This chapter explores the impact of Slovenia’s wage subsidies or employment incentive programmes on various labour market outcomes. Beyond the commonly assessed outcomes in impact evaluations – such as employment probability and duration – it investigates effects on occupational mobility, earnings and migration. Additionally, it contrasts the findings of the counterfactual impact evaluation with those from similar studies conducted in other countries. Finally, the analysis considers subgroups of workers categorised by age, gender, education level, and nationality.
Impact Evaluation of Wage Subsidies and Training for the Unemployed in Slovenia
Abstract
Along with training, the employment incentive or wage subsidy programme is one of the main active labour market policies (ALMPs) used to connect unemployed people with jobs in Slovenia.1 By providing employers with a financial incentive to hire certain categories of jobseekers, employment incentives can facilitate jobseekers’ integration into the labour market. This chapter examines how effective Slovenia’s wage subsidy programmes have been in moving individuals into sustainable employment, how this has affected their career prospects, and how the effects vary across groups of individuals.
The results of the counterfactual impact evaluation (CIE) suggest that employment incentives have large and statistically significant effects on the probability of individuals being in employment. Compared with the results of other studies of similar programmes in other countries, the estimated effects for Slovenia are generally slightly larger over the time horizons examined (up to 48 months after initial entry into the programme). These large employment effects are observed without negatively affecting occupational mobility over the longer term. In addition, examining an outcome that previously has not been the subject of analyses in ALMP CIEs, the subsidies are shown to decrease emigration from Slovenia.
The chapter is structured as follows. The next section presents the overall results of employment incentives in terms of the main outcomes studied: employment probability and duration, earnings, occupational mobility and emigration. It also compares the results of the CIE with those of similar studies, both for Slovenia and for other countries. The next section compares the outcomes observed for employment incentives across subgroups of workers based on age, gender, education level and nationality, comparing these effects with those found in similar studies in other countries. The chapter concludes by analysing how the programme’s effects vary according to the year participants entered the programme.
The following sections present the aggregate results of wage subsidies on selected labour market outcomes and compare these findings with results from other studies. The effects of employment incentives are estimated using the propensity score matching approach outlined in Chapter 3 of this report. The next sections describe the aggregate results for employment incentives on selected labour market outcomes and compare them to the results from other studies.
5.2.1. Positive employment effects of employment incentives decline over time but remain present after three years from the programme start
Before analysing the counterfactual effects of employment incentive participants – that is, the estimated difference between the outcomes they achieved through the programme and those they would have experienced without participation – it is useful to first examine the employment outcomes of participants and the control group (those with similar characteristics). This provides reference values for outcomes that are broadly comparable to those found in analytical reports used for monitoring purposes in Slovenia. However, it is important to note that the control group here includes only a subset of the unemployed, as it only includes jobseekers who are matched with jobseekers participating in wage subsidy programmes.
Figure 5.1, Panel A compares the employment rates of individuals entering employment incentive programmes with comparable individuals who did not participate (at least not during the same calendar month). The analysis excludes the majority of registered jobseekers who were not comparable to programme entrants, as well as a small number of participants for whom suitable non-participants could not be identified. The note accompanying Figure 5.1 explains how treatment and control group individuals were matched, while further details on the econometric approach are provided in Section 3.5 of Chapter 3 and the accompanying technical report (OECD, 2025[1])
Share of individuals in employment and percentage point effect on employment probability
Note: Panel A shows the percentage of employed persons among the treatment and matched control group. Panel B plots the treatment effect (i.e. the difference between the treatment and control group shown in Panel A). The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering ALMPs in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. The confidence intervals in Panel B are shown at the 5% level of significance and represented by the whiskers delimiting the dotted lines on the charts.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
Initially, nearly all employment incentive participants are recorded as employed in the employment database maintained by the Statistical Office (SRDAP). By contrast, individuals in the control group are, by definition, not employed during the month they are matched with an employment incentive participant. The initial small decline in the proportion of employed participants largely reflects individuals leaving the wage subsidy programme prematurely, despite all the analysed programmes being designed to last 12 months. In interpreting the figures, it is worth noting that effects are only measured in 3‑month intervals – this means that, for example, the effects are not measured in months 10 and 11 (the changes between months 9 and 12 in Figure 5.1 in fact occur mostly after month 11). After 12 months, 65% of employment incentive participants remain employed, with the proportion declining to 58% at 36 months.2 In comparison, the share of employed individuals in the control group rises from 0% at the start of the observation period to 28% by month 12, eventually reaching 32% by month 36.
The difference between the outcomes of employment incentive participants and those of the control group – represented by the gap between the two lines in Figure 5.1, Panel A – provides the estimates of the counterfactual impact. The results indicate strongly positive, statistically significant effects (as illustrated in Figure 5.1, Panel B). These findings demonstrate that employment incentives significantly and substantially increase individuals’ probability of being employed. After 36 months, employment incentive participants were 25.6 percentage points more likely to be in employment than their matched controls. Given that the subsidy period lasts for 12 months, these effects capture unsubsidised employment.
