The danger of long-term unemployment and measures for its reduction: the case of Croatia*

Abstract Long-term unemployment (when a person is jobless for more than 12 months) can have serious negative consequences for the individual, society and its economic system. People who are unemployed for a long-term period find it more difficult to be employed as time goes by. Long-term unemployed persons often have limited employability, due to their low level of educational attainment, loss of skills and the substantial expenditures that are necessary to restore these skills. They also face significantly lower earnings and restricted career prospects. The participation rate of the long-term unemployed in employment programmes is usually very low. The article considers the significance of long-term unemployment in Croatia and measures for its reduction.


Introduction
Long-term unemployment (when a person is jobless for more than 12 months) can have serious negative consequences for the individual, society and its economic system. People who are unemployed for a long-term period find it more difficult to be employed as time goes by. They also face significantly lower earnings and restricted career prospects. Workers' human capital (whether actual or as perceived by employers) may deteriorate or become obsolete during a long period of unemployment. Furthermore, the time and will dedicated to searching for a job typically decline. Both elements suggest that the probability of leaving unemployment decreases significantly with its duration, leading to an increased possibility of remaining unemployed. Over time as unemployment continues, long-term unemployed persons probably leave the labour force and retire, enrol in disability programmes, or simply become discouraged workers. At the national level, a high share of long-term unemployment endangers overall employment-policy goals, lowers the ability to allocate the right jobs to the right people and lowers professional and geographical mobility.
In general, long-term unemployment is more or less a determinant of social exclusion. It is a multidimensional process that weakens the links between an individual and the community. The characteristics of exclusion are related to access to the labour market, basic social services and the social network. One of the most important forms of deprivation is limited access to the labour market.
The long-term unemployment rate in Croatia is significantly higher than the EU average. In 2012, the share of long-term unemployment on total unemployment peaked at 63.7% and decreased to 50.7% in 2016 (European Commission, 2017). This was still above the EU average of 46.4%. In 2016, eight other member-states had a share of longterm unemployment in total unemployment higher than the EU average. They were Greece (72%), the Slovak Republic (60.3%), Bulgaria (59.1%), Italy (57.4), Portugal (55.4%), Slovenia (53.3%), Ireland (53.2%) and Belgium (51.6%). In 2016, the share of very long-term unemployment (those who were waiting for more than two years) in total unemployment in Croatia was 33.3%, which was higher than in other EU countries. Regarding age, the highest share of the long-term unemployed in Croatia is in the 55-59-year-old age group. Regarding educational attainment, long-term unemployment is recorded more by unemployed persons with only elementary education (36%).
The paper commences with the section dedicated to the significance of employment and unemployment, followed by a section oriented towards the literature review. The third section describes characteristics of long-term unemployment in Croatia, while section four explains realised measures for improvement. The last part contains conclusions and recommendations. The authors wanted with this paper to improve existing knowledge on the situation of long-term unemployment in Croatia and present measures for the improvement of such a problem as well as their effects.

The importance of employment and danger of long-term unemployment
Employment has an essential role in all societies. People often define themselves and are frequently demarcated by others by what they do for a living. Economic and sociological research and studies underline that employment is not only the most important determining factor of status and human well-being, but it is also vital for generating the feeling of one' meaning of life, social stability and securing participation in society (Jahoda, 1982). Employment is deemed 'the glue that holds society together'. If people are working, it is unlikely that they will be poor. Nonetheless, some of those employed are also poor due to their low educational attainment and low income. Consequently, efforts are made for people to work and earn a salary, and thus be in a better position than those who obtain benefits and assistance in the unemployment and social-welfare systems.
Unemployment is particularly dangerous if it lasts for a long time (for one year or longer). Long-term unemployment not only lowers employability and the likelihood of finding a job in the future, but also causes people to become isolated and sick (Spermann, 2015). The loss of a social network because of dismissal or the impossibility of finding an adequate job causes a serious shock. It can even force otherwise selfassured persons to become unconfident when applying for jobs. Nichols et al. (2013) explain that continuing unemployment decreases both the potential wages of the unemployed person and the chances of finding a new job. That leads to an erosion or depreciation of human capital, while long-term unemployment will probably also reduce a person's social capitalthe network of acquaintance and business contacts that enable the easier finding of a good, new job. Social capital may decrease with a longer duration of unemployment because the social network established through work contact can decline when work contact stops, or because being out of work is increasingly stigmatising the longer a person cannot find new employment. The erosion of social capital means that the longer a worker is unemployed, the less likely he or she is to find a new job (Layard et al., 2009).
