑288 on Vocational Training, Employment, and Social Democracy (La loi du 5 mars 2014 relative à la formation professionnelle, à l’emploi et à la démocratie sociale), which established the CPF and integrated its main features into the Labour Code.
Law No. 2018‑771 on the Freedom to Choose One’s Professional Future (LOI n° 2018‑771 du 5 septembre 2018 pour la liberté de choisir son avenir professionnel), which introduced major reforms to the CPF, including converting entitlements from hours to Euros, launching the MonCompteFormation platform, establishing France Compétences, and introducing the Qualiopi certification system (Perez and Vourc’h, 2020[25]).
In Lithuania, the Law on Non-Formal Adult Education and Continuing Education establishes the legal foundation for the ILA as “a single window” for accessing non-formal adult education and continuing education programmes. Implementation details are set out in a series of ministerial orders linked to this primary law. For example, Order No. V‑599 outlines the development of the KURSUOK platform and defines key features such as financing. Order No. V‑1 218 sets the quality assurance framework, including quality criteria and the learner feedback questionnaire. Order No. V‑1 515 specifies further details, including a model contract between training providers and learners that specifies the obligations of each party.
In the Netherlands, the details of the STAP voucher scheme were defined in the Regulation on the STAP Budget Subsidy Scheme. While this regulation was linked to the Social Affairs and Employment Subsidies Act, it did not involve any amendments to the law itself.
In Czechia, by contrast, the legal basis for the ILA pilot is minimal. The Employment Act allows for the piloting of the ILA scheme within the framework of active labour market policies.
Similarly, in Croatia, while the ILA pilot is linked to several legal acts, none explicitly references it. The Adult Education Act – covering primary and secondary education as well as non-formal and informal learning – and the Higher Education Act – which enables higher education institutions to offer lifelong learning programmes – form the broader legal context in which the pilot operates. Additionally, the Labour Market Act defines education and training governed by both laws as education for employment, further reinforcing the legal foundation for the pilot.
ILAs operate at the intersection of education and labour market policies and involve a broad range of stakeholders. To ensure that ILA initiatives are relevant, effective, and widely supported, it is critical to engage all relevant stakeholders from the earliest stages of development. This includes fostering a sense of shared ownership by bringing together all relevant ministries and public bodies, and establishing mechanisms for co‑ordinated dialogue with employers, training providers, and learners. During implementation, clearly defined roles and responsibilities are essential to promote accountability and ensure effective co‑ordination.
ILAs are underpinned by a legal framework. Countries typically follow one of two approaches: either establishing detailed provisions in legislation that prescribe the design and operation of the ILA scheme, or providing a broader legal mandate that empowers public authorities to manage the initiative or authorise providers to deliver adult learning programmes eligible for ILA funding.
While robust governance structures lay the groundwork for effective ILA schemes, their success ultimately hinges on active learner participation. Digital platforms play a pivotal role in enabling access to adult learning opportunities, acting as the main gateway for individuals to explore, select, and manage their learning paths. To ensure that all individuals – particularly those traditionally underrepresented in adult learning – can fully benefit, it is essential to pair digital accessibility with proactive outreach and tailored offline support. This inclusive approach helps bridge digital divides and ensures equitable access to learning opportunities.
Digital platforms as the backbone of ILA Implementation
ILA digital platforms serve a dual purpose. First, they act as aggregators of training opportunities, bringing together a wide range of learning offers that are accessible to all individuals – regardless of whether they receive funding. In this way, they serve as a one‑stop shop for adult learning, supporting anyone seeking to upskill or reskill. Second, ILA platforms function as a gateway to funding, facilitating the purchase of training while reducing the administrative costs associated with processing applications.
All six reviewed countries have an online portal dedicated to administering ILAs. In some cases, a single platform functions as a one‑stop shop for both learners and providers, while in others, separate portals exist for each group, connected on the back end (Table 3.2).
In France, Mon Compte Formation operates as a one‑stop shop for both learners and providers, with providers accessing a dedicated space called Espace des Organismes de Formation. Similarly, Croatia’s Vaučeri, Czechia’s Jsem v Kurzu and Lithuania’s KURSUOK serve as unified platforms.
In Singapore, MySkillsFuture serves as the interface for learners, while Training Partners Gateway serves as the one‑stop portal for training providers to access government grants to support continuing education and training. Employers access relevant information through the SkillsFuture for Business site, hosted on a separate platform, GoBusiness Singapore. In the Netherlands, the training registry acts as an interface for providers, while users access a separate platform that extracts course information from the registry.
ILA platforms typically display information about training courses, including course descriptions, eligibility requirements, duration, costs, mode of delivery, language of instruction, and contact details of training providers. They may also display the schedule of upcoming sessions and often offer the option for individuals to book or register for training directly through the platform.
Source: Information collected by the OECD project team.
In all six countries covered in this report, individuals access ILA platforms using an ID authentication system. For example, in Lithuania, individuals log into the platform through the e‑Government Gateway (e‑Valdžios). Once their identity is verified, the system automatically determines eligibility for funding and whether budget funds are still available for that year. This platform also displays employment data and participation records from various education registers. In Singapore, learners and training providers log in using SingPass and CorpPass, respectively. SingPass is a national digital identity for citizens and residents. Training providers are required to use CorpPass to transact with SkillsFuture Singapore.
The procedures for using ILA credits vary across countries, depending on whether public authority approval is required. In France, ILA entitlements are an individual right, and credit holders can use their funds without requiring approval from public authorities. In Lithuania, applications submitted through the KURSUOK portal during fixed application periods undergo automatic eligibility checks. Once approved, individuals have full autonomy to choose from eligible training courses. In Croatia, Czechia, and the Netherlands, learners must follow an application process requiring formal approval. However, this process is largely administrative, with most applications being accepted. In Czechia, individuals need to register with the Labour Office online to apply for ILA funding. The application is subsequently assessed by the Labour Office, which may exceptionally request additional documents before approving the application. In Croatia, learners must register with the Employment Service, obtain a quote from education providers, and then apply through the Vaučeri portal. Only 5% of applications were rejected. In the Netherlands, learners must first register with a provider before applying for funding through the STAP portal during set application periods.