5.2.2. The employment incentives also improve labour market outcomes other than employment probability
In parallel with the positive effects on employment probability, the impact evaluation also reveals a decline in registered unemployment and inactivity over the entire 36‑month observation period (Figure 5.2). Given that employment incentives influence registered unemployment and inactivity during the initial months of participation by construction, it is particularly insightful to focus on their effects at the end of the 36‑month period. Examining the effects at the end of the observation period shows that participants in employment incentive programmes have significantly lower rates of both registered unemployment and inactivity compared to comparable individuals who did not participate.
Specifically, the results show that roughly half of the net increase in employment among employment incentive participants can be accounted for by individuals who would have otherwise been registered as unemployed. Another way to interpret the effects is to separately compare the outcomes between the employment incentive participants and their matched controls (Annex Figure 5.A.1), as was also done in Figure 5.1, Panel A for the probability of employment. After 36 months, 58% of employment incentive participants are employed, 25% are registered as unemployed, and 14% are inactive. In contrast, among comparable non-participants, 33% are employed, 39% are registered as unemployed, and 26% are inactive.
Although employment incentives reduce the proportion of individuals in registered unemployment, they also result in an increase in the share of unemployment benefit recipients. In absolute terms – focusing solely on the outcomes of employment incentive participants – the proportion of individuals in the treatment group receiving unemployment benefits begins to rise after nine months (Annex Figure 5.A.1). This timing aligns with the point at which some participants complete the period during which employment incentives are provided. While the majority of participants remain employed even after the incentives end, a minority transition back into unemployment. These individuals may subsequently qualify for unemployment benefits, highlighting a potential trade‑off: while employment incentives successfully reduce registered unemployment, they may also increase benefit claims among a subset of former participants who return to unemployment.
Percentage point effect on being in different labour market states
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
The positive effects of employment incentives on employment probability are also reflected in the employment duration results: employment incentives participants were employed for significantly longer periods than jobseekers who did not engage in subsidised employment (Figure 5.3, Panel A). Over the three‑year observation period, participants were employed for 464 additional days compared to non-participants. This statistic includes both the days worked during the subsidised period – the first 12 months – and the subsequent period when no employment incentives were provided. Approximately two‑fifths of the additional days worked in this three‑year period are attributable to additional days worked during the period after the initial 12 months, when individuals have exhausted their subsidies. During this period, they were employed for 195 days more than individuals in the control group. These findings highlight the sustained employment benefits of the programme, extending well beyond the subsidy period.
Cumulative days worked, cumulative earnings and change in occupational index for those who found a job
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. The confidence intervals are shown at the 5% level of significance and represented by the whiskers delimiting the dotted lines on the charts.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
The increased days worked by programme participants contributed to substantial additional earnings (Figure 5.3, Panel B). Over the three years following their entry into the employment incentive programme, participants earned a total of EUR 33 833 – EUR 21 086 more than their counterparts in the control group during the same period (that is, EUR 12 746). The trajectory of these earnings gains, while consistently positive, shows diminishing marginal increases over time, mirroring the pattern of employment effects, which also taper off in magnitude over time. These earnings effects are particularly significant when considered in the context of counterfactual earnings. Over the 36‑month period, slightly over half of the additional earnings attributable to the employment incentive programme – EUR 11 730 – occurred during the first 12 months of participation, underscoring the immediate financial impact of the programme on participants’ income.
To what extent can the employment incentive programmes be considered cost-effective? While a detailed cost-benefit analysis is outside the scope of this report, the administrative data include information on the exact subsidy amounts, allowing for unsubsidised earnings to be included as an outcome variable.3 During the period examined, employers hiring participants through the employment incentive scheme could receive subsidies ranging from EUR 5 000 and EUR 7 000. These subsidy amounts represent only one‑quarter to one‑third of the additional cumulative earnings attributed to participation in all employment incentive programmes. This suggests that, at least in terms of earnings gains relative to subsidies, the programmes may offer a favourable cost-benefit balance even without taking into account saving from lower social assistance benefit payments.4
The positive employment effects of employment incentives in Slovenia are especially remarkable in light of the additional finding that they did not adversely affect individuals’ occupational mobility. The difference in the average occupational index for jobseekers entering subsidised employment and their matched controls was not statistically significantly different from zero (Figure 5.3, Panel C).5 While these aggregate results do mask substantial heterogeneity across different groups of individuals – detailed further in Section 5.3 – they nonetheless indicate that jobseekers are not, on the whole, accepting lower-paying roles in exchange for employment.
