Poverty Alleviation: The Case of Croatia Poverty Alleviation: The Case of Croatia

In economic literature, poverty is usually defined as the inability to satisfy basic material needs, particularly adequate nutrition, but also ensuring satisfactory housing, means of production and other assets. The goal of the chapter is to present the current situation of poverty in Croatia with the intention to identify measures for the enhancement of pov erty alleviation. From the multidimensional perspective, what matters is a focus on the opportunities—such as a possibility for education and employment, adequate contact to markets and so on—that are available to people. If a person does not possess sufficient capabilities or endowments, he or she has a limited possibility to escape from the unfa vourable situation. Poverty in Croatia is stagnant—those who become poor need a long period to escape from poverty. The inactive and persons unemployed are the dominant groups of the poor in Croatia. The current social protection system is a mix of old and new programmes and it has been adjusted in response to altering social needs and oppor -tunities. Successful poverty reduction is associated with the improvement of the labour market, a consistent increase in decentralisation of financial sources and services, the reduction of corruption, carefully reallocating expenditures and improving coordination among existing social programmes.


Introduction
Although, at first glance, it looks easy and unequivocal to define poverty, its inherent complexity makes this not at all easy. In economic literature, poverty is usually defined as the inability to satisfy basic material needs, particularly food, but also ensuring satisfactory housing, means of production and other assets. In the daily lives of the very poor, poverty becomes a network of detriments, each one aggravating the other. The outcome is generation after generation of persons who lack access to decent nutrition, adequate education, health care, satisfactory housing and proper sanitation. They are the most vulnerable to natural disasters and systems of political and economic oppression.
Very often, poor people are incapable of improving their condition and circumstances. Such conditions frequently ruin social relationships cause a dysfunctional family and generate low self-esteem and a lack of self-reliance. They are usually excluded and marginalised from active participation in society and events that are the usual norm for other citizens. Urban residents frequently tend to equate poverty with the impossibility to take advantage of the cultural or social activities that they previously enjoyed. Because of their poverty, such persons experience multiple disadvantages and unsatisfactory characteristics that are transmitted to the next generation. Regardless of the exiting economic situation, child poverty is one of the most significant topics. All governments need to take seriously because children are society's future and will shape it very soon. Poverty in itself breeds unrest and discomfort and creates a lack of hope. This chapter examines the situation of poverty in Croatia and measures for its reduction. The goal is to present the current situation with the intention to identify measures for the enhancement of poverty alleviation in Croatia. After these introductory notes, Section 2 contains the theoretical framework and the concept of multidimensional poverty. Section 3 deals with the measurement approaches to poverty, with a particular attention towards possible statistical problems and errors. Section 4 presents the issue of poverty statistics in Croatia, while Section 5 deals with characteristics of poverty in Croatia. Anti-poverty policies in Croatia are presented in Section 6, whereas Section 7 contains conclusions and recommendations for the improvement of policies for poverty alleviation.

The concept of multidimensional poverty
To comprehend poverty, it is vital to observe the social, economic and political context, including the state institutions, the government at various levels, markets (particularly labour market), communities and households. The persistence of poverty is related to its interlinking multidimensionality: it is a complex, dynamic, institutionally entrenched and a gender-and locationrelated phenomenon. The forms and characteristics of poverty differ by countries, social groups and locations. Poverty varies with respect to age groups, location (rural versus urban), ethnicity, gender and income source. Regularly, children and females in households often suffer more than males. Poverty outcomes are the consequences of complex relations between various stakeholders, institutions and policies in the social, political and the economic spheres [1]. Poverty depends not only on the level and structure of the national income but also more fundamentally on how redistributive policies and changes in the national income cause the variations in households' income and consumption [2].
Poverty can have grave consequences on the health of affected people. It can endanger their psychological stability and cause the feeling of fear, anxiety, crisis, anxiousness, stress and frustration. Poor persons often have significantly worse health conditions; they suffer from continuous feelings of exhaustion and endure more physical pain in comparison to the total population. These factors also negatively influence their position on the labour market and can lead them to various forms of addiction like heavy drinking of alcohol for instance [3]. Such addictions can further ruin their employability and captivate them into poverty. Many poor people also underline the psychological feature of poverty because they are aware of their feebleness, weakness and exposure to possible exploitation [4].
Poverty usually causes vulnerability and that enables many representatives of state power and/or other citizens to humiliate the poor and treat them as less than human. Poverty generates a feeling of loneliness, and due to the restricted possibilities of maintaining social links with friends, neighbours and relatives, it thus forms a sense of social exclusion [2]. For deprived citizens, it is essential to maintain existing social and cultural norms, particularly because for a majority of them, social solidarity is one of the most crucial positive available factors. Even though poor people regularly underline their social, economic and financial problems, usually, they complain more of social exclusion or limited capacities to break out of the impasse of destitution. Poverty almost never results from the lack of one factor only but from many related factors that accumulate in poor people's experiences and their impression of destitute. If citizens or households fall into poverty, they face a very hard time finding their way out of it. Very often, their social links and relations with friends, family and neighbours are limited or do not exist [5]. Furthermore, due to their non-existing or weak social binding, their possibilities to work in the informal sector are seriously restricted. Thus, for them, it is difficult to find employment in the formal and in the informal economy.