After participating in the training, the ILA platform can allow individuals to consult their attendance history and download certificates, as is the case in Singapore and Lithuania.
While online platforms facilitate access to ILAs and reduce administrative costs, they can pose challenges for individuals with limited digital skills. To ensure inclusivity and equitable access, alternative support channels may be provided. For example, phone helplines are available in Singapore and Lithuania to assist users in navigating the system, while in Czechia, individuals can receive in-person support at the offices of the public employment service office.
Media campaigns and outreach efforts to promote inclusive participation
ILA schemes can only have an impact if they are widely known and actively used by their target population. Beyond establishing digital portals, public authorities and training providers implement a range of media campaigns and outreach efforts to raise awareness and encourage participation.
The financial incentive embedded in ILA schemes naturally drives training providers to actively promote them, as increased awareness leads to higher enrolment in their courses. The evaluation of the STAP scheme in the Netherlands highlighted the critical role that training providers played in promoting the scheme through online platforms, social media, and traditional media channels (Buiskool et al., 2024[12]). In Lithuania, providers also served as key communication channels by advertising their KURSUOK-eligible courses and promoting the platform through their websites and Facebook accounts.
To further enhance the uptake of ILA schemes, public authorities often conduct their own communication campaigns. In Singapore, SkillsFuture Singapore has consistently promoted lifelong learning and its schemes and services through targeted campaigns. In 2024, a promotional video depicted a father struggling to gain recognition at work, who, inspired by a conversation with his daughter, pursued game development training with support from SkillsFuture Singapore. The campaign ran for two months across television, digital platforms, social media, and out-of-home channels (Campaign, 2024[26]). Additionally, Singapore introduces students to the MySkillsFuture student portal as part of their school career guidance curriculum, while adults receive information about SkillsFuture through the MySkillsFuture workforce portal and career guidance centres. SkillsFuture festival is also organised annually to raise awareness of opportunities in growth sectors and to promote learning offers. The 2024 festival focused on industries that are expected to grow in the future – healthcare, digital and green (SkillsFuture Festival, 2024[27]).
Media campaigns are typically most intense at the launch of an ILA scheme. In Lithuania, the introduction of the system was accompanied by press conferences, press releases, and advertising across television, radio, online news portals, and newly created KURSUOK social media accounts on Facebook and LinkedIn. Similarly, in Czechia, the launch of the ILA scheme was promoted through a media campaign featuring radio spots, 15‑ and 6‑second online video ads, banners, social media advertisements, and print media placements.
The evaluation of the STAP scheme in the Netherlands revealed that the Internet and social media were the most effective promotional channels, with 51% of STAP participants learning about the scheme through these platforms. Television, radio, newspapers, and posters were the second most effective outreach method, reaching 25% of participants, followed by posters at public employment services (20%) and the ILA platform itself (18%). Word-of-mouth also played a significant role, as 14% of participants discovered the scheme through family members and friends (Buiskool et al., 2024[12]).
While many ILA media campaigns target the general population, some are specifically designed for underrepresented groups. In Czechia, promotional efforts focused on seniors active in the labour market, parents on parental leave, students, and the self-employed. Targeted outreach is particularly important for groups less likely to engage with ILA on their own initiative, such as individuals with lower educational attainment and older adults (Box 3.1). Since the composition of underrepresented groups varies across countries and depends on the specific design of ILA schemes, continuous monitoring of participant demographics (see Decision 8) is essential to ensure that media campaigns are effectively tailored to reach those most in need.
A digital platform, which is a key feature of ILAs, can serve as a one‑stop shop for learners, facilitating access to a catalogue of available courses and associated funding. It can also significantly reduce the administrative costs related to processing applications.
Authentication systems used to log in to public services can be leveraged to verify individuals’ identities and ensure secure access to ILAs. In well-designed platforms, once identity is verified, eligibility and the amount of funding are automatically determined. If funding is targeted at individuals with specific characteristics – such as low income or lower educational attainment – administrative data can be used to verify eligibility. Once eligibility is confirmed, individuals typically have the autonomy to select from a catalogue of approved training courses without needing further approval.
ILAs create a direct financial incentive for training providers to actively promote the scheme, as it drives more individuals to enrol in their courses. However, additional communication campaigns remain essential, particularly for reaching groups with lower take‑up rates. The experience of the Netherlands shows that the internet and social media – followed by traditional media channels such as television, radio, and newspapers – are the most effective tools for raising awareness about ILA schemes.
Women tend to be overrepresented among the users of ILA, with 85% of beneficiaries being female in Lithuania, 65% in Czechia and the Netherlands and 49% in France. This pattern is particularly striking for Czechia, where men are generally more likely to participate in training than women (Eurostat, 2024[28]). Age also plays a role in the likelihood of using ILA. Prime‑age workers (30‑49 year‑olds) were the primary beneficiaries of ILAs and were overrepresented among its users compared to the general population, while those above the age of 50 and below the age of 30 were underrepresented. In Czechia and the Netherlands, those with a higher level of education were overrepresented. However, the gap between the ILA participation of those with lower educational attainment and the share of the population they accounted for differed across countries, and it was smaller in France, which suggests that the design features of ILA can influence the participation of those with lower education. The regional composition of ILA participants in the Netherlands and Czechia was proportional to the size of the regions.
Note: The benchmark for age composition is an approximation – for example, in the Netherlands, ILA participation of those aged 55‑67 is compared to the share of the population aged 55‑64 due to data limitations.
Source: Authors’ calculations based on data shared by countries included in the review and Eurostat (2025[29]; 2025[30]).
Ensuring the quality of learning opportunities is essential to the success of ILA schemes. As ILAs grant individuals considerable autonomy in selecting their learning pathways – typically without requiring input from career counsellors or employers – it is crucial that the training options available through ILA platforms are both high quality and relevant. This helps to ensure that learning leads to meaningful labour market and educational outcomes, while also safeguarding public resources against misuse or fraud.
To this end, countries employ various mechanisms to ensure the quality of ILA-supported training, both before and after programmes are launched. Ex-ante quality assurance systems differ in several respects, including whether they are part of a broader quality assurance framework that extends beyond the ILA system, and at what level the quality assurance is conducted (Table 3.3).
