This chapter explores the strategic importance of engaging with jobseekers as a central responsibility of the Bulgarian National Employment Agency (NEA), positioning it as the most significant domain within its remit. The NEA presents several strengths in this area, with a well-established foundation of services and practices aimed at supporting jobseekers. This chapter sets out a vision for the NEA to pursue: a modern, inclusive Public Employment Service (PES) that reaches, engages, and empowers all jobseekers through personalised, data-driven, and results-oriented support. This approach, which the NEA should aim to develop, seeks to improve placement rates, job readiness, and long-term outcomes, especially for disadvantaged population groups. The ultimate goal is to create a responsive, agile, and client-centred employment service system that empowers each and every jobseeker to navigate and succeed on their journey to sustainable employment.
Optimising Processes and Services at Bulgaria’s National Employment Agency
Abstract
The journey of the jobseeker at the NEA towards employment comprises several stages (Figure 3.1). Each stage is associated with specific operational processes and services aimed at supporting jobseekers as they navigate challenges, build skills, and secure work opportunities. A thorough assessment of the NEA’s current system indicated that throughout this journey, there are numerous opportunities for improvement.
Source: Authors’ compilation.
This chapter is organised into six sections. The first section provides an overview of the NEA’s current engagement with jobseekers, outlining both its key strengths and challenges in serving this core client group. The remaining five sections focus on the different stages of the jobseeker journey, providing targeted recommendations to enhance the overall experience and outcomes for jobseekers registered with the NEA. The second section focuses on outreach, the third section on intake and registration, the fourth on meeting with job counsellors, profiling and assessment of needs and conditions and the fifth on referrals and follow-up actions. The sixth section closes the chapter with a discussion on issues related to service provision. While this chapter focuses mostly on operational processes, a thorough discussion about digital and AI-based tools and platforms that can support these processes is in Chapter 5 of this report.
Engagement with individuals navigating the labour market is the most substantial area falling under NEA’s responsibility. NEA clients are a broad and diverse range of individuals; not only jobseekers looking for work at different intensities, but also employees who may be considering a career change or are at risk of job loss (Table 3.1). The support by the NEA to its registered population extends to anyone navigating the labour market, whether they are unemployed, underemployed, seeking (better) opportunities, or needing to upskill in response to changing job demands. Particular emphasis is placed on the most vulnerable populations, including youth, the long-term unemployed, persons with disabilities, older workers, migrants, ethnic minorities, individuals with low levels of education or skills and all those facing the greatest barriers to employment.
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata.
The NEA’s approach to client engagement and support already rests on a well-established foundation. Recent developments mark notable improvements in both approach and delivery. Positive changes include enhanced outreach efforts, the personalised touch and approach during individual counselling sessions, the introduction of a promising fast-track process for well-matched candidates, the implementation of standardised service packages for vulnerable groups as well as consistent enforcement of sanctions for non-compliance with the next steps agreed in the Action Plan. However, challenges persist throughout various stages of the jobseeker journey (Table 3.2). Addressing these will be essential for NEA to evolve into a modern, inclusive PES. This implies that the NEA proactively reaches, engages, and empowers every jobseeker through more personalised, data-driven, and outcome‑focused support. It will ultimately lead to higher placement rates, improved job readiness, stronger and sustainable long-term outcomes for diverse jobseeker groups.
Source: Authors’ compilation.
3.2.1. NEA’s outreach initiatives focus on raising awareness and encouraging registration, with tailored efforts specifically designed to engage inactive individuals
The NEA’s outreach initiatives primarily focus on disseminating information to raise awareness and engagement of unemployed individuals – including the inactive groups that are typically harder to reach – and encourage registration with the NEA. These efforts are usually conducted in close collaboration with key partners, such as the MLSP, the Social Assistance Agency (SAA), local municipalities, NGOs and schools.
The NEA has 106 local labour offices across the country, offering detailed information about the services available and the benefits of registering with the agency. To ensure accessibility of services and resources to all, the NEA has established, in addition to these local offices, mobile labour office workplaces as well as mobile labour offices in remote and underserved areas (OECD, 2022[1]). These mobile offices provide the main services of the local offices, including jobseeker registration, job placement and matching, and information on ALMPs. The mobile teams travel to various regions, where temporary service points have been set up in community centres, municipal buildings, or other public locations.
Furthermore, the NEA employs specialised youth and Roma mediators who focus on engaging inactive individuals from these target groups and encouraging them to register with the NEA. The NEA also provides the “Family Labour Consultant” service to work with families consisting of more than one inactive family member, organises job fairs and information campaigns and runs joint centres with the SAA to reach the most vulnerable groups (OECD, 2022[1]). The agency also operates 14 “activation clubs” located outside its labour offices in regions with high concentrations of inactive people.
In addition, the NEA and the SAA jointly operate Centres for Employment and Social Assistance (CESAs) to streamline access to services and offer comprehensive support through “one‑stop-shops”. This helps activate inactive individuals who wish to claim social assistance. In addition to data sharing and managing joint centres, the two agencies work together to organise meetings and information campaigns aimed at raising awareness of NEA services and improving access for those in need of support.
Information regarding the services can also be assessed through the NEA’s website. The website features in addition a dedicated “e‑Labour office” space, offering direct access to job vacancies posted by employers.1 What is more, the NEA has a YouTube channel as well as accounts on Facebook and LinkedIn, organises events and visits to schools to engage with students, and distributes informational leaflets. Some local labour offices also disseminate information through their Viber accounts.
Despite these and other outreach efforts, a significant share of inactive individuals remains beyond the agency’s reach. In 2023, only 1 in 5 unemployed were registered with the NEA, while a significant share of those without a job remained out of touch (Figure 3.2). Although their use has increased recently, non-traditional outreach channels, particularly digital tools, remain underutilised, highlighting the need for more diverse communication channels, content, and messaging, to effectively engage different demographic groups. Additional challenges include the absence of clear, measurable outreach goals for each target group, as well as the lack of systematic assessment of outreach plans and activities that may hinder the NEA’s ability to adapt and improve outreach and communication efforts based on evidence. The NEA as also improved data-sharing with other public institutions, including the National Revenue Agency, the Ministry of Education and Science, the General Labour Inspectorate, and the SAA (Republic of Bulgaria, 2023[2]). Along with recent amendments to the Employment Promotion Act and the NEA is in a solid position to address these gaps and enhance its outreach efforts in a strategic way.
Registered unemployed as a share of 25‑64 year‑olds who do not work
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata and EU-LFS.
3.2.2. The registration process at the NEA still relies heavily on in-person document submissions and remains largely manual
Intake and registration at the NEA take place in parallel.2 Registrations with the NEA can be made either in person or remotely. However, the latter option is rarely used yet as most jobseekers do not possess the electronic equipment or digital skills to register online or prefer personal interaction with the NEA staff. To register with the NEA, jobseekers must provide – either in person or by mail/post – several supporting documents, including an ID, educational and qualification certificates, a labour booklet (if available), medical disability certificates (if applicable), and any documents related to employment termination. These documents – with the exception of the ID card – are necessary in case the information cannot be directly retrieved from existing registers (Figure 3.3, left panel). To allow for flexibility, the NEA permits jobseekers to submit required documents gradually after their initial registration, ensuring they can complete the process even if they do not have all the necessary paperwork at the time of registration. Upon registration, jobseekers receive general information about the services provided by the NEA and the responsibilities associated with their registration. Jobseekers can also apply for unemployment benefits at the NEA, even though the NSSI is responsible for such payments.3
During the registration process, NEA counsellors enter basic information about the jobseeker – such as the unique citizenship number and contact details – in the National Database (NDB). Whenever possible, counsellors also collect additional information such as educational information, basic information about knowledge and skills, self-declared ethnicity, marital status and work experience. In case of high workloads, counsellors can enter this additional information during a subsequent meeting. While most of the data must be entered manually by counsellors, some parts of the data – such as birth dates and personal identifiers and previous employment records – can be retrieved automatically by accessing RegiX, Bulgaria’s inter-register exchange system (Figure 3.3, left panel).
