Does the welfare state crowd out informal solidarity?

Transcription

1 Does the welfare state crowd out informal solidarity? A multilevel analysis of the objective and subjective effects of welfare state generosity on informal solidarity in a context of modernisation processes Tilburg University School of Social and Behavioral Sciences Extended Master Sociology Iris Bakker Rotterdam, December 10 th 2014

2 The relationship between institutionalized and informal solidarity in a European context Author: Student-ID: Supervisor: Second reader: I.A.J. (Iris) Bakker u prof. dr. R.J.A. (Ruud) Muffels P.M. (Philipp) Lersch, PhD

3 Abstract The financial credit crisis and various processes of modernisation are putting a strain on the welfare state. Therefore some European countries are transforming into a more participatory society, asking people to take more responsibility of their own lives and that of the people around them. In order for this to work, a certain level of solidarity towards each other is required and the question is if the welfare state hasn t crowded this out. Empirical evidence on the crowding out hypothesis is scarce and findings are sometimes contradictory. Therefore the hypothesis is put to test with a multilevel analysis in this thesis, based on micro-level data from the European Social Survey (2008) and macro-level data from Eurostat (2008). Also the effect of people s attitude towards the welfare state is taken into account and two processes that are increasing demographic diversity: ageing of the population and ethnic heterogeneity. No significant relationship was found between formal solidarity ( welfare state generosity ) and informal solidarity. It was found that a more negative opinion about social benefits and social services results in a lower level of informal solidarity, which can be considered as evidence in favour of the crowding out hypothesis. A second finding is that a positive attitude towards (more) governmental protection and (more) governmental responsibilities is positively related to informal solidarity. Unlike our expectations, a large share of elderly people in society has a significantly negative effect on informal solidarity, while more ethnic heterogeneity has a significantly positive effect on informal solidarity. Key words: solidarity; welfare state; crowding out hypothesis; ageing of population; ethnic heterogeneity

4 Index 1. Introduction Situation Research questions Social and scientific relevance Social relevance Scientific relevance Literature review Classical sociology: solidarity theories Mechanical and organic solidarity Communal and associative relationships Solidarity the concept Formal and informal solidarity Shared identity and shared utility Deservingness criteria The welfare state Welfare state regime types Crowding in and crowding out hypothesis Crowding out hypothesis empirically tested Care for elderly Social capital Informal solidarity Welfare state attitudes The effect of demographic diversity Ageing of the population

5 2.6.2 Ethnic heterogeneity Conceptual model Data Methods Formal objective solidarity generosity of the welfare state Formal subjective solidarity welfare state attitudes Demographic characteristics Informal solidarity Control variables Results Description of the variables Informal solidarity Formal subjective solidarity Macro-level variables Null model ( empty model ) Testing the crowding out hypothesis Full-model analysis Conclusion and discussion The effect of welfare state attitudes The contextual effect of increasing demographic diversity Discussion Bibliography

6 1. Introduction When the Dutch king Willem-Alexander mentioned during his 2013 throne speech that the traditional welfare state is slowly turning into a participatory society ( participatiesamenleving ), this was the starting point of a heated social debate. Many people were wondering how this change would affect society and the welfare state as we know it. Anyone who can, we ask to take responsibility for his or her own life 1, the king said. This suggests that in the current situation people are not taking responsibility, but let the state take care of them instead. The Netherlands like all European countries have a welfare state regime type, that plays a key role in the protection and promotion of the economic and social well-being of its citizens. Due to some recent developments the strain that the welfare state places on both the economy and society have become a hot item. 1.1 Situation In the beginning of the 20 th century the occurrence of two World Wars and the years in between, had shown how unstable democracy could be if citizens did not feel like their governments responded to their needs. After World War II the system of the welfare state was introduced, so that social policies could guarantee a satisfactory level of life to every citizen. By many this political system was considered as an ideal solution. In the 1960s and 1970s the economy grew very fast and parallel to this grew the welfare state. From the mid-1970s onwards welfare states were faced with rising unemployment rates, high levels of inflation and low economic growth (Gelissen, 2000). Therefore a continuingly growing percentage of people s incomes was required in order to maintain the welfare state. In 2007/2008 the worldwide financial credit crisis hit Europe, with a serious economic decline as a consequence. Again this resulted in higher costs for the welfare states and thus more discussion about the structure of the current system. Apart from this some other modernisation processes put a strain on the welfare state as well, such as individualisation, the ageing of the population, increasing diversity and heterogeneity and globalisation (De Beer, 2012). Some of these processes have turned out to be more relevant than others, but taken altogether the assumption 1 King Willem-Alexander during his first throne speech (2013). Obtained on from 3

7 rises that a transformation is awaiting the welfare state in order to meet the critics. In some European countries this criticism has led to a shift towards a more social government is, for instance in France, Slovakia, Belgium and Denmark. In other countries a shift towards a more liberal government is going on, for instance in Slovenia and Spain. The Netherlands is one of the countries following the example of a big society as set in the United Kingdom (Blond, 2010) recreating a society in which people take their own social responsibilities. The line between people s own responsibilities and governmental responsibilities is a field of tension, because there is always the question of where the one ends and the other begins. Most people will agree that certain groups in society need extra support, but who is responsible for this? In the Netherlands the idea of a participatory society (also: do-democracy) is considered as the solution to all welfare state related issues. In order to realise such a society, citizens have to be willing to contribute and have a certain level of solidarity towards each other. Now the welfare state itself can also be considered as solidarity institutionalized solidarity : the idea that we are all contributing to a system that we do not all directly benefit from, without expecting anything in return. The question that remains is how institutionalized ( formal ) solidarity and the solidarity the Dutch government is expecting from its citizens ( informal solidarity ), are related. Although the welfare state is considered to have many positive effects, the system may also have some adverse effects. One of these effects is captured in the crowding out hypothesis. According to the crowding out hypothesis more generous welfare states displace informal feelings of solidarity, such as the previously mentioned willingness to help other people (Van Oorschot & Arts, 2005). The crowding out hypothesis suggests that the formal solidarity the welfare state provides, substitutes informal solidarity instead of complementing it. The main interest of this thesis will be to find out whether or not this crowding out effect takes place in Europe. Since different types of welfare states exist in Europe, it is likely that differences will be found in the level of institutionalized solidarity that will allow us to test this hypothesis regarding welfare state effects on informal solidarity. 4

