Welfare Mentality as a Challenge to European Sustainable Development. What Role for Youth Inclusion and Institutions?

: This paper takes a free-market approach to the idea of welfare. That is, the analysis does not reject the role of the welfare state to ﬁght against poverty and inequality but underlines the perils of the welfare mentality’s proliferation. Sustainable development requires more individual responsibility and less dependence on the state and its redistribution function. The aim of this paper is twofold. Firstly, it aims at showing that welfare mentality imposes challenges on sustainable development. The second aim is to identify the factors associated with welfare mentality. We use data from several international databases and apply correlation, principal components, and multiple regression analyses on a sample of 28 European countries. The results of our study show that welfare mentality negatively inﬂuences sustainable development by being positively correlated with the risk of poverty and the percentage of young people not in employment, education, or training (NEET). At the same time, countries such as Sweden, Denmark, and Luxembourg, widely acknowledged as welfare policy heavens, register low values in terms of preference for redistribution. The main determinants of welfare mentality are found to be a high level of NEET and a low level of economic freedom. This result points to the role of youth inclusion and free-market institutions in diminishing people’s welfare expectations and encouraging them to take better control of their own lives to reach prosperity and not depending on state support.


Introduction
During the last thirty years, sustainable development has been one of the most prolific research areas and the main aspects dealt with within it have most often been derived from its ecological perspective. The Western world's increasing awareness and care for reducing pollution and for ensuring environmental preservation have been openly expressed in numerous stances taken by the academia, public opinion, and political decision-makers.
Nevertheless, sustainable development harmoniously combines three perspectives on social evolution, i.e., ecological, social, and economic. According to WCED [1], sustainable development refers to "development that meets the needs of the present without compromising the ability of future generations to meet their own needs". Arrow et al. [2] define sustainable development as the "economic path along which intergenerational well-being does not decline". Sustainable development may also be regarded as "any economic activity that raises social welfare with the maximum amount of environmental degradation allowable within given economic, social, and technical constraints" [3] (p. 5). Since questions such as "what should be sustained" and "what kind of development do we in the welfare mentality concept. Welfare mentality is "dependency on the state" [10] (p. 79). It is based on ideas such as "government support, particularly welfare, fosters laziness and a lackadaisical attitude about life", and "welfare recipients see public institutions as the only source of redress for all kinds of problems including those that are individual, as well as social, in nature" [11] (p. 214). Another definition of welfare mentality is "the tendency to expect all solutions to come from outside the dependent" [12] (p. 72).
The first aim of this paper is to show that welfare mentality imposes challenges on sustainable development since it implies significant economic costs and maintains dependence. The second aim is to identify the factors associated with welfare mentality at the European level. In this sense, using data from several international databases and applying correlation, principal components and multiple regression analyses on a sample of 28 European countries, our objectives are: (i) To analyze the relationship between welfare mentality and sustainable development, (ii) to identify several determinants of welfare mentality. The results show a negative relationship between welfare mentality and sustainable development. Furthermore, welfare mentality is positively correlated with the risk of poverty and the percentage of young people not in employment, education, or training (NEET). The main determinants of welfare mentality are found to be a high level of NEET and a low level of economic freedom.
The remainder of this paper is structured as follows. The next section sets the background of this research, following both the relationship between welfare mentality and sustainable development and the analysis of the factors related to welfare mentality and describes the research hypotheses. Section 3 discusses the research design and data used. Section 4 presents and analyzes the empirical results. The last section concludes and provides a discussion of the results obtained. The limits of this research and further directions to be exploited are also presented.

