This chapter examines the career pathways of MIS beneficiaries with a longer-term lens. It focusses on employment stability, occupational mobility, earnings progression, and participation in Active Labour Market Policies (ALMPs). The analysis shows that additional support is needed to help MIS beneficiaries achieve more stable and quality employment outcomes, reinforcing the case for the development of dedicated digital solutions. In Greece, MIS beneficiaries tend to gradually move away from income support. Nevertheless, employment is often unstable, with limited career progression. Participation in ALMPs also remains low. In Belgium, many MIS beneficiaries remain on income support three years after registration, while about one in four enters regular employment, with outcomes varying by region. While employment rates increase over time, career progression remains limited. The Article 60 programme supports activation among the most disadvantaged recipients, but stronger training and job support are needed to improve long-term employment stability and job quality.
AI and Digitalisation for Employment Support in Belgium and Greece
Abstract
Creating pathways into employment for minimum income scheme (MIS) beneficiaries is crucial to ensure sustainable economic and social integration. The European Pillar of Social Rights calls on Member States to combine adequate income support with inclusive labour market measures to guarantee the active inclusion of people excluded from the labour market. In both Greece and Belgium, activation measures play an important role in the overall support provided to MIS beneficiaries. However, compared to other jobseekers, the labour market integration of MIS beneficiaries is particularly challenging due to the multiple and complex barriers they face.
Building on the register-based data infrastructure described in Chapter 2 and the analysis of employability, labour market entry and job characteristics presented in Chapter 3, this chapter takes a longer-term perspective on MIS beneficiaries’ labour market trajectories. It examines career pathways after MIS registration and entry into employment, focussing on employment stability, occupational mobility, earnings progression, and the role of participation in Active Labour Market Policies (ALMPs). Together, Chapters 3 and 4 provide the analytical foundation for the concept of digital career recommendation and job matching tools discussed later on in Chapter 5.
The findings presented in this report have direct implications for the design of digital tools aimed at supporting MIS beneficiaries’ career development. At a general level, such tools need to account not only for initial job matching but also for employment sustainability and support occupational transitions. By identifying typical career pathways and the role of ALMP participation in supporting progression, the analysis in this chapter helps inform how digital tools can recommend appropriate training, reskilling and upskilling opportunities, with a view to promoting more stable and higher quality employment outcomes.
This chapter is structured as follows. Sections 4.2 and 4.3 present the analysis results for Greece and Belgium, respectively, examining the long‑term labour‑market trajectories of MIS beneficiaries and the role of ALMPs. Each section concludes by highlight implications for policy design and provides relevant inputs for the development of digital career recommendation tools.
This chapter analyses the career trajectories of beneficiaries of the Greek GMI beyond their first employment spell. The results reveal several important findings:
Even among those who enter employment, GMI beneficiaries continue to face challenges in sustaining it. Employment spells tend to be short, particularly among groups already facing greater barriers to labour market integration, such as younger beneficiaries, single parents and people with disabilities.
High occupational mobility does not systematically translate into career progression. Although more than half of GMI beneficiaries who enter employment change occupation within three years, most transitions occur between jobs with similar occupational status, while transitions to higher status occupations remain limited.
Earnings progression remains limited, particularly among older beneficiaries. GMI beneficiaries who remain employed tend to experience increases in weekly working hours and wages over time. However, interrupted employment trajectories constrain cumulative earnings growth for many beneficiaries, contributing to persistent income vulnerability, especially among older workers.
GMI beneficiaries registered with DYPA participate in ALMPs at rates comparable to other jobseekers, and more frequently than other vulnerable groups, particularly in subsidised employment programmes. Nevertheless, participation in ALMPs remains low across all groups of jobseekers.
4.2.1. Data used and analytical approach
This chapter builds on the administrative data from the GMI platform, DYPA’s Integrated Information System (IIS) and ERGANI described in Chapter 2. It analyses additional DYPA data on ALMP participation and describes data preparation steps to enable the analysis of individuals’ labour market trajectories over time.
The analyses presented in Sections 4.2.2 and 4.2.3 rely on the construction of longitudinal panel datasets that follow individuals across registers and over time. To this end, for each individual, a monthly panel is constructed on GMI benefits receipt, ALMP participation, unemployment registration with DYPA, and employment status based on ERGANI records.1 This allows tracking individuals’ labour market states and transitions over time in a consistent manner.
Two main panel samples are used in the analysis. The first is a cohort of GMI beneficiaries, defined as all individuals whose first observed GMI application occurred between 1 February 2017 and 31 December 2019. The second is a comparison cohort of jobseekers registered with DYPA, defined as all individuals whose first observed DYPA registration occurred over the same period. Both cohorts are followed over a five‑year period from the date of first entry into the respective system (GMI application or DYPA registration), allowing for a long-term assessment of labour market trajectories.
For the analysis on career trajectories, the sample is further restricted to individuals who transition into employment within two years of entering the cohort (i.e. within two years of first GMI application for the GMI cohort, and within two years of first DYPA registration for the comparison group). This restriction ensures that individuals can be observed for at least three years following entry into employment, which is necessary to analyse outcomes such as employment stability, occupational mobility, and earnings progression over time.
Finally, the analysis presented in Section 4.2.3 uses the panel of jobseekers registered with DYPA to analyse participation in ALMPs and compare MIS beneficiaries with other jobseekers over the observation period. The ALMP participation data covers individuals registered as unemployed between January 2017 and January 2025 and are linked, for each individual in the panel, to the respective unemployment spell. ALMPs can be broadly grouped into: (i) employment incentives (e.g. subsidised employment), (ii) direct job creation programmes (e.g. creation of new jobs), (iii) job retention schemes, (iv) work experience programmes, (v) labour market services, including group counselling and job search support.
4.2.2. Long-term career pathways of MIS beneficiaries
Understanding whether GMI beneficiaries achieve long-lasting labour‑market integration requires an analysis of labour market outcomes beyond initial entry into employment – which constituted the focus of the previous chapter – to examine longer-term trajectories across employment, unemployment, and benefit receipts. This section analyses beneficiaries’ long-term career pathways, focussing first on transitions between GMI receipt, unemployment, and employment; and second, on the quality and stability of jobs after first entering employment.
Figure 4.1 provides an overview of the labour market trajectories of the cohort of GMI beneficiaries, defined as those whose first observed application falls between 1 February 2017 and 31 December 2019, over the four‑year period following their registration. The Sankey diagram is used to illustrate beneficiaries’ pathways and transitions between different labour market states. Each node represents a snapshot of the beneficiaries’ labour market status at specific points in time, while the flows between nodes capture the transitions of individuals across states over time.
Sankey diagram of snapshot states at 12‑month intervals up to four years after GMI entry
Note: The figure shows GMI beneficiaries’ labour-market or welfare state at fixed follow-up points T₀, T₁, T₂, T₃, T₄, corresponding to 0, 12, 24, 36, and 48 months after GMI entry. The sample consists of GMI beneficiaries whose first observed application falls between 1 February 2017 and 31 December 2019. The states are mutually exclusive and defined using monthly administrative records: employment (ERGANI), unemployment registration (DYPA), referring to individuals recorded in the GMI system but neither employed nor registered as unemployed; and exit, referring to individuals no longer observed in any of the three systems (GMI, DYPA, or ERGANI). Employment refers to formal private‑sector employment recorded in ERGANI. As other forms of employment (e.g. public sector, self-employment, informal work) are not fully captured, some individuals classified as “exit” or “non-employed” may be working outside the observed data.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA) and the employment register ERGANI.
Reliance on GMI declines over time, while employment increases, indicating some positive movement towards labour market integration among beneficiaries. At the time of GMI registration (T₀), 47% of beneficiaries are registered with DYPA as unemployed, 42% receive the GMI without being employed nor registered as unemployed with DYPA, and 11% are formally employed, although their household income still falls below the GMI eligibility threshold. Over the four‑year period, the data point to some positive developments for a substantial share of the individuals in the panel: GMI receipt falls sharply, formal employment doubles from 11% in T0 to 22% in T4 and unemployment decreases modestly.
At the same time, these employment gains remain fragile, as evidenced by continuous flows in and out of employment that point to short or unstable job spells, with many individuals moving back into unemployment. Notably, transitions into employment largely occur through registered unemployment, while direct entries into employment from outside the system (e.g. among individuals who have exited the registries) are comparatively rare. This pattern is consistent with DYPA playing a central role as a key intermediary and gateway into jobs for GMI beneficiaries. At the same time, it may partly reflect institutional features of the GMI system, as beneficiaries who are able to work are required to register with DYPA.
While some individuals exiting the GMI scheme successfully transition into employment, a substantial share remain unemployed (40%) or exit the registries altogether (33%) even four years after first receiving GMI. For the latter group, subsequent outcomes are not directly observable. The absence of continued GMI receipt may reflect a range of situations. In some cases, it may indicate transitions into employment not captured in the available data, such as self-employment, informal work, or other forms of work not registered in ERGANI, or migration. However, exits may also reflect other factors, including non-take‑up among eligible individuals, loss of eligibility due to non-compliance with administrative requirements (e.g. registration with DYPA), or transitions into inactivity without continued engagement with support services. As a result, a non-negligible share of beneficiaries may move out of the system without necessarily achieving stable labour market integration.
Career trajectories beyond initial employment
While Figure 4.1 highlights broad movements between states, it does not capture the stability, quality, or progression of jobs held by beneficiaries who enter employment. The remainder of this section therefore zooms in on the labour market trajectories after the first employment spell following registration for the GMI.
It first examines the risk of returning to unemployment after initial employment by occupations and industries. Then, it analyses occupational mobility following the first post‑GMI job, including whether job changes represent upward, lateral, or downward moves in occupational status. Finally, it shows earnings trajectories over time, assessing whether employment translates into sustained income growth and reduced dependence on income support.