There are various theoretical links connecting duration of unemployment to adverse outcomes, depending on many characteristics of the labour market, government activity, organisations and institutions, roles and power of social partners, labour legislation and so on. Thus, for example, many elements on the supply side on the labour market lower the probability of finding a job for recipients of social welfare, particularly those with low employability. One of the possible explanations can be employers' herd behaviour because they act as others do. An easy explanation can be that the long-term unemployed simply are not selected for employment because employers want candidates with higher education, more qualifications or experience for the jobs that come open. A more gloomy possibility is that long-term unemployment is at least partly self-reinforcing: employers are reluctant to hire someone others have already passed over. To find out, Kroft et al. (2013) performed an experiment in which they applied for 3000 various jobs, mostly administrative, clerical, sales and customer service, advertised online by submitting 12,000 invented CVs. It was intended that applicants with similar education, backgrounds and experience went for the same job. The only difference was how long the applicant had been without a job, a period that ranged from no time at all to as much as three years. The researchers found that the chances of a candidate being called back by an employer declined progressively as the duration of unemployment grew longer, from 7.4% after one month of unemployment to 45% at the eighth month, where the rate for invitations for a job interview stabilised. The authors state that employers could not find some hidden qualitative flaw in the long-term unemployed persons, since the candidates were alike in other respects. Another possible explanation is that employers obviously deem that people make less effort to find a job as their time out of employment lengthens. A third possible explanation is that employers equate prolonged unemployment with obsolete and deteriorating skills, which cause a decline in productivity. While probably all three factors contribute to the less likely employment of the longterm unemployed, there is also an impact of herd (peer) behaviour, so employers are reluctant to hire someone other employers do not want.
According to many international experiences Martinson, 2002a, 2002b;European Commission, 2017;Nichols et al., 2013), even after participation in various employment programmes, the majority of long-term unemployed persons (many of whom are recipients of social-welfare benefits) have very limited possibilities of finding a job. Without doubt, such persons face multifaceted obstacles in their search for employment, mostly have low educational attainment, live in areas with inadequate traffic connections, are of poor health or have difficulties with socialisation. Very often, they suffer from physical incapacities and/or disability, have learning difficulties and psychological problems, have experienced domestic violence, can be addicted to alcohol or drugs and so on.
1.2. Literature review on long-term unemployment and its effects Gallie and Paugam (2000, p. 255) concluded that people whose social network consists of similar unemployed persons have greater vulnerability problems when exposed to psychological and financial difficulties, which make it more difficult for them to escape unemployment and poverty. The Acheson Report (Acheson, 1998) indicates that unemployed people, particularly if they are recipients of social-welfare benefits, have a substantially lower level of psychological stability and are often prone to depression and dissatisfaction, even to the extent of causing self-injury and committing suicide. Long-term unemployment is particularly harmful for health (Nichols et al., 2013). It extinguishes personal capacities and opportunities, not only material, but also mental and social ones.
Additionally, the anxiety of long-term unemployment might influence a person's physical and mental health, family dynamics and the well-being of his or her children. Moreover, the negative impact of unemployment on health accumulate over time. Burgard et al. (2007) found a significant decline in self-reported health status following job loss, even after taking into account various traits of people without a job. Losses are more pronounced among those who lose jobs for health reasons, but job losses for other reasons increase depressive symptoms. Many applications without positive outcomes lead to feelings of discouragement, worthlessness and depression, loss of self-reliance, and disbelief in the future. B€ ockerman and Ilmakunnas (2009) found that the health status of those who end up unemployed is statistically significantly lower than that of the employed. Knabe et al. (2010) compared employed and unemployed people regarding their subjective estimation of emotional affects, as well as differences in the composition and duration of daily activities. Their results show that employed persons are more content with their lives than the unemployed are. Involuntary job loss is a stressful event, creating a variety of problems immediately, and long periods of unemployment can multiply those problems. Long-term unemployment intensely affects various dimensions of jobless people's wellbeing, particularly among the younger population. It decreases overall life satisfaction and lowers optimism about the future.