Note: AP: Action Plan, NDB: National Database, NEA: Bulgarian National Employment Agency.
Source: Authors’ compilation.
Although the NEA has streamlined some aspects of the registration process, the whole process still remains largely manual and paper based. As a result, the administrative burden for both jobseekers and NEA frontline staff remains significant and well above peer PES that have established automated and largely paperless registration systems. As part of its modernisation and digitalisation initiatives under Bulgaria’s Recovery and Resilience Plan (RRP), the NEA is planning to upgrade its IT infrastructure and improve its data management, enabling data exchanges and ensuring system interoperability across various databases and platforms, starting with the NDB.4 This transformation should streamline the client experience, reduce processing times and error during data entry, as well as alleviate administrative burden on NEA staff, freeing up time for core services (see also Chapter 5).
3.2.3. Despite NEA’s personalised approach, jobseekers who require the most support may not be receiving it at the necessary intensity
Counselling is at the core of NEA services. Within one month from registration, jobseekers have their first meeting with a job counsellor. In the case of youth, a meeting with a counsellor occurs directly on registration. Currently, counselling is only offered in person at the NEA local or mobile offices. Even though counselling is mostly offered by general counsellors, some local labour offices have specialised teams working with vulnerable groups. In most cases, at the end of each appointment, jobseekers receive a follow-up appointment.
The initial meeting with the NEA counsellor involves completing several procedures and entering additional data into the NEA’s systems. Any missing personal information from the registration – such as work experience, education, marital status, or health status – is manually added to the NDB. A jobseeker profile is then created using the data collected during registration and the initial meeting with the counsellor. This profile is further refined through a skills assessment interview and the counsellor’s observations regarding the jobseeker’s behaviour during the meeting. The information collected through the skills assessment interview and the counsellor’s observations encompasses skills (technical skills, theoretical skills, creative skills, social skills, business skills, organisational skills and basic skills), behaviour (motivation, initiative, discipline) and the person’s physical condition (work ability, physical appearance). All this additional information and data are entered in the NDB by the counsellor during the first meeting (Figure 3.3, right panel).
The collection of all the necessary information at different times is both time‑consuming and challenging for counsellors, as they have to manually enter the information in the NDB – either in the form of free text or through drop-down fields – while carrying out the interview with the jobseekers. In parallel, they must assess the competences and motivation of jobseekers based on responses to questions without dedicated tools. The absence of dedicated tools means that the assessment process relies heavily on the subjective judgment of counsellors, making assessments less systematic and more dependent on the individual counsellor’s experience and interpretation. This can result in inconsistencies in how jobseekers are assessed, thereby affecting the coherence of the services provided across different counsellors and local offices. In addition, the dual task of manual data entry and assessment by counsellors at the same time increases the risk of errors and incomplete profiles, which can further impact the accuracy of assessments.
Based on parts of the data that are entered in the NDB, a rules-based profiling tool – which is incorporated in the NDB – automatically displays a recommended segmentation “phase” as a proxy of a jobseeker’s job readiness. The profiling tool recommends one of two options (Phase 1 for more job-ready clients and Phase 2 for less job-ready clients). Counsellors can confirm or change the recommendation of the tool according to their own assessment or can chose a third phase (least job-ready). The current tool is relatively simplistic, more than one decade old, and relies only on information that is entered as categorial variables (e.g. age, education level). Data that are entered as free text, including important information gathered during the skills assessment and counsellor’s observations regarding client behaviour/motivation, do not affect the tool’s result. This information is only used by the NEA counsellors to verify the segmentation recommended by the tool and change it if necessary (Figure 3.3, right panel).
The absence of an advanced profiling tool leads to inconsistent profiling results across local labour offices. In some offices, the vast majority of jobseekers is in phase one, while in others, the vast majority of jobseekers is allocated to phase three (Figure 3.4, Panel A). What is more, the share of registrations for which counsellors change the phase recommended by the existing profiling tool varies largely, from close to 0% in some local offices to almost 90% in others, reflecting either a potential over-reliance on the tool or a potential mistrust, respectively (Figure 3.4, Panel B). A plausible explanation is that, due to limitations in the current profiling tool, job counsellors often rely on their own judgement and experience to determine the appropriate support for jobseekers. Against this background, it is a promising first step that the NEA recently started to develop a new and more advanced profiling tool.
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata.
The number and intensity of counselling sessions may vary, depending on the jobseeker’s situation and needs. According to the NEA microdata, at the start of their journey, jobseekers across different phasing groups meet with job counsellors at a relatively similar frequency. However, this pattern tends to change over time as jobseekers remain registered with the NEA. Specifically, those who are closer to the labour market are found to have more frequent meetings, while individuals facing significant employment barriers experience longer gaps between counselling sessions (Figure 3.5). This discrepancy suggests potential cream skimming. Resources thus seem to be disproportionately allocated to those who are already more likely to get integrated into the labour market, rather than reaching those who face greater challenges and barriers to employment. This highlights a key area for improvement in counselling provision and related support services, including profiling. The fact that jobseekers registered with the NEA have a diverse profile and needs – including low digital literacy, limited skills, various underlying problems and often lack of motivation – adds to this challenge. Beyond the most vulnerable jobseekers, this issue is also evident in regions undergoing transition, where efforts to intensively profile and counsel affected individuals remain limited. One of the contributing factors to cream skimming is the counsellors’ heavy administrative duties, which can adversely affect the frequency as well as the duration and quality of the sessions.
Meeting frequency by jobseeker profiling phase
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata.
3.2.4. The NEA establishes action plans for all jobseekers and additional agreements for long-term unemployed
Action plans (AP) are established for almost all jobseekers during the meeting with the counsellor. The AP is typically created within one month of registration to outline the steps towards labour market integration. It includes specific immediate actions (one action to be taken before the next meeting with the counsellor) and is usually developed jointly be the counsellor and the jobseeker. The AP may be agreed upon verbally or documented in writing. Information is recorded in the NDB, though not consistently. The absence of digital tools further hinders the creation and tracking of APs. Additionally, not all counsellors update the APs of their clients on a regular basis.
Failure to follow the steps outlined in the AP can lead to the termination of the jobseeker’s registration with the agency. Registrations are more likely to be discontinued among certain population groups, including Roma, jobseekers under 25, and those with low education levels. Jobseekers whose profile directly matches an open vacancy follow a so-called “fast-track” process and are referred to jobs immediately, even before being profiled or having their AP created. This process is primarily used by the NEA for prime‑aged and older jobseekers, and its use has increased substantially from 7% of cases in 2019 to 15% in 2022.
Additional agreements are established for people who have been unemployed for more than 12 months. The Employment Integration Agreement (EIA) is an additional agreement that complements the AP, and its purpose is to provide more tailored and structured support to long-term unemployed individuals. The preparation of the EIA includes an in-depth assessment to identify challenges, gaps, and external factors affecting employment, and proposes more intensive actions like training, subsidised employment, or specialised workshops. It also proposes services provided by other institutions to overcome barriers to labour market integration. However, collaboration with external institutions and service providers is mostly ad hoc and therefore the relevant referrals are not integrated into the AP and the EIA. The EIA is signed on paper and its results are stored in the NDB. EIAs are often viewed as a challenge by NEA staff, who report redundancies and overlaps with the AP, and raise concerns about the EIA’s added value. Additionally, the process of establishing EIAs, particularly the assessment of obstacles to employment, is seen as inefficient and time‑consuming.
Although referral patterns vary slightly across different jobseeker groups, differences are not substantial. For instance, while job-ready individuals (classified as Phase 1) are more likely to be referred directly to job vacancies, the gap in referral rates between these and jobseekers in Phases 2 and 3 is relatively small. Similarly, while Phase 3 jobseekers – who face the greatest barriers to employment – receive more referrals to training programmes, this does not appear to be sufficient to address their complex challenges.
The current level of personalisation in referrals could be enhanced by more effectively tailoring support to the individual characteristics and needs of jobseekers. In addition, there is currently no use of simultaneous referrals, meaning that jobseekers are often referred to services or programmes in a sequential manner, rather than being connected to multiple resources at once. This lack of co‑ordination can lead to delays and inefficiencies in addressing the diverse needs of jobseekers. Moreover, there are no consistent follow-up interactions after a jobseeker completes an action.
