This chapter assesses the success factors and persistent challenges for Italy’s current system. It begins by synthesising the most relevant lessons highlighting some of the key lessons learned from modern contracted-out employment programmes across the OECD. These lessons provide the context for identifying both the positive features, and the challenges and bottlenecks of the current system in Italy. The discussion uses GOL’s national framework as its foundation and synthesises broader OECD experience as inspiration.
A New Model for Strengthening Public‑Private Partnerships in Employment Services in Italy
Abstract
Recent developments in Italy – including the implementation of the GOL programme – represent a positive step towards more modern, user-centred ALMPs that better support the needs of both jobseekers and employers. At the same time, however, important challenges remain. These include uneven implementation across regions, with some regions struggling to implement the programmes given a low administrative capacity, weak monitoring of provider outcomes that make it difficult to reap the potential benefits of competitive forces, a lack of sustainable funding which inhibits durable investments in the programme on the part of providers. Lessons from other OECD countries point to important changes that could be made in this direction.
This chapter provides an assessment of Italy’s current system from an international perspective. It then assesses positive features of Italy’s approach, focussing on GOL’s national framework: harmonised pathways, unified profiling tools, harmonised essential service levels and reimbursement rates, and a broad ecosystem of providers and partnerships. The chapter also documents bottlenecks observed in implementation, such as uneven administrative capacity, complex and delayed reporting rules, fragmented monitoring, limited information for jobseekers, and payment structures with incentives for sustained job placements. These assessments are intended to lay the foundation for the proposed new model in the next chapter.
Almost half of OECD countries contract out employment services through outcome‑based payment models (Langenbucher and Vodopivec, 2022[1]). They employ a variety of approaches in key aspects such as market design, service provision and payment models. These experiences can offer helpful lessons for Italy as it seeks to introduce a new model of contracted-out employment services. For example, evidence from the European PES Network shows that partnerships with private and third-sector actors are increasingly indispensable to expanding vacancy access, improving job matching and strengthening support for vulnerable groups, provided that co‑operation is underpinned by high-quality data exchange and strategic co‑ordination (European Commission, 2025[2]). Drawing on international experience, this section highlights key features of successful systems – such as well-aligned incentives, transparent performance monitoring, and client-centred service delivery – and considers how these lessons can be applied in the Italian context.
At the heart of successful contracting systems is a strong partnership between the contracting authority and providers. Lessons from other OECD countries show that well-performing models invest in structured engagement through provider councils, regular consultations, and forward guidance on key programme developments to align expectations and respond to delivery challenges (Langenbucher and Vodopivec, 2022[3]; Vodopivec, 2023[4]). For example, the Netherlands organises an annual employment services provider conference that all of their roughly 800 providers are invited to engage in dialogue with the PES. Before trialling and eventually introducing a new employment services model, the province of Ontario (Canada) organised several rounds of consultations with a wide array of stakeholders. Such approaches foster trust and enable providers to invest in their services, including with capable and experienced staff. In Italy, institutionalising dialogue between the central government, regions and providers would help move beyond transactional contracting and towards longer-term partnerships.
A diverse and competitive provider market can enhance service quality and responsiveness. Other countries which also have choice‑based models – such as Australia, the Netherlands and Sweden – have systems that are well-placed to incorporate a diverse provider mix, including not-for-profits and specialist actors, and to create conditions for innovation and client-tailored support (Langenbucher and Vodopivec, 2022[1]). While Italy already has a diverse network of accredited providers, encouraging competition on quality and strengthening monitoring is essential for fully realising all the benefits this wide array of providers can offer.
Informed client choice is a cornerstone of well-performing international models (Vodopivec, 2023[4]). When jobseekers can choose their provider, it increases user engagement and motivates providers to differentiate themselves through quality. However, the choice must be meaningful: simply publishing a list of accredited providers is not enough. Jobseekers must be supported by clear information and guidance that permits comparative assessment. Countries such as Sweden and Australia combine provider profiles, performance data, and caseworker support to help clients make well-informed decisions. Italy’s choice‑based programmes provide a foundation for this, but further improvements in transparency, data and user support are needed to make the choice work in practice.
Robust IT infrastructures provide the data exchange backbone that underpins the most successful models of contracted-out employment services. The United Kingdom (UK), for example, has automated the exchange of data between providers and public authorities to track delivery, monitor outcomes and trigger payments. Its system improves accuracy, reduces administrative burden, and supports real-time oversight. In Italy, digital infrastructure has improved in the period when GOL has been implemented, but further integration would streamline operations and support evidence‑based performance management.
Minimum service standards help ensure consistent support for jobseekers. Without clear requirements, service quality can vary widely across providers or regions. The UK’s experience shows that implementing minimum service requirements can help establish more consistent service for all jobseekers. Countries like Ireland and Sweden specify minimum contact frequencies, personalised action plans, and diagnostic assessments to ensure a reliable baseline of support. In Italy, where regional variation has been a persistent challenge, introducing additional standards on service frequency to complement the existing regulations would strengthen equity while preserving space for local flexibility and innovation.
Digital-first strategies enable a subset of job-ready and digitally literate clients to be referred to self-service online tools and platforms. Such systems, which have been introduced by a number of countries including Australia, France and Sweden, can improve the efficiency of services provided they are carefully designed and targeted. Many systems use profiling tools to direct job-ready individuals with strong digital literacy to self-service digital platforms, freeing counsellors to focus on clients who require more intensive support. However, digital-first does not mean digital-only: safeguards such as opt-out rights, counsellor discretion to override profiling recommendations, and the ability to always access in-person support are vital to protect vulnerable users. High-quality portals and tools are also critical to ensure digital channels are useful and effective. Italy has begun to integrate these elements into its services; expanding the approach could increase efficiency while maintaining support for those most in need.
Outcome‑based payments hold the promise to strengthen provider incentives, but their design matters. Payment models that reward sustained employment and include payments for soft-outcomes such as progress towards specific, preset goals tend to perform best across diverse participant groups, as shown by the example of the Netherlands. Italy’s current outcome‑related payments under GOL are limited in scope and weight. International examples suggest that increasing the share of results-based funding and outcomes for which providers are rewarded could improve the effectiveness of the system. Linking payments to longer-term labour market outcomes can incentivise better job matches. But systems need to be designed carefully to ensure fair client allocation and protection against “creaming” and “parking”.
Finally, a strong evaluation culture underpins continuous improvement. Several OECD countries – including, notably, Canada – embed regular impact evaluations into programme cycles and use findings to adjust design. Italy has made progress in building an evaluation ecosystem but lacks a structured, long-term approach. Establishing a national steering committee, promoting counterfactual evaluations, and systematically publishing results would strengthen accountability and support more effective policymaking.
The strong progress made in providing employment services in Italy means that the country is well on its way towards systematically providing high-quality, tailored support to each of its jobseekers. The GOL programme has laid important groundwork, but systemic reform is needed to consolidate progress and address remaining gaps. By tailoring international good practices to its own institutional context, Italy can build a model that is more consistent, responsive, and outcome‑focussed.
This subsection distils the elements of the current system that appear to work best. It highlights design and delivery features that should be retained in a future model, as well as features of regional models that could be rolled out nationwide.
2.3.1. Common GOL framework
The GOL programme has established, for the first time, a standardised national framework that includes consistent definitions of jobseeker activities, a unified profiling system, and a structured approach to financing via the nationally defined UCS. Prior to GOL, regional disparities meant that some areas had sophisticated ALMPs while others lacked basic infrastructure; furthermore, some regions had not yet employed outsider providers for publicly-funded employment services. The introduction of common Essential Levels of Performance (LEP) has led to progress towards ensuring a baseline of services across all regions. GOL strengthened the use of common profiling tools to systematically assess jobseekers’ needs and introduced pathways that differentiate support according to distance from the labour market, from rapid job-search assistance to intensive training and social inclusion measures. The use of nationally defined UCS has created a transparent and predictable financing framework, reducing uncertainty for providers and improving the similarity of key programme parameters across regions. These elements have been widely welcomed by providers and regional authorities, especially as they have brought much-needed coherence across Italy’s fragmented ALMP landscape.