Theoretical Framework
Welfare mentality, or dependence on the state's protective actions, is an attitude or set of preferences that encourages its redistributive actions. The last thirty years have been extremely prolific in terms of studies regarding individuals' welfare preferences. By no means exhaustive, our research is aimed to identify the recurrent reasons which underlie individuals' positive attitudes towards redistribution, which is an essential element of the welfare state and to analyze the relation between welfare mentality and sustainable development.
The welfare state origin is related to the social protection measures imposed by Otto von Bismarck in Prussia at the end of the 19th century. The spread of welfare policies in Europe begins at the end of World War I when the state starts to play an active role in more and more areas of its citizens' lives. Nowadays, the perceptions of the welfare state's role and the risks determined by the moral hazard it fosters give rise to passionate debates. Even in environments otherwise favorable to free-market mechanisms, where the superiority and increased efficiency of individual actions compared to state actions are widely acknowledged, the concern for the poorest members of society, generally incapable of taking care of themselves, prevails. The absence of redistributive policies is often associated with the "damnation of the poor" [13]. However, few seem to know that, well before the existence of the welfare state, the responsibility for the poor was taken on by society, which intervened either by means of private charity and religious organizations or by mutual support societies. While during the 18th and at the beginning of the 19th century, state dependence was regarded as stigma [14], this perception gradually faded during the second half of the 20th century. Today, dependence on redistribution is widely accepted and also favored by the perpetuation of citizens' welfare attitudes.
People tend to be in favor of welfare state redistributive policies if they consider that poverty is generated by causes beyond individual control [15]. The perception of the relation between efforts and results is an important trigger of the redistributive vision. The attitude towards redistribution and high taxes is less favorable in societies where economic results are considered to be generated by individuals' efforts. However, things are quite different when income growth is thought to be due to hazard or good fortune [16]. State intervention in income allocation is preferred when the source of wealth is not accurately identified. Moreover, when good fortune proves biased, the welfare state's helping hand is favored. The wealth source related to effort versus hazard also depends on political sympathies. Leftists' passion for equalization is reflected in their positive attitudes towards redistribution [17].
Ideology is an important factor in income redistribution by means of state action. Alesina and Fuchs-Schündeln's 2005 study, which focused on the example of Eastern and Western Germany compared to the former Communist Bloc in Eastern Europe, shows a higher preference for redistribution in Eastern Germany, particularly among elderly respondents [18]. On the other hand, Olivera's study [17] carried out a couple of years later, reached a different result. Former communist countries' respondents are less favorable to redistribution. The explanation could reside in the weakening effect of the communist ideology throughout time. Furthermore, the young generations' increasing awareness of the negative consequences that egalitarian policies can cause may be a factor in this sense.
Redistributive preferences/the positive attitude towards the welfare state are not only derived from the history of the society that an individual belongs to but from his personal history as well. Personal failure triggers pessimism regarding the future and creates a tendency to equalize incomes by means of state action [19]. Institutions have a powerful impact on individuals' preferences and attitudes. On the other hand, individual faith and preferences pattern institutions. The lack of trust in various institutions, in others' honesty and support is reflected in the attitudes towards welfare policies. The state is summoned to intervene when it becomes responsible for reducing income inequalities [20].
The preference for income redistribution policies is sourced in the perception of the causes underlying income inequalities, in the attitude towards private property, in cultural values and political orientation, in family structure and history, in education, age, and income.
Attitudes towards the paternalistic state and redistribution may clearly be conflicting [21]. It is intriguing that individuals who picture themselves as kind and generous believe that the poor would starve to death in the absence of the coercive intervention of the state. According to Hrdy's study [22], the care for the others seems to have expanded at the same rhythm as human evolution. Moreover, other researchers identify a series of reasons why individuals seem to wish to help strangers [23]. The relationship between cognitive abilities and the demand for redistribution is analyzed in Mollerstrom and Seim's study [24]. One of the alleged reasons behind redistribution support is economic self-interest. Self-interest is indicated as a source of welfare preferences in the studies carried out by Corneo and Gruner [25] as well as in those authored by Fong [15] either in terms of the support for welfare policies by the unemployed [26,27] or in terms of the preferences of those on low income [26,28]. Beyond personal interest, understanding the causes underlying poverty generates different attitudes towards redistribution [26,28]. Guillaud's study [29] is focused on professional status in relation to the preference for redistribution. Another research [30] considers that preferences for redistribution are determined by belonging to a certain group. Welfare sympathies are affected by the enhancement of migration mobility because individuals in communities faced with an increase in the number of immigrants are less willing to support redistribution [31]. Linos and West's research [32] is focused on social mobility and preferences for redistribution. The differences occurred in terms of individuals' preferences for redistribution rely, to a great extent, on their trust in resource allocation equity in economy, which ultimately depends on their perception of income earned by personal efforts or by hazard [16]. Self-interest, insurance motives, and social concerns about inequality and efficiency are the main determinants of the preferences for distribution in Durante et al. experiment [33]. As some scholars point out, the personal experience of economic recession changes individual preferences [19]. Moreover, the ideological and cultural experience of life in socialist economies alters individuals' perception of the role, level, and areas of state's intervention in the economy [18].
Economic freedom is reflected in less favorable attitudes to welfare policies [34,35]. Social values [30,36] and historical educational levels [37] represent key factors in redistributive measures. One analysis carried out for 58 countries during a 30-year period supports the reduction of income inequality if the economic freedom score overcomes a certain threshold [34]. In the same line of thought, we may assume that the highest scores in economic freedom are achieved simultaneously with reductions in individuals' preferences for redistribution.
To sum up, by its effects, the increase in economic freedom is associated with the preference for redistribution, i.e., income growth reduces the preference for redistribution, inequalities reduction diminishes the preference for redistribution, stimulation of entrepreneurial action and perception of effort as a trigger of success also reduce the preference for redistribution.