3.2.5. Services are not directly linked to segmentation phases
The NEA offers various services for its clients. This range from information on job vacancies and ALMPs, to professional development counselling, recruitment mediation, referrals to jobs, training and subsidised employment programmes (see Table 3.3 for a detailed overview of the NEA’s current portfolio). ALMPs in Bulgaria have been financed through a mix of European and national resources. Each year, according to the Employment Promotion Act, the Ministry of Labour and Social Policy, together with other ministries and social partners, develops a National Employment Action Plan (NEAP). This plan serves as a key instrument for implementing active labour market policies, outlining annual priorities and measures to support employment and skills development. The NEAP is co-financed from the State budget and the European structural and investment funds. Key EU instruments include the European Social Fund Plus (ESF+), in particular the Human Resources Development Programme (HRDP) and the Youth Employment Initiative (YEI), and more recently the Recovery and Resilience Facility (RRF). In addition, the Just Transition Fund and other EU funds are mobilised to address the challenges associated with the green transition.
Nevertheless, both the data and feedback from a survey of NEA frontline staff suggest that the support provided to jobseekers remains overly standardised offering limited scope to address individual or complex situations and lacks the flexibility to respond to evolving labour market needs. Capacity constraints also persist, with demand for certain programmes and services exceeding the number of places available.
Even though the services and measures that are provided to jobseekers differ to some extent depending on a jobseeker’s individual employment barriers and their job readiness, overall differences are relatively small and not directly linked to the outcomes of the current phasing/profiling system. For example, although in principle job-ready individuals (Phase 1) should be referred directly to job vacancies, the gap in referral rates between them and those in Phases 2 and 3 is not large. Similarly, while Phase 3 jobseekers, who face more barriers, should receive more training referrals, in practice they don’t (Figure 3.6). Depending on counsellors’ objective for a jobseeker, they have different types of services at their disposal. The choice and objective of the service is usually discussed with the jobseeker but is ultimately at the discretion of the counsellor, who may or may not take account of the segmentation phase of the jobseeker.
Services provided to different types of jobseekers, by profiling outcome (“phase”) and number of meeting with counsellor
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata.
While the services provided by the NEA are likely adequate for more job-ready jobseekers, there is a need for more targeted and intensive interventions for individuals who face greater challenges in accessing employment. To improve matching between services and measures and individual needs, the NEA developed in 2019 standardised service packages for seven vulnerable groups of jobseekers.5 For instance, as part of such a service package, a very common first service for 15‑24 year‑olds is the provision of information on the Youth Guarantee before proceeding to other types of support. However, the use of these packages is not universal, and many jobseekers of this group receive other types of services. In addition, these standardised packages often do not account for the unique circumstances and requirements of each individual.
There is widespread consensus on the critical need for proactive outreach, particularly to vulnerable and hard-to-reach groups. Existing efforts have been positively received and should be expanded to ensure better visibility. To achieve this, it is essential to develop a comprehensive and dedicated outreach strategy that intensifies efforts and maximises the potential of available channels. This approach should also include broader collaboration with various stakeholders to ensure greater reach and impact.
By building on existing practices, leveraging new tools under development, and taking full advantage of data-sharing related developments, the NEA can significantly strengthen its outreach efforts. This work should be organised around the following five key areas (Figure 3.7):
Source: Authors’ compilation.
1. Develop a dedicated outreach strategy:
Leverage the new data-sharing agreements with other public bodies, to identify inactive and unemployed individuals who are not necessarily registered with the NEA and draw a detailed profile of their needs and barriers. This will enable the NEA to tailor its outreach efforts to less well-served populations who may still benefit from NEA’s support. By relying more on data, the NEA will likely be able to spend less resources in terms of staff time on identifying inactive and unemployed people in need of support and use the freed time to motivate people to register with the NEA and provide more tailored support.
Develop a segmented approach based on the data in its registry and create specific outreach campaigns for different target audiences, such as youth, older workers, long-term unemployed individuals, and those from marginalised communities (e.g. Roma, people with disabilities). Prioritise implementation according to the priorities of the agency.
Set clear and measurable goals for each target group, such as increasing registration, promoting skill development, or improving job placement rates. Make adjustments based on the effectiveness of these goals.
2. Leverage all available channels:
Continue and expand direct outreach efforts, including through the local labour offices, the mobile offices and the activation clubs. Make sure that the mobile offices reach populations with low digital literacy, including by working directly with local communities and community centres in disfavoured areas. Ensure that all the offices are equipped with adequate and specialised personnel to serve all types of clients.
Invest in improving substantially the current website as it serves as the first virtual image of the agency. The website should be designed to be user-friendly, with intuitive navigation and accessibility features to ensure that all individuals, including those with disabilities, can easily access the information. The website should provide clear, comprehensive details about the registration process and the full range of services offered by the NEA. Enriching the website with interactive features, such as live chat support, FAQs, videos with success stories and step-by-step guides, will further enhance the overall user experience.
Introduce and promote the use of a mobile app. Unlike the NEA’s website, which requires users to actively visit and navigate through a browser, a mobile app would remain continuously available on users’ phones, providing instant access with a single tap. This enhances user-friendliness and offers greater accessibility and convenience, particularly for individuals with limited digital skills. A mobile app would allow individuals to access job listings, register for services, receive notifications about employment opportunities, and connect with counsellors – all from their smartphones. This is especially useful for younger jobseekers, those in remote areas, or individuals who may hesitate to visit PES offices in person. More advanced app features such as voice assistance, simplified navigation, and multilingual support can help users with low literacy or language barriers.
Maintain active profiles on all social media platforms, which are particularly impactful for engaging younger individuals. Adapt the messaging to reach various demographics. For example, use social media for engaging younger audiences with short, catchy videos on job opportunities or training programmes. At the same time, ensure presence on traditional media outlets such as TV, radio and print media to ensure it reaches those who may not be as active online or those living in areas with limited digital access, particularly older or more vulnerable populations.
Send tailored regular mail, emails and messages to identified individuals based on their profile, informing them of job vacancies, training opportunities, and upcoming workshops. These could include reminders about deadlines or new opportunities.
Share videos, audio messages, and voice notes instead of text messages for certain target groups with limited digital literacy. WhatsApp and Viber apps can be also used for this purpose.
3. Engage key stakeholders:
Work closely with the data-sharing institutions to ensure smooth collaboration and timely access to data.
Disseminate informational material in local community centres, schools, libraries, and other public spaces. Continue to use these places to organise informational sessions and job fairs.
Further strengthen co‑operation with employers to advertise job opportunities, apprenticeships, and vocational training programmes. Team up with business associations and chambers of commerce to help raise awareness among companies about the NEA services, especially for hard-to-reach groups.
Expand collaborations with educational institutions such as schools, universities, and vocational training centres as well.
Further strengthen the co‑operation between the NEA and the SAA to ensure more effective outreach to vulnerable individuals who are far from the labour market. In the past, the uptake of social assistance in Bulgaria has been low, partly due to long waiting periods and low benefit levels, thereby limiting the pool of inactive individuals that can be reached through the NEA-SAA partnership. An up-to-date analysis of social assistance take‑up rates should be conducted in the first place and – depending on the findings – outreach efforts should be complemented with targeted measures to increase the engagement with social assistance recipients (OECD, 2022[1]).
Co‑operate with NGOs and civil society groups working with vulnerable populations at the local level (e.g. homeless, Roma, or people with disabilities, NEETs), to help bridge the gap between the NEA and individuals who may not be actively seeking employment. Expanding the number of Roma mediators in local labour offices in areas with a high Roma presence, as well as employing mediators for the Turkish minority and other underserved groups, can provide culturally sensitive, tailored outreach and support to these communities.
Work through trusted local organisations or leaders to engage with groups that may be sceptical about registering with the NEA. This could include working with religious organisations, youth groups etc.