8 The second interest of the thesis will cover welfare state attitudes. This is mainly a reflection of the topic of institutionalized solidarity: what are people s opinions about the level of welfare state provisions in their country? This can also be considered as a type of formal solidarity, but while the first part of the study concerns objective formal solidarity, the second part concerns subjective formal solidarity. Besides controlling for the effect of formal objective solidarity, measuring the welfares state attitudes in European countries also provides the opportunity of directly testing how this topic is related to people s level of informal solidarity. The third part of the thesis will consist of including two modernisation processes in order to find out if the extent to which they take place in a country, significantly influences the level of informal solidarity. The level of solidarity and the extent to which people feel like they belong to a group is generally influenced by two concepts: the idea of a shared identity ( we are one ) or a shared utility ( we need each other ) (Van Oorschot, 2000). Two processes that may cause a decrease of shared identity and shared utility are the increasing level of immigrants living in a country and the ageing of the population. In a more diverse country with more ethnic heterogeneity or a country with a higher share of elderly people it is likely that people feel less connected to each other, with a lower level of informal solidarity as a result. Also, it is likely that these two processes have a negative effect on people s attitudes towards the welfare state in general. Therefore the share of elderly and immigrants in a country will be included as main effects in the empirical model. 1.2 Research questions To test the crowding out hypothesis in a context of modernisation processes, three main research questions will be answered in this thesis: 1. To what extent does more formal objective solidarity (i.e. a more generous welfare state) crowd out informal solidarity in European countries? 2. To what extent does more formal subjective solidarity (i.e. a more positive attitude towards the welfare state) crowd out informal solidarity in European countries? 3. What is the effect of a more diverse population regarding ethnicity and elderly people in European countries on informal solidarity and how does it affect the relationship between objective and subjective formal solidarity and informal solidarity? 5

9 1.3 Social and scientific relevance Social relevance Some European welfare states currently find themselves in a time of transformation. Due to certain developments the structure of the system is much debated and government s are open to new models. One of these models is called big society or participatory society. This model is to a large extent based on more responsibilities for citizens and leaning more on your own social environment. Many people have however gotten used to the welfare state providing all kinds of support, because this has been the situation in the past decades. The question is therefore if people are still solidary enough towards each other or if this has decreased under the influence of the welfare state. If the latter is the case, then this seriously complicates the introduction of a new system that is based on solidarity and more responsibilities for one's environment and network. It is therefore quite relevant to investigate the condition of solidarity in European countries and the influence of the welfare state and demographic diversity on informal solidarity. Since due to the complexity of the system many citizens might not even feel like the welfare state has anything to do with solidarity anymore, also the effects of their opinions about services and benefits are included. This way both the objective and subjective effect are taken into account Scientific relevance The strength of the study lies in the discovery of the field of tension between institutionalized solidarity and informal solidarity. The crowding out hypothesis has been tested many times before on several societal characteristics, including informal solidarity. Yet not one-dimensional results can be presented on the topic. The different types of welfare state in Europe offer a nice opportunity of doing it once more to find out how the generosity of the welfare state affects informal solidarity. What this thesis specifically adds to the topic, is the reflection of institutionalized solidarity by also including welfare state attitudes. This can be considered as testing the crowding out hypothesis on the individual level and to my knowledge this has not been done before. A final addition of the thesis it the inclusion of modernisation processes. The welfare state itself is directly affected by society and two main processes that are going on right now are likely to influence the level of shared identity and shared utility and therefore result in a decrease of informal solidarity. Taking into account a measurement of ethnic heterogeneity and share of 6

10 elderly adds a better contextual conceptualisation to the model. 7

11 2. Literature review This chapter contains the theoretical framework. The first part will focus on the different concepts of solidarity in social sciences. The second part is about the crowding in and crowding out hypothesis. After that, the previous studies on welfare state effects on societal concepts will be discussed and finally the possible impact of ethnic heterogeneity and ageing of the population on solidarity. But first to start things of: a classical sociological introduction. 2.1 Classical sociology: solidarity theories Mechanical and organic solidarity Ever since Emile Durkheim s The Division of Labour in Society (1893), solidarity and society have always been connected in social sciences. In this famous piece of work Durkheim describes two types of solidarity: mechanical solidarity and organic solidarity. The idea behind mechanical solidarity is that social cohesion and integration are the result of the homogeneity of individuals. People are connected by their similarities - in work, education, religion and lifestyle. This type of traditional solidarity normally operates in smaller scaled societies, or through kinship ties of familial networks (Durkheim, 1893). Mechanical solidarity is closely connected to the collective consciousness. This is a situation in which individuals share the same fundamental cultural elements, which they use as a basis for recognising and accepting each other as members of the same collectivity (Durkheim, 1893; Van Oorschot, 1998). In situations of collective consciousness, individuals identify strongly with the group because of mechanical solidarity. According to organic solidarity, social cohesion is based not on the similarities of people, but upon the interdependence that arises from specialization of work and the complementarities between people. This is merely what occurs in modern and industrial societies: social cohesion is based upon the dependence individuals have on each other in more advanced societies (Durkheim, 1893) Communal and associative relationships Durkheim analyses solidarity as a characteristic of a collectivity or a society; something that can 8

12 only be studied on a macro level. Another important sociologist with classical influences on solidarity is Max Weber. He approaches the subject on micro level, as a characterisation of social relations between individuals (Weber, 1964). Weber also makes a distinction between two types of solidary relations: communal relationships and associative relationships. Communal relationships occur when the orientation of the action is based on the subjective feeling of both parties that they belong together; either based on affection or tradition. Associative relationships occur when the orientation of the action is focussed on treating each other according to a rationally motivated adjustment of interests, or an agreement similarly motivated. The relationship is of interest to both actors and is thus based on utility. When comparing the theories of both Durkheim and Weber, it cannot go unnoticed that Durkheim s mechanical solidarity resembles somewhat like Weber s communal relationships, while organic solidarity shows similarities with associative relationships. This immediately brings out the two core elements of solidarity: in pretty much every case solidarity is either based on the idea of a shared identity or a shared utility. In the next chapter this will be further specified. Beginning this theoretical framework with the ideas of Durkheim and Weber is not coincidental. Their theories have been criticized often and in many ways, but yet they are still a very important starting point for basically every study on the topic of solidarity. Durkheim and Weber were two of the many social scientists who consider solidarity and social cohesion as a positive phenomenon. On a group level, solidarity seems mainly something that ties people together, like the social glue of society. In that way it should definitely be considered positive. On a larger scale, group solidarity also has its dangers. Groups with for instance strong ethnic or religious beliefs do not only include, but also exclude people or consider other groups their enemy. This situation has in the past led to extreme situations, such as ethnic cleansing or violent suppression. Solidarity should therefore always be studied with a certain cautiousness and never be put as unconditionally positive. 2.2 Solidarity the concept Since the days of Durkheim, the world has changed somewhat and so has the concept of 9

13 solidarity. Apart from the distinction mentioned above, many more distinctions have been made. Let s start with the broad sense and then funnel the way down to the definition that will be most fitting for this thesis. Solidarity is nowadays a favoured topic of the media and this determines to a large extent what image people have of solidarity. However when the media publish something about solidarity, this can mean a wide variety of things, such as giving money to charity, voluntary work, informal care for your family or neighbours, a positive opinion about social benefits, etc. All of these things can be considered solidarity, although very different of character. To at least be able to capture this broad sense, De Beer and Koster (2007) came up with the following definition of solidarity: every action one does that benefits another person, without certainty for the contributor that this action will ever be repaid equally. In this definition nothing is mentioned about whether solidarity is voluntary or mandatory, what motives people have to be solidary and the likeliness of being repaid. In order to properly study the concept of solidarity, such distinctions will have to be made Formal and informal solidarity The most important distinction is between formal and informal solidarity. This refers to the way in which solidarity is organized. Formal solidarity is mandatory solidarity, also referred to as state or macro solidarity. It is based on arrangements of risk sharing and risk pooling such as the welfare state and concerns the transfer of funds by the state (Sol-Bronk & Voorneveld, 2011). The social distance between giver and received is very large. Formal solidarity can be described as cold solidarity, because it concerns involvement with anonymous third parties and is based on policy arrangement instead of warm feelings (De Beer, 2008). In order for people to support these policy arrangements without directly knowing who they support, a connection between people is required. This connection could be based on the fact that one is a member of the same collective and therefore shares the same risks (De Beer, 2008). Shutting out risks exposes the first motivation for this type of solidarity: self-interest. The second motivation is accepted authority people are obliged to pay taxes to the state and with this money formal solidarity is financed (Van Oorschot, 1999). Both motives show that this type of solidarity doesn t have anything to do with warm feelings towards fellow citizens. 10

