Impact Evaluation of Vocational Training and Employment Subsidies for the Unemployed in Lithuania

2. Recent trends in the Lithuanian labour market and active labour market policies

Abstract

The employment rate grew strongly in Lithuania over the last decade and suffered due to the COVID‑19 pandemic less than in other OECD countries. An increased labour demand has encouraged people to enter the labour market and look for a job, but has also increased wages faster than the productivity growth. Furthermore, significant disparities in labour market outcomes exist by education level and geographic location, highlighting the need for active labour market policies (ALMPs). The 2017 reform aimed to make the system of ALMPs more effective, efficient and accessible, yet the coverage of ALMPs has remained low and focused on employment incentives. A fully-fledged evidence informed policy making is needed to make ALMP provision more effective and achieve sustainable funding.

2.1. Introduction

This chapter gives first an overview of the labour market situation in Lithuania, highlighting the key challenges that need to be addressed by active labour market policies (ALMPs) and employment policy more generally. Subsequently the system of ALMP provision is reviewed, with the focus on its potential effectiveness to support the labour market and address its challenges.

The employment rate in Lithuania has increased steadily over the past decade and remained resilient during the COVID‑19 pandemic. Yet, high employment has been accompanied by wage growth beyond productivity growth. Furthermore, the employment rate is still low among people with lower levels of education, people living in remote areas and people with health limitations, suggesting needs for training measures, measures to support mobility, as well as ALMP support targeting the individual needs and employment obstacles in general. The importance of providing opportunities for good jobs for all is further underlined by the projections that working-age population will be declining fast in the coming years, beyond the drop in the total population.

A high share of jobseekers register with the Lithuanian Employment Service (LES), but they are potentially incentivised more by gaining access to health insurance and benefits, rather than ALMPs. The budget for ALMPs has remained about half of the level of other OECD countries even after a major reform in 2017 aiming to redesign ALMPs to increase their effectiveness, efficiency and accessibility (a major budget allocation during the COVID‑19 was for employment maintenance schemes, but not for other ALMPs). A large component of the ALMPs are employment incentives, particularly a scheme supporting hiring people with disabilities in so-called social enterprises that have been widely criticised by the stakeholders. Take‑up of training schemes has remained modest regardless of significant efforts to redesign the schemes. Lithuania needs to implement a fully-fledged evidence‑informed policy making, including conducting counterfactual impact evaluations systematically and disseminating their results. This could help make the system more effective, as well as attract sustainable national funding and rely less on EU funding resources.

2.2. Labour market situation and trends in Lithuania

This section compares the labour market situation in Lithuania with other OECD countries and identifies specific challenges that Lithuania faces which could be addressed by ALMPs.

2.2.1. Strong employment growth characterised the Lithuanian labour market over the past decade

The Lithuanian labour market has steadily improved over the past years and remained resilient through the COVID‑19 pandemic. The employment rate reached the levels seen before the Global Financial Crisis around 2013 and continued to increase until 2019, exceeding the OECD average (73.0% versus 68.7% among 15‑64 year‑olds, Figure 2.1). The drop in the employment rate in 2020 was smaller in Lithuania (1.4 percentage points) than in the OECD on average (2.5 percentage points) and has recovered close to the pre‑pandemic level by 2021 (at 72.4%). Similarly, the labour force participation rate in Lithuania had increased faster than the OECD average before the COVID‑19 pandemic, and continued to grow in 2020, while the OECD average dropped. This high and growing labour force participation rate (and thus low inactivity rate) is also the reason why the unemployment rate has remained higher in Lithuania than the OECD average and increased substantially in 2020 (from 6.5% in 2019 to 8.8% in 2020 among 15‑64 year‑olds, while the unemployment increased from 5.6% to 7.3% in the OECD on average). In 2021, the unemployment rate in Lithuania has decreased (to 7.4%), but also labour force participation dropped (to 78.2%).

The general labour market trends in Lithuania have been over the years similar to the other Baltic countries – Estonia and Latvia. During the years preceding the COVID‑19 outbreak, the employment rate grew slightly faster in Lithuania than in the other Baltic countries. Lithuania’s employment rate surpassed the same figure in Latvia in 2016, but has not yet caught up with the Estonian level (72.4% versus 74.0% among 15‑64 year‑olds in 2021). The labour force participation rate has also grown significantly faster in Lithuania than in the other Baltics, reaching 78.2% in 2021 among 15‑64 year‑olds, being still below the level of Estonia (79.1%), but above the level of Latvia (75.8%) and the OECD (72.4%).

Figure 2.1. The employment rate suffered less in Lithuania than the OECD average during the COVID‑19 pandemic, and the unemployment rate increased partly due to the growing labour force participation rate

Note: OECD is the weighted average of the 38 OECD member countries. Countries ranked by 2019 data.

Source: OECD LFS by Sex and Age – Indicators Database, http://stats.oecd.org//Index.aspx?QueryId=54218.

Statlink

With the right employment policy, including effective active labour market policies, Lithuania has the potential to further increase its employment rate and labour force participation rate, securing labour income and stronger social security for a higher share of its working age population.

The increasing employment rate has been accompanied by quickly rising wages in Lithuania over the past years (Figure 2.2). Lithuania witnessed a higher real wage growth than any other country in the OECD between 2013 and 2020 (real wages grew in total by 49% over this period, while the OECD average increased by 8%). Quickly increasing wages and employment rates have also likely encouraged people to enter the labour force from inactivity, thus increasing the labour force participation rate and keeping the unemployment rate relatively high. While the wage level in Lithuania has been quickly catching up with the OECD average, a significant gap still remains.