4. Adopt clear, accessible and targeted messaging:
Simple Language: Use straightforward, easy-to‑understand language to communicate PES services. Avoid jargon and technical terms that might confuse or alienate target audiences, especially those with limited education or literacy.
Visuals and Infographics: Use engaging visuals such as infographics, charts, and short videos to explain PES services, job search tips, and success stories. Visual content can be shared across digital platforms and displayed in physical spaces like community centres.
Multilingual Resources: Ensure that outreach materials are available in multiple languages, particularly in regions with significant minority populations, to increase accessibility.
Success Stories and Testimonials: Share real-life stories of individuals who have found success through PES services. This can build trust and motivate others to engage with the service.
Tailored communication can help ease any doubts or concerns to registration, particularly for groups that may be reluctant to engage with or trust the NEA.
Make it easy for individuals to take the next step in their job search. Use clear calls-to‑action, such as “Register Now”, “Find a Job”, or “Sign Up for Training”, and explain how easy it is to get started.
5. Assess effectiveness of outreach efforts and adapt them accordingly:
Collect feedback from the target groups to assess the effectiveness of outreach efforts. Use this feedback to refine the approach and adapt messaging, channels, and services based on the evolving needs of the unemployed and inactive populations.
Regularly review the data provided by partner institutions to ensure the PES is focusing on the right individuals and adjusting outreach strategies accordingly.
With respect to information about jobseekers upon their registration, there is a need to go beyond recording personal information and employment history and place emphasis on recording their preferences and aspirations. This would enable counsellors to offer more tailored guidance and make more relevant referrals in later stages. Accepting unemployment applications, introduced in 2020, increases NEA staff workload but significantly reduces administrative barriers for jobseekers and reinforces the NEA’s role as a primary support contact. Therefore, it is important to maintain this practice.
The NEA plans to upgrade its IT infrastructure and improve its data management, which is expected to streamline the client experience, significantly reduce processing times and minimise the potential for human error during data entry. It should also alleviate administrative burden on NEA staff, particularly benefiting job counsellors by freeing up more time for client-focused services (see also data management section under Chapter 5).
In addition to regularising the automatic retrieval of data from other administrative databases, the NEA should actively promote and encourage remote registrations for those who can use them. This could involve raising awareness about the availability and convenience of online registration. The registration process itself should be facilitated through a user-friendly interface and tools for remote access, through the NEA’s website and a potential mobile app. The NEA should provide clear instructions including, for example, through a brief video, explanatory notes for relevant fields, as well as a comprehensive FAQ section to guide users through the process. To further streamline the process, some fields could be pre‑filled through interoperability with other administrative systems. This approach would help alleviate the workload of frontline staff and limit reporting error while providing jobseekers with a more convenient and accessible option.
While advancing digitalisation is important, a careful balance must be maintained to ensure that no jobseeker is left behind in the transition to more digital processes. Digital tools can be effectively used by individuals who are closer to the labour market and more digitally literate. However, in-person support remains essential for certain groups – particularly older jobseekers, those who lack access to digital infrastructure and those with limited digital skills. For these individuals, the option of paper-based registration should be preserved to prevent exclusion and ensure equitable access to NEA services. Another option for such individuals would be to allow for online registrations on site at local labour offices, using available PCs with assistance from NEA staff. Upon registration, the system could automatically generate information and notifications about next steps as well as the first appointment with a counsellor.
Since a local labour office serves numerous clients each day, with or without appointments, effective resource management is essential. Ensuring that staff and resources are efficiently allocated helps to streamline operations, minimise wait times, and maintain a high level of service for all clients. Effective resource management usually relies on comprehensive management systems that can streamline scheduling of appointments, client intake, and service delivery. These systems allow for better allocation of resources, reduce wait times, and ensure that both scheduled and walk-in clients are handled efficiently. Additionally, knowledge exchanges between different local offices could prove to be very informative. By sharing good practices and mutual learning, each office can identify what strategies work most effectively in their context, for continuous improvement in jobseeker support.
3.5.1. Continuous and tailored engagement with jobseekers to avoid labour market detachment is key
The NEA should review and refine its counselling model by implementing a more structured, needs-based approach that takes into account the diverse profiles of jobseekers in its database (Figure 3.8). Each jobseeker requires tailored support based on their unique situation, as they often face multiple challenges beyond unemployment – including health, social, and household issues – which affect their ability to engage with the labour market. Older jobseekers often face low literacy, limited skills and various underlying problems, highlighting the need for more targeted support for this group in the context of an ageing population. In this respect, increasing specialisation among the NEA’s frontline staff to ensure that jobseekers receive the support they need would be highly beneficial. Counsellors should be able to support jobseekers not only in entering the labour market, but also in navigating the necessary steps to prepare for sustainable labour market integration. This includes supporting jobseekers changing trades or specialties if they wish to. Before focusing on the content of counselling, however, it is essential to revise counsellors’ administrative duties, ensuring they have as much time as possible for direct interactions with jobseekers. This would enhance the quality and frequency of individual counselling sessions, allowing for more personalised support.
Increasing the number of group counselling sessions and workshops can enhance the efficiency and reach of employment services by providing support to more jobseekers simultaneously. Such sessions could be focused on topics such as job search, CV preparation, interview readiness, and soft skills, as well as information for the labour market in Bulgaria. Some group counselling sessions could focus on vocational guidance, targeting mainly but not exclusively jobseekers referred to training. These sessions would help participants select the training programmes that best match their skills, interests, and career goals. Group sessions can have multiple benefits for participants. They can foster peer learning, encourage motivation, and create a supportive environment where individuals can share experiences. Participants also gain both practical skills through role‑play and exercises, and soft skills through their interaction with others.
Counselling should be delivered not only in-person but through other channels as well. In most PES, counselling – as well as other job-search support services – are also delivered remotely, either via phone or through digital channels, which can be especially beneficial for individuals with mobility issues or located in remote areas. Some of these digital platforms resemble virtual offices and have embedded features such as interpretation services in sign language or foreign languages ensuring that all jobseekers, regardless of their background or specific needs, can access the support they require (Box 3.1). By diversifying the methods of counselling delivery and investing in the establishment of a fully virtual office or platform, the NEA can expand reach and flexibility, improve accessibility and enhance the overall effectiveness of its counselling services. Thereby, the NEA is better able to serve jobseekers where they are and provide the appropriate level of support.
In December 2020, the Greek PES launched tele‑counselling as an alternative to on-site counselling services through the specially designed myDYPAlive platform (European Commission, 2021[3]). Counselling on this platform is delivered by specially trained counsellors from various local employment offices around Greece. While the platform provides counselling for both jobseekers and employers, it has been particularly beneficial for individuals with disabilities and other vulnerable groups facing significant barriers to labour market integration. There are several new features, such as specific counselling services with interpretation for foreigners and people with hearing disabilities. Its underlying Customer Service Management mechanism has shown that the vast majority of users are highly satisfied with the platform.
Source: European Commission (2021[3]) and presentation by officials of the Greek PES during the international workshop on PES operating models organised on 7 February 2025 as part of this project.
To ease the workload on counsellors and enhance efficiency, the NEA should consider reducing in-person meetings for jobseekers who are likely to secure employment on their own. For these individuals, online assistance could be offered, and a digital AP could be automatically generated by the system. Nevertheless, careful balance with digitalisation is needed. While digital means can be fully deployed by the jobseekers who are close to the labour market, it will be important to retain in-person and more intensive support, for jobseekers facing significant employment barriers – especially older jobseekers and those with low digital literacy.
The NEA should also consider revising the timetable and duration of counselling sessions to ensure that jobseekers receive the appropriate level of support. An effective way to manage this would be to implement a digital tool or calendar that automatically alerts counsellors when a jobseeker may require follow-up sessions. This tool would help counsellors to stay on top of their caseload and proactively check-in with jobseekers who need more frequent guidance, ensuring they remain engaged and motivated throughout their journey towards (re)employment.