14 Informal solidarity refers to familiar or micro solidarity. It concerns non-mandatory solidarity that is often not organized, such as help or support within your family, network or even towards strangers on the street (Sol-Bronk & Voorneveld, 2011). Where the social distance between giver and receiver is very large for formal solidarity, for informal solidarity it is characterised by a small social distance. Informal solidarity can be described as warm solidarity, because motives are mostly affection and equality (De Beer, 2008). While formal solidarity is mainly cash and benefits, informal solidarity can also be gifts or food transfers. Another distinction was made by Paul de Beer and Ferry Koster (2009): between one-sided and two-sided solidarity. One-sided solidarity means someone will do something for another human being without ever expecting anything in return. This kind of help usually runs from the stronger to the weaker, by motives of affection or moral obligation (such as charity). Two-sided solidarity means doing something for another human being and expecting to (ever) get better from it yourself. Motives for this are usually self-interest (such as insurances). Taking into account all different types of solidarity, two will be the centre of this thesis. First of all, formal solidarity. Formal solidarity will be defined as the generosity of the welfare state in a country. Secondly, informal solidarity. Informal solidarity will be defined as the extent to which people feel it is important to help the people around them and care for their well-being Shared identity and shared utility When discussing the motives of solidarity, one assumption is dominating. Solidarity is generally based on two things: shared identity and shared utility. In the case of shared identity, we identify with one another and therefore a feeling of we are one is created. This means that the interest of the other integrates in yourself: your interest is my interest is our interest. According to the other possibility, solidarity is based on shared utility. In this case people share fates because they depend on each other for realising life opportunities. Therefore there is a perception of we need each other (Van Oorschot, 1998). These two perceptions show similarities with the concepts from Durkheim and Weber. Putting the concepts next to each other, it becomes clear that a higher level of solidarity will be found in a social system with more people and groups involved who identify with one another and if they 11

15 are more depending on one another. This is something to keep in mind when creating expectations on solidarity in different countries Deservingness criteria Apart from basic criteria on solidarity, people s solidarity is not quite unconditionally. Some conditions are set in order to decide if people actually deserve the feelings of solidarity from others. Van Oorschot (2000) derived five of these deservingness conditions. First of all, there is control over neediness, which questions if people are needy because of something they did themselves or if they can t be held responsible for their neediness. If the latter is the case, people are considered more deserving. Secondly there s the level of need, simply meaning that people with greater needs are considered as more deserving. Thirdly there s a familiar one: identity. The needy ones who we are closer to are considered more needy than the people that stand further away. A fourth criterion is attitude. This has to do with personal preferences: if we like the attitude of the needy one, so if the needy one is more likeable, grateful, compliant and conforming to our standards, we feel they deserve more solidarity. The final criterion is reciprocity, a very rational criterion based on the idea that people who have contributed to our group before or may be expected to do so in the future, earn our support. Of these five, control seems to be the most important criterion and identity the second most important (Van Oorschot, 2000). Some of these deservingness criteria are relevant to the hypotheses, but that will be specified in a later chapter. 2.3 The welfare state Every European countries has a certain type of welfare state to regulate it s social services. But what exactly can or cannot be considered a welfare state? In order to answer that question, a definition will be presented. According to Offe (1984) the welfare state is: The system of political institutions in a capitalist society which deploys the instruments of monetary and fiscal policy to ensure stable economic growth; which provides education and training programs to facilitate labour market participation and income support programs to protect citizens against the vicissitudes of the market; and whose legitimacy derives from popular consent organized through regular elections. ( ) From this definition we can conclude that the welfare state is more than just a bag of money for 12

16 those who need it. It is a comprehensive system that regulates the economy and people s incomes, in order to protect citizens against the market. It is rather difficult to provide a more specific definition, because there is no such thing as the welfare state. All western countries use the system, but in almost as many different shapes. Since every country is entitled to make its own decisions on how much they want to spend on which social policy areas, quite some differences exist among the different European welfare state. Because of these differences, different types of welfare state regimes can be distinguished Welfare state regime types Distinguishing welfare state regime types was firstly done by Gøsta Esping-Andersen in his famous work The three worlds of welfare capitalism (1990). According to Esping-Andersen, the welfare state is more than just the sum of discrete programmes and social policies. The relation between state and economy in a country is determined by discrete logic of organization, stratification and societal integration. In the degree of decommodification and the kind of stratification they produce in society, a pattern can be witnessed identifying three models of welfare state: conservative, liberal and social-democratic. These three models are considered ideal-types and the real world is therefore likely to exhibit hybrid forms. Basically every country in Europe presents a system mix (Esping-Andersen, 1990; Arts & Gelissen, 2002). The three worlds of welfare capitalism evoked a lot of response, including authors who didn t agree with the typology and wanted to add new or other welfare types. Relevant for Europe seemed to be the Mediterranean or Southern type (Arts and Gelissen,2001; Bonoli, 1997). Other alternative typologies include the East-Asian type and a radical type to distinguish Australia and New Zealand, because their liberal regime differs somewhat from the others (Castles, 1998; Fenger, 2007). Fenger (2007) incorporated post-communist countries in Central and Eastern Europe to the welfare regime typology. He found that there are clear differences in the lines of the governments of traditional Western welfare states and post-communist welfare states. The postcommunist countries show quite some variety themselves and can be divided into three groups: a group of former USSR-countries, a group of rather successful Central and Eastern European countries and a group of developing welfare states. His typology is based on numerous characteristics of governmental programmes, social situation variables and political participation 13