Along with a strong wage growth, gender wage gap has decreased, but remains significant. In 2018, gender wage gap in Lithuania was at 11.7%,1 which was lower that the OECD average (12.7%), but slightly higher than the EU average (11.1%), (OECD, 2022[1]).

Nevertheless, labour productivity has not kept up with the rising wages. In the context of relatively high and increasing employment rate, the employers have been pressured to increase wages, while investments in productivity have been lagging behind.

One factor contributing to labour market tightness in Lithuania has been its decreasing population. Compared to 1990, Lithuania’s population has shrank by 26% by 2020. The drop among the working-age population (15‑64 year‑olds) has been even sharper (29%). This challenge is expected to remain as Lithuania is forecast to lose more of its population by 2050 than any other OECD country – 22% of total population (Figure 2.3) and 31% of working-age population. The particularly fast decline in the working-age population stresses the urgency to support anyone willing and able to work to access good jobs via effective employment policies, including active labour market policies. Addressing such a challenge successfully requires reaching out to and supporting groups beyond the usual target groups of ALMPs, such as discouraged workers and other groups in inactivity who would like and are able to work in case of appropriate support, people in low value‑added jobs and in risk of job loss, as well as people who have reached the retirement age, but would like to continue working.

Figure 2.2. Wages have increased fast in Lithuania, exceeding labour productivity

Note: Gross domestic product (GDP), purchasing power parity (PPP), United States Dollar (USD). Panel B: Countries are ordered according to the increase in average annual wages between 2013 and 2020 (highest increase on the right). OECD is an unweighted average (and excludes Colombia, Costa Rica and Turkey). This dataset contains data on average annual wages per full-time and full-year equivalent employee in the total economy. Average annual wages per full-time equivalent dependent employee are obtained by dividing the national-accounts-based total wage bill by the average number of employees in the total economy, which is then multiplied by the ratio of average usual weekly hours per full-time employee to average usually weekly hours for all employees. Average wages are converted in USD PPPs using 2020 USD PPPs for private consumption and are deflated by a price deflator for private final consumption expenditures in 2020 prices.

Source: Panel A: OECD calculations based on the OECD Productivity Database, Growth in GDP per capita, productivity and ULC Dataset, http://stats.oecd.org//Index.aspx?QueryId=54368 for Labour compensation per hour worked in current prices and GDP per hour worked in constant prices; and OECD Key Short Term Economic Indicators Dataset [Consumer Prices – Annual inflation], http://stats.oecd.org//Index.aspx?QueryId=21757 for consumer price index. Panel B: the OECD Average Annual Wages Database, http://stats.oecd.org//Index.aspx?QueryId=25148.

Statlink

Figure 2.3. Lithuania is forecast to lose close to one‑quarter of its population by 2050

Note: OECD is the weighted average of the 38 member countries.

Source: United Nations World Population Prospects 2019.

Statlink

2.2.2. Large labour market disparities still exist between different groups of population

While employment has increased strongly among people with lower educational attainment, gaps compared to highly educated people have remained. Both the employment gap between people with tertiary and secondary education (19 percentage points among 20‑64 year‑olds in 2021), as well as the employment gap between people with secondary and primary education (18.8 percentage points) are more pronounced than in many other OECD countries (Figure 2.4). Furthermore, the higher than the OECD average total employment rate among the working age population is indeed driven by the high employment rate among people with tertiary education. The average of the 26 OECD countries shown in Figure 2.4 indicates that the employment rate among 20‑64 year‑olds was lower in Lithuania than in the OECD in 2021 for people with primary education (51.5% vs 55.3%) and people with secondary education (70.3% vs 72.9%), but higher for people with tertiary education (89.3% vs 85.2%).

Figure 2.4. The employment gap between tertiary and secondary education in Lithuania is one of the widest in OECD

Note: Employment gap between primary and secondary education: difference in employment rate in between people with less than primary, primary and lower secondary education (ISCED levels 0‑2), and people with upper secondary and post-secondary non-tertiary education (ISCED Levels 3 and 4). Employment gap between secondary and tertiary education: difference in employment rate in between people with upper secondary and post-secondary non-tertiary education (ISCED Levels 3 and 4), and people with tertiary education (Levels 5‑8). The purple markers represent the unweighted average of the 26 countries shown. Data are sorted by the ascending gap size between secondary and tertiary education. Data for Turkey refer to 2020.

Source: Eurostat – Employment by educational attainment level – annual data.

Statlink

There are strong geographic disparities in the Lithuanian labour market. While the employment rate was 81.4% among the 15‑64 year‑olds living in the capital city of Vilnius in 2020, it was only around 50% in the remote and rural areas like Anykščiai, Ignalina, Lazdijai and Šakiai municipalities (Statistics Lithuania data). The OECD computations using the data of the EU Statistics on Income and Living Conditions from 2019 show that geographic obstacles to jobs (living far from the location of jobs) are particularly pronounced in Lithuania (Figure 2.5). About a quarter (23.5%) of 16‑64 year‑olds who were not in employment in 2019 were living in a thinly populated area and in a household without a car. Another prominent employment obstacle in Lithuania tends to be health limitations of working-age people (OECD, 2022[2]; Pacifico et al., 2018[3]), while other reasons like care responsibilities are slightly less severe than in the other EU countries on average.

