The Role of Values for Social Cohesion: Theoretical Explication and Empirical Exploration

Shared values are deemed necessary as a solid foundation for social cohesion by commentators and observers in many countries. However, when examining what kind of values this is based on, answers often come down to platitudes and national clichés. This discussion paper offers some clarification through both a theoretical explication and an empirical exploration concerning the general role of values for social cohesion.<br><br>Values are notions about desirable, trans-situational end-states and behaviors. They fall into two categories, individual and societal values. We provide a critical discussion of the most prominent conceptualizations and their operationalization in the social sciences.<br><br>Values affect social cohesion in three possible pathways: First, when they are shared; second, when they promote behavior per se conducive to social cohesion and third, through their effect on policy choice and institutional design. We review evidence provided by the research literature for each of these pathways.<br><br>We further explore the third pathway by deriving from the research literature the conjecture that a cultural value emphasis on egalitarianism makes a universalistic scope of welfare institutions more likely, which in turn increases social and political trust. We first examine this conjecture with a series of regression models, and then run a mediation analysis. The results show that <br><br>(1.) egalitarian values are moderately strongly and positively linked to universalistic welfare institutions, but that <br><br>(2.) welfare institutions mediate the association of egalitarian values with social trust only to a small extent and that <br><br>(3.) more universalistic welfare institutions counteract a negative association between egalitarian values and institutional trust.


Preface
This Discussion Paper is part of DIE's research project "Preconditions for Sustainable Development: Social Cohesion in Africa". Social cohesion -or social solidarity -within societies is a key success factor for sustainable development in Africa. However, social cohesion is also particularly under pressure in African societies and other world regions. The DIE team aims at identifying patterns of social cohesion in Africa, analyses factors that influence the degree of social cohesion (or its absence) and identifies domestic and international policies that contribute to the creation and consolidation of social cohesion. The team addresses five issue areas: 1. Measuring social cohesion in African societies across countries; 2. Effects of tax systems and social policy on strengthening social cohesion in Africa; 3. Interdependence of financial systems design (small and medium-sized enterprises) and social cohesion 4. Relevance of values, democracy and political institutions for social cohesion 5. Influence of external peacebuilding, political institutions and individual attitudes on societal peace and social cohesion. This research is funded by the German Federal Ministry for Economic Cooperation and Development (BMZ).
We hope that DIE research will help to better understand the drivers of social cohesion and to formulate policies that contribute to cohesive societies worldwide.

Abstract
Shared values are deemed necessary as a solid foundation for social cohesion by commentators and observers in many countries. However, when examining what kind of values this is based on, answers often come down to platitudes and national clichés. This discussion paper offers some clarification through both a theoretical explication and an empirical exploration concerning the general role of values for social cohesion.
Values are notions about desirable, trans-situational end-states and behaviours. They fall into two categories, individual and societal values. We provide a critical discussion of the most prominent conceptualisations and their operationalisation in the social sciences.
Values affect social cohesion in three possible pathways: First, when they are shared; second, when they promote behaviour per se conducive to social cohesion and third, through their effect on policy choice and institutional design. We review evidence provided by the research literature for each of these pathways.
We further explore the third pathway by deriving from the research literature the conjecture that a cultural value emphasis on egalitarianism makes a universalistic scope of welfare institutions more likely, which in turn increases social and political trust. We first examine this conjecture with a series of regression models, and then run a mediation analysis. The results show that (1.) egalitarian values are moderately strongly and positively linked to universalistic welfare institutions, but that (2.) welfare institutions mediate the association of egalitarian values with social trust only to a small extent and that (3.) more universalistic welfare institutions counteract a negative association between egalitarian values and institutional trust.   Tables   Table 1: Overview of value concepts 7   Appendix Tables   Table A1: Overview of variables and sources 43  In the UK, the department of education issued guidelines for schools to teach fundamental British values after unverified rumours that Islamists were indoctrinating students in Birmingham (Vincent & Hunter-Henin, 2018); in Germany, the former minister of the interior revived a debate on how German "core culture", or "Leitkultur", can be defined two years after the start of Europe's so-called "migration crisis" (Zeit Online, 2019); in the US, the editorial board of USA Today issued a statement that President Trump's "America First" policy doctrine betrays "core American values" (USA Today, 2019); and in Kenya, the government launched the National Cohesion and Integration Commission with the mandate to "promote national identity and values" as a response to the electoral violence in -2008(NCIC, 2019. In times of polarisation, many societies turn to debating their common values as a basis of unity. The basic premise of such debates is that common and shared values are the essential cultural core that unites societies. This conviction often manifests itself during public discourse in the media, or in statements by politicians. But it can also be found in the social sciences, where it has a long history dating back to the works of Emile Durkheim and Talcott Parsons (Kolstø & Tyldum, 2006).
Nevertheless, many questions are left in the dark when politicians, the media or academics speak of values as a basis for social cohesion. Why should values transform a group of unrelated individuals into a cohesive social entity? What are the mechanisms by which values create social cohesion? What do we mean by values? And how do they come about?
Research efforts in different fields have begun to rob the concept of values of its mystique. By summarising and synthesising these research efforts, this discussion paper provides a general theoretical explication in its first part, and in its second part, presents a tentative empirical exploration of the role values play for social cohesion.
The first part, the theoretical section, starts with a conceptualisation of social cohesion. We define social cohesion as formed by the interlinkage of the three attributes inclusive social identity, trust and cooperation for the common good. We then proceed to values as a driver for social cohesion and present an introduction to the value concept and different value types. Values are socially transmitted notions of what end-states and behaviours are desirable across different situations and fall into two categories, individual and societal values. Many different value types in different stages of conceptualisation have been brought forward by researchers. Prominent ones include political values, Schwartz' 'basic values' as well as Inglehart and Welzel's 'self-expression values' among personal-level values. Schwartz' 'cultural values' and Hofstede's 'cultural value dimensions' are the most prominent ones for society-level values. We close the introduction of the different value types with a critical discussion of the universality assumption implicit in much of value research.
We then move on to the question of how values might affect social cohesion. We describe two possible pathways as presented by the literature, and contribute a third based on the relationship between values, and institutions and policy regimes. First, values, of whatever nature, are conducive to social cohesion when they are shared by many members of a society. Dissent on which values matter most leads to polarisation that undermines social cohesion. Value consensus, the opposite of dissent, is thus conducive to social cohesion. Second, certain values promote behaviours and end-states that are conducive to social cohesion per se, such as egalitarianism and individualism. In this pathway, values have a direct influence on social cohesion through the content they promulgate. Third, some values affect the formation, change and design of social, political and economic institutions and policies, which then influence social cohesion. There is nothing natural about which institutions and policies a society implements. Instead, institutions and policies are "humanly devised" (North, 1990, p. 3), created and changed by people who, in creating institutions and policies, draw on their values, attitudes, ideas, notions, preconceptions and other mental notions. Institutions and policies in turn may undermine or strengthen social cohesion depending on whether they marginalise and polarise, or include and unify societal groups. Thus, values influence social cohesion indirectly through their link to institutions and policies. We review evidence provided by the research literature for each of these pathways.
We further explore the third pathway in the subsequent empirical part. In this exploration, we focus on welfare policy regimes as a mediating variable of values on social and institutional trust. A well-established literature draws a distinction between means-tested welfare policies that allocate social assistance to a particular beneficiary group (e.g. social transfers for the very poor) and universalistic policies that cover the entire population according to uniform rules of eligibility (Rothstein, 2008). Means-tested welfare policies are associated with societal stratification and stigmatisation of the poor. Universalistic welfare policies in contrast draw on a universal rights-based approach and are associated with more societal solidarity. From the research literature, we derive the conjecture that a cultural value emphasis on egalitarianism makes a universalistic welfare policy regime more likely, which in turn increases social and political trust. Using cross-sectional data from 63 countries that represent each continent with at least one country, we first investigate this conjecture with a series of regression models, and then run a mediation analysis.
Our paper contributes to the literature by synthesising evidence for the influence of values on social cohesion, and explicating possible pathways of how values exert this influence. Besides drawing on pathways that are already presented in the literature (via shared values and value content) we add a third, in which values influence the formation and change of institutions and policies which then in turn affect social cohesion. Our preliminary findings suggest general implications for the centrality of values as well as institutions and policy regimes for social cohesion: The right institutions and policy regimes can improve social cohesion, but their formation, durability and effectiveness hinge on the underlying societal values. However, future research needs to further explore the mediating effect of institutions and policies on the relationship between values and social cohesion. In choosing welfare policies, we pick a particular type of policies. This raises the general question whether the mediating effect also holds true for other types. In addition, we do not explore whether similar relationships exist for other attributes of social cohesion such as societally inclusive identities and cooperation for the common good. Another direction of future research should be to more specifically parse out the theoretical and empirical relation of the direct and indirect effect of values on social cohesion.