Employment duration tends to be short after GMI registration
A central dimension of long-term labour‑market integration is the extent to which beneficiaries can sustain employment spells over time. The curves presented in Figure 4.2 show the probability of remaining in employment over time after individuals enter their first job following registration for the GMI. The “start time” for each individual corresponds to the moment they enter employment, so all beneficiaries begin with a probability of still being employed equal to 1. The curve then traces how this probability declines over time as individuals exit employment, providing a dynamic view of job retention.
The results show that employment retention declines sharply over the months following job entry, particularly during the first year. Overall, within three months of starting their first job, around 79% of beneficiaries remain in employment. This share decreases to 59% after six months and below 50% after nine months. After one year, only 38% of beneficiaries remain continuously employed without interruption. The decline continues over the following months, although at a slower pace. After two years, approximately 22% of beneficiaries remain continuously employed.
Employment stability varies somewhat across the occupations and industries in which GMI beneficiaries find their first job after registering for the GMI. For instance, beneficiaries entering clerical support occupations have above average levels of retention, with around 80% remaining employed after three months and 45% after a year. By contrast, those entering service and sales occupations, and especially elementary occupations, experience more rapid exits from employment. Out of all individuals finding their first employment in elementary occupations, only one‑third of beneficiaries remain in employment after one year. Service and sales occupations display intermediate outcomes, but still substantially lower retention than clerical occupations throughout the observation period.
Probability of remaining employed by month since job start for GMI beneficiaries, by occupation and industry
Note: The y-axis shows Kaplan-Meier estimates of the probability of remaining in employment over time since entry into the first job after GMI registration (Kaplan and Meier, 1958[1]). The sample consists of GMI beneficiaries whose first observed application falls between 1 February 2017 and 31 December 2019, and who started a job at least three years before the end of the five‑year observation window. Occupations (ISCO major groups) and industries (NACE Rev. 2 level 1) are shown for categories covering at least 10% of GMI beneficiaries.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA) and the employment register ERGANI.
While the previous results show that employment retention varies according to firm and job characteristics, it is also informative to examine how employment retention varies across demographic characteristics of GMI beneficiaries, such as gender and age. Figure 4.3 reports the hazard of leaving first employment after applying for the GMI by combinations of age and gender, after controlling for other factors such as occupation and education. A higher hazard is associated with a greater likelihood of leaving employment and therefore lower employment retention.
Figure 4.3 shows that employment retention varies both by age and gender. GMI beneficiaries aged below 29 display the lowest employment retention, with younger women facing slightly lower retention than their male counterparts. This suggests that, once in employment, younger GMI beneficiaries face a higher hazard of leaving employment than older beneficiaries. Among beneficiaries aged 30‑39, employment retention is higher for both men and women compared to their younger counterparts, although the gender gap is the widest within this age group. Among GMI beneficiaries aged 40 and above, the age gradient differs between men and women: while men aged 40‑49 display higher employment retention than men aged 50 and above, the opposite pattern is observed among women.
While these results highlight important differences in employment retention across gender and age groups, understanding the underlying mechanisms would require a more in-depth analysis. Importantly, the model controls for the industry and occupation of the job entered by the GMI beneficiary, implying that differences across gender and age groups do not reflect differential selection into occupations or industries. However, selection into employment may also depend on unobservable characteristics not captured in the analysis, which could partly explain these differences.
Estimated hazard ratios of leaving first employment after applying for the GMI, by gender and age group
Note: The table reports estimated hazard ratios from a Cox model (Cox, 1972[2]) of exit from first employment after registration for the GMI. Reported values are combined hazard ratios for each age – gender group, calculated from the relevant main effects and interaction terms. In addition to age and gender, the model controls for nationality (Greek/non-Greek), education (low, medium, high, other/missing), disability status (yes/no), single‑parent household status (yes/no), occupation (ISCO major group) and industry (NACE). The reference group is comprised of men aged below 29. Values above 1 indicate a higher hazard of exiting employment compared to this reference group, while values below 1 indicate a lower hazard.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA) and the employment register ERGANI.
In addition to age and gender, other individual characteristics of GMI beneficiaries may also affect employment retention. Figure 4.4 shows how retention in first employment after GMI registration varies across a range of characteristics. Somewhat surprisingly, Greek nationals face higher hazards of leaving employment than non-Greek beneficiaries. GMI beneficiaries with disabilities also face higher risks of exiting employment, suggesting that individuals with disabilities may face not only greater difficulties in entering employment, but also in retaining it once employed. A similar pattern is observed among single parents, possibly reflecting the care responsibilities they face. Overall, these results suggest that GMI beneficiaries facing greater barriers to employment continue to face higher risks of leaving employment even after successfully entering the labour market.
Estimated hazard ratios of leaving first employment after applying for the GMI, by various characteristics
Note: The figure shows estimated hazard ratios and confidence intervals from a Cox model (Cox, 1972[2]) of exit from first employment after registration for the GMI. In addition to the characteristics displayed, the model controls for gender, age, occupation (ISCO major group) and industry (NACE). For each characteristic, hazard ratios are expressed relative to the corresponding reference category: non-Greek nationals, low level of educations, individuals without disability, and individuals not living in a single‑parent household. Values above 1 indicate a higher hazard of exiting employment compared to the relevant reference group, while values below 1 indicate a lower hazard.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA) and the employment register ERGANI.
High occupational mobility after initial employment does not translate into upward progression
The short duration of initial employment spells among GMI beneficiaries is accompanied by a high degree of occupational mobility. More than half of those who enter employment experience at least one occupational change within the three years following their first post-GMI job.2 Such mobility may support the accumulation of experience, acquisition of skills, and facilitate movement towards higher-quality employment. However, it may also reflect instability, with individuals cycling between jobs of similar or lower quality.
To better understand the nature of these career moves, occupational transitions are analysed using the ISCO classification. The incidence of occupational change is measured as the share of individuals who hold jobs in more than one four‑digit ISCO occupation during the three‑year observation window, while the nature of these transitions is assessed by comparing the occupational status of jobs held three years after entry with those of the initial job, using the International Socio-Economic Index of Occupational Status (ISEI) (Ganzeboom, De Graaf and Treiman, 1992[3]), which provides a standardised ranking of occupations based on their associated earnings and educational requirements.
Figure 4.5 presents the direction of occupational transitions, comparing the ISEI score of occupations held three years after entry with those in the initial job, by initial occupation.3 Most transitions occur within the same level of occupational status. While such moves may still involve some degree of skill acquisition or adaptation, they do not necessarily translate into clear upward progression within the observation period.
Upward mobility is observed for some beneficiaries. Around 17% of GMI beneficiaries hold jobs with a higher occupational status at the end of the three‑year window, indicating some scope for upward progression, particularly among those initially employed in elementary occupations and service jobs. However, this partly reflects the structure of opportunities: individuals starting at the lower end of the occupational distribution face limited scope for downward mobility and greater scope for moving to occupations with higher status.
Share of GMI beneficiaries by direction of occupational transitions and by initial occupation
Note: Occupation refers to the first job held after GMI registration, classified according to ISCO major groups. Occupational status is defined based on the International Socio-Economic Index of Occupational Status (ISEI), which ranks occupations according to their associated earnings and educational requirements. “Better status”, “same status” and “worse status” refer to the relative position of occupations held three years after entry compared to the initial job. “Not in employment” refers to individuals not observed in registered employment at the end of the three‑year follow-up period. The sample includes GMI beneficiaries with at least three years of follow-up after entry into employment.
Source: OECD calculations based on the employment register ERGANI and Greek Guaranteed Minimum Income (GMI) records.
Downward mobility is of a similar order of magnitude, remaining below 20% for most occupational groups. While not predominant on average, downward transitions are particularly informative when combined with high levels of occupational mobility. Beneficiaries who frequently change occupation and at the same time move to lower-status jobs appear to face greater challenges in sustained progression. These trajectories may signal difficulties in skill matching, job stability, or progression pathways, and point to a need for continued re‑skilling and up-skilling support even after labour market entry.
Figure 4.6 provides a more granular perspective on these dynamics by jointly considering the incidence of occupational change and the risk of downward mobility across detailed ISCO‑2 occupations. This representation helps identify occupations in which beneficiaries are simultaneously more likely to change occupation and more likely to move to lower‑ranked jobs. Such combinations characterise employment pathways marked by high mobility but limited progression, and in some cases by repeated exits from and re‑entries into employment. Occupations such as customer service clerks, sales workers, and personal service workers, stand out in this respect, combining relatively frequent occupational changes (50% or more people change occupation at least once during the three‑year-window), with sizeable shares of downward transitions (27% for sales workers and 39% for customer service clerks).
Finally, low downward mobility among those starting in elementary occupations does not signal better outcomes. It largely reflects the limited scope for further downward movement from already low-ranked jobs. beneficiaries whose first job after GMI registration is in elementary occupations face weaker employment prospects overall, with around 47% no longer in employment after three years.
Occupational mobility incidence, and share of downward transition, GMI beneficiaries by initial occupation
Note: Occupation refers to the first job held after GMI registration, classified according to the ISCO 2‑digit occupational groups. Occupational status is defined based on the International Socio‑Economic Index of Occupational Status (ISEI), which ranks occupations according to their associated earnings and educational requirements. “At least one occupation change” indicates whether individuals experienced one or more occupational transitions during the three‑year follow‑up period. “Lower ranked occupation” refers to cases in which the occupation held three years after entry has a lower ISEI score than the initial job. Bubble size reflects the proportion of GMI beneficiaries entering each occupation. Labels are displayed only for occupations accounting for at least 3% of observations. The sample includes GMI beneficiaries with at least three years of follow‑up after entry into employment. A complete list of occupational mobility incidence and share of transitions by ISCO‑2 occupations is reported in Annex Table 4.A.1 (Annex 4.A).