Due to the substantial labour market disparities, structural unemployment is likely a significant challenge in Lithuania. In 2019, the natural rate of unemployment (non-accelerating wage rate of unemployment – NAWRU) was estimated to be 7.2% in Lithuania, i.e. slightly higher than the observed unemployment rate of 6.5%, as well as higher than the EU average NAWRU at 6.9% (Figure 2.6). Although the level of natural unemployment is forecast to decrease in Lithuania in 2022, data from the past years suggest that there is still a lot of scope to address the education/skills and geographic mismatches on the labour market, as well as to support non-employed people to overcome the specific obstacles they face in accessing jobs. This underlines the importance of training measures, measures to support geographic mobility, as well as individualised support more generally that should be provided by the system of ALMPs in Lithuania.

Figure 2.5. Geographic distance is a severe barrier to employment for the out-of-work people in Lithuania

Note: Geographic distance barrier is defined as living in a thinly populated area and in a household without a car. Due to data comparability across countries the “self-defined” measure of out-of-work is used in this chart. See the methodology in OECD (2022[2]), Data refer to 2018 for Ireland and Italy.

Source: OECD calculations based on the European Union Statistics on Income and Living Conditions (EU-SILC), 2019.

Statlink

Figure 2.6. Structural unemployment in Lithuania is potentially higher than the EU average

Note: The natural unemployment rate refers to the non-accelerating wage rate of unemployment (NAWRU), i.e. the rate of unemployment consistent with constant wage inflation. The natural rate of unemployment consists of the frictional and structural components. If the observed unemployment rate is close to the natural unemployment rate, we can assume that cyclical component in the observed unemployment rate is low and most of it can be explained structural unemployment.

Source: AMECO Database of the European Commission’s Directorate General for Economic and Financial Affairs, https://dashboard.tech.ec.europa.eu/qs_digit_dashboard_mt/public/sense/app/667e9fba-eea7-4d17-abf0-ef20f6994336/sheet/2f9f3ab7-09e9-4665-92d1-de9ead91fac7/state/analysis.

Statlink

2.3. The system of active labour market policies in Lithuania

This section provides an overview of the institutional and organisational set-up of ALMP provision and the ALMP composition in Lithuania, aiming to provide a general idea of how the ALMP system functions. The evaluation of specific ALMPs is presented in next chapters.

2.3.1. The institutional set-up of ALMP provision has improved

The current institutional set-up of ALMP provision in Lithuania is a result of a wider labour market reform introduced in July 2017. The introduction of the so-called “social model” aimed to strengthen flexicurity as well as ALMPs, copying some aspects of the Danish “golden triangle” (OECD, 2016[4]). The preparation of the reform set a good example how policy makers involved researchers tightly in the process and designed the new institutional set-up based on available evidence and good practices from other countries.

Regarding ALMPs, the new social model aimed to increase their effectiveness and accessibility. To achieve that, a number of changes were introduced in ALMP design that also meet well the labour market challenges identified in the previous section of this chapter:

• Targeting ALMPs more according to the individual characteristics of the jobseekers;

• Increasing the importance of training measures among ALMPs, as well as workplace‑related components in training;

• Dropping ALMPs that were not considered effective in helping people to integrate into the primary labour market, such as job rotation and public works schemes. Public works schemes were entirely transferred to municipalities;

• Introducing possibilities for more varied support to jobseekers, such as supporting mobility.

With the 2017 reform, the organisational set-up of ALMP provision got centralised and was modernised. Instead of the previous decentralised Lithuanian Labour Exchange, ALMPs are implemented by the Lithuanian Employment Service (LES) under the Ministry of Social Security and Labour. In addition to fundamental changes in the structure and management of the LES, its operating model, processes and infrastructure have been continuously modernised over the past years (European Commission, 2019[5]).

The introduction of the social model has strengthened the role of the social partners in ALMP design. The new model included the topic of ALMPs in the discussions of the Tripartite Council, which is a channel for the social partners to provide strategic advice for the LES and the Ministry of Social Security and Labour. The social partners discuss regularly in the Tripartite Council the organisation of ALMP design and provision, as well as the LES strategies (Lauringson and Lüske, 2021[6]). Although this committee has only an advisory role and the social partners are not involved as extensively as in the Danish system of ALMP provision (OECD, 2021[7]), the committee’s work has been assessed positively by all stakeholders involved and is believed to bring ALMP provision closer to the actual needs of jobseekers and employers.

The involvement of the social partners in ALMP design is expected to improve the LES image, enabling to reach out to additional segments of jobseekers and employers. As the majority of registered jobseekers have no higher qualification (73.1% of registered jobseekers had up to secondary education in 2020), employers are currently reluctant to contact the LES to fill vacancies requiring higher qualification. As such, registered jobseekers have fewer opportunities to move to good jobs even after upskilling, and jobseekers with higher skills might be reluctant to contact the LES expecting there are no matching vacancies available for them. The LES has been actively aiming to engage with employers providing vacancies for high-skilled jobs over the past years also bilaterally in addition to discussions in the advisory committee. Furthermore, the LES has dedicated employers’ counsellors since 2017 aiming to meet the employers’ needs better (European Commission, 2017[8]). Nevertheless, while the image of the LES is getting better among employers, it has still scope for further improvements.