Long-term unemployment can likewise have indirect long-lasting adverse outcomes. While a person is jobless, her family income falls due to the shortage of regular earnings, and that loss of income (which becomes more serious with prolonged duration of unemployment) can influence not only the unemployed, but also other members of her household. The loss of income will probably endanger the quantity and quality of goods and services the family can purchase. Being unemployed for a long time is more likely to put stress on a relationship with a partner than to improve it. More than 20% of the long-term unemployed state that their joblessness has hurt their relationship with their spouse or partner, while less than 10% say the troubles have brought them closer (Geewax, 2011). Bratberga et al. (2008), using data from Norway, analysed the effects of worker displacement in the period 1986-1987 on their future children's earnings and concluded that job losses have a negative effect on earnings and the employment of those affected. Furthermore, dealing with the loss of income can exacerbate stress or motivate family members to participate in some informal or illegal activities. Long-term unemployment also includes a lack of expectations for future income. Therefore, the impossibility of paying for necessary medical treatment and drugs can easily endanger the health conditions of the long-term unemployed and members of their families. Lindo (2011) showed an adverse impact of paternal unemployment on infant birth weight. Rege et al. (2011) determined that paternal job loss lowers children's school performance, and the harmful consequences of paternal unemployment have not diminished by income obtained by the spouse's employment, marital termination or residential relocation. Stevens and Schaller (2011) studied the relationship between parental job loss and academic achievement by their children using data on job loss and grade retention (keeping a child in the same grade for more than one year) from 1996, 2001 and 2004. They found that a parental job loss increases the probability of children's grade retention by around 15%. Wightman (2012) estimated the effect of parental job loss on children's educational attainment. He focused on the role of parental cognitive and non-cognitive ability, and found that a parental job loss during childhood lowers the probability of an offspring obtaining any post-secondary education (by age 21) by a minimum of 10% and a maximum of 50%. Therefore, the negative consequences of long-term unemployment easily transfer to the next generation.
Considering the causal systems through which long-term unemployment relates to effects both during and after unemployment is crucial in proposing and implementing effective employment policy measures to reduce the adverse consequences of job loss. Policy responses to the problems of long-term unemployment may have different outcomes depending on which of the possible explanations is dominant in accounting for any particular outcome. Given the many levels of disadvantage that long-term unemployed people are more likely to face, a multidimensional policy response is required, one that includes holistic innovative new approaches in policy design and implementation. Such an approach includes elements such as individualised counselling, mentoring and appointment to specialised support, suitable education, training and job placements, as well as flexible and sustained support through all stages of the programme. An efficient active employment policy may help in alleviating (reducing) the barriers between knowledge and skills, or long-term unemployed persons and employers' needs.
Economic theory deems that an active labour-market policy has almost no immediate effect on overall employment (de Koning and Mosley, 2001;Stupnytskyy, 2004;Eurofound, 2017), but that the efficiency of labour supply would increase if those long-term unemployed or recipients of assistance in the social-welfare system could find a job. Consequently, the prevention of long-term unemployment or the inclusion of long-term unemployed persons in the world of work is without doubt worth considering. Although the introduction of such measures may increase the likelihood of negative impacts, the risk is acceptable if the overall impact is a reduction in the average duration of unemployment. Educational, training and employment programmes have the strongest impact and the highest social return for those with the least previous experience in the labour market and for those that are the most vulnerable (Bj€ orklund et al., 1991;Spermann, 2015).
Long-term unemployed persons often have limited employability due to their low level of educational attainment, loss of skills and the substantial expenditures that are necessary to restore these skills. The younger the unemployed individual and the shorter his or her unemployment period is the more likely he or she will be included into an employment programme. The participation rate of long-term unemployed in employment programmes is usually very low (European Commission, 2017; ICON Institut Public Sector GmbH and Quantos S.A, 2017). The reasons for the low participation of this group in various employment programmes are manifold and mostly caused by their lower motivation related to their narrow and/or obsolete knowledge and skills acquired in formal education or no formal education whatsoever. Furthermore, some of them are unwilling to disturb their present inactive lifestyle because they are scared that with participation in employment programmes and possible employment, they will end up in a worse position. Such passivity in their participation in training programmes, search for a job and reluctance to accept jobs is in some cases reasonable. In some cases, they are unwilling to accept available low-paid jobs that bring very limited or no financial benefits, but it is mostly incorrect because a person can soon be promoted to more interesting and better-paid jobs.