The NEA’s approach vis-a-vis counselling should be regularly reviewed and remain flexible. For example, in areas undergoing transitions where mass-layoffs are anticipated, the NEA should engage with at-risk workers early, offering counselling and related assessments before they become unemployed. This proactive approach would better prepare these workers for the challenges ahead and help prevent delays in accessing support services. In parallel, this strategy would help alleviate some of the pressure on local labour offices and job counsellors, as the caseload would be distributed more evenly.
Source: Author’s compilation.
3.5.2. Advanced profiling and assessment tools will help the NEA to get a more accurate picture of jobseekers’ skills, needs, and barriers to employment
Most PES are progressively adopting statistical and AI-driven profiling tools to increase predictive accuracy and facilitate the work of job counsellors (see the case of Estonia in Box 3.2). The NEA would also greatly benefit from a more modern and advanced profiling tool to streamline jobseeker segmentation and ensure consistency across local labour offices. A more elaborate profiling tool with a closer match to the needs of the jobseekers would help ensure that the segmentation achieves its desired objectives, i.e. that jobseekers receive support which closely aligns with their needs. To address the cream skimming issue mentioned earlier, it’s key to refine the profiling system to ensure that it accurately identifies individuals facing the greatest challenges and most in need of support, particularly those who may otherwise be overlooked or excluded from services. Subsequently, this would also ensure that resources are more equitably distributed to those who need them most. To be effective, the tool must be trained on a rich, individual-level dataset covering a broad range of relevant information (Box 3.3).
Even with the introduction of a statistical or AI-based profiling tool, it will be key to preserve a degree of discretion for counsellors, as they are the ones in daily contact with jobseekers and best positioned to apply their expertise and professional judgment in making final decisions about the phasing and support provided. The current tool’s recommendations should be seen as indicative only, with counsellors using their judgment to ensure tailored support (see Chapter 5 for a detailed discussion about a new profiling tool). Targeted training for counsellors is essential to ensure the effective use of a new profiling tool, alongside involving them in the design and development process, as they are a key end-user group. Experience from other PES shows that a lack of stakeholder involvement from the outset – particularly of counsellors – has often hindered successful implementation and led to public criticism. It is also crucial to communicate about its benefits, and to have a robust ethical framework in place.
The Estonian PES follows a work-focused counselling approach, where counsellors and clients collaborate to support the client’s path to suitable and sustainable employment. This includes analysing the jobseeker’s profile and factors that influence employment opportunities, setting goals, developing and regularly updating an individual Action Plan and guiding jobseekers in exploring labour market opportunities and expanding personal work- related choices. The proposed employment measures consider both individual and labour market needs. Counselling occurs at least every 30 days – more often if needed – with each session reviewing progress and setting next steps. A gradual sanction system applies in cases of non-compliance without valid reason. This counselling system operates through a two‑tier model: case managers provide 45‑minute sessions for 100‑120 clients, while job mediation counsellors offer 25‑minute sessions for around 200‑250 clients. While all counsellors provide employment counselling, job mediation, ALMP referrals, and process benefit claims, case managers also handle casework, networking, follow-up support, and workability assessments, whereas job mediation counsellors manage jobseeker registrations. Counselling is delivered through a blended approach, including face‑to-face meetings, phone, Microsoft Teams, and the PES e‑channels, with options for assisted or independent job search.
The Estonian PES counselling model has recently been enhanced by the introduction of new digital tools. One such tool is a data warehouse dashboard, which provides counsellors with a comprehensive overview of jobseekers’ profiles and key information within their specific PES office. Additionally, digital data exchange is facilitated through X-Road® – an open-source software and ecosystem that enables secure, standardised communication between different information systems of public and private sector organisations in Estonia.
Counselling has also been enhanced by a decision support tool based on machine learning that helps counsellors efficiently assess jobseekers’ needs and prioritise support for those requiring the most assistance. By enabling more targeted interventions, the tool helps reduce inefficiencies – avoiding unnecessary support for those likely to find employment independently and ensuring timely guidance for those needing more intensive help. The tool has been implemented across all PES offices since October 2020, following testing, end-user training, and a pilot phase in five offices. Since its implementation, the tool has been improved through user experience seminars, feedback collection, and testing, along with ongoing training for users. As of 2023, the system has also integrated Gradient Boosting techniques and enhanced its ability to interact with other data sources to improve decision-making accuracy.
The decision support tool is now fully integrated into the daily workflow of counsellors. The underlying model assesses each job seeker’s probability of entering the labour market and risk of returning to unemployment – both rated on a scale from very low to high and also indicates the top 10 factors affecting the probabilities. For the estimation of these probabilities, the underlying model uses 60 variables that encompass jobseeker characteristics along with local labour market conditions – including the number of vacancies and jobseekers. Based on these results, the tool positions the jobseeker within the counsellor’s portfolio, supporting more informed and efficient case management.
The results of the tool help employment counsellors in determining the best channel of support, as well as the optimal frequency of contact, and recommending tailored services and interventions based on each jobseeker’s needs. It also provides counsellors with a comprehensive overview of their client portfolio and associated risk levels, aiding in workload planning and schedule management. Despite its high accuracy, the tool serves only as a support mechanism as final decisions are always made by counsellors, not determined by the model alone.
Source: Presentation by officials of the Estonian PES during the international workshop on PES operating models organised on 7 February 2025 as part of this project.
To ease the burden from counsellors, the NEA could consider enhancing counselling services through the adoption of automated assessment tools. These tools help jobseekers identify their skill profiles and increase their self-awareness about their abilities and preferences. At the same time, they allow job counsellors to gain deeper insights into their clients’ skills, needs and conditions, enabling them to provide more targeted and tailored support (OECD, 2023[4]). This would reduce human error, improve consistency, and speed up the process of compiling accurate jobseeker profiles (see Chapter 5 for a detailed discussion about assessment tools).
The NEA is currently exploring the creation of a modern profiling tool to improve predictive accuracy and better support job counsellors in delivering tailored assistance to jobseekers. To achieve its intended objectives, such a tool must be trained on a comprehensive individual-level dataset that captures a wide range of relevant information.
Despite the fact that the available data offers a solid foundation for developing an initial version of a statistical profiling tool, significant limitations remain. One of these is the incompleteness of existing registers, which lack critical variables that could enhance the precision and relevance of profiling. Missing variables include job preferences (such as preferred industry, salary expectations, preferred working hours or willingness to relocate), transportation barriers (e.g. access to public transit or car ownership), and skill gaps. Additionally, even within the available datasets, data quality issues pose a concern. Certain variables suffer from inconsistent reporting or are systematically missing. For instance, variables such as the field of study, reasons for leaving previous jobs, and behavioural engagement on digital platforms are either only sporadically recorded or unavailable altogether. These data gaps can undermine the completeness and reliability of the profiling tool, potentially skewing model outputs or limiting its applicability to specific subpopulations and therefore should be carefully considered when interpreting model outcomes. The profiling tool could also benefit from additional information on local labour market demand. Although job vacancy data is available, it frequently suffers from generic or incomplete job descriptions, making it difficult to extract meaningful insights about the specific skills, qualifications, or experience required.
Future iterations should aim to address the identified gaps, improve the quality and granularity of key variables, and where feasible enhance data collection procedures to better capture key information. In this respect, recent changes to the Employment Promotion Act and improved data-sharing arrangements with other public institutions, including the NRA, are expected to significantly improve the NEA’s data collection efforts and processes. When the new profiling tool is developed, particular attention should be paid to missing values and data inconsistencies throughout both the modelling and implementation phases. The profiling data are also expected to feed into the new matching tool, currently under development by the NEA’s external contractor, acting as criteria to search for and match relevant opportunities.
Should the NEA move towards the development and implementation of an AI-based profiling tool, additional requirements would need to be considered. Such requirements may indicatively include higher data volume and diversity to ensure model training quality and avoid biased predictions, more frequent updates and real-time data feeds to maintain model relevance over time, extensive feature engineering to create variables that capture complex behavioural and contextual patterns, clear protocols for model explainability and transparency, ensuring that AI-based decisions can be interpreted and justified according to legal and ethical standards, ongoing performance monitoring and regular retraining of the underlying model to adapt to labour market changes etc.