Jobseekers are motivated to contact and register with the LES, although their incentives to register might lie above all in becoming eligible for benefits and health insurance, rather than accessing ALMPs. Lithuania has one of the highest rates of jobseekers contacting the public employment service across the OECD countries, reaching 86.4% in 2020 (Figure 2.7). The stakeholders believe that eligibility for health insurance is the main reason behind that, although this might not fully explain that high rate as eligibility for health insurance has not induced the same rate of registration in other countries where similar conditions are applied. For example in Estonia where registration with the PES also provides health insurance coverage, only 67.2% of jobseekers were in contact with the public employment service in 2020, and even lower in the past years before the Work Ability Reform that made work ability allowance (one type of disability benefits) conditional on registering with the PES (OECD, 2021[9]). Furthermore, the stakeholders of the ALMP system (the authorities in charge of designing and implementing ALMPs as well as the social partners) in Lithuania tend to focus on the downside of this set-up – having additional clients who are not interested to get the LES support – rather than seeing it as an opportunity – being able to get in contact with and motivate people furthest from the labour market to engage and support them in job search. Furthermore, the question in the Labour Force Survey that these statistics are based on, specifically ask jobseekers about their contacts to public employment service to seek employment, suggesting the jobseekers in Lithuania might be more interested in the LES support and entering employment than the authorities think. Private employment services do not play a significant role currently as only few jobseekers contact them to seek employment.

Figure 2.7. A very high share of jobseekers contact the public employment service in Lithuania

Note: The purple bar represents the unweighted average of the 26 European countries shown.

Source: OECD calculations based on the EU Labour Force Survey dataset: Methods used for seeking work- Percentage of unemployed who declared having used a given method, by sex.

Statlink

2.3.2. The package of ALMPs has been revised, but positive changes are not yet visible

The introduction of the so-called social model has not made ALMPs more accessible in Lithuania. In 2017‑19, the drop in ALMP expenditures was in fact steeper than improvements in the labour market indicators (Figure 2.8). Furthermore, as ALMPs are mostly financed via European Social Fund (ESF) funding, the availability of ALMPs for jobseekers has fluctuated according to the ESF financing cycles rather than the needs of the labour market.

Lithuania spends less than half of the average of OECD countries on ALMPs (0.21% versus 0.45% of GDP in 2019). Allocations to the traditional package of ALMPs increased only marginally in 2020 (to 0.23% of GDP). Simultaneously, a massive funding was allocated to a job maintenance incentive (EUR 546 million, while the rest of the ALMP package received in total EUR 110 million), in addition to further allocations to income maintenance schemes (passive labour market policies). Lower allocations to ALMPs also mean lower accessibility for support for jobseekers and people at risk of job loss. In 2019, 1% of labour force participated in ALMPs in Lithuania, while this indicator stood at 5% in the OECD.

Figure 2.8. Lithuania invests little in active labour market policies relative to other OECD countries

PES: public employment service. GDP: gross domestic product.

Note: OECD average is an unweighted average. 2019 data for Australia and New Zealand regarding employment incentives refers to budget year July 2018 to June 2019 and not July 2019 to June 2020 unlike for the other ALMPs as this category was highly affected by the exceptional measures taken to address the challenges of COVID‑19. Data for Lithuania in 2020‑22 excludes the measure “Subsidies for wage after downtime” as an exceptional measure to tackle specific challenges caused by the COVID‑19 outbreak and not comparable with other measures through the years (i.e. the figure depicts actual data for employment incentives without the exceptional measures in 2020, and an estimation of costs without the exceptional measure for 2021‑22).

Source: EC-OECD Labour Market Policies Database, https://stats.oecd.org/Index.aspx?DataSetCode=LMPEXP; OECD LFS by Sex and Age – Indicators Database, http://stats.oecd.org//Index.aspx?QueryId=54218; Economic Outlook No 110, December 2021 https://stats.oecd.org/index.aspx?DataSetCode=EO; and the OECD Questionnaire on Policy Responses to the COVID‑19 Crisis (responses from Lithuania).

Statlink

Regarding the composition of ALMP package, Lithuania is very much focused on employment incentives rather than tackling actively individual employment obstacles via training measures and PES support. While employment incentives can be effective to support particularly vulnerable groups and during economic downturns (Card, Kluve and Weber, 2018[10]), Lithuania has spent somewhat more generously than the OECD average on these measures particularly during better labour market conditions. Furthermore, half of these expenditures cover wage subsidies for so-called social enterprises, which are not enterprises following necessarily a social objective, but are legally defined as enterprises that hire a certain level of people with disabilities. This scheme has been criticised by the stakeholders of the ALMP system as these enterprises might not support people with disabilities to get good jobs, but rather abuse the generous benefit scheme. Social enterprises have a tendency to hire the most employable from the overall target group, leaving those in need of support behind. The Lithuanian authorities have been revising the law regulating social enterprises already for a few years, aiming to prevent abuses and target the support better to those who need it (Pacifico et al., 2018[3]) with more significant changes to improve targeting implemented in 2020 (OECD, 2020[11]), but the challenges of the scheme have not yet been entirely solved. As of May 2022, the government has submitted to the parliament (Seimas) a draft amendment to the Law on Employment with the aim to improve labour market opportunities for people with disabilities, above all via support aiming at primary labour market integration rather than employment incentives for social enterprises. The next chapters of the current report evaluate the effectiveness of employment incentives that are targeted to the vulnerable groups to be integrated in the primary labour market and have thus higher potential to be effective.