The GOL programme also expanded the role of private providers in employment service delivery in many regions, which had not engaged outside providers for employment services (e.g. job matching) in the past. This inclusion fostered public-private synergies, enabling a more robust support network and better sharing of information and resources. Private providers became integrated partners, helping to deliver tailored services while benefiting from clearer rules and responsibilities within the national framework.
Another strength of the GOL framework is its dual emphasis on minimum service standards and policy customisation, allowing for adaptations to sectoral and territorial employment challenges and to clients’ specific needs and circumstances. It has enabled better alignment between training and employment services, helping to close skill gaps by tailoring training offers to labour market demand. The approach has allowed for individualised pathways defined through detailed assessments, enabling more accurate and targeted interventions that have the potential to meet clients’ specific needs.
2.3.2. Flexibility for regions to adapt the programme
While standardisation has been a key strength of GOL, the programme also permits a high degree of flexibility across regions and, to a lesser degree, over time. Regions have been able to tailor interventions to their specific labour market contexts, allowing for local innovations and responsiveness. For example, Valle d’Aosta was able to add workers employed in the tourism sector to their list of target groups given the precariousness of their seasonal work. Regions were also able to design their training programmes according to their local labour market needs. For example, Bolzano integrated additional regional services into the programme, including language courses in Italian, German, English, and French and specialised skills training in its key sectors. In the PARs updates for 2024, some regions have also noted that extending training access to all pathways, including Pathway 1, enabled all participants to benefit from training opportunities. This flexibility to adapt the programme’s parameters to both regions’ unique challenges and to broader labour market developments has been a remarkable achievement.
2.3.3. Ambitious targets for participation combined with strong advertisement and outreach campaign
The GOL programme has set ambitious targets for participation: the 3 million participants who are envisioned to participate by the end of the programme in 2025 represent roughly 12% of Italy’s labour force. This is particularly ambitious given the underwhelming uptake of some past programmes, such as the previous national system of contracted-out employment services, the Reintegration Voucher, that was first initiated in 2015. Although the launch of the GOL programme was initially met with implementation challenges, Italy is on track to have met the participation targets by the end of 2025. This has been facilitated by strong co‑ordination between the central government and the regional authorities, who meet regularly to compare and monitor progress in achieving their GOL participation targets through a key co‑ordination body, the Conference of Regions and Autonomous Provinces (Conferenza delle Regioni e delle Province autonome).
Unlike earlier national employment initiatives, GOL benefited from a robust communication campaign that improved visibility and engagement across the country. The programme was well promoted, and regions were actively involved in outreach efforts, leading to high levels of awareness among jobseekers and encouraging participation. For example, Abruzzo created a network of GOL information points involving municipalities, libraries, schools, universities, youth centres, third-sector organisations, and accredited providers. These information points provide GOL-related information free of charge and prioritise access in peripheral and disadvantaged areas. To facilitate access, the locations of these information points are shown on a publicly available map. Similarly, in Lombardy, local authorities and not-for-profit organisations are able to provide basic labour market information services without requiring accreditation. These efforts have contributed to the successful uptake of the programme.
2.3.4. Combination of process and outcome‑based payments
One of the challenges of the prior national model of employment services in Italy, the Reintegration Voucher, was its very heavy focus on outcome‑based payments. Almost all the payments in that scheme – roughly 90% – were contingent on participants achieving employment outcomes. The extremely strong financial incentive to place participants into employment arguably resulted in undesirable behaviour from some providers, leading them to devote less resources in assisting more difficult-to-place clients. In GOL, by contrast, the payment model places a much stronger emphasis on process-based payments. Consultations with stakeholders indicate that many welcomed the introduction of a payment model that combines process and outcome‑based elements, as it can better mitigate the risk of cherry picking. The process-based element allowed providers to spend the required time understanding the participant’s needs. These process-based payments are especially important given the characteristics of many of the GOL clients, especially in Pathway 4, who may have multiple barriers to employment.
Another qualified strength of the GOL system is its flexibility for regions to adapt their payment models to regional preferences. For example, while in the majority of regions the payment for the accompaniment to work service is process-based, Campania incorporates a results-based component. Lombardy has adapted its payment model for training courses to incentivise dual-mode courses, which combine classroom-based training with practical work experience, by increasing the share of funding reimbursed based on the successful implementation of work placements. This regional experimentation with different payment models provides the possibility for rigorously evaluating the relative merits of different payment models in order to inform the design of future national schemes.
2.3.5. Enhanced IT platforms and digital tools
Italy has also made remarkable progress in developing and implementing IT platforms and digital tools to provide some cutting-edge digital solutions to facilitate the job-matching process, identify individuals’ training needs and facilitate the monitoring of whether jobseekers are fulfilling their activation requirements, which has proven useful in the implementation of GOL. At the national level, the newly-implemented SIISL platform promises to facilitate information sharing across the network of employment service providers. SIISL links unemployment benefits to activation requirements by automatically enrolling beneficiaries into required training and prompting early engagement (e.g. completing a CV within 15 days). If beneficiaries fail to act, alerts are sent to the CPI, triggering follow-up. This system addresses longstanding inefficiencies, such as the “parking” of benefit recipients, by linking social and unemployment benefit receipt with active engagement. SIISL has also introduced a clearer interface for both providers and participants, with AI-powered tools that support CV completion, skill profiling, and job-vacancy matching.
Regions have also adopted various solutions to expand digital access to services. For example, Lombardy has enabled self-profiling through its IT platforms allowing users to register with chatbot assistance and access training and guidance tools once validated. Tuscany has deployed 40 digital facilitators, available remotely or in person, to help users navigate the Toscana Lavoro portal, issue the DID, schedule CPI appointments, attend online interviews, and sign the Personalised Service Pact digitally.
Digital job portals are being enhanced in several regions to streamline services. For example, Tuscany’s Toscana Lavoro portal allows CPIs and accredited providers to use a shared appointment system, as well as manage job postings and track applications in real time. The portal’s CV search function is powered through the INTUITION engine, which creates logical matches between job requests and CV content, including CVs uploaded by people not registered with PES and people outside of the region. Another example is Veneto’s ClicLavoro Veneto portal, which facilitates remote access, job matching, CV publishing, and real-time service tracking. Veneto also developed an integrated labour and social services platform (Sistema Informativo per l’inclusione Sociale e Lavorativa, SILS) that connects public employment services, social services and private providers, enabling fast and co‑ordinated responses to users in need of economic support and job-search assistance. It ensures interoperability with national systems, such as the Unified Information System of Social Services (Sistema Informativo Unitario dei Servizi Sociali, SIUSS).
Updated information systems have also facilitated the tracking of participants’ adherence to GOL participation requirements. Consultations with stakeholders indicated that these systems were used to follow up with, and potentially sanction, participants who failed to sign up to services as agreed in their individual action plans. While some stakeholders also indicated that jobseekers’ adherence to job-search requirements was not subject to the same stringent monitoring, enforcing participation requirements represents an important step towards more rigorously enforcing mutual obligations requirements for jobseekers.
2.3.6. Collaboration between employment services and social services
Several regions have established systems that formalise collaboration between employment services and social services to support individuals with multiple barriers to employment. One example of a promising practice is from Sicily, which has established a Single Access Point (Punto Unico di Accesso, PUA) within CPIs, where dedicated counsellors conduct in-depth assessments of vulnerable participants. Based on identified needs, individuals are directed to both labour market measures (e.g. job orientation, training) and relevant social or healthcare services. Referrals work in both directions: social services can also refer individuals back to employment services when employability potential is identified. A formal Memorandum of Understanding (MoU) is being drafted to institutionalise co‑operation between employment and social services.
Even in cases where regions are not adopting a “one-stop shop” approach to serving clients, several regions have adopted a more co‑ordinated approach between employment and social services. Piedmont employs a collaborative model whereby CPIs work directly with social welfare operators during the assessment and planning phase. A “planning booth” is established, bringing together employment and social actors to design personalised intervention plans. In addition to employment pathways, individuals may participate in social inclusion internships, which help them gain work experience in a supportive context. Similarly, Tuscany applies a multi-professional intake process for vulnerable individuals. CPI counsellors work alongside social and healthcare services to co-develop tailored socio‑occupational integration plans, ensuring a co‑ordinated, cross-sector response.
