This chapter presents a proposal for a revised model of contracted-out employment services in Italy. The proposal builds on the strengths of the current GOL framework while addressing existing limitations identified through national experience and international comparisons. The chapter outlines key design principles, details mechanisms to enhance efficiency and targeting, and introduces changes to strengthen provider incentives, accountability and performance.
A New Model for Strengthening Public‑Private Partnerships in Employment Services in Italy
Abstract
Despite the significant progress Italy has made in scaling up contracted-out employment services – particularly through the GOL programme – important structural challenges remain. The current system operates within a complex multi-level governance framework, with variation in implementation approaches, funding cycles and provider engagement practices across regions. Outcome‑based incentives remain relatively weak in practice, information for participants is limited, and administrative processes can be burdensome for both public authorities and providers. At the same time, the expiry of PNRR funding creates uncertainty about the long-term sustainability of recent gains. Addressing these issues does not require a wholesale redesign of the system, but rather a targeted refinement that strengthens coherence, improves operational effectiveness and reinforces accountability while preserving the core strengths of the existing model.
Against this background, the reform priorities can be grouped around three interlinked themes. First, there is a need to establish a more robust and predictable framework for contracted-out provision – including greater funding stability, structured engagement with providers, and clearer national reference standards that support regional implementation while respecting autonomy. Second, improvements in efficiency and targeting are required to ensure that resources are directed where they add the greatest value, administrative burdens are reduced, and differentiated service pathways are used more strategically. Third, the incentive and accountability architecture can be strengthened so that providers are more consistently rewarded for sustained outcomes, performance information is more transparent, and authorities are better equipped to monitor, evaluate and continuously improve the system.
This chapter describes a new model for contracted-out employment services in Italy structured around these three pillars. It builds on the elements of the current system that function well – including the LEP standards, structured profiling and participant choice – while introducing targeted enhancements designed to reinforce the institutional framework, improve efficiency and targeting, and strengthen incentives and accountability. The proposed approach positions the national level as a facilitator of system improvement, provides regions with practical implementation tools with minimal additional requirements, and embeds flexibility and modularity so that new elements – such as provider ratings or enhanced participant reporting – can initially serve informational purposes, with scope for gradual extension where appropriate.
Delivering contracted-out employment services depends on a stable and coherent framework that provides clarity to all actors involved. Past experiences in Italy show that service continuity, provider engagement, and basic delivery safeguards were not always consistent across regions, contributing to uneven implementation and variable support for jobseekers.
International examples demonstrate that strong contractual systems are characterised by predictable funding flows, structured dialogue with providers, and clearly defined minimum meeting frequencies to ensure that all participants receive a reliable standard of quality support. These are essential features of collaborative models underpinned by trust.
This section presents a set of elements for strengthening the architecture of Italy’s contracted-out model. It begins by considering how stable funding can support service continuity across regions, before proposing a more structured engagement with providers to improve programme design and operational problem-solving. It then sets out proposals for minimum service requirements that safeguard quality across territories. Together, these elements create the underlying framework within which a more effective and consistent system of contracted-out employment services can develop.
3.2.1. Ensuring stable funding and service continuity
Contracted-out employment services are an essential pillar of Italy’s system to connect people with jobs and should therefore be embedded within a structural governance framework alongside public employment services. Rather than functioning as an ad hoc or programme‑specific component, they need to be recognised as a stable and integral part of the system’s overall architecture, embedded in routine operations with clearly defined rules, rather than attached to specific initiatives.
Italy has historically invested modestly in ALMPs compared to other OECD countries, limiting both the scale and the quality of services that could be delivered.1 The GOL programme, however, has shown that well-structured measures, supported by generous funding, can be effective. The recent assessment of Recovery and Resilience Facility measures by the European Commission led to the extension of the GOL programme to 2026, signalling confidence in the approach.2 The proposed new model for contracted-out services presented in this note aims to build on this experience with the GOL programme by consolidating what works and addressing the weaknesses revealed during GOL implementation.
Taking a long-term perspective, however, requires a more strategic approach that ensures the continuity of core functions and standards of contracted-out employment services, beyond any extension of GOL as a programme. Stable and predictable financing is essential to give labour market actors the certainty they need to plan investments, staffing and service capacity over time. Without such stability, short funding cycles and fragmented resources make it difficult for providers and regions to maintain capacity, invest in quality and deliver consistent services.
Although Italy receives substantial support from the ESF+, the PNRR and other EU instruments, these sources do not offer long-term stability, as they are time‑limited and subject to shifting priorities. Reliance on them alone cannot sustain the system in the long run. For this reason, Italy should establish a more predictable and stable, long-term financing framework that co‑ordinates national resources with multi-year European funding (see Box with Proposals 1). Such a framework would guarantee continuity for providers and jobseekers, improve predictability, and ensure efficient use of all available funds.
A key objective should be to guarantee a minimum operational level at all times, ensuring that the system continues functioning, even at reduced scale, between major EU funding cycles. Maintaining this foundational layer prevents the loss of capacity that would otherwise be costly and slow to rebuild and would undermine trust in the system. Within such a framework, national funds could finance baseline services, while EU funds could be used to support specific target groups or pathways and enable scaling up the provision when the need arises. Italy could also explore alternative funding mechanisms, including inter-ministerial initiatives that pool resources across employment, education, social inclusion and skills development to align policy goals across sectors. Crucially, these different funding streams should be well-aligned and strike an appropriate balance between ensuring adequate coverage and avoiding overly complex reporting and administrative requirements associated with managing multiple sources of funding.
To maximise the value of EU as well as national investments, Italy should ensure rigorous and continuous evaluation of contracted-out services (see Section 3.4.4). This will improve programme effectiveness, help sustain political support for long-term funding and promote evidence‑based adjustments.
To support providers in planning and maintaining adequate service capacity, national and regional authorities should offer regular and transparent guidance on expected funding levels over a multi-year horizon. Providing forward-looking budgetary indications, even if subject to adjustment, would enable providers to make informed decisions on staffing, investments and programme planning. This would reduce uncertainty, strengthen financial stability among providers, and contribute to a more resilient service delivery system.
Finally, alongside building a stable financial framework, Italy should streamline and expedite reimbursement procedures for service providers. This includes addressing delays in formalising monitoring and reporting requirements and ensuring that guidance materials are provided in a timely manner, are user-friendly, clear and avoid unnecessary administrative burden. Italy should also set firm deadlines to the local CPIs for settling payments that providers are eligible to receive. Expediting payment processes is crucial to reduce financial uncertainty, support providers’ liquidity, and help safeguard continuity of services.
1.1. Establish stable and predictable long-term financing
Create a multi-year funding framework that co‑ordinates national resources with EU instruments to ensure continuity, support long-term planning, and reduce fragmentation.
1.2. Guarantee a minimum operational level of contracted out employment services
Ensure that core service capacity is maintained at all times, using national funds for baseline provision, modifying target groups based on fund availability, and scaling up with available external funds when needed.
1.3. Ensure continuous evaluation to inform and justify future investment
Conduct rigorous evaluation of contracted-out services to guide adjustments, demonstrate effectiveness and sustain political support for long-term funding.
1.4. Offer providers regular guidance on expected funding levels
Provide transparent multi-year funding indications to help providers plan staffing, investments and service capacity, reducing uncertainty and supporting financial stability.