Although the social model aimed to prioritise and improve training provision for jobseekers, expenditures on training relative to GDP have slightly decreased since 2017. Training measures have also been more extensively redesigned in the past three years, but not all of these schemes have started to function and taken up as anticipated. For example, an apprenticeship programme to target the needs of employers has not been as appealing for employers as expected. Not many companies have been prepared and willing to conduct work-based training, and rather opt for using employment subsidies to recover for some training on the job during the beginning of the employment period (i.e. employment subsidies are financially more attractive to the employers than the apprenticeship scheme). Also, targeting training to those with lower skills and introducing modular training has taken some time and efforts, as initially in 2019 modular training was introduced only to people who had higher education levels, forcing people with lower education to undertake only longer education programmes and not enabling them to integrate into the labour market quickly. Furthermore, COVID‑19 outbreak hindered employers to commit to tripartite training agreements with the LES and jobseekers to provide jobs after successful completion of training. Digital training solutions during the pandemic were not well taken up by the jobseekers and training providers and trainings that used to be previously fully in classroom suffered in quality when conducted digitally, also due to gaps in digital skills. The next chapters of this report evaluate one of the key training programmes – the vocational training programme – provided by the LES and provide recommendations on scaling it up and redesigning to better meet the needs of the labour market as presented in the previous section.

Lithuania essentially does not allocate resources for sheltered and supported employment and rehabilitation, as well as direct job creation (public works) schemes. Indeed, evaluations in other countries have shown that public works do not have positive effects on the participants’ labour market outcomes, or can even harm them (Card, Kluve and Weber, 2018[10]). Very general sheltered and supported employment schemes might not be effective measures either, although these could be a solution in case these are accompanied by a broader set of support together with training and job search assistance and when having integration into the primary labour market as the final objective (OECD, 2021[12]). Limiting the support to the most vulnerable groups currently through the wage subsidies for social enterprises is not likely addressing the pathways to labour market for the most vulnerable sufficiently well.

With the introduction of the social model in 2017, Lithuania abolished dedicated measures to encourage entrepreneurship as the existing measures were considered ineffective (yet, there were no counterfactual impact evaluations conducted about these specific measures (PPMI, 2015[13])). Nevertheless, taking up self-employment is currently supported via employment incentives (above all covering wage costs), possibilities to receive vocational training (similarly to other jobseekers), and basic training on business. In 2020, Lithuania introduced a temporary innovative measure for self-employed in response to COVID‑19 to support them change their economic sector and so continue being employed during the pandemic (OECD, 2021[14]; European Commission, 2021[15]). Yet, systematic support to encourage entrepreneurship, develop business plans, and offer coaching and training during the initial phases of entrepreneurship is not provided. A more comprehensive business start-up support can be effective for some smaller groups of jobseekers, such as people living in areas where suitable vacancies for them are missing.

The social model also aimed to modernise and strengthen the LES, but this does not reflect in the level of expenditures on PES administration and the number of PES staff. A public employment service with more efficient and effective processes and administration and a modern infrastructure would need indeed less resources to achieve good results. Nevertheless, achieving a modernised employment service would need first investments, such as in the IT infrastructure and staff skills, as well as in building partnerships with employers and other stakeholders.

Sufficiency of resources for ALMPs in Lithuania is looking more promising for the few years ahead with the help of additional European Union funding (Resilience and Recovery Facility, ESF+). Additional allocations were made also in the 2021 and 2022 budgets for the LES and most of the ALMP measures, although barely keeping up with inflation and GDP growth. Nevertheless, the budget increases in 2022 have the potential to create an ALMP package that matches better the labour market needs, as the budget is most notably strengthened for the LES (10.3%), training (50.8%) and sheltered and supported employment and rehabilitation (174%, but starting from a very low level, so a significant difference from the OECD average level remains).

2.3.3. Lithuania needs to strengthen its system of ALMP monitoring and evaluation to attract sufficient and sustainable funding for ALMPS that are effective

Low spending on ALMPs in Lithuania is linked to the low priority of ALMPs for policy makers. As such, ALMPs receive only minor allocations from the national resources and are currently financed mostly using EU funding (i.e. national funding is above all filling the role of mandatory co-financing to be able to use EU funding). This financing mechanism makes the available resources fluctuate with the EU funding cycles, inflexible to take labour market changes into account quickly, and is not sustainable in the long run.

Fully fledged evidence‑informed policy making needs to be developed in the system of ALMPs to ensure that policies that are effective in supporting jobseekers and employers achieve sustainable funding. Having credible evidence on the effectiveness of ALMPs and the LES would help the Ministry of Social Security and Labour and the LES communicate this evidence to the public and policy makers and attract the resources needed to provide ALMPs. Evidence‑informed policy making needs to be systematic and involve the whole cycle of designing, monitoring and evaluation frameworks, generating knowledge, disseminating knowledge, adjusting policies based on evidence, as well as evaluating the knowledge generation process itself and adjusting the monitoring and evaluation framework accordingly. Knowledge generation needs to involve ex-ante evaluations in designing policies, monitoring frameworks to enable agile overviews on policy implementation, and ex-post process and impact evaluations to understand what works, for whom and how. Credibly evaluating the impact of policies allows identifying the need to adapt or terminate inefficient policies and boost the efficient ones. Process evaluations help to design more efficient policy implementation practices. Generating evidence and designing policies based on evidence is not important only regarding specific labour market services (such as the LES counselling services) or measures (training, employment incentives), but also across the tools, processes and approaches that the LES uses. The European Commission has highlighted the importance of such CIEs, including of measures implemented with ESF and ESF+ support, and the collection and use of administrative data (see Box 2.1).