To increase the efficiency and effectiveness of the current process regarding the Action Plans and referrals, the NEA could consider the following (Figure 3.9):
Promoting the use of more forward-looking Action Plans would benefit jobseekers by making their employment journey easier to anticipate and plan accordingly. By outlining clear, progressive steps, jobseekers would gain a better understanding of their pathway to employment, reducing uncertainty and increasing motivation.
To enhance effectiveness of Action Plans, the NEA should ensure these are consistently documented, updated to reflect new conditions, actions taken, and outcomes. This information needs to be diligently recorded in the NDB. Clear guidelines should be provided to all counsellors regarding the recording of information and on the frequency of updates.
The fast-track process is a promising measure and should be thoroughly assessed. If it has generated positive results, it could be extended to ensure that it is not limited to prime‑aged or older jobseekers but also to other target groups such as younger jobseekers who may match open vacancies. Developing a clear, consistent set of criteria for fast-tracking could ensure more equitable use of this process and improve matching efficiency. Information about the fast-track option should be more widely promoted to ensure jobseekers are aware of this opportunity from the outset.
With respect to the differentiation between EIAs and Action Plans, the NEA could consider several strategies. Eliminating EIAs and incorporating them within the standard Action Plan framework with specific dedicated features and actions for long-term unemployed and vulnerable jobseekers would be a first option. If the NEA wants to retain EIAs as a standalone step, efforts should focus on reducing the administrative burdens that EIAs imply for NEA staff. Improvements should include eliminating redundancies between the assessment of labour obstacles that are already included in the NDB, and the assessments made for the EIA, automating routine updates of the EIA and standardising common steps for addressing typical employment barriers. Additionally, the EIA process should be simplified to ensure that it adds clear value and does not duplicate steps already covered in the Action Plan.
The frequency and type of follow-up interactions between jobseekers and the NEA’s job counsellors should be adapted based on the jobseeker’s trajectory. For example, after the completion of a training programme, jobseekers may benefit from periodic check-ins to support independent job search efforts and refer them to vacancies. Informal post-placement check-ins, such as brief phone calls, could be introduced for more challenging cases to ensure ongoing support. In any case, counsellors should have access to data on referral outcomes to inform their post-referral guidance.
The NEA could introduce features that automatically generate synthesised, employment-relevant summaries for employers (like simplified CVs) based on Action Plans, actions taken, and other important information relevant to employers. Additionally, employer testimonials could be incorporated into referrals, particularly when jobseekers apply for positions in the same occupation or sector.
Digital tools to assist both the creation and tracking of Action Plans should also be introduced. By automating parts of the documentation and assessment processes (e.g. gathering of standard information from existing registers), the NEA can reduce the time spent on administrative tasks and improve the efficiency of service delivery. A centralised digital platform could be used to track progress and automatically notify jobseekers and counsellors of necessary updates.
A detailed survey of both counsellors and jobseekers could help uncover the root causes behind the observed lack of differentiation in referral patterns across jobseeker groups. Such a survey would shed light on whether this is due to a limited range of available actions for referrals, which would require the NEA to expand service offerings and make sure they are continuously available. Counsellors may also lack access to key information, such as skills (both technical and soft skills), which are not systematically recorded, limited insights into local labour market conditions as well as the performance of different programmes and services (what works well and for whom). To address these challenges, the NEA – in addition to a new profiling tool – should invest in skills assessment tools, digital tools that can offer real or near-real time insights into local labour market conditions and trends, as well as monitoring mechanisms to track the success rates of the NEA’s different programmes. Such digital tools could greatly improve the situation, providing insights to counsellors so they can make more well-informed referrals.
Some interventions can be offered simultaneously for greater efficiency. For example, a technical training programme could be paired with language, digital or basic/foundational skills training, professional counselling, or job search preparation.
A further avenue to be considered are strengthened partnerships, along with formal collaboration agreements, with external institutions and service providers that could typically offer services that the NEA cannot and incorporate these referrals into the AP. Establishing stronger partnerships with relevant institutions at the local level would ensure that jobseekers who require multiple forms of support receive comprehensive assistance tailored to their individual needs, while avoiding duplication. For example, strengthening partnerships between the NEA and other entities (both public and private) that offer literacy training or other relevant support could help identify needs early and co‑ordinate interventions effectively. In addition, these referrals should be incorporated into the AP and provided to jobseekers early in the process without having to wait until they become long-term unemployed. To facilitate this, the NEA should develop a comprehensive, regularly updated mapping of external service providers at the local level. Once jobseekers complete these external actions and services, institutions should refer them back to the NEA to update their APs and continue their employment journey in collaboration with the NEA’s counsellor.
Note: AP: Action Plan, EIA: Employment Integration Agreement.
Source: Authors’ compilation.
To enhance the effectiveness of its services, the NEA should move away from offering a standardised menu of services for all, and instead tailor its ALMPs to better meet the diverse needs of different clients – both jobseekers and workers who may wish to change job/specialty. This expanded portfolio of services should be available on a continuous basis (see Table 3.3. for indicative examples of additional services). The NEA’s clients are diverse, so it is crucial to assess its current portfolio of ALMPs and update it based on the specific needs of target groups, including women, older workers, individuals with disabilities, and youth. This may involve removing outdated services or introducing new ones that address emerging challenges, such as those created by industry transitions or changing labour markets. A detailed analysis of registered jobseekers can provide valuable insights into their profiles and needs and thereby inform the design of new programmes. For instance, some jobseekers may face language challenges, and therefore could benefit from targeted language training. The outcomes of a new profiling tool can also help categorise jobseekers and offer them the support they need.
Early intervention is crucial to avoid that jobseekers fall into prolonged unemployment or uncertainty. By implementing services and programmes as soon as possible, the NEA can provide timely support, maintain motivation, and increase the chances of a quick and sustainable return to work. A standout example of early intervention at the Swedish PES is the “Matching from Day 1” initiative, which reflects the agency’s commitment to providing intensified support for jobseekers who are furthest from the labour market (Box 3.4).
Matching from Day 1 (MD1) is a proactive and personalised intervention developed by the Swedish PES to support jobseekers with lower chances of entering the labour market, particularly newly arrived refugees and their family members who have been in Sweden for no more than four years, with a focus on women. The methodology is built around three core elements: (i) a thorough mapping of jobseekers’ formal qualifications (such as education and language skills), informal competencies gained through life experiences, interests, and personal motivation and preferred occupation, (ii) recruitment of employers and vacancies according to jobseekers’ preferred occupation with demand for those specific skills, and (iii) matching between jobseekers and employers based on the identified fit.
The mapping of jobseekers is implemented from the very first meeting and, in some cases, can be completed within a single session. All jobseekers are asked to respond to the same set of questions to avoid bias. The PES staff identify and assess potential employers, also applying diversity and inclusion criteria to reduce the risk of discrimination in hiring. Employers are considered not only for current vacancies but also for their growth potential. The MD1 model has proven effective, with a large randomised controlled trial (RCT) involving over 7 000 participants showing positive impacts on subsidised employment outcomes. In fact, the evaluation showcased a 25‑30% increase in employment transitions, particularly benefiting jobseekers with the lowest employment chances. There was no evidence of displacement effects, and the costs were comparable to other matching programmes, including those delivered by private providers of job search assistance services.
For an ALMP system to remain effective and responsive to the evolving demands of the labour market, it must be flexible – able to retain programmes that deliver results while adapting or phasing out those that do not. Adaptation of ALMPs should also involve key stakeholders and reflect local labour market conditions and specific needs of key groups. If resources are limited, the NEA could consider launching pilot programmes – with a randomised controlled trial (RCT) component from the outset, and then scale/roll-out based on the results. A clear example of this flexibility comes from the PES in Slovenia with the launch of a ChatGPT training programme in May 2024. In response to the growing importance of AI for jobseekers, the PES introduced innovative eligibility criteria for instructors, requiring only prior experience in delivering relevant workshops, rather than formal qualifications (OECD/European Commission, 2025[5]).