What is social cohesion?
Social cohesion is notoriously hard to define. However, a consensus on its core characteristics has crystallised in the social science and policy literature in recent years. In broad strokes, this consensus conceptualises social cohesion as a latent characteristic of a social entity, such as a group or society, that emerges from the micro, meso and macro level, and that is constituted of a number of dimensions (e.g. Schiefer & van der Noll, 2016). Narrowing down which dimensions exactly constitute social cohesion, is currently a central point of debate in the literature. Despite still being up for debate, three lowest common denominators emerge.
First, at its minimum core social cohesion provides for emotional connectedness, or a "sense of belonging" (Chan, To, & Chan, 2006, p. 290), by defining who is part of the social community (Jenson, 1998, p. 4). A constitutive part of social cohesion thus is community or society membership. This takes both the form in which it is perceived at the micro level by individuals, but also the form in which membership is negotiated at the macro level within the community or society (Friedkin, 2004).
Most social scientists go beyond this by arguing that, second, individuals in a socially cohesive society need to have positive and robust interpersonal ties. Ties that are characterised by mutual trust, tolerance and reciprocity are usually conceptualised as being an inherent part of social cohesion (Chan et al., 2006;Friedkin, 2004;Green, Janmaat, & Han, 2009;Langer, Stewart, Smedts, & Demarest, 2017;Schiefer & van der Noll, 2016).
Third, a number of social scientists emphasise that social cohesion is marked by the notion that society and community are more than just the sum of the individual members. According to this view, a sense of constituting a larger social collective needs to prevail. For social cohesion to emerge and endure, individuals thus need to orient their actions and behaviour towards the common good of the society (Green et al., 2009;Langer et al., 2017;Schiefer & van der Noll, 2016).
These three most common aspects of social cohesion -belonging, positive interpersonal ties and an orientation towards the common good -also find their way into the communities of policy-makers, although policy documents on social cohesion usually include a number of additional aspects, such as economic inclusion, equality or political participation (e.g. Dragolov, Ignácz, Lorenz, Delhey, & Boehnke, 2013;Lefko-Everett, 2016).
Our conceptualisation builds upon Chan et al. (2006). However, we choose to extend it in order to incorporate the notion of orientation towards the common good. Accordingly, we define social cohesion as Relations which are the glue to hold society together. They comprise the vertical and horizontal relations among members of society and the state as characterised by a set of attitudes and norms that includes trust, an inclusive identity and cooperation for the common good.
This definition incorporates the consensus towards which the social science literature on social cohesion has converged so far. However, to bring in the vertical aspect of social cohesion more explicitly, we include the relationship between individuals and the state. In our conceptualisation, trust in the form of generalised social trust towards "people outside one's familiar or kinship circles" (Mattes & Moreno, 2018, p. 1) and trust towards state institutions signifies the need for positive ties within society. An inclusive identity mirrors the need for individuals to have a positive national identification with their society if the society is to be socially cohesive. Finally, cooperation for the common good is necessary for the emergence and endurance of social cohesion as something that exists not within individuals, but as a societal phenomenon among them.
As stated previously, no consensus exists yet on the different dimensions of social cohesion.
In the past, many social scientists chose to include values as an additional dimension of social cohesion (Jenson, 1998;Kearns & Forrest, 2000;Maxwell, 1996). However, this has changed in the recent literature for two interrelated reasons.
First, as Chan et al. (2006, pp. 292-293) point out, it is not logically true that a certain set of values, liberal values for instance, or a high degree of shared values per se lead to social cohesion. In contrast, certain values may not lead to cohesion but rather to conflict. At least intuitively, values such as 'achievement' or 'power' might pit people against one another rather than foster harmony. Additionally, human societies tend to evolve a variety of values, each of which limits the scope of the other. For instance, the conflict between the values 'achievement' and 'equality' gives rise to the distinction between 'equality of outcome' and 'equality of opportunity' that is prominent in different schools of liberal thought (Morris, 2015, p. 125). Value diversity thus must constrain the strength of commitment to values, as otherwise it would lead to conflict within society. This suggests the idea that only a 'loose' commitment to values rather than general value consensus might foster social cohesion (Mann, 1970). Overall, these examples illustrate that the relationship between values and social cohesion is more complex than truisms along the lines of 'when A, then B' would suggest.
Second, it follows that the relationship between values and social cohesion is not a constituent one, but a relationship of cause and effect. Chan et al. (2006) and Schiefer and van der Noll (2016) both stress that so far, all statements on the relationship between values and social cohesion are still in the state of being hypotheses rather than empirically grounded findings. To really understand this relationship, it is important to treat values and social cohesion as conceptually distinct in order to investigate the effect of either one on the other. Note that not only may values affect the state of social cohesion in a society, but social cohesion itself may also feed back on values and influence value consensus and value change in a society. For now, we will focus on the influence of values on social cohesion. There are several ways how this relationship may play out.
First, there might be particular sets of values that are conducive to social cohesion. Chan et al. (2006) and Schiefer and van der Noll (2016), however, are unanimous in that this is unlikely and that in fact, different sets of values are conducive to social cohesion under different socio-economic conditions. For instance, while the values of 'hierarchy' and 'respect for tradition' might have been conducive to maintain social cohesion in 18th century agricultural societies in Europe, this might not apply to industrial societies (Chan et al., 2006). Morris (2015) provides a deeper historical narrative of a similar argument.
Second, as Schiefer and van der Noll (2016) discuss, apart from the content of specific value sets, the extent to which there is consensus on the importance of a specific set of values within a society could possibly influence social cohesion positively or negatively. However, value dissent does not necessarily impede social cohesion if either the general commitment to values, that is, the degree to which one feels morally obligated and emotionally impelled to defend a value, is low and value dissent hence not salient, or if dissent is peacefully resolved. We might add, that the causal direction of the relationship between values, value consensus and social cohesion is also not clear yet, as social cohesion itself may create, or at least reinforce, certain sets of values and/or a value consensus.
For these reasons, we aim to contribute to the empirical exploration of the relation between values and social cohesion and, just like the literature on social cohesion, leave values out of our conceptualisation of social cohesion. As outlined above, the relationship between values and social cohesion is more complex than hitherto presented in much of the social science literature. To start investigating this relationship, we first need to grasp an understanding of what values are. The next section surveys the current state of the literature on this question before proceeding to more in-depth hypotheses of how values relate to social cohesion.

2.2
What are values?
The general predicament in defining and describing values is that they can never be observed directly but only inferred indirectly (Halman, 2007). The literature agrees that values are "stable meaning-producing superordinate cognitive structures" (Rohan, 2000, p. 257). That is to say that they are cognitively represented and stored in some form in the minds of people. They are nonetheless expressed only in an evaluative action, be that actual behaviour or a statement. This renders the concept of values elusive.
Most authors agree that a form of desirability is the core defining feature of values. Hence, values express what actions (van Deth & Scarborough, 1998), or what end states and behaviours, are worth seeking and doing. This in itself does not sufficiently separate values from other cognitive-evaluative concepts, such as attitudes. Although a commonly agreedon definition is missing, the psychological literature describes an attitude as "a summary evaluation of a psychological object captured in such attribute dimensions as good-bad, harmful-beneficial, pleasant-unpleasant and likeable-dislikeable" (Ajzen, 2001, p. 28). An important distinction between attitudes and values is that attitudes are an evaluative reference to a specific 'attitude object' which can be either a person, place, thing or event (Fishbein, 1963). While an attitude refers narrowly and closely to a particular object, a value is broader and more general in its application (Schwartz, 2006). Hence, values may inform attitudes and attitudes may have the function to express general values in the specificity of a particular situation (Ajzen, 2001). Therefore, scholars agree that besides specifying the desirable, values are also inherently trans-situational (Rohan, 2000). They apply across a number of different life situations as "abstract principles with which action is to conform" (van Deth & Scarborough, 1998). Being trans-situational differentiates them from attitudes and preferences that are bound to specific situations. As such, values are more basic than, and often provide the foundation for, attitudes and preferences (Rokeach, 1968, p. 124, as cited in Halman, 2007. Since the seminal work of the social psychologist Milton Rokeach, there is also consensus that individuals can hold different values at the same time in a hierarchically ordered cognitive value structure, which may sometimes give rise to value ambivalence (Jacoby, 2006).
Although they share many commonalities, there are two approaches to measuring values across social science disciplines. They differ substantially only regarding the conceptual terms employed in items measuring respondents' values. However, this choice in wording can create a crucial difference (Sokolov, 2018). First, the literature on cross-cultural and social psychology predominantly follows the measurement technique as pioneered by Rokeach (Rohan, 2000). Values are measured by stating an abstract, trans-situational term that represents a value, adding a brief description for clarification, and then asking the respondent to indicate how important this value is as a guiding principle in their life. In order to minimise the abstractness inherent in measuring values like this, Schwartz (2011) modified Rokeach's technique and constructed written portraits of persons in line with different values. Respondents were then asked to indicate how much they resemble the portrayed persons on a Likert scale. The portrait for the value achievement, for instance, reads as "It is very important to her to show her abilities. She wants people to admire what she does" (Schwartz, 2011, p. 191). Note that despite specifying the value through a description, it remains an abstract, trans-situational notion. This contrasts with the second value measurement approach which is often used in the political science and sociology literature. Here, respondents are commonly asked to indicate their agreement with a specific, less abstract and less trans-situational statement. One item for the widely used Emancipative Values Index 1 , for instance, reads as "When jobs are scarce, men should have more right to a job than women" (Welzel, 2013). The content validity of such items, in other words to what extent they measure the theoretical concept they are intended to measure, is debatable. If values are defined as abstract notions of trans-situational, desirable actions and end-states, it may well be argued that this item is too specific in describing a particular situation. Although attitudes may serve the function of expressing and applying values with reference to specific 'attitude objects', such as persons, places, things or events, they may still build upon multiple values. Inferring values inductively from a survey battery of attitudinal items risks arriving at loose and not clearly defined or distinct values (Schwartz, 2006, p. 144).
Regardless of the wording of items, the construction of value indices from a number of single survey items follows a general common approach. Data from various items are bundled into broader value categories via statistical techniques, such as factor analysis or multidimensional scaling. While this statistical approach is in accordance with the theoretical conceptualisation of values, it also inflates the number of conceptual value variants that circulate in the literature. This inflation has led to conceptual stretching (Sartori, 1970) in which sometimes distinct, sometimes overlapping, sometimes more abstract, sometimes more specific value concepts are occasionally specified under the same or similar names within different social science disciplines.
Nevertheless, a major fault line exists between those values that belong to individuals and those that are located at the societal level. To provide some orientation in this "forest of values" (Haidt & Kesebir, 2007), this section gives an overview of different value concepts that are used in the literature (see Table 1 for an overview).