Box 2.1. Useful resources from the OECD and the European Commission for countries building their capacity to conduct counterfactual impact evaluations of ALMPs

The OECD and the Directorate General for Employment, Social Affairs and Inclusion of the European Commission (EC, DG Employment), in co‑operation with the Competence Centre on Microeconomic Evaluation (CC-ME) of the European Commission’s Joint Research Centre (JRC), are working on a multi-year project that aims to help countries build or strengthen their analytical capacity and their use of linked administrative and survey data (OECD, 2022[16]). The current report presenting the impact evaluation results for vocational training and employment subsidies in Lithuania was prepared in the framework of this OECD-EC joint project.

The OECD-EC project consists of two phases running between 2019 and 2024. The main report of the first phase in 2019‑20 finds that in the majority of the 34 EU and OECD countries studied, administrative data on registered unemployment and labour market policies can be linked with data on employment outcomes (OECD, 2020[17]). However, most countries still need to make significant investments in linking data from registers containing information on income, social assistance and incapacity benefits. In addition, the report provides practical advice on how to use impact evaluations to assess labour market policies, and illustrates this with several country examples and best practices.

The second phase of the project, which started at the end of 2020, includes country-specific work in at least five other EU and OECD countries besides Lithuania (Finland, Greece, Ireland and another country to be determined in 2022). In the same overall framework, the OECD has also just carried out an assessment of the system of ALMP impact evaluation in Canada, with funding provided by Employment and Social Development Canada, which provides many good practices regarding conducting high-quality CIEs by public administrations internally (OECD, 2022[18]). During 2023‑24, the OECD-EC project will offer peer-learning opportunities via a technical workshop, a high-level policy exchange, as well as a synthesis report sharing lessons and good practices of the EU and OECD countries in conducting CIEs of ALMPs using linked administrative data and using the evidence for better policies.

The OECD-EC project builds on recent and ongoing related projects undertaken by the OECD and the EC, some of which can also be used as guidelines when conducting CIEs of ALMPs, particularly by national authorities. For example, the Centre for Research on Impact Evaluation (CRIE) of the CC-ME of the JRC has published a guideline for advanced CIE methods (European Commission, 2019[19]), as well as guidelines tailored to national authorities evaluating the impact of ESF (European Commission, Directorate-General for Employment, Social Affairs and Inclusion, 2020[20]; European Commission, 2020[21]). Furthermore, CRIE has supported many countries to conduct CIEs of ESF interventions via the Data Fitness Initiative for Counterfactual Impact Evaluation in 2016‑18 (for example in Flanders (Belgium), Ireland, Latvia, Portugal and Umbria (Italy)), and its successor Quality Assurance Support for CIE launched in 2019 that promotes CIEs of ESF funded interventions and goes beyond the data related aspects. Furthermore, elements of monitoring and evaluation using administrative data are often parts of projects relating to ALMPs that the OECD implements with funding from and in co‑operation with the European Commission’s Directorate‑General for Structural Reform. These support individual EU Member States to design and implement resilience‑enhancing reforms, but provide learning opportunities also for other countries (see for example an Impact Evaluation Framework tailored for Spain (OECD, 2020[22]).

Source: European Commission (2020[23]), Data Fitness Initiative for CIE, https://knowledge4policy.ec.europa.eu/microeconomic-evaluation/data-fitness-initiative-cie_en; European Commission (2020[20]), Counterfactual impact evaluation of European Social Fund interventions in practice: guidance document for managing authorities, https://data.europa.eu/doi/10.2767/721497; European Commission (2020[21]), How to use administrative data for European Social Funds counterfactual impact evaluations: a step-by-step guide for managing authorities, https://data.europa.eu/doi/10.2767/721497; European Commission (2019[19]), Advanced counterfactual evaluation methods: guidance document, https://data.europa.eu/doi/10.2767/464242; European Commission (2019[24]), Quality Assurance Support (QAS) for CIE, https://knowledge4policy.ec.europa.eu/microeconomic-evaluation/quality-assurance-support-qas-cie_en; European Commission, Joint Research Centre (2020[25]), JobsPlus evaluation, https://data.europa.eu/doi/10.2760/986782; European Commission, Joint Research Centre (2020[26]), Active labour market policies in Flanders: evaluation of the ESF “Work Experience for Young Persons” programme, https://data.europa.eu/doi/10.2760/623819; European Commission, Joint Research Centre (2020[27]), The evaluation of the youth employment initiative in Portugal using counterfactual impact evaluation methods, https://data.europa.eu/doi/10.2760/368100; European Commission, Joint Research Centre (2017[28]), Counterfactual impact evaluation of “Work Experience Laureati e Laureate – WELL” (Work Experience for Graduates): the impact of an ESF-funded intervention in Umbria region, https://data.europa.eu/doi/10.2760/01166; OECD (2022[18]), Assessing Canada’s system of impact evaluation of active labour market policies, https://doi.org/10.1787/27dfbd5f-en. OECD (2022[16]), OECD-EC project on policy impact evaluation through the use of linked administrative and survey data, https://www.oecd.org/els/emp/impact-evaluation-linked-data.htm; OECD (2020[17]), Impact evaluation of labour market policies through the use of linked administrative data, https://www.oecd.org/els/emp/Impact_evaluation_of_LMP.pdf; OECD (2020[22]), Impact Evaluations Framework for the Spanish Ministry of Labour and Social Economy and Ministry of Inclusion, Social Security and Migrations, https://www.oecd.org/els/emp/Impact_Evaluations_Framework.pdf.