Note: *Within the STAP scheme, quality assurance was maintained through several mechanisms. For providers recognised by the Ministry of Education, Culture and Science (such as primary and secondary schools and higher education institutions) quality assurance was ensured through registration in the Registry of Institutions and Training Courses. For other providers, different mechanisms applied, including obtaining the NRTO Quality Mark or having their courses classified against the Netherlands Qualifications Framework by the National Co‑ordination Point.
**In Lithuania, the programme evaluation by the KPMPC is the main mechanism for ensuring the quality of ILA-supported training. However, before undergoing this evaluation, providers must complete administrative registration processes at both the provider and programme levels.
***In France, higher education institutions are exempt from obtaining the Qualiopi certification. Similarly, in Lithuania, formal education providers (such as vocational education and training schools and higher education institutions) are exempt from the programme evaluation conducted by the KPMPC.
Source: Information collected by the OECD project team.
A general trend is to integrate ILA schemes into a broader quality assurance framework – as seen in Croatia, France, the Netherlands and Singapore. However, some countries, such as Czechia and Lithuania, have developed quality assurance mechanisms specifically tailored to ILA-supported training. This approach may be appropriate when existing systems are not well aligned with the specific objectives of ILAs – for example, when quality assurance mechanisms are fragmented and non-formal providers operate under limited oversight.
In countries with more mature systems – such as France and Singapore, where ILAs have been in place for over a decade – quality assurance mechanisms are applied at both the provider and programme levels. In contrast, other countries typically focus on quality assurance at the programme level, although some formal providers may be subject to provider-level quality assurance outside the ILA framework.
In France, quality assurance under the CPF scheme draws on broader national frameworks, notably Qualiopi, the RNCP, and the RS. Quality assurance is primarily conducted at the provider level, but a mechanism also exists to regulate content at the programme level.
At the provider level, the Qualiopi certification serves as the main quality assurance mechanism. It is a national quality certification system for providers seeking public funding, including CPF funding. Since 2022, Qualiopi certification has been mandatory for all publicly funded providers offering training, skills assessments, validation of acquired experience, and apprenticeships – excluding higher education institutions, which are exempt. Certification is granted by accredited or authorised bodies under the French Accreditation Committee (Cofrac) and is based on seven quality criteria, comprising 32 indicators. These cover: transparent public information, clear service objectives tailored to target audiences, appropriate service delivery and support methods, adequate educational and technical resources, qualified personnel, provider engagement in the professional environment, and responsiveness to stakeholder feedback. To maintain certification, providers must undergo a renewal process every three years.
At the programme level, CPF-supported training must be linked to a qualification registered in either the RNCP or RS (see Box 2.3). Training providers must obtain authorisation from the certification holder of the respective RNCP or RS qualification before applying for CPF eligibility. For example, the Ministry of Agriculture and Food Sovereignty holds the certification for the “Higher Agricultural Technician Certificate – Quality, Food, Innovation and Health Control” and authorises 32 organisations to offer training leading to this qualification as of April 2025. Both registries are managed by France Compétences, and each certification entry is valid for up to five years.
Building on these frameworks, the Caisse des Dépôts et Consignations assesses whether candidate providers meet the eligibility requirements, including possessing Qualiopi certification and authorisation to offer training leading to RNCP- or RS-recognised qualifications. Approved providers are then registered in the Training Organisations Space (Espace des Organismes de Formation, EDOF) under Mon Compte Formation. As of 2023, over 12 000 applications for registration have been submitted, of which approximately 10 000 were rejected and 1 700 approved. A total of around 5 400 different training organisations have applied. Applications are rejected for various reasons, with 71% failing to meet the eligibility criteria (Caisse des Dépôts, 2024[9]).
Singapore’s SFC initiative builds on established quality management frameworks – at both the provider and programme levels – that existed prior to its launch. All training providers must complete an organisation registration process to demonstrate financial stability and robust corporate governance. This includes an on-site visit to assess the provider’s organisational structure, systems and processes, as well as facilities and equipment supporting training delivery. At the programme level, courses are evaluated against a set of criteria that includes requirements for adult educators, relevance of the skills covered, quality of the learning design (including adherence to andragogical and assessment principles), and the presence of clearly defined learning outcomes.
The Dutch STAP scheme also leveraged existing quality assurance frameworks by offering multiple eligibility pathways. Providers recognised by the Ministry of Education, Culture and Science – including primary and secondary schools and higher education institutions – qualified through course registration in the Registry of Institutions and Training Courses, which is used beyond the STAP scheme. Registration procedures vary by education level. Other providers could qualify for STAP funding by meeting alternative quality standards, such as holding a Quality Mark from the Dutch Council for Training and Education (NRTO) or having their courses classified in line with the Netherlands Qualifications Framework by the National Co‑ordination Point. These quality designations typically remain valid for four to six years. Providers meeting one or more of the approved standards could then apply to the Education Executive Agency to have their courses listed on the STAP portal.
Croatia similarly integrates its ILA initiative within a broader qualifications framework. Training programmes eligible for voucher funding must align with competencies or learning outcomes listed in the CROQF Register (see Box 2.2). Once this alignment is established, providers submit programme applications for approval to the relevant authority: the Ministry of Science, Education and Youth for formal education providers, and the Ministry of Labour, Pension System, Family and Social Policy for non-formal providers.
In contrast, Lithuania adopts a mixed approach by establishing an ILA-specific quality assurance system – but only for non-formal education and training providers, who are generally subject to less stringent oversight. The KPMPC ensures that these providers meet minimum requirements (Box 3.2) and evaluates programmes against four quality criteria: 1) the clarity and coherence of programme objectives in relation to intended competencies; 2) the relevance and comprehensiveness of programme content; 3) the effectiveness of learning methods; and 4) the clarity and demonstrability of learning outcomes. Regulations stipulate that the entire evaluation process should not exceed four months, although in practice it is usually completed within one month. Formal providers such as vocational schools and higher education institutions are exempt from this process, as they are already subject to quality assurance mechanisms outside the ILA system.
Czechia’s ILA pilot operates under a distinct framework. The Ministry of Labour and Social Affairs sets minimum requirements for programmes (Box 3.2), which are verified prior to inclusion in the ILA training registry. These requirements apply to both formal and non-formal providers, although formal providers remain subject to their own external quality assurance systems.