The last outcome examined in this evaluation concerns the impact of employment incentives on the probability of migrating abroad. The results indicate that participation in employment incentives reduces the likelihood of migration (Figure 5.4). Although the absolute effect size is modest – reaching a maximum reduction of 0.75 percentage points approximately 21 months after participants enter the subsidy – the relative effect is substantial. Specifically, at 21 months, 0.93% of individuals in the control group have migrated abroad, compared to just 0.18% of the treatment group, representing an 80% reduction in the migration rate. Over subsequent periods, the relative effect size moderates slightly, but the migration rate remains two‑thirds lower for the treatment group even after four years (based on results not shown in the figure).
The decline in migration probability is initially driven by members of the control group migrating abroad during the first year, when virtually all employment subsidy recipients are employed by an employer receiving the subsidy in Slovenia. Thereafter, a small number of former employment incentives participants do move abroad. However, there is a general trend of moving back to Slovenia, among both the treatment and control groups, from month 27 onwards. This can be explained by the onset of the COVID‑19 crisis in March 2020, which prompted many Slovenes to return from abroad.6
Percentage point effect on emigration probability
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. The confidence intervals are shown at the 5% level of significance and represented by the whiskers delimiting the dotted lines on the charts.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
5.2.3. The estimated boost to employment probability by Slovenia’s employment incentive programmes compare favourably with estimates from other studies
To assess how the results of the CIE of Slovenia’s employment incentive programmes compare with findings from similar studies conducted internationally, this section situates them within the context of two meta‑analyses and a CIE of older programmes. The first meta‑analysis, conducted by Card, Kluve and Weber (2017[2]) synthesises findings from over 200 impact evaluations of ALMPs across 49 countries. Of these, 15 evaluations provide specific point estimates of employment effects for employment incentive programmes. The second meta‑analysis focuses on projects funded by the European Social Fund (ESF). It includes results from 17 studies on employment incentives and 14 studies on mixed interventions, which combine measures such as incentives with training components (European Commission, 2023[3]). These comparative analyses provide a useful comparison for how the employment effects of Slovenia’s programmes align with international benchmarks.
The first part of this section focuses exclusively on employment outcomes. As highlighted in the discussion of training outcomes in Chapter 4, the meta‑analysis by Card, Kluve, and Weber (2017[2]) does not include estimates for other outcomes examined in Slovenia, such as earnings, days worked, or occupational mobility. Although the meta‑analysis of ESF programmes includes some estimates for measures such as earnings, the limited number of such estimates prevents meaningful comparisons across programme types. Therefore, the analysis here concentrates solely on employment effects, where sufficient data allow for robust comparisons.
Compared with the results of the meta‑analysis, the estimated effects for Slovenia are much larger, particularly over shorter time horizons (Figure 5.5). The estimated short-term effect for Slovenia, 34 percentage points, is considerably higher than the median of 0 and 10 percentage points found in the 2018 and ESF meta‑analyses, respectively. The long-term effect, of 23 percentage points, is also slightly above or on par with the 23 and 19 percentage point median of the 2018 and ESF meta‑analyses, respectively.7
Percentage point effect of employment incentives on employment probability
ESF: European Social Fund.
Note: Short, medium and long-term effects respectively refer to effects up to one year, 1‑2 years, and more than two years after programme completion. For Slovenia, results refer to 12, 24 and 48 months after beginning the programme. As such, the observation periods are similar, but potentially not fully aligned. The box and whisker plots for the other studies refer to the 5th, 25th, 75th and 95th percentiles of the estimates (i.e. the black line at the bottom refers to the 5th percentile and the bottom of the blue box refers to the 25th percentile). Point estimates are included in the chart even if they are statistically insignificant. The studies presented adopt various research designs and econometric techniques – the results for Slovenia use nearest-neighbour propensity score matching (for details, see Chapter 3).
Source: Card, Kluve and Weber (2017[2]), “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations”, Journal of the European Economic Association, Vol. 16/3, pp. 894‑931, https://doi.org/10.1093/jeea/jvx028, European Commission (2023[3]), Meta‑analysis of the ESF counterfactual impact evaluations,. https://data.europa.eu/doi/10.2767/580759 and OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
When interpreting the results, it is important to note that, while the point estimates in the comparison studies are generally positive, they are not always statistically significant, even in the long term (in contrast to the present study). Figure 5.5 displays all the point estimates from the studies included in the meta‑analysis Card, Kluve and Weber (2017[2]). Interestingly, a slight majority (58%) of the studies in this meta‑analysis do not report effects that are both positive and statistically significant over the long term. This highlights the variability in the effectiveness of employment incentive programmes across different contexts and underscores the importance of considering the broader evidence base when evaluating their outcomes.
The results of the impact evaluation of this report corroborate the results of an older evaluation of Slovenia’s employment incentives programme, which found similarly positive effects (Burger et al., 2021[4]). This evaluation looked at the 11 228 entrants into the Employ.me programme during the 2009‑12 period. At that time, the employment subsidy amount was relatively more generous, amounting to roughly 50% of programme participants’ gross annual wages (compared to 38% in the current evaluation) and lasting up to two years. The estimated employment effects were also slightly higher: the effects on employment at 36 months were approximately 30 percentage points, compared to 26 percentage points in the current evaluation.