Lithuania has improved its monitoring and evaluation framework of ALMPs significantly in the context of the 2017 reform and the introduction of the social model. Since 2017, the LES has made continuous efforts to generate more knowledge on ALMPs, has gained access to more data from other administrative registers that support monitoring and evaluation, and is improving its IT infrastructure to support data management better more generally. Nevertheless, there is still a lot of scope for improvement as impact evaluations are not conducted systematically, and data exchange with other registers focuses on operational purposes and does not fully take into account the needs for data analytics. Furthermore, the IT infrastructure has no modern solutions to support data analytics well as the main IT systems do not include Data Warehouse or similar solutions, but only limited built-in queries directly to the operational database. There are also no dedicated solutions yet to facilitate data access for researchers, and such data exchanges are implemented via ad-hoc queries and file sharing solutions.

Most importantly, the introduction of the social model created a strong legal basis for ALMP monitoring and evaluation. The Law on Employment implemented in 2017 puts the task of generating and disseminating knowledge on labour market and ALMPs on the LES and states that the LES (and potentially other organisations implementing ALMPs) need to evaluate the effectiveness of ALMPs they provide following the procedures set by the government, as well as make the evaluation results public.

A decree by the Minister of Social Security and Labour (Lietuvos Respublikos socialinės apsaugos ir darbo ministerija, 2017[29]) and an order by the Director of the LES (Užimtumo tarnyba prie Lietuvos Respublikos socialinės apsaugos ir darbo ministerijos, 2020[30]) set the processes and methodology for ALMP evaluation activities, limiting these to monitoring the gross impact of key ALMPs. This methodology observes employment rates and rates of registered unemployment of participants in training programmes (in total six different policies) in intervals up to two years after participation, job maintenance rates of employment incentives (three different policies) up to four years of creating the jobs, and customer surveys (jobseekers and employers) to assess the LES service provision. These indicators are important to provide some knowledge on the labour market outcomes of ALMP participants, but do not provide credible evidence whether the labour market outcomes are affected by participation in ALMPs. The evidence on ALMP impact is only possible to generate using counterfactual impact evaluations (CIE) comparing the outcomes of ALMP participants to credibly comparable non-participation (see more details in Chapter 3). Furthermore, the current monitoring framework gives some indication on labour market integration and employment sustainability, but the results are not easily comparable over time as the labour market situation changes, and do not cover well job quality, such as the aspects of income and career progression.

Although the current methodology to evaluate the impact of ALMPs does not cover CIEs, the LES and the Ministry of Social Security and Labour are aiming to build the capacity to conduct CIEs of ALMPs systematically and have conducted a few CIEs in co‑operation with research organisations in the past. These evaluations have most often covered training measures and employment incentives, as well as to some extend other schemes (see the most recent evaluations by ESTEP (2019[31]; 2016[32]) and PPMI (2015[13]), the latter providing also an overview of previous evaluations). The more recent evaluations tend to find that employment subsidies have positive impact on the participants’ labour market outcomes, while the effects of training programmes are more mixed. All of these reports call for more systematic impact evaluations, and further improvements in available data to enable improving the credibility of future impact evaluations. The evaluation of vocational training and employment subsidies presented in the next chapters of this report aim to support the LES and the Ministry of Social Security and Labour to start conducting similar CIEs regularly and across ALMPs.

References

[10] Card, D., J. Kluve and A. Weber (2018), “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.

[31] ESTEP (2019), Aktyvios darbo rinkos politikos priemonių bei tarpininkavimo įdarbinant ir įdarbinimo tvarumo stebėsenos ir vertinimo sistemos sukūrimo paslaugos.

[32] ESTEP (2016), 2007–2013 m. ES struktūrinės paramos poveikio užimtumui ir kitiems Lietuvos ūkio makroekonominiams rodikliams vertinimas, https://www.esinvesticijos.lt/lt/dokumentai/2007-2013-m-es-strukturines-paramos-poveikio-uzimtumui-ir-kitiems-lietuvos-ukio-makroekonominiams-rodikliams-vertinimas.

[15] European Commission (2021), Measure to support self-employment for those changing their economic activity, https://ec.europa.eu/social/BlobServlet?docId=23517&langId=en.

[23] European Commission (2020), Data Fitness Initiative for CIE, https://knowledge4policy.ec.europa.eu/microeconomic-evaluation/data-fitness-initiative-cie_en (accessed on 8 June 2022).

[21] European Commission (2020), How to use administrative data for European Social Funds counterfactual impact evaluations : a step-by-step guide for managing authorities, Publications Office, https://data.europa.eu/doi/10.2767/721497 (accessed on 11 October 2020).