From the multidimensional perspective, what matters is an orientation to the opportunities-like a possibility for education, employment, access to markets and so on-that are normally available to other better-off people. If a person does not possess enough and adequate endowment or capabilities, such as a basic education, or does not have the possibility to obtain them, he or she will have a weak or reduced opportunity to escape from poverty. However, in poverty alleviation, the important issue is how to measure it.

The measurement approaches to multidimensional poverty
It is very essential to derive how to measure poverty because it defines (or confuses) attempts to design adequate and efficient policy measures for aiding the poor and for redistribution of income. The most universally used way to measure poverty is based on incomes or consumption levels. The main indicator, absolute poverty, measures the proportion of a population surviving on less than a specific amount of income per day. This specific amount is the poverty line-probably the most contentious topic in the discussion. There are two types of poverty lines: absolute and relative. Absolute lines are related to an absolute minimum living standard and are usually based on a fixed basket of nutritional substances (considered to epitomise the minimum nutritional consumption required for good health) plus an increase for other outlays (such as lodging and clothing) [6]. Absolute lines are usually different in various countries, depending on the structure of the consumption basket. While there is some arbitrariness in determining what is appropriate and adequate, the concept of a poverty line still is a valuable benchmark [7].
Adequate attention and different methods are used to assess the diet and non-diet parts of the poverty line. Although alimentary necessities are obtained on a scientific basis according to physical needs, the non-diet part of the poverty threshold is not assessed in such way. The procedure recognises households whose spending on food is the same as the cost of the diet specified as the poverty line. After that, one should assess how much households really spend on non-food items that are observed. Thus, the food part is evaluated according to needs, while the non-food part mirrors the poverty of the poor with no guarantee that all basic nonfood needs are, or could be, successfully satisfied [4]. This is a serious disadvantage that could neglect the importance of other basic needs such as housing, education, health, information and participation in social and political life.
Regardless of absolute needs, people may deem themselves poor when their living standards are considerably below those of other friends or neighbours. This type of poverty is presented by relative poverty lines that express poverty relative to living standards on the national level. Relative poverty lines are convenient for international comparisons of the features of the worst-off persons in observed countries. Relative poverty lines are usually expressed as a fixed percentage of median or mean equivalent household income and most often uses 50 or 60% of median income as a base for international comparison (the other two are one-third and two-thirds of median income or consumption).
Data on consumption and income are acquired through sample surveys, during which households provide answers to detailed questions on their sources of income and spending habits. Such surveys are realised almost recurrently in most countries. Mentioned sample survey data collection methods are to a greater extent added by participatory methods. The poor citizens should say what their most important needs are and how their life in poor conditions is. Interestingly, new surveys show a high degree of concordance between poverty lines based on objective and subjective assessments of needs.
It is important to have in mind the conventions used to construct these figures and their limits, primarily related to the definition of income [8]. First, very often in household surveys, only cash receipts are included and/or they dominate, while in-kind receipts-payments to individuals in commodities or services-are neglected. One major form of in-kind income is the value of time adults devote to their households. Second, these figures often ignore taxes and in-kind transfers from the government. Third, income is measured over some period (usually annually), although lifetime income should be measured and included. Lifetime income ("permanent income") would be ideal, but it is hard to estimate. Using the measure of welfare for a longer period (mentioned lifetime income) than annual income would probably lower the share of households in poverty by several percentage points. Fourth, consumption data may provide a better assessment of well-being; the official data on poverty are usually based on income. In today's world, many people work on the fixed terms and/or on various projects for a definite time. When a person or household's income pattern is not stable, current income is an incomplete approximation of an individual's standard of living. People tend to soften consumption over time, spending in advance the money that they will obtain in future [9].
Fifth, there are problems in defining the unit of observation because most people live with others, so there is a question of whether income distribution should be measured over households or individuals? If economies (primarily lower costs per capita) are achieved by living together, should they be taken into account in computing an individual's income? What is an appropriate equivalence scale: how to calculate the cost of second and/or third household's member, how to reckon the child and so on? Furthermore, the equivalence scale does not capture the true economies of scale for this household. Sixth, there is the apparent increase in under-reporting of income; this is what significantly underestimates the incomes of those at the lower end and overestimates of the extent of poverty. Seventh, there is a measurement error, for example, respondents may forget certain expenditures or income component or include ones that should be excluded, and errors may occur in data entry.
The main indicator is the at-risk-of-poverty rate, which shows a share of persons with the equivalised total available income below the at-risk-of-poverty threshold (the poverty line). The at-risk-of-poverty rate does not represent the real number of poor people but rather how many of them have an income (after social transfer) below the at-risk-of-poverty threshold. The at-risk-of-poverty threshold is mostly determined as 60% of the median value of the equivalised disposable income of all households. This indicator does not measure wealth or poverty but low income in comparison to other citizens in the observed country, which does not necessarily indicate a low standard of living.
