Abstract
Social scientists and development practitioners are increasingly recognizing the significance of social cohesion as a prerequisite for and part of sustainable development. At the same time, the COVID-19 pandemic poses severe challenges to the cohesion of societies. However, the specific role of people’s religiosity in their attitudes and behaviors relevant for social cohesion has rarely been empirically-quantitatively investigated.
Using primary data from an online survey in Germany, this article addresses three research questions: 1) How can the construct “social cohesion” be empirically and quantitatively measured? 2) What dynamics of social cohesion were evident in Germany during the COVID-19 pandemic (in 2020/2021)? 3) What specific role did people’s religiosity play in social cohesion in that context?
These questions are answered within a theory-led empirical analysis. Social cohesion is conceptualized as a multidimensional construct that manifests itself in horizontal (citizen–citizen) and vertical (citizen–state) relationships. It is measured by eight subdimensions: social/institutional trust, social inclusive/national identification, social/institutional responsibility, and social/political engagement. Religiosity is specified with three dimensions: belonging (religious affiliation), believing (faith intensity/content), and behaving (prayer/service attendance).
The empirical results for Germany show that the people surveyed in the period from July 2020 to January 2021 tended to hold together more. Several cohesion dimensions, particularly institutional trust and national identification, strengthened. In contrast, social trust weakened. The role of religiosity is ambivalent. It is not the intensity but the content and practices of faith that make the difference. Exclusivist religious beliefs, a punitive image of God, and private prayer practice are negatively related, while a loving image of God and service attendance are positively related to social cohesion.
1 Introduction
In view of current challenges around the globe, such as polarization, state collapse, and civil wars, the ageless question of what holds societies together is particularly relevant. Being a fundamental prerequisite for a functioning state, social cohesion is regarded as an essential political goal (Fonseca, Lukosch, and Brazier 2019; Langer et al. 2017; Schiefer et al. 2012). Consequently, various research projects, institutes, and organizations have increasingly recognized the relevance of social cohesion and integrated corresponding concepts and policies into their work, especially in the field of sustainable development. Current initiatives include the Social Cohesion Projects1 of the United Nations Development Programme (UNDP), the recently launched Social Cohesion Hub2 of the German Institute of Development and Sustainability (IDOS), and the Social Cohesion Radar3 of the Bertelsmann Stiftung. In the context of the Sustainable Development Goals (SDGs), social cohesion is considered an essential part of or pathway to achieving SDG 16: “Promote peaceful and inclusive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels.”4
However, the global COVID-19 pandemic has posed major challenges to the cohesion of societies. Have their members stuck together (because they “sit in the same boat”) or have they drifted apart (due to restrictions, e.g., concerning personal contacts/social distancing)? The literature proposes two possible scenarios: “coming-together” vs. “coming-apart” (Borkowska and Laurence 2021; Delhey et al. 2021). In Germany, for example, both tendencies have become visible. Numerous initiatives for solidarity and neighborhood assistance were launched on the ground. However, protest movements and anti-Covid demonstrations (e.g., by the “Querdenker”) became louder and various conspiracy theories circulated (Kühne et al. 2020b; Schulz and Faus 2022; Yendell, Hidalgo, and Hillenbrand 2021).
So, the term “social cohesion”5 has become a buzzword: “[E]veryone talks about social cohesion, but everyone means something different by it, and hardly anyone can really say what it is all about” [own translation] (Vopel and Unzicker 2012, 8). However, “[d]espite its importance, social cohesion is rarely quantified and measured. Yet measurement is needed if we are to investigate causes and consequences of sc [social cohesion] empirically” (Langer et al. 2017, 323). Against this background, this article considers how the abstract construct “social cohesion” can be empirically and quantitatively measured and analyzed by presenting a theory-led empirical research design and applying it exemplarily to the German context.6 To this end, I use primary data from a (non-representative) comprehensive online survey that we, at the Cluster of Excellence “Religion and Politics” (WWU Münster)7 carried out in cooperation with the Research Institute of Social Cohesion (University of Leipzig)8 between July 2020 and January 2021.
While several studies about social cohesion in Germany during the COVID-19 pandemic have been published recently (e.g., Bertelsmann Stiftung 2022; Brand et al. 2021; Delhey et al. 2021; Follmer, Brand, and Unzicker 2020; Schulz and Faus 2022), they have rarely looked at religious factors (see section 2.2). Therefore, I specifically examine the role of people’s religiosity in aspects of social cohesion in a comprehensive and differentiated manner. Within the limited scope of this article, I concentrate on the German context, focus on the micro level (individuals as units of analysis) and situate myself in quantitative attitudes and survey research.9 Specifically, I answer the following research questions:
1) How can the construct “social cohesion” be empirically and quantitatively measured?
2) What dynamics of social cohesion were evident in Germany during the COVID-19 pandemic (in 2020/2021)?
3) What specific role did people’s religiosity play in social cohesion in that context?
In the first part, I outline the theoretical framework, in which I provide a literature review corresponding to the three research questions: the conceptualization and measurement of social cohesion (2.1), studies on social cohesion in the context of the COVID-19 pandemic in Germany (2.2), and the role of religiosity in social cohesion (2.3). Then, I lay out my methods and data (3), as well as the operationalization of the variables (4). In the empirical part, I present key descriptive statistics (5.1), multivariate regression analyses (5.2), and finally discuss the results (5.3). This will lead to my final conclusion (6).
2 Theoretical Framework
Within this theoretical framework, I provide an overview of the literature relevant to my research questions and derive the specification of the central concepts as well as the working hypotheses.10 For that, I draw on the research design that is outlined in Hillenbrand (2020a).11
2.1 Conceptualizing and Measuring Social Cohesion12
There is no standard definition, nor unified measurement, for social cohesion (Fonseca, Lukosch, and Brazier 2019; Schiefer and van der Noll 2017). Different terms like “social integration” (Schnabel and Grötsch 2012, 2014), or sub-concepts like “social capital” (Pickel and Gladkich 2011; Putnam 2001; Traunmüller 2012; World Bank 2002), are used.
Systematic approaches to comprehensively quantify and measure the concept of social cohesion are still rare and quite recent (Langer et al. 2017, 322). Early attempts to make this construct measurable and thus comparable (across different countries) were made in the 2000s, for example, by the World Bank (Colletta and Cullen 2000), the OECD (2011), or the United Nations Economic Commission for Latin America and the Caribbean (ECLAC 2007). In addition, engagement with the conceptualization and measurement of social cohesion has increased in academia (e.g., Chan, To, and Chan 2006; Dickes, Valentova, and Borsenberger 2010; Langer et al. 2017). For example, authors commissioned by the Bertelsmann Stiftung systematized the cohesion literature and developed a comprehensive measurement concept which they applied to various countries (first from the West and later from Asia) (Bertelsmann Stiftung 2018; Dragolov et al. 2013, 2016; Schiefer et al. 2012; Schiefer and van der Noll 2017). Similarly, a research team at the German Institute of Development and Sustainability has recently been working on a measurement concept for social cohesion that is particularly applicable to the African continent (Leininger et al. 2021).
Despite differences in these approaches, a consensus seems to be emerging on the following key points:
Social/societal cohesion is essentially about relationships between specific actors, groups, and institutions that make up a society (Langer et al. 2017; Leininger et al. 2021; Schiefer et al. 2012).
Several relationship levels can be differentiated. Primarily, a horizontal level (citizen – citizen relations) is distinguished from a vertical level (citizen – state relations) (Chan, To, and Chan 2006; Leininger et al. 2021; Schiefer et al. 2012).13
Social cohesion is understood as a multidimensional phenomenon. However, there is no consensus on the exact number or content of these dimensions. Earlier concepts were much broader and also included socioeconomic factors like inequality, poverty, social mobility, etc. These were measured at the macro level using objective factors (Colletta and Cullen 2000 ECLAC 2007; OECD 2011). However, the more recent approaches focus on subjective variables. This means that social cohesion and the quality of relationships are measured in terms of perceptions, attitudes, and behaviors of the society members (toward others/the state). For such a perceptions-based social cohesion index, surveys are usually used (e.g., the European or World Values Survey, the Afro-/Euro-/Asia-/Latinobarometer, etc.) (Hino et al. 2019; Langer et al. 2017; Leininger et al. 2021). Widely recognized pillars of cohesion involve trust, a sense of belonging, and forms of prosocial behavior such as solidarity or civic engagement (Chan, To, and Chan 2006; Leininger et al. 2021; Schiefer et al. 2012).
I follow the latest approaches of IDOS (Leininger et al. 2021), the Bertelsmann Stiftung (2018), Delhey et al. (2018), Dragolov et al. (2016), Langer et al. (2017), and Hino et al. (2019) and concentrate on perceptions-based indicators, i.e., on citizens’ attitudes and behaviors – in line with this article’s focus on the micro level and survey research. I refrain from blending them with aggregated, objective measures such as a country’s poverty/inequality rate. This approach has the advantage of avoiding conceptual overstretching and achieving higher discriminatory power. Moreover, when socioeconomic factors are not included as components but as consequences or prerequisites of social cohesion, further studies will be able to investigate their interdependencies and provide insights into influencing or output variables (Chan, To, and Chan 2006; Dragolov et al. 2016).