[19] European Commission (2019), Advanced counterfactual evaluation methods : guidance document., Publications office, https://doi.org/10.2767/464242.

[24] European Commission (2019), Quality Assurance Support (QAS) for CIE, https://knowledge4policy.ec.europa.eu/microeconomic-evaluation/quality-assurance-support-qas-cie_en (accessed on 8 June 2022).

[5] European Commission (2019), Success Story: Organisation development at the Lithuanian Employment Services, https://ec.europa.eu/social/BlobServlet?docId=20750&langId=en.

[8] European Commission (2017), Targeted service provision for employers – employer specialist counsellors, https://ec.europa.eu/social/BlobServlet?docId=18822&langId=en.

[20] European Commission, Directorate-General for Employment, Social Affairs and Inclusion (2020), Counterfactual impact evaluation of European Social Fund interventions in practice : guidance document for managing authorities, Publications Office, https://data.europa.eu/doi/10.2767/55495.

[26] European Commission, Joint Research Centre (2020), Active labour market policies in Flanders : evaluation of the ESF “Work Experience for Young Persons” programme, Publications Office, https://doi.org/10.2760/623819.

[27] European Commission, Joint Research Centre (2020), he evaluation of the youth employment initiative in Portugal using counterfactual impact evaluation methods, Publications Office, https://doi.org/10.2760/368100.

[25] European Commission, Joint Research Centre (2020), JobsPlus evaluation, Publications Office, https://doi.org/10.2760/986782.

[28] European Commission, Joint Research Centre (2017), Counterfactual impact evaluation of “Work Experience Laureati e Laureate – WELL” (Work Experience for Graduates) : the impact of an ESF-funded intervention in Umbria region, Publications Office, https://doi.org/10.2760/01166.

[6] Lauringson, A. and M. Lüske (2021), “Institutional set-up of active labour market policy provision in OECD and EU countries: Organisational set-up, regulation and capacity”, OECD Social, Employment and Migration Working Papers, No. 262, OECD Publishing, Paris, https://doi.org/10.1787/9f2cbaa5-en.

[29] Lietuvos Respublikos socialinės apsaugos ir darbo ministerija (2017), Dėl Darbo rinkos stebėsenos sąlygų ir tvarkos aprašo patvirtinimo, https://e-seimas.lrs.lt/portal/legalAct/lt/TAD/f4bb4fc0578611e78869ae36ddd5784f.

[18] OECD (2022), Assessing Canada’s System of Impact Evaluation of Active Labour Market Policies, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/27dfbd5f-en.

[1] OECD (2022), Gender wage gap (indicator), https://doi.org/10.1787/7cee77aa-en (accessed on 9 June 2022).

[16] OECD (2022), OECD-EC project on policy impact evaluation through the use of linked administrative and survey data, OECD, Paris, https://www.oecd.org/els/emp/impact-evaluation-linked-data.htm.

[2] OECD (2022), Reaching Out and Activating Inactive and Unemployed Persons in Bulgaria, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/7b91154a-en.

[12] OECD (2021), “Building inclusive labour markets: Active labour market policies for the most vulnerable groups”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/607662d9-en.

[14] OECD (2021), “Designing active labour market policies for the recovery”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://doi.org/10.1787/79c833cf-en.

[9] OECD (2021), Improving the Provision of Active Labour Market Policies in Estonia, Connecting People with Jobs, OECD Publishing, Paris, https://doi.org/10.1787/31f72c5b-en.

[7] OECD (2021), Institutional and regulatory set-up of active labour market policy provision in Denmark, OECD, Paris, https://www.oecd.org/els/emp/Denmark_ALMP_Institutional_set-up.pdf.

[17] OECD (2020), Impact evaluation of labour market policies through the use of linked administrative data, OECD, Paris, https://www.oecd.org/els/emp/Impact_evaluation_of_LMP.pdf.

[22] OECD (2020), Impact Evaluations Framework for the Spanish Ministry of Labour and Social Economy and Ministry of Inclusion, Social Security and Migrations, https://www.oecd.org/els/emp/Impact_Evaluations_Framework.pdf.

[11] OECD (2020), OECD Economic Surveys: Lithuania 2020, OECD Publishing, Paris, https://doi.org/10.1787/62663b1d-en.

[4] OECD (2016), Back to Work: Denmark. Improving the Re-employment Prospects of Displaced Workers, OECD Publishing, Paris, https://doi.org/10.1787/9789264267503-en.

[3] Pacifico, D. et al. (2018), “Faces of Joblessness in Lithuania: A People-centred perspective on employment barriers and policies”, OECD Social, Employment and Migration Working Papers, No. 205, OECD Publishing, Paris, https://doi.org/10.1787/3657b81e-en.

[13] PPMI (2015), Final Report on Counterfactual Impact Evaluation of ESF Funded Active Labour Market Measures in Lithuania, https://doi.org/10.13140/RG.2.2.12469.83689.

[30] Užimtumo tarnyba prie Lietuvos Respublikos socialinės apsaugos ir darbo ministerijos (2020), Dėl darbo rinkos stebėsenos vykdymo, https://www.e-tar.lt/portal/lt/legalAct/f234e860f80511eaa12ad7c04a383ca0.

Note

← 1.  The gender wage gap is defined as the difference between median earnings of men and women relative to median earnings of men. Data refer to full-time employees on the one hand and to self-employed on the other.