17 variables. Of the 23 countries studied in this thesis, 19 were part of Fenger s analysis as well. They were distinguished as follows: I: Conservative-corporatist type (Belgium, France, Germany, Greece, The Netherlands and Spain) II: Social-Democratic type (Finland, Denmark, Norway and Sweden) III: Liberal type (United Kingdom) IV: Former-USSR type (Estonia, Latvia) V: Post-communist European type (Bulgaria, Czech Republic, Hungary, Poland and Slovakia) VI: Developing welfare states type (Romania) These regime types show an extensive pattern of differences among countries, for example on total government expenditures, expenditures on health, education or social protection, inequality, inflation, unemployment, female participation and level of trust. What lacks in his research is how all these regime types are related to societal characteristics. Arts and Gelissen (2001) did use the welfare state typology by Esping-Andersen (1990) to investigate whether people s notions of solidarity are related to the type of welfare state regime they live under. According to their findings an actual relationship exists between the welfare regime type of the country in which people live and their views about which level of solidarity should be achieved. The differences they found between the studied welfare states were however quite small, which they explain by stating that the fact that there is a welfare state regime in a country already implies a certain level of solidarity. The fact that Europe can so clearly be divided into different types of welfare states confirms the idea that there is enough variation to use it as a case for a comparative study on the effects of welfare state generosity and welfare state attitudes Crowding in and crowding out hypothesis Apart from the question what the welfare state really is and which types can be distinguished, there is also the question of how the welfare state effects civil society. This is where the distinction between formal and informal solidarity becomes particularly relevant, because this distinction forms the basis of the hypothesis of crowding in and crowding out. 14

18 The hypothesis of crowding out is originally an economic concept, meaning that increased public sector spending replaces or drives down private investment. Higher government expenditure increases the demand for goods and services which causes a raise in interest rates. Therefore, capital will become more expensive and private investment will be reduced. The crowding out hypothesis is the traditional point of view (Aschauer, 1989; Ahmed & Miller, 2000). This negative relationship is also known as the substitution hypothesis (Van Oorschot & Arts, 2005). The crowding in hypothesis is the non-traditional point of view and assumes the exact opposite, namely that higher government expenditure would stimulate private investments of different kinds. Crowding in generally occurs under different economic circumstances: when resources are un- and underemployed (Admed & Miller, 2000). This positive relationship is also known as the reinforcement hypothesis. In a situation with more formal solidarity citizens will be given more space for personal investments, based on the assumption that the government creates conditions in which this is allowed (Sol-Bronk & Voorneveld, 2011). In social sciences the theory about crowding in and crowding out focuses on the effect of government expenditure (i.e. the welfare state) on societal characteristics. It is a recurrent critique of the welfare state that in spite of or even because of its good intentions, the system has some unintended, negative social and moral consequences. The welfare state is said to have a crowding out effect upon the social capital of the society it serves (Van Oorschot & Arts, 2005). The effect of social expenditures and comprehensive social programmes is to crowd out informal caring relations and social networks. Because of the state taking responsibility for this kind of care, familial communal and occupational systems of self-help and reciprocity seem unnecessary. As a consequence, the welfare state fosters social isolation, anomie and selfcentredness. Altogether this leads to a general decline of commitment to civil norms, participation in civil society and trust in fellow citizens and social institutions (Fukyuama, 2000; Putnam, 2000; Van Oorschot & Arts, 2005). Opponents of this theory support the crowding in hypothesis. They argue that a country with a well-developed welfare state is a positive environment for a civil society. Because of structural and cultural conditions, comprehensive welfare states create a society in which people have enough financial resources and free time to actively develop their social capital. Also, the state sets an example of taking responsibility for the well-being of the people around you. Because of 15

19 this, people are said to behave more solidaristically and impartially (Rothstein, 2001; Van Oorschot & Arts, 2005). 2.4 Crowding out hypothesis empirically tested Several authors have empirically tested the effects of the welfare state on certain elements of society. Studies on the effects on just informal solidarity are scarce and therefore also studies on some other elements of society will be described, to find out which mechanisms form the basis of this relationship Care for elderly Care for elderly is a favoured topic in the literature on the crowding out hypothesis. A strong welfare state is claimed to reduce the willingness of families to provide financial support and services to their aged parents. The supply of public pensions and services for elderly people assumingly crowds out family solidarity. Because the state already provides adequate incomes, services and care for the elderly, their family or neighbours feel less need to help them out. This is considered a shift in responsibility from the family to a public solidarity system (Künemund & Rein, 1999). The question is now to what extent this shift is empirically tested. Künemund & Rein (1999) explored the relationship between different levels of welfare spending and the level of support for elderly people. They studied five developed countries: the United States, Canada, the United Kingdom, West-Germany and Japan. Their findings were two-sided. One the hand their international comparison did not support the crowding out hypothesis. By looking at the times elderly people received help when they were ill, their house needed taking care of, help with transport or helping out with money they found that in most of these situations, elderly people could count on a certain level of support. They also studied the impact of giving on receiving and found that the giving of services by older people to their adult children increases the probability that they will receive help from them themselves. This indirectly supports the crowding in theory: in generous welfare states people still receive family solidarity. However what s missing in this study is the comparison with countries who are not as developed when it comes to the welfare, so the findings cannot be put in such a perspective. Motel-Klingebiel, Tesch-Roemer and Von Kondratowitz (2005) did study countries with 16

20 different types of welfare regimes with different types of care provision: Norway, England, Germany, Spain and Israel. Spain and Germany have a more familiaristic orientation than the other countries. In Norway and Israel the welfare state is more generous than in the other countries. A first thing they found is that in countries with a familiaristic orientation and a weaker infrastructure of social services, family help is significantly higher. So: family help is lower in more generous welfare states. This supports the crowding out hypothesis. However when including personal and household attributes to the analysis, the results were no longer significant. This could mean two things: either the effects of the welfare state are less relevant than personal and household attributes, or the welfare state effects these personal and household attributes also. Their second relevant finding was that the highest aggregate provision of help and support was found in generous welfare regimes. This supports the hypothesis of mixed responsibility : in the generous welfare state regimes, help from formal services alone or in combination with family help was greatest (Motel-Klingebiel et al., 2005) Social capital Another topic that the welfare state is said to influence, is social capital. Putnam (1993) wrote that social capital in a country is a resource that improves government performance. The question is however if a certain governmental performance also improves social capital; are social capital and government substitutes or might they complement each other? Kumlin and Rothstein (2005) took the case of Scandinavia as a starting point for their analysis. Scandinavia has extensive welfare state arrangements and such arrangements are likely to crowd out social capital. Instead of people organizing themselves in associations that reach out to fellow citizens in distress, people rely on the welfare state to take care of those people. Therefore this type of welfare state should have a negative effect on families, communities and social networks (Wolfe, 1989). Sweden however has a wealth of social capital. This paradox is what Kumlin and Rothstein wanted to research. Their findings were mixed: the welfare state could have both a positive and a negative effect on social capital. Positive is that experience with universal institutions tends to build trust. Negative is that personal experience with welfare state institutions seem to reduce interpersonal trust. So again, no clear outcome. Kääriäinen and Lehtonen (2006) also studied the occurrence of social capital in 21 countries 17