2.3.7. Diverse network of providers offering potential for specialised services that meet clients’ and employers’ specific needs
Italy’s diverse ecosystem of employment and training providers – which include CPIs, private employment agencies (APLs), not-for-profit organisations, specialist training providers (including ITS Academies), and labour consultants – offers the capacity to match support to varied participant needs and employer demands. This diversity can provide sector-specific and local knowledge – often from providers who are already engaging with other relevant actors through their other lines of business – that can tap into the existing personal and professional networks that are most commonly used in the job matching process in Italy. They can also harness the competencies of a varied set of individual providers through different delivery modes, from intensive guidance, group sessions and work-based learning. This can enable services to be personalised across GOL pathways, including for groups facing multiple barriers. Finally, public-private partnerships and consortia (e.g. ATS/ATI) can bundle counselling, skills development and employer outreach into coherent packages that may be otherwise difficult to achieve with smaller, more specialised providers. The result is a broader menu of interventions – short upskilling, longer reskilling, targeted inclusion measure – better aligned with both jobseekers’ circumstances and employers’ vacancies.
This section takes stock of the main challenges and bottlenecks that undermine the effectiveness of the existing system of employment services in Italy. These include the uneven administrative capacity to administer the regional programmes, complex reporting rules, and limited provider monitoring. The section also examines how payment structures, lack of information for jobseekers to make an informed and deliberate choice of providers, and long referral periods can allow clients – especially those with complex needs – to disengage. The section sets the basis for the specific solutions proposed to address these challenges through a new model of employment services in Italy (see Chapter 3).
2.4.1. Using counterfactual impact evaluations to learn and guide policy development
A related challenge on using data to provide insight for Italy is the absence of systematic, regular counterfactual impact evaluations to guide policy development. While counterfactual impact evaluations are currently being conducted, their results will not be available until the current GOL programme has concluded. This means they will be useful in designing future schemes, but that they cannot be used to inform changes to the existing scheme – even though short-term evaluations could have informed parametric changes to the system, given that the system was first launched in 2022. Regularly conducting rigorous evaluations – using experimental or quasi‑experimental methods and leveraging administrative data – would provide timely evidence on what works, for whom, and at what cost. This, in turn, could support adjustments to key design parameters (e.g. payment models, service mixes, targeting rules) and help compare performance across pathways, providers and regions.
Introducing a more systematic approach to impact evaluation and evidence‑building would permit better evidence‑generation and progressively build evidence on policy effectiveness and implementation. This could include the use of randomised control trials (RCT) to build evidence on specific policy design (such as contract design- payment thresholds, payment groups, “soft” outcome payments), building in a degree of automation to policy evaluation (for example automatic provider data feeds to enable streamlined counterfactual evaluation), and a region-led analytical work programme to strategically build evidence on contracted-out programmes.
For example, in Canada legislation mandates regular ALMP evaluation. Provinces and territories have local responsibility for ALMP implementation and evaluation, but have agreed an evaluation workplan that is implemented federally by the employment ministry and managed by an evaluation committee of all participating provinces and territories.1 This brings the advantage of economies of scale and consistency across all the evaluations (OECD, 2022[5]).
Careful RCT design can also be used to provide high quality evidence on the dynamics of policy delivery itself. In Sweden, the RCT design of Rusta och Matcha allowed the investigation of policy effectiveness across several dimensions. The RCT permitted an evaluation of the performance of private providers relative to the PES delivering the same service in-house. In addition to this, extra treatment arms (participants operating under different policy conditions) allowed investigators to evaluate how changing the performance‑related payments to providers impacted their behaviour and participants entry into employment (Egebark et al., 2024[6]). Designing the trial in this manner allowed policymakers in Sweden to look not only at programme effectiveness, but the relative benefits of contracting-out and how contract design influenced effectiveness. Similarly, an RCT of contracted-out services in the Netherlands added an additional treatment arm, to look at two variations of programme design – a more and less intensive variation. This found that the more intensive counselling variation did achieve better outcomes, though without any additional cost-benefit analysis it did not conclude which service offered better value for money (de Hek, Toorn and Chazan, 2024[7]).
2.4.2. Simplifying provider reporting and strengthening their monitoring of providers
A key challenge in the GOL programme concerns the monitoring of providers’ activities and their clients’ outcomes. Few regions currently operate robust IT systems that generate reliable, timely, and provider-specific data on participant progress and results – information that is essential for steering strategy and allocating resources. This has important parallels to the discussion on provider ratings and relative performance comparisons on the participant side. Tracking of service delivery and engagement also remains weak, particularly for private providers whose activities often remain a “black box”. Recognising the importance of monitoring for improving the performance of the system, some regions themselves have highlighted the need to strengthen monitoring data on service results.
From the perspective of providers, one important challenge relates to the reporting requirements – the information they need to provide to the CPI. Delays in formalising monitoring and reporting rules created significant uncertainty for providers, especially private actors, during the GOL programme’s rollout. The official GOL reporting manual, published by the Ministry of Labour and Social Policies, was officially posted only three years after the launch of GOL, even though it had been applied informally with regional input in the interim. Fragmentation across funding streams has added to the complexity: GOL reporting requirements differ from those of other ALMPs, increasing administrative burdens and contributing to provider hesitancy, particularly among organisations new to the system or serving complex participant groups.
Delays in the formalisation of monitoring and reporting rules created significant uncertainty for providers, particularly private actors, during the programme’s rollout. The reporting manual was only officially ratified three years after programme launch, although it had been applied informally in co‑ordination with regions during that time. The lack of unified reporting mechanisms across funding streams has further complicated matters. GOL reporting requirements differ from those associated with other active labour market measures, adding to the administrative burden and contributing to provider hesitancy, particularly among actors newly entering the system or working with complex participant groups.
The legal and bureaucratic complexity of the GOL programme has also increased the workload for both CPIs and providers. Digitalisation of procedures – such as electronic signatures, automated forms, pre‑filled templates, and interoperable data exchanges – could substantially reduce this burden. Increasing the reporting requirements for jobseekers on their job-search and other activities – as part of the activation requirements – could also serve to shift the burden of reporting requirements.
Reporting requirements can impose a significant administrative burden on service providers (as seen during the implementation of GOL), which may discourage some providers from engaging in the market for employment services. One way to alleviate this burden is to shift part of the reporting obligations onto (some groups of) jobseekers, introducing a requirement for them to report regularly their job-search activities and participation in activation measures through an online portal. Providers would then have an option to review these entries, add comments, or make amendments as needed. The requirement could be applied only to certain groups of jobseekers who have necessary skills and characteristics to be able to meet such obligation (e.g. have adequate digital skills). This can be implemented using an approach similar to what is done in Sweden in terms of jobseeker reporting.
Several programmes include monitoring metrics on participant compliance, which provides a better understanding on participant engagement with services. For example, in Employment Placement in New Zealand, providers are required to submit attendance logs to the PES, which show the extent to which participants are engaged in the services. Where programme attendance is mandatory, this data sharing is required so that PES can enforce any sanction decisions that may need to be made for non-compliance with obligations (such as in the Work Programme and Restart Scheme in the United Kingdom).
To reduce administrative reporting burden, countries should also consider how to automate data capture with providers. In the United Kingdom, the Department for Work and Pensions (DWP) leverages data from the tax authority in an automated digital system that enables efficient contract management and supports longitudinal tracking of outcomes post-placement (for 18 months in Restart, though with the possibility to extend this period to any desired length required by the particular programme). The advantage of this direct data ingest is that it also helps to minimise reporting error and potential fraud, whilst simultaneously enabling prompt payment to providers.
2.4.3. Improving jobseekers’ ability to make a deliberate and informed choice
An important precondition for harnessing the full potential of Italy’s diverse ecosystem of providers is ensuring that jobseekers make a deliberate and informed choice of provider. This is important for several reasons. Having a jobseeker pick the most suitable provider for them can improve that individual’s own outcomes, as it can provide them with support from a provider that is best placed to match their career aspirations or help them overcome their personal barriers to employment (Vodopivec, 2023[4]). Similarly, having jobseekers channelled towards better performing providers can improve overall outcomes directly, through e.g. better job outcomes, but also indirectly, through increased competitive pressures: once providers realise they need to offer better services in order to obtain clients, they have an additional incentive to offer better services.