Source: OECD calculations using the employment register ERGANI and Greek Guaranteed Minimum Income (GMI) records.
Combination of short employment spells and limited upward occupational mobility translates into modest and uneven earnings progression among GMI beneficiaries
While previous sections document transitions into and within employment, analysing earnings progression is essential to determine whether these transitions translate into sustained income growth and reduced economic vulnerability.
Table 4.1 shows that, among GMI beneficiaries who enter employment, job quality, as measured by hours worked and hourly wages,4 improves gradually over time for those who remain employed. Conditional on being employed, average weekly working hours increase from 26 to 29 hours, while real hourly wages rise from EUR 4.4 to EUR 4.7 over the three years following the first post-GMI job. At the same time, average days worked decline from 217 days in the first year to around 164‑169 days thereafter, reflecting the prevalence of short employment spells (see Figure 4.2). While part of the improvement observed between the first and second year may reflect selection effects, i.e. beneficiaries in a worse job-match being more likely to exit employment, employment levels stabilise between the second and third year, as reflected in similar numbers of days worked. Over this period, further gains in working hours and real hourly wages are observed, suggesting that beneficiaries who sustain employment for longer durations also experience more favourable progression in job quality. However, these improvements remain modest and may partly reflect broader wage dynamics, including changes in minimum wages and price levels over the period, rather than solely within-job or within-worker progression.
Labour market indicators by time since first job (GMI beneficiaries)
Note: T0 is defined as the month of the first registered job after GMI enrolment. The sample covers GMI beneficiaries. The analysis is restricted to individuals whose first post-GMI job occurs at least 3 years before the end of the observation window, allowing full 3‑year follow-up.
Source: OECD calculations using the employment register ERGANI, Greek Guaranteed Minimum Income (GMI) records and Harmonised Index of Consumer Prices from the ECB data portal.
These dynamics are also reflected in cumulative earnings trajectories. Among GMI beneficiaries, cumulative earnings increase modestly over the three‑year period following entry into employment. Figure 4.7 illustrates this pattern and places it in context by comparing outcomes with DYPA jobseekers. Even when classified as vulnerable, DYPA jobseekers consistently accumulate higher earnings than GMI beneficiaries.
Cumulative earnings by time since first job, GMI beneficiaries and DYPA jobseekers
Note: For the GMI series, T0 is defined as the first registered job after GMI enrolment. The sample covers GMI beneficiaries only. For the DYPA series, T0 is defined as the first registered job after DYPA registration. The sample includes DYPA jobseekers registered in the same period, excluding individuals receiving GMI. “DYPA (vulnerable)” includes jobseekers identified by DYPA as facing physical, mental or social barriers to labour market participation, including those in special categories. “DYPA (non-vulnerable)” refers to all other registered jobseekers. The samples for the GMI and DYPA series are restricted to individuals whose first job occurs at least three years before the end of the observation window. Earnings are measured cumulatively over the three years following first employment and expressed in euros and real terms (2015 = 100).
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA), the employment register ERGANI and Harmonised Index of Consumer Prices from the ECB data portal.
Earnings trajectories also differ substantially across demographic groups among GMI beneficiaries (Figure 4.8), and broadly reflect the patterns observed in the wider labour market.5 A gender gap in cumulative earnings persists and widens throughout the follow-up period with men consistently accumulating higher earnings than women. After three years, men earn approximately EUR 10 400 cumulatively, compared to around EUR 9 000 for women. Earnings trajectories also vary significantly by age at job entry. Beneficiaries aged from 30 to 39 at the time they enter their first job have the strongest earnings progression, reaching over EUR 10 500 after three years. In contrast, older beneficiaries (aged 50 and above) experience substantially weaker earnings accumulation, with cumulative earnings remaining well below those of other age groups.
Overall, while some individuals achieve sustained earnings growth, a substantial share experiences interrupted employment trajectories, limiting their ability to accumulate income over time.
Cumulative earnings by time since first job, by gender and age, GMI beneficiaries
Note: Starting time is defined as the month of the first registered job after GMI enrolment. The sample covers GMI beneficiaries and is restricted to individuals whose first post-GMI job occurs at least three years before the end of the observation window, allowing full follow-up. Earnings are measured cumulatively over the three years following first employment and expressed in euros and real terms (2015=100).
Source: OECD calculations using the employment register ERGANI, Greek Guaranteed Minimum Income (GMI) records and Harmonised Index of Consumer Prices from the ECB data portal.
4.2.3. Participation in ALMPs
ALMPs can improve the matching of jobseekers with employers and support transitions into employment. For example, wage subsidies enable firms to hire jobseekers who may not yet have the required skills, while compensating for the initial period needed to reach full productivity, and training programmes can help jobseekers upskill or reskill to meet the requirements of a new job and support occupational mobility. Overall, ALMPs can help bridge the gap between jobseekers and employers and facilitate entry into employment. However, those furthest from the labour market, who may stand to benefit the most, are often harder to reach and less likely to participate in ALMPs, including in Greece (OECD, 2026[4]).
This section examines the extent to which MIS beneficiaries participate in ALMPs. It considers the types of programmes they access and how their participation compares with that of other jobseekers. To do so, the data used in the previous section – linking information from DYPA, the GMI platform and ERGANI – are augmented with data from DYPA’s IIS on ALMPs. These data contain individual participation in four different ALMP measures: three different forms of employment incentives (subsidised employment, retention schemes of new jobs and work experience programmes6) and direct job creation (creation of new job positions). In addition, the data also include information on participation in ALMP services in form of group counselling. Group counselling covers practical aspects of job search and career development, such as preparing a CV, identifying and presenting skills, using social networks and preparing for interviews. Data on another key ALMP, training, are not included, as they are held in a separate register managed by the external institution Diofantos and were not made available for analysis at the time of carrying out this analysis.
The analysis is based on a sample of jobseekers registered with DYPA whose first observed registration occurred between 1 February 2017 and 31 December 2019. These individuals are followed over a five‑year period from their initial registration. Linking these data with information from the GMI platform allows for the identification of three groups: GMI beneficiaries registered with DYPA, other (non-GMI) jobseekers flagged by DYPA as vulnerable, and all other jobseekers. Comparing these groups provides insights into how ALMP participation among GMI beneficiaries differs from that of other jobseekers who are neither GMI beneficiaries nor classified as vulnerable.
Figure 4.9 shows that GMI beneficiaries registered with DYPA participate in ALMP measures and group counselling at rates similar to other jobseekers, and substantially more often than other (non-GMI) vulnerable jobseekers. However, the figure also highlights that overall participation in ALMPs remains low: fewer than one in ten jobseekers participate in at least one ALMP (excluding training). Participation in group counselling is even lower (below 1%). This pattern is similar across GMI beneficiaries and other jobseekers.
Participation rates in active labour market policies (ALMPs), by jobseeker type
Note: The category “Other ALMPs” includes employment incentives (subsidised employment, retention schemes of new jobs and work experience programmes) and direct job creation (creation of new job positions). The sample consists of individuals who registered with DYPA between 1 Feb. 2017 and 31 Dec. 2019, with participation observed for up to five years after registration. GMI beneficiaries include all jobseekers registered with DYPA who registered for the GMI either before or after their DYPA registration. Other vulnerable jobseekers refer to those flagged as vulnerable by DYPA who are not GMI beneficiaries. “Other jobseekers” includes all remaining jobseekers registered with DYPA.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform and the Greek Public Employment Services (DYPA).
Although ALMP participation rates among GMI beneficiaries are comparable to those of other jobseekers, the types of ALMPs in which they participate vary substantially across groups (Figure 4.10). GMI beneficiaries are more likely than other jobseekers to participate in subsidised employment (similarly to other vulnerable non-GMI jobseekers), but less likely to take part in the creation of new job positions and job retention schemes for new jobs. Given that public works are typically targeted to those furthest from the labour market, such as the long-term unemployed, this pattern may appear surprising. However, evidence from counterfactual impact evaluations suggests that wage subsidies are often more effective than public works in supporting jobseekers access employment (OECD, 2024[5]; OECD/European Commission, 2025[6]), and may therefore provide stronger support for GMI beneficiaries.
Subsidised employment requires matching between jobseekers and employers, taking into account both the needs and barriers of GMI beneficiaries and the needs of employers. The relatively higher participation of GMI beneficiaries in subsidised employment reinforces the need for job matching tools that effectively match GMI beneficiaries with suitable jobs, also in the context of such programmes.
Regarding retention schemes for new jobs, the analysis in the previous section shows that GMI beneficiaries often experience non-linear career trajectories, characterised by occupational changes and alternating periods of employment, unemployment and inactivity. This suggests scope to expand the use of retention schemes of new jobs and, more broadly, in-work support to help GMI beneficiaries remain in employment and reduce repeated spells of unemployment or inactivity.
Share of jobseekers participating in different active labour market policies (ALMPs), by jobseeker type
Note: Participation in multiple ALMPs is taken into account. The sample consists of individuals who registered with DYPA between 1 Feb. 2017 and 31 Dec. 2019, with participation observed for up to five years after registration. GMI beneficiaries include all jobseekers registered with DYPA who registered for the GMI either before or after their DYPA registration. Other vulnerable jobseekers refer to those flagged as vulnerable by DYPA who are not GMI beneficiaries. “Other jobseekers” includes all remaining individuals registered with DYPA.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform and the Greek Public Employment Services (DYPA).
An important question is whether GMI beneficiaries who participate in ALMPs experience different career trajectories from non-participants. While answering this question would require a rigorous counterfactual impact evaluation focussing specifically on GMI beneficiaries, which goes well beyond the scope of this project, the available data allow for a descriptive assessment of whether ALMP participation is associated with better labour market outcomes.