1.5. Streamline reimbursement and administrative procedures
Simplify and speed up verification and payment processes, supported by clear and user-friendly guidelines on monitoring and reporting to reduce administrative burden and uncertainty.
1.6. Set explicit deadlines for processing and settling payments due to providers
Ensure that regions and CPIs make timely payments to providers, possibly as a condition for the disbursement of subsequent funds.
3.2.2. Strengthening engagement with providers
The current GOL programme has significantly expanded the scale of public-private partnerships in Italy, but its implementation has also revealed systemic weaknesses that a new contracting model will need to resolve. Across regions, despite a consistent approach being outlined in the LEP standards, programme implementation can vary. This can lead to inconsistent involvement of providers in questions of programme design and implementation, and fragmented monitoring arrangements. In several regions, providers have highlighted that changes to rules or procedures are often communicated late in the process, with little opportunity for structured consultation or feedback. This has contributed to variability in delivery quality and in how jobseekers experience services across different territories.
International experience shows that high-performing contracted-out employment systems establish formal, ongoing mechanisms for provider engagement. In many OECD countries, structured dialogue between ministries, PES and provider associations plays a central role in refining programme rules, aligning expectations, and ensuring that operational issues emerging on the ground are addressed early. This engagement not only improves programme stability and transparency but also supports a more collaborative, less transactional relationship between providers and the public authorities. Italy’s proposed new model can draw on these lessons by institutionalising a more systematic approach to consultation and shared problem-solving with providers (see Box with Proposals 2).
Introducing a national Provider Council to facilitate policy implementation
The new model elaborated in this note proposes that the Italian authorities consider establishing a national Provider Council, which would be formally housed within the Ministry of Labour and Social Policies and supported by a dedicated secretariat in Sviluppo Lavoro Italia. The Council would serve as the central forum for structured, ongoing dialogue between national authorities, regions, and provider networks.
Membership would include representatives from the Ministry of Labour and Social Policies, Sviluppo Lavoro Italia, regional authorities and selected providers or provider association(s) from each region. Regions would retain autonomy in determining how they appoint their representative(s), with the expectation that individuals be drawn from either senior management within accredited providers or from regional or sectoral industry bodies. This flexibility acknowledges regional diversity in market structures, while ensuring that representation reflects those with a direct understanding of operational realities. The representatives should be able to reflect any views from their regional networks, and take any issues for consideration from the national Council back to their regional networks for review.
The Council’s terms of reference would cover:
Advising on the creation and structure of new ALMPs;
Discussing ongoing programme design and implementation issues;
Identifying emerging operational barriers and solutions;
Supporting greater alignment and mutual learning across regions; and
Advising on adjustments to accreditation, monitoring, and payment systems.
To institutionalise its role, any proposed changes to programme rules or national guidance would be subject to mandatory consultation with the Council. Meetings would take place on a quarterly basis, with extraordinary meetings convened if needed, for example during major policy transitions.
Despite having a formal role in the system, the Council’s mandate would be limited to an advisory function, meaning that the autonomy of the existing governing bodies in the system of designing and implementing ALMPs would be fully retained. The Council’s role as an advisory, consultative body would also help mitigate conflicts of interest, given that the governing bodies would not be under any obligation to accept proposals put forth by the Council. The composition of the Council would be determined by the Ministry of Labour and Social Policies, balancing the representation to include a diverse set of perspectives from amongst the various providers.
Consider also introducing regional-level provider networks to support the national Council
To support information gathering at the national and regional levels, it would be helpful for each of the regions to have a mechanism in place to provide a structured forum where the providers could express their opinion on the design and implementation of contracted-out programmes within their region. Therefore, it is suggested that regions be required – or, alternatively, strongly encouraged – to put in place a structure that allows provider feedback to be gathered and co‑ordinated. Regions should have freedom as to the exact specification and structure their network takes. For individual providers, participation in the regional-level provider network would be voluntary. However, each region’s solution should include Sviluppo Lavoro Italia as a member, given the institution’s role in co‑ordinating regional delivery. Similar to the national-level provider Council, the provider networks would formally only have an advisory role (although they could put forth specific proposals for consideration by the relevant contracting authority in their region).
2.1. Establish a national Provider Council as the central consultation body
Create a formal national Provider Council housed within the Ministry of Labour and Social Policies and supported by Sviluppo Lavoro Italia, ensuring structured, regular consultation with providers on programme rules, operational issues and system development.
2.2. Introduce mandatory consultation on programme rule changes
Require that all significant modifications to national guidance, payment rules or operational procedures undergo formal consultation with the Provider Council prior to adoption.
2.3. Strengthen regional feedback mechanisms through provider networks
Encourage each region to maintain a structured provider network to gather operational feedback, share implementation issues, and feed these insights to the national Council.
2.4. Support cross-regional learning and alignment
Use the Council to identify delivery challenges, compare regional practices and promote consistent standards, reducing variability in implementation across territories.
3.2.3. Operational safeguards in the form of minimum meeting frequencies
The absence of clearly defined minimum meeting frequencies in GOL creates the possibility of variation in the baseline service level experienced by participants and increases the risk that some participants do not receive the same level or quality of support. Providers and regional authorities indicated that variability in expectations has resulted in uneven quality and weak comparability across regions. This includes variability in client contact frequency, diagnosis of barriers, and personal action planning.
International evidence suggests that minimum standards, when co-designed with regions and providers, help ensure a consistent baseline of service intensity while still allowing flexibility for local adaptation. Countries such as Ireland, Sweden and Canada (Ontario) apply standardised requirements for early engagement, diagnostic profiling, and minimum frequency of advisor contact, which ensure that harder-to-place groups are not “parked” and that service quality is more predictable across jurisdictions.
Introduce minimum meeting frequencies
To address this, the proposed new model would introduce national minimum meeting frequencies, set through a renewed State‑Regions agreement (see Box with Proposals 3). These standards would apply across all contracted providers and ensure a baseline level of service irrespective of geography. They would be loosely based on requirements currently included in most similar types of programmes in other OECD countries.
These new minimum meeting frequency requirements would include:
Mandating that the first meeting with a participant takes place within two weeks of the date when a participant selects a provider (upon referral to the programme by the CPI counsellor),
Mandating at least one monthly meeting with participants assigned to Pathways 1‑3, and
Mandating at least two monthly meetings with participants assigned to Pathway 4 (of which one must be a one‑on-one meeting).
Individuals may be exempt from the last two requirements if they are enrolled in training courses or for other legitimate reasons, which are to be discussed and formally acknowledged during the signing of PSP. Information on training would be recorded within the reporting mandated in the new proposed reporting framework, where responsibility for reporting would be shared between the jobseeker and the provider (see Section 3.3.1 for details).
Each required contact would involve a substantive interaction, such as updating the PSP, verifying job-search activity, providing referrals or taking other concrete steps. Meaningful interaction at every meeting would be ensured by maintaining the content of these meetings as set out in the existing LEP standards. The primary purpose of these meetings is to ensure that jobseekers engage meaningfully with providers, and the proposed minimum frequency of meetings would be in line with requirements in similar models in other countries. Regions would remain free to strengthen standards further, but not to fall below the nationally agreed minimum service requirements. Providers would be monitored on their compliance with these requirements. Those not able to provide the requisite frequency of meetings would need to turn away participants and would be required to provide the CPI with written justification for why they cannot accept the participant. In addition to helping provide service continuity for participants, this requirement could help ensure that providers are not accepting higher levels of participants than they can adequately serve. Providers who are not adhering to the minimum service requirements could be temporarily prevented from accepting new clients (or, in the extreme case, have their clients referred to other providers).