Regarding the specific cohesion dimensions, I aim to develop a relatively parsimonious concept that can be broadly applied to diverse (country) contexts in future studies. Therefore, I focus on the core dimensions in which most existing approaches agree. I primarily follow the recently developed concept of IDOS because it brings together the current state of cohesion research and was developed in a profound research process with diverse scholars and development practitioners.14 Based on this concept, I define four dimensions (or eight subdimensions, depending on the relationship level: horizontal/vertical) which are summarized in table 1 (further argumentations can be found in Hillenbrand 2020a or the literature cited in the table).15 This results in the following working definition:
The cohesion of a society manifests itself in social and political relationships among citizens (horizontal level) and between citizens and the state (vertical level): the greater the trust, inclusive identification, responsibility and engagement in a country, the stronger its cohesion (Hillenbrand 2020a, 26; see also Leininger et al. 2021, 3).
Conceptualization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a)
Conceptualization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a)Conceptualization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a)2.2 Social Cohesion in the Context of the COVID-19 Pandemic
The global COVID-19 pandemic has been an impactful event affecting the individual, social, and political lives of people and societies worldwide. In relation to how crises such as this pandemic influence social cohesion, two possible scenarios have been described in the relevant literature (e.g., Borkowska and Laurence 2021; Delhey et al. 2021; Kritzinger et al. 2021).
The “coming-together” scenario, on the one hand, assumes a unifying effect. An exogeneous shock could bring people in a country closer together, to jointly resist the shared threat or common “enemy” (Delhey et al. 2021, 4; Kritzinger et al. 2021, 1207–08). Collective experiences of hardship (such as the COVID-19 pandemic) could alter people’s preferences, encourage the sharing of burdens and resources, and promote prosocial behavior expressed, for example, through trust. This scenario assumes that trust functions as a resource and coping mechanism in situations of uncertainty (Delhey et al. 2021, 4; Goerres and Vail 2021, 6). Part of this scenario is the “rally-round-the-flag” effect16 that is prominent in political science. Threats would trigger powerful emotions and desires, for example, for security, which citizens often seek to obtain from public institutions and officials. Thus, citizens would show greater support for their governments, which they expect will protect them and resolve the crisis (Delhey et al. 2021, 4–6; Goerres and Vail 2021, 5–7; Kritzinger et al. 2021, 1205–08).
On the other hand, the “coming-apart” scenario assumes social cohesion and bonds among society members will weaken through crises. During the Covid pandemic, people would feel isolated and less connected while social interactions are restricted and would develop mistrust and fears (e.g., that every fellow citizen can transmit the virus and therefore poses a potential threat). Trust in public institutions could also decrease, for instance, when mistakes in crisis management occur (Borkowska and Laurence 2021, 5618–22; Delhey et al. 2021, 4).
With regard to these scenarios, and in line with this article’s research focus, I reviewed empirical studies analyzing the dynamics of social cohesion during the COVID-19 pandemic, particularly in the German context (e.g., Bertelsmann Stiftung 2022; Borbáth et al. 2021; Brand et al. 2021; DCV 2022; Delhey et al. 2021; Follmer, Brand, and Unzicker 2020; forsa 2022; Goerres and Vail 2021; Höltmann and Hutter, n.d.; Krause, Gagné, and Höltmann 2020; Kühne et al. 2020a, 2020b; Liebig 2021; Schulz and Faus 2022). Most empirical results to date support the “coming-together” scenario, at least for the early stages of the pandemic (in 2020).17 Often, stronger solidarity, helpfulness, and political trust was found: “[T]here is mounting evidence that the pandemic initially drew societies together rather than apart” (Delhey et al. 2021, 2). From that, I derive my first working hypothesis (as an expected outcome regarding my second research question on the dynamics of social cohesion):
H1) The people surveyed in Germany between July 2020 and January 2021 tended to hold together more, i.e., showing higher levels of trust, identification, responsibility, and engagement.
However, as the pandemic is still ongoing at the time of writing, it remains to be seen how the social cohesion dimensions will develop and change in the long term. Recent data and results from late 2021 and early 2022 paint a less optimistic picture and the results are more ambivalent (e.g., Bertelsmann Stiftung 2022; DCV 2022; forsa 2022).
Moreover, some studies point to the following paradox (Brand et al. 2021; Schulz and Faus 2022). If people are asked about the specific cohesion dimensions (e.g., how much they themselves trust others/institutions, show solidarity, or are socially engaged), the picture is quite positive – and overall, social cohesion in Germany seems to have remained stable over the years and may have even improved in the early stages of the pandemic. However, when people are asked about their assessment/feeling of whether social cohesion has generally improved or deteriorated, the diagnosis is much more negative: More people see social cohesion eroding and society polarizing.
Looking at these empirical studies on social cohesion in Germany in the context of the COVID-19 pandemic, the religious factor has rarely been taken into account. Very few publications have considered the role of people’s religiosity and when they have done so, they only looked at religious affiliation and differentiated between two or three main groups (DCV 2022; Follmer, Brand, and Unzicker 2020; forsa 2022).18 Contributing to filling this research gap, I investigate religiosity’s specific role in the following section.
2.3 The Role of Religiosity in Social Cohesion
Before formulating hypotheses about the role of religiosity in social cohesion, I need to clarify the terminology – what I mean by “religiosity” in this article and how I specify this variable so that I can empirically measure it. The aim and strength of this study is that “religiosity” is conceptualized in a comprehensive and differentiated way.
First of all, there is no universal definition for “religion” (Pollack and Rosta 2017). In line with this article’s focus on individuals as units of analysis, I concentrate on “religiosity” (as a characteristic of individuals), which is defined as a person’s beliefs and behaviors toward the transcendent (Sherkat 2015, 377).19 “Transcendence” relates to a reality that “exceed[s] the differently defined area of the empirically comprehensible” (Pollack and Rosta 2017, 45). What precisely this means is not predefined but left to the interpretation of the individuals who are surveyed about their beliefs within this article’s framework of attitude research.
According to my empirical approach, I conceptualize various measurable dimensions of religiosity. For that, I draw on religious dimension research, especially the current systematization by Pollack and Rosta (2017), and important empirical studies (Basedau, Gobien, and Prediger 2018; Olson and Warber 2008; Schnabel and Grötsch 2014) that specify three central dimensions – the widely used “three b’s” (see table 2). From this, the following working definition can be derived: “Religiosity refers to the beliefs and behaviors of individuals towards the transcendent, expressed in the intensity and contents of their faith (believing), religious practices (behaving) and religious affiliation (belonging)” (Hillenbrand 2020a, 27).
Conceptualization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Basedau, Gobien, and Prediger (2018); Hillenbrand (20 20a, 2020b); Olson and Warber (2008); Schnabel and Gr ötsch (2014); Traunmüller (2012)
Conceptualization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Basedau, Gobien, and Prediger (2018); Hillenbrand (20 20a, 2020b); Olson and Warber (2008); Schnabel and Gr ötsch (2014); Traunmüller (2012)Conceptualization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Basedau, Gobien, and Prediger (2018); Hillenbrand (20 20a, 2020b); Olson and Warber (2008); Schnabel and Gr ötsch (2014); Traunmüller (2012)What role do these different aspects of religiosity play in social cohesion? As there is still a lack of current, empirical studies on the specific role of religiosity in social cohesion during the COVID-19 pandemic (see section 2.2), I rely on the general empirical literature about the nexus of religion and social cohesion (for the German context) to derive hypotheses about their relationships (e.g., Pickel and Gladkich 2011; Pollack and Müller 2013; Schnabel and Grötsch 2012, 2014; Traunmüller 2012). Based on this literature, I generally regard “social cohesion” as the dependent variable and “religiosity” as the independent variable. However, as the survey design applied in this article cannot prove causality (see section 5.3), I rather speak of “role/relationships” than of “causal effects.” Given this article’s limited scope, the hypotheses will only be briefly outlined in the following. They stem from Hillenbrand (2020a) and should be regarded as working hypotheses for this specific research design.20
H2) Belonging dimension and social cohesion
People with a religious affiliation are embedded in religious communities, i.e., social networks, where certain values, social norms, and rules of behavior are transmitted to them that also shape their attitudes and behaviors relevant for social cohesion (Preston, Salomon, and Ritter 2014, 161–63; Traunmüller 2012, 61–63). There is no clear empirical evidence on the general cohesive vs. disruptive effects of specific religious groups per se. Rather, the status of a religion in the particular society and resulting ingroup vs. outgroup dynamics seem to be decisive. Members of the majority religion may perceive members of the minority group as a threat and devalue them, while minority religious groups may feel excluded and react by forming subgroups (Pollack and Müller 2013; Tausch, Hewstone, and Roy 2009). From this, it follows that (Hillenbrand 2020a, 30):
H2) Religious affiliation is related to social/institutional trust, inclusive social/national identification, social/institutional responsibility, and social/political engagement (which constitute the cohesion of a society).21 A religious minority status is negatively related to dimensions of social cohesion.