Individual values
This section provides an overview of different individual-level value concepts and theories. It starts with less well-established and less clearly conceptualised frameworks, and proceeds towards more prominent and well-developed theories. Individual-level values are those values which are thought to be latent characteristics exclusively applicable to individuals.
Political and democratic values are an example for particularly fuzzy conceptualisations, despite their prominence in the research literature. They are usually simply conceptualised as any values with content pertaining to political questions and issues (e.g. Halman, 2007;Sleat, 2016). Among them, democratic values are of special interest in the political science Fuchs (2007), however, argues that democratic values are distinct from attitudes that provide support for a particular type of democracy. Sack (2017) follows him on this but acknowledges that he cannot measure democratic values. He therefore argues to measure "value orientations" based on attitudinal data, hence, individual orientations towards democratic values. But the value orientations he measures still only differentiate between support for two different styles of democracy, the liberal and the socialist model of democracy. Hence, his measurement is still closer to what Fuchs (2007) argues are attitudes in support of different democratic regime types rather than democratic values per se. Thomassen (1998Thomassen ( , 2007) constitutes a similar case; he claims to measure support for democratic values but ultimately his measurement is closer to attitudes towards the role of state intervention and welfare institutions in democracies. The general conundrum that values cannot be measured directly -but only indirectly through attitudes -certainly complicates the issue. However, the bottom line of our critique is that the literature investigates attitudes towards varieties of democracies rather than democratic values themselves. Consequently, the concept of democratic values remains elusive as long as we do not compare it with its informative antagonist. To really grasp what constitutes democratic values, we need to contrast them with their non-democratic equivalent: autocratic values 2 .
Besides democratic values, the research literature has advanced additional kinds of political values. Breidahl, Holtug and Kongshøj (2018) for instance, investigate "community values". According to them, these are values concerning who legitimately belongs to the state community. To measure such values, they extract four indices from survey data on attitudes about citizenship and cultural heterogeneity. As a result, their value indices provide information on attitudinal dimensions towards what constitutes a national identity, such as conservative nationalism or multiculturalism. Similar to democratic values, the question arises again whether what they measure fulfils the criteria for values, specifically whether it denotes notions about some desirable and trans-situational behaviour or end-state that is latent to attitudes, or whether it is still closer to an attitudinal index that simply aggregates attitudes according to topic.
A comparable criticism, albeit to a lesser degree, can be extended to the concept of 'core political values'. This concept originated in the social and political psychology literature and denotes "overarching normative principles and belief assumptions about government, citizenship, and society" 3 (McCann, 1997). Besides having a strong focus on American 2 To date, there is no such concept as autocratic values. The next closest thing is provided with the concepts of right wing authoritarianism (RWA) and social dominance orientation (SDO). These concepts emerged in psychological research after World War Two inspired by Adorno's concept of the authoritarian personality. Right wing authoritarianism broadly captures readiness of submission, while social dominance orientation largely describes an attitudinal preferences for either equal or hierarchical intergroup relations (Caprara & Vecchione, 2013).
politics, a major deficit of this concept is that its delineation to political attitudes is blurry. Researchers often use the terms 'core political attitudes' or 'beliefs' interchangeably (Caprara & Vecchione, 2013, p. 16). Core political values are also often conceptually conflated with attitudes and beliefs. Attitudes on more specific questions are frequently conceptualised as core political values together with less specific and more abstract notions. According to Goren (2005), for instance, both the more abstract traditional family values and moral tolerance as well as the more specific attitudes towards limited government and equal opportunity fall under core political values. In his operationalisation of core political values, Feldman (1988) also includes attitudes towards free enterprise in his operationalisation of core political values, just like Schwartz, Caprara, and Vecchione (2010) who also include blind patriotism. In contrast, McCann (1997) sticks with the more abstract concepts of egalitarianism and moral traditionalism.
However, despite the blurry delineation between core political values on the one hand, and beliefs and attitudes on the other, there is some general notion in the social and political psychology literature, that core political values are one layer of less foundational and more flexible values that are informed by underlying personal values (Caprara & Vecchione, 2013;Feldman, 2013).
One likely candidate for more foundational personal values that inform core political values is given by Schwartz' influential and prominent theory of basic values (Schwartz, 1992;. Based on considering what goals lie in the biological and social nature of human beings, Schwartz deductively conceptualised ten basic personal values that are supposed to be universal among all people (Schwartz, 2011). To test this, he developed two cross-cultural value surveys, the Schwartz Value Survey and the Portrait Values Questionnaire. Based on these data sets, Schwartz explored the complementary and opposing relationships between the ten basic values, and showed that they order along two dimensions. The first dimension contrasts values of openness, such as stimulation and hedonism, against values of conservation, such as security and conformity; the second dimension contrasts values that put the focus on benefits for oneself, such as power and achievement, with values that put the focus on the well-being of others, such as universalism and benevolence. An important limitation of conceptualising values deductively instead of inferring them inductively from attitudinal items, is that it might miss values that exist, but were not conceptually deduced beforehand.
Compared to the concepts of political values, and core political values, Schwartz' basic personal values are much more trans-situational and general, often described by a highly abstract notion, such as 'achievement', 'conformity' or 'universalism' (see Table 1). Hence, they are generally thought to underlie behaviour, attitudes, beliefs and less foundational values, such as core political values (Feldman, 2013). Caprara, Schwartz, Capanna, Vecchione and Barbaranelli (2006) and , for instance, find that the impact of Schwartz' basic personal values on voting behaviour was fully mediated by core political values. Similarly, Caprara et al. (2017) find that basic personal values predict to which political ideology people adhere.
Inglehart and Welzel (2005)  For instance, tolerance is a self-expression value, while security is a survival value. The second dimension is between secular-rational values, such as separation of religion and authority, and traditional values, such as deference to authority, as its dimensional opposite. Inglehart and Welzel's theory of values and value change is complemented by Inglehart's (1971) (Inglehart, 2007;. According to Inglehart and Welzel, another important development in this broader survival/self-expression and traditional/secular-rational value change is a shift towards emancipative values. Emancipative values put a strong emphasis on the decisionmaking freedom of individuals and correlate strongly with democracy (Welzel, Inglehart, & Deutsch, 2005). According to their theory, the shift to emancipative values thus is an important predictor and stabiliser for democratisation and democracy (Welzel, 2007(Welzel, , 2014. Despite the centrality of values and value change, Inglehart and Welzel's theory nevertheless is a revitalisation of modernisation theory, which proposes socio-economic change to be a major driver of political institutional change. They argue that "economic development tends to transform people's basic values" (2005, p. 5) from which political institutional change then follows. They hence leave little room for endogenous value change. Scholars of other disciplines have picked up this theory. Morris (2015), for instance, provides a historical narrative from a point of view of modernisation theory. According to his narrative, the entire human history can be broadly characterised by the three distinct value sets of hunter-gatherer values, farmer values and values of fossil fuel-based societies. Which values prevail at what time largely depends on the socio-economic conditions within each historical period.
The research literature largely agrees on the conceptualisation and definition of personal values. However, the practice of measuring personal values might still not meet this conceptualisation. Some researchers make heavy use of a limited number of attitudinal items in which the 'attitudinal object' often is a specifically described situation that might not differ much from item to item. These attitudes are then inductively aggregated into values. This approach comes with the risk of creating index scores of similar or related attitudes rather than measuring values. Other researchers present respondents with deductively conceptualised abstract notions and ask them whether these notions serve as a guiding principles in their life. This approach minimises the risk of creating attitudinal indices rather than measuring values, but it might not measure existing values because they were not deductively conceptualised beforehand. Value researchers need to find a way which minimises both risks and reconciles the advantages of both approaches.