Services and programmes take regional labour market specificities into account, as needs tend to differ largely across regions. In particular, areas undergoing economic transition or dealing with a seasonal workforce should receive specialised programmes that address immediate labour market demands. In contrast, urban areas with high unemployment or regions with long-term unemployment may benefit from comprehensive employment packages, offering a blend of support measures to help individuals return to work. Bringing in a region-specific perspective and design programmes to address specific regional challenges will also allow for a more responsive and tailored approach. To ensure that ALMPs are truly responsive to current challenges, the NEA should conduct a comprehensive assessment of regional labour market needs. This assessment should include identifying the specific skills required by employers and addressing the immediate demands of regions undergoing transitions, where workers may face displacement due to industry changes.
Mapping of stakeholders in target areas will be key to providing effective services. This includes identifying key industries, local authorities, training providers and educational institutions with the capacity to deliver the necessary training and support. Partnerships with these stakeholders will help the NEA expand its reach and improve the quality of services offered. One good example in this respect is the PES in Lithuania. The PES website includes a mapping of all social services available, organised by municipality. In turn, caseworkers provide intensive and personalised counselling support for the most vulnerable jobseekers referring them to external partners (municipalities, NGOs, social partners) who provide such services.
The NEA could establish specialised teams or one‑stop shops dedicated to serve and support specific target groups – such as youth or persons with disabilities. These teams or one‑stop shops could offer a comprehensive range of services in addition to employment services, including legal and financial advice, and housing assistance while making formal referrals to the appropriate social services. To ensure accessibility, they should be open without the need for appointments or preconditions. An good practice example in this respect are Welfare Plus Centres in Korea, which are organised as one‑stop-shops to support vulnerable jobseekers, bringing together staff from partner organisations – such as local governments and NGOs – providing ALMPs and social services (OECD, 2024[6]). Similarly, Finland has created one‑stop-shops for all young people under 30, including marginalised young people with major employment barriers, providing training and a wide range of other types of support (OECD, 2024[6]). Building on this experience,
A lot of important initiatives take place at the local level. The Swedish PES tailors its efforts to meet specific regional needs and challenges. The most prominent example is met in Northern Sweden, where the regional PES office in Norrbotten is trying to address the significant labour shortages, while also focusing on supporting the most vulnerable population groups such as the long-term unemployed individuals, people with disabilities, and those with low educational attainment (Box 3.5). Meanwhile, in the south of the country, initiatives such as the Inclusive Job Search pilot in Malmö are designed to remove digital barriers for jobseekers, in line with the agency’s broader efforts of integrating sustainability into all its operations and processes.
The PES in Sweden maintains a close collaboration with municipalities1 – both at a strategic and operational level – to deliver integrated support to jobseekers. This partnership allows for a holistic approach, ensuring that individuals facing employment obstacles receive not only employment services, but also social support, such as housing, healthcare, and other welfare services as well as education-related services. The PES with the municipalities establish agreements on the structure and organisation of their co‑operation. The latter involves mapping and analysing local and regional labour market needs, as well as the composition and needs of target groups. It also includes mapping joint measures and vocational education resources. Annual follow-ups are conducted to assess progress toward agreed objectives, identify challenges, and implement corrective actions to improve outcomes.
The collaboration between the municipalities and the Swedish includes the Local Job Tracks (Lokala Jobbspår), designed to help jobseekers find employment while at the same time ensuring that employers’ skills needs are met. They are usually addressed to individuals distanced from the labour market and include a chain of combined interventions. The development of Local Job Tracks is guided by a structured process comprising four key phases: (i) the planning phase during which the PES and the municipality identify employer needs, define support measures, and establish recruitment strategies; (ii) the recruitment phase during which jobseeker’s suitability is assessed while employers are actively engaged to carry out recruitment activities; (iii) the implementation phase which focuses on delivering the planned interventions. During this phase continuous co‑ordination and monitoring are essential to make adjustments; and (iv) the closing phase which marks the completion of the track, and during which non-hired participants receive follow-up employment support. Additionally, all tracks are assessed to evaluate their outcomes and identify areas for improvement (Arbetsförmedlingen, 2025[7]).
One of the most pressing challenges in the region of Norrbotten, Northern Sweden is tackling the enormous labour shortages. Despite having the lowest unemployment rate in the country, the region faces a rapidly ageing population and a consistently high demand for workers – Swedish, EU as well as third-country nationals – across all sectors including due to high investment inflows.2 To address this, the PES offices in the region3 implement targeted programmes and works in close partnership with employers, municipalities, and other public agencies to attract and support individuals willing to relocate from the South of the country or from abroad and contribute to the local labour market. Other key priorities include strengthening support for the long-term unemployed (especially those out of work for more than 24 months), increasing employment among people with disabilities, and encouraging individuals with low educational attainment to pursue further studies. Jobseekers are rarely met in person, as communication is primarily conducted over the phone; however, if there are concerns about missing or hidden information, an in-person meeting is arranged. Employees can also register with the PES offices.
The portfolio of the office includes:
Prepare and Match: A service designed to help jobseekers quickly move into work or education. Participants choose their provider, and the support they receive is tailored based on their individual distance from the labour market. Providers work within regulated frameworks, and their compensation is performance‑based. The PES determines eligibility and ensures contract compliance.
Professional Training and Education: This helps jobseekers acquire the skills needed in the labour market. A hub in the municipality of Övertorneå has been established to provide training for high-demand occupations.
Internships and Work Experience: Provide hands-on exposure to real job environments, enhancing practical skills and employability.
Support to Start a Business: Assists jobseekers interested in entrepreneurship with advice and resources.
Occupational Rehabilitation: Supports individuals with health-related challenges in returning to the labour market.
Supported Employment: Offers targeted assistance to those with disabilities or significant barriers to employment.
Introduction Programme: Specifically targeted to newcomers, offering language training, civic orientation, work experience, and education at various levels, helping them build the skills needed to enter the Swedish labour market. Participants also receive a small monthly compensation.
In addition to the programmes and interventions above, at Service Centres (Servicekontor), jobseekers can receive help with simpler administrative matters, such as using online services, reporting activities, writing job applications, or accessing and printing forms.
1. In addition to their co‑operation with the PES, municipalities are involved in a wide range of labour market initiatives. Such initiatives include for example youth follow-up and support (known as KAA), activation of individuals on social assistance, local job matching and training programmes in collaboration with employers, as well as language and work introduction programmes for newly arrived migrants.
2. North Sweden has been an Investment Hub including large climate investments spanning multiple sectors, including energy production, space industry, mining, steel production, ICT/technology/data centres, forestry/pulp and paper, automotive/vehicle testing, food production, and tourism in Swedish Lapland.
3. The Norrbotten region encompasses 14 municipalities. The largest PES office is located in Luleå while smaller local offices are spread across the region, often staffed by just 2 to 4 employees, ensuring a local presence and service in this sparsely populated and geographically dispersed area.
In the longer term, the NEA may wish to explore whether certain services – particularly those for disadvantaged groups – could be more effectively delivered through partnerships with specialised external providers. Outsourcing has the potential to improve outcomes by leveraging the expertise and tailored approaches of organisations experienced in addressing the unique needs of these groups. Outcome‑based payment schemes would incentivise providers to deliver high-quality services thereby improving outcomes, while ensuring quality through proper oversight (Langenbucher and Vodopivec, 2022[8]). However, international experience shows that not all outsourcing models successfully meet their intended objectives (see for example the PES in Sweden).6 Challenges include, among others identifying which services should be provided in-house versus which should be outsourced and determining a balance It is also important to monitor and control providers, ensure effective performance monitoring, and design clear and efficient service and reimbursement structures. Given the current reservations among Bulgarian stakeholders – mostly shaped by prior negative experiences with outsourcing, particularly in efforts to serve vulnerable groups –, any consideration of such a model should be gradual, selective, and based on clear evidence of added value. The NEA should thus only consider contracting out some services where it lacks the necessary expertise or resources, while maintaining oversight and ensuring quality standards are met.