Table 4.2 presents one of the outcomes analysed in the previous section, cumulative earnings, measured up to three years after entry into the first job, for GMI beneficiaries who have participated in an ALMP at some point. It shows that GMI beneficiaries who have participated in an ALMP have higher cumulative earnings over the three years following first employment than non-participants. This provides a first descriptive indication that ALMPs may support GMI beneficiaries’ labour market outcomes. However, as noted above, this difference cannot be attributed solely to ALMP participation, as it may reflect the influence of other confounding factors. A rigorous counterfactual impact evaluation would therefore be required to establish causal effects.
Cumulative earnings by active labour market policy (ALMP) participation status and jobseeker type
Note: Jobseeker type is defined at the individual level as follows: “GMI” includes individuals who at one point received GMI; “Other vulnerable” includes individuals never in GMI but identified as vulnerable at any point; “Other job seekers” includes all remaining individuals. “Years after job” refers to time since the first recorded employment following registration with DYPA. ALMP participants are individuals who participated in at least one ALMP during the observation window. Values represent cumulative earnings over time expressed in real terms (2015 = 100).
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek Public Employment Services (DYPA) and Harmonised Index of Consumer Prices from the ECB data portal.
4.2.4. Reliance on income support gradually declines, yet MIS beneficiaries in Greece still need more tailored support to secure more stable and better-quality jobs
This chapter examined the long-term labour market and welfare trajectories of GMI beneficiaries in Greece, focussing on employment sustainability, occupational mobility, earnings progression and participation in ALMPs. By following beneficiaries over several years after registration for the GMI, the analysis provides insights into whether labour market entry leads to more stable and improving outcomes over time, and where additional support is most needed. The findings presented in this chapter should be interpreted within the context of a significantly improved labour market environment in Greece compared to previous years, characterised by rising employment rates, declining unemployment and sustained labour market recovery. Remaining challenges therefore include both the successful and sustainable labour market integration of the most vulnerable groups and the improvement of the overall labour market performance.
Based on the analysis, the key findings for Greece are as follows:
Reliance on GMI declines substantially over time, alongside a gradual movement into employment. Four years after entry into the scheme, only around one in five beneficiaries continue to rely on GMI, while a growing share transitions into employment, either with or without continued benefit receipt. This highlights both the potential for labour market integration and the need to support those who struggle to make or sustain this transition. Digital tools could support pathways into work, while also assisting those already employed in progressing towards more stable and better-quality jobs.
Most beneficiaries who enter employment do so from registered unemployment. Digital solutions should therefore be closely integrated with DYPA processes and data, supporting job matching and career guidance at key transition points.
Employment spells following GMI registration tend to be short, pointing to the need for support beyond initial job placement. Many beneficiaries experience early exits from employment, particularly during the first year after job entry. These patterns reflect, in part, the characteristics of the jobs and sectors in which beneficiaries are employed but also the barriers and vulnerabilities they face in sustaining employment. Together, these factors the importance of career progression tools that focus not only on access to jobs, but also on employment retention and stabilisation over time.
High occupational mobility does not systematically translate into upward progression. More than half of GMI beneficiaries who enter employment change occupation within three years, but most transitions occur at similar status levels. While some horizontal moves may be associated with re‑skilling, a subset of beneficiaries experience frequent mobility combined with limited or downward mobility. These patterns point to groups that may require more intensive and sustained re‑skilling and up‑skilling support even after labour‑market entry.
Limited employment stability and lack of upward occupation mobility translate into modest and uneven earnings progression, particularly for older beneficiaries. Among those who remain employed, job quality improves gradually in terms of working hours and wages. However, interrupted employment trajectories limit cumulative earnings growth for many beneficiaries, contributing to persistent income vulnerability, especially among older workers.
GMI beneficiaries registered with DYPA participate in ALMPs at rates comparable to other jobseekers and more often than other vulnerable groups, particularly in subsidised employment programmes. Nevertheless, participation in ALMPs remains low across the different groups of jobseekers. Descriptive evidence suggests that beneficiaries who participate in ALMPs tend to experience stronger subsequent earnings outcomes, highlighting the potential for ALMPs to support more stable employment pathways, and the importance of embedding ALMP and training recommendations within digital career guidance tools.
This chapter analyses the longer‑term employment trajectories of Belgian minimum income scheme (MIS) beneficiaries (Revenu d’intégration / leefloon), with a particular focus on their participation in active labour market policies (ALMPs) and, in particular, the Article 60 programme. It complements the analysis in the preceding chapter by moving beyond initial exits from MIS to document how beneficiaries fare over time in the labour market and how programme participation shapes the stability, quality, and progression of their employment.
Understanding these longer-term dynamics is of interest both from policy and data perspective. From a policy perspective, Belgium’s Article 60 programme has been operating for nearly five decades and absorbs substantial CPAS resources. From a data perspective, the availability of CBSS’ linked administrative records spanning a ten‑year intake window creates an opportunity to document medium-term trajectories for an unusually large and nationally representative cohort of MIS beneficiaries – an opportunity that has not previously been taken up systematically at this scale. The analysis therefore aims to inform both the design of digital career recommendation tools (Chapter 5) and broader policy debates on the activation of the most disadvantaged jobseekers in Belgium.
The chapter is organised into three main sections:
Section 4.3.2 charts the medium‑ and long‑run trajectories of MIS beneficiaries, describing patterns of benefit receipt, participation in ALMPs (including Article 60), and transitions between MIS, unemployment insurance benefits and employment over three years. It highlights how these trajectories differ across regions, between socio‑demographic groups and by prior labour market attachment.
Section 4.3.3 examines beneficiaries’ longer-term outcomes in work, documenting their sectoral and occupational mobility, earnings dynamics, and risks of returning to joblessness, and comparing the trajectories of MIS beneficiaries to those of unemployment benefit recipients with similar characteristics.
Section 4.3.4 finally then zooms in on ALMP participation, with a particular emphasis on Article 60, analysing how different participant groups fare after their work placement and comparing them to unemployment beneficiaries who have not participated in Article 60.
4.3.1. Data used and analytical approach
This analysis draws on aggregate tables produced from administrative data compiled by the CBSS, focussing on individuals entering MIS and following their trajectories over one and three years after entry. As in Chapter 3, the tables map recipients’ profiles and pathways across key characteristics such as region, age, gender, household composition, migration background and education. However, the emphasis here is on medium‑term dynamics in benefit receipt, participation in ALMPs and movements into employment.
A distinctive feature of this chapter is the use of the same administrative infrastructure as carried out in Chapter 3 to identify and follow MIS beneficiaries who participate in Article 60 programmes, allowing a systematic comparison of their trajectories with those of otherwise similar beneficiaries who do not enter such schemes. For both groups, outcomes are observed one year and three years after MIS entry, making it possible to assess not only initial exits into employment or other benefits but also the stability of these exits over time, including the risk of returning to MIS or unemployment benefits. As before, the CBSS cross‑tabulations are constructed from linked microdata from CPAS, regional public employment services and national social security registers and are complemented by recent statistical reports and research on Article 60 participation in Belgium.
For this analysis, the CBSS generated purpose‑built cross-tabulations from linked microdata, constructing a cohort of all new individual spells of minimum income receipt (or Article 60 placement) in any quarter between 2011 and 2020. Spanning a ten‑year intake window ensures that the sample is large enough to permit reliable breakdowns by socio-demographic subgroups while protecting the confidentiality of individual records. Outcomes are tracked at two fixed time horizons: four quarters (one year) and 12 quarters (three years) after the start of each spell, enabling an assessment of both short-term transitions and the durability of any change in status over the medium term.
The use of aggregate cross-tabulations rather than individual-level microdata reflects the access conditions under which these data were made available for this project: the CBSS provided pre‑specified tabulations rather than a research extract of the underlying microdata. This approach is well-suited to tracking population-level trajectories and comparing outcomes across clearly defined subgroups, but it carries inherent limitations. Because the analysis is based on grouped summaries, causal inference – for instance, whether Article 60 participation itself improves employment chances relative to a valid counterfactual – is not possible. Patterns reported in the comparison between Article 60 and non-Article 60 unemployment benefit recipients are therefore descriptive: they document differences in observed outcomes between two groups that differ in their prior trajectories, rather than estimating a programme effect.
Complementary insights are drawn from secondary sources, including statistical reports from the Federal Public Service for Social Integration (SPP IS/POD MI), regional monitoring reports published by the VDAB, Le Forem, and Actiris, and published academic research on MIS trajectories and Article 60 in Belgium.
4.3.2. Long-term career pathways of MIS beneficiaries
Longitudinal evidence on MIS beneficiaries in Belgium and other OECD countries shows that entry into social assistance often marks the beginning of a complex, multi‑year sequence of benefit receipt, ALMPs and intermittent work, rather than a short, one‑off spell (Immervoll, Jenkins and Königs, 2015[7]; Königs, 2017[8]). Evidence from Belgium is consistent with this picture: a longitudinal study tracking a cohort of social assistance entrants in 2004‑2005 over four years found that while the majority of beneficiaries ultimately exited the scheme, only around 12% achieved durable employment within two years of entry, and repeated re‑entry following short work episodes was common, particularly among those with a migration background (Carpentier, Neels and Van den Bosch, 2017[9]; Carpentier, Neels and Van den Bosch, 2014[10]). Many recipients experience repeated movements between MIS, unemployment benefits and low‑paid or unstable jobs, reflecting both structural barriers in the labour market and the limited capacity of existing instruments to secure durable exits from social assistance.
Against this backdrop, this chapter provides updated evidence on the trajectories of MIS beneficiaries in Belgium, examining how new entrants progress over the medium term and the extent to which they move into employment, contributory benefits or other activities three years after first claiming support.