Societal values
It is a straightforward and a consensual assumption within value research that personal values are cognitive properties of individual people. However, most researchers agree that values exist conceptually not only on the individual micro level, but that groups and societies can also be characterised by specific value sets. The nature of such societal values is a more complicated issue in which the literature has not yet settled on a common conceptualisation.
Early work in value research did not explicitly differentiate between values at the individual and societal level. While social psychologists early on clearly focused on values at the individual micro level, researchers in anthropology focused more on societal values that made human communities distinct. As different disciplinary research communities did not explicitly address levels of conceptualisation for decades, this led to some confusion within the literature (Rohan, 2000;Schwartz, 2011). However, more recent work has begun to clear up this confusion.
One of the most prominent conceptualisations of society-level values stems from Hofstede (2003) and his theory of cultural dimensions. Being a researcher with the IBM corporation, Hofstede developed and conducted a survey among IBM employees in a number of different countries. From his data, he distilled six different value dimensions that characterise cultures around the world (see Table 1). The most prominent of the six value dimensions is probably the individualism-collectivism dimension. It broadly distinguishes between societies where an orientation towards oneself and an emphasis on rights over duties dominates, versus societies in which a 'We'-orientation and conformity to group norms and the common good prevail.
Schwartz (2014a) developed another influential and commonly used conceptualisation of societal values. Based on his theory of personal basic values, Schwartz expounded his cultural values theory that argues that societies around the world can be distinguished by their relative emphasis on one of seven cultural value dimensions (Schwartz, 2006). Just as in his theory of basic personal values, Schwartz' cultural values are ordered in a dimensional spectrum in which two opposing values roughly characterise the dimension. The masteryharmony dimension, for instance, is defined by the opposition of the following values: Mastery describes the self-assertion to be able to direct and change one's social and natural environment; the value of harmony in contrast emphasises being and staying part of the world as it is without changing it. A salient and much researched dimension in Schwartz' cultural values concerns the embeddedness-autonomy dimension in which autonomy is subdivided into affective and intellectual autonomy. Embeddedness describes a worldview that predominantly sees and treats people as belonging to a human collective, while autonomy refers to a worldview that rather perceives people as bounded entities that emphasise their own preferences and opinions. In its meaning and conceptualisation, Schwartz' embeddedness-autonomy dimension is very close to Hofstede's collectivismindividualism dimension.
How to conceptualise societal values is an important question. This is still a point of debate in the literature (Schwartz, 2014b). According to Hofstede's (2003, p. 9) conceptualisation, societal-level values are the "collective programming of the mind". He thus implies that societal-level data somehow also resides within individual minds and can be inferred from individual survey results. However, he does not explicitly expound on whether he thinks that societal-level data are internal or external to individuals. Schwartz, in contrast, argues that culture is "the latent, normative value system, external to the individual, which underlies and justifies the functioning of societal institutions" (Schwartz, 2014b, p. 6). He argues that, as the underlying value system shapes and gives rise to societal institutions, individual society members are 'imprinted' by the cultural values of their society. Accordingly, cultural values can be deduced as an aggregate from survey data on individual society members. The large variation within a given society that such survey data reveal is just a residual to these aggregate cultural values caused by differences in personal basic values (Fischer & Schwartz, 2011).
The conceptualisation of societal values is important because it is intricately linked to their measurement. Most conceptualisations of societal values assume that societal values can be derived from individual-level data. Both Schwartz and Hofstede use survey data to measure societal values. Rohan (2000), however, stresses that the operationalisation and measurement of societal-level data is still an open question. He suggests that they might also be measured through analysing the presentation of ideal and prototypes. Content analysis, for example, could measure under which headings and using which key words issues of public debate are discussed within the media to infer the societal ideals and goals, and hence, ultimately, the values of a society. Even though both Hofstede's and Schwartz' conceptualisations are close enough to each other to justify measuring societal-level values through survey data, at least Schwartz' conceptualisation also justifies measuring societallevel data through content or discourse analysis as suggested by Rohan (2000). To the extent that public discourse and similar societal-level processes are what Hofstede implies with the "programming of the mind", his conceptualisation does so as well.

Universality of values
Are the values identified by the literature universal for all people across time and space? Almost all value theories presented so far assume this. Discussing this assumption is highly relevant for two reasons. First, by building upon an assumption of universality, value theories rule out endogenous cultural change, that is to say self-motivated cultural change. This means in consequence that people and human societies cannot evolve new values. The plausibility of this conclusion is questionable. Second, it is well known that social science has suffered from a euro-centric bias in the past, and universalism is a particular form of eurocentrism in the social sciences (Wallerstein, 1997). Any claim to universality thus needs to be evaluated carefully. But as Rohan (2000) points out, value research rarely makes this assumption explicit, let alone discusses it.
The following paragraphs investigate whether the universality assumption is justified. They start by exploring neuroscientific foundations of how values are represented in the mind, and how the capacity for having values probably is a universal product of human evolution. They then show that most value theories assume, contrary to this, that the actual content of values, not only the capacity for having values, are universal. Finally, we discuss why this assumption might be wrong, and how it might constrain future research.
Although often perceived as abstract, research on how values and general morality are cognitively represented in the mind suggests that they are not entirely part of pure abstract reasoning. Recent neuroscientific research has investigated the areas of the brain where neurons are activated when individuals reflect on values. These studies have shown that value reflection is connected to neuronal activities between and within the superior temporal lobe (a region connected to abstract reasoning), the prefrontal cortex (connected to action planning), but also to basal forebrain regions, such as the hypothalamus, which are responsible for generating emotions (Zahn et al., 2008). This indicates that values are not purely abstract cognitive concepts but in some way have an emotional association in the human brain. Other studies additionally found that value reflection stimulates areas of the brain that activate during economic reasoning, such as when people compare prices (Brosch & Sander, 2013).
Overall, this suggests that the capacity for cognitive evaluation is an essential product of human evolution and that at least this capacity must be universal among people (Evers & Changeux, 2016). As our closest living relatives, the great apes, also exhibit forms of morality, it is likely that human morality evolved for seven to eight million years since the split from the last common ancestor shared by humans and great apes (Langergraber et al., 2012;Tomasello & Vaish, 2013). Based on a review of morality research, Haidt and Joseph (2004) put forth the argument that during this time, humans evolved innate "moral intuitions" that form the basis of human values 4 . They suggest that these moral intuitions are universal for all people, but leave sufficient space for the emergence of differing value variants across and within human societies.
This is contrary to the conceptualisations of individual and societal values introduced here so far, almost all of which assume universality of values across human societies. Schwartz, for instance, argues that both his conceptualisation of basic and of cultural values are based on universal problems that human societies have always had and always will have to deal with. Hence, he concludes that the resulting types and dimensions of values are also universal, although different individuals and different societies may emphasise them differently. Even though their choices may be different and distinct, they all choose from the same never changing set of universal values. Schwartz formulates it most explicitly in his value theory, but almost all other value concepts imply the same (e.g. Hofstede, 2003;. Many of the values and value dimensions identified by the literature do indeed overlap or are at least similar to one another. The conceptual similarity between Hofstede's individualism-collectivism dimension and Schwartz' autonomy-embeddedness dimension has already been noted. Inglehart (2018, pp. 50-51) argues that there even is a value superdimension. Putting forth strong correlations between his and Welzel's measure of survival and self-expression values, Hofstede's individualism-collectivism value dimension, and Schwartz' autonomy-embeddedness dimension, he argues that they "all tap a common underlying dimension". This super-dimension reflects "variation to the extent to which given societies allow people a narrow or broad range of free choice" (Inglehart, 2018, p. 52). Because all three dimensions were conceptualised and measured separately, this could indicate their universality across human societies, but it could also indicate that values simply converge across societies.
Additional research supports the idea that certain values are universal to all people. Research by Fischer and Schwartz (2011) shows that variation in values within countries is larger than between countries. This suggests that the effect of societal culture on individual values is weaker than thought so far, and that the individual value set conceptualised by Schwartz is at least to some degree and for some values universal.
Despite these research findings, the assumption of value universality is open to debate. Methodologically, it is doubtful to assume that the values that research has identified so far are universal across all human societies. First, the results are all based on data, in which the scope for responses of respondents was conceptually limited beforehand. Respondents can only choose from the value items that the surveys offer as responses, ipso facto they could not provide alternative values if they wanted to. Second, universality implies that the understanding of the survey items used to measure values does not vary between societies.
Only conditional on this assumption would value surveys measure the same concept in different societies, a property which is called measurement equivalence in cross-national research (Davidov, Meuleman, Cieciuch, Schmidt, & Billiet, 2014). Recent studies show that this measurement equivalence is not given for many value theories proposed so far. The Portrait Value Questionnaire of Schwartz' basic values theory was reformed after such findings, and suggestions for changes are now also being proposed for Welzel's Index of Emancipative Values (Cieciuch, Davidov, Algesheimer, & Schmidt, 2018;Sokolov, 2018). However, the important implication is that the universality assumption, at least for some value concepts, does not hold in the face of a lack of measurement equivalence.
An additional important implication of the universality assumption is that it restricts the focus of future research. There are tendencies to include the notion of cultural evolution into the research of values, e.g. Inglehart (2018), Morris (2015) and Welzel (2014). Cultural evolution is the transmission of social facts, such as ideas, attitudes, beliefs, but also technologies, within and between generations (Richerson & Boyd, 2001). In short, it describes cultural change. The theoretical framework of cultural evolution might thus aid to explain value change. However, the idea of a universal pre-determined set of individual and societal values is irreconcilable with the dynamics of evolution.
For methodological and theoretical reasons, the assumption that certain values are universal thus needs to be critically investigated and tested. In addition, besides focusing on the aggregation of values into broader value indices and dimensions in order to subsume societies into the same, research needs to examine more of the cultural dynamics of values. More research is needed that investigates value change and the formation of new values within human societies.

2.3
How might values affect social cohesion? The three pathways As briefly outlined before, many theorists regard values as central to social cohesion. In discussing the role of values, it makes sense to differentiate between two distinct pathways: the intrinsic and the extrinsic pathway (Holtug, 2017). Based on the literature on values and institutions, we further suggest that the extrinsic pathway branches into a direct nonmediated effect, and an indirect effect in which institutions mediate the effect of values on social cohesion (see Figure 1). In this pathway, the content of a particular value, such as egalitarianism, is only of minor interest. Rather, it is of interest whether members of a society agree on and share the same values, that is, whether there is value consensus within a society. The premise behind this pathway is that "the real threat to the moral order is not ethnic diversity, it is moral diversity -a lack of consensus on basic values" (Haidt & Kesebir, 2007, p. 226). It is based on the idea that shared values forge a common identity, and, thus, "imagined communities" (Anderson, 1983;Holtug, 2017).
So far, empirical evidence for this positive effect of shared values is scarce and inconclusive. Concerning identity, qualitative analyses suggest that shared values at least play some role in discourses on national identity, but these studies seldom also shed light on their ultimate effect on social cohesion (Henderson & McEwen, 2005). Most quantitative studies on the effect of shared values focus on trust. A cross-national analysis undertaken by Beugelsdijk and Klasing (2016) finds that higher polarisation in terms of values is moderately associated with less generalised trust. However, the value measure they construct from the World Values Survey is based on a range of attitudinal survey items, concerning attitudes towards, for instance, government intervention and prostitution, which clearly presents not a transsituational context. It thus captures attitudes towards political ideology and so-called morally debateable behaviour rather than values as defined by Hofstede (2003), Rohan (2000) or Schwartz (2006). Contrary to their finding, other studies do not find the same moderately robust effect. Breidahl et al. (2018) explored whether the perception that others agree with their community values influences Danish respondents' trust levels. They only find a weak negative correlation that did not prove to be robust or consistent across different outcome measures. Given that Breidahl et al.'s (2018) results are inconclusive, and that Beugelsdijk and Klasing's (2016) findings concern attitudes rather than values, and the lack of research on the effect of shared values on cooperation and identity, more evidence is clearly needed.