21 with different kinds of welfare state regimes. They were looking for a systematic variation of social capital and found mainly evidence for the hypothesis that social capital emerges particularly in societies with strong welfare state institutions. Their final conclusion was however also that the welfare state both prevents and promotes the development of social capital. The different forms of social capital are connected to welfare state regimes in different ways (Kääriäinen and Lehtonen, 2006). Taking into account these different studies, the idea that the type of welfare state regimes matters grows stronger. Most results show an effect; different effects for different types of welfare regimes. In some cases this supports the crowding in hypothesis, in some cases the crowding out hypothesis and in other cases it is more a matter of mixed responsibility. The direction of the effect remains therefore unclear Informal solidarity A generous welfare state can be considered as state-organised solidarity. Taking into account the mixed findings on both the topics of care for elderly and social capital, it may not come as a surprise that studies on the effect of the welfare state on informal solidarity also show quite some inconsistency. Van Oorschot, Arts and Halman (2005) studied welfare state effects on both social capital and informal solidarity. Between the European countries they studied, they didn t find much variation in average levels of informal solidarity. In order to try and explain this variation, they grouped the countries together based on welfare state regime types, but their results showed that between-type variation was smaller than within-type variation. This could be possible evidence for the assumption of Künemund and Rein (2001) that regime types are blurring and that hybrid types are proliferating. Another finding of Van Oorschot et al. (2005) was that welfare state effort is negatively related to informal solidarity regarding disabled people and unemployed people, thus supporting the crowding out hypothesis. Additionally, there is a high positive correlation between a country s level of expenditure on social protection and social capital. So on this topic, their findings confirm the crowding in hypothesis. A change in welfare state spending involves a trade-off between feelings of informal solidarity on the one hand and voluntary activities and social trust 18

22 on the other (Van Oorschot et al., 2005). Welfare state spending has an effect on either one of the two and not both. This is confirmed by the fact that the two concepts are not statistically correlated. Altogether their study cannot be considered conclusive evidence for either the crowding in or the crowding out hypothesis. Paskov and Dewilde (2012) studied the relationship between income inequality, a macro-level characteristic, and solidarity of Europeans. Income inequality is directly affected by the arrangements of the welfare state; after all the welfare state is an instrument that prevents extreme poverty and limits extreme wealth because higher incomes are subject to higher taxations. Paskov and Dewilde defined solidarity as informal solidarity, towards different groups. They conceptualised it with the following question: Would you be prepared to actually do something to improve the conditions of: a)people in your neighbourhood/community; b) elderly in your country; c) sick and disabled people in your country; d) immigrants in your country. Their results show that in more unequal societies, people are significantly less likely to engage in improving the living conditions of members of their community, elderly people and the sick and disabled. This is the case for people living in both low-income households and high-income households. Solidity towards immigrants is not significantly related to inequality, which is a first incentive to the possible effects of a large share of immigrants. This effect will be further discussed in the next chapter. Taken altogether, most studies aim in the direction of the crowding out hypothesis: social expenditures displace informal caring relations, social networks and occupational systems (Van Oorschot and Arts, 2005; Halevy, Chou and Galinsky, 2011). A safety net is already created on state level and therefore people feel less obliged to solidarity in their everyday lives. The expected effect of formal solidarity on informal solidarity will therefore be negative. This results in the first hypothesis: H1: In countries with a higher level of social expenditure, the level of informal solidarity among citizens will be lower (i.e. crowding-out hypothesis) 2.5 Welfare state attitudes Apart from the actual objective state of the welfare state in a country there are also the opinions 19

23 of people on this topic. This can be described as the formal subjective solidarity, or in other words: the attitudes of people in a country towards their welfare state. Perhaps even better than the previous, these attitudes actually capture the feelings of people and their perception of the level of formal solidarity. The perceived level of formal solidarity is the result of both situational and ideological factors. The understandings and beliefs of people about national social problems and about the relationships between individuals, the state and other institutions influence popular beliefs: what kind of policies should the state pursue and who should benefit (Blekesaune & Quadagno, 2003)? Both situational and ideological factors are important. If the situation in a country is bad, for instance because of economic decline, high unemployment rates, high taxes or (speculated) corruption, this is likely to influence people s attitudes towards the welfare. On the other hand: if people in a country are very much pro egalitarianism or very much liberal, this is also likely to affect their attitudes towards the welfare state. The situational characteristics of the welfare state are already captured in the concept of formal solidarity. Therefore, the basic expectation of people agreeing or disagreeing with the level of social expenditure in their country will be based on ideology (i.e. political preferences). Measuring people s feelings about topics such as income inequality, unemployment benefit expenditure and old age benefit expenditure says a lot about their political views at the left or right side of the political spectrum. On the left side a more progressive and redistributive view of the role of the state can be found, while at the right side a more conservative and market-oriented opinion exists (De Vries, Hakhverdian and Lancee, 2013). It might be expected that leftist people feel more strongly solidaristic than rightist people. In other words: leftist people are likely to think the level of social expenditure in their country is justified or even too low, while rightist people are more likely to feel like the level of social expenditure is too high. Previous studies have shown that leftist people have a more positive attitude towards equality, social rights and welfare support (Coughlin, 1980; Papadakis and Bean, 1993; Van Oorschot, 2000). These findings were confirmed by Van Oorschot and Arts (2005) who found that political values are reflected in people s welfare attitudes. These feelings directly reflect a certain level of solidarity at individual level. When someone is likely to sympathize with people in their country they don t even know, and feel like they need to be taken care of; they will also be more likely to care for the people around them. This results in 20

24 the second hypothesis: H2: People with a more positive attitude towards the generosity of the welfare state of their country, will have a higher level of informal solidarity. 2.6 The effect of demographic diversity Welfare states were established in a period when equality was considered important and services were directed at a relatively large working class with fairly homogenous preferences. Nowadays there is a way more diversity because of which welfare states have to take account of more heterogeneous needs and expectations (Esping-Andersen, 1996). Certain groups in society need far more support than others and the question is what effect this has on the solidarity with each other. Two current processes in modern society will be included in the analysis as main effects: the ageing of the population and the ethnic heterogeneity in a country Ageing of the population One of the greatest challenges the welfare state is currently facing is the ageing of the population. Low birth rates combined with higher life expectancy are the causes of an extremely high old-age dependency ratio; 27.5% in 2014 for the EU-28. The old-age dependency ratio is calculated as the ratio of dependent old people (over 65) compared with the population aged 15 to 64 years old (Eurostat, 2014). The relative number of people who are retired is expanding, while the (relative) number of people of working age is declining. In the coming decades this will increase even more, because of the share of post-war baby boomers who are about to retire. For those still working this will lead to an increased burden of the social expenditure required by the ageing of the population (Eurostat, 2014). Clearly it is safe to state that the population of Europe is ageing rapidly and this puts a strain on the welfare state. An increasing share of the population is affected by the risks of this problem, such as risk of poverty or risk of poor health. Addressing these risks challenges the intergenerational solidarity. The question relevant to us is how this affects people s feelings about the welfare state and their level of informal solidarity. Motel-Klinebiel, Tesch-Roemer and Von Kondrawitz (2005) studied the cases of Norway, England, Germany, Spain and Israel to see if older people received more help in countries with a stronger infrastructure of formal services. No decisive evidence was found for either the 21