Figure 4.11 summarises how new MIS entrants move across main labour‑market and benefit states, looking at their benefit and employment status at two points in time after entry (T0): one year later (T1) and three years later (T2). By construction, at T0 (the moment of entry), all people in the sample are in the MIS scheme, i.e. the dark-blue MIS bar reaches 1. By T1 (one year after entry), the dark‑blue MIS bar has shrunk, showing that around half of the cohort still depends on minimum income benefits, while the remaining half has moved into other situations. About 14% of former MIS beneficiaries are at T1 in employment, followed by roughly 9% who are receiving unemployment benefits (UB), and around 6% who participate in the Article 60 programme. Together, these figures indicate that a non-negligible share of former MIS beneficiaries has either entered employment after one year, or is covered by insurance‑based unemployment benefit schemes.
By T2 (three years after entry), the MIS share has contracted further to about one‑third, signalling a gradual but incomplete exit from MIS for the cohort. The shares in unemployment insurance (12%) and employment (25%) grow compared to T1, indicating that more beneficiaries have transitioned into contributory benefits or work over the medium term. At the same time, the “Other” segment expands, suggesting that an increasing fraction of the cohort moves into states outside both MIS and standard employment or unemployment insurance, such as inactivity, other social programmes, or unobserved statuses.
Activity status of new MIS spells, 4 quarters (T1) and 12 quarters (T2) after registering, 2011-2020
Note: T0 denotes the quarter in which individuals first enter MIS. T1 shows their main status one year after entry, and T2 three years after entry. Shares are expressed as proportions of the original MIS entry cohort and sum to 1 (100%) at each point in time. “MIS” refers to continued receipt of minimum income, “Art 60” to participation in Article 60 programmes, “UB” to receipt of unemployment insurance benefits, “Work” to registered employment, and “Other” to all remaining observed statuses (including inactivity and other benefits).
Source: OECD calculations based on administrative data received from the Crossroads Bank for Social Security (CBSS).
Three years after registration, persistence in MIS declines markedly across all groups compared with the one‑year benchmark (see Chapter 3), but important differences across demographic profiles remain (Figure 4.12). Overall, around one‑third of beneficiaries are still receiving the MIS after three years (as indicated by the red line), with some gaps narrowing and others becoming more pronounced over time:
Regional disparities persist, although at lower levels than one year after registering for MIS. Beneficiaries in Brussels (40.5%) and Wallonia (34.6%) continue to display higher persistence on MIS than the national average, while those in Flanders (23.2%) are substantially less likely to remain in the scheme. Compared with the one‑year results, the relative situation in regions is unchanged, but the gap between Brussels and Flanders remains sizeable even three years after entry.
Gender differences remain modest but slightly more pronounced than after one year. Women are more likely to stay on the MIS three years after registration (34.9%) than men (29.6%). While the difference was limited after one year, it appears to widen slightly over time.
Educational attainment continues to be strongly associated with lower longer-term persistence. Individuals with at below upper secondary education (ISCED 0‑2) are more likely to remain on the MIS after three years (34.3%) than those with upper secondary or post-secondary non-tertiary qualifications (ISCED 3‑4) (25.2%) or tertiary education (ISCED 5‑8) (22.1%). Compared with the one‑year benchmark, educational gradients appear to strengthen over time, as higher-educated recipients exit the scheme more rapidly.
Age‑related differences become less marked three years after entry. Young adults aged 18‑29 still show slightly higher MIS persistence (33.6%) than other age groups, but rates are broadly similar across those aged 30‑39 (31.0%), 40‑49 (30.0%) and 50‑64 (30.5%). Compared with the one‑year results, age gradients flatten over time, suggesting more similar long-term trajectories across age groups once initial exits have occurred.
Further analysis also revealed that:
Household composition remains a modest differentiating factor, with persistence rates converging somewhat compared with one year after entry. Beneficiaries living with others (45.0%) and those living with dependent children (43.2%) show the highest persistence after three years, while individuals living alone remain slightly less likely to stay in the scheme (40.2%). The relatively lower persistence observed among single parents after one year is no longer evident at the three‑year horizon.
Differences by migration background remain pronounced. First-generation foreign-born recipients continue to exhibit the highest persistence in the MIS (48.2%), followed by second-generation recipients (44.4%) and first-generation native‑born individuals (44.0%). In contrast, third-generation or later recipients are less likely to remain on the MIS after three years (37.2%). While persistence declines for all groups compared with the one‑year benchmark, gaps by migration background remain substantial over time. These patterns are consistent with longitudinal evidence for Belgium showing that migrants take longer to exit social assistance, face higher re‑entry risks after short work episodes, and are less likely to achieve durable employment – findings that hold even after accounting for educational attainment and household composition (Carpentier, Neels and Van den Bosch, 2017[9]).
Share of MIS beneficiaries who remain in MIS three year after registering, all new spells between 2011-2020
Note: The horizontal line indicates the population share of beneficiaries who remain in the minimum income scheme (MIS) after three years. Educational attainment is based on International Standard Classification for Education (ISCED) codes (0‑2, 3‑4 and 5‑8 respectively).
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS).
Three years after registering for the MIS, transitions into regular employment increase substantially compared with the one‑year benchmark, though also here marked differences across groups persist (Figure 4.13). On average, around one‑quarter of recipients have entered regular employment by this point, with outcomes continuing to vary strongly by region, education and age, while gender differences remain limited. These patterns closely mirror those observed for persistence in MIS (Figure 4.12).
Regional disparities widen over time. In Flanders the share of beneficiaries who access regular employment within three years is the highest (33.2%). In contrast, labour market integration rates of former MIS remain considerably lower in Brussels (21.6%) and Wallonia (20.8%). Compared with one year after entry, all regions record higher employment rates, but the relative advantage of Flanders becomes more pronounced over the longer horizon.
Educational attainment remains a strong predictor of successful labour market integration, with gaps increasing over time. Nearly four in ten beneficiaries with higher education (ISCED 5‑8, 39.1%) transition into regular employment within three years, compared with 32.1% among those with mid-level qualifications (ISCED3‑4) and 22.0% among those with the lowest levels of education (ISCED 0‑2). While education-related differences are already visible after one year, they widen further as higher-educated recipients continue to exit the MIS more rapidly.
Gender differences remain small. Men (25.8%) are only slightly more likely than women (23.9%) to enter regular employment three years after registration, suggesting that gender plays a limited role in shaping employment outcomes over both the short and medium term.
Employment rates of MIS beneficiaries by characteristic, after 12 quarters, all new spells between 2011-2020
MIS: minimum income scheme.
Note: The horizontal line indicates the population share of beneficiaries who are in regular employment after three years. Educational attainment is based on International Standard Classification for Education (ISCED) codes (0‑2, 3‑4 and 5‑8 respectively). Employed in formal employment.
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS).
4.3.3. Employers who hire MIS beneficiaries are predominantly in the service sector
Service sector jobs are the main entryway to employment for MIS beneficiaries
The sectoral distribution of jobs taken up by MIS beneficiaries remains strongly service‑oriented both one year and three years after registration, although some shifts emerge over time (Figure 4.14). At both horizons, employment is concentrated in service activities, with only limited participation in construction and goods-producing sectors.
One year after entry, employment among former MIS recipients is dominated by “other services” and administrative and support services, together accounting for a large majority of observed job matches. Three years after entry, this overall pattern persists, but the relative weight of sectors evolves slightly. The share of employment in “other services” increases, while administrative and support services represent a smaller, though still substantial, proportion of placements. Trade, transport and hospitality remain an important source of employment at both horizons, with only modest changes over time. By contrast, construction and industry and goods production continue to absorb only a small minority of MIS beneficiaries, even three years after registration.
These patterns suggest that while a growing number of beneficiaries enter employment over time (Figure 4.11), longer-term integration does not substantially diversify the sectors in which jobs are found, though a small increase in the share of other services can be observed. Instead, transitions into work continue to rely heavily on service activities characterised by relatively low formal entry requirements and high labour turnover. The persistent prominence of administrative and support services points to the sustained role of intermediary functions such as temporary agency work, cleaning and facility management in facilitating labour market entry, particularly in the earlier phases following registration.
Compared with the structure of employment in the Belgian economy as a whole, MIS beneficiaries remain overrepresented in administrative and support services and underrepresented in construction and industry at both time horizons. The limited expansion into goods-producing sectors over time indicates that barriers related to skills, certification requirements or working conditions are not fully overcome even several years after entry into the MIS. These patterns closely mirror those documented in Chapter 3 at the one‑year horizon, confirming that the sectoral footprint of MIS employment is not simply an artefact of short observation windows: concentration in service activities with low formal entry requirements persists over the medium term. The reliance on administrative and support services, which notably include temporary agency work and cleaning, is consistent with the broader literature on minimum income beneficiaries in Belgium, which highlights these sectors as the primary gateway to employment while noting their association with short contract durations and limited upward mobility (Carpentier, Neels and Van den Bosch, 2014[10]).
Sectors of employment 12 months and 36 months after MIS (2011‑2020), benchmarked against 2024 national averages
Note: Data for former minimum income scheme (MIS) spells between 2011-2020 and tracks their status 1 year and 3 years after registration.
Source: OECD calculations based on administrative data from CBSS, and StatBel Statistics on establishment units https://statbel.fgov.be/en/themes/datalab/statistics-establishment-units for the national average.
… and most MIS beneficiaries who find employment do manual jobs
The type of jobs accessed by MIS beneficiaries further underlines the segmented nature of post-MIS employment (Figure 4.15). Both one year and three years after registration, manual occupations account for a large share of employment, particularly among those entering work shortly after exit from the scheme. Over time, however, the composition of employment shifts modestly towards non-manual positions.
Three years after entry, non-manual employment represents a larger share of observed jobs than at the one‑year mark, suggesting some upward movement or delayed access to less physically demanding roles. Civil servant positions remain marginal at both horizons, indicating that transitions into public sector employment are rare for former MIS recipients. Overall, while longer time horizons are associated with higher employment rates (Figure 4.11), the distribution of job types points to continued concentration in lower-skilled and often less secure forms of work, with only small diversification over time.