While the empirical results do suggest that Slovenia’s employment incentives are effective, the large magnitude of the results should be interpreted with caution. Several additional factors could conceivably play an important role in the evaluation of the employment incentive programmes:
Deadweight effects. These effects related to an unintended effect of the subsidies and occur when subsidies unintentionally support hiring that would have otherwise occurred without them. Concerns about such effects are common to all employment incentive programmes, and empirical research has documented that they are indeed present and often sizable (see Brown and Koettl (2015[5]) for an overview of the empirical evidence). Appropriately targeting such employment incentives – for example, by limiting eligibility to long-term employed, as is generally the case in Slovenia – can help limit such effects: a recent evaluation of Lithuania’s employment incentives, for example, suggested any such deadweight effects were minimal (OECD, 2022[6]).
Distortionary effects of subsidies on non-participants. These can include two types of effects affecting individuals and firms (Calmfors, 1994[7]). Substitution effects refer to job positions created for a certain category of workers replacing jobs for other categories because relative wage costs have changed. Displacement effects refer to jobs created by the subsidised programme in firms receiving the subsidies resulting in reduced hiring of such workers in firms not receiving the subsidies (may also be referred to as “crowding out” effects). The empirical evidence on these effects is far from conclusive, possibly due to differences in policy parameters and how they are implemented (Boockmann,, 2015[8]; Desiere and Cockx, 2022[9]; Lombardi, Skans and Vikström, 2018[10]).
Mismeasurement of counterfactual outcomes due to undeclared work. The large, estimated effect could be partly attributable to measurement of outcomes for control group individuals, comprised of individuals with otherwise similar observed characteristics but who could conceivably be engaging in undeclared work while registered as unemployed. Such undeclared work has the effect of decreasing the observed employment rate of the control group, artificially inflating the difference in the rates of actual employment – declared or undeclared – between the two groups. This is because participants in wage subsidies are, by construction, in formal employment, making it also more likely that they remain in formal employment over the longer term. While significant progress has been made in addressing the prevalence of undeclared employment in Slovenia – the European Labour Authority (2023[11]) estimates it decreased by 2.4 percentage points between 2013 and 2019, to 10.7% – it arguably remained a relevant feature of the labour market for much of the period when the programmes were analysed.8 Individuals engaging in undeclared work have strong financial incentives to register with the PES, given that it is a condition for receiving of financial social assistance benefits. On the other hand, benefit conditionality is enforced in Slovenia – in the first 10 months of 2024, for example, 5% of individuals who exited the unemployment registry did so because they were not fulfilling the activation requirements (ESS, 2024[12]). These sanctioned individuals represent 6% of the average monthly number of unemployed during that same period.
An additional, complementary interpretation of the results is that the employment incentives programmes are encouraging people in Slovenia to transition from undeclared work into formal employment. Individuals who engage in undeclared work have a financial incentive to register with the ESS – for example, in order to receive social benefits and free health insurance. While the vast majority of those targeted by the subsidies are indeed likely to be from the most vulnerable groups, a subset of them may be engaging in undeclared work. Participation in employment incentives programmes provide a financial incentive for entering the formal labour market and serve to counteract the financial disincentives to formal employment that arise whenever earnings are subject to taxes or social contributions. This interpretation of the results highlights the potential for employment incentives to contribute to further formalising the labour market in Slovenia.
This section discusses how the results of the employment incentives vary across sub-groups of jobseekers. It begins by discussing the results on employment and then discusses earnings, occupational mobility and emigration.
5.3.1. Employment incentives are effective for all groups, but particularly for the long-term unemployed
Given that the results above have documented the generally positive effects of employment incentives programmes in aggregate, an interesting additional set of questions concerns their effects across different characteristics of subgroups of unemployed. Paralleling the analysis of training in Section 4.3 of Chapter 4, the subsequent analysis provides separate estimates for the results along several dimensions: (i) gender, (ii) age, (iii) level of education, (iv) nationality, and (v) long-term unemployment status. While the estimated results do differ slightly across some of these characteristics, it is worth bearing in mind that the positive effects on outcomes such as employment probability are present for all groups examined.
Men and women tend to experience similarly high boosts to employment probability (Figure 5.6). Three years after beginning subsidies, the effect size for men is not statistically significantly different from the effect for women. However, this masks considerable differences by age within gender: while both older men and women experience smaller boosts to employment compared to younger men and women, this difference is particularly large for women over 50. This difference in the effect sizes between women and men over 50 years – which amounts to 6.6 percentage points – is especially remarkable given that they otherwise have similar re‑employment rates in aggregate (as discussed in Section 3.3). However, the relatively modest employment effects for older women should not draw attention away from the broader positive programme outcomes older women experience from participating in the programme. As will be discussed further below, differences in earnings across demographic groups are notably smaller, suggesting that the issue may be more confined to employment probability rather than overall labour market outcomes.