3.1. Set minimum contact frequencies to ensure consistent engagement.
Require at least one meaningful interaction every month for participants of Pathways 1‑3 and twice a month for participants of Pathway 4, with regions free to strengthen (but not reduce) this baseline.
3.2. Allow regional flexibility above the national floor.
Permit regions to add additional intensity or requirements in line with local needs, while maintaining uniform minimum safeguards nationwide.
A well-functioning employment services system must not only be well-designed but also efficient in its operations and responsive to differing participant needs. Evidence from GOL suggests that administrative burdens on providers, uneven resource availability across CPIs, and complex customer journeys may contribute to delays in service delivery and inconsistent support for jobseekers. At the same time, international practice highlights the importance of aligning administrative processes, streamlining referral pathways, and ensuring that frontline staff have the tools and capacity needed to target support effectively.
This section considers several operational improvements aimed at strengthening the efficiency and targeting of the proposed new model. It develops a new set of reporting requirements for providers, and outlines how CPIs can be better resourced to manage their core functions. Utilising customer profiling to streamline and delineate journeys can help match jobseekers to appropriate support more quickly. Finally, it proposes interim outcomes as important guardrails for vulnerable groups, ensuring progress is recognised even when employment may take longer to achieve. Together, these measures aim to reduce friction in the system and ensure that resources are focussed where they can generate the greatest impact.
3.3.1. Simplifying reporting for providers
While conditionality requirements for jobseekers and benefit recipients on participation in ALMPs and active job search are already embedded in the Italian legislation, their implementation may vary substantially across regions, CPIs, and providers. Building on the existing requirements, the new model proposes to introduce a shared reporting responsibility between providers and jobseekers, without altering current conditionality, to encourage participants to take a more proactive role and greater ownership in their activation journey (see Box with Proposals 4). Reporting could also incorporate opportunity for jobseekers to provide feedback on the quality of the services they receive and the performance of providers. Providers, in turn, could be given an option to amend or comment on participants’ reports, enhancing the accuracy and completeness of collected information.
At least in the initial stage, reporting requirements should serve primarily monitoring and management purposes. They should be kept minimal and should not carry legal consequences. Failure to submit a report could trigger an invitation to a follow-up meeting rather than sanctions. In the longer term, and subject to the development of suitable information and communications technology (ICT) systems, the framework could be extended to support more rigorous conditionality. If so, responsibilities for assessing compliance, the use of reported information and the respective roles of CPIs and providers in validation and escalation should be clearly defined.
Reporting requirements should apply only to a subset of jobseekers with the necessary skills (such as adequate digital literacy) and the capacity to meet these obligations effectively. It is essential that the system takes the needs of vulnerable groups into account, as these participants may struggle to see the value of the activities they undertake, find online tools difficult to navigate or perceive reporting as sensitive or burdensome. The identification of jobseekers suitable for reporting requirements could be based on similar profiling tools and digital literacy screening, as described in Section 3.3.2. Vulnerable participants currently assigned to Pathway 4 of the GOL programme should be excluded from these requirements.
To facilitate buy-in and usability of the new reporting system, the format and content of reporting templates could be co-designed with regional authorities, with optional involvement of selected CPIs and providers to integrate their experience and perspective. This process should lead to a creation of standardised national templates that, to the extent possible, accommodate local requirements related to e.g. region-specific procedures, information systems and operational constraints. The national government could support the adoption of these common tools through dedicated funding and technical assistance. Where appropriate, a formal agreement between the state and regions could mandate the use of a national reporting tool for essential core data, while allowing regions to add additional functionalities through their own add-ins.
Reporting templates should be defined in sufficiently general terms to ensure that neither providers nor jobseekers feel compelled to disclose sensitive information. For example, jobseekers could submit monthly summaries of their job-search efforts and participation in ALMPs, including key dates and the names of the programmes in which they have participated. Ensuring that even minimal data is shared by providers and regions, for instance the date of assignment to a provider, activity start/end dates, can substantially improve monitoring and enable better oversight of both providers and participants.
Consider developing a platform or portal for shared data‑exchange
To ensure that reporting is as straightforward and efficient as possible, this proposal recommends developing a shared data‑exchange portal accessible to all direct stakeholders: regions, providers and jobseekers. Although technically demanding, such a portal would substantially reduce the reporting burden on providers, enable consistent data collection across regions, and facilitate national-level monitoring. Particular attention should be paid to ensuring that the portal is user-friendly and easily accessible via mobile phones, recognising that many participants may not have access to a computer.
While such a portal would be extremely useful, it would not be absolutely essential for introducing the new reporting requirements. An alternative interim solution would be to introduce reporting through other means, including email. This approach would be feasible given that the initial system does not entail sanctions and that existing communication channels are to be used for formal exchanges, including those related to benefit conditionality.
Mandate exchange of data on provider-participant assignments between regions and central government
One important challenge in the existing GOL programme relates to establishing employment outcomes for participants, which should be addressed in the future model. Mandating a minimum set of data requirements on participant referrals to specific providers would greatly simplify reporting for providers while offering a host of additional benefits in terms of monitoring outcomes and evaluation. For this reason, the new model would mandate that regions provide, on a monthly basis, the following individual-level information to the national agency identified by the Ministry:
participant ID,
employment services provider ID,
referral date, and
programme pathway.
The monthly exchange of this (relatively simple) information would have several benefits, enabling the following:
Verifying employment outcomes for outcome‑based payments to providers (see Section 3.4.1 for details), with the central government providing individual-level information on employment outcomes to regions on a monthly basis;
Publishing provider performance ratings, with employment outcomes (after accounting for participant characteristics) playing an important role in the quantitative indicators (see Section 3.4.2 for details);
Facilitating the monitoring of providers, based on their performance (see Section 3.4.3 for details); and
Conducting regular and systematic counterfactual impact evaluations (see Section 3.4.4).
Consultations with stakeholders indicate that some regions have already been reporting this information during the implementation of GOL, which should facilitate the introduction of this measure within the new model. However, in order to be truly useful, the reporting must be systematic and cover all programme participants in every region.
4.1. Introduce a shared reporting responsibility for providers and jobseekers
Building on existing conditionality requirements, introduce joint activity reporting by providers and jobseekers, thereby strengthening engagement, ownership and data accuracy, but without necessarily linking reporting to legal consequences.
4.2. Integrate a feedback loop into reporting
Enable (or, optionally, mandate that) jobseekers comment on service quality and provider performance, while enabling providers to amend or comment on participant reports to ensure accuracy.
4.3. Target reporting to suitable participants
Apply reporting requirements only to jobseekers with adequate digital skills and capacity to meet reporting obligations, while safeguarding vulnerable groups.
4.4. Collectively design national reporting templates
Co-design standardised reporting templates with regions (and optionally with CPIs and providers), taking into account local requirements while ensuring a common national core.
4.5. Define reporting templates in sufficiently general terms
Request simple monthly summaries on basic activities to ensure minimum data availability nationwide, while minimising the need to share sensitive information.