In the current GOL system in Italy, the information available for participants to make an informed choice is rather limited. Few clients have the possibility to make a truly well-informed choice in practice, with choices very often based on geography or service type rather than provider quality. Enhancing participant choice would require better information and transparency mechanisms, such as provider rating systems. It would also entail offering support and encouragement for jobseekers to make effective use of this information, as experiences from other countries show that jobseekers are often reluctant to dedicate sufficient time in making a deliberate choice, often defaulting to choosing the closest provider.
Several regions indicated in their Regional Implementation Plans (PAR) that they are working to develop such rating systems, such as the one Veneto had in place for its previous system of employment services. For example, Puglia plans to introduce a public rating system for accredited providers to support informed user choice, which may include participant feedback mechanisms, similar to online consumer ratings platforms. Calabria is developing a performance rating model for providers, focussing on their effectiveness in job placements and service delivery. Abruzzo also aims to introduce a regional rating system, enabling jobseekers to evaluate providers, particularly based on job placement rates. Tuscany currently records qualitative information on participant reviews of services and may use this data to structure a rating system.
Although some regions are exploring these tools, a co‑ordinated national effort could help leverage the considerable information available at the central level to compile updated provider ratings that could help them make a better-informed choice. Such a monitoring framework, in the longer-term, could also help automate the outcome‑based payments made to providers, making them more timely while reducing the administrative burden. It would also have the added potential benefit of simplifying reporting requirements, between both regions in relation to the central government, as well as between providers and the CPIs.
For client choice to be effective, jobseekers must be supported with information, guidance, and quality assurance mechanisms to avoid inequities and gaming. Many of the programmes offering participants choice in OECD countries have readily available metrics to judge provider performance. A new system for Italy could include a model for rating providers based on the employment outcomes of participants, and/or using participants’ assessments of providers (see Box 2.1 as an example). This would have two purposes. Firstly, monitoring providers to help target oversight efforts towards the least effective providers. Secondly helping jobseekers make an informed choice. This system would require individual-level information on placements of participants with individual providers (including the dates and pathways). Some regions already have such a system in place (e.g. Veneto).
Alongside relative performance metrics using quantitative and qualitative metrics provided online, provider choice can be facilitated using expert counsellor judgement to guide jobseekers. In the Netherlands, provider choice is facilitated by conversations with counsellors, who have in depth knowledge of the provider landscape and can advise potential participants on which providers might be suitable for them. This model aims to empower clients with the necessary knowledge, guided by experts, to self-select service providers. After agreeing on a reintegration plan, clients choose from a list of pre‑approved providers, supported by the PES counsellor and detailed provider profiles. These profiles encompass contact information, service areas, specialisations, methodologies, historical performance, and labour expert feedback, promoting a competitive environment that motivates providers to differentiate themselves based on quality and outcomes (Langenbucher and Vodopivec, 2022[3]).
In Sweden, jobseekers who are assessed as needing moderate support (groups A, B and C) are referred to Rusta och matcha, a system of contracted-out employment services. They are offered freedom of choice among approved providers, with the Swedish public employment service – Arbetsförmedlingen – publishing performance ratings (a four‑star system) to enable jobseekers to make an informed provider selection, which accounts for their performance.
Arbetsförmedlingen using a survival analysis model with 11 jobseeker characteristics to determine the predicted likelihood of becoming long-term unemployed. This score determines which group a jobseeker is assigned and also allows Arbetsförmedlingen to determine expected job outcomes for providers. Because the model uses elapsed duration of unemployment claim, it can adjust probabilities based on time in unemployment. This allows the later re‑classification of individuals in who were initially classified as being able to self-serve, but who do not find work quickly.
Expected provider outcomes then account for their participant mix. Providers with a large proportion of individuals who are determined as high risk of long-term unemployment, would not be expected to place as many individuals into work quickly. Once a provider has enough clients (more than 18), Arbetsförmedlingen calculates their relative performance rating, compared to expectations. It does this for all providers and then ranks them. Providers are given star ratings which depend on how far above or below average their performance is. Providers with above average performance are given 3, or 4 (more than one standard deviation above the mean) stars. Providers with below average performance are given 1 (more than one standard deviation below the mean), or 2 stars. This permits a relative assessment, which does not use absolute thresholds (it is practically impossible for all providers to be given 4 stars) and instead evaluates providers on performance relative to other providers.
These ratings are then published on the website that participants use to choose providers, allowing participants to filter and sort providers by rating. This helps to provide objective performance metrics to participants so that they can choose providers based on their ability to help individuals into work.
For countries that have not invested in the resources for expert advice on provider choice delivered by employment counsellors, comprehensive, well-presented and easily accessible performance data are essential to guide participant decision making. Since moving to fully contract out employment services in 1998, Australia has also developed several iterations of its provider performance rating system for its contractors. It now implements a mixed approach that aims to provide individuals with a demonstrated progress towards employment allied to a high-quality service delivery. Workforce Australia replaced an earlier provider “Star ratings” to include a broader set of performance measures. Views of participants and employers are combined with metrics on outcomes across five individual modules. High, medium and low ratings are provided for each of the individual modules, and for the individual metrics comprising each module. A final overall rating then aggregates these individual modules, giving most weight to the sustained employment module (which measures 12‑ and 26‑week employment outcomes) (DEWR, 2025[10]). Ratings are published quarterly and used to guide both jobseeker choice and government decisions on contract continuation (DEWR, 2025[11]). This transparent, data-driven system enhances accountability and provides ongoing performance incentives (see Box 2.2 for more details on this approach).
With the rollout of its newest iteration of contracted-out employment services, Australia has implemented a comprehensive Provider Performance Framework (PPF) that broadens the scope compared to the “Star ratings” of the previous system, which used regression analysis to measure provider’s employment outcomes after accounting for the composition of their clients. The new system now includes a range of performance dimensions intended to provide a more holistic assessment of employment service providers. The framework, introduced with the new Workforce Australia employment services, is organised into five modules:
Sustained Employment: success in placing jobseekers into sustained jobs, measured at 12 and 26 weeks;
Progress Toward Employment: providers’ effectiveness in improving participants’ employability through training, referrals and other progress measures;
Quality of Service to Participants: the quality and responsiveness of services delivered to jobseekers, gauged via participant feedback and independent service assessments;
Quality of Service to Employers: the quality of engagement and services provided to employers, measured through employer feedback and compliance checks; and
Licensing Standards: the provider’s adherence to contract requirements and expected standards of conduct.
For each module – and for individual indicators within them – providers receive a rating of High, Medium or Low performance. These feed into an overall performance rating for each provider (or each provider “license” in a region). Importantly, the overall score gives the greatest weight to the sustained employment outcomes module, reflecting the programme’s core goal of durable job placements.
The Department of Employment and Workplace Relations (DEWR) compiles and publishes the provider ratings quarterly. By making performance data transparent, the system empowers jobseekers to make informed choices. For example, a jobseeker can view which local providers are rated “High” overall or on service quality, and may opt to register with a top-performing provider. At the same time, the government uses the ratings as a management tool. Providers with persistently poor (Low) ratings may face closer scrutiny or loss of business share, whereas those with High ratings gain credibility and may be entrusted with expanded contracts. Indeed, DEWR explicitly links the PPF to contract decisions: performance ratings inform whether a provider’s license is extended.
Source: Select Committee on Workforce Australia Employment Services (2023[12]), Rebuilding Employment Services: Final report on Workforce Australia Employment Services.
2.4.4. Managing efficiently the resources available to some CPIs would help alleviate capacity shortages
The shift from voluntary participation under the Youth Guarantee to mandatory participation for some target groups under GOL has led to an increase in caseloads in many regions. Somewhat paradoxically, CPI counsellor workloads often increased even though employment services were contracted out, due to a dramatic increase in the intensity and scale of services provided to jobseekers, limiting the ability to e.g. follow-up with jobseekers and more intensively monitor the work of providers. Because GOL funding could not be used to expand public capacity (e.g. by hiring additional staff), many CPIs have had to manage higher volumes with existing resources.