Percentage point effect on employment probability at 36 months after starting employment incentives
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. Unemployment duration is measured either as the duration before ALMP entry (for ALMP participants) or entry into the control group (for their matched controls).
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
In terms of participant characteristics, the employment effects are strikingly consistent across individuals with varying levels of education and unemployment duration. Across all categories examined, the estimated effects on employment fall within a narrow range of 22 to 24 percentage points. One interpretation of this limited variation is that the subsidies are effectively targeted along these dimensions. The primary purpose of employment incentives programmes is to provide employers with an incentive to hire individuals they might otherwise be reluctant to employ, due to factors such as a lack of relevant prior experience or other significant barriers to employment. These barriers can exist for individuals regardless of their education level or unemployment duration, even though, at an aggregate level, those with higher education and shorter unemployment spells are statistically more likely to re‑enter employment. The lack of heterogeneity in effects across education levels aligns with the findings of studies of similar programmes from recent OECD evaluations in Lithuania and Greece (OECD, 2022[6]; OECD, 2024[13]).
In terms of nationality, non-EU nationals experience a larger boost to their employment prospects than Slovenes and other EU nationals. The difference in the effect size – 7 percentage points – is substantial. One plausible explanation for this disparity is that employment incentives programmes help mitigate uncertainties employers may face when hiring foreign workers. These uncertainties often stem from imperfect information about the quality of their education, skills, and prior work experience. Over time, as these individuals demonstrate their capabilities during employment, the quality of the job match is often revealed to be better than initially expected, resulting in higher observed subsidy effects. An additional explanation is that employment incentives support non-EU nationals in maintaining their residence within Slovenia and the EU, as securing employment is frequently a requirement for their continued residency.
To a large degree, the effects of employment incentives on the cumulative earnings of different sub-groups of workers reflect their employment effects (Annex Figure 5.A.2). Groups experiencing relatively larger boosts to their employment probabilities – such as the long-term unemployed – also generally experience a larger effect in terms of increased earnings. The one exception to this trend relates to women over age 50 – the relatively lower employment effect for this group is not reflected in lower cumulative earnings. This can be partly explained by a difference in the trajectory of the employment effect for this group: compared to other groups of workers, employment incentives are observed to have an especially positive employment effect in the first year after individuals enter the programme. This boosts their cumulative earnings in the first year, with these additional earnings still relevant in the cumulative earnings after three years.
How do the effects of employment incentives on emigration vary across different groups of individuals? With the notable exception of non-EU nationals, the results generally show strong decreases in emigration probability across all the categories examined (Annex Figure 5.A.3). The point estimates for all the groups are negative, although not all are statistically significant. Notably, they are negative and statistically significant for individuals aged 30‑50, who are among the most likely to emigrate – in 2023, 1.5% of Slovene residents in this cohort migrated abroad (SORS, 2024[14]).9 The finding that employment incentives appear to contribute to higher emigration rates among non-EU nationals is the only anomaly in the results. This may be linked to the fact that longer periods of employment can help non-EU nationals secure permanent residency status or Slovene nationality, thereby facilitating their ability to work freely within the EU.
5.3.2. Employment incentives have a positive effect on the occupational mobility for young men who become employed
Another interesting dimension for examining the effects of employment incentive participation relates to occupational mobility. Empirical evidence for other countries has documented the “scarring” effects of job loss (for example, Lachowska, Mas and Woodbury (2020[15])). Interestingly, in the case of Slovenia, the occupational indices of individuals becoming re‑employed are on average only slightly adversely affected by job loss. Looking at the individual-level data used in the analysis among all individuals who are observed to have been employed both before and after a spell of registered unemployment, slightly under one‑third (32%) did not experience any change in their occupational index. Of the remaining 68%, the proportions of individuals experiencing negative mobility was 1.7 percentage points greater than those who experienced positive mobility.
The counterfactual impact evaluation results on the effects of employment incentive participation on occupational mobility show strong differences in the profiles across the groups examined. Figure 5.7 shows the average differences in the occupational index for individuals who became employed (at which point roughly one‑third of the control group and three‑fifths of the treatment group are employed).
Older men and women experience greater downward occupational mobility compared to their younger counterparts, although the differences are not statistically significant in all cases. Interestingly, young men may experience a slight boost to their occupational mobility, though the coefficient is only marginally insignificant. Surprisingly, across all age and gender groups, individuals in the control group – those not receiving employment incentives – are employed in occupations with similar average characteristics. These occupations typically pay wages approximately 10 percentage points below the average.
The findings for Slovenia align with similar analyses of employment incentives programmes conducted in Lithuania and Greece (OECD, 2022[6]; OECD, 2024[13]). Using a comparable methodological approach, the results for Lithuania indicated that men and women under 30 who became employed generally experienced increases in their occupational index, irrespective of whether they received employment incentives. In contrast, individuals over 50 experienced downward occupational mobility, a trend that employment incentives helped mitigate in the case of men. In Greece, positive effects on occupational mobility were observed for men under 30.