While, in terms of redistributive preferences, the works of Alesina and Giuliano [38] and Bowles and Gintis [39] indicate factors such as ideological preferences and faith in social justice, according to Kerr's study [40], these very much rely on income inequality. The levels of trust set individual relationships and expectations regarding others' actions. Trust creates perceptions related to the provision of public goods and solutions to social dilemmas [41]. Trust in governmental policy efficiency is particularly important in terms of the key role that trust plays in welfare models [42]. Government efficiency and corruption are proven to shape redistributive preferences in 24 OECD countries [43]. Some researchers show that, by only approaching the welfare generosity by primarily outlining three sets of determiners: Income, beliefs, and cultural preferences, the result will not be as significant as with trust [44]. They demonstrate that the differences regarding redistribution support, in other words, the differences in welfare mentality derive from the different degrees of trust among people, in society as a whole and in government efficiency [44]. Government quality and perception of corruption are relevant for the opportunity cost of tax payment as well as for taxpayers' support of welfare policies. If there is a high perception of governmental spending inefficiency and officials are suspected of corruption, the preference for redistribution falls [45].
Education is related to welfare policy support in several scientific works [17,46,47]. The way that education influences individuals' preference for redistribution may be seen as a reflection of their own level of education as well as of the historical level of education of society as a whole. The way people relate to each other, the degree of trust and social values are historically determined by education. The result of the study carried out by Yamamura [37] confirms the hypothesis according to which the education during the first years of nation-states' development in the 19th century significantly contributed to individuals' unfavorable attitude towards redistributive policies. The effect of individuals' current education on either favorable or unfavorable mentality towards welfare policies is rather circumstantial. The U-curve may be applied in this case as well because redistribution support, most often encountered among the less educated, is also found among the preferences of the most educated members of society as a consequence of their leftist sympathies. This context may be partially explained by a shift in beliefs in line with changes occurring in social structures [48]. Educational degrees are associated with additional efforts to achieve social progress. Superior incomes earned by additional efforts determine a reduced preference for redistribution among the educated. However, if the most educated members of society consider that access to education is limited by hazard-related elements, low income, or family background. Their preferences will most likely include support for redistribution.
The welfare state development is dominated by high heterogeneity as a result of "the mediation between economics, politics, and culture" [49]. Despite all differences in terms of approaches to the welfare state, the policies it implies may be identified within three directions-production, reproduction, and distribution [49]. The social policies determined by the distribution function address some situations closely related to income loss or reduction, i.e., unemployment, illness, invalidity, or old age. The welfare state acts in two ways: Either as a piggy bank or Robin Hood [50]. As a piggy bank, the state offers income distribution so as to ensure the population against risks such as unemployment, disability or old age. As Robin Hood, the welfare state will distribute resources among the poorest members of society so as to reduce inequality.
The traditional social function of the welfare state is confirmed by the sustainable development programs promoted by the UN or the World Bank. The fight against inequality is one of the key points on the Sustainable Development Goals (SDGs) agenda. Out of 17 goals, 11 are focused on various aspects related to equality, equity and/or inclusion, and one (Goal 10) deals exclusively with inequality reduction [51]. In order to achieve these objectives on the sustainable development agenda, the consolidation of the welfare state, as well as an extension of its redistribution function, are required.
However, redistribution has a negative impact on trading behavior and efficiency and, thus, on generating economic wealth [52]. Inequality reduction implies economic efficiency costs, though these might not be immediate: "Much of the gain from a redistributive program is immediate, while many of the efficiency costs are delayed and, indeed, less obvious than the gain" [53] (p. 148). In the preface to Okun's book, Kermit Gordon makes one of his most famous statements: "At many points along the way, society confronts choices that offer somewhat more equality at the expense of efficiency or somewhat more efficiency at the expense of equality. In the idiom of the economist, a trade-off emerges between equality and efficiency" [53].
Although sustainable development is a rather ethical-political term than an economic one and despite the fact that its objectives are rather focused on "social justice" than on "economic growth" [54], the social function extension supported by distribution preferences may be threatened on the long run in terms of economic growth capacity. The welfare state is historically related to developed capitalist countries [55]. It is, thus, hard to imagine that the redistribution it implies could actually be possible independently from economic performances. In its attempt to reduce inequalities by means of redistributive measures, the welfare state actually deepens the gaps in the labor market, thus leading to a rather scarce and poorly trained workforce [56]. Moreover, if the law of diminishing marginal utility is taken into account, the increase in the consumption level leads to a decrease in the welfare determined by additional consumption [57].
Redistributive actions are meant to decrease the inequalities and to increase the welfare of the citizens. Sometimes, their results are not the intended ones, or their efficiency very much depends on the context. Several studies support this aspect. When investigating the major determinants of governmental redistribution, Gründler and Köllner show that governments do not incorporate the objectives of the poorest in society when determining the amount of redistribution [58]. The study of Scervini reached a conclusion that "gains from redistribution are negligible for the middle class" and "the amount of redistribution targeted to the middle class is lower in more asymmetric societies" [59] while Milanovic shows that the middle-income groups are losers in the redistribution process when comparing cash transfers and taxes [60]. Neustadt and Zweifel [61] found that the Swiss welfare state does not appear to be sustainable due to the fact that the citizens' estimated marginal willingness to pay is below the current value of their contribution. By analyzing the case of Denmark, the country which has registered the best results in terms of poverty reduction, Esping-Andersen [62] considers that "the truly effective remedy" did not consist in welfare state income redistribution but in policies that promote women's employment.
Social welfare implies a redistribution of responsibility among the market, family, and government [63]. In order to avoid endangering the welfare of future generations, social sustainability implies the necessity to change this culture of dependency. With this purpose, Garces et al. [64] (p. 214) propose to "change the welfare culture away from a conception of the state as the sole tutor of welfare and towards co-responsibility in welfare supply through taking out private sector services in a wider market." In this context, a very important role is played by education, which turns out to be the unique force able to shift the cultural features of a nation, including the welfare state dependency. Education is an important resource for a sustainable society [65]. Education is considered crucial "for advancing the ability of the people to address sustainable development issues" [66]. Therefore, education is both a welfare mentality determiner and a major resource for sustainable development. To a great extent, the expectations related to the redistributive role of the state, as well as individual actions meant to harmoniously combine all the economic, social, and environmental issues which affect each and every one of us, depend on education.
Starting from the state's social function, which can be found in the traditional welfare state model, as well as in the sustainable development directions, and from the preference for redistribution (welfare mentality) determinants, the hypotheses of our research are the following: There is a negative relationship between welfare mentality and sustainable development, in all its three dimensions-economic, social, and environmental.