Despite progress, regional disparities persist in the quality of administrative capacity for overseeing the system of contracted-out employment services. Several regions have indicated that administering the GOL programme was a considerable challenge due to staffing shortages. For example, in Puglia, the regional authorities were able to hire a considerable number of counsellors in the period prior to GOL adoption, increasing the number of front-line counsellors from roughly 300 to 1 300 across their 44 local CPIs. Nevertheless, during the period of GOL implementation, these counsellors have had high workloads. To maintain service continuity, some activities previously delivered in-house, such as specialised orientation, were contracted out to external providers. This experience highlights the importance of ensuring the administrative capacity by the contracting authority, with sufficient resources in place to guarantee the delivery of quality services.
The limited administrative capacity of some CPIs, particularly in the context of increasing workload for CPI counsellors, may lead to regional disparities in service quality and longer waiting times for jobseekers. To make better use of limited resources, some activities that are currently delivered in-house by CPIs, could be considered for contracting out to accredited external providers. For example, in Lombardy, external providers can serve as initial contact points or “gatekeepers” for some groups of jobseekers. They perform the role of CPIs, such as intake and initial interviews, reducing the workload for CPI counsellors and enabling them to focus on clients with more complex needs. This approach could potentially be considered in other regions, provided that there is an increased workload for CPIs, adequate capacity to deliver necessary services by external providers and no legal or regulatory barriers to outsourcing these functions.
2.4.5. Using profiling tools to streamline jobseekers’ customer journeys and refocus the work of CPI counsellors
The current system does not exploit the full potential of the existing profiling tools to help streamline jobseekers’ customer journey. Currently, the first step of the profiling process – the quantitative profiling – is delivered on-site with jobseekers in the presence of counsellors. This makes it less useful for the counsellors, because they are already in direct contact with the clients – meaning that many of them will already get a sense of their clients’ potential barriers to employment through direct interactions with them. In addition, having meetings with all registered jobseekers – including those who are likely to become quickly employed without additional in-person assistance – means that counsellor resources are diverted from e.g. counselling those clients who would need the most assistance.
Part of the reason for the current setup is the focus on maximising participation in GOL to meet participation targets. This has arguably resulted in jobseekers for whom self-service job search tools would have sufficed, being channelled into the system. The relatively high share of referrals who never start activities suggests scope to streamline intake, reduce avoidable in-person contacts, and concentrate resources where they are needed the most.
The use of existing profiling tools can help to streamline jobseekers’ customer journey by better targeting the support provided and ensuring that CPI counsellors focus their time on those groups of jobseekers who need it most. Not all jobseekers require the same level of assistance: some are likely to find work relatively quickly on their own without additional in-person counselling. For these job-ready individuals, access to self-service job-search tools, such as online vacancy databases and job-matching platforms, may be sufficient to facilitate their return to employment. This approach is used in both Australia and Sweden, where more job-ready individuals are first referred to digital services (see Box 2.3 for more detail on Australia), leaving more resources to help those with greater need.
Online profiling tools for digitally-literate clients in Italy could be used to identify individuals who fall into this job-ready category and allow them to self-serve initially. This would limit the tendency, driven by participation targets, to channel quickly employable people into GOL pathways that add little value. Triaging the more job-ready would also enable CPI counsellors to refocus their work on jobseekers with greater barriers to employment, such as long-term unemployed and other vulnerable groups. To account for varying levels of digital literacy and personal preferences, opt-in/opt-out options could be provided to give identified job-ready jobseekers a choice to either use self-service tools or to request more personalised counsellor support.
Launched in July 2022, Workforce Australia Online represents a wholesale shift to digital-first employment services for jobseekers closest to the labour market. Under this model, jobseekers assessed as job-ready are placed in an online self-service stream managed centrally by the DEWR, while those with higher support needs go to contracted external providers.
Only some participants enter Workforce Australia Online and everyone can opt out
Upon registering for unemployment benefits, individuals are profiled using the Job Seeker Snapshot questionnaire. They are then assigned to the self-service Workforce Australia Online, or referred to provider-led Workforce Australia Services. If severe barriers are indicated, they have a specialist assessment prior to referral.
The model builds in mechanisms to transition individuals out of the online stream if needed. Many jobseekers naturally exit early: nearly two‑thirds of those who left online services in the first year did so because they found employment – most within 6 months. If a participant remains unemployed at the 12‑month point (or 18 months in some cases), they are automatically moved to a provider in Workforce Australia Services for more intensive support. Jobseekers also have the right to opt out of the digital service earlier if they feel they need greater support. Roughly one‑quarter chose to transfer to a provider within the first 3 months of the programme. These transfers (whether voluntary or at the 12‑month mark) ensure that the online cohort does not indefinitely retain struggling individuals.
Activation requirements are still monitored for those referred to Workforce Australia Online
Participants in Workforce Australia Online agree to a personalised Job Plan through the website (without meeting a coach). The plan outlines required monthly tasks, quantified through a points system. Each activity (job applications, training modules, workshops, etc.) carries a point value, and the jobseeker must accumulate a target number of points each month (typically 100 points, including at least 4 job applications) to satisfy their mutual obligation requirements. Individuals have flexibility to choose activities from a menu to reach the total. They report completions online and track progress on their dashboard. Crucially, compliance is automatically monitored: the system sends a reminder at Day 15 if the person is not on track, and another warning five days before the month’s end if the target is still not met. If the points target is achieved, the requirement is met (and any extra points can be partially carried over). If not, the system triggers a formal compliance process under predefined rules. Intervention by DEWR staff is minimal, with no regular meetings. A Digital Services Contact Centre (operated by the Department) is available if participants need help or have questions.
Australia uses feedback to iteratively develop the model
Australia also demonstrates the virtue of using feedback to continuously improve policy implementation and outcomes. Changes to the model were introduced in March 2026 to improve the link between eligibility and digital literacy. Previously there was potential for those with few digital skills being referred to digital self-service tools. However, Australia has addressed this issue by giving the digital capability assessment a direct impact on eligibility. Those with low digital skills are no longer eligible for referral to digital self-service tools.
Source: Select Committee on Workforce Australia Employment Services (2023[12]), Rebuilding Employment Services: Final report on Workforce Australia Employment Services; DEWR (2025[13]), Overview of the Points Based Activation System, https://www.dewr.gov.au/points-based-activation-system-pbas-workforce-australia
2.4.6. Long placement periods risk having jobseekers get “lost” in the system, especially for vulnerable clients
A recurring challenge concerns services for those who need the most support – in the case of GOL, the Pathway 4 (Work and inclusion) participants. These clients are typically referred to a provider, but follow-up from the CPIs is often insufficient, particularly over long placement or training periods. Without regular checkpoints and clear ownership of each case, participants can disengage and, effectively, “fall between the cracks”.
In practice, some Pathway 4 participants have dropped out because training pathways are lengthy, training allowances are too low to sustain basic living costs, or counsellor follow-up is limited. To address this and reduce drop-out rates, some regions have adopted mitigatory measures. Tuscany reduced classroom hours and expanded online delivery to ease access for people in remote areas and those with caring responsibilities. Lombardy strengthened continuity by setting a minimum contact frequency (at least one meeting every 20 days). Such adjustments help sustain engagement and reduce attrition.
A more modular approach with shorter referral periods and shorter milestones would further reduce the risk of clients getting “lost”. Recognising that vulnerability varies within Pathway 4 – as in Valle d’Aosta’s sub-categorisation – allows the intensity and mix of support to be matched more closely to need. Facilitating switches between pathways where circumstances change (e.g. health, care duties, or readiness for work) can also ensure that employment plans remain relevant and motivating.