For groups like the long-term unemployed that are in an unfavourable position in the labour market, the government provides interventions through various employment measures. Measures are a temporary support with the goal of activating the unemployed, to help people move from involuntary inactivity into employment, or maintain the jobs of persons threatened by unemployment. Programmes oriented towards long-term unemployed persons consist of four groups of activities (Bonoli, 2010;ICON Institut Public Sector GmbH and Quantos S.A., 2017). The first group is publicly financed training that by enhancing the job skills of jobless people aims to improve their employability and enhance their chances of finding employment. The second are employment incentives that facilitate the recruitment of unemployed persons. The third are direct job-creation measures that create additional jobs, usually of a community benefit or which are socially useful, in order to activate and employ the long-term unemployed. The fourth group are start-up incentives that promote entrepreneurship by encouraging the unemployed to start their own businesses or to become self-employed.
In order to avoid inadequate results, an employment policy should adjust to the changed economic conditions, and focus on long-term unemployed persons, improving their employability, skills and motivations. Consistency and coordination of employment policy and social-welfare measures are particularly important, so that they can mutually reinforce each another. Of course, coordination is desirable, but difficult to achieve. The next section provides information on the situation with longterm unemployment in Croatia.

The situation with long-term unemployment in Croatia
Declining production in Croatia in the early 1990s and the consequent reduction in the number of employed due to the Homeland War and economic transition were not surprising. However, it is less clear why high, particularly long-term, unemployment has remained so constant. The main concern for the economic analysis and policy measures is not so much why until recently unemployment increased at very high rates, but why some people find it so difficult to get out of unemployment.
According to administrative data collection carried out in 2017, long-term unemployed persons (LTU) in Croatia who wait for a job more than one year corresponded to 6.1% of the active population. Of the mentioned rate, 1.3% are persons unemployed for 12-18 months and 4.8% are persons unemployed for more than 18 months. These rates were significantly higher than those for the EU but have decreased since 2013. The transition rate of LTU to employment (based on Labour Force Survey) was 10.3%. The probability of remaining unemployed between successive quarters was 60.9%. Only 1.4% of LTU participated in education and training, while in Poland the figure is 2%, in Lithuania 2.2% and in the Czech Republic 2.9%. The EU average was 7.5%, with almost 50% of LTU in Sweden and more than 25% in Denmark participating in education and training. Most LTU in Croatia use employment incentives. The at-risk-of-poverty rate for this population in Croatia was 47.3%, while that of persons in work was 5.8%. Only 7.4% of all long-term unemployed were receiving benefits and social benefits (all data are from ICON Institut Public Sector GmbH and Quantos S.A., 2017).
Educational attainment has a significant impact on the possibility of finding a job as well as on the duration of unemployment. For instance, at the onset of the crisis in 2008, the share of long-term unemployment among individuals with primary education or less was almost 75%. These shares were lower for individuals with three-and fouryear secondary education (54% and 48% respectively), and even lower for those with college and higher educational qualifications (43% and 40% respectively). In 2016, regarding the duration of unemployment in terms of the level of education, persons with lower levels of education account for the greater share of long-term unemployed persons. Thus, the share of long-term unemployed persons (one year or more) in the total number of unemployed persons with no schooling or uncompleted basic school was 71.4%. The share of long-term unemployed persons was smaller in the groups of persons with post-secondary non-university and university and postgraduate education (30% and 29.7%). Although the number of unemployed persons with postsecondary non-university and university and post-graduate education increased significantly, it can be concluded that the level of education has a significant influence on the duration of unemployment. Long-term unemployment is more frequent in the case of women than in the case of men. This difference in the share of men and women in long-term unemployment is particularly evident in the segment of long-term unemployed persons who had been unemployed for up to three months (more men) and more than three years (more women) (Croatian Employment Service, 2015).
From the position of long-term unemployed persons who are social-benefit users and recipients of unemployment benefit, the existing model of social benefits may create work disincentives because their amounts steeply decrease when an individual becomes employed. This discourages (re)employment and/or activation of inactive persons. For Croatia, Bejakovi c, Urban and Bezeredi (2014) have shown that almost all inactive people who are beneficiaries of social welfare would have benefits from employment, but these benefits are relatively small. It was found that there are cases and types of families for whom the incentives are almost negligible due to very high marginal effective tax rates when the move from unemployment to employment occurs. According to calculations in Croatia regarding 'does work pay', vulnerable families are those in which a member who is employed can only earn a low wage, families that receive all types of available benefits and assistance, and families with only one adult member who works.