H3) Believing dimension and social cohesion
According to the widely held religious prosociality hypothesis, believers are expected to show stronger prosocial behavior such as solidarity, charity, responsibility, etc. (Galen 2012, 876; for details on the effect mechanisms see Preston, Salomon, and Ritter 2014, 158–63). In contrast, the religious intolerance hypothesis considers strong faith to be accompanied by intolerance and the devaluation of members of other religions and non-believers (Batson 2013, 88–100; Berggren and Bjørnskov 2011, 462–63). To examine the direction of this relationship, I follow and test the prominent religious prosociality hypothesis with the next working hypothesis (Hillenbrand 2020a, 28):
H3a) The intensity of faith is positively related to the dimensions of social cohesion.
Regarding the content of faith, empirical research on religion, especially the concept of bridging vs. bonding in the social capital literature, points to the importance of inclusive vs. exclusivist beliefs (Pickel and Gladkich 2011, 83/85; Traunmüller 2012, 20/58–61). “Exclusivist” beliefs that place one’s own religion absolutely, above other religions and fields (e.g., science, politics, etc.), are obstructive to the cohesion of a country as a whole with its usually diverse groups. Moreover, it is suggested that an image of a loving, benevolent God facilitates prosocial and cohesive behaviors, while a punitive conception of God hinders them. Based on this, I hypothesize that (Hillenbrand 2020a, 29):
H3b) Exclusivist religious beliefs are negatively related to the dimensions of social cohesion.
H3c) An image of a punitive God is negatively related to the dimensions of social cohesion, while an image of a loving God is positively related.
H4) Behaving dimension and social cohesion
The pertinent literature often ascribes different effect directions to private vs. social religious practices. Extensive private prayer would consume a lot of time, leaving a person with fewer resources for other sociopolitical commitments (Basedau, Gobien, and Prediger 2018, 1113–17; Traunmüller 2012, 61/126/158). On the other hand, social religious practices such as collective rituals and shared experiences in religious services would socially embed the individuals and connect members of society from various generations and social statuses, strengthening social networks and ultimately the cohesion dimensions (Durkheim 2007, 72–75/625–27; Norris and Inglehart 2011, 192–94; Traunmüller 2012, 41–42/61–63). This leads me to my last set of hypotheses:
H4a) Private prayer practice is negatively related to the dimensions of social cohesion.
H4b) Social religious practice (expressed in service attendance) is positively related to the dimensions of social cohesion.
3 Methods and Data
To answer the underlying research questions, quantitative empirical survey and attitudes research as well as the application of statistical methods represent suitable methodological approaches because they make it possible to include large numbers of cases and examine correlations between the variable dimensions, while controlling for relevant third variables (age, gender, education, etc.) (Lauth, Pickel and Pickel 2015; Pickel 2011).
Primary data from a comprehensive online survey in Germany serves as the data basis (Hillenbrand n.d.).22 Researchers at the Cluster of Excellence “Religion and Politics” (WWU Münster) conducted this survey in cooperation with the Research Institute of Social Cohesion (RISC) (University of Leipzig). It is important to note that we did not have several waves (no panel data), but there was one survey period from July 2020 to January 2021. Our case selection was not based on a random principle (no representative study) because one of the central goals (and strengths) of our survey was the focus on the religious factor and its comprehensive and differentiated measurement (which has so far been neglected in the current social cohesion research in the context of the COVID-19 pandemic). We wanted to be able to make comparisons and gain deeper insights into various religious people and groups as well as into different types of religiosity that impact relevant political and social attitudes and behaviors. That is why we, for example, actively mobilized specific religious groups and minorities to participate, including Jews, Muslims, and Evangelicals, so that the case numbers for them would be sufficiently large for statistical analyses. In representative surveys in Germany, the samples for these religious groups are usually not large enough (< 30) for inferential statistics. Moreover, Evangelicals and Protestants are often not distinguished from each other. However, it has been shown that their attitudes can strongly diverge on central sociopolitical attitudes (e.g., regarding their affinity for conspiracy theories in the context of the Covid pandemic; see Hillenbrand and Pollack forthcoming). We wanted to investigate whether this is also the case for the social cohesion dimensions.
For these reasons, we used an online link for our data collection, which was widely distributed, for example, via the broad networks of the Cluster of Excellence in Münster and the RISC in Leipzig (press releases, websites, interviews, research communication, etc.). In addition, we reached out to various religious communities and organizations in Germany (e.g., Zentralrat der Muslime, Zentralrat der Juden, Jüdische Hochschule), which encouraged their members to participate. Moreover, our survey received widespread attention in the media,23 which also called for participation. In addition, we distributed printed survey forms in, for instance, socially poorer districts and neighborhoods, and handed them out to churches and parishes, who distributed them widely.
In this way, we reached a total of 2,373 respondents, with sufficient representatives from diverse religious groups in Germany (e.g., Catholics, Protestants, Evangelicals, Muslims, Jews, and spiritual and non-religious people). The survey included items related to social cohesion, COVID-19, and conspiracy theories as well as religion and spirituality.24
4 Operationalization
When operationalizing the variables and creating our questionnaire we were guided by high-quality standard data surveys, especially the World Values Survey,25 the Bertelsmann Stiftung Religionsmonitor,26 and the International Social Survey Programme.27 We were also in contact with an international research team that was carrying out similar surveys on COVID-19 and religiosity in other European countries (Spain, Italy, Finland, and Poland). To facilitate future comparative analyses, we used their wording and coding for some central items. Due to this article’s limited scope, I refrain from a detailed description of all the respective items in the text.28 Instead, I provide overviews by listing them in tabular form. Table 3 displays the operationalization of the cohesion dimensions.29
Operationalization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)
Operationalization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)Operationalization of social cohesion
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)As we did not have several survey waves (and thus do not have panel data or pre-Covid data), we integrated questions about changes within the questions so that we could track dynamics and analyze whether the social cohesion dimensions have become weaker or stronger during the COVID-19 pandemic. We let respondents themselves evaluate possible changes that might have occurred. In order to explore the paradox emerging in previous research (see section 2.2), we asked questions about the assessment of the relevant behavior of others (e.g., to what extent people (generally) show solidarity or hold together) in addition to questions about attitudes and behaviors of oneself (e.g., to what extent the respondents themselves trust others more/less, feel more/less connected, etc.). Figure 1 shows this battery of questions (respondents used a slider to indicate whether the cohesion dimension had increased, decreased, or remained the same).30
Operationalization of social cohesion (dynamics during the COVID-19 pandemic)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Operationalization of social cohesion (dynamics during the COVID-19 pandemic)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Operationalization of social cohesion (dynamics during the COVID-19 pandemic)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)The operationalization of the independent variable “religiosity” is summarized in table 4.
Operationalization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)
Operationalization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)Operationalization of religiosity
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: Hillenbrand (2020a); LTCP (2021)Finally, table 5 provides an overview of the additional control variables included in the multivariate analyses.
Operationalization of control variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Operationalization of control variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Operationalization of control variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)5 Empirical Results
In this section, I discuss the results of my quantitative empirical analysis. First, I outline the descriptive statistics regarding central variables and the dynamics of the social cohesion dimensions (second research question). Then, I use multivariate regression models to examine the relationships between the religion and cohesion dimensions (third research question).31
5.1 Descriptive Statistics
First, it is important to have an overview of the sample composition. As this is not a representative sample for the entire German population (due to our specific research focus; see section 3), certain groups are overrepresented and others are underrepresented.32 Of the 2,373 respondents, 57% identify as female, 42% as male, and 1% as gender diverse. Persons of younger or middle age, those with higher education and belonging to the lower-middle or upper-middle class, as well as people from western German states, are more strongly represented. In addition, religiously affiliated people (including minority religious groups) are overrepresented, because our study aimed to zoom in on different religious groups and investigate certain forms, types, and patterns in terms of their cultural, political, and social attitudes and behaviors. According to the central limit theorem (see Weigand 2009, 221–25), statistical analyses are appropriate from a case number of 30. We achieved this for the following groups, which can, consequently, be considered in the subsequent empirical analyses: Catholics (999), Protestants (504), Evangelicals (219), Muslims (87), Jews (48), other religions (67), spiritual persons without religion (92), and non-religious people (314).
The following tables 6, 7, and 8 summarize key descriptive statistics of all the variables integrated in the empirical analysis (independent and dependent variables).