4.6. Develop a new portal for jobseekers, providers and CPIs – optional
Create a single, mobile‑friendly platform for all stakeholders to reduce reporting burdens and ensure consistent national monitoring, with the potential to support conditionality in the long term.
4.7. Mandate the exchange of basic, individual-level participant information with central government
Require that regions report monthly on individual-level programme starts that contain provider identifiers which are a precondition for provider ratings, monitoring and evaluation.
3.3.2. Streamlining customer journeys and refocussing CPI work through a digital-first approach
The use of existing profiling tools can help streamline jobseekers’ customer journey by better targeting the support provided and ensuring that CPI counsellors focus their time on those groups 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. The use of the current statistical profiling tools can help identify job-ready participants and direct them to a digital-first pathway, enabling CPI counsellors to concentrate on those requiring more intensive, personalised support, including vulnerable groups and the long-term unemployed (see Box with Proposals 5).
A digital-first pathway would rely on access to self-service job-search tools, such as online vacancy databases, job-matching platforms and other digital resources, which are likely to be sufficient to support the return to employment for this group of jobseekers. A subset of jobseekers currently assigned to Pathway 1 in the GOL programme could tentatively be considered job-ready, using predicted unemployment duration as a more accurate indicator to define this group. Those with an expected relatively short unemployment spell (or alternatively a very low probability of becoming long-term unemployed) can be regarded the most suitable candidates for a digital-first approach.3
The profiling tools currently used within the GOL framework could already serve to identify individuals in this job-ready category. In GOL, the automated quantitative profiling system generates an initial assessment of participants by classifying them into one of three employability risk classes based on estimated probabilities of long-term unemployment derived from a logistic regression model. With appropriate testing and fine‑tuning, these estimates can be used to select individuals who would benefit from a digital-first approach. In addition, it would be important to verify that selected individuals have adequate digital skills, for example by introducing a digital literacy test or a set of screening questions.
The criteria for defining the digital-first group can remain flexible, with parameters that can be adjusted over time and across regions as needed. Once the model is established, the target group for digital-first support can be progressively refined based on emerging evidence and local labour market conditions. At the same time, it is important to allow everyone to opt out of the digital-first pathway if they prefer more personalised face‑to-face counsellor support to account for varying levels of digital literacy and individual preferences.
If individuals selected for the digital-first pathway do not secure employment within the expected timeframe (in this proposal, three months) they would be redirected to the standard (Pathway 1) approach and proactively contacted by a counsellor for a face‑to-face meeting at the CPI. Similarly, an early return or escalation mechanism should be in place where there is evidence of non-progression or non-compliance, to ensure that misclassified participants are not kept in a digital-first track longer than necessary and are promptly returned to the standard pathway. The proposed approach only delays participants’ entry into the standard system (with a possible opt-out and early return options) and therefore reduces the risk of producing any legally binding decisions or significant effects on participants. The process should always be reviewed by a human, with clear and transparent decisions about each participant’s journey. In this context, given the non-binding nature of guidance and the presence of human oversight, additional General Data Protection Regulation (GDPR) related requirements may not be necessary.
5.1. Introduce a digital-first pathway for job-ready participants
Identify jobseekers likely to find work independently and direct them to self-service digital tools (vacancy databases, matching platforms, online resources), enabling CPI counsellors to focus on those groups who need more intensive support.
5.2. Use statistical profiling and digital literacy checks to identify digital-first participants
Use the GOL statistical profiling model refined through testing, alongside a simple digital literacy check, to identify job-ready individuals.
5.3. Keep digital-first eligibility criteria flexible
Allow the criteria for defining the “digital-first” group to be adjusted over time and across regions based on evidence and labour-market conditions.
5.4. Allow opt-out and early return options
Ensure jobseekers can opt out from the digital pathway and access in-person counsellor support, reflecting different levels of digital literacy and personal preferences. Introduce an early return mechanism in cases of non-compliance to promptly redirect participants to the standard pathway.
5.5. Redirect participants to standard pathways if employment is not secured
If jobseekers in the digital-first pathway do not find employment within a reasonable expected timeframe, move them back to the standard pathway and schedule a counsellor meeting.
3.3.3. Introducing interim outcomes as guardrails for the most vulnerable groups
Jobseekers who face more complex barriers to employment may lose motivation if they are placed in lengthy training or other employability-enhancing measures without regular follow-up. To address this, the new model proposes revisiting long placement periods, which risk leaving clients “lost” in the system, by introducing shorter, modular placements for vulnerable groups, with smaller interim outcomes toward employment (see Box with Proposals 6). This approach is similar in spirit to the “participation ladder” used as the basis for payments to providers in the Netherlands. In that system, jobseekers’ gradual progress is tracked from social isolation to engagement in community and training activities and ultimately to autonomous employment (Vodopivec, 2023[1]).
The new model proposes to split up the existing referral of Pathway 4 participants, who are assessed to be requiring the most in-depth assistance. As in the current GOL programme, the participant and provider will have several meetings after the participant is referred to the provider, including meetings constituting the specialised orientation (LEP-E). Following the conclusion of meetings under this activity, the new model foresees a mandatory trilateral meeting between provider, participant and CPI counsellor, with the optional involvement of social services where appropriate. The purpose of this meeting would be to determine whether the participant is making satisfactory progress with the provider after the specialised orientation has been completed – for example, agreeing on a specific training course or other ALMP, for which the participant is motivated and eager to attend. Following this discussion, the CPI counsellor would have the discretion to ascertain whether suitable progress on deciding on specific goals has been made:
In case of a positive assessment of the CPI counsellor, the participant would sign an updated PSP agreeing on updated reintegration goals. These would then serve as the basis for possible outcome‑based payments to the provider (based on the subjective determination of the CPI counsellor – see Section 3.4.1 for details).
In case of a negative assessment of the CPI counsellor, the participant would be given the choice to select another provider, beginning the process anew.
This approach would very broadly mirror the one adopted by Valle d’Aosta in the current GOL programme, where Pathway 4 was divided into two sub-pathways, “Work” and “Inclusion”. While not foreseen in the current proposal, Pathway 4 could be further split into different modular components, with similar periodic check-ins with CPI counsellors to assess progress and establish that a participant should continue to be placed with a given provider.
6.1. Introduce trilateral meetings between participants, providers and CPI counsellors with optional involvement of social services
Replace lengthy placement periods on Pathway 4 with shorter modular steps and interim outcomes toward employment, drawing on the Dutch “participation ladder”.
6.2. Allow CPI counsellors to mandate changing providers if progress is not achieved
If pre‑determined, specific goals (“soft outcomes”) are not achieved within the expected timeframe, give CPI counsellors discretion to require that individual change provider to ensure more effective support.
3.3.4. Ensuring adequate resources for CPIs (optional)
The limited administrative capacity of some CPIs, particularly during periods of increased workload, 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.
The Italian national framework assigns core gatekeeping functions, such as initial intake, provision of information, profiling and signing the PSP, to the public employment service. This may create bottlenecks, as many CPIs are understaffed. Even when CPIs outsource the delivery of other services to private providers, they may still be unable to manage high inflows of jobseekers. Recruiting new counsellors and administrative staff is operationally complex due to the public nature of employment and existing regulatory constraints. An extraordinary staffing plan that has been in place since 2019 remains incomplete in several regions.