Croatia is, no doubt, a country with a high share of long-term unemployment, so efforts aimed at its reduction are highly important. Obviously, the suppression of (particularly long-term) unemployment is a demanding and expensive task, which becomes even more expensive if these efforts are delayed. In the further text, activities in Croatia regarding the employment of the long-term unemployed are explained.

Overview of the employment measures for long-term unemployed persons
Out of all the beneficiaries of financial assistance (mostly support or maintenance allowance) in the social-welfare system, unemployed persons able to work account for around half. No doubt they have a certain (although limited) level of employability and their activation in the world of work would be an important, although complex, task. To ease the transition to work, structural mismatch needs to be reduced first of all through employment policies targeting those groups of individuals with lower employability levels and/or those more likely to become long-term unemployed.
The Croatian Employment Service (CES) systematically helps inactive persons who seek employment. The aid consists of various forms of information and counselling, while workshops for developing skills for an active job search are organised. The workshops are organised through three modules: how to seek a job actively; how to introduce oneself to the employer; and self-assessment methods.
In the activation of unemployed persons, CES also implements Professional Plans. They contain a list of jobs and professions that the jobless persons can perform according to his or her defined professional, working and personal abilities and the labourmarket requirements. Furthermore, Professional Plans have a list of activities the unemployed person will carry out in order to find employment and the deadlines for their execution, for example, applying to employers and answering job adverts. The Professional Plans specify the time for contact between the unemployed person and the employment counsellor. The Professional Plans are signed within 60 days of the registration and both the employment counsellors and the unemployed persons sign them.
Employment-policy programmes have been continuously implemented in accordance with the Guidelines for Development and Implementation of the Active Labour Market Policy. With the goal of including long-term unemployed persons and to prevent entry into long-term unemployment, CES financed the training of unemployed people to address labour-market needs. Long-term unemployed and special groups of the unemployed had the advantage of inclusion in the training. For example, in 2011, a total of 13,788 persons were included in training, 202% more than in 2010. The ratio of the number of participants in the training programmes in the course of the year and the number of long-term unemployed has constantly increased since 2008, but it is still a relatively small number (less than 10% of long-term unemployed).
In May 2012, the Act on Employment Incentives was adopted, introducing benefits for the employment of long-term unemployed. The employer who hires a very longterm unemployed person (over two years) is exempt from the payment of contributions on his/her salary for a duration of up to two years. Henceforth, the Ministry of Labour and the Pension System, in collaboration with its implementing body CES, has continuously promoted fiscal relief for hiring long-term unemployed persons.
In the period 2013-2015, the 'Training for Employment for Workers Facing Unemployment and Long-term Unemployed Persons' project was realised. It included surplus employees, workers facing the threat of lay-off and long-term unemployed persons. The main objective of the project was to provide support in the implementation of active employment measures. The Project's purpose is to increase the level of employability of surplus workers and long-term unemployed persons through vocational training for employment.
Furthermore, for some groups of unemployed persons, like the long-term unemployed, people over 50 years of age and people with disabilities, there are employment incentives. Such incentives are approved to employers if the employment with an incentive leads to a net increase in the number of employed compared to the average number of employed within the last 12 months, as well as to employers who have not had an employment increase in the last 12 months, but who justly fill vacant positions by employing with an incentive. The amount of incentive depends on the type of employers and the educational attainment of the unemployed person (primary, secondary or tertiary level). SME employers receive in total HRK 19,450 (e2,593) for employing an older person with primary education, HRK 29,480 (e3,931) for those with secondary education, and HRK 36,283 (e4,838) for those with tertiary educational attainment. Bigger employers receive HRK 11,670 (e1,556), HRK 17,688 (e2,358) and HRK 21,769 (e2,903) respectively. There are no conditions regarding the length of registration in the Croatian Employment service register. The contract should last for a minimum of 12 months, while there are no conditions regarding the type of contract (indefinite or fixed-term contracts are both eligible). The job has to be full-time (40 hours per week).