It is useful to know how far the poverty population lies below the poverty line. This measure is the poverty gap that shows how much income would have to be given to the poverty population to bring every household's income to the poverty line (supposing the transfers had no effects on the recipients' work effort). Regarding the duration of poverty, standard measures of poverty, however, are only "snapshots". They are based on people's current income status and do not account for their history of poverty. However, most people dip into poverty for a longer period, which aggravates the problem of their material and social position [10].
Poverty measurement and researchers have increasingly oriented to measures of poverty intensity which do account jointly for the number of poor persons, depth of poverty and inequality among the poor. For practical purposes, the percentage of change in poverty intensity can be approximated as the sum of the percentage changes of the poverty rate and the average poverty gap ratio. As the parameter rises, more and more weight is attached to the lowest income levels or, in other words, to the well-being of the poorest individuals.

Poverty statistics in Croatia
Knowledge about the scope, problems and incidence of poverty in Croatia was limited until the research on poverty and vulnerability was realised by the government in collaboration with the World Bank [2]. The research was based on the first post-Homeland war (1992-1995) household expenditure survey in Croatia realised by the Croatian Bureau of Statistics. According to the obtained results, poverty in Croatia is relatively low, that is, lower than in most transition states in the region (except for Slovenia). Only 4% of the population lived on less than US$ 4.30 a day at the purchasing power parity (internationally comparable standard used across transition economies) and about 10% lived on less than US$5.30 a day, which the study recommended as an appropriate absolute poverty line for Croatia [11].
In the past, in Croatia, the income poverty indicators were calculated by using data from the total net income of a household and all household members obtained through the Household Budget Survey (HBS). The survey was realised on the random sample of private households in such a way that the sample was separately determined for each year. In 2007, the at-risk-ofpoverty rate for the whole population was 17.4% (with income in kind) and 19.0%, (without income in kind), and for employees, it was significantly lower, 4.1 and 4.2%. However, the situation was quite opposite for self-employed persons, whose at-risk-of-poverty rates in 2007 were 17.9 and 24.1%. At-risk-of-poverty rates were much higher for the unemployed (42.8 and 43.7%, respectively) and other economically inactive persons (28.2 and 30.7%, respectively).
In the period from 2005 to 2008 before the economic crisis, there was a stagnant trend in the total at-risk-of-poverty rate (around 17%). However, at the same time, there was an obvious trend of growth of the mentioned rate among single households (39.8 in 2008 in comparison with 31.1% in 2005) and particularly among older single households (47.8 in 2008 in comparison with 35.2% in 2005). For pensioners, the at-risk-of-poverty rate was around 30% higher than the national average, but for elderly people who do not receive pension benefits the at-risk-of-poverty rate was 170% higher than the national average. Croatia has a public pension insurance (pay-as-you-go) scheme of Bismarck type so if the person did not pay pension contribution and does not have a family (survival) pension, he or she is not entitled to the pension rights. It is estimated that in Croatia, there are roughly 13% of people above 64 years of age who do not receive any pension benefits. These are mostly women that account for more than 95% of this cohort [12]. Elderly people without pension who live in single households are at the most adverse economic position.
The at-risk-of-poverty rate started to grow in the beginning of the economic recession (18 in 2009 in comparison to 17.4% in 2008). The Croatian at-risk-of-poverty rate in 2010 was 20.6%, which is considerably higher than the EU-27 average (16.4%). The rate in Croatia increased also in 2011 to 21.1%. In 2010 and 2011, there were no significant changes in the poverty profile. In addition to the oldest age group, children at the age of up to 15 years had the at-riskof-poverty rate above the national average. While in 2009, the relative poverty rates among the children and youth grew, in 2010, the at-risk-of-poverty rate for children in the group 0-17 (20.5%) was at the national average level. On the other side, at-risk-of-poverty rates for youth in the age 18-24 (21.6%) and persons over 64 years (28.1%) were above average.
In the period from 2006 to 2009, the difference between the at-risk-of-poverty rates for males and females increased. In 2010, the difference was 1.6 percentage points, which is significantly less than in 2009, when it was 3.8 percentage points. The greatest difference between the at-risk-of-poverty rates for males and females is in the eldest age group (65+). The mentioned difference continued to grow from 2006 to 2009 (from 4.4 percentage points to 12.8 percentage points, respectively). In 2010, the at-risk-of-poverty rate for elderly women was by 8 percentage points higher than the rate for men.
In 2009, in comparison with 2008, the at-risk-of-poverty rates augmented expressively among the jobless persons, single households with older citizens (as mentioned mostly women) and families with three or more children. In 2009, the highest growth in at-risk-of-poverty rates in comparison with 2008 was recorded for households with three or more children (by 12.8 percentage points). During the observed period, the at-risk-of-poverty rates for children (0-15) increased by 3.8 percentage points, for the unemployed by 4.8 percentage points and for subtenants by 5.4 percentage points. The at-risk-of-poverty rates for households with two adult persons and one or two children, as well as of other households with dependent children, increased slightly in 2009 but were still below the national average. The mentioned rates for pensioners increased moderately in 2009 when compared to 2008 (less than 1 percentage point) and was 35% higher than the national average [13].