The NEA should implement data-driven approaches to continuously assess the effectiveness of its ALMPs. By leveraging detailed data analytics, the NEA can identify which programmes yield the best results, allowing for the optimisation of resource allocation and programme design (Box 3.6). Improving the effectiveness of programmes is essential to support jobseekers in gaining meaningful employment. However, it is observed that many jobseekers withdraw from training courses, often due to a lack of clarity about course expectations or insufficient motivation. Addressing this challenge requires better communication and guidance prior to enrolment, as well as sustained support throughout the programme duration. Enhancing the role of counsellors in providing this support could help reduce dropout rates and increase completion. The potential use of proportionate sanctions for early withdrawal might also be worth exploring. Additionally, counsellors are central to service delivery; hence their insights should be taken into considerations in designing and implementing new programmes.
Counterfactual impact evaluations (CIEs) are a powerful tool to develop a reliable and in-depth understanding of the effectiveness of active labour market policies (ALMPs) and can also be used to determine their cost-effectiveness. By comparing outcomes between those who participated in a programme and a carefully selected control group consisting of people that did not participate in a programme, it becomes possible to accurately estimate the true impact of an intervention: did participants benefit from the intervention (relative to what would have happened had they not participated in the intervention), and to what extent? If these two groups are comparable, one can conclude that any difference between the group of beneficiaries (what happened) and the comparison group (what would have happened) is entirely due to the policy and not to any other factors because, in the absence of this policy, these two groups would have evolved exactly the same way. Such evaluations provide evidence‑based insights that inform policy decisions, ensuring that public resources are allocated to the most effective ALMPs.
Depending on the context, different key methods exist to conduct counterfactual impact evaluations including randomised controlled trials (RCTs) and quasi‑experimental techniques. Carrying out CIEs requires collecting and/or using data that are already available, such as administrative or survey data. Administrative data come in general from public institutions and may be very large and rich but are not in general collected for research purposes nor are easily available for research purposes. They may require extensive work to clean and organise them, but they offer a far greater population coverage that is often crucial to conduct CIEs. Survey data come from experimental surveys that can be conducted as part of an RCT but can also come from observational surveys, like social surveys, that have nothing to do with the evaluated policy. Other types of research including process evaluations could complement CIEs to assess whether the evaluated policy was indeed implemented as intended.
The prerequisite of any successful CIE of ALMPs is the availability of the necessary data. The aim of ALMPs is to improve labour market participants’ chances of finding employment, increasing their earning capacity, but they can also aim to contribute to social outcomes (e.g. better health, less dependence on social benefits, etc.). As such, to conduct CIEs of ALMPs, detailed information is needed on unemployment spells, including the duration of unemployment, entry and exit dates, and in-depth information on ALMP participation during unemployment. In addition, background information on the individual is required, including both previous employment history before becoming unemployed, and a wide set of personal characteristics, such as age, gender, education etc. Finally, data is needed on the outcomes after participation (or non-participation) in the ALMP: Did jobseekers find employment, and if yes, how rapidly and what types of employment? Is it possible to observe other outcomes, such as health outcomes? Considering the wide range of information that is needed for CIEs, it is usually necessary to link data from several administrative registers to evaluate ALMPs (OECD/European Commission, 2025[9]).
The NEA possesses a solid dataset, but additional data are needed to be able to effectively evaluate ALMPs (Figure 3.10). In Bulgaria, the PES register contains data on jobseekers’ characteristics, ALMP participation, unemployment spells and the reasons for leaving the unemployment register. The Social Assistance Agency register and the register of the National Social Security Institute, which can be linked to the PES data, complement the PES register with information on social benefits (notably social assistance payments) and unemployment benefits, respectively. Furthermore, the employment register, which is held by the National Revenue Agency, includes information on labour market outcomes of ALMP participants and other jobseekers, such as employment contracts, profession and remuneration. However, in Bulgaria, as in many other countries, data from these registers is not fully linked. For example, in Bulgaria, the PES register is not systematically linked with the tax register.
Source: Authors’ compilation based on work undertaken as part of the OECD-EC project on policy impact evaluation through the use of linked administrative and survey data.
An CIE of the NEA’s professional development counselling services – one of its core ALMPs – conducted as part of this project aimed to rigorously assess their effectiveness on key employment outcomes, specifically examining reemployment probabilities, job stability, and job quality. The results reveal a short‑term negative impact of NEA’s counselling services on employment outcomes. Counselled jobseekers initially experience a lower probability of reemployment compared to their uncounselled peers – a finding consistent with the “lock-in effect” observed in similar interventions (Figure 3.11, Panel A). This effect, where time spent in counselling temporarily delays job search activities, appears to be transient: within a year of counselling, reemployment rates between counselled and uncounselled jobseekers converge. However, anticipated improvements in job quality – measured by metrics such as contract duration and wage levels – were not observed. On average, counselled jobseekers did not secure more stable or higher-paying jobs compared to uncounselled individuals, suggesting that the time invested in counselling did not result in better job matches (Figure 3.11, Panel B). These findings align with mixed evidence in the literature on job search assistance programmes, which often enhance reemployment rates, and have limited effects on job quality (van den Berg and Vikström, 2014[10]). From a policy perspective, the findings of this impact evaluation highlight the need to rigorously assess counselling methods to ensure that goals are realistic and achievable, thereby preventing unrealistic expectations and minimising potential lock-in effects.
Source: OECD calculations based on Bulgarian National Employment Agency (NEA) microdata.
References
[7] Arbetsförmedlingen (2025), Local Job Tracks (Lokala Jobbspår), https://arbetsformedlingen.se/for-kommuner/overenskommelser-for-samverkan/lokala-jobbspar (accessed on 10 April 2025).
[3] European Commission (2021), PES practices database, MyDYPAlive: Tele-counselling services to jobseekers and employers, https://ec.europa.eu/social/main.jsp?catId=1206&langId=en (accessed on 30 October 2024).
[8] 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.
[6] OECD (2024), Strengthening Active Labour Market Policies in Korea, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/44cb97d7-en.
[4] OECD (2023), Workshop “Digital tools to profile skills”, organised in cooperation with the EU Directorate-General for Structural Reform Support (DG-Reform), and Latvia’s Ministry for Welfare and State Employment Agency, https://www.oecd.org/en/about/directorates/directorate-for-employment-labour-and-social-affairs.html.
[1] OECD (2022), Reaching Out and Activating Inactive and Unemployed Persons in Bulgaria, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/7b91154a-en.
[9] OECD/European Commission (2025), “Counterfactual impact evaluations of active labour market policies: Lessons from using linked administrative data”, OECD Publishing, Paris, https://doi.org/10.1787/bea7860e-en.
[5] OECD/European Commission (2025), Impact Evaluation of Wage Subsidies and Training for the Unemployed in Slovenia, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/47098a5e-en.
[2] Republic of Bulgaria (2023), Employment Promotion Act.
[10] van den Berg, G. and J. Vikström (2014), “Monitoring Job Offer Decisions, Punishments, Exit to Work, and Job Quality”, The Scandinavian Journal of Economics, Vol. 116/2, pp. 284-334, https://doi.org/10.1111/sjoe.12051.
← 1. As of 03 January 2025, there were 5 222 job vacancies reported at the labour offices.
← 2. Intake involves engaging with the jobseekers and collecting relevant information, whereas registration focuses on recording and verifying that information.
← 3. The NEA only submits the unemployment benefit applications to the NSSI, but does not manage the applications, or verify a person’s eligibility.
← 4. The NEA is in the process of renewing the NDB with support from an external contractor, moving from a rigid, traditional database system to a modern, more flexible data infrastructure.
← 5. These groups are: unemployed young people, including those from specialised institutions or receiving social services in residential communities who have completed their education; the long-term unemployed; individuals with primary or lower education and no vocational qualifications; unemployed persons with permanent disabilities; single parents or adoptive parents with children under the age of five; unemployed individuals over 50; and those who have previously served a prison sentence.
← 6. A recent evaluation found that vocational training is more cost-effective for long-term unemployed and disadvantaged groups (such as persons with disabilities, individuals aged 55 and above, and those with low levels of education), whereas private providers of job matching and coaching are not overall more effective than in-house services – while being more expensive.