Next to the incentives for and training of long-term unemployed persons, public work is community service that takes place for a relatively short period and offers cofinancing and financing of the employment of the long-term unemployed persons. Public works have a two-fold role in the range of employment-policy measures. Their first specific feature is that they are used to increase and maintain a certain level of social inclusion of those unemployed who have not worked for a prolonged period. Their second task is to 'create' jobs that probably would not have been created without this measure, i.e., funds allocated for its implementation, in order to open up space for the activation of these persons, whereby these jobs must necessarily be for the general well-being.
In recent years, activities have increased, and a respectable part of them has been directed to the social-welfare beneficiaries who have participated in public works. Public works in Croatia are not used for building large infrastructure facilities, but for some socially useful community activities, such as care for the elderly, reconstruction, cleaning and preparation of land for cultivation and so on. The majority of these activities are not profitable and therefore do not raise interest from profit-oriented companies; the bodies mostly competent for them are state and local authorities. The main benefit of public works is not immediate economic gain, but the greater encouragement given to unemployed persons, for whom there is very low demand in the labour market to join the labour force.
Since 1999, there has been a subsidy for self-employment for long-term unemployed persons called 'Your initiativeyour working place'. The start-up subsidy is intended for people registered with the CES who show interest in self-employment, and they are provided with professional help from the Employment Service, where they can also receive all relevant information on activities connected to the realisation of the entrepreneurial idea. The measure consists of favourable loans in the amount of HRK 25,000 (e3,400) (which increased to HRK 55,000 or e7,333 in 2018). It in itself is no warranty for business success, but in the words of the participants, a nudge that enables the initial acceleration of the entrepreneurial idea.
Succinctly, there are relatively many activities for the preparation and employment of long-term unemployed persons in Croatia. It is praiseworthy that in the last few years there were evaluations of realised activities, whose results are presented in the following text.

Evaluations of realised employment activities for long-term unemployed
A study by Dorenbos et al. (2002) on the public-work programme concludes that participation in the programme does not directly increase the participants' probability of finding a job, but at least it does not reduce it. Almost all of the respondents (96.9%) said that the public works helped them in some way. The majority of the participants stated that the programme aided them in ameliorating their financial situation. Unfortunately, all other impacts, like strengthening confidence and acquiring knowledge and expertise and assistance in establishing social contacts, were less reported.
Although expenditures for ALMP in Croatia have increased, there are serious doubts about the efficiency of these efforts (Babi c, 2003), where the problems are related to their short duration and relatively low consistency. Various programmes started and have yielded better than expected results, but they were interrupted, mostly due to a lack of money. In addition, greater attention and financial support was given to the co-financing of employment, and not to the improvement of knowledge and expertise of job seekers.
In evaluation of Active Labour Market policies in 2009 and 2010 in Croatia, Matkovi c, Babi c and Vuga (2012) observed effects of participation in measures realised by the Croatian Employment Service. Effectiveness was analytically evaluated by the application of matching techniques, as they compared outcomes of participants with a control group comprised of unemployed persons with similar observable characteristics who did not participate in measures. They evaluated five measures for which appropriate matching could be enacted, among others, longterm unemployed and public works. The results of this quasi-experimental evaluation approach do not indicate that participation bears a particularly strong effect with respect to the observed outcome. Participants in subsidy programmes were less likely to be in unemployment than controls for the first two years after subsidies ceased, but the advantage of participants was declining over time. The probability of being unemployed for participants of training programmes was about the same (or minimally lower) than for comparable non-participants. Participation in public-work programmes was estimated to have increased the mid-term unemployment risk for participants, but this has to be understood through both programme effect and selection of the most vulnerable unemployed in public works. Additionally, the higher probability of being in unemployment among public work participants when compared to controls can be interpreted in terms of activation: if they did not participate in public work, more of them might have left unemployment for inactivity (which likely happened to many statistical twins with whom they were matched).
Similar results were achieved with the last realised evaluation of employment programmes (Ipsos Puls and Croatian Employment Service, 2016). The impact of employment incentives for long-term unemployed people through the evaluated period 2010-2013 was quite positive. While 71% of participants were employed in the period of six months after the exiting the measures (70% after 12 months and 68% after 18 months), in the five-times-bigger control group (with almost identical qualification, age and gender structure) the mentioned shares were 29%, 33% and 31%, respectively. Thus, the differences were 42% after six months, 37% after 12 months and after 18 months.