According to the data for 2010, single households, single-parent families and families with two adults and three and more children had the at-risk-of-poverty rate that was above average. Just like in previous period, single persons recorded the greatest risk of poverty. Their at-riskof-poverty rate was 44.5%. The situation was particularly unfavourable for single households with elderly household members, where every other person was at risk of poverty. Single parents with one or more dependent children had the at-risk-of-poverty rate around 68% (or 14 percentile points) higher than the national average). Households with two adults and three or more children also had a very high at-risk-of-poverty rate (33.1%), which was the result of an adverse impact of the economic crisis on income and the situation on the labour market characterised by the high unemployment rate. At-risk-of-poverty rates for households with two adult persons and one or two children and other households with dependent children were mostly under or at the level of the national average. In 2010, the at-risk-of-poverty rate for children in the ages from 0 to 17 (20.5%) years was at the national average level, while at-risk-of-poverty rates for youth in the age group 18-24 (21.6%) and persons over 64 years of age (28.1%) were higher than average. It should be mentioned that the at-risk-of-poverty rates of the unemployed (44.7%), and especially unemployed men (51.4%), were particularly high [14]. In 2011, the at-risk-of-poverty-rate increased to 21.1%. Higher at-poverty-risk rates in 2011 were for the unemployed (42.5%), followed by the economically inactive (33%) and the retired (22.2%). The at-risk-of-poverty rate for 2012 was 20.5%, while exposure to poverty was higher in rural areas (38.1%).
For 2014, mentioned rates were calculated using the Survey on Income and Living Conditions (SILC). The indicators are based on a concept of relative poverty, which takes into consideration the disposable household income, the number of household members (household size) and the income distribution within the population. According to the survey data, the at-risk-of-poverty rate 1 in 2014 was 19.4%. In 2014, the lowest at-risk-of-poverty rate by the most frequent activity status was obtained for employees, 4.7%. It amounted to 5.8% for male employees and 3.5% for female employees. 1 All other data on poverty trends and comparison in Croatia are available on http://www.dzs.hr/default_e.htm, http:// povertydata.worldbank.org/poverty/country/HRV; http://ec.europa.eu/eurostat/statistics-explained/index.php/People_at_risk_of_poverty_or_social_exclusion. Analysis and official estimations show that the profile of poverty has changed under the influence of the crises and that the risk of absolute poverty has increased for the children and the youth population under the age of 30 [15]. The highest at-risk-of-poverty rate by the most frequent activity status was for unemployed persons, and it amounted to 43.2%. It was 49.5% for the unemployed men and 37.4% for the unemployed women. For the self-employed, the at-risk-of-poverty rate was 14.9% (male 16.5% and female 11.6%) [16]. Regardless of the impossibility to compare directly the data from mentioned surveys due to the different methodology, the results from 2007 and 2014 show that the risk of poverty in Croatia is directly related to exclusion from the world of work. In 2014, the highest atrisk-of-poverty rate was indicated for unemployed persons and amounted to 43.2%, while for employees it was 4.7%.
For Croatia, different poverty rates were calculated by EUROMOD [17] than by the official SILC data by the Croatian Bureau of Statistic. In 2011, EUROMOD recorded a larger poverty rate for the group "65+ years", by 2%, than the external source. This is probably due to use of pensioner contributions in EUROMOD, whereas SILC data might not assume that form of contributions. For all other age groups, poverty rates are lower in EUROMOD, and the discrepancy mostly lies in the difference between EUROMOD and SILC regarding the coverage of social assistance benefits, which significantly affects the bottom income decile. The most remarkable difference in 2011 is for the group "0-15 years", where EUROMOD indicates 13% lower poverty rate than the external source. This may be caused by the fact that social welfare benefits (particularly in that time applied subsistence benefit) had a relatively more pronounced effect on families with children.
According to the small area estimation model and consumption approach, the estimated atrisk-of-poverty rate in Croatia amounted to 17.1% in 2011, while the at-risk-of-poverty threshold was HRK 23,919 (€ 3190) for a one-person household. It is assessed that 17.1% of persons had equivalised consumption below the threshold in 2011. At-risk-of-poverty rate estimates at the NUTS 2 level in Continental Croatia amounted to 19.4 and to 12.6% in Adriatic Croatia. Looking at the estimates at the level of counties, they range from 5.9 to 34.3% [18].
Of the population in the EU-28, 24.4% were at risk of poverty or social exclusion. More than one-third of the population was at risk of poverty in the three EU member states: Romania (40.2%), Bulgaria (40.1%) and Greece (36.0%), while Croatia with 29.9% is sixth in Europe and significantly above the at-risk-of poverty rate before social transfers EU-28 (26.1%). The at-risk-of poverty rate after social transfers for Croatia in 2014 was 19.4%, which is 2.2 percentage points above the at-risk-of poverty rate after social transfers for EU-28 (17.2%) [19].