Some regions, however, have defined specific legal frameworks that allow them to “delegate” certain gatekeeping activities to private providers. For example, in Lombardy, accredited external providers can serve as initial contact points or gatekeepers for some groups of jobseekers. They perform functions normally carried out by CPIs, such as intake and initial interviews, thereby reducing CPI workload and enabling counsellors to focus on clients with more complex needs. Accredited providers must agree to accept all participants requesting support, deliver the required services, maintain continuous communication with the CPIs and conduct rigorous monitoring. This approach could potentially be considered in other regions if CPI workloads increase, especially in times of crisis or higher inflows of jobseekers (see Box with Proposals 7).
The involvement of external providers can therefore represent an important mechanism to address CPI capacity shortages. However, it is likely to be most effective when applied to a well-defined subset of jobseekers, identified, for example, through initial statistical profiling (see Section 3.4.2) to ensure that the approach is appropriate for the selected group. Crucially, if this solution is used for outreach or intake activities, robust and transparent standards for the provision of these services must be established.
7.1. Delegate selected gatekeeping functions where appropriate
During periods of increased workload, consider delegating certain gatekeeping tasks, such as intake and initial interviews, to accredited providers to help address CPI capacity shortages, following the examples of some Italian regions.
7.2. Apply delegation only to carefully selected groups
Use statistical profiling tools to identify suitable jobseeker groups for provider-led intake, ensuring the approach is appropriate and effective.
7.3. Establish clear standards for delegated functions
If intake activities are contracted out, implement robust, transparent service standards to guarantee consistency, quality and accountability by e.g. requiring accredited providers to maintain regular communication with CPIs, accept all eligible participants, deliver all required services, and conduct rigorous monitoring.
A key objective of the new model is to ensure that contracted-out providers are incentivised, monitored and evaluated in ways that promote good jobs and sustained employment outcomes. Under GOL, the limited weight placed on outcome‑based payments, combined with fragmented monitoring and a lack of systematic evaluation, reduced the ability of authorities to drive performance improvements or identify effective practices. Internationally, more mature contracting systems combine well-designed payment structures, transparent provider information, and regular evaluation processes to create a virtuous cycle of accountability and learning.
This section outlines how Italy can strengthen the incentives and performance framework within its contracted-out model. It begins with proposals to refine payment models so that providers are rewarded for sustained employment and meaningful progress for harder-to-place groups. It then sets out options for improving jobseekers’ ability to make informed choices through enhanced provider information and performance reporting. Finally, it presents a proposal for an impact evaluation ecosystem, including cyclical evaluations managed by a Steering Committee. These elements together create the foundations for a more transparent, accountable and evidence‑driven system.
3.4.1. Refining the payment models
Italy’s current payment model under-utilises outcome‑based incentives. As discussed in Section 1.4.6, the share of payments linked to sustained employment under GOL is relatively low compared with many programmes in other countries, such as Australia, the United Kingdom, and Sweden (Vodopivec, 2023[1]). This limits the incentive for providers to prioritise job quality and sustained labour market attachment. More sophisticated payment regimes abroad place weight on multi-stage outcomes past six months and incorporate rewards for progression for groups furthest from work. A redesigned model for Italy should reflect these practices by strengthening outcome‑based payments and aligning incentives more closely with longer-term job retention (see Box with Proposals 8).
Increasing performance‑related payments related to employment outcomes
To support stronger employment attachment, the new model proposes to rebalance payments to increase the share tied to employment outcomes. It proposes to keep the broad structure as is currently in place, with differential payments across individuals assigned to different pathways, but with the following modifications (for details, see Table A A.1.):
For employment on open-ended contracts and apprenticeship contracts of level I and III, the existing payment would be split into two: half of the payment would be paid at the beginning of employment, and half after seven months if the person is still employed at the same employer. This would ensure that participants remain employed at the same employer past the longest statutory probation period for permanent contracts (which is six months by law, although it can be shorter based on collective agreements).
For employment on fixed-term contracts of 6‑12 months, the existing (relatively low) payment would be retained, but equal amounts would also be paid in addition at 13 months if the individual is still employed at the same employer.
These payments would reward providers for sustained employment outcomes, providing incentives to both find jobs that are likely to be durable as well as possibly provide some in-work support to ensure that individuals stay with the employer, rewarding stability.
Introduce payments for progress towards employment for those with multiple barriers to employment
The proposed new model introduces an important element to strengthen the incentives for providers to deliver high-quality services to the most disadvantaged groups, currently assigned to GOL’s Pathway 4 (Work and Inclusion).
Upon the completion of all services with process-based payments for a given participant, a trilateral meeting between provider, participant and CPI counsellor would be used to determine whether the participant has made sufficient progress towards becoming employed. This assessment would be based on the updated PSP, which would be jointly agreed upon at the prior trilateral meeting (see Section 3.3.2 and Figure 3.2 for details). In case the CPI counsellor deems the progress is sufficient, the provider would be eligible for an additional payment: based on current UCS levels, this additional payment would amount to roughly EUR 400, which would be in addition to the EUR 915 process-based payment they would have received for their work with the client.
Modifying the payment model to incentivise training participation and completion
Under the current GOL implementation, in some regions only a small share of these participants begin or complete training, with many finding employment before training starts or experiencing delays in accessing it. For example, analysis of training completion in Emilia-Romagna shows that completion rates for Pathways 2 (Upskilling) and 3 (Reskilling) are roughly 20% and 15%, respectively. In this context, initiating and completing a training course could already be considered meaningful interim results for many participants.
In order to improve the incentives for engaging in training, the revised payment model would split the current payments associated with the specialised orientation (LEP-E) during which providers help participants decide on a suitable training course. Currently, providers are reimbursed with process-based payments for either 6 hours of counselling (for Pathways 1 and 2) or 10 hours of counselling (for Pathways 3 and 4). This would increase to 9 and 15 hours, respectively, for which providers would theoretically be reimbursed. However, the split would now be as follows:
3 hours would be reimbursed based on process-based payments (5 hours for Pathways 3 and 4),
3 hours would be reimbursed upon a participant’s enrolment in training (5 hours for Pathways 3 and 4), and
3 hours would be reimbursed upon a participant’s successful completion of training (5 hours for Pathways 3 and 4).
The revised system would help address the challenge of the low take‑up rate of training.
Harness the improved data exchange to help automate outcome‑based payments to providers
Based on the improved data exchange proposed under Section 3.3.1, the process of making payments to providers that are based on employment outcomes could be considerably simplified. Monthly information on the employment outcomes, taking into account the precise parameters necessary to trigger payments (as detailed in Table A A.1), could be shared with regions to facilitate their administration of payment to providers. This would help ensure that the payments are more accurate and timely. Initially, such data exchanges could be used for automating the data exchange process also at the level of individual regions.
Optional: Incentivising high-performance with additional bonuses
In addition to payments related to the overall outcomes, Italy should consider whether it is feasible in the new model to use further incentives for providers to drive up performance. In systems with multiple providers, best practice from the OECD sees countries using information about relative performance of providers to provide additional payments to those providers that perform well in terms of securing sustainable employment to participants. As Italy implements better information about the performance of its providers into the new system, this opens up the possibility to reward those doing an outstanding job for their participants.