There are no separated evaluation results for long-term unemployed for training and incentives for self-employment. Generally, the latter measure could be considered as efficient if observed from the perspective of the number of subjects who continued their business even after it expired. Evaluation results indicate that three quarters of business entities launched (users are not distinguished if they are inactive or longterm unemployed) were still active and owned by measure participants, of whom less than half consider their current business situation as good, very good or excellent (Ipsos Puls and Croatian Employment Service, 2016). The available data indicates purposefulness of the funds invested in the start-up incentive programmes, while two additional data point to a wider social function of the incentive as a measure for employment policy. First, one third of still active business entities employ at least one person, apart from the founder. Second, the collected data point to an improvement of the socio-economic situation of the households in which the measure participants live, in comparison with the period before using the incentive.
One can presuppose that mostly the long-term unemployed participate in public works . Participants in public works indicated that their participation is important to them because it generates a feeling that they are not forgotten and alone, that society has not abandoned them, but rather that it cares about them, which had a positive influence on their self-confidence. In terms of social inclusion of the unemployed, public works have accomplished significant success. The socialisation benefit from public works is a specific emotional fulfilment. Some participants underlined precisely the 'positive emotional effect' as the main positive outcome of their participation in public works. It is also linked to the sense of pride in participating in the works they did. Nearly 90% of them evaluate the work they did as useful. Only a minority of participants (around 1/6 or less) in the measure had a feeling of social stigmatisation while doing public works. Besides the activation of the role, this measure also has a very important financial aspect for participants. Again, their employability does not increase very much, but some established a valuable social network that is very important in the process of job finding. More evidence that this measure was relatively well targeted and that it accomplishes the goal of activation is the fact that approximately half the participants stated they were almost forced to participate in this measure because of the 'threat' of being deleted from the CES records if they refused to participate in the measure. In the end, it was a quite positive experience for them.
Croatia is fully aware of the problems related to long-term unemployment and in its most important development documents, like Strategy for Combating Poverty and Social Exclusion in the Republic of Croatia (2014-2020), being long-term unemployed is an important cause of exclusion based on economic status. Furthermore, adequate attention is dedicated to improving the situation in the labour market: In order to additionally emphasize the support to employment of especially vulnerable groups ( … the long-term unemployed), it has been envisaged to further implement the Employment Promotion Act, which ensures the possibility of using reliefs for the employment of unemployed persons. (p. 22)

Conclusion and recommendation
There is evidently a strong inter-relation among education, long-term unemployment, employability and poverty. The problems that many of the long-term unemployed have are mostly multidimensional, and frequently include a low level of educational attainments and motivation. Croatia already has active and passive measures of assistance to the unemployed. Unemployment and low activity rates in Croatia are mainly the consequence of insufficient demand for labour due to low job creation in the private sector and the mismatch in labour supply and demand. Relatively high real wages, institutional rigidities and wide-spread skills mismatches appear to be major impediments for a more dynamic labour-market performance. A high level of longterm unemployment reduces the employability of the work force due to a lack of skills, barriers to employment or a combination of some other factors. Long unemployment spells, particularly for the young, may have long-lasting effects in terms of future employment and wages. This may lead to withdrawal from the labour market, which in turn lowers labour-force participation.
The evaluation results indicate that participation in various forms of employment programmes in Croatia may be considered as a successful way of overcoming financial problems, but also a possibility to achieve new knowledge, skills and attitudes needed for the labour market. The increased scope and expenditure of employment programmes was also followed by additional orientation towards the long-term unemployed. For the majority of the long-term unemployed, it is very difficult to emerge from this unfavourable status and poverty, so they are forced to stay almost permanently as benefit recipients in the social-welfare system.
We could only reiterate and add to some of the most important recommendations from the evaluation by Ipsos Puls and CES (2016). There is an obvious need to try to secure financial sustainability of the measure according to previous positive experience. It is important to stabilise the design and the conditions of the majority of measures at least in the medium-term, if not in the long-term, period. Responsible CES should not change the contents and the criteria of measures (for example, needed and additional documentation) during one cycle of implementation. Adequate attention should be directed to improve the integrity and unambiguity of the provided information to avoid the risk of incorrect, incomplete and false information. It is essential that complete information is constantly available on the CES internet page, but also that all included staff have printed brochures available, which can be given to clients during the meeting. It will be good to consider the possibility of simplifying the overall procedure for submitting documentation. Finally, future evaluation of the realised employment measures will also be more than useful.