The extrinsic pathway I: value content
In contrast to the intrinsic pathway of shared values, the content of a value plays a central role in the extrinsic pathway (Figure 1). In this pathway, the effect of values on social cohesion does not depend on whether the values are shared. Instead, the values are per se conducive to social cohesion because of their content (Holtug, 2017). An individual value of conformity or a societal value of hierarchy, for instance, might maintain a (minimal) degree of social cohesion by legitimising an unequal distribution of economic and political power. Findings from the research literature suggest that different value content can indeed have a direct extrinsic effect.
First, concerning trust as one attribute of social cohesion, studies show that the content of some values is especially conducive to higher social trust. Collectivism, for instance, puts a higher emphasis on one's in-group, thus it is associated with lower social trust while individualism generally relates to higher social trust levels (Allik & Realo, 2004;Van Hoorn, 2015). As discussed above, Welzel and Inglehart's self-expression values, and Schwartz' societal-level value of embeddedness correlate strongly with Hofstede's individualism-collectivism dimension. Thus, it is not surprising, that they also associate values with greater social trust (Schwartz, 2006;Welzel, 2010). In addition, Schwartz's (2006) results suggest that egalitarianism is associated with higher interpersonal trust, as egalitarianism socialises individuals to transcend their self-interest, thereby leading to the expectation that individuals care for one another. In a similar way, community values that stress multicultural diversity, notions of active citizenship and civic engagement correlate moderately positively with social trust, as well as trust towards in-group members and immigrants. In contrast, nationalist community values that stress a common shared culture are linked to less trust towards immigrants, while the more conservative version of nationalist community values associate weakly with higher in-group trust (Breidahl et al., 2018). Turning from social trust to trust towards institutions, studies show that people holding values of tradition, conformity and security have more institutional trust, while individuals for whom the value of autonomy dominates tend to mistrust institutions (Devos, Spini, & Schwartz, 2002;Morselli, Spini, & Devos, 2012). In sum, values that promote an individualistic lifestyle and values that promote civic engagement correlate positively with higher generalised social trust, negatively with institutional trust, while it is the other way around for values that promote a lifestyle emphasised by security and one's role and position in society.
Second, concerning cooperation for the common good as an attribute of social cohesion, study findings seem to depend strongly on the level of analysis. While some studies focus on cooperation in group exercises, other studies investigate aggregate measures of cooperation, such as civic engagement, and results differ accordingly. Early research showed that in meso-level group exercises (e.g. Wagner III, 1995), members of collectivist societies tend to cooperate more than those of individualist societies. A meta-analysis carried out by Marcus and Le (2013) verifies this finding for cooperation at the work place.
However, the same meta-analysis shows that cooperation in group exercises does not correlate meaningfully with individualism-collectivism as a societal phenomenon on the macro level. Turning from cooperation in group exercises to measures that capture cooperation for the societal common good, such as civic engagement, research finds an almost diametrically opposite relationship. Communities and societies that are characterised by greater individualism also show more political and civic engagement (Allik & Realo, 2004). A positive relation also appears between peaceful political collective action, such as participating in protests, and self-expression values (Welzel, 2010). Schwartz reports a number of findings concerning basic and cultural values as well as cooperation among individuals (Schwartz, 2006). Among other relationships, egalitarianism correlates positively with engaging in social activities, embeddedness associates negatively with membership in voluntary organisations and autonomy correlates positively with political activism, mirroring the relation found by Welzel (2010). In sum, these findings suggest that the super-dimension of individualism-self-expression-autonomy values (Inglehart, 2018) correlates positively with cooperation for the common good at a societal level, while the effect of values on cooperation within a group as part of group exercises is more complex.
Third, concerning an inclusive identity, the effect of values is probably most understudied when compared to other attributes of social cohesion. Nevertheless, within the social and cross-cultural psychology literature, some studies link Schwartz' theory of individual-level basic values to theories of identity formation and change, such as Social Identity Complexity (Roccas & Brewer, 2002) and Identity Process Theory (Bardi, Jaspal, Polek, & Schwartz, 2014). These studies consistently report that basic personal values are important in identity formation and perception of what threatens identity. First, individuals that embrace the values of conservation as opposed to openness to change tend to form less complex identities (Roccas & Brewer, 2002). Such identities are less tolerant towards ambivalent group memberships and flexible group boundaries. Hence, individuals that value conservation tend to form identities in which their perception of group memberships overlaps entirely and exclusively: For instance, being German automatically means being Christian, and anyone who is not a Christian, cannot really be a German -or is at least somehow 'less German'. The values of conservation in contrast to values emphasising openness to change are also associated with a stronger national identification (Roccas, Schwartz, & Amit, 2010). However, as Roccas and Brewer (2002) hypothesise, this probably comes at the cost of greater identity complexity and inclusiveness. Finally, individuals that hold conservation values are more prone to perceive their identity as being under threat and fear losing it when facing changes in their life (Bardi et al., 2014). In sum, the evidence that already exists suggests that values of conservation, such as security, conformity and tradition, create narrow and exclusive national identities, while values such as hedonism, stimulation and self-direction correlate positively with broader and inclusive national identities.