25 crowding in or crowding out hypothesis. Their results supported the hypothesis of mixed responsibility : in non-familistic welfare state regimes help from families and welfare state services accumulate, but in familistic welfare states this is not the case. To derive an hypothesis about the effect of ageing is not easy. There is one starting point that provides some clarity: the deservingness criteria by Van Oorschot (2000). Elderly people fall into the first category control over neediness and are likely to deserve solidarity because they have no control over their neediness: they can t help being old. Also the second category level of need works in their favour: most older people really do need help with things in their everyday lives. Thirdly, there is identity. The older people in one s country are likely to have been living there for quite some time and thus will it be easier to identify with them. The fifth criterion reciprocity is again fairly important when it comes to elderly people; they are likely to have contributed themselves in the past and therefore deserve our solidarity now. This does not directly imply that ageing of the population will also increase the level of informal solidarity, but taken altogether the presence of a large share of (deserving) elderly is likely to have a positive effect on informal solidarity. This result in the third hypothesis: H3: In countries with a larger share of elderly people, informal solidarity will be higher Ethnic heterogeneity Ethnic heterogeneity in relation to solidarity is not a topic that has been able to count on a lot of positive attention, nor by the media or by scientists. In recent years it has become a muchdebated issue - the increasing diversity and heterogeneity amongst citizens may threaten the sustainability of the welfare state. Shared identity is an important condition for the system, because we are more willing to pay for people who remind us of ourselves and our habits. Putnam (2007) claims that more ethnic diversity does not just weaken the ties between different ethnic groups, but also within ethnic groups. He states that natives living in an ethnic diverse neighbourhood tend to hunker down and lose the connection to the other natives. When again taking into account the deservingness criteria by Van Oorschot (2000), this doesn t seem to improve the effects of an ethnic diverse country. Immigrants don t do all too well on these five criteria - they have a certain amount of control over their neediness, there is no shared identity, there has not been reciprocity and also attitude could be problematic, because of the cultural differences. It s only their level of need that might make them more deserving, but in 22

26 this case it will have more to do with pity and less with actual warm feelings. The presence of (non-deserving) immigrants is thus likely to have a negative effect on the level on informal solidarity. This results in the fourth hypothesis: H4: In countries with a higher level of ethnic heterogeneity, informal solidarity will be lower. Apart from the element of deservingness, there is another element affected by the presence of immigrants and elderly. Since they are both partly needy groups, they put an extra strain on the costs of the welfare state (OESO, 2013). The extra costs that this brings along, is likely to have a (latent) effect on the relationship between people s attitude towards the welfare state and informal solidarity. Therefore an interaction effect will be included in the analysis, hypothesizing the following: H5: In countries with a more diverse demography considering elderly and immigrants, the (positive) effect of formal subjective solidarity on informal solidarity will be weaker than in countries with less demographic diversity. Formal subjective solidarity can also be considered as a mediator variable, because it explains the relationship between two variables: formal objective solidarity and informal solidarity. The state of the art of the welfare state directly influences the attitudes of people towards the welfare state and thus the level of informal solidarity. For the mediator it can be hypothesized that: H6: The (negative) relationship between formal objective solidarity and informal solidarity can be explained through the effect of formal subjective solidarity. 2.7 Conceptual model The hypotheses about the different effects of the independent variables on informal solidarity are presented in the conceptual model in figure

27 Figure 2.1: Conceptual model regarding the hypothesised effects explaining informal solidarity 24

28 3. Data The data source to be used for the thesis will be the 2008 round of the European Social Survey (ESS). The ESS is an academically driven cross-national survey that has been conducted every two years across Europe since The survey measures the attitudes, beliefs and behavioural patterns of diverse populations in more than thirty nations. Every round consists of a core section and a rotating section. The 2008 round includes a section on welfare and concerns the attitudes towards, and perceptions and evaluations of welfare policies in the broad sense. Also it focuses on attitudes towards welfare state provision, views on taxation, attitudes towards service delivery and future dependence on welfare (ESS, 2008). This provides a unique opportunity of studying the attitudes of European citizens towards the welfare state. In 2008, the European Union consisted of 27 member states. 22 of these states are included in the 2008 wave, namely: Belgium, Bulgaria, Cyprus, Denmark, Germany, Estonia, Finland, France, Greece, United Kingdom, Hungary, Ireland, Latvia, The Netherlands, Poland, Portugal, Romania, Slovenia, Slovakia, Spain, Czech Republic and Sweden. Austria, Italy, Lithuania, Malta and Luxembourg are not included in the ESS and can therefore not be part of the analysis. Two non-eu members that are included in the ESS and they will therefore be added to the other 22 countries: Switzerland and Norway. This would have resulted in an analysis with a total of 24 European countries. During the research it turned out that most required data for Cyprus could not be obtained and therefore it was decided to exclude this country from analysis. The final analysis is thus done on 23 countries. When using ESS-data, analyses should always be conducted using weights. Weighting the data is important, because then results will be based on samples using estimates which take into account how likely each respondent is part of the sample. Only design weight will be applied, because all countries will be separately compared to each other. 25

29 4. Methodology In this chapter the conceptualisation and operationalization of the different variables will be discussed. This includes the construction of different scales for formal objective and formal subjective solidarity, the implementation of various macro-variables and the construction of control variables. 4.1 Formal objective solidarity generosity of the welfare state Social expenditure as a percentage of GDP The most well-known and possibly also the most important indicator for measuring the generosity a welfare state is gross total social expenditure as a percent of GDP. Which percentage of the total GDP a country is willing to spend on social expenditure tells a lot about how social the regime is structured. This percentage is the sum of the spending on various domains of social policy, including social benefits, administration costs and other expenditure linked to social protection schemes. Eurostat provides the exact percentages each year in their Social protection expenditure (Eurostat, 2008). Social benefits are "gross", which means that their value represents the total amount of payment, before deduction of taxes on income or other obligatory levies paid by recipients. Looking at this percentage only would be a slightly unsatisfactory conceptualisation of the generosity of the welfare state. After all, expenditures are the results of two underlying factors that also provide information about the generosity of the welfare state: the volume and the average level of social benefits. The volume is on the one hand determined by coverage: how universal or strict are the policies that determine who is and who isn t entitled to social benefits. On the other hand the situation of the market is of influence; in times of economic prosperity fewer people will be in need of social benefits than in economically bad times and thus will the percentage of GDP spend on social expenditures be lower. The level of social benefits is determined by the generosity of the replacement rates. A replacement rate is the ratio of an individual s benefit as a percentage of previous earnings in a given time period (OECD, 2005). The higher the replacement rates in a country, the more 26

30 generous the welfare state is. Another element to take into account is the complexity of national tax and benefit systems. Governments have two opportunities of supporting their citizens: cash benefits and benefits in kind. Cash benefits are the financial resources the government provides. Benefits in kind are the benefits employees receive that cannot be converted into cash but do have a cash value, such as private use of a company car, free or subsidised accommodation and preferential loans. Unemployment replacement rates An unemployment replacement rate is the ratio of unemployment benefits a worker receives relative to the worker s last gross earning. For unemployment replacement rates only cash benefits are of relevance, since unemployed people are no longer employees and cannot profit benefits in kinds. In many countries the replacement rate depends on the housing situation. For instance if someone has no children and is a single person or if two people are a one-earner married couple family, they could receive more or less money than someone who has two children and belongs to a lone parent or a one-earner married couple family. The OECD provides overall-averages of the benefits of these family types and these will be part of the analysis. Another thing that is of interest is the period people will receive the benefits - in some cases this is only a few months and in other cases it could be years. The OECD accounted the average of net replacement rates over 60 months of unemployment. Finally this number is after tax and including unemployment and family benefits. Social assistance and other means-tested benefits are assumed to be available subject to relevant income conditions (OECD Tax-Benefit Model, 2008). Pension replacement rates Pension replacement rate shows the level of pensions in retirement relative to earnings when working. In Pensions at glance (2009) the OECD provides information about the ratio in various countries. The table Gross pension replacement rates by earnings contains the percentages of the previous incomes that people receive when they are retired. The table contains this information for the median income, the mean income, and half, three quarters, one and a half or two times the mean income. The replacement rate for the mean income provides most information about the overall generosity of the welfare state by covering the expenses for both 27