Hypothesis 2 (H 2 ).
In those countries with a high preference for welfarism, the poverty risk is higher, and the level of education and government effectiveness are lower than in the countries with a low welfare mentality.

Hypothesis 3 (H 3 ).
Good institutions play a significant role in diminishing welfare mentality.

Hypothesis 4 (H 4 ).
The inclusion of young people in education or labor market will most likely lead to a diminishing overall welfare mentality at the country level.

Material and Methods
To test our hypotheses, we used data from international databases, covering 28 European countries in 2014. Their description and source are presented in Table 1.

Indicator Notation Description Source
Welfare mentality (%) WELFARE_MENTAL The percentage of the Europeans in each country who "agreed" with the statement "The Government should take more responsibility to ensure that everybody is provided for".

Special Eurobarometer 418
Real GDP per capita Real GDP per capita The indicator is calculated as the ratio of real GDP to the average population of a specific year.

Long-term unemployment rate
The indicator measures the share of the economically active population aged 15 to 74 who has been unemployed for 12 months or more.

Social Progress Index Social Progress Index
The average of the scores for three dimensions: Basic Human Needs, Foundations of Wellbeing and Opportunity, each disaggregated into four components (Nutrition and basic medical care, Air, water, and sanitation, Shelter and Personal safety; Access to Basic Knowledge, Access to Information and Communications, Health and Wellness and Ecosystem Sustainability; Personal rights, Access to higher education, Personal freedom, and choice and Equity and inclusion)

Social Progress Imperative
At-risk-of-poverty rate (%) RISK_POVERTY The share of people with an equivalized disposable income (after social transfer) below the at-risk-of-poverty threshold, which is set at 60% of the national median equivalized disposable income after social transfers.

Eurostat
The expenditure on social protection (%) SOCIAL_EXP The share of expenditure on social protection in GDP. Eurostat Table 1. Cont.

Indicator Notation Description Source
Impact of social transfers (excluding pensions) on poverty reduction (%)

IMPACT_SOC_TRANSF
Reduction in the at-risk-of-poverty rate in % due to social transfers, calculated as the percentage difference between the at-risk-of-poverty rate before and after social transfers.

Eurostat
Index of economic freedom (index) EC_FREEDOM The index captures ten economic freedoms, graded from 0 to 100, grouped in four categories (rule of law, limited government, regulatory efficiency, and open markets).
The higher the score a country gets, the higher the level of economic freedom.
Heritage Foundation

Government effectiveness (index) GOV_EFFECTIV
A governance indicator computed by the World Bank. It ranges from −2.5 to +2.5, and it reflects the "quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies".

World Bank Governance Indicators
Quality of the education system (index) QUAL_EDUC_SYST The weighted average of the evaluation of the education system in each country from the perspective of meeting the needs of a competitive economy (the scale ranges from 1 = not well at all to 7 = extremely well).

Quality of primary education (index) QUAL_PRIM_EDUC
The weighted average of the evaluation of the quality of primary schools in each country (the scale ranges from 1 = extremely poor-among the worst in the world to 7 = excellent-among the best in the world).

The Global Competitiveness Report
Young people neither in employment nor in education and training (%)

NEET
The percentage of the total population (in the same age group) of young people aged 15 to 24 who are not employed and who have not received any education or training in the four weeks preceding the survey.

Eurostat
Personal freedom(index) Personal freedom It includes two equally weighted parts: Legal protection and security, made up of Rule of Law, and Security and Safety, and specific personal freedoms: Movement, Religion, Association, Assembly, and Civil Society, Expression, and Relationships.

The Cato Institute
Source: Authors' compilation, using the definitions provided by the respective data sources.
The methods employed in the paper are descriptive statistics, factorial analysis, bivariate correlation, and linear regression modeling.
To test our first hypothesis, we analyzed the correlation in order to measure the intensity and significance of the relationship between the welfare mentality and: The economic dimension of sustainable development, measured by real GDP per capita and the long-term unemployment rate, the social and environmental dimensions of the sustainable development of a state, measured by the Social Progress Index. The consistency of the relationships was verified by controlling for the risk of poverty as a measure for the social vulnerability of a country. The considered indicators cover all aspects of sustainable development, i.e., they cover all 17 goals from the 2030 Agenda for Sustainable Development [8]. Real GDP per capita and long-term unemployment rate capture outcomes related to Goal 8-Decent work and economic growth [67], while the Social Progress Index captures outcomes related to the other 16 of the 17 SDGs [68].
The second hypothesis was tested using descriptive statistics, correlation analysis, and principal components analysis. The variables taken into consideration were welfare mentality, at-risk-of-poverty rate, the expenditure on social protection, the impact of social transfers, index of economic freedom, government effectiveness, quality of the educational system, and the percentage of young people neither in employment nor in education or training.
Descriptive statistics are useful in knowing the variation in the data and in identifying possible problems that outliers or the missing data might impose on the outcome. Summaries of the variables include central tendency, variance, and shape of the distribution indicators. To assess the relationships between the variables, we used the principal component analysis (PCA) and the Pearson bivariate correlation coefficient. Given the nature of our data, the PCA helps in reducing the complexity of the initial data by determining the association between the variables and the formed components, which must explain a minimum of 70% of the total variation in the data. Furthermore, this method allows to group the countries in the sample in clusters that manifest the same behavior regarding the variables. The bivariate correlation adds to the information required in finding the direction and intensity of data variation. This step is necessary in order to identify the predictors that may have a significant effect on the variation of welfare mentality.
In order to verify the third and fourth research hypotheses, we identified the main determinants of welfare mentality by building an econometric model. The multiple regression model is built in order to evaluate the significant influences that determine the behavior of welfare mentality. We controlled for two variables that reflect the country's political background, as an important specificity of the countries, and the personal freedom, as a variable that assesses the individual autonomy of the people, as an important characteristic that influences the welfare mentality's relationship with the determinant factors. The estimated values of the regression coefficients will give insight into the quantitative response of welfare mentality under the variation of its determinants. The results provided by this method will answer questions regarding the future actions required to positively change the welfare mentality. Figure 1 below captures the Europeans' approach to welfare. As previously stated, welfare mentality is proxied by the respondents' agreement to the statement "The Government should take more responsibility to ensure that everybody is provided for", from the Special Eurobarometer 418-Social Climate/Wave EB81.5, 2014. As shown, welfare mentality varies substantially across countries, and it is highly valued by half of the European citizens (the EU-28 average is 48%). In thirteen of the European countries (of which eight are former communist countries), the values are above the average, indicating that the welfare status quo enjoys high public appreciation. The Netherlands shows the least support for welfare (23%), while people in Greece are the "champions" in taking a maximalist approach. In order to verify the first hypothesis, we analyzed the correlation to measure the intensity and significance of the relationship between the welfare mentality and:

Results and Discussions
the social and environmental dimensions of the sustainable development of a state, measured by the Social Progress Index; the economic dimension of sustainable development measured by real GDP per capita and the long-term unemployment rate.
The consistency of the relationships was verified by controlling for the risk of poverty as a measure for the social vulnerability of a country.
The correlation matrix ( Figure 2) suggests correlation patterns between all variables. More importantly, all of the three variables of interest (real GDP per capita and long-term unemployment and Social Progress Index) show a medium to strong correlation with the control variable, the risk of poverty. Aside from the economic logic, the data shows that this control variable can be used to demonstrate that the relationships between welfare mentality and the interest variables are not an outcome of the relationships between these variables and the social vulnerability of a state, measured by the risk of poverty. 23  In order to verify the first hypothesis, we analyzed the correlation to measure the intensity and significance of the relationship between the welfare mentality and: the social and environmental dimensions of the sustainable development of a state, measured by the Social Progress Index; -the economic dimension of sustainable development measured by real GDP per capita and the long-term unemployment rate.
The consistency of the relationships was verified by controlling for the risk of poverty as a measure for the social vulnerability of a country.
The correlation matrix ( Figure 2) suggests correlation patterns between all variables. More importantly, all of the three variables of interest (real GDP per capita and long-term unemployment and Social Progress Index) show a medium to strong correlation with the control variable, the risk of poverty. Aside from the economic logic, the data shows that this control variable can be used to demonstrate that the relationships between welfare mentality and the interest variables are not an outcome of the relationships between these variables and the social vulnerability of a state, measured by the risk of poverty. A zero-order, partial and semi-partial correlation coefficients have been calculated in order to test the consistency of the relationships' intensity between the welfare mentality and the three interest variables.
The testing hypotheses for the significance of the three coefficients are as follows: H0: The zero-order correlation coefficient is not significant (There is no significant correlation between the two variables, ignoring all other influences that can affect the strength of the relationship) H0: The partial correlation coefficient is not significant (There is no correlation between the two variables when excluding the effect of the control variable on the two interest variables) H0: The semi-partial correlation coefficient is not significant (There is no correlation between the two variables when excluding the effect of the control variable on the factor variable) Tables 2-4 present the results from estimating and testing the significance of the three correlation coefficients, using the R package, ppcor. Table 2. Zero-order, partial and semi-partial correlation between welfare mentality and Social Progress Index.

Zero-Order Partial Semi-Partial
Social Progress Index The results show a consistent medium intensity correlation between welfare mentality and Social Progress Index for all correlations. The fact that the correlation coefficient does not suffer abrupt changes in value from zero-order to semi-partial correlation, and the correlation coefficients remain significant is a sign of a consistent relationship between the two variables, under the control of the risk of poverty. A zero-order, partial and semi-partial correlation coefficients have been calculated in order to test the consistency of the relationships' intensity between the welfare mentality and the three interest variables.
The testing hypotheses for the significance of the three coefficients are as follows: H0: The zero-order correlation coefficient is not significant (There is no significant correlation between the two variables, ignoring all other influences that can affect the strength of the relationship).
H0: The partial correlation coefficient is not significant (There is no correlation between the two variables when excluding the effect of the control variable on the two interest variables).
H0: The semi-partial correlation coefficient is not significant (There is no correlation between the two variables when excluding the effect of the control variable on the factor variable).
Tables 2-4 present the results from estimating and testing the significance of the three correlation coefficients, using the R package, ppcor. Table 2. Zero-order, partial and semi-partial correlation between welfare mentality and Social Progress Index.  Table 3. Zero-order, partial and semi-partial correlation between welfare mentality and real GDP per capita.  Table 4. Zero-order, partial and semi-partial correlation between welfare mentality and long-term unemployment rate.