One way to reduce drop-out rates and maintain engagement is to introduce shorter (modular) placements for jobseekers with clearly defined specific targets (including soft outcomes, e.g. engagement in regular social, volunteering or training activities), regular checkpoints and shorter milestones. CPI counsellors would monitor the process and assess whether the jobseeker is making adequate progress; if they are not, they can be referred to another provider or pathway better suited to their needs. If they are making adequate progress, they would continue, potentially triggering outcome‑based payments for providers. This approach draws on both the Dutch and Australian models (see Box 2.4), in which jobseekers’ progress is tracked step by step from social isolation to engagement in community and training activities, and ultimately to autonomous employment.
Netherland’s “participation ladder” creates smaller supportive steps towards employment
The Netherlands has implemented a creative employment programme that incorporates its “participation ladder”. Providers working in this set-up are eligible for payments for specific steps which moves participants with complex needs closer to the labour market. There are six discrete steps which start when the participant has little physical contact outside of the home, and end in paid employment. Steps in between include being able to participate in structured activities, such as volunteering, and reaching the position to be ready to start work or training.
Job counsellors provide a mechanism to quality assure and audit the delivery of these steps to employment. They independently determine whether an individual has made sufficient progress to move towards the next step, removing any conflict of interest from providers or participants self-reporting progress. Building progressively in this manner allows individuals to build confidence reaching milestones and gradually remove their barriers to work.
“Progress payments” in Workforce Australia also gradually break down employment barriers
Similarly, Workforce Australia introduces several types of “progress payments” for specific outcomes which improve an individual’s chance of finding employment. For example, providers receive a payment when a participant completes activities such as a six‑month accredited education course, short vocational training, four weeks of unpaid work or volunteering, or non-vocational interventions like addiction treatment, mental health support, or securing stable housing. These payments acknowledge meaningful steps toward employment and are limited to one per participant every two years.
Source: Vodopivec, (2023[4]), Raising the bar: Designing and implementing innovative contracted-out employment services in OECD countries, https://doi.org/10.1787/c7a819e8-en; DEWR, (2025[14]), Workforce Australia Guidelines, https://www.dewr.gov.au/workforce-australia/resources/workforce-australia-guidelines-part-b-workforce-australia-services.
2.4.7. Operational safeguards such as minimum service requirements and stronger engagement from CPI counsellors would facilitate more consistent service
Although GOL introduced a common framework, service definitions and delivery standards still vary markedly across regions, especially for clients with complex needs (e.g. Pathway 4). Current requirements are often high level; for example, they specify the total maximum number of hours but not how often counsellors should meet clients. This looseness leads to uneven practice, gaps in follow-up, and a higher risk of disengagement among the most vulnerable participants. A parallel challenge is the lack of regular follow-up between CPI counsellors, jobseekers and their chosen employment services provider.
One contributing factor to the limited follow-up with participants is the target-oriented design of the programme. The focus on achieving measurable outputs, such as training completions or job placements, has in some instances reduced the emphasis on process monitoring and mid-course adjustments. As a result, CPI counsellors may have less capacity to provide ongoing support, especially for participants on longer or more complex pathways. This has implications for tracking service continuity and ensuring that participants receive timely support if barriers to participation arise during the programme cycle.
Many countries use minimum service standards to ensure a baseline service level for participants. Such requirements can, for example, prevent providers from focussing their attention only on the most readily employable groups and minimise the risk of “parking” (see Box 2.5). Whilst Italy already has some of these elements there is a possibility to strengthen others to ensure good participant service levels.
Australian evidence shows that where providers have few minimum requirements some participants received little support
Evidence from Australia’s early employment services indicated that black box models (those with very few service requirements) can lead to “parking”, where providers focus on clients who are easier to place in employment and neglect those who are harder to serve. In Australia, some job-ready participants were placed too quickly into unsuitable employment, with fewer hours, whilst other clients with greater barriers were left with little support.
Australia has learnt from its previous provision to design policy that supports quality for all its participants. The current Workforce Australia system implements clearer minimum service expectations, better safeguards and strengthened performance monitoring.
The United Kingdom has also moved away from “black box” provision
Informally, this synopsis is supported by the experience of the United Kingdom, who have gradually moved away from the black box freedoms of the Work Programme to slightly more “grey box” requirements in the recent Restart Scheme, with minimum service requirements imposed upon providers.
Source: O’Halloran, Farnworth and Thomacos, (2020[15]), Australian unemployed workers’ experiences of being parked and creamed by employment providers, https://doi.org/10.1002/ajs4.131.
To ensure that individuals are given a tailored service from providers, the majority of contracted-out programmes across the OECD stipulate that an individual action plan must be completed for individuals. Most often this action is mandatory only at the start of the placement with the provider. However, in Wisconsin Works, the programme implemented by the State of Wisconsin since 1997, this requirement is time bound, such that an updated action plan is necessary every six months (Langenbucher and Vodopivec, 2022[3]). This condition aims to ensure that providers are not “parking” individuals and efforts are made continually to review the participant’s needs and continue the journey towards work.
Providers are typically given more flexibility in how they manage ongoing contact with their participants, though even here many programmes stipulate requirements on minimum contact frequency with individuals. Around one‑third (35%) of the programmes put some kind of requirement on minimum meeting frequency. The most common minimum meeting frequency is monthly (every 4 weeks). Providers may meet more frequently with their clients, but are bound by the minimum frequency. In New Zealand, minimum frequency in Employment Placement is related to its segmentation model and differed between monthly, fortnightly and weekly dependent on the level of participant needs (see Langenbucher and Vodopivec (2022[3]), for more information).
Alongside requirements on meeting frequency, some programmes place additional requirements on providers. For example, in an effort to ensure adequate capacity for counsellors to work with clients, the private operators’ scheme in France imposes a maximum capacity of 40 clients per counsellor. This sits alongside its requirement for weekly meetings, imposing quite an intensive regime on providers. Arguably this gives providers slightly less room for manoeuvre and innovation (it is difficult to envisage meetings much more frequently than weekly) but limits the scope for inequitable provision across individuals. Ireland imposed a variation on the updated action plans required in Wisconsin Works, by mandating quarterly “in-depth reviews” of progress towards employment for participants in its JobPath programme. Meanwhile in Restart qualitative requirements on diagnostics assessments for participants are given, to ensure that action plans for individuals aim to holistically cover needs (DWP, 2025[16]).
Operational safeguards could be introduced in Italy to help address these challenges. These include minimum service standards for providers (e.g. at least twice monthly meetings, with remote and group meeting options); a dedicated CPI worker to co‑ordinate services; progress checkpoints tied to concrete goals; as well as early-warning triggers (missed sessions, repeated non-starts) that prompt a joint review with the provider and CPI counsellor. Increasing jobseeker reporting requirements – with frequent reports that could be viewed and commented on by the private provider before being shared with the CPIs – could also strengthen engagement of clients while lowering the administrative burden on providers. Together, such measures could shorten feedback loops, clarify responsibilities, and help ensure that clients – especially the most vulnerable ones – remain engaged and supported throughout their journey. To a certain extent, many of these features could be conceivably integrated into existing IT platforms.
2.4.8. Strengthening co-creation of key programme parameters through better engagement with employment service providers
A well-functioning contracted-out employment-services system depends on robust, structured relationships between the contracting authority and providers. This means consulting with providers early and routinely when designing key programme parameters – accreditation criteria, UCS/payment rules, service standards and reporting templates – through predictable consultation cycles. Clear contact points, agreed service‑level expectations, and shared performance dashboards create fast feedback loops and a common evidence base. This decrease in uncertainty can improve the operating environment for private providers, reducing implementation risks and administrative frictions, and enables providers to plan capacity and invest in quality – all of which are critical to the system’s success.
While the Ministry of Labour and Social Policies maintains regular bilateral contact with providers, broader stakeholder interaction remains limited. Establishing structured forums that regularly bring together regions, CPIs, and private providers (including social economy actors and training organisations) would enable lesson-sharing and joint problem-solving. Standardised feedback channels – consultation calendars tied to key policy milestones, calls for comment on accreditation/UCS updates – would make exchange systematic and transparent. Giving providers a formal role in co-design and implementation (e.g. thematic working groups on accreditation, data/reporting and payment models) would ensure that frontline experience feeds directly and promptly into national and regional policy adjustments.