Italy could make use of information provided by a rating system (see proposals in Section 3.4.3 for more details) to deliver additional performance‑related pay to providers. Broadly, two options exist for performance‑related bonuses – bonuses that are paid if a provider outperforms their own expectations, or bonuses that are paid to the top providers, relative to other providers. For example, providers could be paid if they exceeded their own expectations by a certain threshold (using information from the statistical model to determine performance thresholds), or providers could be ranked on their performance and the top X per cent awarded performance bonuses. The advantage of using the rating model to determine performance is that it can adjust for local labour market characteristics and differences in participant profiles between providers to allow valid comparisons between them. It is suggested to first let any new rating system bed in, before evaluating whether and how to introduce performance‑related payments at the provider level.
8.1. Reward sustained employment outcomes for permanent contracts
For employment on open-ended contracts and apprenticeship contracts of level I and III, the existing payment would be split into two: half of the payment would be paid at the beginning of employment, and half after seven months if the person is still employed at same employer.
8.2. Introduce additional payments for extensions of fixed-term contracts
For employment on fixed-term contracts of 6‑12 months, the existing (relatively low) payment would be retained, but equal amounts would also be paid in addition at 13 months if the individual is still employed at the same employer.
8.3. Introduce structured progress payments for harder-to-place groups
Reward verified interim steps for Pathway 4 participants relative to pre‑agreed, specific targets (“soft outcomes”), ensuring providers remain incentivised to support those furthest from work.
8.4. Introduce payments for training enrolment and completion
To incentivise training participation and completion, modify the payment model to give providers incentives to ensure participants complete training programmes.
8.5. Use improved data exchange to help automate outcome‑based payments to providers
Data provided by the central government on participant employment outcomes could be used as a basis for outcome‑based payments.
8.6. Include an optional performance bonus for top-performing providers
Once performance ratings are established, allow regions to offer additional bonuses to the best-performing providers (for example, those exceeding their expected performance by one standard deviation, or the top 10% of providers) based on risk-adjusted results.
8.7. Ensure transparent documentation of payment eligibility thresholds
Publish clear criteria for outcome verification, data requirements and acceptable evidence to ensure consistent regional implementation and prevent administrative disputes.
3.4.2. Improving jobseekers’ ability to make a deliberate and informed choice
Italy has a multitude of providers, offering jobseekers a wide range of potential service providers to choose from. However, there is currently no way for jobseekers to access data that would meaningfully allow jobseekers to assess relative provider performance in advance of making their choice on who to partner with. Other OECD countries with multiple providers endeavour to provide jobseekers with information to make better choices. Australia’s provider rating system, Sweden’s quality dashboards, and the Dutch provider profiles show that transparent and regularly updated performance monitoring systems build accountability and permit jobseekers to make better informed provider choices.
There are different ways to summarise information on provider performance. Both Australia and Sweden use some form of data analysis to determine how well a provider is rating relative to its expected performance. This means a provider with harder-to-place jobseekers, or operating in more challenging local labour markets, would not be expected to generate as many successful outcomes as those providers with more job-ready participants, or in more dynamic labour markets. Australia augments this statistical analysis through the direct incorporation of participant feedback into their rating metric, whilst Sweden aims to offer such assessment alongside its statistical rating. Providing these data into easily accessible public websites then allows jobseekers to filter providers by criteria that allow them to determine which providers are best placed to meet their individual needs.
Introducing a national portal on provider performance
The proposed new model (see Box with Proposals 9.) would establish a national provider information portal, which would publish comparable information on all accredited providers. This would utilise a range of different information to provide a rounded view on provider performance which will help jobseekers to make choices that depend upon the prospective provider’s ability to support them into work.
This approach aligns with practices in countries such as Australia and Sweden, which use publicly available performance information to support informed participant choice and generate competitive pressure within the market.
Quantitative data should be used to calculate relative provider performance
Italy can use information on jobseekers’ characteristics and local labour market performance, to determine the statistical performance of different providers. Following the regression-based methods used in Australia and Sweden, this would involve a regression which would show provider performance in terms of employment outcomes after accounting for the composition of their individual clients. This would leverage the detailed administrative data available at INPS, including information on past employment and employment spells, wages, unemployment benefits, employer characteristics, as well as socio-demographic data and information on geography.
Performing this assessment at the national level has several advantages. It enables a consistent application of statistical modelling across all regions. This is particularly helpful for providers who work across multiple jurisdictions. It also facilitates comparisons and learning across regions, and standardises processes. This reduces administrative burdens and/or duplication, as the same processes are applied to all jobseekers, regardless of where in Italy they reside.
Including regional information in the estimation will allow performance indicators to be calculated for individual regions, enabling regions to manage their local providers more effectively, and allowing jobseekers to compare performance of providers within their area. Once a performance calculation is made, these data can then be used to provide information on performance of providers, relative to their peers.
Qualitative data should be used to supplement and contextualise performance
Alongside the quantitative calculations made, Italy should implement a feedback system to incorporate participant feedback into provider assessments. The introduction of a post-placement participant survey, with some scored questions on performance, will allow a numerical scale to be produced that can add the views of participants to the overall assessment of providers. Some countries have gone further, for example, Sweden is introducing the ability for text-based comments from users to appear, providing the possibility of more contextual feedback from participants. The recent advances with large language models also offer the possibility to develop this further and to synthesise participants qualitative assessments into an abridged summary, that whilst not “scored” per se, might allow participants to determine a more sophisticated qualitative understanding of performance.
The new portal should summarise information into an accessible format
A national performance information portal would bring Italy in line with OECD best practice, and support both PES counsellors and jobseekers in making better-informed decisions. The portal should include:
Performance metrics based on sustained employment outcomes, progression indicators, and service quality measures;
A statistical adjustment using existing profiling data to estimate participants’ risk of long-term unemployment, enabling the publication of risk-adjusted performance scores;
A quantitative summary of survey information from participants post-participation, and
Provider-supplied information outlining their services, specialisations, locations, and contact details.
9.1. Develop a national provider performance portal to support informed choice
Create a national online portal, offering comparable data on all accredited providers, including performance results, specialisations and service locations.
9.2. Publish risk-adjusted performance metrics based on objective data
Use profiling information and local labour market trends to calculate expected versus actual provider outcomes, ensuring fair comparison across providers working with different client groups and regions.
9.3. Integrate participant feedback into provider assessments
Introduce a standardised post-placement participant survey feeding into the published performance indicators, providing jobseekers with insights into service quality and client experience.
9.4. Ensure information is user-friendly and accessible
Present key performance information clearly through summary indicators, visual dashboards and simplified explanations to support jobseekers of all literacy and digital skill levels. Enable users to sort providers by performance scores, specialisms, geographic proximity, client group focus and employer engagement strength, supporting personalised and evidence‑based choice.
9.5. Provide space for qualitative insights and contextual information
Include provider-supplied descriptions of their methodologies and strengths, alongside synthesised qualitative feedback from jobseekers, offering a more rounded understanding of provider performance.
9.6. Use the portal as a tool for continuous improvement and accountability
Ensure counsellors, regions and national authorities use the portal to support better referrals, identify high-performing practices, and increase transparency across the system.
3.4.3. Strengthening monitoring of providers
Robust monitoring and transparent reporting (see Proposals 4) are essential to maintaining accountability, supporting client choice, and managing provider performance. Currently, the absence of transparent and comparable information on provider performance in Italy impedes informed choice, restricts benchmarking, and hinders national and regional oversight of providers. Introducing provider ratings that combine qualitative and quantitative metrics can help improve transparency in the market (see Box on Proposals 10.). In addition, expanding the toolkit of mechanisms available to manage provider performance can help Italy drive up provider performance. Italy can build on the good platform provided by the SIISL to strengthen its monitoring of providers further by incorporating information that will be developed on relative provider assessments into its performance dashboards.