The material deprivation rate presents the share of persons who live in households thatbecause of insufficient financial resources-cannot afford at least three of nine deprivation items. For Croatia, the rate of material deprivation in 2014 was 33.8% [16].
Succinctly, in Croatia areas most affected by poverty often have high unemployment and inactivity rates of their population. The education attainments of local population are mostly low; while poor areas can provide only lower incomes, lower living standards and poorer dwelling conditions contribute to the significant depopulation to other parts of the country or abroad. In the following section, attention is oriented towards more detailed characteristics of poor citizens in Croatia.

Characteristics of poverty in Croatia
There are several prevailing groups among the poor in Croatia, principally the unemployed and inactive persons. In that way, in-work poverty is not a particular problem in Croatia because employment (and self-employment according to the data from 2014) is a relatively secure protection from poverty. The situation did not change during the economic crisis, and the main traits of poverty have stayed the same. In Croatia, there has been a division of the society: relatively securely employed (insiders) and the unemployed persons (outsiders) of whom a substantial share is the long-term unemployed [20]. A significant part of long-term unemployed persons is with low employability. Thus, they have small chances of finding a job and are trapped to live in the poverty.
According to the presence of the risk of in-work poverty, Croatia is below the EU-28 average, regardless of the type of labour contract. Thus, 4% of employees with contracts for an indefinite time and 9% of employees with fixed-term contracts were at risk of poverty. In the last few years, about 6% of employees were living in poverty. However, there are various groups of employees that are probably more exposed to in-work poverty, primarily people without education and qualification and people with lower educational attainments, employees working only on seasonal work, self-employed employers, involuntary part-time employed and temporary agency workers.
Although the unemployed and inactive represent a relatively small share of the poor population in the Croatia, they are seriously exposed to danger from poverty, while working status-particularly permanent and full-time employment-is a reliable shelter against poverty. Almost three-fourths of the poor citizens live in families whose main member has only a primary education or even incomplete primary education. These people have reduced possibilities of finding work if they are jobless and/or of having higher earnings if they work. The risk of poverty is overall high when low educational attainment is linked with inactivity or unemployment. Citizens that live in households where the household head is an unemployed or inactive are around three times more likely to be trapped in poverty than the total population. The probability of poverty problems is even greater for the unemployed persons with small children.
Poverty in Croatia is stagnant-those who are poor have limited possibilities and need a great deal of time to exit from poverty. There is no recent data on poverty duration, but one older study by Šućur [21] provides some insights into the social, economic and demographic characteristics of unemployed welfare recipients in Croatia. The first goal of the analysis was to make a distinction between subgroups of users. The author takes into the account their working participation or inactivity and how long they had been receiving welfare benefits. The second goal was to find out which factors were predominant in determining the duration of the period of obtaining benefits. Šućur makes a distinction between beneficiary of social aid (the individual person who is the applicant for assistance) and user (includes formal applicant and members of his or her family that used the received benefit). In the analysis, there was a sample of 501 social beneficiaries. According to their employment status, welfare recipients consist of two dominant groups: the unemployed persons and the disabled. In comparison to other transitional countries, Croatia has a higher share of unemployed persons in all welfare recipients. Around 70% of welfare recipients have no qualification or have only primary education attainment. Probably, many do not have basic skills of literacy and numeracy, while some have had no formal education at all.
The length of receiving the welfare assistance varies in statistical terms, depending on the level of education attainment, beneficiary's age and region where the person lives. The less educated and the older recipients are more likely to remain longer as social welfare benefit recipients. The average length of benefit-receiving is quite long (almost 5 years). The average length of welfare scheme usage is 7.3 years for users without education and around 3.5 years for persons with tertiary education attainment. Regarding the age, it is 2.5 years for the younger welfare recipients (between the age of 19 and 29) and 7.6 years for those older than 60 years of age. According to regression analysis, one can predict that welfare assistance is received for a longer period by applicants of senior age, of lower level of education, those who are not married and living in incomplete families and who receive other benefits available under the social welfare system.
Those who do not have remunerative employment due to low levels of education attainment are also likely to transfer their limited opportunities for their children. Available data show that the access of children from poor families to upper secondary and tertiary education level is very limited. The children of the poor are more likely to drop out of the schooling system early, and differences in access to tertiary education are obvious. The poor have access to university only through relatively few highly competitive scholarships. The lack of access to levels of education that are highly valued on the market tends to prolong existing inequalities in earning prospects between the poor and non-poor and to create the probability for the intergenerational transfer of poverty [11].