In addition, ex-post quality evaluations are conducted, often through learner feedback, and in some cases, supplemented by audits and assessments. Learner feedback is a key measure of training quality and relevance, typically collected through structured questionnaires administered immediately after training, a few months later, or both. Participants rate programmes or specific aspects often on a five‑point scale, with results typically published on digital platforms to help prospective learners make informed decisions. In some systems, learner feedback plays a significant role in quality assurance. For example, in Singapore, low ratings or response rates can lead to course removal from the MySkillsFuture portal (see Decision 8 for more on learner questionnaires).
Beyond learner feedback, audits and assessments help maintain training standards and drive continuous improvement. These are often triggered by poor learner evaluations. In Czechia, for example, the Labour Office conducts monitoring visits and public administration inspections of ILA-supported training courses when course ratings fall between one and two out of five. Monitoring visits assess aspects such as teaching quality, course content, textbooks, class size, attendance records, and exam participation. Authorities track the number of visits and ensure that corrective measures imposed on providers are implemented. Public administration inspections focus on administrative compliance, financial returns, tax reporting, and legal obligations.
In France, audits and assessments are a standard part of the quality assurance process, regardless of feedback results. In 2025, France strengthened its auditing framework, with the Caisse des Dépôts et Consignations evaluating providers selected in a randomised way. The first level of audit is conducted remotely, assessing providers against 13 criteria across three key areas: training quality (content, educational tools, and trainer expertise), organisational quality (logistics, human resources, and trainee support), and alignment with professional goals (relevance to career objectives and overall impact). Providers that do not meet the required quality threshold undergo a second-level audit, which includes an on-site visit, a broader review covering 17 criteria, and interviews with stakeholders, including learners. In 2025, the Caisse des Dépôts et Consignations plans to audit approximately 1 000 providers (Mon Compte Formation, 2025[31]).
Singapore employs a similar approach through the Training Provider Quality Assessment (TPQA), conducted by SkillsFuture Singapore. This evaluation examines administrative and governance processes, course quality assurance, and training outcomes to ensure effective content design, delivery standards, and measurable impact. It includes a one‑day on-site visit, and a poor TPQA grade can result in funding restrictions or loss of funding eligibility. There is also a reporting system for whistle‑blowers in place.
ILA-supported training is typically subject to minimum quality requirements, which are often followed by additional evaluation processes. Training providers are generally expected to have prior experience in delivering educational activities, and instructors are usually required to hold relevant qualifications. Additional conditions are also commonly applied to course content and delivery to ensure that ILA-supported training is fit for purpose. Table 3.4 presents the minimum requirements established specifically for the ILA schemes in Czechia and Lithuania.
Source: Information collected by the OECD project team.
Given the substantial involvement of public funding, ILA initiatives have faced challenges related to fraud, abuse, and price inflation. In response, governments have implemented a range of measures to mitigate these risks.
Inappropriate promotional activities
In the Netherlands, there were instances where providers engaged in inappropriate promotional activities and exploited the definition of eligible expenses to attract more learners. For example, some providers offered incentives such as trips or gifts to entice potential participants. To prevent such practices, these promotional activities were prohibited starting from the fifth round of the eight‑round STAP scheme. Furthermore, the original definition of eligible expenses – which included tuition, examination fees, and teaching materials – was revised following a review of the scheme’s first year. Specifically, the category of “teaching materials (lesmateriaal)” was replaced with “literature and resources (literatuur of middelen),” narrowing the scope to essential items such as books and safety equipment, and helping to limit the misuse of public funds (Onderwijsland.com, 2023[32]).
Price inflation and market concentration
Price inflation was another significant concern. A price monitoring exercise conducted two years after the scheme’s launch found that, on average, the prices of courses eligible for STAP increased more sharply ‑ by approximately 5 percentage points – than those of comparable courses not covered by the scheme (Buiskool et al., 2024[12]). However, this trend varied widely across providers, with observed increases ranging from 0 to 15 percentage points. Stakeholders attributed part of this increase to the additional administrative burden and implementation costs faced by training providers. For example, the scheme required learners to register for courses before receiving confirmation of funding. As a result, many participants cancelled their bookings when their applications were unsuccessful, leading to financial uncertainty and added complexity for providers.
The STAP scheme also revealed concerns regarding the concentration of public funding among large private providers. During the first six rounds, three major providers accounted for over one‑third of all STAP participants, while another seven providers collectively served more than 40% of learners (Buiskool et al., 2024[12]). Although the prominent role of private providers was anticipated given their strong presence in the Dutch adult learning sector, concerns about market dominance and equitable access led to corrective measures. From the fifth application round, a cap of 300 awards per training activity was introduced to prevent an excessive share of participants enrolling in a limited number of programmes (Onderwijsland.com, 2023[32]). From the ninth round onward, eligibility was further restricted to courses and providers officially recognised by the Ministry of Education, Culture and Science (Buiskool et al., 2024[12]).
Non-completion of courses
Another potential form of abuse arises when training providers offer courses to participants who do not actually attend. In the absence of a co-payment mechanism, the full cost of such training would be borne by the state. To mitigate this risk, some countries have introduced safeguards. For example, in Lithuania, providers are not compensated if the training is not completed, while in France, payments are made in proportion to the portion of the training attended (Mon Compte Formation, 2024[33]). However, these measures can place a disproportionate burden on honest providers, particularly when participants fail to complete training for personal reasons. Although, in theory, providers in Lithuania could attempt to recoup the costs directly from the participants, doing so is often difficult in practice.
Quality assurance is fundamental to the success of ILA schemes. It provides the necessary foundation to support learner autonomy by ensuring that all training options available through the scheme meet predefined quality standards.
Ex-ante quality assurance is typically integrated into broader national frameworks. However, when existing systems are fragmented or not well aligned with the specific goals of ILAs – such as in the case of non-formal providers operating with limited oversight – countries may establish ILA-specific mechanisms to ensure adequate quality control.
Ex-post quality assurance typically relies on learner feedback, and in some cases is complemented by audits or external assessments to evaluate the impact and effectiveness of supported training programmes. Going forward, advanced data analytics and artificial intelligence may also help detect patterns of fraudulent use of ILAs.