Distribution of independent variables I (categorical)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of independent variables I (categorical)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Distribution of independent variables I (categorical)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of independent variables II
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of independent variables II
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Distribution of independent variables II
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of dependent variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of dependent variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Distribution of dependent variables
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Knowing the sample composition, we now analyze the distribution of the variables related to possible changes in the social cohesion dimensions during the COVID-19 pandemic in order to answer the second research question about the dynamics of social cohesion. The relevant descriptive statistics are illustrated in figure 2 and discussed in the following.
Distribution of variables related to the change in the social cohesion dimensions due to the COVID-19 pandemic in Germany
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Distribution of variables related to the change in the social cohesion dimensions due to the COVID-19 pandemic in Germany
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Distribution of variables related to the change in the social cohesion dimensions due to the COVID-19 pandemic in Germany
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)We see that social trust remained the same for most respondents (61%), while 26% said that their trust in others decreased and 13% reported an increase. Institutional trust increased for 41% of respondents and decreased for 23% (for 35% it remained unchanged). Social connectedness also increased (for 46%) rather than decreased (for 17%) – for 38% it remained the same. A large majority of 74% were satisfied living in Germany during the pandemic – while only 11% were dissatisfied. Moreover, 64% said they showed more solidarity and helpfulness during the pandemic (only 4% showed less) and around half of the respondents declared that they complied with the legal regulations more (only 11% complied less). In addition, two-thirds of respondents considered social engagement an important area of their life during the Covid period. In comparison, the figure for political engagement was significantly lower at 36%.
Overall, we can conclude that the people surveyed in our study in the period from July 2020 to January 2021 tended to stick together more (confirming H1). This finding fits very well with previous studies that speak of a “coming-together” scenario (Borowska and Laurence 2021), a “rally-round-the-flag effect” (Kritzinger et al. 2021) or “honeymoon” (Goerres 2022) of social cohesion during the COVID-19 pandemic.
However, our survey also reveals that there can be differences depending on the specific cohesion dimension, which is why a differentiated analysis is important. In line with the “rally-round-the-flag effect,” it was mainly institutional trust and national identification that were strengthened. We also saw greater solidarity and helpfulness (as found in other studies, e.g., Borbáth et al. 2021; Follmer, Brand, and Unzicker 2020; Höltmann and Hutter, n.d.). Social trust, on the other hand, rather weakened.
In addition to the respondents’ statements about their own cohesion-relevant attitudes and behaviors, we asked them about their assessment of the behaviors of others in society in general. In this regard, 48% said that people generally showed more solidarity during the pandemic, 19% perceived less solidarity, and 33% saw no change. In contrast, 40% felt that people were drifting further apart in the pandemic; one-third said that they were holding together more and 28% perceived no change. This points to the paradox cited in the literature (see section 2.2). On the one hand, people generally feel that cohesion is deteriorating and rate the behaviors of others in this regard rather negatively. On the other hand, stability or slight improvement tends to emerge when people are asked about their own attitudes and behaviors that constitute a society’s cohesion. Of course, response tendencies such as the phenomenon of “social desirability” could also cause a bias here. This needs to be explored in more detail in further studies.
5.2 Multivariate Regressions
To empirically test the hypotheses derived in section 2.3, I conducted multivariate regression analyses (linear ordinary least squares regressions). The detailed regression tables and the distributions of all variables included in the models are presented in the appendix. It is important to note that the dependent variables used for the regression modelling are not the items related to the change/dynamics of social cohesion (which were relevant for the second research question). Instead, the general dimensions of social cohesion (as operationalized in table 3) were used in order to remain congruent with the theoretically derived conceptual specification of social cohesion (see section 2.1) – and because answering the third research question is about the extent to which different dimensions of a person’s religiosity are related to his/her general sociopolitical attitudes and behaviors that constitute overall cohesion in a society.33
In the following, I present the key results of the regression models, structured according to the hypothesis testing.34
With regard to religious belonging (H2), various significant results for several religious groups become apparent, which in the following must always be interpretated in comparison to the non-denominational reference group. In our survey, both Catholics and Protestants demonstrate higher institutional trust and national identification, but lower social responsibility (than the non-denominals). Moreover, belonging to evangelical or free churches is related to higher national identification but lower social responsibility and social engagement. The identification with Islam is associated with higher inclusive identification but lower social trust and social and political engagement, which is reflected in a lower overall cohesion index. The Jewish people in our survey show lower social trust, social identification, and institutional responsibility, resulting in a lower social cohesion index. Those who report belonging to a religion other than those mentioned have lower identification (in all three areas: social, inclusive, national) as well as lower social and institutional responsibility and political engagement – which translates into a lower social cohesion index. Finally, people who describe themselves as spiritual (but without a religious affiliation) trust other people and political institutions less, help others less, and are less compliant with the legal regulations – resulting in a lower overall index of social cohesion.
All in all, belonging to a religious community seems to make a difference in certain areas of social cohesion. What difference it makes concretely depends on the specific religious identity and the cohesion subdimension. In general, there is a tendency among those respondents belonging to larger religious communities (Christians in Germany) to show stronger ties to the vertical level, especially concerning institutional trust and national identification. Besides that, several negative correlations emerge, particularly regarding minority groups such as Muslims, Jews, other religions, and spiritual persons.
This supports H2), that religious identity indeed makes a difference and that, in particular, the minority status of a religious group is negatively associated with social cohesion dimensions. The reasons for this would require further research and more in-depth study. Other studies have shown that socially disadvantaged (e.g., socioeconomically weaker) groups often feel less cohesion (e.g., Brand et al. 2021; DCV 2022; Follmer, Brand, and Unzicker 2020). This could be due, for example, to experiences of discrimination (e.g., from the majority) or to certain ingroup vs. outgroup dynamics. As a consequence, feelings of exclusion could translate into lower trust, a weaker sense of belonging, and less motivation or opportunities to participate in society (e.g., Pollack and Müller 2013; Tausch, Hewstone, and Roy 2009).
Furthermore, I find that the self-reported intensity of faith does not make a significant difference regarding the social cohesion dimensions (no empirical evidence for H3a). Instead, the faith content proves to be relevant. People who believe that their own religion is the only acceptable one show lower social trust and inclusive identification, which also results in a negative effect on the overall cohesion index. Those who value their religion over science have lower institutional trust. So, H3b) about the negative role of exclusivist beliefs can be partly supported. Similarly, H3c) finds empirical confirmation for several cohesion dimensions: Experiencing a punishing God relates negatively with social trust, inclusive identification, institutional responsibility, and the overall cohesion index – whereas the image of a loving God goes hand in hand with higher levels of social trust, social and inclusive identification, social responsibility and engagement, and the overall cohesion index.
In relation to religious practices, I find that private prayer practice is negatively related to institutional trust, social identification, and the overall cohesion index (H4a is partly affirmed). In contrast, service attendance shows positive correlations with almost all cohesion dimensions (except for two insignificant results with social identification and national identification). Thus, there is strong evidence for H4b).
With regard to the control variables, higher levels of identification and responsibility in all areas can be found among the women in our survey, while the men trust institutions more and are more politically engaged. Higher education is positively related to social trust, inclusive identification, social and political engagement, and the overall index. Older people show higher levels of social and national identification, institutional responsibility, and engagement at both levels – this also translates into a positive correlation with the overall cohesion index. In contrast, younger people demonstrate a higher inclusive identification. Political leftist orientations go hand in hand with higher levels of social trust, inclusive identification, social responsibility, and social and political engagement – resulting in a positive correlation with the total cohesion index. In comparison, politically right-wing people feel socially more connected, and both identify more strongly with Germany and are more law-abiding. In eastern Germany, social engagement is lower (than in western Germany). Finally, the degree of conspiracy mentality shows strong significant influences that are mainly negatively related: with both forms of trust, inclusive and national identification, social responsibility, political engagement, and the overall cohesion index. One slightly positive correlation is found with social responsibility.
A final look at the model fit shows that the different models vary in their performance to explain the variance. The adjusted R2 lies between 0.077 and 0.416 (see appendix B). This means that between 7.7% and 41.6% of the total unexplained variance of the respective dependent variable can be explained by the independent variables integrated in the individual model. The variance explanation for the overall cohesion index, for example, is about 30%.35
5.3 Discussion
Overall, the empirical findings presented show that people’s religiosity does make a difference in terms of their attitudes and behaviors that relate to societal life, even, or perhaps especially, in secular societies like Germany. However, this role becomes apparent, only or primarily, when religiosity is understood as a comprehensive concept of which different dimensions – believing, belonging, behaving – are considered. It is, therefore, crucial to analyze and understand what people specifically believe, what their relationship to the transcendent is like, how they practice their faith, and how they construct their religious identity.