The similarity of findings across these countries underscores a common pattern: while employment subsidies have positive effects on employment outcomes, they are often accompanied by older workers moving into lower-paid occupations – particularly for older women. Conversely, younger workers, especially men, may benefit from both increased employment probabilities and upward occupational mobility. These findings highlight the dual nature of employment incentives, with benefits and drawbacks varying across demographic groups.
Employment incentives appear to have a negative impact on upward occupational mobility for jobseekers with lower levels of education, non-EU nationals, and the short-term unemployed. These three groups overlap significantly, as the design of the incentives allows individuals unemployed for less than 12 months to qualify only if they have completed primary education or less – a category in which non-EU nationals are disproportionately represented. While the employment incentives programme provides a substantial boost to employment and earnings for these groups, this is largely in occupations that are lower paid. Furthermore, for the short-term unemployed, entering the programme very early in their unemployment spell may lead them to accept jobs they otherwise may not have given more time to sample job offers or prepare for the right vacancy (although in the past this arguably was not a relevant feature in Slovenia – see van Ours and Vodopivec (2008[16])).This points to the potential trade‑offs that may be involved in ALMP participation.
Percentage point effect on occupational index 36 months after entering programme
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
One intriguing question is whether the effects of Slovenia’s employment incentives programme have changed over time during the period examined. Given that there were not important changes to the programme’s parameters in this period, the relevant question is whether it was affected by changing economic conditions. The period under review, 2016‑19,10 was characterised by an exceptionally rapid improvement in the labour market situation in Slovenia (as discussed in detail in Chapter 2). Following a peak in 2013, unemployment fell to 8.0% in 2016 and 4.4% in 2019 (for reference, the comparable decrease in the OECD-wide unemployment rate during this period was from 6.5% to 5.4%). This period of rapidly improving labour market conditions could conceivably have an effect on the effectiveness of the employment incentive programmes.
Examining the outcomes of the employment incentive programmes with entrants (cohorts) from individual years in the 2015‑19 period shows that all cohorts have very similar outcomes (Figure 5.8). The profile of the effects almost exactly matches the profile from the results presented in Panel B of Figure 5.1: large effects in the first year, when employment is subsidised, followed by a lower (but still strong) effect from month 12 onwards.
The finding that there are very few differences across the cohorts also provides reassurance that the results are not strongly affected by other important events, such as the COVID‑19 crisis. As in other countries, Slovenia instituted a variety of job-retention schemes during that period, including short-time work schemes (OECD, 2022[17]). Given that the results of Slovenia’s employment incentive programmes show a decline in effect over time elapsed since programme entry (consistent with the results of evaluation of similar programmes in other countries), such schemes could serve to artificially inflate the estimated effects of employment incentives by retaining workers in employment. However, given that the results are similar also for cohorts who were not subject to the COVID‑19 measures over the three‑year time horizon depicted – i.e. the 2016 and 2017 cohorts – any effects from these measures are likely to be negligible. Furthermore, the results are interesting in their consistency also in light of the gradual increase in labour market tightness over the period observed. In contrast to theoretical predictions and empirical evidence that such subsidies may be less effective in tight labour markets (see e.g. Albanese, Cockx and Dejemeppe (2024[18])), the evidence for Slovenia is not consistent with this notion.
Effect of employment incentives on employment probability by year of entry into programme
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. All point estimates are statistically significantly different from zero.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
A key question in designing employment incentive programmes is whether specific parameters, such as the duration or generosity of subsidies, influence their effectiveness. Unfortunately, the lack of significant variation in programme parameters in Slovenia during the analysed period limits the ability to investigate these questions. The programmes in the period analysed all provided subsidies for a fixed duration of 12 months. While subsidy amounts varied, this variation was tied to objective factors such as the employability of participants. This makes it challenging to isolate the impact of subsidy amounts from the influence of jobseeker characteristics. Consequently, while the programme’s overall effectiveness can be assessed, determining how adjustments to its parameters might affect outcomes remains an open question requiring further research with more diverse programme designs.
To examine the effect of programme parameters in the future, such evaluations would ideally be factored into the design of programmes. Several options are potentially available (Smith et al., 2024[19]). One option for examining the effects would be through a regression discontinuity design (RDD), where changes in eligibility relate to a continuous variable, such as a statistical profiling score which assigns the level of individual subsidies based on a combination of factors (but where subsidies are set in discrete amounts, as in the existing programmes). An RDD would examine the effects for individuals immediately below and above a certain threshold. If suitable numbers of individuals were observed in such groups – and the data contained information on precisely which group each individual belonged to – this could be used to examine the differences in the programme’s parameters. Another option would be to conduct a randomised controlled trial which would examine the effect of a key parameter, such as the subsidy’s level, duration, or the presence of additional feature, such as providing in-work, post-placement support. Such an intervention would have to be carefully designed and may require legislative changes to establish that different levels of subsidies may be provided to similar jobseekers, purely for the purposes of establishing the policy’s causal effect.