Strong quality assurance systems are essential not only for training effectiveness but also for safeguarding public funds by preventing fraud, abuse, and price inflation. However, overly burdensome or rigid systems can hinder innovation and responsiveness – particularly in fast-moving sectors like ICT or when the goal is to encourage a diverse range of providers, especially during the early stages of implementation.
An “enabling framework” refers to a set of complementary policies and practices that help remove barriers to adult learning participation and support individuals to use their ILA entitlements effectively. Many of these entitlements are embedded in the structure of adult learning systems and reflect how education and training are regulated in each country. Key complementary policies to ILAs include (i) paid training leave, (ii) career guidance, and (iii) the recognition of prior learning, as highlighted in the 2022 EU Council Recommendation on individual learning accounts (Figure 3.2).
Source: Authors’ elaboration based on the Council of the European Union (2022[7]), Council Recommendation of 16 June 2022 on individual learning accounts, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=oj: JOC_2022_243_R_0003.
Training leave to overcome time barriers to learning
Training leave may support the take‑up of ILAs by allowing employees to take time away from their regular job responsibilities to engage in skill development. While ILAs cover (some of) the costs of training, training leave is designed to free up individuals’ time – a frequent barrier to adult learning participation. None of the countries covered in this report has a systematic entitlement to paid training leave, such as those in Austria (Bildungskarenz) or Sweden (ledighet för utbildning), but in all the countries covered, training leave is available under certain circumstances (see Table 3.5 for an overview).
Note: *The programme does not give eligibility to training leave, but a training allowance for individuals combined with wage subsidies for employers.
Source: Republic of Lithuania (2023[34]), XII‑2 603 Republic of Lithuania Law on Approval, Entry into Force and Implementation of the Labour Code, https://e-seimas.lrs.lt/portal/legalAct/lt/TAD/6eea48d294ea11eea70ce7cabd08f150?jfwid=-k3id7tf7e (accessed on 28 February 2025);
Service‑Public.fr (2024[35]), Projet de transition professionnelle (PTP), www.service-public.fr/particuliers/vosdroits/F14018 (accessed on 28 February 2025); Government of Singapore (2025[36]), SkillsFuture Level-Up Programme, www.skillsfuture.gov.sg/level-up-programme (accessed on 28 February 2025).
In France, although a statutory entitlement for training leave (congé individuel de formation, CIF) was repealed in 2018, individuals still have a right to take leave to participate in training aimed at changing jobs or occupations. Employees wishing to use their ILA budget to finance training during working hours must obtain approval from a public body (Transitions Pro) under their professional transition project (projet de transition professionnelle).1 To be eligible, individuals must have been employed for at least two years, including one year with the same company. They must send a written request for leave to their employer, who can defer the leave under certain conditions.2 During the leave period, employees earning less than twice the minimum wage continue to receive 100% of their salary. Those earning more receive 90% of their salary during the first year of training and 60% in subsequent years. Salary costs are covered by Transitions Pro (Service-Public.fr, 2024[35]). Special provisions apply to public servants who can take a special leave called congé de formation professionnelle dans la fonction publique d’État (FPE) if they have at least three years of service. This leave can last up to three years over the course of a career. Public employees are entitled to return to their jobs and receive a monthly allowance equivalent to 85% of their previous salary, capped at around EUR 2 800 per month for one year, or two years for priority groups with lower educational attainment (Service-Public.fr, 2024[37]).
In Lithuania, employees can take education leave of up to five days per year for non-formal training, conditional on informing their employer at least 20 working days in advance, and up to 30 days for formal education. Education leave is remunerated by employers at a minimum of 50% of the employee’s income if the individual participates in formal learning or job-relevant non-formal training, and has more than five years of tenure (Republic of Lithuania, 2023[34]).
In the remaining countries, training leave is not a national entitlement but is instead dependent on collective agreements between employers and employees. Nevertheless, paid training leave is supported through various mechanisms. In Singapore, for instance, a SkillsFuture Mid-Career Training Allowance is available to Singapore citizens aged 40 or above who undergo substantive full-time upskilling or reskilling. While their jobs are not protected, eligible participants receive 50% of their monthly income for up to 24 months, with a maximum allowance of SGD 3 000 (EUR 2000) per month. From 2026, the scheme will be extended to include selected part-time training, helping mid-career workers to defray incidental training expenses. Under this extension, participants will receive a flat-rate allowance of SGD 300 (EUR 200) per month (Government of Singapore, 2025[36]).
Career guidance to support informed training choices
Career guidance is a complementary policy to ILAs, helping individuals choose training pathways that align with their skills and labour market needs. It often includes a skills assessment to identify an individual’s competencies and interests.
In many countries, career guidance is connected to ILA schemes. For example, in France, a detailed career guidance and skills assessment service (bilan de compétences) can be financed through the CPF scheme (Service-Public.fr, 2025[38]). The bilan de compétences can also be initiated by employers as part of a skill development plan to adapt to changing skills needs and maintain their staff’s employability. In 2023, it accounted for 6% of activities funded by the CPF (Caisse des Dépôts, 2024[9]). In Singapore, individuals can use SkillsFuture Credit to pay for approved career guidance programmes by endorsed providers (Government of Singapore, 2025[39]). In the Netherlands, the STAP scheme was accompanied by a separate subsidy for career development advice, but only 2% of STAP participants made use of this support (Dutch Ministry of Social Affairs and Employment, 2022[40]; Buiskool et al., 2024[12]). In Croatia, an advisory and informational interview is conducted before individuals receive the education voucher, either by the Croatian Employment Service or an adult education institution (Tomac, Zovko and Herceg, 2023[41]).
Beyond ILA-linked support, many countries offer free career guidance services. In France, the conseil en évolution professionnelle (professional development advice), which was introduced jointly with CPF, is a separate instrument but one of its objectives is to facilitate access to training, including training funded through CPF (Perez and Vourc’h, 2020[25]). While the bilan de compétences carried out through the CPF is fee‑based, generally limited to 24 hours, and includes detailed assessments of skills acquired through formal and informal learning, the conseil en évolution professionnelle is free of charge, has no time restrictions, and primarily focuses on providing information (Service-Public.fr, 2025[42]). Lithuania has since 2022 set up a free career guidance service for students, adults, and job seekers through 14 regional career centres (Ministry of Social Security and Labour of the Republic of Lithuania, 2024[43]; Karjeras, 2025[44]). Similarly, Czechia has a career guidance service for job seekers and employed individuals in place, which is provided through the Labour Office and other providers. It is free for adults however, stakeholders mentioned that there might be capacity restrictions.