The ambivalent empirical results of this and other studies (e.g., Liedhegener et al. 2019; Pickel and Gladkich 2011; Traunmüller 2012; Yendell, Hidalgo and Hillenbrand 2021) stimulate further research, such as on different religious types. For example, with respect to conspiracy theories, Pollack and Hillenbrand (forthcoming) suggest distinguishing between at least two different types of religiosity. On the one hand, an “exclusivist-privatist-punitive religious type” refers to believers whose image of God is characterized by fear, guilt, and punishment, who pray intensely in private and who attribute exclusive, absolute truth to their religion. On the other hand, the “inclusive-social-loving religious type” involves people who have a primarily loving image of God, are more inclusive toward other religions and areas of life, and practice their faith more socially with others. These types of religiosity need to be explored further, for instance, through cluster or latent class analyses. In addition, the assumption that the first type is negatively related and the second type is positively related to the cohesion dimensions could be further tested.
A great strength of this study was that the religious factor was measured and analyzed in a comprehensive and differentiated manner, including religious minorities, which often do not receive their own consideration in quantitative studies (due to insufficient case numbers). The empirical results presented here indicate that particularly with regard to religious minorities in Germany relevant results (concerning social cohesion) can be expected, and there is need for further research in this area.
However, this study also has several limitations. As our survey is not representative, the empirical results are not to be understood as general statements for the entire population in Germany (even though our empirical results fit well with the relevant research on cohesion aspects during the pandemic in Germany, e.g., Brand et al. 2021; Delhey et al. 2021; Follmer, Brand, and Unzicker 2021; Goerres and Vail 2021). There may be possible biases. Given the underrepresentation of people with lower education, socioeconomic status and from former “East Germany”, and that other representative studies for Germany have found more negative or pessimistic tendencies regarding cohesion among socioeconomically disadvantaged groups (e.g., Brand et al. 2021; DCV 2022; Follmer, Brand, and Unzicker 2020), the strength of the cohesion dimensions might be overestimated here. In further research, different socioeconomic and political groups could be considered and compared in terms of the different cohesion dimensions.
Moreover, we face the general drawbacks of quantitative research and survey designs in the social sciences, including omitted variable bias, reverse causation, and the response tendency of social desirability (Gehring and Weins 2009; Kohler and Kreuter 2012). In the absence of multiple survey waves and pre-Covid data, clear links between social cohesion and the COVID-19 pandemic cannot be established (the links are based on self-reported changes). To address these challenges, longitudinal studies, for example, and more in-depth, qualitative (case) analyses could follow, or complex models involving mediations, moderations, or structural equations could be tested.
Finally, I could only provide a snapshot of the different dimensions of social cohesion in the period from July 2020 to January 2021. Most recent studies suggest that as the pandemic continues, the “honeymoon” period of greater social cohesion fades away and the “coming-apart” scenario becomes stronger (e.g., Bertelsmann Stiftung 2022; DCV 2022). It remains to be seen and further researched how the pandemic will affect the cohesion of societies in the long run.
6 Conclusion
The theory-based empirical analysis undertaken in this article answers the research questions as follows. First, the abstract construct of “social cohesion” can be measured empirically and quantitatively by conceptualizing core dimensions on two relationship levels (horizontal: citizen – citizen / vertical: citizen – state): social/institutional trust, social inclusive/national identification, social/institutional responsibility, and social/political commitment. These can then be operationalized, for example, using survey items, which I did by carrying out a (non-representative) online survey for the German context during the COVID-19 pandemic between July 2020 and January 2021. Based on the key findings I answer the second and third research questions as follows: Regarding possible dynamics of social cohesion in Germany during the Covid pandemic, the results have shown the tendency of strengthening cohesion dimensions; the people surveyed tended to hold together more in the period studied. However, there were variations depending on the cohesion subdimension, which is why it is important to measure social cohesion in a differentiated manner. Strengthening is particularly evident in the realm of institutional trust and national identification, which provides empirical support for the popular “rally-round-the-flag effect” (Kritzinger et al. 2021). Solidarity and helpfulness also showed an upswing – but social trust declined. These empirical findings are in line with other studies on the Covid pandemic in Germany (e.g., Brand et al. 2021; Delhey et al. 2021; Follmer, Brand, and Unzicker 2020; Goerres and Vail 2021). However, this “coming-together” scenario (Borowska and Laurence 2021) can be identified especially for the initial phases of the pandemic. The later developments and final impact remain to be studied.
Finally, multivariate regression analyses have revealed that people’s religiosity plays an ambivalent role in social cohesion, thus highlighting the importance of measuring the multidimensional religious factor in a comprehensive and differentiated way. The findings show that a person’s religious identity can make a difference in terms of certain cohesion aspects. This depends on the specific religious group and especially on its minority status, which is often negatively related to the cohesion dimensions. Further research is needed, for example, on the underlying reasons or certain ingroup vs. outgroup dynamics. In addition, the specific content and practices of someone’s faith are decisive (rather than the strength of their religiosity). Exclusivist religious beliefs about the superiority of one’s religion (over other religions or areas like science) as well as an image of a punitive God and purely private prayer practice prove to be obstacles to social cohesion. In contrast, the belief in a loving God and social religious practices (service attendance) develop cohesive forces.
Overall, with the research design and empirical results for the German context presented here, I have laid out one possible way to empirically measure the abstract concept of social cohesion and to explore the role of religiosity in it. However, subsequent studies could explore and compare other ways, also including other country contexts. Ultimately, I aim to stimulate further research in this highly topical, sociopolitically relevant field, on questions of what holds societies together, even in times of crisis, and what role religion plays in this.
References
Basedau, Matthias, Simone Gobien, and Sebastian Prediger. 2018. “The Multidimensional Effects of Religion on Socioeconomic Development: A Review of the Empirical Literature.” Journal of Economic Surveys 32(4): 1106–111133. doi:10.1111/joes.12250.
Batson, C. Daniel. 2013. “Individual Religion, Tolerance, and Universal Compassion.” In Religion, Intolerance, and Conflict, edited by Steve Clarke, Russell Powell, and Julian Savulescu, 88–106. Oxford: Oxford University Press.
Berggren, Niclas, and Christian Bjørnskov. 2011. “Is the Importance of Religion in Daily Life Related to Social Trust? Cross-Country and Cross-State Comparisons.” Journal of Economic Behavior / Organization 80(3): 459–4480. doi:10.1016/j.jebo.2011.05.002.
Bertelsmann Stiftung. 2018. What Holds Asian Societies Together? Insights from the Social Cohesion Radar. Gütersloh: Verlag Bertelsmann Stiftung.
Bertelsmann Stiftung. 2022. “Umfrage Februar 2022: Erschöpfte Gesellschaft: Auswirkungen von 24 Monaten Pandemie auf den gesellschaftlichen Zusammenhalt.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/Umfrage_Erschoepfte_Gesellschaft_Feb2022.pdf.
Borbáth, Endre, Sophia Hunger, Swen Hutter, and Ioana-Elena Oana. 2021. “Civic and Political Engagement during the Multifaceted COVID-19 Crisis.” Swiss Political Science Review 27(2): 311–324. doi:10.1111/spsr.12446.
Borkowska, Magda, and James Laurence. 2021. “Coming Together or Coming Apart? Changes in Social Cohesion during the COVID-19 Pandemic in England.” European Societies 23(S1): 618–636. doi:10.1080/14616696.2020.1833067.
Brand, Thorsten, Robert Follmer, Jana Hölscher, and Kai Unzicker. 2021. “Gesellschaftlicher Zusammenhalt in Zeiten der Pandemie: Ergebnisse einer Längsschnittstudie in Deutschland 2020 mit drei Messzeitpunkten.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/ST-LW_Studie_Zusammenhalt_in_Pandemie_2021.pdf.
CEPAL (Comisión Económica para América Latina y el Caribe). 2010. “Cohesión Social en América Latina y el Caribe: Una Revisión de Conceptos, Marcos de Referencia e Indicadores” (LC/G.2420). Santiago de Chile: CEPAL. https://repositorio.cepal.org/bitstream/handle/11362/2978/1/S2010800_es.pdf.
Chan, Joseph, Ho-Pong To, and Elaine Chan. 2006. “Reconsidering Social Cohesion: Developing a Definition and Analytical Framework for Empirical Research.” Social Indicators Research 75(2): 273–302. doi:10.1007/s11205-005-2118-1.
Colletta, Nat. J., and Michelle L. Cullen. 2000. “The Nexus Between Violent Conflict, Social Capital and Social Cohesion: Case Studies from Cambodia and Rwanda.” Social Capital Initiative Working Paper No. 23. https://documents1.worldbank.org/curated/en/318921468743671186/pdf/multi0page.pdf.
DCV (Deutscher Caritasverband e.V.). 2022. “Caritas-Umfrage: Corona lässt das Wir-Gefühl in der Gesellschaft schwinden: Ehrenamtliches Engagement und Anbieter von sozialen Hilfen Stifter von Zusammenhalt – Politik abgeschlagen.” https://www.caritas.de/presse/pressemeldungen-dcv/caritas-umfrage-corona-laesst-das-wir-gefuehl-in-der-gesellschaft-schwinden-ecdfb986-eb9d-493b-bf30-3cee60f88442.