As shown in the impact evaluation results discussed in this chapter, Slovenia’s employment incentive programmes for the unemployed are effective in helping jobseekers move into employment, leading to long-lasting and meaningful improvements in other labour market outcomes as well. The increased employment rates of employment incentive participants are also reflected in increased earnings as well as days worked. They also contribute to strongly decreased rates of emigration abroad for Slovene nationals. Employment incentives also provide a small boost to the occupational mobility of men aged 30‑50, but do not affect the occupational mobility of other groups over the longer-term. While employment incentive programmes are effective for all groups of jobseekers examined, their effectiveness is particularly high for certain groups of jobseekers, such as those aged 30‑50. These results could be used to better inform the targeting of the programmes.
References
[18] Albanese, A., B. Cockx and M. Dejemeppe (2024), “Long-term effects of hiring subsidies for low-educated unemployed youths”, Journal of Public Economics, Vol. 235, p. 105137, https://doi.org/10.1016/j.jpubeco.2024.105137.
[8] Boockmann,, B. (2015), “The effects of wage subsidies for older workers”, IZA World of Labor, https://doi.org/10.15185/izawol.189.
[5] Brown, A. and J. Koettl (2015), “Active labor market programs - employment gain or fiscal drain?”, IZA Journal of Labor Economics, Vol. 4/1, https://doi.org/10.1186/s40172-015-0025-5.
[4] Burger, A. et al. (2021), “A comprehensive impact evaluation of active labour market programmes in Slovenia”, Empirical Economics, Vol. 62/6, pp. 3015-3039, https://doi.org/10.1007/s00181-021-02111-6.
[7] Calmfors, L. (1994), “Active Labour Market Policy and Unemployment: A Framework for the Analysis of Crucial Design Features”, OECD Economic Studies No. 22.
[2] Card, D., J. Kluve and A. Weber (2017), “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations”, Journal of the European Economic Association, Vol. 16/3, pp. 894–931, https://doi.org/10.1093/jeea/jvx028.
[9] Desiere, S. and B. Cockx (2022), “How effective are hiring subsidies in reducing long-term unemployment among prime-aged jobseekers? Evidence from Belgium”, IZA Journal of Labor Policy, Vol. 12/1, https://doi.org/10.2478/izajolp-2022-0003.
[12] ESS (2024), Mesečne informacije [Monthly information bulletin], https://www.ess.gov.si/fileadmin/user_upload/Trg_dela/Dokumenti_TD/Mesecne_informacije/MI_2024_10.xlsx (accessed on 2 December 2024).
[3] European Commission (2023), Meta-analysis of the ESF counterfactual impact evaluations – Final report, Publications Office of the European Union, Directorate-General for Employment, Social Affairs and Inclusion, Pompili, M., Kluve, J., Jessen, J. et al, https://data.europa.eu/doi/10.2767/580759.
[11] European Labour Authority (2023), Factsheet1 on undeclared work – Slovenia, https://www.ela.europa.eu/sites/default/files/2024-02/SI-UDW_factsheet-2023-fin.pdf.
[21] Haepp, T. and M. Serrano Alarcon (2024), Meta-analysis of the European Social Fund counterfactual impact evaluations: Brief update with alternative measures, European Commission, Ispra, https://publications.jrc.ec.europa.eu/repository/handle/JRC137571.
[15] Lachowska, M., A. Mas and S. Woodbury (2020), “Sources of Displaced Workers’ Long-Term Earnings Losses”, American Economic Review, Vol. 110/10, pp. 3231-3266, https://doi.org/10.1257/aer.20180652.
[10] Lombardi, S., O. Skans and J. Vikström (2018), “Targeted wage subsidies and firm performance”, Labour Economics, Vol. 53, pp. 33-45, https://doi.org/10.1016/j.labeco.2018.04.002.
[1] OECD (2025), “Technical report: Impact Evaluation of Employment Incentives and Training for the Unemployed in Slovenia”, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/about/projects/technical-reports-and-presentations-dg-reform/slovenia/Technical-Report-Evaluation-of-active-labour-market-policies-in-Slovenia.pdf.
[13] 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 (2022), Impact Evaluation of Vocational Training and Employment Subsidies for the Unemployed in Lithuania, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/c22d68b3-en.
[17] OECD (2022), Riding the waves: Adjusting job retention schmes through the COVID-19 crisis, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/publications/reports/2022/03/riding-the-waves-adjusting-job-retention-schemes-through-the-covid-19-crisis_9c5e78fe/ae8f892f-en.pdf.