The extrinsic pathway II: value content mediated
Some of the extrinsic influence of institutions on social cohesion might be indirect. It is plausible to theorise that institutions and the policies they produce mediate this part ( Figure  1). This is indeed a corner stone of Schwartz' theory of values, according to which "national differences on cultural value dimensions may influence both national policies and the change or maintenance of political, economic and demographic conditions in a society" German Development Institute / Deutsches Institut für Entwicklungspolitik (DIE) (Schwartz, 2011). In his framework, cultural values are products of history and ecology, and influence societal institutions such as political and legal systems, economic markets, or family and kinship structures. Through their expression in institutions and policies, crossnational value differences are reproduced in institutions (Schwartz, 2011, p. 8). Institutions and policies, in turn, shape individuals' basic values through socialisation which then feed back on institutions, since institutions are "the humanly devised constraints that shape human interaction" (North, 1990, p. 3). In this way, cultural values, basic values, institutions and individuals are connected in a circle of constant emergence and change according to Schwartz' framework.
Evidence supports the importance of values for institutions. Using a simple path analysis, Schwartz (2006) shows that the cultural value dimensions of embeddedness-autonomy and hierarchy-egalitarianism correlate significantly cross-nationally with levels of democracy. More sophisticated statistical analyses suggest that societal-level values indeed influence institutions. Licht, Goldschmidt and Schwartz (2005) provide evidence that a cultural value of harmony lessens the tendency to solve economic conflicts through litigation as part of national corporate governance. Similarly, Licht, Goldschmidt and Schwartz (2007) find that a value of embeddedness negatively influences the extent of rule of law, non-corruption and democratic accountability. Both studies employ instrumental variables to exclude reversed causation.
The relatedness between Schwartz' value of embeddedness and Hofstede's value of collectivism has been noted earlier. The fact that there are similar results for collectivism thus further corroborates these findings. Gorodnichenko and Roland (2015) construct a formal model according to which a societal-level value of individualism creates greater chances for the maintenance of democratic institutions, while a societal-level value of collectivism increases the likelihood for coups, but at the same time also increases the likelihood for failure of regime transition from autocracy to democracy. Employing a different instrumental variable than Licht et al. (2007), they test their model empirically and find that collectivism indeed decreases the likelihood for a democratic transition. Gorodnichenko and Roland's (2015) work belongs to the particular strand of literature generated by New Institutional Economics. Initially demonstrating the influence of institutions on economic growth and development often with a focus on long-term history, the New Institutional Economics literature now also explores and emphasises the role of culture in the form of social norms, informal institutions, values, ideas and beliefs as antecedents of institutions (e.g. Acemoğlu & Robinson, 2016). Leading scholars of this strand of literature also point to the importance of cultural evolution as endogenous cultural change (e.g. Mokyr, 2014;North, 2005).
If institutions indeed mediated the effect of values on social cohesion, they would also need to correlate with social cohesion. There is evidence for this especially concerning trust. Institutions, such as stable property rights, strong welfare state institutions and good governance as operationalised with the World Governance Indicators as well as strong government capacity are positively associated with trust (Hutchison & Johnson, 2011;Robbins, 2011Robbins, , 2012Tsai, Laczko, & Bjørnskov, 2011). In contrast, the direction of the association between trust and democracy, as a particular type of political institution, is not yet fully clear. Older studies identified democratic institutions as important predictors of trust (Brehm & Rahn, 1997;Rothstein & Uslaner, 2005;Simpson, 1990). But according to more recent studies, democracy levels are either not clearly (Robbins, 2011) or negatively associated with both social (Robbins, 2012;Tsai et al., 2011;Van der Meer, 2017) and political trust (Hutchison & Johnson, 2011). Bjørnskov (2006) also does not find support for the idea that democracy is related to higher trust levels, and suggests that it is not democracy that facilitates trust but trust that facilitates democracy. Another strand of research has focused on electoral institutions and consequently the function of elections in creating trust (Listhaug & Jakobsen, 2018;Van der Meer, 2017). Yet, the impact of electoral institutions on trust remains unclear: Some studies find that trust is higher in countries with more proportional systems, whereas others support the opposing view that proportionality lowers trust. Overall, there is no consistent evidence on the relationship between democracy and trust. The variety of institutional settings subsumed under the heading of democracy is probably too large to show a clear picture. However, country cases in which exclusionary institutions marginalise large segments of the population, and thus cause civil conflict (Dixon, 2009), suggest that generally inclusive institutions are important for social cohesion.
Taken together, the literature on both values and institutions suggests that values affect social cohesion not only directly but also indirectly through institutions and the policies they generate. This premise is important for two reasons. First, it suggests that social cohesion can be generated and maintained if the right institutions are in place. Second, it proposes that values play a central role in influencing and explaining institutional change. In the next section, we turn towards empirically exploring the links between values, institutions and social cohesion. Theory and previous findings suggest that institutions and policies mediate the effect of values on social cohesion. In this section, we present a first preliminary exploration of this triangular relationship. This exploration is limited in focus and scope. First, of those effects that values may have on social cohesion, we choose to focus on the extrinsic direct and indirect effect, and leave aside the intrinsic effect of shared values (Figure 1). Second, to keep the empirical exploration in scope, we choose to focus on only one dimension of social cohesion, namely trust. We choose trust, as this is one attribute of social cohesion for which the effect of institutions is well studied and robust (see above). Finally, as we argue further below, we limit our focus on the particularly relevant nexus of one type of values, namely egalitarian values, with one particular policy regime, namely welfare policies. Extending and deepening our exploration will be a task for future efforts.
We define welfare policies as policies that aim at providing a country's population with a secured minimum living standard over a range of different life stages and conditions. A welfare state regime, or, interchangeably, welfare policy regime, is the network of the social institutions that create welfare policies. A welfare policy regime encompasses, for instance, educational policies, health care schemes, retirement schemes as well as unemployment and poverty programmes. Today's differences in welfare policy regimes evolved historically, and differ primarily in terms of their scope, or degree of eligibility and inclusion (Esping-Andersen, 1990). The liberal type of welfare policy regimes is at one end of the spectrum. It evolved from the European poor-relief policies of the 19th century, and focuses on meanstested social support. In these welfare policy regimes, entitlement to a social service is strictly dependent on need, which is assessed according to fixed rules. Typical examples are the welfare policy regimes of the UK or Australia. At the other end of the spectrum lies the social democratic, or universalistic, regime type. It evolved as a response of socialist parties to the liberal welfare regime type at the turn of the 19th to the 20th century and builds upon the notion of universal eligibility (Esping-Andersen, 1990). Here, welfare policies that benefit the entire population without prior means-testing dominate. A typical example of such policies are universal child care policies (Rothstein, 2008). The Scandinavian countries provide prominent examples for universalistic welfare regimes (Vrooman, 2012).
Welfare institutions are important factors for social cohesion in general, and for trust in particular. The literature is consistent on the effects of welfare institutions and points out how important they are in generating and maintaining social cohesion (Edlund & Lindh, 2017;Rothstein, 2001). According to the literature, both types of welfare institutions have distinct effects on trust in a society. Universalistic social welfare institutions are conducive to trust by treating everyone according to the same standards, while means-tested institutions risk increasing perceived social distinctions by emphasising inequality, thus undermining trust (Edlund & Lindh, 2015;Rothstein & Uslaner, 2005). First, concerning universalistic welfare policies, Brewer, Oh and Sharma (2014) show that a higher total social welfare expenditure, which is known to be correlated with a higher proportion of universalistic welfare policies, is associated with greater trust, even when controlling for reversed causality. Furthermore, exploiting the health care reform in the US as an event in the course of which the US welfare policy regime became more universalistic, Mewes and Giordano (2017) show that universalistic welfare policies mitigate factors that usually explain lower social trust. Second, compared to universalistic welfare policies, liberal means-tested welfare policies tend to emphasise social distinctions by stigmatising beneficiaries (Esping-Andersen, 1990;Stuber & Schlesinger, 2006). More recent research shows that means-tested welfare policies do not automatically lead to a "welfare stigma" (Moffitt, 1983), but are conditional on societal moralistic perceptions on who the 'deserving' and the 'undeserving poor' are (Calnitsky, 2016). Although the stigmatising effect of means-tested welfare policies is conditional in this way, they may nonetheless increase social tensions. In addition, means-tested welfare policies may increase a sense of social alienation among beneficiaries by violating their personal integrity through the necessary intrusive procedures involved in assessing their economic and social situations (Calnitsky, 2016;Rothstein, 2008).
Having characterised the general association between different types of welfare policy regimes and trust, the next question concerns which role values play for the scope of welfare policies. A minor strand of the literature on welfare policies has begun to focus on the role of cultural factors for the support of different policy regimes, but has not yet made the connection to value concepts from social psychology (Van Oorschot, 2007;Vrooman, 2012). Among these values, Schwartz' egalitarian values seem to be most in line with universalistic welfare policies. They "socialize people to internalize a commitment to cooperate, to feel concern for the welfare of all, and to act voluntarily to benefit others. Important values in such cultures include equality, social justice, responsibility, help, and honesty" (Schwartz, 2014a, p. 551). In theory, it should thus be expected that societies with a strong emphasis on egalitarian values are more likely to create and maintain universalistic welfare institutions. Indeed, individuals in countries where egalitarian as opposed to hierarchical values are emphasised, more strongly support redistribution through welfare institutions. The same effect holds for embeddedness in contrast to autonomy values. This suggests that individuals in egalitarian societies are more supportive of universal welfare institutions, while societies with an emphasis on embeddedness values are more supportive of means-tested institutions (Arikan & Bloom, 2015).
Taken together, although the evidence is still slim, theory and already extant empirics provide for the hypothesis that egalitarian values increase the likelihood that a society adopts a welfare policy regime in which universalistic programmes dominate, which in turn increases both social and institutional trust. Figure 2 illustrates the conjectured relationships in a path diagram. It should be carefully noted that the causal relationship between values, welfare policies and trust can run in either direction. Indeed, this is still an unsolved issue in the literature on welfare policies (e.g. Bergh & Bjørnskov, 2014). We discuss this issue further in more detail below. In the next sections, we explore whether the conjectured relationships between egalitarian values and universalistic welfare institutions on the one hand, and between universalistic welfare institutions and trust on the other hand hold. Lastly, we also assess whether a universal scope of welfare institutions takes on a mediating role in the relationship between egalitarian values and trust.

Figure 2: Conjectured relationship between egalitarian values, welfare institutions and trust
Source: Authors

Data and operationalisation
In order to explore empirically the triangular relationship between values, institutions and social cohesion, we drew on cross-sectional secondary data from large-scale surveys, indexes from expert ratings, as well as national statistics. To account for the macro-societal nature of some of the main concepts employed in this research (e.g. cultural values, welfare institutions), all individual data were aggregated at a country level. In total, data on all variables was available for 63 countries across all inhabited continents of the world. An overview of the countries can be found in the Appendix.
In our empirical exploration, we chose to examine two potentially important country-level antecedents of social cohesion: the countries' cultural value emphasis on egalitarianism and the ratio of universalistic to means-tested social welfare institutions. In order to keep confounding effects as small as possible, we included multiple control variables, including the pervasiveness of political corruption, level of democracy and economic development.
Political corruption and democracy have been found to be strong determinants of social and institutional trust (Rothstein & Uslaner, 2005) and were thus included as control variables. We also controlled for educational equality and freedom of discussion, as social trust is likely to be fostered in environments where information can flow freely (Fisman & Khanna, 1999), and where equality of opportunity in areas such as education prevails (Rothstein & Uslaner, 2005). Since ethnic heterogeneity might influence the sense of belonging to a larger social order (Licht et al., 2007), it was also integrated as a control variable. Lastly, we added economic development and government revenue to rule out differences between countries in terms of economic strength, which is usually associated with a more benign social policy framework and affects how extensively governments engage in distributive expenditures (Muinelo-Gallo & Roca-Sagalés, 2011) 5 . An overview of all variables and their measurement is provided in the Appendix.
We retrieved data on egalitarianism from the Schwartz Value Survey (SVS) 6 . In the SVS, participants are asked to rate the importance of different values as guiding principles for their lives on a Likert scale ranging from -1 (opposed to my values), 0 (not important) to 7 (of supreme importance). Importance ratings on six values (equality, social justice, loyalty, honesty, welfare and responsibility) are used for determining a country's cultural value emphasis on egalitarianism (Schwartz, 2014a). Since individuals and cultural groups might use the value scales differently, the individual-level raw scores were corrected for scale use before aggregation for the cross-country analyses. In this we followed the instructions provided in the SVS.
Recall that our mediating variable is the ratio of universalistic to means-tested social welfare policies. For this, we used the means-tested vs. universalistic policy indicator from the database Varieties of Democracy (Coppedge et al., 2019), which measures the quality of the welfare state on an interval scale ranging from 0 to 5. A score of 0 implies that there are no, or extremely limited, welfare state policies, whereas a score of 5 means that almost all welfare state policies are universal in character (V-Dem, 2019). More fine-scaled indicators on social welfare programmes are rare and would have necessitated a major data collection effort. Because of this and due to its broad coverage in terms of countries and time, this indicator was deemed most suitable for the purposes of our explorative analysis.
With regard to the outcome variables, social trust was measured by the binary 'mostpeople'-question ("Generally speaking, would you say that most people can be trusted or you can't be too careful in dealing with people?"). This question has been included as an item in all waves of the World Value Survey (WVS), as well as other regional surveys 7 . In their review of trust measurement, Bauer and Freitag (2018) conclude that the dichotomous most-people measure is still mainly used to assess social trust in spite of several shortcomings. These include limited scale length and interpersonal comparability, as people from different cultures may interpret the most-people question differently. Since more nuanced data are not available for a broad range of countries, we decided to stick with the binary most-people question to approximate countries' levels of social trust. Reversing the scores from the WVS and European Value Survey (EVS), ensured that for all countries a higher social trust score indicates that people trust each other more.
In order to gauge institutional trust at the country level, we calculated an additive individuallevel index based on the measurement of trust in the police, the courts and the parliament on a Likert scale from 1 to 4, which was then aggregated at the national level as a simple average. The scores of the WVS and EVS were reversed, so that for all countries a higher institutional trust index translates into more trust in the respective institutions.
Availability of the data posed a major challenge. The SVS data are available only until 2005, and the WVS and EVS often provided no data earlier than 2005. Hence, checking for the time difference between the SVS data and the data from the other surveys was especially important. Even though cultural values are thought to be quite stable (Schwartz, 2011), some authors found them to change inter-generationally (Inglehart & Abramson, 1994). Therefore, we compiled data only for those countries where the data collection of all surveys took place within a one-generation period of approximately 20 years, that is to say 10 years before and 10 years after the cultural value measurement. Countries to which this did not apply (Ethiopia, Indonesia, Singapore and Zimbabwe) were given an additional dummy variable in order to control for periods between data collection exceeding a one-generational time span. The analysis revealed that, in fact, the longer periods between data collection have no significant effect on the outcome variables social trust and institutional trust. Hence, there is no statistical support for excluding the four countries mentioned above from the analyses. They were thus kept in the sample for all analyses.