Delhey, Jan, and Klaus Boehnke. 2018. “Conceptualizing Social Cohesion in Asia.” In What Holds Asian Societies Together? Insights from the Social Cohesion Radar, edited by Bertelsmann Stiftung, 29–49. Gütersloh: Verlag Bertelsmann Stiftung.
Delhey, Jan, Klaus Boehnke, Georgi Dragolov, Zsófia S. Ignácz, Mandi Larsen, Jan Lorenz, and Michael Koch. 2018. “Social Cohesion and its Correlates: A Comparison of Western and Asian Societies.” Comparative Sociology 17(3–4): 426–455. doi:10.1163/ 15691330-12341468.
Delhey, Jan, Leonie C. Steckermeier, Klaus Boehnke, Franziska Deutsch, Jan Eichhorn, Ulrich Kühnen, and Christian Welzel. 2021. “A Virus of Distrust? Existential Insecurity and Trust During the Coronavirus Pandemic.” DSS Working Paper 80. https://www.soz.ovgu.de/isoz_media/downloads/arbeitsberichte/80_Delhey_et_al_2021_A_Virus_of_Distrust-download-1.pdf.
Devine, Daniel, Jennifer Gaskell, Will Jennings, and Gerry Stoker. 2020. “Trust and the Coronavirus Pandemic: What are the Consequences of and for Trust? An Early Review of the Literature.” Political Studies Review 19(2): 274–285. doi:10.1177 /1478929920948684.
Dickes, Paul, Marie Valentova, and Monique Borsenberger. 2010. “Construct Validation and Application of a Common Measure of Social Cohesion in 33 European Countries.” Social Indicators Research 98(3): 451–473. doi:10.1007/s11205-009-9551-5.
Dragolov, Georgi, Zsófia Ignácz, Jan Lorenz, Jan Delhey, and Klaus Boehnke. 2013. “Radar Gesellschaftlicher Zusammenhalt: Messen was verbindet: Gesellschaftlicher Zusammenhalt im internationalen Vergleich.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/GP_Radar_Gesellschaftlicher_Zusammenhalt.pdf.
Dragolov, Georgi, Zsófia S. Ignácz, Jan Lorenz, Jan Delhey, Klaus Boehnke, and Kai Unzicker. 2016. Social Cohesion in the Western World: What Holds Societies Together: Insights from the Social Cohesion Radar. Springer Briefs in Well-Being and Quality of Life Research. Cham, Switzerland: Springer International Publishing.
Durkheim, Émile. 2007 [1912]. Die elementaren Formen des religiösen Lebens. Taschenbuch / Verlag der Weltreligionen 2. Frankfurt am Main: Verlag der Weltreligionen.
ECLAC (Economic Commission for Latin America and the Caribbean). 2007. “Social Cohesion: Inclusion and a Sense of Belonging in Latin America and the Caribbean.” Santiago del Chile: ECLAC. https://repositorio.cepal.org/bitstream/handle/11362/31966/1/S2007450_en.pdf.
Esaiasson, Peter, Jacob Sohlberg, Marina Ghersetti, and Bengt Johansson. 2020. “How the Coronavirus Crisis Affects Citizen Trust in Institutions and in Unknown Others: Evidence from ‘the Swedish Experiment’.” European Journal of Political Research 60(3): 748–760. doi:10.1111/1475-6765.12419.
Follmer, Robert, Thorsten Brand, and Kai Unzicker. 2020. “Gesellschaftlicher Zusammenhalt in Deutschland 2020: Eine Herausforderung für uns alle: Ergebnisse einer repräsentativen Bevölkerungsstudie.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/ST-LW_Studie_Gesellschaftlicher_Zusammenhalt_2020.pdf.
Fonseca, Xavier, Stephan Lukosch, and Frances Brazier. 2019. “Social Cohesion Revisited: a New Definition and How to Characterize it.” Innovation: The European Journal of Social Science Research 32(2): 231–253. doi:10.1080/13511610.2018.1497480.
forsa (Gesellschaft für Sozialforschung und statistische Analysen mbH). 2022. “Gesellschaftlicher Zusammenhalt und soziales Engagement.” https://www.caritas.de/cms/contents/caritas.de/medien/dokumente/forsa-ergebnisberich/forsa-ergebnisbericht_gesellschaftlicher_zusammenhalt_ergaenzt.pdf?d=a/f=pdf.
Galen, Luke W. 2012. “Does Religious Belief Promote Prosociality? A Critical Examination.” Psychological Bulletin 138(5): 876–906. doi:10.1037/a0028251.
Gehring, Uwe W., and Cornelia Weins. 2009. Grundkurs Statistik für Politologen und Soziologen. Lehrbuch. 5th ed. Wiesbaden: VS Verlag für Sozialwissenschaften.
Goerres, Achim. 2022. “Die Coronakrise und das Verhältnis zwischen Bürger*innen und Wohlfahrtsstaat.” Präsentation Forschungsinstitut für Gesellschaftlichen Zusammenhalt.” http://achimgoerres.de/wp-content/uploads/2022/01/Goerres_Institut-fur-gesellschaftlichen-Zusammenhalt_202201120.pdf.
Goerres, Achim, and Mark I. Vail. 2021. “How National Models of Solidarity Shape Public Support for Policy Responses in the COVID-19 Crisis.” Working Paper. http://achimgoerres.de/wp-content/uploads/2022/01/Goerres-Vail-National-Models-of-Solidarity-and-the-Covid-19-Crisis_Version20211217.pdf.
Hillenbrand, Carolin. 2020a. “Religion, a Bridge or Barrier in Society? A Research Design for Empirical Analyses About the Role of Religiosity in Social Cohesion.” In Religiosity in East and West: Conceptual and Methodological Challenges from Global and Local Perspectives, edited by Sarah Demmrich, and Ulrich Riegel, 19–41. Wiesbaden: Springer Fachmedien Wiesbaden.
Hillenbrand, Carolin. 2020b. “Religion als Treiber oder Bremse für Bedrohungsgefühle? Eine empirische Analyse der Zusammenhänge zwischen Religion und migrationsbezogenen Bedrohungsgefühlen in Deutschland.” Zeitschrift für Religion, Gesellschaft und Politik 4(1): 45–79. doi:10.1007/s41682-020-00057-7.
Hillenbrand, Carolin. n.d. “Study on Religion and Social Cohesion in the Time of the Corona Pandemic.” Cluster of Excellence Religion and Politics, WWU Münster. Accessed 17 May 2022. https://www.uni-muenster.de/Religion-und-Politik/en/aktuelles/schwerpunkte/epidemien/03_thema_verschwoerung.html.
Hillenbrand, Carolin, and Detlef Pollack. Forthcoming. “Do Religiosity and Spirituality Curb or Stimulate Belief in Conspiracy Theories? An Empirical-Quantitative Analysis Conducted During the COVID-19 Pandemic in Germany.” German Society and Politics.
Hino, Hiroyuki, Arnim Langer, John Lonsdale, and Frances Stewart, eds. 2019. From Divided Pasts to Cohesive Futures: Reflections on Africa. Cambridge / New York: Cambridge University Press.
Höltmann, Gesine, and Swen Hutter. n.d. “Viel Hilfsbereitschaft, viel Protest: Polarisierung und Zusammenhalt liegen in der Pandemie nah beieinander.” Accessed 17 April 2022. https://www.wzb.eu/de/publikationen/wzb-mitteilungen/polarisierung-und-gesellschaft/viel-hilfsbereitschaft-viel-protest.
Kohler, Ulrich, and Frauke Kreuter. 2012. Datenanalyse mit Stata: Allgemeine Konzepte der Datenanalyse und ihre praktische Anwendung. 4th ed. München: Oldenbourg Verlag.
Krause, Laura-Kristine, Jérémie Gagné, and Gesine Höltmann. 2020. “Vertrauen, Demokratie, Zusammenhalt: wie unterschiedlich Menschen in Deutschland die Corona-Pandemie erleben.” Berlin: More in Common. https://www.moreincommon.de/media/gu2h03pc/more-in-common_studie-corona-zusammenhalt.pdf.
Kritzinger, Sylvia, Martial Foucault, Romain Lachat, Julia Partheymüller, Carolina Plescia, and Sylvain Brouard. 2021. “‘Rally Round the Flag’: The COVID-19 Crisis and Trust in the National Government.” West European Politics 44(5–6): 1205–1231. doi: 10.1080/01402382.2021.1925017.
Kühne, Simon, Martin Kroh, Stefan Liebig, Jonas Rees, and Andreas Zick. 2020a. “Zusammenhalt in Corona-Zeiten: Die meisten Menschen sind zufrieden mit dem staatlichen Krisenmanagement und vertrauen einander.” DIW Berlin, Nr. 49. https://www.diw.de/documents/publikationen/73/diw_01.c.793195.de/diw_aktuell_49.pdf.