Zero-Order Partial Semi-Partial
Long-term unemployment rate The results show a consistent medium intensity correlation between welfare mentality and Social Progress Index for all correlations. The fact that the correlation coefficient does not suffer abrupt changes in value from zero-order to semi-partial correlation, and the correlation coefficients remain significant is a sign of a consistent relationship between the two variables, under the control of the risk of poverty.
When excluding the effect of poverty risk, the relationship between welfare mentality and real GDP per capita suffers a slight reduction, but the relationship remains significant at a 0.1 significance level.
The strongest and robust relationship is between welfare mentality and long-term unemployment rate, with the coefficient having the highest value and the highest significance across the three indicators.
Overall, the results show a significant, negative, and robust relationship between the welfare mentality and the three dimensions of sustainable development.
To test the second research hypothesis, a descriptive analysis was performed at the beginning in order to obtain an overview of the variation of the variables. In Table 5 are presented the main descriptive indicators. The largest differences between countries are observed for government effectiveness, followed by the impact of social transfers and NEET. The smallest differences between countries are recorded for economic freedom. For government effectiveness, the impact of social transfers and economic freedom most countries have values higher than the mean, since there is a negative skewness, as opposed to NEET, where the high values are uncommon, most countries recording values that are smaller than the mean.
Using correlation analysis and factorial analysis, we highlighted the nature of the relationships between variables.
The correlation matrix, presented in Table 6, shows that there are significant correlations between welfare mentality and the rest of the variables, except for the social expenditure. The welfare mentality is positively correlated with the risk of poverty and NEET and negatively correlated with the rest of the variables. As already shown in other studies [17,43], we have found no connection between welfare mentality and social protection expenditure. We have identified both countries whose citizens are in favor of redistribution in the context of reduced social expenditure and so-called welfare-paradise countries, where the increase in the redistribution level is no longer supported by citizens. Some authors no longer consider social spending increase a valid option and believe that reforms meant to stimulate economic growth are clearly more likely to increase income and reduce inequality [69].
A high risk of poverty is associated with an increase in redistribution preferences and welfare mentality proliferation. The NEET category individuals perceive the welfare state as a convenient solution to the problems they face. This aspect is also supported by the fact that the increased unemployment rate among the young brought about by the 2008 crisis has been shown to have led to an increase in the number of supporters of state intervention in the process of income allocation in society [70].
Welfare mentality is negatively associated with social transfers. Another study in the field points to the same result, related to a redistribution paradox defined as "mismatches between the sources of social policies' effectiveness and the sources of social policies' popularity" [71]. Moreover, in line with previous research [43,72], welfare mentality is negatively associated with governmental efficiency. Lack of trust in the state, perception of a high degree of corruption as well as inefficient governmental spending/expenditure contribute to diminishing support of redistribution policies. The relation between the quality of the educational system and support of redistribution is also negative. Individuals' preference for redistribution decreases in line with their educational level [47] and the historical levels of education nevertheless affect the perception of inequality and welfare policies' support [37]. Equal opportunities regarding access to high-quality education reduce income inequalities and decrease support of redistributive policies.
In order to explain the pattern of correlation between our set of variables, we performed a principal component analysis. The information presented in Table 7 gives us an extracted solution of two components that considerably reduce the complexity of the data. Together, the two extracted components explain almost 75% of the total variance of the initial data. By analyzing the data from Table 8 and Figure 3, we found that all the variables are contributing to the formation of the first component, which is most highly correlated with government effectiveness. The only variable that is strongly correlated with the second component is social expenditure. The display of the variables leads us to the result that countries with high values for welfare mentality, NEET, and risk of poverty have small values for government effectiveness, the quality of education, impact of social transfers, economic freedom, and social expenditure. In Figure 4 we plotted the component scores for the two factors and identified the groups of countries with similar or opposite behavior, considering the set of variables. Two groups of countries can be observed: One is formed by Romania, Bulgaria, Croatia, Spain, Greece, Italy, and it is opposed to the group composed by Finland, Denmark, Netherlands, Belgium, Sweden, Ireland, UK, and Luxembourg. The display of the variables leads us to the result that countries with high values for welfare mentality, NEET, and risk of poverty have small values for government effectiveness, the quality of education, impact of social transfers, economic freedom, and social expenditure.
In Figure 4 we plotted the component scores for the two factors and identified the groups of countries with similar or opposite behavior, considering the set of variables. Two groups of countries can be observed: One is formed by Romania, Bulgaria, Croatia, Spain, Greece, Italy, and it is opposed to the group composed by Finland, Denmark, Netherlands, Belgium, Sweden, Ireland, UK, and Luxembourg.  The first group has high values for welfare mentality, NEET and risk-at-poverty rate, and low government effectiveness, low quality of education and low economic freedom, small expenditures on social protection, and a small impact of social transfers. The second group of countries has opposite behavior to the first group, with high quality of education and low level of welfare mentality.
A more in-depth classification in the existing literature [73] points out to five sub-models within the European social model: The Scandinavian one, including Denmark, Finland, Sweden, and the Netherlands; the Anglo-Saxon sub-model, specific to the UK and Ireland; the Continental one, grouping France, Germany, Austria, Belgium, and Luxembourg; the Southern sub-model characteristic to Greece, Italy, Spain, Portugal and Cyprus, and the Catching-up sub-model, which is specific to Central and Eastern European countries.
Various studies have identified the existence of gaps in terms of preferences for redistribution and social policies. For example, there is a strong preference for redistribution in the context of low social expenditure in Greece and Portugal [43]. At the same time, countries such as Sweden, Denmark, and Luxembourg, widely acknowledged as welfare policy heavens, register low values in terms of preference for redistribution. The same aspect is observed by Olivera [17] who shows that Nordic states, traditionally in favor of welfare, mark a low level in their preference for redistribution. This can be explained by the specificity of the Nordic model of welfare, one combining in a unique way liberty and equality [74] (p. 36). Focusing on the Swedish welfarism, Peter Antman observes that "few welfare States are as consistently based on the idea of individual autonomy as is the Swedish. Virtually all of our welfare programs are tied to the individual person, not to the family or to the job as is the norm in other western countries… the struggle for full employment… follows the principle that each person should have power over his or her own life…" [75] (p. 24).
Further, in order to verify the third and fourth research hypotheses, we identified the main determinants of welfare mentality by building an econometric model. We built a multiple linear model, and we estimated the regression coefficients by using OLS. We used the stepwise selection method, and we found that NEET and Economic Freedom significantly influence the welfare mentality. We controlled for two variables that reflect the country's political background, as an important specificity of the countries, and the personal freedom, as a variable that assesses the The first group has high values for welfare mentality, NEET and risk-at-poverty rate, and low government effectiveness, low quality of education and low economic freedom, small expenditures on social protection, and a small impact of social transfers. The second group of countries has opposite behavior to the first group, with high quality of education and low level of welfare mentality.