Source: Labour Office of the Czech Republic (2021[45]), Informační a poradenské středisko [Information and advisory centre], www.uradprace.cz/informacni-a-poradenske-stredisko-up-cr (accessed on 4 March 2025); Euroguidance (2024[46]), Regional Work Centers (RWC): A new step in career guidance, https://euroguidance.eu/regional-work-centres-rwc-a-new-step-in-career-guidance (accessed on 28 February 2025); Government of Singapore (2025[39]), Career Guidance Programmes, www.wsg.gov.sg/home/individuals/career-matching-guidance/career-guidance‑programmes (accessed on 28 February 2025); Karjeras (2025[44]), Kas vyksta Karjere [What’s happening in the Career], https://karjerastau.lt/events?category=6 (accessed on 28 February 2025); Service‑Public.fr (2025[38]), Bilan de compétences d’un salarié du secteur privé, www.service-public.fr/particuliers/vosdroits/F3087 (accessed on 28 February 2025); Service‑Public.fr (2025[42]), Conseil en évolution professionnelle (CEP), www.service-public.fr/particuliers/vosdroits/F32457 (accessed on 28 February 2025).
In addition to career guidance provided by a counsellor or career coach, digital tools that provide career information or allow self-assessment of skills and interests may be valuable for adults’ training choices. In Croatia, for example, the online platform supporting the ILA pilot scheme – managed by the Croatian Employment Service ‑ includes a self-assessment questionnaire designed to help users identify educational programmes that align with their skills and interests (Croatian Employment Services, 2025[47]). In Singapore, the MySkillsFuture portal offers a suite of self-assessment tools that help users explore suitable upskilling and career options based on career interests, skills confidence, and work values (Government of Singapore, 2025[20]). Complementary resources are also available under the MyCareersFuture initiative: CareersCompass provides career-related knowledge and tools, while the CareersFinder feature leverages government data and AI to deliver personalised job recommendations based on users’ skills and experience, along with training opportunities aligned with their career goals (Government of Singapore, 2025[48]; MyCareersFuture, 2025[49]). The Ministry of Education and Science of Lithuania runs an online career guidance information system called Aikos, which makes data on education, statistics, and labour registers available to a wide range of customers, including career guidance service providers (Ministry of Education, Science and Sports of the Republic of Lithuania, n.d.[50]).
Recognition of prior learning to personalise training pathways
Formally recognising the skills that adults have acquired through informal and non-formal learning can allow individuals to participate in courses that address their skill gaps, offering shorter and more personalised learning pathways (Meghnagi and Tuccio, 2022[51]) and lowering the costs of training funded via ILAs. Recognition of prior learning (RPL) processes include an assessment of skills acquired on the job or through other means and can lead to full or partial qualifications, credits that count towards a qualification, or a certificate of labour market competencies.
In practice, RPL is often separate from ILA schemes and functions as one of several skills policies that can make ILAs more effective. France is one of the countries where RPL processes, called validation des acquis de l’expérience or VAE (validation of acquired experience), are connected with the ILA scheme. The CPF in France can be used to pay for the validation process, including the preparation, accompaniment, and assessment costs through Mon Compte Formation. The VAE process has several stages and usually takes between 8 and 12 months (France VAE, 2025[52]). Once obtained, a VAE certification has the same value as a certification or qualification obtained through training.
In the Netherlands, the option to fund the RPL process was originally meant to be integrated into the STAP scheme, but this was not achieved, and instead, a temporary subsidy scheme was created. Applicants could receive a contribution of up to EUR 1 000 towards the cost of the RPL process. The scheme was open to those who were at least 18 years old and had not reached retirement age (Dutch Ministry of Social Affairs and Employment, n.d.[53]).
In Czechia, a system for validating non-formal and informal learning enables individuals to obtain certification of their skills by passing examinations for (partial) qualifications, aligned with the National Register of Vocational Qualifications (NSK). This system allows qualifications within the NSK framework to be awarded through RPL. Currently, the cost of these examinations cannot be covered by the ILA budget. However, there is an ongoing policy discussion on expanding the ILA’s scope to include support for the RPL scheme.
An ‘enabling framework’ – including paid training leave, career guidance and RPL – can help remove barriers to adult learning participation and support individuals to use their ILA entitlements effectively.
Funding from ILA could be used to pay for career guidance and RPL, as is the case in France. However, whether this is desirable will depend on the existing career guidance and RPL provision in the country.
Establishing ILAs is a complex process, and while there are some general lessons that can be drawn from the experience of other countries, the exact design features of ILAs that work best will also depend on the specific context in which they operate. Therefore, it is critical that countries take an experimental approach to introducing ILAs, which involves monitoring their implementation, evaluating their impact and continuously adapting the scheme based on the results. This requires establishing a monitoring and evaluation framework that helps answer the following key questions:
1. Are ILAs having the desired impact on the target population?
2. Is the impact achieved in a cost-effective way?
3. What features of ILA design should change to increase their effectiveness?
The first step in developing a monitoring and evaluation framework that helps answer these questions is to explicitly define the desired impact of ILAs, for example, better employment outcomes or better ability to cope with everyday tasks for the beneficiaries (see Decision 1) and to have clarity about how ILAs will lead to this impact. The causal pathway from the establishment of ILAs to the impact can be summarised using the Theory of Change approach (Figure 3.3) which describes how the Inputs, such as funds spent on ILAs, lead to Outputs, such as participation in training, which ultimately result in desired Outcomes and Impact, for example, individuals developing new skills that help them become more productive at work and earn higher wages.
Note: This infographic presents a non-exhaustive list of inputs, outputs, outcomes and impact of ILAs.
Source: Authors’ elaboration.
What information is collected?