Using ratings to monitor providers can permit better risk-based interventions
Once Italy has established a system of provider ratings, it can use these to manage provider performance and audits. Provider ratings which assign quantitative scores permit the ability for individual or relative risk-based management. For example, individual intervention criteria would allow authorities to intervene if a provider’s performance was below a certain percentage of its expectation. This could in theory mean that authorities could judge 100% or 0% of providers as requiring intervention, dependent on whether or not they outperformed their own individual performance expectation. Alternatively, providers could be benchmarked relative to other providers. In this case, the bottom X percentage of providers, when ranked by their performance, could be performance audited. This has the advantage of driving continuous improvements but also means providers could be subject to scrutiny even where they meet their own pre‑defined individual expectations. Relative performance could be evaluated at the national level, or within regions. A hybrid option could rank providers, but only intervene if those below the set threshold (relative to peers) are also under-performing, relative to their own expectations.
The model proposes a blended approach that addresses these issues and balances minimum performance standards with relative performance improvements. Rather than a strict threshold for e.g. the “bottom 10%” in each region, the model will apply performance criteria that consider both relative and absolute performance. For instance, providers falling in the lowest tier nationally or under a fixed benchmark will be targeted for improvement actions. This blended approach prevents cases where high-performing regions still have a “bottom 10%” despite strong results, or low-performing regions overlook providers that, in absolute terms, need improvement. This ensures consistent oversight for similar performance levels across all regions.
Initial performance monitoring could comprise of communications on performance expectations, provider assessments of delivery and implementation issues, and a system of further written warnings if performance does not improve. Once the system is fully embedded, Italy could consider whether ratings and performance management could be used as a tool for setting temporary caps on provider participant numbers (for example, look at average flows to providers in the previous six months and reducing this by one‑third) whilst the providers implement improvement plans. Sweden is currently taking such an approach by using its performance rating distribution to automatically de‑license providers in the bottom 10% of the distribution. It retains the legal authority to conduct such exercises every six months, with flexibility to do fewer removals if it judges the system to be performing well. Sweden is also considering introducing additional payments for the top performing providers while lowering the baseline, fee‑for-service payments.
10.1. Introduce a national provider rating system to support transparent oversight
Develop a risk-adjusted rating framework using employment outcomes, progression indicators and participant feedback to generate comparable performance assessments across all providers.
10.2. Use performance ratings to inform proportional, risk-based monitoring
Allow regions to prioritise monitoring and audits based on rating levels, focussing resources on providers displaying weaker results or early warning signs of delivery problems.
10.3. Implement a graduated performance improvement process
Introduce stepped interventions including performance reviews, improvement plans and written warnings, escalating only when providers fail to improve within defined timeframes.
10.4. Use ratings to temporarily cap participant flows where necessary
Allow regions to temporarily reduce referrals to consistently underperforming providers while improvement plans are implemented, helping safeguard service quality for jobseekers.
10.5. Integrate monitoring tools into existing digital systems (e.g. SIISL)
Embed performance dashboards and risk signals within SIISL to support real-time monitoring and enable both national and regional authorities to track provider performance consistently.
10.6. Promote transparency through regular publication of performance insights
Publish aggregated findings from provider monitoring, highlighting common implementation challenges and successful practices to support learning and improve market functioning
3.4.4. Developing a counterfactual impact evaluation ecosystem
While evaluations are being conducted with increasing frequency and scope, Italy does not yet have a systematic, multi-year cycle for counterfactual impact evaluations. Counterfactual evaluations enable assessment of the impacts of ALMPs on participants, after accounting for what would have occurred had they not participated (the counterfactual outcomes). For example, how does a training programme impact how quickly a person finds work, and how much they earn once in work? Having a systematic programme of work to complete such analyses will enable Italy to precisely identify the effects of specific ALMPs – i.e. what works for whom. Given additional administrative data relating to costs, it could also potentially help identify the costs and benefits of any new contracted-out employment model, including the relative costs and benefits of different pathways, and for different groups of individuals (for example young and old, or males and females). This learning will then allow Italy to determine how the programme should evolve to better serve the needs of Italian jobseekers.
Lessons from other OECD systems highlight that countries with mature ALMP systems institutionalise evaluation through dedicated steering bodies, regular evaluation cycles, and published findings that feed directly back into programme redesign. For example, Canada, which has a similarly de‑centralised ALMP delivery model has a federal mandate for its regional jurisdictions to conduct regular impact evaluations of ALMPs. Provinces and territories are then given the option to join together and allow central analytical services in its employment ministry to conduct analysis on their behalf – 12 out of 13 jurisdictions choose this option. This has the benefit of economies of scale, which allows the employment ministry to coalesce analytical resources and replicate analysis using linked administrative data. Embedding a similar approach in Italy would support continuous improvement and allow the new model to evolve on the basis of robust, independent evidence.
Cyclical evaluation cycles managed by a Steering Committee
To strengthen evidence‑informed policymaking, the new model proposes a structured framework for continuous counterfactual impact evaluation (see Box on Proposals 11). A new legal instrument should be introduced to mandate “regular” impact evaluation by the regions.
Funding should be made available by central government for regions to conduct this ongoing evaluation. Funding should be set at a level that would enable a central analytical team to conduct counterfactual evaluation across all regions using administrative data. Regions would then be free to dispense this funding in the joint analytical evaluation, or use their proportion to fund their own separate research programme. The advantage of joint evaluation is that it reduces overlap compared to regions conducting evaluation separately, and ensures fully comparable results. This comes at the expense of less flexibility to consider specific regional needs (though this can be mitigated somewhat by collective evaluation objective setting, and ongoing cycles of evaluation to progressively build evidence).
Separately, for those regions willing to participate in a joint evaluation, an Evaluation Steering Committee should be established, comprising representatives from the regions, the Ministry of Labour and Social Policies, employer confederations, trade unions, and provider associations. This mirrors international good practice in governance of evaluation systems and ensures that methodological decisions are collectively owned. This Steering Committee should determine the duration of the regular cycle of evaluation.
Programmes would be evaluated on a cyclical basis, with the cycle length determined by the Committee – either aligned with the funding cycle or through a mutually agreed timeline. All evaluation results would be published, reinforcing transparency and accountability.
Regions would have flexibility to:
Conduct their own evaluations, or
Join multi-region consortia that delegate evaluation to a third party.
The Steering Committee would set the evaluation work programme for each cycle, identifying the core policy questions, priority thematic areas (for example, outside of core impact assessment this could be employer engagement, quality of training, effects on vulnerable groups etc), and appropriate methodologies. This ensures evaluations are not mechanical exercises, but meaningful inquiries aligned with policy needs and implementation challenges.
Anticipating programme participation-related data needs to facilitate the evaluations
One important aspect of the new model would be to establish the data that regions would be required to maintain in order to facilitate the subsequent evaluations. Accurately identifying the causal impact of programme participation requires information on certain important parameters that would naturally be acquired by CPIs during the course of the programme’s administration – the key point would be to ensure that this information would be collected and suitably stored.