As a conclusion, we can reiterate that a considerable number of Croatian citizens, particularly the unemployed and inactive persons, suffer from a poor financial situation, adverse social and living conditions and absence of suitable access to public goods and services. Due to the longlasting economic recession from 2008 to 2014, increased unemployment and budgetary problems, there was further deterioration of the social position of significant number of citizens. In order to alleviate the consequences of the crisis, the government is proposing and implementing various measures regarding economic, fiscal and particularly social welfare policy. The current social welfare protection system in Croatia is a mix of old and new programmes. It has been frequently changed due to the altering social opportunities and conditions with the intention to ensure more efficiently the provision of social transfers and needed services. Croatian social welfare system consists of three basic components: cash aids, benefits and services in kind and a variety of foster care and residential programmes. According to the statistics maintained by the Ministry of Demography, Family, Youth and Social Policy, there are various financial transfer benefits as well as numerous types of in-kind assistance provided. Furthermore, most local government institutions and bodies as well as many NGOs provided financial aid and services for various groups of citizens like poor, disabled and infirm persons, the elderly and so on. Money aids financed on a much larger scale than in-kind benefits and services, although there are a relatively large number of recipients of the latter. In fact, this is deceptive because the main in-kind programme is the assistance in covering costs for public and communal services. The evaluation [22,20] has shown that financial outlays inside the large programmes (pensions) are less directed towards helping the poor, while those programmes on which fewer expenditures are spent (welfare benefits) are more strongly oriented to the impoverished.

Anti-poverty policies in Croatia
Total social costs in Croatia at the beginning of millennium were around 21% and were considerably lower than in EU-27. Nevertheless, said outlays in Croatia were higher than in other transition countries (with the exception of Hungary and Slovenia) and similar to the ones in the Czech Republic and Poland. In Croatia, there was an obvious trend of a gradual reduction of total social outlays after 2001 (to 19% of GDP in 2008) due to a relatively high economic growth until 2009 and a decrease in pension expenditures. Since 2009, because of economic crisis and deterioration of the situation on the labour market, there was an opposite trend of growth in the absolute amount of the mentioned expenditures and increase of their share in GDP (to 21.5% in 2010). Comparing data on the share of social protection expenditures in the national GDP with the share of such expenditures in the GDP of the other EU member countries, Croatia ranks 18th, lagging by 8.6 percentage points behind the average of the EU-27.
Previously, the most important social welfare form of aid was subsistence benefit (pomoć za uzdržavanje, stalna pomoć, minimalna zajamčena naknada). It was a means-tested benefit, intended for households whose income was below the "means of subsistence". The means of subsistence represented the amount of money necessary to satisfy the basic needs of the observed household. They were defined as a sum of individual contributions depending on the traits of the household members. When calculating means of subsistence, children, the elderly and disabled people were automatically accounted; working-able persons in the age between 18 and 65 years were not included into account unless they did satisfy the conditions for being treated as unemployed. According to Urban and Bezeredi [17], the poorest decile seems to have benefited the most from changes in means-tested benefits and in particular from the increase in the subsistence benefit levels. On the other hand, the simultaneous introduction of a cap in this benefit, set at the level of gross minimum wage, had a negative impact on the second income decile.
Subsistence benefit was replaced with a single basic Minimum Income scheme which is termed Guaranteed Minimum Benefit (Zajamčena minimalna naknada-ZMN) and introduced in a law which came into effect on 1 January 2014, merging the basic social assistance scheme with two smaller benefits. The scheme has been amended in September 2015 with some changes taking effect immediately, and the others set were implemented on 1 June 2016. The ZMN is a national scheme, means-tested and asset-tested. Croatia spends about 0.4% of GDP on this scheme annually. Anyone may be entitled to ZMN providing his or her income is not sufficient to guarantee person's subsistence. A number of sources of income, including a range of other benefits, are not taken into account when considering eligibility for ZMN. For those capable of work, receipt of ZMN is conditional on being available for and actively seeking work. Although subject to review and adjusted in the light of changing circumstances, there are now no time limits of the receipt of ZMN, even for those who are capable of work, a change introduced in September 2015. Changes are following: (a) the contribution of children in the calculation of the "means of subsistence" is significantly lowered, (b) additional conditions for working-able individuals were introduced and (c) the wealth test is stiffened and so on. Those who find work can continue to receive ZMN, in decreasing amounts, during the first 3 months of employment. Mentioned changes from 2015 have improved the situation for single persons and single-parent households but couples with children, even if in receipt of child benefits, receive less under the ZMN scheme than under the previous social assistance scheme. ZMN covers only between 32 and 46% of subsistence needs of different model households, defined as 60% of median income. Coverage of the ZMN scheme is low and has not increased significantly during the crisis. Currently, the scheme reaches only about 12% of those who are at risk of poverty, but leakage of the scheme to higher-income groups is extremely low and, indeed, targeting appears to have improved over time.
There is no specific arrangement for the transition from unemployment-related benefits to ZMN benefits. Unemployment benefits received in the previous 3 months are included as income to be taken into account when assessing a claim as is income received from active labour market policies. There is, however, provision for a tapered withdrawal of benefits on return to employment. Those who find work while claiming the benefit may continue to receive ZMN for the first 3 months of employment in decreasing amounts from 100% in the first month, to 75% in the second month and 50% in the third month [23].