Types of occupations 12 and 36 months after MIS entry (2011‑2020), benchmarked against 2024 national averages
MIS: minimum income scheme.
Source: OECD calculations based on administrative data from CBSS, and StatBel Statistics on establishment units https://statbel.fgov.be/en/themes/datalab/statistics-establishment-units for the national average.
4.3.4. Participation in ALMPs: Trends and trajectories of Article 60 beneficiaries
Belgium’s labour market policy encompasses a broad suite of ALMPs delivered by regional PES, designed to help jobseekers move into sustainable employment. Across the country, regional PES such as Actiris in Brussels, VDAB in Flanders, Forem in Wallonia, and ADG in the German-speaking Community provide a mix of job search assistance, training and upskilling, employer engagement and recruitment incentives for unemployed workers, including tailored vouchers, counselling, and placement services. Some programmes, such as locally organised training vouchers or employer hiring subsidies, are explicitly targeted at long-term jobseekers or young people or intended to address specific skill gaps in the labour market (e.g. training vouchers and ICT or skills cheques in Brussels).
Research on Belgian CPAS services suggests that, in practice, co‑ordination between local welfare offices and regional employment services remains uneven. CPAS case workers differ significantly in how they assess work readiness and which activation pathways they offer to clients, partly reflecting resource and capacity differences across municipalities (De Wilde and Marchal, 2019[11]).
Within this diverse ALMP ecosystem, the Article 60 programme occupies a central role for recipients of the MIS in particular. Unlike most PES measures, which are generally aimed at the unemployed registered with a regional employment service, Article 60 is specifically embedded in the social assistance framework: it is implemented by CPAS and support MIS recipients by offering subsidised work experience and, crucially, giving them a chance to acquire the social insurance contributions needed to establish eligibility for unemployment benefits. While other ALMPs such as subsidised job incentives or training are important for the job search and upskilling of a broad set of unemployed individuals, Article 60 is one of the main tools available to support and activate the most disadvantaged MIS beneficiaries, helping them move off social assistance and into the unemployment insurance system and, subsequently, into employment. Article 60 is the principal activation instrument available to CPAS, accounting for around 14% of all MIS trajectories in Flanders and approximately 23 000 placements nationally in recent years, by far the largest single activation measure directly operated by welfare offices (Caldarini, 2021[12]; Union des Villes et Communes de Wallonie – Fédération des CPAS, 2023[13]). Given its specific importance for MIS beneficiaries, the remainder of this section focusses on Article 60 when analysing ALMP participation among the MIS population, while acknowledging the complementary role of PES-delivered measures for beneficiaries who have transitioned to unemployment insurance.
Design and eligibility of the Article 60 programme
Belgium’s Article 60 (Art. 60 §7 of the Organic CPAS/OCMW law) is a subsidised work‐experience programme run by local CPAS first introduced in 1976 (Ministère de l'Intégration Sociale, 2024[14]). It allows a CPAS to place a MIS recipient into a job in the public or social economy under a fixed‐term contract, with the aim of rebuilding that person’s contributions and eligibility for unemployment insurance.
Article 60 work can be organised in two different ways. One option is that the beneficiary is employed directly by the municipal welfare centre in roles supporting CPAS operations, including administrative functions, maintenance, community services, etc. Alternatively, the CPAS can place the programme participant for work into a “host” organisation or employer. This can include:
Municipal administrations or local government services other than the CPAS,
Public utilities and water companies,
Hospitals and health services,
Educational institutions (schools, libraries, training centres),
Non-profit associations and civil society organisations or
In some regions, semi-public companies and social economy enterprises.
Also in this case, the CPAS remains the legal employer, “lending” the worker to the “host employer”, with wages and social‐security contributions being fully paid by the CPAS.
The host employer enters into a service agreement with the CPAS, committing to provide structured work experience and, typically, on-the‑job skills development. Payment for the worker’s services is negotiated between the CPAS and the external user. As of 2025, such placements can no longer be free, signalling a policy shift toward cost-sharing for subsidised employment (ADAS, 2021[15]; Bruxelles-J, 2025[16]). The work contract typically lasts until the worker has accrued the minimum days of employment required to qualify for unemployment benefit, usually a maximum of two years, though frequently shorter. In the meantime, Article 60 provides a subsistence wage – usually around the minimum wage, on‐the‐job experience and the contributions needed to build the entitlements that will eventually permit the programme participant to transition off social assistance into the unemployment insurance system.
Article 60 contracts are available only to persons already receiving the CPAS‐administered MIS. In practice, CPAS counsellors refer welfare recipients – often long‐term unemployed individuals or people with disabilities – into Article 60 placements. Recent practice (e.g. in Brussels) also allows Ukrainian refugees under temporary protection who lack other entitlements to take Article 60 positions (safe.brussels, 2025[17]).
In sum, Article 60 is an ALMP embedded within Belgium’s minimum income policy, that requires recipients to work. Importantly, its primary objective is to restore social insurance rights (qualification for unemployment benefits), which are generally higher than social assistance payments and can then act as a springboard back to stable employment (including through access to PES support).
Article 60 participation increased rapidly since the early 2000s
The Article 60 programme experienced rapid expansion in the early 2000s. The number of placements nearly doubled between 2000 and 2004, reflecting both a rise in beneficiary numbers, driven by growing caseloads and a stronger policy emphasis on activation. From 2004 onwards, programme participation stabilised, with annual placements slightly declining and then plateauing at around 19 000 – 20 000 nationally.
Regional differences are pronounced. Wallonia has consistently accounted for 40‑45% of Article 60 placements. This over-representation reflects the region’s persistently higher unemployment and inactivity rates.
Brussels-Capital, by contrast, has seen the most substantial proportional growth. Its share of national placements roughly double from between 2000 and 2008, from 12% to 23%, with growth further accelerating in subsequent years. This trend is linked to an expanding EU migrant and refugee population in Brussels, which drove both subsistence wage caseloads and demand for Article 60 placements.
Annual programme participants, 2000-2023
Note: Regional participation figures were reconstructed from published administrative charts.
Source: SPP Intégration Sociale Baromètre de l’intégration sociale (SPP-IS) – Article 60 §7 placements by region, administrative data, stat.mi‑is.be.
Comparison between UB beneficiaries with and without previous participation in Article 60
The analysis compares unemployment benefit (UB) recipients who entered UB through participation in the Article 60 programme with those who entered through other channels.
UB beneficiaries who previously participated in the Article 60 programme have substantially lower employment rates than those without prior programme participation (Figure 4.17). One year after entering UB, approximately 45% of non-Article 60 participants are employed, compared to only 25% of prior Article 60 participants. This gap persists over time: three years later, around 50% of non-participants are in work versus 35% of prior participants. Correspondingly, beneficiaries with Article 60 experience remain on unemployment benefits at higher rates. Three years after UB entry, roughly 45% of prior participants still receive benefits, compared to 35% of those without Article 60 history. The share in “Other” status (including inactivity or other benefit schemes) is broadly comparable across both groups at approximately 15‑20%.
These patterns suggest that prior Article 60 participation is associated with less favourable labour market trajectories, though the gap narrows slightly between the one‑ and three‑year marks as some prior participants transition into employment. It is important to note that this comparison is by design not a causal impact assessment. Article 60 targets the most disadvantaged MIS recipients: those with weak or absent contribution histories and significant barriers to employment. Meanwhile, “standard” unemployment benefit entrants have demonstrated a stronger prior attachment to the labour market, fulfilling work history requirements. The lower employment rates observed among former Article 60 participants therefore partly reflect this negative selection into the programme rather than any effect of the programme itself. The aim of this study is to document where these two populations end up and thus not serve as a robust impact evaluation.
Note: Data for former minimum income scheme (MIS) spells between 2011-2020 and tracks their status 1 year and 3 years after registration.
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS).
Unemployment benefit recipients with prior Article 60 experience face weaker employment outcomes
The following section compares transitions into employment three years after entering unemployment benefits, comparing recipients with and without prior Article 60 experience:
Regional variation is particularly pronounced. Employment is highest in Flanders (57.8%), followed by Wallonia and Brussels, suggesting that regional labour market conditions strongly shape re‑employment prospects. Prior Article 60 participants show a similar regional ordering, though at a lower level (45.5% in Flanders), with gaps of around 10‑12 percentage points (p.p.) across all regions.
Gender differences are modest among both groups. Men and women without Article 60 experience show near-identical employment rates (51.8% vs. 50.0%), and the same holds for prior participants (36.5% vs. 34.9%). Gender appears to play little role in shaping employment outcomes in either group.
Educational attainment shows a clear positive gradient among standard UB recipients, with higher-educated beneficiaries considerably more likely to find employment after three years. Notably, this gradient is largely absent among prior Article 60 participants, who show similar employment rates across education levels (35‑38%). This suggests that Article 60 may be particularly effective at supporting the labour market participation of lower-educated groups, narrowing the education gap relative to standard UB recipients.
Age emerges as the most consistent factor across both groups. Younger beneficiaries (18‑29) have the highest employment rates in both panels, and rates decline steadily with age. Among those aged 50‑64, only 20.4% of prior participants are employed after three years, compared to 33.2% of standard UB recipients. Unlike the education gradient, the age gradient is not meaningfully reduced among Article 60 participants.
Overall, Article 60 participants show lower employment rates than standard UB recipients across all demographic groups. The regional and age patterns are broadly similar between the two groups. The key exception is education: Article 60 is associated with a compression of the education gradient, pointing to a potential role in boosting employment prospects for lower-educated recipients. Although this advantage does not extend to older workers, where the gap between the two groups persists.
Note: The horizontal line indicates the population share of beneficiaries who are in regular employment after three years. Educational attainment is based on International Standard Classification for Education (ISCED) codes (0‑2, 3‑4 and 5‑8 respectively).