Methodology
For our quantitative exploration, we employed two different types of data analysis strategies: multiple regression and mediation analyses.
First, we used multiple regression analyses to indicate the strength and direction of the relationship between egalitarian values and the proportion of universalistic vs. means-tested social welfare programmes. Similarly, we ran regressions to test the relationships between the proportion of universalistic vs. means-tested social welfare programmes and social trust (see Figure 2: upper path), as well as institutional trust (see Figure 2: lower path).
German Development Institute / Deutsches Institut für Entwicklungspolitik (DIE) Second, we ran mediation analyses to further investigate the triangular relationship between values, institutions and trust. A mediation model seeks to identify and explain the mechanism underlying the relationship between an independent and a dependent variable via the inclusion of a third variable, the so-called mediator (Hayes, 2009). The utility of mediation analyses 8 stems from their ability to clarify the nature of the relationship between an independent and dependent variable by assessing the extent to which the effect of the independent on the dependent variable is direct or indirect through the mediator (Iacobucci et al., 2007). In their review of statistical methods for mediation analysis, Iacobucci et al. (2007) underline the superiority of Structural Equation Modelling (SEM) for determining mediation effects. SEM makes it possible to examine how well a process model linking a predictor variable to an outcome variable through one or more intervening pathways fits the observed data (Hayes, 2009). In a traditional mediation analysis, the model would be fit by a series of linear regression analyses as outlined in Baron & Kenny (1986). The advantage of using SEM is that it fits a single model and provides estimations of the indirect and total effects. However, it should be noted that a non-correlation of the variables' error terms is assumed when using SEM as a tool for mediation analyses.
Before we proceed to the interpretation of our results, we want to issue a cautionary note on statistical analysis. Statistical analysis testing has recently been criticised in some parts of the social science literature. Within psychology, for example, this debate has been prominent, and was accompanied by a general 'replication crisis' (for an overview, see: Earp and Trafimow (2015); Aarts and Lin (2015)). The major thrust of the critique on statistical significance and p-values is that they are falsely understood as reporting the odds at which a statistical result is true, whereas they actually report "the probability of obtaining the results in hand, assuming that the statistical null hypothesis is true in the population" (Lambdin, 2012, p. 74). One consequence of this is, that p-values are often presented in a definitive manner even for non-randomly drawn samples or for whole-population samples.
However, for such samples the meaning of statistical significance is restricted, as the first case does not allow inferences on the entire population and the obtained coefficients are already those of the entire population in the second case. Some therefore argue that statistical significance testing should not be done for whole-population samples or nonrandom samples (Figueiredo Filho et al., 2013;Frick, 1998). Others, like Gigerenzer, Krauss and Vitouch (2004) propose a more informed and transparent use and presentation of statistical significance. We want to stress that we also do not have a randomly taken sample, but run our explorative analysis based on the totality of the available data. We further discuss the nature of our data below in a section on the limitations of our analysis. For now, we want to stress that the interpretation of our results refers to our sampled countries, and inferences beyond our availability-sample are indicative at best. For maximum transparency, we nevertheless follow Gigerenzer et al.'s (2004) best-practice suggestion and present exact p-values accompanied by confidence intervals.

Results
The multiple regression analyses revealed that egalitarian values positively relate to the proportion of universalistic vs. means-tested social welfare programmes, even if other explanatory variables are taken into account (β=0.235, p=0.111, see Table 2). In other words, with every increase of one standard deviation in the cultural value emphasis on egalitarianism, the proportion of universalistic vs. means-tested social welfare programmes increases by almost a quarter standard deviation if the other variables are held constant. Surprisingly, the multiple regression analysis revealed that out of all variables, only freedom of discussion is statistically significantly linked to the proportion of universalistic vs. means-tested social welfare programmes (β=0.53, p=0.013). This implies that countries with more freedom of discussion have a larger proportion of universalistic vs. means-tested social welfare programmes.
With regard to the relationship between the scope of social welfare programmes and institutional trust, the results of the multiple regression confirm that the proportion of universalistic vs. means-tested social welfare programmes is significantly positively related to institutional trust (β=0.296, p=0.021, see Table 3). This implies that countries with a larger proportion of universalistic social welfare programmes experience significantly higher levels of institutional trust even if other explanatory factors are controlled for. Additionally, the results reveal significant negative relations between institutional trust and political corruption (β=-0.738, p<0.01) as well as electoral democracy (β=-0.6, p=0.014). These results suggest that countries with a high pervasiveness of political corruption and high levels of electoral democracy, experience lower levels of institutional trust.
Contrary to the statistically significant positive association between the scope of social welfare programmes and institutional trust, the multiple regression analysis showed that a country's proportion of universalistic vs. means-tested social welfare programmes is not connected to its degree of social trust (β=0.083, p=0.533, see Table 4). In other words, if all covariates are held constant, the social trust level rises by merely one tenth standard deviation with every increase of one standard deviation in the proportion of universalistic vs. means-tested social welfare programmes. Hence, the results of the multiple regression suggest that countries with more universalistic social welfare programmes do not experience either statistically significantly or substantially higher levels of social trust. In fact, the analysis revealed that out of all variables, only electoral democracy (β=-0.368, p=0.14), government revenue (β=0.345, p=0.017), GDP per capita (β=0.316, p=0.054), and freedom of discussion (β=0.291, p=0.175) are moderately strongly connected to social trust, with effect sizes of about one third of a standard deviation in social trust for every one standard deviation change in the independent variable. Electronic copy available at: https://ssrn.com/abstract=3553340 To test for mediating pathways in the relationship between egalitarian values and the different types of trust, we conducted Structural Equation Modelling, and subsequently, assessed the indirect effects and their statistical significance. Based on the results of the multiple regression reported in the preceding paragraph, it was likely that a statistically significant indirect effect might not occur due to the lack of a statistically significant relationship between the independent variable (egalitarianism) and the suggested mediator (proportion of universalistic vs. means-tested social welfare programmes). However, some researchers have argued that the use of mediation analyses should not be preconditioned on significant paths linking X to M or M to Y (Hayes, 2009).

Source: Authors
The first part of the mediation analysis examined the triangular relationship between egalitarian values, the scope of social welfare programmes and institutional trust. Structural Equation Modelling indicated that the model sufficiently fits the data, 9 but reveals no significant indirect effect (βindirect effect=0.075, SE=0.032, p=0.141). Interestingly, the direct and indirect effect have opposite directions, so that, ultimately, the total effect is smaller in size than the direct and indirect effects. The indirect effect counters the direct effect, and is almost two thirds as large. Hence, the negative relationship between egalitarian values and institutional trust is largely, although not significantly, reversed by the positive indirect effect that goes via the proportion of universalistic vs. means-tested social welfare programmes. Figure 3 illustrates the path coefficients for the first mediation model.
Similarly, for the triangular relationship between egalitarianism, proportion of social welfare programmes and social trust, the model showed a sufficient model fit, 10 however effect estimations showed that the indirect effect is small and not statistically significant (βindirect effect=0.015, SE=0.01, p=0.630). The analysis implies that the positive relationship between egalitarian cultural values and social trust is only to a small extent mediated by the 9 We assessed model fit using absolute and comparative measures of fit, including the chi-square test, the comparative fit index (CFI), root mean square error of approximation (RMSEA) and the standardised root mean square residual (SRMR). When the model fits well, the chi-square test is non-significant, CFI>0.95, RMSEA<0.06 and SRMR<0.08. The first model showed the following indices of fit: Chisquare(17)=71.21, p=0.00, RMSEA=0.00, CFI=1.00, SRMR=0.00. One possible explanation for the significant chi-square test lies in the size of the model, as models with many variables and non-normal data tend to inflate the chi-square measure. 10 Chi-square(17)=79.69, p=0.00, CFI=1.00, RMSEA=0.00, SRMR=0.00.
proportion of universalistic vs. means-tested social welfare programmes. To facilitate the interpretation of the effects, we computed the ratios for the different types of effect. These showed that the indirect effect is around one eighth the size of the direct effect with about 12 per cent of the total effect being mediated. Figure 4 shows the path coefficients for the second mediation model. To summarise, the results of our explorative quantitative analysis revealed that the scope of social welfare policies statistically significantly predicts institutional but not social trust across countries. More specifically, they show that countries with a large proportion of universalistic social welfare programmes experience higher levels of institutional trust. We find a similar association for social trust, but it is neither statistically significant nor substantially strong. Similarly, the findings reveal that countries with a cultural value emphasis on egalitarianism have more universalistic and less means-tested social welfare programmes. This relationship is sizable, but narrowly misses a significance level of 0.1. Based on the results of the direct effects, it was likely that no statistically significant indirect effect would occur through the proportion of universalistic vs. means-tested social welfare programmes. The analysis proved that the association between emphasis on cultural values and trust is indeed neither statistically significantly nor substantially mediated by the scope of social welfare programmes. Nevertheless, the results show that the proportion of universalistic vs. means-tested social welfare programmes influences the relationship between cultural values and the two types of trust in different ways. For social trust, the scope of social welfare programmes marginally mediates and thereby adds to the positive relationship between egalitarian values and social trust, whereas for institutional trust, it offsets a weak negative effect of egalitarian values.