Pursuant to the provisions of the Job Placement and Unemployment Insurance Act, unemployment insurance in Croatia includes among other unemployment benefit (UB). Unemployed persons acquire the right to UB provided they had worked 9 out of 24 months prior to termination of employment and that such termination was not caused through their own fault or will. Depending on the total years of service, an unemployed person may claim UB for a period from 90 to 450 days. The basis for determining the UB amount for a person whose work relationship has been terminated is the average salary earned in the 3-month period preceding the termination of employment reduced by statutory contributions. During the first 90 days, UB amounts to 70% of the base amount. During the remaining period in which UB is received, the recipients get 35% of the base amount. During the first 90 days, UB may not exceed 70% of the base amount. In the remaining period, it may not exceed 35% of the average salary paid in the Croatian economy in the previous year. The highest unemployment benefit in 2015 amounted to HRK 3873 (€ 516), whereas the average unemployment benefit amount was HRK 1817 (€ 242). Thus, the net replacement rate (the share of the average unemployment benefit in the wage before unemployment) for the 6-month period was 52.5% which is among the lowest share in the EU, while the coverage rate (the number of the unemployment-benefit recipients in total unemployment) in 2015 was 16.9% [24]. In comparison with other EU member states, UB in Croatia has relatively low replacement rate and very low coverage rate [25].
There is no particular group that the policies target, but some more theoretical attention and discussion have been directed towards the model "does work pay". According to calculations, in Croatia regarding this issue, vulnerable families are those in which an employed member earns a low wage or salary, families that receive all types of available benefits and social assistance, families with only one adult member who works and families with more children. The most significant determinants of the marginal effective tax rates are social welfare subsistence and unemployment benefit.
There are some local benefits like housing benefit and lump sum assistance directed towards the poor population. Housing benefit (pomoć za podmirenje troškova stanovanja) is received for covering the costs of rent, electricity, gas, heating, water and other housing bills. Local self-government (LSG) finance and disburse such benefit and determine income tests and amounts of benefit. The maximum amount is equal to 50% of the amount of subsistence benefit. Since January 2014, it is explicitly defined that the beneficiaries must be receivers of subsistence benefit (now Guaranteed Minimum Benefit). The recipients of such benefit also have the right to claim the benefit for covering the wood-heating costs, which is also administered and financed by LSG units.
Although there are various measures oriented towards the poor and vulnerable, combination of such measures targeted towards in-work poor is still missing. Thus, Nestić et al. [26] conclude that in Croatia, employment does not protect one from poverty for those with minimum wage and badly paid jobs.
Despite relatively high expenditures, most of the benefits in the social welfare systems are low. This is particularly so in the case of social assistance benefits (today Guaranteed Minimum Benefit) and unemployment benefits. The social protection system is mostly public, but there is an increasing privatisation in some social protection sectors (the pension and health systems and service provision in the social services system) [27].
As a part of active labour market policy, in the period 2013-2015, the Project Training for Employment for Workers Facing Unemployment and Long-term Unemployed Persons and other groups unfavourably positioned in the labour market was realised. It included surplus employees, workers facing the threat of layoff and long-term unemployed persons. The main objective of the project was to provide support in the implementation of active employment measures, and its purpose is to increase the level of employability of surplus workers and long-term unemployed persons through vocational training for employment.
The Ministry of Labour and Pension System, in collaboration with its implementing body which is the Croatian Employment Service, during 2015, promoted fiscal relief for hiring longterm unemployed persons in accordance with the stipulations of the Employment Incentives Act. Active labour market policy programmes are continuously being implemented in accordance with the Guidelines for Development and Implementation of the Active Labour Market Policy. Due to the circumstances in which employment of long-term unemployed persons is becoming more difficult, a special focus of measures is oriented towards their activation through additional education or employment based on facilitated criteria and greater aid intensity. Contracting of active labour market policy projects financed by the European Social Fund commenced at the end of 2015.
The Strategy for Combating Poverty and Social Exclusion in the Republic of Croatia (2014-2020) indicated priorities and corresponding primary objectives within eight strategic areas. Education and life-long learning are specified as an area of long-term priority, followed by employment and access to employment; adequate housing and availability of energy; access to social benefits and services; access to the health system; efficient care for the elderly people; the fight against indebtedness and promotion of financial independence and balanced regional development. Each strategic area includes several main strategic activities and the methods of their implementation, which are presented in more detail within a 3-year implementation programme on the national level.
In accordance with the main goal of the Strategy Europe 2020, Croatia aims to reduce the number of persons at the risk of poverty by 1,50,000 by 2020. This is based on the assumption that the economy will grow, the trends on the labour market will be positive and that there will be conditions for the opening of new workplaces. Furthermore, it has planned to implement the labour policy measures oriented toward the long-term unemployed and other vulnerable groups (the youth, persons with disabilities, the Roma). Additional efforts are dedicated to improve the adequacy of social benefits in the social welfare system, the accessibility and uniform availability of the various social rights and services like education, health care, social welfare and others. Finally, attention is given on the availability of housing, the reduction of regional differences and the prevention of indebtedness and financial dependence.