The next step is deciding what information should be collected to measure whether the desired impact of ILAs is achieved in a cost-effective way, and to identify what elements of the process described in Figure 3.3 might be hindering the achievement of ILAs’ objectives.
All countries covered in this report continuously collect some information about the Inputs, such as the cost of the scheme, and some Outputs, such as the number of beneficiaries participating in training funded via ILAs. In addition, many countries collect more detailed data about the characteristics of individuals who participate in training, to learn whether ILAs benefit the desired population. This includes information about beneficiaries’ gender, age, educational attainment, employment status, and place of residence. Similarly, many countries monitor whether the training was completed and the type of training that individuals undertook, such as the name of the course and provider, training duration, cost, topic and mode of delivery. This type of information is necessary to identify whether funding for the ILA scheme was spent in accordance with the legislation regulating ILAs, but it is not sufficient to understand whether ILAs have had the desired impact.
While most countries attempt to collect some information about the Outcomes and Impact of the use of ILAs, the data is often collected less frequently and tends to be less reliable than in the case of Inputs and Outputs (Table 3.7). For example, in Singapore, upon course completion, participants are asked if the course will help them progress in their work or career, as well as whether the skills and knowledge acquired in the course will lead to improvements in their performance at work. In addition, three months later, in a follow-up survey, beneficiaries are asked whether the training had an impact on their career progression, performance at work, and salary. In Czechia, training participants were asked three to six months after completing the training whether it had an effect on their likelihood of finding a job. In Lithuania, regular surveys of ILA users are conducted but focus on the Outcomes of the scheme, such as whether the individual developed new skills or applied what was taught during the course at work rather than its Impact. In France, a study of ILA users, conducted eight to nine months after training,3 included questions about the Impact of training participation on the likelihood of being in employment, but it was implemented as a one‑off survey rather than a regular data collection mechanism. In all the above‑described countries, if information about Impact is collected, it is self-reported and therefore less reliable. The exception is the Netherlands, where comprehensive information about the Impact of ILAs was collected by linking the individuals who participated in the STAP scheme to administrative sources, such as tax records, to collect an objective measure of the impact of the scheme on the likelihood of being employed and hours worked.
Note: Objective measures refer to those that are collected through administrative data, instead of self-reported by the participant.
Source: Information collected by the OECD project team.
Finally, some countries collect information about satisfaction with the mechanism to access funding and the quality of courses (referred to as the Process in Figure 3.3), which helps identify the parts of the process of implementing ILAs that need to be improved. For example, in Singapore, information about satisfaction with the training programmes is collected through the learner survey administered upon completion of training. In this survey, training participants are asked about overall satisfaction with training and various aspects of course quality, such as the alignment of content with its description, relevance to work, quality of the teaching materials and methods, and whether the course was good value for money (Government of Singapore, 2025[54]). In Lithuania, similar questions about the quality of teaching and materials are asked in a survey conducted upon completion of training. In addition, participants are asked whether they would recommend the course to others (Ministry of Education, Science and Sport of the Republic of Lithuania, 2023[55]). In Czechia, a follow-up survey distributed after the training includes questions about the channels through which individuals learned about the scheme, whether the co-payment was a barrier to participation, whether the access to funding was easy to navigate, and what could be improved about the scheme. Such information is useful not only to ensure a high-quality course offer (see Decision 6) but also to identify the aspects of ILA implementation that might be hindering the achievement of the desired Outcomes and Impact.
How is the information collected?
The ideal data collection mechanisms depend on the type of information needed. However, countries often face constraints related to the information systems in place and may need to resort to other methods of data collection instead of the preferred ones.
Note: * In Czechia, participants are asked to evaluate the course on a scale from 1 to 5, but no additional questions are asked.
Source: Information collected by the OECD project team.
The ILA portal and training providers can be a source of rich data about the training programmes that individuals participate in, such as their cost, duration, topic, and mode of delivery. These data sources are used, to varying extents, in all countries covered in this report (Table 3.8). In addition, in some countries, the personal identification number used to register on the platform contains information about the gender, age or other basic characteristics of the individual. The unique identification number can further be used to link the registered individual with other administrative data, such as tax records or education registers. In this way, reliable information about an individual’s employment status and income can be collected first at the point of registration to characterise the individual but also at a later stage to measure the impact of the scheme, as was done in the Netherlands.
Satisfaction of the individual with the ILA and the training is best measured through learner surveys, which in some countries, such as Singapore and Lithuania, are administered upon completion of the programme. Surveys can also be used to measure the impact of ILAs that either go beyond employment outcomes recorded in administrative data or in cases where linked administrative data is not available. For example, in Singapore, a follow-up survey conducted 3 months after training completion is designed to collect information about the course’s impact on the ILA participants’ jobs and careers (Box 3.4). However, data collected through a survey will necessarily be self-reported and, therefore, less reliable. Moreover, the response rate to surveys about ILA is low in some countries, and if individuals are more likely to respond when they are satisfied with the training, then the results of the survey may be biased.
Additional surveys of training providers and qualitative interviews with stakeholders can be helpful in identifying how the ILA platform and the rules regulating training delivery could be improved. Such surveys have been conducted in the Netherlands. Moreover, surveys of the general population to monitor awareness of the scheme can inform decisions about additional information campaigns and their design.
In Singapore, two surveys are administered to trainees who use SkillsFuture Credit as part of a broader Training Quality and Outcomes Measurement initiative introduced in 2018 to capture trainees’ feedback on the quality and career impact of publicly funded courses.
The first survey focuses on course quality and includes questions about the overall satisfaction with the course, alignment of course content with the course description, its relevance for work, the quality of the course materials and teaching, and whether the course was good value for money. In addition, participants are asked if the course would help them progress in their careers and improve performance at work (SkillsFuture Singapore, 2024[56]). Training providers are obliged to administer the survey at the end of the training programme via a QR code. Learners who do not complete the survey via the QR code (e.g. if they are absent) are invited to complete the survey via either email or text. The scores are published on the MySkillsFuture Course Directory to help course seekers in their selection of courses. From June 2026, course renewal requirements will be strengthened: providers must achieve a minimum survey response rate of 75% and must not fall within the lowest quantile of course quality ratings (SkillsFuture Singapore, 2024[56]).