An additional benefit of moving to the digital-first approach outlined above (see Proposals 5) is the creation of a natural setting to evaluate the effectiveness of in-person support. By shifting job-ready participants to self-service channels, it would become possible to compare outcomes between those who receive in-person assistance and those who rely primarily on self-service digital tools. For example, the outcomes of individuals who fall just above or below the eligibility threshold for self-service could be compared with those referred into self-service tools to identify differences in results (using a so-called Regression Discontinuity Design, RDD). Alternatively, this could be done by randomly varying assignment to the digital-first approach for some individuals close to the eligibility threshold for self-service. Conducting such evaluations would generate robust evidence on the effectiveness of the offered services and support the development of an impact evaluation ecosystem.
To support the evaluation strategy outlined above, requirements on individual-level information include the following:
Mandatory information:
Profiling scores (initial and final)
GOL programmes assigned and attended by each jobseeker
Optional information:
Information used to complete qualitative assessment (barriers to employment, digital literacy, etc.)
ID of employment services and/or training provider.
Such information could then be merged with the rich and detailed information within the full administrative archive – on e.g. employment spells, wages, unemployment benefits, employer characteristics.
11.1. Establish a national Evaluation Steering Committee
Create a multi-stakeholder committee (Ministry of Labour and Social Policies, INAPP, SLI, regions, providers, unions, employers) to define evaluation priorities, approve methodologies and oversee implementation cycles.
11.2. Mandate cyclical counterfactual evaluations for all regions
Introduce a legal requirement for regions to participate in regular impact evaluations, either by conducting their own assessments or by joining national multi-region consortia.
11.3. Develop a multi-year funded evaluation cycle aligned with programme phases
Adopt a predictable evaluation calendar (e.g. every 2‑3 years) covering core programme impacts, thematic priorities and implementation questions. Assign funding for regions to conduct analysis.
11.4. Standardise key data requirements to enable robust evaluation
Require all regions and CPIs to record and archive essential participation data (profiling scores, programme assignments, activity records) for linkage with INPS administrative datasets.
11.5. Use evaluations to subsequently rigorously examine various aspects of the programme design
Leverage embedded aspects of the programme, such as strict eligibility thresholds, to conduct rigorous counterfactual impact evaluations and tweak key programme parameters or inform subsequent changes to customer journeys.
11.6. Publish evaluation findings to strengthen accountability
Commit in advance to make all completed evaluations publicly available to inform policy refinement, support transparency and build an evidence base for future system reforms.
11.7. Promote methodological diversity and high-quality analytical standards
Encourage use of randomised or quasi‑experimental approaches where feasible (e.g. RDD through digital-first thresholds), while allowing flexibility for regions with smaller sample sizes.
As noted in the discussion in the preceding sections, the proposed new model would retain the key positive features of the existing national legal framework, as well as its current implementation within e.g. GOL, while introducing some new elements such as mechanisms for strengthening accountability. This section brings together various elements of the proposed new model in several diagrams. This includes a diagram on the proposed introduction of a digital-first customer journey, the customer journey within the new model (for the subset of clients referred to outside providers), and the proposed changes to the payment model. The diagrams are intended to complement the detailed proposals provided in the preceding sections.
Figure 3.1 shows the first steps of a jobseeker’s customer journey in the new model. Jobseekers assessed as job-ready are routed to online self-service tools, freeing counsellors to focus on those needing intensive in-person support. Such an approach follow the examples of countries such as Australia and Sweden, which have adopted this type of approach (see Section 2.4.5 in Chapter 2 for more detail on Australia).
Customer journey of jobseekers from registration to programme participation in the proposed new model
Note: The precise thresholds for referrals to self-service tools could be set by the regions, possibly within a range of parameters set at a central level. For individuals not receiving income supports, referrals to the proposed new model would also be contingent on the participant being suitably motivated to participate in the programme.
Figure 3.2 shows the customer journey for those individuals who are referred into the proposed new model. An important point to emphasise is that many of the individuals referred to Pathway 1 of GOL (work reintegration) would not be referred to the new proposed model of contracted-out employment services, but rather would use self-service tools. This could decrease the number of referred individuals considerably: in the GOL programme, roughly half of participants were assigned into Pathway 1.
Customer journey of jobseekers once they are referred into the new programme
Note: Only a subset of jobseekers would be referred to contracted services in the new model – see Figure 3.1 and Proposals 5 for details.
Figure 3.3 illustrates the revamped payment system. Under the new design, outcome payments are split into multiple instalments: for example, a portion is paid upon job placement and the rest only after sustained employment (such as seven or 13 months). The model also adds incentives for milestones like training enrolment and completion. Many OECD programmes reward sustained outcomes beyond six months, which underpinning the choice of splitting payments at the 6‑7‑month mark. Similarly, models in countries such as Sweden reward payment for enrolment in education – motivating a similar type of payment in Italy. The shift to a higher outcome‑based payment share is supported by comparisons to models in countries like the United Kingdom and Australia, where outcome payments form a larger portion of provider compensation (Vodopivec, 2023[1]).
Payments according to services stipulated in GOL
Note: Each block refers to one (or more) services as stipulated in the LEP, the minimum set of services and standards mandated by GOL.
Details on the proposed payment model are included in Table A A.1. The changes to the payment model are intended to slightly increase the share of payment for outcomes. The existing model in Italy places relatively little weight on outcomes – roughly 20‑30% – compared to roughly half in similar models in other OECD countries (Vodopivec, 2023[1]).
Several important aspects on what is not included in the new model proposal are worth highlighting:
Specifications on employment services or training accreditation requirements. Harmonising accreditation requirements could arguably improve the market for these services in Italy by simplifying operations for providers wishing to operate across different regions and helping lower barriers to entry, thus facilitating competition. However, Italy’s devolved governance structure means that accreditation requirements is in the legal jurisdiction of the regions. As such, this proposal does not propose any specific harmonised accreditation requirements. The proposed new model does propose new mechanisms to facilitate the introduction of more harmonised accreditation requirements, by introducing formal consultation mechanisms for providers to be involved in programme design, both at the national and regional levels.
Exact prices for specific services or outcomes. The proposed model does not propose specific levels of prices (the exact numbers specified presume the levels in place in January 2025, for reference, but would need to be updated). These prices are currently outlined in the LEP standards – most notably, by setting the hourly rate for individual counselling sessions (currently set at EUR 39.94). The model does propose slightly tweaking some of the maximum number of hours different services and introducing some new outcome‑based fees (with the amounts linked to the hourly rates of individual counselling sessions). The goal is to slightly increase the outcome‑based share of payments to strengthen financial incentives for providers.
Reference
[1] 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. For example, in 2019 pre‑COVID spending on ALMPs was 0.27% of GDP in Italy versus 0.63% on average across the OECD, with the gap narrowing in subsequent years (OECD Data Explorer Labour Market Programmes).
← 2. European Commission COM(2025) 212 and COM(2025) 310.
← 3. Evidence from some regions in Italy suggest such clients represent sizable proportion of the CPI’s caseloads: in Trento, for example, according to statistics on GOL cohorts from programme start to 30 April 2024, 38% of participants became employed within 3 months. While this statistic cannot be generalised to other regions in Italy and itself at least partly reflects the effects of GOL participation, it does suggest that there is potentially scope for reducing the administrative burden involved in helping such job-ready individuals in Italy.