Regional variation in accreditation processes creates barriers for some private providers seeking to operate at the national level. Many systems prioritise physical infrastructure over quality of service and staff qualifications. Greater alignment between national and regional accreditation frameworks would help to streamline procedures and reduce the heavy administrative burden. However, it is important to acknowledge that transitioning to a unified national system may be difficult to achieve. Many regions maintain their own accreditation systems to better address local needs, institutional contexts, and workforce dynamics. Progress towards better alignment will need to respect regional autonomy and seek consensus on common standards where possible.
Establishing a systematic means to engage with providers offers the prospect of better policy feedback, and a more direct means to influence market provision of services. This could include funded initiatives to strengthen provider collaboration, and ensuring that there are formal communication channels for providers to input into market design. For example, providers in the United Kingdom fund transversal research into employer engagement to provide guidance and advice on market development (see Box 2.6).
In the United Kingdom’s ongoing Restart programme, which provides intensive employment support for the long-term unemployed, the employment services providers established an industry-led funding partnership known as ReAct. Six of the eight prime providers collectively contribute a portion (0.01%) of programme revenues to fund the ReAct continuous improvement network.
The partnership is housed and managed by the social partners (the Institute for Employment Studies, working with the Institute of Employability Professionals and the Employment Related Services Association).
ReAct uses its pooled resources to conduct research, share best practices, and pilot innovations across the Restart programme. The partnership focusses on developing a shared employer engagement approach, so that employers see a consistent approach to employment services across jurisdictions. It also reviews the evidence of “what works” in effective employer engagement more generally, to drive performance improvements in the programme on the likelihood of participants finding good, sustainable jobs. It has conducted a desk-based review of the existing evidence and interviewed employers, operational staff and field experts to collate evidence to drive forward programme implementation.
This effectively acts as an internal levy on providers that is operated at no direct cost to government or participants that creates a stable platform for quality improvement and innovation in service delivery. By reinvesting a slice of providers’ earnings into programme development, the partnership sustains ongoing improvements and helps ensure the programme’s long-term effectiveness.
Source: Institute for Employment Studies (2025[17]), ReAct publications https://www.employment-studies.co.uk/research-collections/react-and-restart-publications.
2.4.9. Refining the payment models could better incentivise providers and account for regional differences
Consultations with stakeholders indicate that the process-based element of GOL’s payment model has allowed providers to spend more time to understand participants’ needs, which the previous employment services scheme – where over 90% of payments were outcome‑based – had previously neglected. While the previous national scheme arguably placed too much emphasis on outcome‑based payments, the current model may have gone into the other extreme. Outcome‑based payments in GOL provide only weak financial incentives to reward providers for placing their clients into employment.
Looking ahead, a more balanced mix of incentives could strengthen performance. For the most readily-employable groups, one option would be to increase the outcome‑based share while rewarding longer-term employment outcomes. For more vulnerable groups, rewarding providers for outcomes other than employment is a sensible option to consider, given that only a small share of such clients become employed. Outcomes to be considered include either other “hard” outcomes such as entry into education or “soft” outcomes such as subjectively-determined (by a CPI counsellor) success in progressing toward pre‑defined, client specific goals. For the social economy and other mission-driven actors, a differentiated model with a higher process-based share (even with a lower total potential envelope) may be more appropriate, reflecting their role with complex cases and integrated inclusion services. For training, combining milestone payments (enrolment, completion, certification) with placement bonuses where training is explicitly linked to vacancies could better align incentives with labour market outcomes.
In addition to the share of outcome‑based payments, a related challenge pertains to the absolute levels of the maximum payment amounts, determined by the UCS values. There are no regional adjustments to UCS values, despite differences in operating costs across regions. Some providers have reported financial losses in delivering services under the current UCS. For example, while the national UCS sets the price of one hour of orientation at EUR 39.94, private agencies often pay counsellors a greater amount per hour, excluding the cost of support staff, infrastructure, and overheads. Additionally, providers note that each hour of direct participant contact typically requires further time dedicated to administrative or compliance‑related activity, which is not separately compensated.
In response to these challenges, the Ministry of Labour and Social Policies has initiated a process to revise the UCS levels. In 2024, a technical working group, supported by regional committees and representatives from employer associations such as Assolavoro, conducted a comprehensive cost analysis using internal financial data and standardised templates across seven regions. The analysis took into account factors such as recent increases in labour costs linked to the renewal of collective agreements, and the full administrative load borne by providers. The analysis is intended to support a discussion between the ministry and the European Commission on the potential revision of the UCS rates. The payment model for training provision also needs to be revised to better reflect the real costs of service delivery, especially for sector-specific training that requires expensive resources.
A final consideration relates to the timeliness of the payments made to providers. Stakeholders noted that in several regions, payments to providers were made with considerable delays and subject to uncertainty. The exact causes for these delays are unclear, but they are likely tied to the administrative procedures involved in making payments to providers. Such delays not only threaten the viability of providing services for existing providers, but also make it less likely that new entrants will decide to participate in the market.
Although GOL includes elements of performance‑related payment, further consideration over their precise structure could better align provider incentives with customer outcomes. For example, extending the employment period necessary to generate an outcome‑based payment could help to stimulate better job matches (see Box 2.7 for some discussion and examples).
In addition, the alternative outcome payment triggers already discussed (such as steps towards employment) could help to better engage jobseekers with greater barriers, given that a relatively low share of such jobseekers may be expected to enter employment.
Alongside variation in overall proportion of performance‑related in providers total payments, their release to providers over time may also vary significantly. Vodopivec (2023[4]) provides several examples of how payment profiles vary, dependent on both scheme and the dynamics of job placement (looking at the timing and duration of participants’ employment. One set of examples is re‑produced below to demonstrate the variability inherent across different contracted-out programmes. Figure 2.2 shows the payment schedule that providers receive when they place a participant in employment in month five of their programme, and that individual then stays employed for 12 months.
Source: Adapted and updated from Vodopivec (2023[4]), “Raising the bar: Designing and implementing innovative contracted-out employment services in OECD countries”, https://doi.org/10.1787/c7a819e8-en, Figure 2.1.
Each of the four programmes employs a very different payment schedule. Flanders’ flat payment structure is notable. Its providers are paid an initial (comparatively high) placement fee at the beginning of the contract, but do not receive any further payment until the individual has reached one full year of employment. This contrasts to both Australia and New Zealand, where providers receive lower initial attachment fees (relative to Flanders), but are paid upon job placement (immediately in New Zealand and after four weeks tenure in Australia) and then progressively paid through the months of attachment. Meanwhile Sweden requires a job attachment of three months before additional payments are made.
The payment structure is important because it affects how providers work with clients and their ongoing cashflow. To avoid bad quality job matches payments can be tied to sustained job attachment. This helps to focus providers on achieving good matches and incentivises them to continue to work with participants, even after they have secured employment. For example, Australia incentivises its objective for sustainable employment by paying its final performance‑related payment to providers after six months of sustained job tenure.
Source: Vodopivec (2023[4]), “Raising the bar: Designing and implementing innovative contracted-out employment services in OECD countries”, https://doi.org/10.1787/c7a819e8-en
2.4.10. Ensuring stable funding and service continuity
The time‑limited funding of the GOL programme brings into question broader issues of service continuity and how planning can be facilitated to allow regions (and their private delivery partners) to make long-term investment decisions to support their jobseekers and improve labour market matching. Uncertainty around future financing can discourage providers from expanding services, recruiting staff, or investing in infrastructure, all of which are essential to sustain the progress achieved under GOL.
Some regions are already taking steps to maintain and expand employment services beyond the GOL programme, embedding GOL principles into broader policy frameworks.
Calabria intends to build on its GOL experience, either by integrating lessons learned into national or regional programmes. The region is actively participating in national planning discussions on post-GOL active labour policies. Liguria also plans to preserve the structure established under GOL and improve it further by integrating it more closely with traditional vocational training.
Emilia-Romagna has made strong commitments to continue support even after the conclusion of the GOL programme through various policy frameworks, including the Labour and Climate Pact (Patto per il Lavoro e per il Clima), which promotes sustainable development and enhanced public-private collaboration in employment services, as well as the ESF+ Regional Programme 2021‑2027, which supports access to employment, particularly for young people, via measures like career guidance, traineeships, self-employment support, and skills development in strategic sectors (e.g. green, digital, and blue economy).