The second survey focuses on the impact of the course and is administered 3 months after the end of the course. Respondents are asked whether the course helped them develop new skills, progress in their careers, improve their performance, and whether it contributed to changes in responsibilities and salary increases. The survey also includes questions about participants’ motivation to participate in the course and gives individuals the opportunity to provide suggestions on how the course could be improved (Government of Singapore, 2025[54]).
The set of survey questions was developed based on SkillsFuture Singapore’s research and data needs in consultation with training providers (SkillsFuture Singapore, 2024[56]).
How is the information used?
The collected information about ILAs is commonly used to track the achievement of the target Outputs levels, such as the number of participants in the scheme, their characteristics and the type of training they participated in. These key statistics are typically used to prepare monitoring reports, which can be either public, like in France, or used internally by the managing institutions, like in Lithuania and Singapore. The frequency with which monitoring reports are issued differs across countries. In the Netherlands, the monitoring report was prepared after each application round, i.e. every two months, to track the effectiveness of the scheme, and additional extended reports were published at a lower frequency. In France and Lithuania, monitoring reports are prepared on a yearly basis, while in Czechia, the first monitoring report will only be prepared towards the end of the pilot scheme, i.e. after more than two years since implementation began.
Beyond monitoring reports that focus on descriptive statistics about participation levels and self-reported impact, evaluation reports that present findings from counterfactual impact evaluations are desirable, but these are rarely prepared. For example, if the desired impact of ILAs is to boost the productivity and wages of participants, simply observing that wages increased after the use of ILAs is not enough to conclude that this was caused by the scheme rather than, for example, a result of an improvement in the country’s economic performance. Counterfactual impact evaluations can help establish a causal link between ILAs and the outcomes of participants by comparing the outcomes of individuals participating in the programme with the outcomes of individuals similar in all respects to the treated ones except for participation in the scheme. Among the countries covered in this report, only the Netherlands carried out an impact evaluation, which used administrative data to compare the employment status and hours worked of beneficiaries of STAP to a control group (Box 3.5).
In the Netherlands, a comprehensive evaluation of the STAP scheme was conducted to determine not only the type of beneficiaries who participated in the training but also to test the scheme’s effectiveness and cost-efficiency. For this purpose, a combination of data collected through surveys and administrative data was used. The impact of STAP on labour market outcomes was analysed primarily using administrative data by comparing STAP beneficiaries to individuals with similar characteristics who did not participate in the scheme. Survey data was used to learn about the impact on individuals’ self-confidence, attitudes and job satisfaction. The key findings of the evaluation were:
About half of STAP participants indicated that they would not have followed the chosen training course without the scheme, and this was particularly the case among those with lower levels of education (56%) than among medium and highly educated individuals (50%).
STAP participants reported that training boosted their self-confidence (75% responding affirmatively) and increased their interest in further learning (77%).
Those who were not in employment at the start of STAP-funded training were more likely to be employed 12 months after starting training (29%) than the control group (24%), and the difference was statistically significant. The impact was greater for those with less education and for older individuals (55+). Six out of ten participants who found a job since the start of the STAP-funded training believed that training participation helped them achieve that. However, STAP participants found jobs with, on average, lower hourly wages (by EUR 0.6) and fewer hours (by three hours per month) compared to the control group. Some catch-up was observed over time, and the authors of the evaluation speculate that the initial negative impact on hours and wages may be a result of changes in jobs/sectors after participation in training.
Those who were employed at the start of the training course were 2 percentage points less likely to be employed 12 months after starting training than the control group. Moreover, this group did not experience improvement in their number of hours worked, hourly earnings or monthly earnings compared to the control group. These effects were partly driven by individuals who participate in longer training who might have experienced a lock-in effect, as training participants typically spend less time and effort on job search activities than non-participants, and further analysis over a longer time period is needed to verify if such effects persist.
Source: Buiskool et al. (2024[12]), Evaluate en monitoring STAP-regeling: Doeltreffendheid en doelmatigheid STAP-regeling [Evaluate and monitor STAP scheme: Effectiveness and efficiency of STAP scheme], https://open.overheid.nl/documenten/023f6e9f‑400c‑4baf‑9d46‑17a4105fa49a/file (accessed on 15 May 2025).
Finally, if the detailed information collected through the monitoring process is made publicly available, it can help individuals make better training decisions and support training providers in adapting their offer to the learners’ needs. In France, relatively detailed data on the use of CPF is shared publicly. For example, detailed information about the most popular training courses is available online and can help training providers decide how to adjust their training offer to meet the needs of learners (Caisse des dépôts, 2024[57]). In Singapore and Lithuania, the ratings calculated based on the learner survey are displayed on the portal next to each training scheme to help guide the training decisions of individuals. They also serve as feedback for providers on satisfaction with their courses. In addition, in Singapore, detailed results of the learner survey are shared with training providers.
It is critical that countries take an experimental approach to introducing ILAs, which involves monitoring their implementation, evaluating their impact and continuously adapting the scheme based on the results. Countries considering establishing ILAs should first have clarity on how the inputs, such as funds spent on ILA, lead to outputs, such as participation in training, which ultimately result in desired outcomes and impact, for example, individuals developing new skills that help them become more productive at work and earn higher wages.
Measuring the performance of ILAs at all these stages, and not just the inputs and outputs, as is sometimes the case, is critical. To do so, countries can leverage data collected through the ILA portal, information transferred by training providers, as well as linked administrative data from other governmental sources. These can be complemented by learner surveys.
To assess whether the scheme achieves its objectives, counterfactual impact evaluation should be conducted to complement the standard monitoring reports.
← 1. According to Article L6323‑17‑1 of the French Labour Code.
← 2. Employers have 30 days to respond to the request by employees. If they do not respond within this time period, authorisation for training leave in the context of the professional transition project is considered to be granted. Employers may defer the employee’s request for a maximum of 9 months, but this must be justified by negative consequences of the employee’s absence for the company’s operations. Rules vary according to the size of the company; for companies below 100 employees, 1 employee at a time can take the training leave, while for companies above 100 employees, 2% of employees can take training leave at a given time (Service-Public.fr, 2024[35]).
← 3. The data collection took place from the beginning of July to mid-August 2022 and included people who completed a training course in November 2021.