31 low and high incomes, so this percentage is included in the analysis. In Poland and Switzerland pension replacement rates differ between men and women. In Poland the replacement ratio is 61.2% for men and 44.5% for women and in Switzerland the ratio is 58.3% for men and 59% for women. To take account of this difference, the mean of these two will be taken and included in the analysis. Individual income tax rate Another ratio that provides information about the extent to which formal solidarity is present in a country, is the individual income tax rate. This percentage reveals the relative percentage of tax paid by the different income groups in society. Because taxes are the main source of income for the welfare state they are a direct measurement of the generosity of this system. Eurostat captured this information in the table Total tax revenue in percentage of the GDP. The 2008 numbers of this table are included in the analysis. Gini-coefficient The income inequality in a country can be measured by using the Gini-coefficient: a widely used measure that ranges from 0 (everyone has the same income) to 1 (one person owns all the income). Since the welfare state is also a system of redistribution, the Gini-coefficient provides information about how well a country succeeds in this. The Gini-coefficient can be derived from the OECD ilibrary (2008). For the countries missing in this table, the data are completed with information from the Worldbank. The Gini-coefficient is measured at disposable income, post taxes and transfers. This way of the welfare state is most visible. A higher score of the Gini-coefficient normally means more inequality, while a higher score should here represent a more generous welfare state. Therefore the coefficient will be recoded, so that a higher score equals more equality. The five items measuring welfare state generosity are standardized and show a one-dimensional outcome, which means that they are measuring one concept. The Cronbach s Alpha for the new scale is.787 that makes it almost perfectly reliable. Because some countries have missing data for one of the items, the minimum of present items is set at four. This way, all countries will still be included in the analysis. 28

32 4.2 Formal subjective solidarity welfare state attitudes The welfare state attitudes of people are a reflection of both the actual condition of the welfare state and people s own welfare state preferences. For measuring welfare state attitudes I have created three different scales. The three different scales measure the following concepts: government s responsibility, effect of social benefits/services and governmental protection. Government s responsibility Government s responsibility is a concept that measures to what extent people feel like the government is responsible for facilitating social protection of some type. The scale of government s responsibility is created with the following six items: Government s responsibility to ensure a job for everyone who wants one Government s responsibility to ensure adequate health care for the sick Government s responsibility to ensure a reasonable standard of living for the old Government s responsibility to ensure a reasonable standard of living for the unemployed Government s responsibility to ensure sufficient child care services for working parents Government s responsibility to provide paid leave from work for people who temporarily have to care for sick family members Scores are measured from 0 to 10, where score 0 means not government s responsibility at all and 10 means entirely government s responsibility. A high score equals a positive attitude towards government s responsibility and therefore towards formal solidarity. The Cronbach s Alpha is which makes it a highly reliable scale. Effect of social benefits/services Effect of social benefits and/or services is a concept that measures to what extent people feel like social benefits/services have a positive or negative effect on the economy or people s own efforts in life. The scale of government s responsibility is created with the five following items: Social benefits/services place too great a strain on economy Social benefits/services cost businesses too much in taxes/charges Social benefits/services make people lazy Social benefits/services make people less willing to care for one another Social benefits/services make people less willing to look after themselves and/or their 29

33 family Scores are measured from 1 to 5, where 1 means Agree strongly and 5 means Disagree strongly. A high score equals a positive attitude towards social benefits/services and therefore towards formal solidarity. The Cronbach s Alpha = 0.783, which makes this scale very reliable as well. Governmental protection Governmental protection is a concept that measures to what extent people feel that it is the governments job to protect society from poverty and inequality, but also to improve the workfamily balance. The scale of governmental protection consists of the following three items: Social benefits/services prevent widespread poverty Social benefits/services lead to a more equal society Social benefits/services make it easier to combine work and family Scores are measured from 1 to 5, where 1 means Agree strongly and 5 means Disagree strongly. According to this distribution a high score would equal a negative attitude towards formal solidarity and this is unwanted. Therefore the scores of this items will be recoded so that a higher score represents a more positive attitude on social benefits/services and thus on formal solidarity. The Cronbach s Alpha = which is large enough to consider it an acceptably reliable scale. When controlling for the reliability of all the items together, the Cronbach s Alpha is This is high, indicating that the scale represents a good conceptualisation of the presumed underlying concept that is formal subjective solidarity. To deal with missing data, each scale is created with instructions on how many variables are allowed to be missing. This number was set at the total number of items minus 1. Also, because the three scales were not measured on the same rank of scores, the final scale variables are standardized as z-scores. 4.3 Demographic characteristics Ethnic heterogeneity Ethnic heterogeneity or ethnic diversity can be defined in different ways. The percentage of 30

34 foreign born people in a country captures this concept, because this simply refers to residents of a country who were born in another country. In general the people who are born in their country of residence are called citizens, while the people born outside are non-citizens or residents. Under certain circumstances foreign-born residents can change their status and become citizens through naturalization. The population can thus be divided into three categories: native-born citizens, foreign-born citizens and non-citizens (Grieco, 2002). Now these last two groups will determine our concept of ethnic heterogeneity. The OECD provides this information for 2008 in their table Stocks of foreign-born population in OECD-countries - it shows the percentage of foreign-born as a percentage of the total population. For Belgium, 2008 was not included and therefore the numbers of 2007 were used. Unfortunately eleven of the countries in our analysis are not included in this table at all. Therefore a second table was used: Foreign-born population, 2010 (Eurostat, 2010). This percentage of foreign-born population is not from 2008 but is the closest to be found. For Germany, Estonia, Greece, Hungary, Latvia, Poland and Slovenia data from 2010 was used. Ageing of the population Ageing of the population is measured by using the Old Age Dependency Ratio. This indicator is the ratio between the total number of elderly persons of an age when they are generally economically inactive (aged 65 and over) and the number of persons of working age (from 15 to 64). Eurostat provides data for the ratio of 2008 in the table Old-age-dependency ratio. 4.4 Informal solidarity The dependent variable is informal solidarity. This item is conceptualised with a variable in the ESS measuring people s opinion on how much they are like the following person: It's very important to her/him to help the people around her/him. She/he wants to care for their well-being. Scores are ranked 1 to 6, where 1=very much like me and 6=not like me at all. The aim is to have a higher score represent a higher level of informal solidarity and to accomplish this the scores are switched around. 31