Tuscany plans to extend and consolidate the organisational models tested under the GOL programme. It aims to transition from a time‑limited, project-based framework towards a stable and structural ALMP system. This strategic direction seeks to ensure the continuity of service delivery beyond the lifecycle of the PNRR, while orienting policies toward greater inclusion and responsiveness to evolving labour market dynamics. The region aims to expand coverage to new target groups and increasing the involvement of private providers from the initial intake phase. A number of emerging priority groups have been identified for future interventions, such as facilitating the school-to-work transition for young people, strengthening dual training for young people in apprenticeships, providing skills certification and language training for foreigners under special protection. Furthermore, the integration between the GOL programme and the regional justice and social reintegration networks will make it possible to build training and work placement paths to foster rehabilitation and employability of prisoners who have been admitted to alternative sentences than imprisonment.
While these regional initiatives are promising, a stronger and more centralised strategy is needed to provide stable funding and guidance, ensuring continuity and reducing the risk of fragmented or uneven provision once GOL resources are phased out. While it is understandable that the levels of financing will vary depending on political priorities and the labour market situation, constantly having a programme running – even if with a relatively low number of participants – can ensure the infrastructure is in place so that it can be quickly ramped up should the need arise.
A stable and continuous supply of finance can help countries to provide support for jobseekers, regardless of the economic conditions. This argument holds whether talking about ALMPs more generally, or when considering those ALMPs contracted out to external partners. Research has demonstrated that a greater degree of stability for ALMP funding helps to ensure greater impacts of ALMPs in reducing unemployment (Sahnoun and Abdennadher, 2024[18]). Another dimension for funding stability in contracted-out programmes is the creation of resilient markets of providers. Stability of funding helps to ensure that there is a ready and willing pool of providers that can tender for contracts when new programmes are conceived, or existing ones are renewed.
One way of ensuring funding stability is to create policy rules that automatically scale spending to account for demand. Creating rules removes the need to make ad hoc adjustments depending on economic circumstances and ensures equity between different cohorts of jobseekers. Despite this, the practice is still relatively sparsely practised in OECD countries. Only Belgium (Flanders) and Switzerland have some kind of link between the number of jobseekers in the country and the amount of ALMP spending (Lauringson and Lüske, 2021[19]). The advantage of this approach is that countries can seamlessly scale expenditure as new jobseekers arrive. This is particularly useful in times of increases in the rate of unemployment, to allow the swift implementation of support to help jobseekers. In the context of contracted-out programmes, which can have longer lead times due to necessity to tender contracts, this automation is particularly useful to reduce feasible implementation timescales.
The time‑limited funding of specific programmes such as GOL creates uncertainty regarding service continuity and complicates long-term planning and investment decisions for regions, CPIs and private service providers. To address this, alternative funding models could be explored to provide greater transparency and predictability, and ensure the continuity of quality service delivery.
The current system of ALMPs combines a mix of funding sources, with some sources subject to relatively large fluctuations that make it difficult to build a robust ecosystem of providers. Having in place a framework where a programme similar to GOL is implemented over the longer term, even at a vastly smaller scale, can help it be more rapidly scaled up should the need arise and additional funds to become available.
References
[9] Arbetsförmedlingen (2023), Arbetsförmedlingens nya modell för bedömning av arbetssökandes stödbehov, https://arbetsformedlingen.se/download/18.6546760718a1c207b931ed0/1694504706728/arbetsformedlingens-nya-bedomningsmodell.pdf.
[8] Arbetsförmedlingen (2022), Betygsmodellen i Rusta och matcha, Arbetsförmedlingen, https://arbetsformedlingen.se/download/18.29c19c8e1948841b16628cb/1740479312036/beskrivning-betygsattning-leverantorer.pdf.
[7] de Hek, P., A. Toorn and A. Chazan (2024), Effectevaluatie Dienstverlening Wga: Tweede voortgangsrapportage, SEOR Erasmus School of Economics, https://www.uwv.nl/assets/files/600b6984-f27d-4489-9158-a73c8de97e37/effectevaluatie-dienstverlening-wga-tweede-tussenrapportage.pdf.
[13] DEWR (2025), Overview of the Points Based Activation SystemThe Points Based Activation System (PBAS) for Workforce Australia, https://www.dewr.gov.au/points-based-activation-system-pbas-workforce-australia.
[14] DEWR (2025), Workforce Australia Guidelines, Department of Employment and Workplace Relations, https://www.dewr.gov.au/workforce-australia/resources/workforce-australia-guidelines-part-b-workforce-australia-services.
[11] DEWR (2025), Workforce Australia Services Deed of Standing Offer 2022 - 2028, Department of Employment and Workplace Relations, https://www.dewr.gov.au/download/13970/workforce-australia-services-deed-standing-offer-2022-2028/40012/workforce-australia-services-deed-standing-offer-2022-2028/pdf.
[10] DEWR (2025), Workforce Australia Services Provider Performance Ratings, Department of Employment and Workplace Relations, https://www.dewr.gov.au/workforce-australia/workforce-australia-information-providers/workforce-australia-services-provider-performance-ratings (accessed on 2 July 2025).
[16] DWP (2025), Restart Scheme provider guidance, https://www.gov.uk/government/publications/restart-provider-guidance/restart-scheme-provider-guidance.
[6] Egebark, J. et al. (2024), En effektutvärdering av arbetsförmedling med fristående leverantörer, IFAU, https://www.ifau.se/globalassets/pdf/se/2024/2024-17-en-effektutvardering-av-arbetsformedling-med-fristaende-leverantorer.pdf.
[2] European Commission (2025), PES partnership with private employment services, state of play and trends (Survey Report), https://ec.europa.eu/social/BlobServlet?docId=27955&langId=en.
[17] Institute for Employment Studies (2025), ReAct publications, https://www.employment-studies.co.uk/research-collections/react-and-restart-publications/ (accessed on 4 December 2025).
[3] Langenbucher, K. and M. Vodopivec (2022), Paying for results: Contracting out employment services through outcome-based payment schemes in OECD countries (Online Annex), OECD Publishing, Paris.
[1] Langenbucher, K. and M. Vodopivec (2022), “Paying for results: Contracting out employment services through outcome-based payment schemes in OECD countries”, OECD Social, Employment and Migration Working Papers, No. 267, OECD Publishing, Paris, https://doi.org/10.1787/c6392a59-en.
[19] Lauringson, A. and M. Lüske (2021), “Institutional set-up of active labour market policy provision in OECD and EU countries: Organisational set-up, regulation and capacity”, OECD Social, Employment and Migration Working Papers, No. 262, OECD Publishing, Paris, https://doi.org/10.1787/9f2cbaa5-en.
[15] O’Halloran, D., L. Farnworth and N. Thomacos (2020), “Australian unemployed workers’ experiences of being parked and creamed by employment providers”, Australian Journal of Social Issues, Vol. 56/4, pp. 596-611, https://doi.org/10.1002/ajs4.131.
[5] OECD (2022), Assessing Canada’s System of Impact Evaluation of Active Labour Market Policies, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/27dfbd5f-en.
[18] Sahnoun, M. and C. Abdennadher (2024), “Active Labor Market Policies and instability of their allocations: A changing relationship in the case of regions of Tunisia”, Regional Science Policy and Practice, Vol. 16/6, p. 100026, https://doi.org/10.1016/j.rspp.2024.100026.
[12] Select Committee on Workforce Australia Employment Services (2023), Rebuilding Employment Services: Final report on Workforce Australia Employment Services, https://www.aph.gov.au/Parliamentary_Business/Committees/House/Former_Committees/Workforce_Australia_Employment_Services/WorkforceAustralia/Report.
[4] Vodopivec, M. (2023), “Raising the bar: Designing and implementing innovative contracted-out employment services in OECD countries”, OECD Social, Employment and Migration Working Papers, No. 301, OECD Publishing, Paris, https://doi.org/10.1787/c7a819e8-en.
← 1. Regions can also opt out. Currently Quebec discharges its evaluation requirements via its own internal evaluation programme.