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS).
Two‑in-three former Article 60 participants enter manual employment, while industries show less variance
Among former UB recipients who have found work three years after entry into UB, occupational outcomes differ notably between those who participated in Article 60 programmes and those who did not. Participants in Article 60 programmes show a substantially higher concentration in manual work (approximately 70%) compared to non-participants (around 45%) and the national average (roughly 30%). Conversely, non-manual employment is markedly lower among Article 60 participants (approximately 30%) than among non-participants (about 55%) and the national workforce (around 60%). The proportion entering civil service positions remains relatively small across both groups, while it makes up a significant share among all workers in Belgium on average (Figure 4.19).
Types of occupations 36 months after unemployment benefit entry (2011‑2020), benchmarked against 2024 national averages
Note: Data for former unemployment benefit (UB) includes spells between 2011-2020 and tracks their status three years after registration.
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS) and StatBel Statistics on establishment units https://statbel.fgov.be/en/themes/datalab/statistics-establishment-units for the national average.
The variation in sectoral distribution between the two groups is less pronounced. Indeed, both cohorts show broadly similar patterns of employment across other services, administration and support, trade and transport, industry and goods, and construction sectors (Figure 4.20). Article 60 participants demonstrate marginally lower representation in construction and industry and goods sectors compared to non-participants.
Sectors of employment 36 months after MIS or unemployment benefit entry (2011‑2020), benchmarked against 2024 national averages
Note: Data for former unemployment benefit (UB) includes spells between 2011-2020 and tracks their status three years after registration.
Source: OECD calculations based on administrative data from the Crossroads Bank for Social Security (CBSS) and StatBel Statistics on establishment units https://statbel.fgov.be/en/themes/datalab/statistics-establishment-units for the national average.
Evidence from recent evaluations of Article 60 effectiveness
The available evaluation evidence on Article 60’s labour market effects is limited and points to a mixed picture: Article 60 is effective at achieving its institutional objective of providing access to insurance‑based benfits, notably UB, but its impact on sustained employment in the open labour market appears limited and uneven across groupsIt is also important to note that the comparison groups differ across studies and from the analysis presented in this chapter. Studies in the evaluation literature typically compare Article 60 participants to non-participants within the social assistance population, asking whether participation leads to better outcomes relative to remaining on MIS. The analysis in this chapter takes a different approach: it compares UB recipients who previously participated in Article 60 to those who entered UB through other routes, and is therefore best read as descriptive evidence on where these two populations end up, rather than as an assessment of programme effectiveness.
Quasi‑experimental studies using administrative data confirm that Article 60 raises transitions from social assistance into the unemployment insurance system, rather than directly into employment, with the largest effects among those with weak or absent contribution histories who would otherwise face the greatest difficulties qualifying for UB independently (Le Forem, 2024[18]). Further, an evaluation by the National Employment Office shows that post-Article 60 unemployment trajectories are long and complex: on average, former participants spend more than six years in the unemployment system before achieving a durable exit, with only one in seven leaving within the first year; however, those who do exit unemployment typically do so into employment, suggesting that the programme does eventually serve its bridging function for a share of participants (RVA/ONEM, 2023[19]). A Flemish process evaluation of the reformed TWE‑OCMW scheme, which integrated Article 60 into a broader temporary work experience framework from 2017 onwards, found more positive short-term results, with around 59% of participants employed three months after completing their trajectory, though stricter entry criteria also meant that fewer of the most disadvantaged beneficiaries accessed the programme than under the previous Article 60 arrangement (De Coen, Valsamis and Vanoeteren, 2018[20]). Together, these findings suggest that Article 60s effectiveness is sensitive to the quality of the placement, the support provided during and after the programme, and the characteristics of participants.
The patterns presented in this study are descriptive and should be interpreted with care. Rather than estimating a programme effect, they describe the employment trajectories of former Article 60 participants after they have entered unemployment benefits. Within this group, employment rates three years after entering UB are higher among younger participants and decline sharply with age, consistent with patterns documented in Belgian administrative data (Le Forem, 2024[18]; Irfam, 2023[21]). Whether these differences reflect characteristics of the participants themselves or the nature of their Article 60 experience cannot be determined from the data available.
Qualitative evaluations emphasise the importance of varying local practices: structured placements with supervision and skills yield better trajectories than internal or sheltered tasks, implying that devoting greater financial resources to the implementaiton of the programme likely makes a real difference. Many jobs involve low-skilled services, explaining concentrations in manual roles (Le Kiosque ASBL, 2019[22]).
Qualitative evidence further suggests that the quality and content of Article 60 placements vary substantially across CPAS: structured placements in external organisations with active supervision and skills development are associated with better employment trajectories than internally focussed tasks, while insufficient accompaniment is linked to revolving-door dynamics in which participants cycle back to social assistance shortly after their contract ends (Caldarini, 2021[12]). An evaluation of the reformed Flemish TWE‑OCMW framework similarly noted that investment in coaching and employer-facing support made a measurable difference to post-programme employment rates (De Coen, Valsamis and Vanoeteren, 2018[20]). These implementation dimensions cannot be captured with administrative data but are important context for interpreting the sectoral and occupational outcomes documented above.
4.3.5. Improving pathways from Article 60 to stable labour market integration is crucial, as is the development of digital tools that support employment sustainability and career transitions
This chapter documents the longer-term trajectories of MIS beneficiaries in Belgium, with a particular focus on the role of Article 60 in shaping subsequent labour market outcomes. Three main conclusions emerge.
First, exits from MIS are gradual and, for some groups, incomplete. Three years after entry, around one‑third of beneficiaries remain on the scheme, and many others cycle between social assistance, unemployment benefits and unstable employment. While employment rates increase over time, especially among higher-educated beneficiaries, long-term persistence remains high for groups facing cumulative factors related to migration background, household composition and regional labour market conditions.
Second, Article 60 plays a central role in the activation of the most vulnerable MIS recipients, but its contribution is best understood as institutional rather than employment-centred. The programme is effective at moving beneficiaries from social assistance into unemployment insurance and at restoring contribution records. At the same time, former participants display weaker employment outcomes once in the unemployment system, reflecting both negative selection into the programme and limited progression into stable, higher-quality jobs.
Third, the structure of post-MIS employment remains highly segmented. Former beneficiaries, and especially those with Article 60 experience, continue to concentrate in manual occupations and service‑sector activities characterised by low entry requirements and high turnover. While some occupational upgrading occurs over time, transitions into non-manual roles, goods-producing sectors or public employment remain rare even three years after entry.
From a policy perspective, these findings point to the importance of strengthening the link between Article 60 placements and subsequent labour market opportunities. This includes improving the quality and relevance of work experience, tailoring placements more closely to local labour demand, and combining Article 60 with training and post-placement support. Without such complementary measures, the programme risks functioning primarily as a mechanism for benefit transitions rather than as a lever for durable labour market inclusion. These conclusions are broadly consistent with the wider Belgian evidence base. Tracking data from the National Employment Office and research on CPAS activation alike point to long post-programme spells, highly variable outcomes depending on the quality of support, and a persistent risk of cycling between benefit states for the most disadvantaged (Bogaerts et al., 2010[23]; Sempels, Goffin and Vanoeteren, 2023[24]; RVA/ONEM, 2023[19]). The present analysis adds medium-term trajectory evidence at national scale to this picture, reinforcing the case for more systematic monitoring of post-programme outcomes and for strengthening the link between Article 60 placements and subsequent labour market support.
A final limitation of this analysis concerns data availability on the content and quality of activation measures, in particular training and Article 60 placements, which also has implications for the future use of data-driven job matching and career recommendation tools. While the CBSS-linked administrative data allow participation in Article 60 and broad labour market states to be tracked over time, they contain only limited information on the tasks performed, skills acquired, training content, supervision arrangements or employer exposure associated with individual placements, though some of it can be retrieved from ONSS records. Similarly, information on training participation is fragmented across regional public employment services and is not consistently integrated with MIS trajectories at the national level. These gaps constrain the ability to assess which programme features support durable employment and to distinguish programme effects from selection into different activation pathways.
From a forward-looking perspective, richer and more integrated data are a precondition for designing effective job matching and career recommendation tools for minimum income recipients. Tools that aim to guide beneficiaries towards suitable occupations, training pathways or employers require detailed information on prior work experience, skills gained during programmes such as Article 60, and subsequent employment outcomes. Without such data, algorithmic or rule‑based matching risks reproducing existing segmentation by repeatedly directing disadvantaged beneficiaries towards a narrow set of low-quality jobs. Strengthening data linkages between CPAS, regional employment services and social security records would therefore not only improve evaluation capacity but also enable more personalised, evidence‑based activation strategies that support progression rather than repeated cycling through temporary or low-skilled employment.
Lastly, it should be noted that Belgian unemployment benefit eligibility has recently undergone a significant reform. Under the federal government agreement of January 2025, unemployment benefits will be capped at a maximum of 24 months for most recipients (with an exemption for those aged 55 and over), reforming Belgium’s long-standing system of time‑unlimited UB. This reform has potential implications for the role that social assistance benefits play in the Belgian income support system, and – more specifically – for Article 60 as a pathway to income security. First, MIS, which so far has played a rather residual role in the Belgian benefit system, will gain in importance, as a potentially larger number of long-term UB recipients will flow into the system upon reaching the end of their UB entitlement period. Along with rising recipient numbers, also the profiles and specific support needs of MIS beneficiaries will change, which may call for a revised targeting of the Article 60 programme. Second, in light of evidence that former Article 60 participants who move into the UB system often remain on UB for several years, a significant share of them may – under the new rules – exhaust their UB entitlements before finding regular employment, and hence return to the MIS. In other words, there is a risk that certain jobseekers may end up “cycling” between the MIS, Article 60 participation and UB receipt without managing to find regular employment and achieving self-sufficiency. However, as the reform is recent and its effects have yet to materialise, additional evaluation may be needed to assess its impact on MIS beneficiary numbers and profiles, benefit trajectories, and the effectiveness of Article 60 as a tool to activate jobseekers and support their transition into regular employment.