A more in-depth classification in the existing literature [73] points out to five sub-models within the European social model: The Scandinavian one, including Denmark, Finland, Sweden, and the Netherlands; the Anglo-Saxon sub-model, specific to the UK and Ireland; the Continental one, grouping France, Germany, Austria, Belgium, and Luxembourg; the Southern sub-model characteristic to Greece, Italy, Spain, Portugal and Cyprus, and the Catching-up sub-model, which is specific to Central and Eastern European countries.
Various studies have identified the existence of gaps in terms of preferences for redistribution and social policies. For example, there is a strong preference for redistribution in the context of low social expenditure in Greece and Portugal [43]. At the same time, countries such as Sweden, Denmark, and Luxembourg, widely acknowledged as welfare policy heavens, register low values in terms of preference for redistribution. The same aspect is observed by Olivera [17] who shows that Nordic states, traditionally in favor of welfare, mark a low level in their preference for redistribution. This can be explained by the specificity of the Nordic model of welfare, one combining in a unique way liberty and equality [74] (p. 36). Focusing on the Swedish welfarism, Peter Antman observes that "few welfare States are as consistently based on the idea of individual autonomy as is the Swedish. Virtually all of our welfare programs are tied to the individual person, not to the family or to the job as is the norm in other western countries . . . the struggle for full employment . . . follows the principle that each person should have power over his or her own life . . . " [75] (p. 24).
Further, in order to verify the third and fourth research hypotheses, we identified the main determinants of welfare mentality by building an econometric model. We built a multiple linear model, and we estimated the regression coefficients by using OLS. We used the stepwise selection method, and we found that NEET and Economic Freedom significantly influence the welfare mentality. We controlled for two variables that reflect the country's political background, as an important specificity of the countries, and the personal freedom, as a variable that assesses the individual autonomy of the people, as an important characteristic that influences the welfare mentality's relationship with the determinant factors.
Political background refers to the communist or capitalist past of the analyzed European countries. As can be seen from Figure 1, eight of the countries registering above the average values for welfare mentality are former communist states. As previously shown, ideology was proven as an important factor in explaining the preference for redistribution [18]. Indoctrination, state dependency, and poverty provide explanations for a higher preference for redistribution in some post-communist countries. Political regimes seriously affect preferences. The amputation of private initiative and its substitution with coercive governmental action for almost half a century has left a deep mark on the individual-state relationship. The absence of private property and the prevalence of common wealth illusion can only be synonymous with the equal distribution of economic actions' results. Communism meant the extinction of the effort-result relationship.
In the countries with a capitalist past, welfare is the result of a capitalist way of development where the market economy has worked by all its rules since "respect for private property was intertwined with the unhampered initiative of entrepreneurs, creating new opportunities for capital accumulation and production" [76]. On the other hand, in some of the former communist ones, people still carry the burden of the past and call in for the welfare state to solve their social issues rather than assuming individual responsibility [77].
Personal freedom, as a variable that assesses the individual autonomy of the people, is used as a control variable since, as previously exemplified by the Swedish case, there are countries in our sample which adequately balance welfare state and great individual autonomy while others display less individual autonomy and more reliance on state support.
The political background was defined by a dummy variable taking the value 0 for the post-communist countries in the sample and 1 otherwise.
The values for personal freedom (PF) were recorded into four variation intervals, using as cutpoints the means and standard deviation of the variable, as follows: low personal freedom countries with values between the minimum and one standard deviation below the mean (x − s); -medium-low personal freedom countries with values between one standard deviation below the mean and the mean; -medium-high personal freedom countries with values between the mean and one standard deviation above the mean; -high personal freedom countries that have values higher than one standard deviation above the mean. -This categorical variable with four categories was transformed into three dummy variables: -Medium_low_PF-it takes the value 1 for countries with medium-low personal freedom and 0 otherwise; -Medium_high_PF-it takes 1 for countries with medium-high personal freedom and 0 otherwise; -High_PF-it takes 1 for countries with high personal freedom and 0 otherwise.
For non-colinearity purposes, no dummy variable was built for the low personal freedom category. The linear combination of the two significant factors explains 75.5% of the total variation of welfare mentality (Table 9). The estimated value of the coefficients shows that, for the same level of NEET, the same political background and the same level of personal freedom, an increase with one point in the Index of economic freedom determines a decrease on average of 1.05% of the percentage of people that manifest a welfare mentality. Furthermore, when the percentage of young people neither in employment nor in education and training increases by 1%, the percentage of the people that have a welfare mentality increases on average by 1.563%, all other things being equal. The values of the standardized coefficients show that the most important factor in explaining the welfare mentality is the percentage of young people neither in employment nor in education and training (Table 10). This implies that if we want a diminishment in welfare mentality, the fastest way to achieve it is by reducing the NEET rather than increasing economic freedom. The welfare mentality is more sensitive to youth inclusion. The hypotheses of the regression model are not violated. The errors are normally distributed (Table A1), are homoskedastic (Table A2), are not autocorrelated (Table A3), and there is no collinearity present between the independent variables (Table 9). Therefore, we can assume that the model accurately fits the data.
Our results are in accordance with the ones shown by previous studies on the role of young people inclusion in order to reach a "balanced and sustainable society" and point out to the need to design specific policies to promote youth engagement in society [78]. Moreover, in line with the results of previous research [34,35], our study points to the fact that a high degree of economic freedom is associated with a less favorable attitude towards welfarism. More economic freedom is equivalent to a more powerful role played by the forces of the market economy and to lower expectations regarding the role of the state in providing for the citizens. As Nikolaev and Bennet argue [79], individuals in countries with high levels of economic freedom tend to manifest better control of their own lives. This result may be related to these individuals' low levels of preference for having their own lives controlled by the welfare state. Economic institutions, such as economic freedom, protection of property rights, etc., provide incentives for productive economic behavior and a rational allocation of resources [80]. Moreover, among institutions, economic freedom was proven the most important of them in all development stages [81].
The relation between economic freedom and welfare mentality can be regarded from two perspectives. On the one hand, there is a direct relation between economic freedom and development. The countries with the highest scores in terms of freedom hold the top positions in terms of economic growth. The general increase in income is equivalent to a decrease in the number of those dependent on the state's social protection expenditure. Moreover, a less intrusive state presence in the economy, associated with an increased preference for risk-taking, typical to entrepreneurship-friendly societies, is equivalent to a decreased pressure for redistributive intervention. To a certain extent, individuals' perception of income achievement can also be altered by increased economic freedom. If individual effort and not hazard triggers success, the preference for redistribution falls. Secondly, economic freedom is associated with income inequality reduction after reaching a certain level. We particularly refer to Kuznets's Inverted-U hypothesis. The first steps towards economic freedom are followed by an increase in income inequality because, due to market mechanisms, income allocation primarily favors the individuals in the richest groups. Subsequently, the living standard improvement and income increase within the whole society enable the path to economic freedom to coincide with the one towards income inequality reduction.