Kühne, Simon, Martin Kroh, Stefan Liebig, Jonas Rees, Andreas Zick, Theresa Entringer, Jan Goebel, Markus M. Grabka, Daniel Graeber, Hannes Kröger, Carsten Schröder, Jürgen Schupp, Johannes Seebauer, and Sabine Zinn. 2020b. “Gesellschaftlicher Zusammenhalt in Zeiten von Corona: Eine Chance in der Krise?” SOEPpapers on Multidisciplinary Panel Data Research 2091. https://www.diw.de/documents/publikationen/73/diw_01.c.793201.de/diw_sp1091.pdf.
Langer, Armin, Frances Stewart, Kristien Smedts, and Leila Demarest. 2017. “Conceptualising and Measuring Social Cohesion in Africa: Towards a Perceptions-Based Index.” Social Indicators Research 131: 321–343. doi:10.1007/s11205-016-1250-4.
Lauth, Hans-Joachim, Gert Pickel, and Susanne Pickel. 2015. Methoden der vergleichenden Politikwissenschaft: Eine Einführung. 2nd ed. Grundwissen Politik. Wiesbaden: Springer VS.
Leininger, Julia, Francesco Burchi, Charlotte Fiedler, Karina Mross, Daniel Nowack, Armin von Schiller, Christoph Sommer, Christoph Strupat, and Sebastian Ziaja. 2021. “Social Cohesion: A New Definition and a Proposal for its Measurement in Africa.” Discussion Paper 31. Bonn: German Institute of Development and Sustainability (IDOS). https://www.die-gdi.de/uploads/media/DP__31.2021.v1.1.pdf.
Liebig, Stefan. 2021. “Während der Corona-Pandemie halten mehr Menschen als zuvor ihre Mitmenschen für vertrauenswürdig, fair und hilfsbereit – Ost- und Westdeutsche unterscheiden sich dabei aber.” SOEP CoV Spotlight #06. https://www.soep-cov.de/.cm4all/uproc.php/0/Spotlight_6.pdf?_=17a3d9692a0&cdp=a.
Liedhegener, Antonius, Gert Pickel, Anastas Odermatt, Alexander Yendell, and Yvonne Jaeckel. 2019. “Wie Religion ‘uns’ trennt – und verbindet: Befunde einer Repräsentativbefragung zur gesellschaftlichen Rolle von religiösen und sozialen Identitäten in Deutschland und der Schweiz 2019”. Forschungsbericht. sLuzern / Leipzig. https://resic.info/wp-content/uploads/2019/12/KONID-Bericht-2019_Religion-trennt-verbindet.pdf.
LTCP (Our Life in Times of the COVID-19 Pandemic). 2021. Online Survey. Accessed 18 May 2022. https://umfrage.ivv1.uni-muenster.de/corona-umfrage2020/. The full data set is available upon request from the author.
Manrique, Gilberto A., Araceli Sánchez, and Enrique M. Campa. 2015. “La Cohesión Social y los Límites de los Hogares en México (2008–2012).” Acta Universitaria 25(4): 48–64. doi:10.15174/au.2015.775.
Norris, Pippa, and Ronald Inglehart. 2011. Sacred and Secular: Religion and Politics Worldwide. 2nd ed. Cambridge Studies in Social Theory, Religion, and Politics. Cambridge: Cambridge University Press.
OECD (Organization for Economic Co-operation and Development). 2011. Perspectives on Global Development 2012: Social Cohesion in a Shifting World. Paris: OECD Publishing. https://www.un.org/esa/socdev/ageing/documents/social-cohesion.pdf.
Olson, Laura R., and Adam L. Warber. 2008. “Belonging, Behaving, and Believing.” Political Research Quarterly 61(2): 192–204. doi:10.1177/1065912907313075.
Pickel, Gert. 2011. Religionssoziologie: Eine Einführung in zentrale Themenbereiche. Lehrbuch. Wiesbaden: VS Verlag für Sozialwissenschaften.
Pickel, Gert, and Anja Gladkich. 2011. “Säkularisierung, religiöses Sozialkapital und Politik – Religiöses Sozialkapital als Faktor der Zivilgesellschaft und als kommunale Basis subjektiver Religiosität?” In Religion zwischen Zivilgesellschaft und politischem System: Befunde – Positionen – Perspektiven, edited by Antonius Liedhegener, 81–109. Politik und Religion. Wiesbaden: VS Verlag für Sozialwissenschaften.
Pollack, Detlef, and Carolin Hillenbrand. Forthcoming. “Umkämpftes Vertrauen: Zur Rolle von Religion, Verschwörungstheorien und Wissenschaft in der Corona-Krise.” In Zwischen Gottesstrafe und Verschwörungstheorien: Deutungskonkurrenzen bei Epidemien von der Antike bis zur Gegenwart, edited by Marcel Bubert, and André Krischer. Frankfurt am Main: Campus Verlag.
Pollack, Detlef, and Olaf Müller. 2013. “Religionsmonitor: Verstehen was verbindet: Religiosität und Zusammenhalt in Deutschland.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/GP_Religionsmonitor_verstehen_was_verbindet_Religioesitaet_und_Zusammenhalt_in_Deutschland.pdf.
Pollack, Detlef, and Gergely Rosta. 2017. Religion and Modernity: An International Comparison. Oxford / New York: Oxford University Press.
Preston, Jesse L., Erika Salomon, and Ryan S. Ritter. 2014. “Religious Prosociality: Personal, Cognitive, and Social Factors.” In Religion, Personality, and Social Behavior, edited by Vassilis Saroglou, 149–169. New York / London: Psychology Press.
Putnam, Robert D. 2001. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.
Schiefer, David, and Jolanda van der Noll. 2017. “The Essentials of Social Cohesion: A Literature Review.” Social Indicators Research 132(2): 579–603. doi:10.1007/s11205-016-1314-5.
Schiefer, David, Jolanda van der Noll, Jan Delhey, and Klaus Boehnke. 2012. “Kohäsionsradar: Zusammenhalt Messen: Gesellschaftlicher Zusammenhalt in Deutschland – Ein Erster Überblick.” Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/GP_Kohaesionsradar_Zusammenhalt_messen.pdf.
Schnabel, Annette, and Florian Grötsch. 2012. “Religion und soziale Integration – Europa als Anwendungsfall einer Mehrebenen-Analyse.” In Religiöser Pluralismus im Fokus quantitativer Religionsforschung, edited by Detlef Pollack, Ingrid Tucci, and Hans-Georg Ziebertz, 371–388. Veröffentlichungen der Sektion Religionssoziologie der Deutschen Gesellschaft für Soziologie. Wiesbaden: Springer VS.
Schnabel, Annette, and Florian Grötsch. 2014. “In God We Trust – the Role of Religion for Cohesion and Integration in Current European Societies.” European Journal of Cultural and Political Sociology 1(4): 375–398. doi:10.1080/23254823.2015.1057752.
Schraff, Dominik. 2020. “Political Trust during the COVID‐19 Pandemic: Rally around the Flag or Lock-Down Effects?” European Journal of Political Research 60(4): 1007–1017. doi:10.1111/1475-6765.12425.
Schulz, Leonie, and Rainer Faus. 2022. “Gesellschaft im Corona-Stresstest: Sorgen und Hoffnungen in Zeiten der Pandemie.” FEZ diskurs. https://library.fes.de/pdf-files/a-p-b/18758.pdf.
Sherkat, Darren E. 2015. “Religiosity.” In International Encyclopedia of the Social / Behavioral Sciences, edited by James D. Wright, 377–380. 2nd ed. Amsterdam: Elsevier.
Tausch, Nicole, Miles Hewstone, and Ravneeta Roy. 2009. “The Relationships Between Contact, Status and Prejudice: An Integrated Threat Theory Analysis of Hindu-Muslim Relations in India.” Journal of Community and Applied Social Psychology 19(2): 83–94. doi:10.1002/casp.984.
Traunmüller, Richard. 2012. Religion und Sozialkapital: Ein doppelter Kulturvergleich. Wiesbaden: VS Verlag für Sozialwissenschaften.
Urban, Dieter, and Jochen Mayerl. 2011. Regressionsanalyse: Theorie, Technik und Anwendung. Lehrbuch. 4th ed. Studienskripten zur Soziologie. Wiesbaden: VS Verlag für Sozialwissenschaften.
Vopel, Stephan, and Kai Unzicker. 2012. “Einleitung.” In Kohäsionsradar: Zusammenhalt Messen: Gesellschaftlicher Zusammenhalt in Deutschland – Ein Erster Überblick, edited by David Schiefer, Jolanda van der Noll, Jan Delhey, and Klaus Boehnke, 8–11. Gütersloh: Verlag Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/GrauePublikationen/GP_Kohaesionsradar_Zusammenhalt_messen.pdf.
Weigand, Christoph. 2009. Statistik mit und ohne Zufall: Eine anwendungsorientierte Einführung. 2nd ed. Berlin: Physica-Verlag.