35 4.5 Control variables Gender is turned into a dummy variable where score 1 is male and female is the reference category. Employment status originally has nine answer categories, but this will be limited to six. The first category consists of employed people who do paid work and thus make their own income. The second category consists of people who are following an education. These people are generally quite young and are likely to receive support from either their parents or the government to pay for their education. The third category consists of unemployed people (who are either actively looking for a job or not) and permanently sick or disabled people. These are the people who need social support in order to get by. The fourth category is people who are retired. They also are likely to receive a certain amount of money (pension) from the state. The fifth category consists of people who spend their days doing housework or looking after the children. The sixth category consists of people who do community or military service or other. Employment status is turned into five dummies with reference category other, since this is the largest category. The age of the respondents will be limited from 15 to 95. People with age higher than 95 were not included in the analysis. Because the direction of the effect may vary per age group, three categories for age were created. The first category contains people aged 15-29, the second category and the third category Education level is distinguished into five categories. The first category means Less than lower secondary education. The second category means Lower secondary education completed. The third category included people with Upper secondary education completed. The fourth category means Post-secondary non-tertiary education completed and the final fifth category represents Tertiary education completed. Education level is turned into four dummies with the lowest category as reference. Household income is divided into deciles 1 to This number shows how income is distributed: how much of the total income in a country is earned by certain earning groups. These 2 The values of the different deciles vary per country and can be found here: 32

36 values can thus be copied. Some people chose refusal when asked for their household income. For this group a dummy was created, because refusing to provide information about income is generally done for a reason. Also this way the amount of missing data could be limited. The final control variable is macro-level and is GDP per capita, which is taken from Eurostat (2008) and added to the data. 33

37 5. Results In this chapter the results of the analysis will be presented. To work with both the individuallevel variables as well as with the country-level variables, multilevel (mixed models) analysis is used for a stepwise procedure estimating various models. 5.1 Description of the variables Before presenting the result of the actual analysis, first an impression of the values and distribution of the dependent variable and the Eurostat/OECD macro-variables Informal solidarity The dependent variable level of informal solidarity is measured on a scale from 1 to 6, with 6 representing the highest level of informal solidarity as possible. In table 5.1 an overview is provided of the means of informal solidarity per country. Also the N per country is reported and for the 24 countries this is quite equal, ranging from 1254 (Slovenia) to 2741 (Germany). Table 5.1: Mean of informal solidarity per country Mean (M) SD N Belgium Bulgaria Switzerland Czech Republic Germany Denmark Estonia , Spain Finland France United Kingdom Greece Hungary Ireland Latvia Netherlands

38 Norway Poland Portugal Romania Sweden Slovenia Slovakia Source: ESS Total N=43533 As the table shows and the figure below will confirm the scores of the different countries do not show extreme variations, but yet they do differ substantially the lowest score for informal solidarity is (Czech Republic) while the highest score is (Spain). On a scale from 1 to 6 this is quite a difference and therefore worth explaining. The relationship between country and informal solidarity is significant p<.01. The mean of all countries together is translated to the meaning of the answering possibilities this number holds the middle between like me and somewhat like me on the question of how much you are like a person who finds it very important to help the people around him/her and wants to care for their well-being. Figure 5.1: Image of mean of informal solidarity per country 6,0000 5,0000 4,0000 3,0000 2,0000 1,0000 0,0000 Source: ESS (2008) Informal solidarity 35

39 5.2.2 Formal subjective solidarity The concept formal subjective solidarity is measured with three scales: government s responsibility, governmental protection and the effect of social benefits/services. The scales are standardized, because they did not have the same range in scores. The values of the three standardized scales per country are presented in the three separate figures 5.2, 5.3 and 5.4. Figure 5.2: The standardized scores of government s responsibility per country 1,00 Government s responsibility 0,50 0,00-0,50-1,00 Government s responsibility Figure 5.3: The standardized scores of governmental protection per country 0,50 Governmental protection 0,00-0,50-1,00-1,50 Governmental protection 36

40 Figure 5.4: The standardized scores of the effect of social benefits/services per country 0,80 0,60 0,40 0,20 0,00-0,20-0,40-0,60 Effect social benefits/services Effect social benefits/services In the first figure regarding government s responsibility it shows that a positive attitude towards this is lowest in Switzerland (-.81) and highest in Latvia (.72). For both governmental protection and effect of social benefits/services the value for most negative attitude was found in Hungary (-1.07 and (-.53). The highest value of a positive attitude towards governmental protection was found in Belgium (.39) and for effect of social befits/services it was again Latvia (.57) Macro-level variables The different macro-level variables are obtained from either Eurostat or OECD-iLibrary. An overview of the numbers on each variable regarding formal objective solidarity is presented in table

41 Table 5.2: Descriptives of the country-level variables measuring Formal objective solidarity Social expenditure as % of GDP Income tax ratio Pension ratio Unemployment ratio Gini (reversed) Belgium Bulgaria Switzerland Czech Republic Germany Denmark Estonia Spain Finland France United Kingdom Greece Hungary Ireland Latvia Netherlands Norway Poland Portugal Romania Sweden Slovenia Slovakia Source: Eurostat/OECD-iLibrary (2008) The social expenditure as a percentage of GDP varies from 12.7 (Latvia) to 31.3 (France). Since this is an important indicator for the generosity of the welfare state, it is right away clear that the levels of formal objective solidarity differ substantially between these European countries. The income tax ratio is lowest in Switzerland (28%) and highest in Denmark (48.6%). The pension replacement rates are lowest in Czech Republic (26.3) and highest in Greece (95.7%). Then the unemployment replacement rates: these are lowest in Greece (21%) and highest in Ireland (80%). It is remarkable that Greece scores highest on the first replacement ratio and lowest on the second, but this is a typical example of the large differences that exist between welfare states. Finally the Gini-coefficient (reversed) is lowest in Portugal (0.641) and highest in Slovenia (0.764). In Slovenia the extent to which income or consumption is distributed among individuals or households is thus most equal. 38

42 The other two macro-variables are ethnic heterogeneity and ageing of the population. The values of those two are presented in figure 2. Figure 5.5: Demographic diversity as a result of old-age dependency ratio and ethnic heterogeneity 60,0000 Demographic diversity 50, , , , ,0000 0,0000 Old-age dependency ratio Ethnic heterogeneity Source: Eurostat/OECD-iLibrary (2008). NB: for ethnic heterogeneity data from 2010 was used for Germany, Estonia, Greece, Hungary, Latvia, Poland and Slovenia Romania and Bulgaria have missing data. By adding the two characteristics up, it shows how diverse the countries are when taking both processes into account. Separately, the old-age dependency ratio is lowest in Ireland (15.6%) and the highest share of elderly people lives in Germany (30.4). For ethnic heterogeneity the lowest share of foreign-born people lives in Poland (1.2%) and the highest share in Switzerland (25.8%). When combining forces of the two demographic characteristics, Switzerland and Poland are also the ones with the least and most diverse demography in general. 5.2 Null model ( empty model ) The null-model or empty model is meant to gain insight on how much of the variability in the outcome of informal solidarity can be explained by the clustering within countries. 39