[22] OECD (2022), The scope and comparability of data on labour market programmes, OECD, Paris, https://www.oecd.org/content/dam/oecd/en/topics/policy-issues/employment-services/almpdata-scope-and-comparability.pdf.
[20] Schneider, F. (2022), “New COVID-related results for estimating the shadow economy in the global economy in 2021 and 2022”, International Economics and Economic Policy, Vol. 19/2, pp. 299-313, https://doi.org/10.1007/s10368-022-00537-6.
[19] Smith, D. et al. (2024), “The Application of Randomized and Quasi-Experimental Designs in Federal Government Program Evaluation”, in Contributions to Economics, Public Policy Evaluation and Analysis, Springer Nature Switzerland, Cham, https://doi.org/10.1007/978-3-031-67604-8_5.
[14] SORS (2024), Emigrants by country of next residence, age groups and citizenship, Slovenia, annually, https://pxweb.stat.si/SiStatData/pxweb/en/Data/Data/05N3210S.px/.
[16] van Ours, J. and M. Vodopivec (2008), “Does reducing unemployment insurance generosity reduce job match quality?”, Journal of Public Economics, Vol. 92/3-4, pp. 684-695, https://doi.org/10.1016/j.jpubeco.2007.05.006.
Share of individuals in different labour market states
Note: The analysis presents gross outcomes from nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match.
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
Change in cumulative earnings at 36 months after starting employment incentives
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. Unemployment duration is measured either as the duration before ALMP entry (for ALMP participants) or entry into the control group (for their matched controls).
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
Percentage point effect on emigration probability 36 months after entering programme
Note: The analysis presents nearest-neighbour propensity score matching results which matches individuals based on several characteristics: each individual’s employment history (earnings, occupation), unemployment information (duration, unemployment benefit level and remaining duration), employability rating (based on ESS counsellor judgement), disability status, demographic characteristics (age, education, gender, nationality, marital status), and location. For every individual in the treatment group, the matching is conducted based on the values of these characteristics in the calendar month when the individual enters the programme. The control group is comprised of individuals with similar characteristics not entering active labour market programmes in that same calendar month. For paired individuals in the treatment and control groups, this calendar month is then the reference point after which outcomes are measured. Some individuals in the treatment group are dropped as they do not have a corresponding match. Unemployment duration is measured either as the duration before ALMP entry (for ALMP participants) or entry into the control group (for their matched controls).
Source: OECD calculations based on data from the Statistical Office of the Republic of Slovenia and the Employment Service of Slovenia.
← 1. Under the OECD classification of ALMPs (OECD, 2022[22]), Slovenia’s Employ.me (Zaposli.me) wage subsidy programme falls under Category 4 (Employment incentives), specifically, Subcategory 4.1 (Recruitment incentives includes measures which are payable for a limited period only). In this report, the terms wage subsidies and employment incentives are used interchangeably.
← 2. In results not reported in the figures in this report, we find that the employment effects remain similar even after 48 months: the employment rates of matched employment incentive participants stand at 57.5%, compared to 58.3% after 36 months. Also note that the effects are calculated in one‑month intervals for the first three months and in 3‑month intervals thereafter. Thus, the actual employment rates of participants do not markedly decrease in months 10 and 11 after entering the programme.
← 3. A value‑for-money analysis of these will be conducted as part of a subsequent evaluation being conducted by the OECD.
← 4. While a more detailed value‑for-money analysis is outside the scope of this report, it will be conducted by the OECD as part of an ongoing project with the OECD.
← 5. This occupational index, described in Section 3.7 of Chapter 3, is calculated for each detailed occupational code using data on the wages and employment of individuals in Slovenia and measures the extent to which individuals move in generally better-paid occupations (with an index value one unit higher than another corresponds to an average real monthly wage that is 1 percentage point greater relative to the overall average wage).
← 6. Note that the period for which participants in the analysis enter the programmes spans from June 2016 through December 2018. An individual entering the programme in December 2017 would be 27 months into the programme upon the onset of the COVID‑19 crisis.
← 7. The meta‑analyses of the ESF-funded studies refers to results that includes more than one result per study. Including only one result per study, the “balanced” average effects for wage subsidies are slightly higher (Haepp and Serrano Alarcon, 2024[21]).
← 8. A parallel measure of undeclared work is the size of the informal economy. By some estimates, this remains above the EU average. For example, Schneider (2022[20]) estimates that Slovenia’s informal economy amounted to 22% in 2018, still considerably above the EU average of 17%.
← 9. The only age group with notably higher emigration rates is individuals aged 25‑29, among whom the rate reached 3.6%.
← 10. This period is in contrast to analyses in the other sections, which captures potential participation during 2015‑18 (given the lack of the Employ.me programme in 2015, the analysis begins with entrants in June 2016). This is to enable an up to 4‑year time horizon for tracking outcomes. The analysis in this section tracks individuals for only up to 36 months after they enter the employment incentives programme and refers to the period from 2016‑19. This is done so as to provide results from another cohort of participants.