References
[15] ADAS (2021), «Article 60» : Un vrai contrat de travail ou un sous-statut?, https://www.adasasbl.be/2021/01/28/fiche-info-article-60-un-droit-une-obligation/.
[23] Bogaerts, K. et al. (2010), Activering bij werkloosheid en recht op maatschappelijke integratie [Activation measures for the unemployed and the right to social integration].
[16] Bruxelles-J (2025), Le CPAS et le contrat de travail article 60/61, https://www.bruxelles-j.be/droits-sociaux/obtenir-de-laide-du-cpas/le-cpas-propose-des-contrats-de-travail-particuliers-les-articles-60-61/.
[12] Caldarini, C. (2021), Avantages et inconvénients des contrats « article 60 » pour les travailleurs étrangers, https://www.irfam.org/avantages-et-inconvenients-des-contrats-article-60-pour-les-travailleurs-etrangers/.
[9] Carpentier, S., K. Neels and K. Van den Bosch (2017), “Do First- and Second-Generation Migrants Stay Longer in Social Assistance Than Natives in Belgium?”, Journal of International Migration and Integration, Vol. 18/4, pp. 1167-1190, https://doi.org/10.1007/s12134-017-0519-z.
[10] Carpentier, S., K. Neels and K. Van den Bosch (2014), “How Do Exit Rates from Social Assistance Benefit in Belgium Vary with Individual and Local Agency Characteristics?”, in Safety Nets and Benefit Dependence, Emerald Group Publishing Limited, https://doi.org/10.1108/s0147-912120140000039004.
[25] Cazes, S., A. Hijzen and A. Saint-Martin (2015), “Measuring and Assessing Job Quality: The OECD Job Quality Framework”, OECD Social, Employment and Migration Working Papers, No. 174, OECD Publishing, Paris, https://doi.org/10.1787/5jrp02kjw1mr-en.
[2] Cox, D. (1972), “Regression Models and Life-Tables”, Journal of the Royal Statistical Society Series B: Statistical Methodology, Vol. 34/2, pp. 187-202, https://doi.org/10.1111/j.2517-6161.1972.tb00899.x.
[20] De Coen, A., D. Valsamis and V. Vanoeteren (2018), Activering van leefloongerechtigden via tijdelijke werkervaring: Eindrapport. IDEA Consult [Activating recipients of the living wage through temporary work experience: Final report. IDEA Consult], https://publicaties.vlaanderen.be/view-file/28921.
[11] De Wilde, M. and S. Marchal (2019), “Weighing up Work Willingness in Social Assistance: A Balancing Act on Multiple Levels”, European Sociological Review, Vol. 35/5, pp. 718-737, https://doi.org/10.1093/esr/jcz028.
[3] Ganzeboom, H., P. De Graaf and D. Treiman (1992), “A standard international socio-economic index of occupational status”, Social Science Research, Vol. 21/1, pp. 1-56, https://doi.org/10.1016/0049-089x(92)90017-b.
[7] Immervoll, H., S. Jenkins and S. Königs (2015), “Are Recipients of Social Assistance ’Benefit Dependent’?: Concepts, Measurement and Results for Selected Countries”, OECD Social, Employment and Migration Working Papers, No. 162, OECD Publishing, Paris, https://doi.org/10.1787/5jxrcmgpc6mn-en.
[21] Irfam (2023), Avantages et inconvénients des contrats « article 60 » pour les travailleurs étrangers., https://www.irfam.org/avantages-et-inconvenients-des-contrats-article-60-pour-les-travailleurs-etrangers/.
[1] Kaplan, E. and P. Meier (1958), “Nonparametric Estimation from Incomplete Observations”, Journal of the American Statistical Association, Vol. 53/282, pp. 457-481.
[8] Königs, S. (2017), “Micro‐level dynamics of social assistance receipt: Evidence from four European countries”, International Journal of Social Welfare, Vol. 27/2, pp. 146-156, https://doi.org/10.1111/ijsw.12279.
[18] Le Forem (2024), Analyse du profil et de l’insertion des demandeurs d’emploi connus du Forem et ayant bénéficié des dispositifs d’insertion dans l’emploi des CPAS (article 60, 61 et économie sociale) : Sorties 2021-2022., https://www.leforem.be/content/dam/leforem/fr/documents/chiffres-et-analyses/analyses/analyse-profil-insertion-article60-61.pdf.
[22] Le Kiosque ASBL (2019), La remise au travail « article 60 »., https://kiosqueasbl.be/wp-content/uploads/2020/11/remise-au-travail-article-60.pdf.
[14] Ministère de l’Intégration Sociale (2024), Loi du 26 mai 2002 concernant le droit à l’intégration sociale (Coordination officieuse – mise à jour le 11 avril 2024).
[4] OECD (2026), Strengthening Individualised Support for Jobseekers Furthest from the Labour Market in Greece, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/e732645e-en.
[5] OECD (2024), Impact Evaluation of Training and Wage Subsidies for the Unemployed in Greece, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/4b908517-en.
[6] 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.
[19] RVA/ONEM (2023), Het traject van werklozen toegelaten na tewerkstelling bij de OCMW’s (artikel 60§7) [The procedure for unemployed persons admitted following employment with the OCMWs (Article 60(7))].
[17] safe.brussels (2025), Brussels help Ukraine - What is Article 60?.
[24] Sempels, Y., K. Goffin and V. Vanoeteren (2023), Activering van leefloongerechtigden: Noden, aanpak en lokale samenwerking [Empowering those entitled to income support: Needs, approach and local cooperation], https://publicaties.vlaanderen.be/view-file/64763.
[13] Union des Villes et Communes de Wallonie – Fédération des CPAS (2023), Insertion socioprofessionnelle: Articles 60/61, https://www.uvcw.be/insertion/accueil.
Share of GMI beneficiaries by occupational transition status three years after entry into employment, by initial ISCO 2‑digit occupation group
Note: Occupation refers to the first job held after GMI registration, classified according to ISCO 2‑digit groups. “At least 1 occ. change” indicates the share of individuals who experience at least one occupational transition during the three‑year follow-up period. Status levels are defined based on the International Socio-Economic Index of Occupational Status (ISEI), which ranks occupations according to their associated earnings and educational requirements. “Same”, “better” and “worse” skill levels refer to the relative position of occupations held three years after entry compared to the initial job. “Not in employment” refers to individuals not observed in registered employment at the end of the follow-up period. “No score” indicates cases where skill level comparisons could not be computed due to missing ISEI values. The sample includes GMI beneficiaries with at least three years of follow-up after entry into employment.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform and the employment register ERGANI.
Share of GMI beneficiaries by occupational transition status three years after entry into employment, by selected individual characteristics
Note: “At least 1 occ. change” indicates the share of individuals who experience at least one occupational transition during the follow-up period. Status levels are defined using the International Socio-Economic Index of Occupational Status (ISEI), which ranks occupations according to their associated earnings and educational requirements. “Same”, “better” and “worse” status levels refer to the relative status of the occupation held three years after entry compared with the initial job. “Not in employment” refers to individuals not observed in registered employment at the end of the follow-up period. “No score” indicates cases where status comparisons could not be computed due to missing ISEI values. Education levels are defined according to ISCED: low education corresponds to ISCED 0‑2, medium education to ISCED 3‑4, and high education to ISCED 5‑8. The sample includes GMI beneficiaries with at least three years of follow-up after entry into employment.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, and ERGANI registry.
Share of GMI beneficiaries by direction of occupational transitions, by initial occupation
Note: Skill levels are defined based on the International Socio-Economic Index of Occupational Status (ISEI), which ranks occupations according to their associated earnings and educational requirements. “Better”, “same” and “worse” skill levels refer to the relative position of occupations held three years after entry compared to the initial job. “Not in employment” refers to individuals not observed in registered employment at the end of the three‑year follow-up period. The GMI sample includes with at least three years of follow-up after entry into employment. DYPA sample includes jobseekers with the same follow up period, and excluding individuals receiving GMI.
Source: OECD calculations based on data from the Greek Guaranteed Minimum Income (GMI) platform, the Greek public employment service (DYPA) and the employment register ERGANI.
← 1. ERGANI records do not capture all forms of employment, including parts of the public sector, self-employment and informal work.
← 2. Occupational change is defined as a transition between jobs with different four‑digit ISCO codes. In this context, experiencing at least one occupational change implies that an individual holds jobs in at least two distinct occupations during the observation window.
← 3. Further analyses disaggregating the results by individual characteristics and jobseeker type are presented in the Annex. Annex Table 4.A.2 presents the same occupational transition indicators by selected individual characteristics, showing that younger beneficiaries are the most occupationally mobile, while gender differences are limited. Annex Figure 4.A.1 reports the corresponding distributions for the overall GMI population and for DYPA jobseekers (distinguishing vulnerable and non‑vulnerable groups), GMI beneficiaries and DYPA vulnerable jobseekers show broadly similar patterns of occupational mobility.
← 4. Based on the OECD Job Quality Framework, in which earnings quality is a key determinant of overall job quality (Cazes, Hijzen and Saint-Martin, 2015[25]).
← 5. Gender wage gaps persist in Greece, with women earning around 14‑15% less than men in recent years (2018-2024). Earnings differences by age follow standard life‑cycle patterns, with both younger and older workers earning less than prime‑age workers. See OECD, Gender wage gap and Wage gap by age, (OECD, 2026[4]).
← 6. Work experience programmes are on-the‑job training programmes aimed at providing jobseekers with practical work experience.