In the achievement of the strategy goals and the implementation of related measures and activities, further principles are respected and followed: a multidimensional approach to dealing with the problem of poverty; solidarity and the principle of non-stigmatisation; subsidiarity principle; objectivity in the planning of measures based on clear indicators; coordination and equal and synergetic participation of all stakeholders in strategy preparation and implementation and an innovative social policy approach. The government is fully aware of the importance of lifelong learning as a precondition for successful employability and sustainability on the labour market, the fulfilment of personal potential and development of active citizenship. A central role in the implementation of the lifelong learning policy belongs, among others, to adult education, which represents a key component of the lifelong learning continuity. The share of persons between the ages of 25 and 64 participating in education and training in Croatia is only 2.9%, whereas the European average is 8.9%, which is an obvious confirmation of the need for additional attention and investment in this area.
The government believes that employment is the best route out of poverty. It is committed to reduce poverty and dependency on social welfare benefit system and to help unemployed people in finding work and support themselves and their families. Continuation of the implementation of reform measures in this area aims at enhancing the efficiency of the system of social benefits, further stimulation of employment of unemployed and activation of inactive persons, strengthening transparency of the system and increasing the adequacy of social welfare benefits. This is achieved with consolidation of social benefits through the development of the mentioned Guaranteed Minimum Benefit and the One Stop Shop.

Conclusions and policy recommendations
In Croatia, there is still a deeply rooted and/or highly widespread paternalism-the belief that the government is obliged and can solve all crucial financial and other problems of its citizens. However, a substantial body of research has shown that that the governments are not able to solve mentioned issues on their own. What is more, there is no single model for poverty reduction that is optimal for and applicable in all societies.
What the government should and could provide is a stable legal framework, social infrastructure that, with the co-operation of its citizens, would establish the rule of law. Croatia, as each and every other country, has to find and develop constitutional and legal arrangements that best suit its own historical, social, cultural and economic situations, conditions and possibilities. Otherwise, the poor will suffer most from the lack of a clear legal framework and unwillingness of other parts of the society to respect the rule of law. It is important to enforce and strengthen the current legal framework rather than pursue further changes in the legislation and the organisational structure. However, in order to make the laws work, political will and leadership commitment is vital, just as important is the empowerment of citizens and their full participation in the political and decision-making process.
The society and state have to be committed to the reduction of poverty. The right way to solve problems of destitute citizens is not by punishing the rich but by doing more to support the poor to become more affluent. Regarding labour activity, for the poor citizens, it is crucial to create preconditions for equal and successful access to the labour market. This can be realised with activation and integration as well as adequate support and targeted efforts in enhancing social inclusion and mitigating adverse personal and social consequences of unemployment. This means not only empowering the bonds of trust and mutual respect among all citizens in the society but also equipping poor citizens by increasing their social capital and employability. This can be accomplished by enhancing their skills, knowledge, attitudes and values, which will enable them to be successful in equal participation on the labour market. A primary goal of economic and social policies must be to get people into work or back into work. For majority of welfare benefit recipients, dependency on social welfare support cannot provide an adequate alternative to regular employment regarding material well-being and psychological satisfaction.
One of the most important priorities of economic policy in Croatia currently represents the increase in employment, with the goal to solve the problem of low employment rate. There is a strong link between low educational attainment, inadequate employability, long-term unemployment and poverty. Many of the poor and long-term unemployed have multidimensional problems that often include a low level of education attainment and motivation. The issue of long-term unemployment in Croatia is very serious and although there was a respectable improvement in active and passive labour market policies, there is still a scope for their further enhancement. Sometimes, it seems that unemployment problems could be resolved immediately, if only there were enough available financial resources for implementation of various employment measures. However, it is more useful and important to invest in "quality" than in quantity of active labour market policy measures. That means to insure appropriate emphasis and orientation on the long-term unemployed and those with a low level of employability.
Furthermore, a coherent framework of employment measures is essential and needed with clearly defined components, which empower each other in resolving the issue of long-term unemployment and help people who can become long-term unemployed. To improve the design and implementation of the active labour market policy measures, the evaluation of effects of performed activities can be of great benefit. Thus, available financial resources are directed to those participants and programmes where there is the biggest benefit and the largest investment return. Next to the improvement of employment opportunities, there is a need to enhance reconciliation of family life and work, primarily related to the inclusion of kids into adequate and affordable preschool institutions.
Furthermore, it is essential to reinforce the social safety net and welfare system through a better design, amended targeting and monitoring the effects of social welfare measures. Finally, it is vital to empower non-governmental sector in the provision of social services and decentralise some social services to the lower level of the government. The priority should be to insure adequate support to the development of social services and to the increase capacity of their providers in combating poverty. Croatia's prospects for economic growth and job creation look good, presuming that fiscal retrenchment and reforms continue. Enjoying political stability, the Croatian Government should continue to promote and improve the current poverty alleviation policies. To be taken into account in designing these policies are various risks, primarily related to the internal path dependency and/or policy slippage as well as vulnerability to adverse external developments and political influences.

Predrag Bejaković
Address all correspondence to: predrag@ijf.hr The Institute of Public Finance, Zagreb, Croatia