Implications of main results
The findings of our empirical exploration show that a strong cultural value emphasis on egalitarianism translates into a larger proportion of universal welfare programmes, although this finding does not prove statistically significant according to conventional levels. In the regression model, egalitarian cultural values are only the fourth strongest in terms of effect size, and a number of other factors (e.g. educational equality) are more significant for explaining differences in the scope of social welfare policies. Our results thus suggest that underlying cultural value emphases, such as egalitarianism, express themselves at least moderately in a society's policies and institutions.
Our results also reveal that institutional efforts to tackle social inequalities influence the levels of institutional trust. So far, only a few publications have researched the relationship between the scope of social welfare programmes and trust. For instance, Kumlin and Rothstein (2005) examined individual trust levels of beneficiaries from means-tested and universalistic social welfare programmes in Sweden. However, they draw on individuallevel data and focus on only one country case. This study extends the scope of the sample to 63 countries. Our empirical exploration shows that the scope of institutions is essential for the level of institutional trust that they generate, and thereby for their contribution to social cohesion (Chan et al., 2006;OECD, 2011).
The mediation analysis showed only comparatively small indirect effects of cultural value emphases on trust via the scope of social welfare programmes. However, interestingly, the proportion of universalistic vs. means-tested social welfare programmes mediates the relationship between cultural values and the two trust types in different ways. It would therefore be worthwhile to investigate the indirect effects further. Regarding social trust, the positive effect of egalitarian values on social trust increases due to the positive indirect effect from the scope of social welfare programmes. As cultural egalitarianism seeks to socialise people to view one another as equal and to internalise a commitment to cooperate, it seems logical that people trust each other more in countries where egalitarian values prevail (Rothstein & Uslaner, 2005). According to our results, welfare institutions channel some part of this effect. As they increase the likelihood that welfare institutions have a universal scope, egalitarian values raise social trust within societies.
For our second mediation model, the scope of social welfare programmes has a positive indirect effect on institutional trust that offsets the negative relationship between egalitarian values and institutional trust. One plausible explanation for the negative relation between egalitarian values and institutional trust originates from critical attitudes towards institutions and their ability to guarantee social justice, equality and welfare to every citizen, which all represent characteristics that lie at the core of cultural egalitarianism. This is in line with findings in Schwartz (2006), showing that egalitarian values predict greater political activism as well as more critical political attitudes. Since the direct and indirect effect have different directions, introducing the proportion of universalistic vs. means-tested social welfare policies as a mediating variable largely reverses the negative effect of egalitarian values on institutional trust. In other words, countries with a strong cultural value emphasis on egalitarianism experience lower institutional trust, unless they have a large proportion of universalistic vs. means-tested social welfare programmes.
Overall, our results imply that the scope of social welfare policies mitigates the negative relation between egalitarian values and institutional trust, whereas for social trust, it reinforces the positive effect of egalitarian values.

Limitations
Several limitations of this empirical exploration must be considered. First, even though the time difference between the assessment of the different variables was kept to a minimum, time ranges between data collection still vary across countries from zero to fifteen years. This is mostly down to the age of the Schwartz Value Survey, with some data dating back to 1985 while the most recent entries were made in 2005. Therefore, more recent data is imperative for a better insight into cultural values, especially in fast changing societies.
Second, some of the country-level variables were highly aggregated from individual-level data. In almost all social sciences, the loss of information that goes along with the aggregation of micro-level to macro-level data has been a prominent issue (Clark & Avery, 1976). Additionally, this quantitative exploration draws on data from 63 countries, which constitutes a relatively small sample. In small samples, the obtained regression coefficients fluctuate more easily, thereby providing less robust results. Although the sample includes a limited number of countries and parts of the variables are aggregated, there is a wealth of data behind every country (e.g. N=14996 for the Schwartz Value Survey). Given this large amount of data backing up every country variable, our analysis still provides valid results, despite the aforementioned limitations.
Third, European and Western countries are clearly overrepresented in the sample of this explorative analysis, as quantitative data from developing countries remains sporadic. Especially in case of the Schwartz Value Survey, it is necessary to expand its cross-cultural application across all continents, which implies increased efforts to collect reliable primary data.
Some measurements used in our exploration are also subject to methodological limitations. For instance, the binary most-people question to assess social trust has shortcomings, such as limited scale length and interpersonal incomparability. The latter may also apply to the institutional trust measure, as people from different cultures might interpret the questions and answer scale differently. Furthermore, it may also be problematic to use countries as units of cultural analysis since cultural differences within countries are often larger than between countries, especially when it comes to values (Fischer & Schwartz, 2011). Up to now, Schwartz' cultural values have been widely used for measuring cultural differences, next to the common binary distinction between individualism and collectivism. However, more recent theories such as the dimensional tightness-looseness concept introduced by Gelfand et al. (2011), might shed new light on cultural differences and broaden the ways of 'unpacking' culture.
German Development Institute / Deutsches Institut für Entwicklungspolitik (DIE) Finally, an important limitation of our empirical exploration is that its findings are based on correlational data, and thus are not indicative of causality. Future research that combines qualitative and quantitative data and that uses different methodological approaches is needed to establish whether and to what extent causality exists between the variables discussed here. Along the same lines, our explorative analysis only looked at a limited amount of possible relationships between the variables, but for instance, did not consider analysing reversed causal direction. For instance, social strain stemming from a growing inequality might hinder the implementation of universalistic policies, leaving countries stuck in a sort of social trap (Rothstein & Uslaner, 2005). The logic behind this is the following: If trust levels are low due to prevailing social inequality, universalistic public policies that could counter this situation are hard to establish exactly because there is no trust. Therefore, future research looking at the effects of trust and social inequality on the implementation of universalistic social welfare programmes is needed, to better understand the interlinkages between the factors and their position and direction in the causal chain.

Conclusion
In times of crisis and polarisation, societies discuss and debate their values as common denominators that bind them together. Yet, the label of values often hides a patchwork of different ideas and concepts. The research literature offers a general understanding that might help to more clearly delineate this. Within the literature, values describe desirable and trans-situational end states and behaviours. A major fault line divides personal from societal values. Here in particular, more conceptual work is needed to make the distinction between the two clearer and easier to understand. In addition, the research literature has to engage in more basic empirical groundwork on what values are and how they are expressed by individuals and societies. Maybe most urgently, the research literature has to critically examine its own implicit assumption of the universality of values. To achieve this will require research with a broader geographical scope.
A key premise in research and public discourse is that values play a central role for social cohesion. Three different pathways channel this role of values. First, consensus on values may create social cohesion beyond the end states and behaviours these shared values actually promote. But evidence for this intrinsic pathway of values is inconclusive. What is dearly needed, is a commonly acknowledged way to measure value consensus and value polarisation. Second, the particular end state or behaviour, that is to say the content promoted by certain values, may foster social cohesion. Although evidence supports this conjecture for some types of values, such as egalitarianism and individualism, this evidence does not look at social cohesion as an emergent overall concept, but investigates only the distinct sub-attributes of social cohesion, such as trust, identity and cooperation. To gauge the influence of values on social cohesion, research needs to transcend the focus on these sub-attributes and investigate social cohesion as a stand-alone concept. Third, values may foster social cohesion through their influence on a society's institutions. Although expounded in the literature, the trilateral relationship between values, institutions and social cohesion has been neglected so far.
After explicating the theoretical foundation of values, we empirically explored this last pathway, in which institutions mediate the effect of values on social cohesion. To do so, we used the example of how egalitarian values relate to universalistic welfare institutions as well as social and institutional trust. Our empirical exploration revealed a complex relation. Egalitarian values are plausibly and moderately connected to more universalistic welfare institutions. However, their positive relation to social trust is only minor. The presence of universalistic welfare institutions adds to this slight, but positive relation. In contrast, the relation between egalitarian values and institutional trust is negative, and in this case, the trust-increasing influence of universalistic welfare institutions offsets the initial negative relationship. This suggests that people are more critical of their institutions in societies with an emphasis on egalitarian values, but that institutions in accordance with these egalitarian values can make a crucial difference for citizen confidence in the state.
Our empirical exploration opens several alleyways for further research. Due to issues related to scope, we only focused on trust. However, the relationship between egalitarian values and welfare institutions should also be investigated focusing on other attributes of social cohesion, such as identity and cooperation, as well as overall social cohesion. In addition, our exploration was limited in geographical scope. Data for more countries is clearly needed, both quantitative and qualitative. Moreover, the relationship between values and social cohesion should also be investigated using different types of institution, like political bodies such as parliaments, or economic institutions such as labour regulations. Sustainability institutions represent another type of institution that is becoming more relevant to investigate. Depending on the type of institution, it might also make sense to relate them to specific kinds of values. For instance, hierarchy and embeddedness values might be more relevant for parliaments and labour regulations, whereas value emphases on mastery and harmony could be more important for sustainability institutions. A final implication of our empirical exploration is that value and institutional change coevolve. But how exactly does this happen? Often, social movements are at the forefront when it comes to propagating new values. But how does this translate into demands and the formation of new institutions? These questions are applicable to a wide range of topics, from democratisation to gender and LGBTQI equality to social movements promoting sustainability. Electronic copy available at: https://ssrn.com/abstract=3553340