Yendell, Alexander, Oliver Hidalgo, and Carolin Hillenbrand. 2021. The Role of Religious Actors in the COVID-19 Pandemic: A Theory-Based Empirical Analysis with Policy Recommendations for Action. ifa Edition Culture and Foreign Policy. Stuttgart: ifa (Institut für Auslandsbeziehungen). doi:10.17901/akbp1.10.2021.
Appendix A
List of all variables in original German wording
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Appendix B
Multivariate regression table about the relationships between religious and social cohesion dimensions
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)
Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)Citation: Religion and Development 2023; 10.1163/27507955-20230003
Source: LTCP (2021)For example: https://www.iq.undp.org/content/iraq/en/home/social-cohesion.html; https://www.mw.undp.org/content/malawi/en/home/projects/social-cohesion-project.html (Accessed June 17, 2022).
www.socialcohesion.info (Accessed June 17, 2022).
https://www.bertelsmann-stiftung.de/en/our-projects/social-cohesion/ (Accessed June 17, 2022).
https://sdgs.un.org/goals/goal16 (Accessed June 17, 2022).
In German, the term “Gesellschaftlicher Zusammenhalt” is widely used.
I conceptualize “social cohesion” as a multidimensional construct on horizontal and vertical relationship levels, including four core dimensions: trust, identity, responsibility, engagement (see section 2.1).
https://www.uni-muenster.de/Religion-und-Politik/en/index.shtml (Accessed June 17, 2022).
https://www.fgz-risc.de/das-forschungsinstitut/standorte/leipzig (Accessed June 17, 2022).
I will conduct a more extensive, cross-national comparative analysis, considering both micro and macro levels, in my forthcoming dissertation thesis. The meso level is difficult to integrate in quantitative empirical research designs due to the lack of corresponding data for groups/communities (especially when making comparisons across countries and religions). Further studies, e.g., using other (qualitative) methods and data, could shed more light on this. However, I indirectly consider the group level, for instance, with the dimensions
of religious affiliation (to which religious communities the individuals belong) and social engagement (in which associations/organizations the individuals participate).
In line with this article’s focus, the literature overview concentrates on quantitative empirical studies (related to the German context). Further publications could analyze what has been researched about the relationship between religious and cohesion aspects in other contexts, or using other (qualitative) methods (for sub-Saharan Africa see, e.g., Hino et al. 2019 and Langer et al. 2017; for Latin America see, e.g., CEPAL 2010, Manrique, Sánchez, and Campa 2015, or the various research of COES: https://ocs-coes.netlify.app/). It is important to keep in mind that the term “social cohesion” was primarily introduced by Western organizations such as the EU, OECD, or the World Bank (Delhey and Boehnke 2018, 36). Further analyses could take into account similar (indigenous) concepts from other parts of the world (e.g., Buen Vivir in Latin America or Ubuntu in Southern Africa).
In Hillenbrand (2020a) I developed a research design for analyzing the role of religiosity in social cohesion from an empirical, cross-country perspective. In this article, I apply the developed theoretical framework to the German context during the COVID-19 pandemic and test it, using primary data.
Comprehensive overviews of social cohesion research and measurement are provided, e.g., by Fonseca, Lukosch, and Brazier (2019), OECD (2011), Schiefer and van der Noll (2017), and Schiefer et al. (2012).
Less commonly, a third level is added that refers to relationships with groups/organizations (Fonseca, Lukosch, and Brazier 2019).
The definition of IDOS is: “Social cohesion refers to the vertical and horizontal relations among members of society and the state that hold society together. Social cohesion is characterized by a set of attitudes and behavioral manifestations that includes trust, an inclusive identity and cooperation for the common good” (Leininger et al. 2021, 3). In line with other studies (see table 1), my concept further specifies the dimension “cooperation for the common good” with the dimensions “responsibility” and “engagement” in order to make it quantitively measurable.
All the dimensions specified in this article are to be understood as analytical differentiations, while in reality, interconnections exist, and boundaries are fluid.
With the dimensions and concept specifications I give in this article, I do not claim to capture the complex phenomena in their entirety, nor to give universal definitions (which lies beyond this article’s scope).
“During international crises, trust in government is expected to increase irrespective of the wisdom of the policies it pursues. This has been called a ‘rally-round-the-flag’ effect” (Kritzinger et al. 2021, 1205).
The “rally-round-the-flag” effect is empirically found, for example, for Sweden (Esaiasson et al. 2020), the Netherlands (Schraff 2020), Austria (Kritzinger et al. 2021), and Germany and Great Britain (Delhey et al. 2021). A comprehensive literature overview on the Covid pandemic and trust is given by Devine et al. (2020).
Follmer, Brand, and Unzicker (2020) found that religiously affiliated people have higher values on their social cohesion index compared to non-affiliated people. In the study by Caritas (DCV 2022; forsa 2022) around 61% of both Catholics and Protestants say that they have been involved in voluntary work – among the non-denominational group, the figure is only 47%.
I focus on the classical area of “religion,” whereas broader, spiritual, or alternative forms would need to be explored in subsequent studies.
As there is no sophisticated theory of the role religiosity plays in social cohesion, and the empirical results so far remain ambiguous, I here summarize arguments and presumptions about possible relations. However, arguments can often be found for both effect directions. In such cases, I choose one side and formulate directed hypotheses – which can be empirically tested (and if necessary revised) in future empirical studies (Hillenbrand 2020a).
For reasons of space and better readability, I no longer mention the subdimensions individually in the following hypotheses, but the summary term “dimensions of social cohesion” implies this differentiated proceeding in each case.
Because I later find divergent effects depending on the cohesion subdimension in my empirical analyses, I recommend further specifying the hypotheses in subsequent studies.
Our online survey, entitled “Our Life in Times of the COVID-19 Pandemic,” will be cited in the following with “LTCP 2021.”
For example: https://www1.wdr.de/mediathek/audio/wdr5/wdr5-diesseits-von-eden/audio-glaube-pandemie-und-verschwoerung-100.html; https://www.mdr.de/religion/studie-religion-verschwoerungsmythen100.html; https://www.deutschlandfunk.de/umfrage-der-uni-muenster-kann-religion-in-der-coronakrise-100.html; https://zeitung.faz.net/faz/feuilleton/2021-01-07/debe4e24ea1837a9153741d280d26329/ (Accessed February 2, 2022).
It took around 20 minutes to answer the questionnaire. The survey was administered using a standard tool (“SoSci Survey”) with an integrated consent form and a note that the data would be used for academic research purposes only and be treated with the highest discretion and confidentiality in accordance with standard ethics protocols.
https://www.worldvaluessurvey.org/ (Accessed February 2, 2022).
https://www.bertelsmann-stiftung.de/de/unsere-projekte/religionsmonitor/ (Accessed February 2, 2022).
http://www.issp.org/menu-top/home/ (Accessed February 2, 2022).
A comprehensive table with the operationalization of all variables used (in the original German wording) can be found in appendix A.
Regarding scaling, we mostly adopted the widely used five-point scale, from which a quasi-metric scale level can be assumed (Urban and Mayerl 2011). Here and in the following, I applied explorative factor analyses for the respective (sub)dimensions to check whether the items load on the same factor and can therefore be summarized in one index. The factor analyses for the cohesion dimensions empirically justified the theoretically systematized subdimensions, with one exception: The items for the subdimension social/inclusive identification (2.1) loaded on two factors, so that I differentiate between 2.1a) social identification and 2.1b) inclusive identification.
For the change in social and political engagement, we asked separately how important the respondents consider each of these in their life since the outbreak of the COVID-19 pandemic (see appendix A).
The LTCP (2021) data will be comprehensively evaluated in the author’s doctoral thesis (probably published in 2024).
Possible biases regarding this sample composition are discussed in section 5.3.
Alternatively, I also computed regressions with the variables related to the changes of the cohesion dimensions. However, their presentation would exceed the scope of this article. They can be obtained from the author on request. Overall, similar relationships as in the models reported here are found.
I concentrate on the statistically significant results that are relevant for testing the working hypotheses. However, for space reasons, I cannot elaborate on, e.g., the comparison of effect sizes. For that, I additionally computed the standardized beta coefficients (for exact calculation and interpretation see Kohler and Kreuter 2012, 268), which can be obtained on request.
In the social sciences, these values can be considered as a moderate model explanation. However, I would like to emphasize that my research aim was not to explain the dependent variables as comprehensively as possible – which would entail searching for as many relevant influencing factors as possible – but to examine the specific relationships between religious and cohesion dimensions.
Regarding regression diagnostics, I checked for multicollinearity and heteroskedasticity. For all models, no worrying multicollinearity was found via the variance inflation factor. However, since heteroscedasticity (non-constant variance of the error term) was detected, I estimated the robust standard errors in all models because they take this issue into account (Kohler and Kreuter 2012).