Conclusions
The main objectives of this paper were to analyze the relationship between welfare mentality and sustainable development and to identify several determinants of welfare mentality.
We took a free-market approach to the idea of welfare. That is, our analysis does not reject the idea of the welfare state to fight against poverty and inequality, but the paper defended against the proliferation of the welfare mentality and in favor of sustainable development based on freedom and individual responsibility. This means that the individuals should not wait to be provided by the state with a better life but to take responsibility for their life, knowing that they can rely on the state support in case they need it.
The results show that priority should be given to young people's inclusion and economic freedom when it comes to diminishing the welfare mentality. Besides, both youth inclusion in education and the labor market and economic freedom have been acknowledged as key elements in reaching sustainability goals [78].
At the same time, although economic freedom has been proven as an important factor for development [81][82][83], the rising inequalities cannot be left to be solved only by the market. The presence of the state is necessary to support, through a mix of policies, the implementation of sustainability goals and to build a balanced and equitable society. In today's unequal world, "we may not accept that relying on market outcomes is enough to address poverty, so some form of redistribution is required. Moreover, we should also focus on the inequality of income, wealth, or other measures of wellbeing. To address either poverty or inequality may therefore require some level of "taking" from the rich to give to the poor. Redistribution to address inequality might be considered important as a goal in itself, or because unlike the classic idea of a trade-off between growth and redistribution, addressing inequality is seen as instrumental to achieving growth" [84] (p. 6).
As put by Kohler, the present context requires "renewed thinking about opportunities to foster redistributive institutions and policies through domestic reforms and international cooperation is required in order to advance towards a framework enabling redistributive policies for equity and sustainable development" [85] (p. 25). Some researchers argue in favor of a "new sustainability approach, linked to a broad sense of economic, ecological, and social development, with an internalized core belief in the essential value of the social policy for the comprehensive development and the transformation of a society" [86] (p. 13). Considering our research results, we add that the new approach should also empower individuals, especially young people, and make them aware that they are responsible for providing for themselves.
All these aspects point to the idea of a trade-off, a proper balance between the classical principles of the free market, which have proven to be a sure path to prosperity and progress, and the need for flexible and efficient governments to support the free market and to implement effective policies to encourage people to provide for themselves. It is our firm belief that a "give me a fish" type attitude is an unsustainable one, and in order to reach the 2030 Agenda objectives, the government should rather support a "teach me how to fish" mentality.
In our opinion, this study brings certain contributions. To our knowledge, no prior study has analyzed the relationship between welfare state, welfare mentality or preference for redistribution and sustainable development, seen in an integrated way, by reaching a dynamic balance between the objectives of creating wealth as a basis of any economic activity, on one side, and the social and environmental conditions, on the other [87] (p. 93). Moreover, our results add to a mounting body of literature on the preference for redistribution. In this direction, our research mainly contributes to explaining the cross-country differences in welfare mentality and to providing reasons for these differences.
Furthermore, the findings carry normative implications. When investigating the determinants of the preference for redistribution on the basis of macro data, we cannot differentiate between individual preferences, but we are able to account for existing socio-economic and institutional differences between countries. Our analysis provides evidence of the significant influence several factors have on the preference for welfarism, which, in turn, can be inspiring if a government's aim is to support welfare but not encourage its citizens' dependence on the state. This research also emphasizes the need for a trade-off between the principles of the free market and the principles of welfare to reach sustainable development goals.
However, this research carries several limitations deriving from the data we used. Firstly, the study is based on past data due to the lack of a more recent opinion study on Europeans' approach to welfare. Secondly, it is data subjectivity which sheds a certain limitation to the study. Furthermore, the analysis is cross-sectional, all variables being measured at one point in time. Further research should take into consideration the above-mentioned limitations. More in-depth country-level longitudinal studies might shed deeper knowledge on this topic. Moreover, this research operates with aggregates, as variables, and some researchers may argue that it carries some ecological fallacy. Related to this, we believe that the ecological design is appropriate for our study since state intervention, through different kind of welfare policies, may affect all people in a society, regardless of their opinion about welfare and redistribution and thus, this intervention can be conceptualized as a group-level attribute [88]. That is, the individual level analysis is not our real concern, and the conclusions are not inferred to individuals [89]. Actually, we share the belief that "the individual may not be the most appropriate unit of analysis since inferences are to be drawn about whole societies rather than about individuals" [90] (p. 208). Our purpose was not to discover something about the behavior or the nature of the individuals but rather a macroeconomic one, i.e., to suggest several ways for the welfare state to act in order to be effective and to contribute to reaching sustainable development goals. That is to encourage economic freedom, to invest in education, to design and implement policies for young people inclusion in education and/or labor market, etc.
Since welfare policies to address sustainable development goals are complex, and they have varying redistributional effects, further research may evaluate the redistributive effects of specific policies in order to increase governments' effectiveness in spending the existing resources on improving people' lives and reaching eradication of poverty and hunger (SDG 1 and 2), ensuring healthy lives and promoting the well-being at all ages (SDG 3), universal access to education (SDG 4), employment and decent work (SDG 8), and reducing inequalities.