1. Introduction
The recent influx of refugees into Europe, and Germany in particular, has led to intensified debates on the concomitant challenges of immigration. Manuel Castells (Reference Castells1996) has long argued that, in examining the integration of immigrants into host societies, it is necessary to consider not only government-led initiatives but also a range of other social processes that may aid immigrant social participation. Immigrant integration refers to immigrants’ equal access to resources that allow their active participation in social, cultural and economic life (Phillimore, Reference Phillimore2012). Volunteering has been suggested as one path to enhance immigrant integration (Baert and Vujic, Reference Baert and Vujic2016; Handy and Greenspan, Reference Handy and Greenspan2009).
While prior research has consistently found that volunteer participation assists immigrant acculturation, the rates at which immigrants volunteer remain significantly lower compared to native-born populations (Couton and Gaudet, Reference Couton and Gaudet2008; Manatschal, Reference Manatschal2015; Wang and Handy, Reference Wang and Handy2014). This gap raises two questions: is the introduction of policies to facilitate volunteering a useful way to accelerate immigrant integration? And, if so, where should the policy emphasis be placed? These questions may at first seem paradoxical given that volunteering is considered an autonomous social process that may not be amenable to policy interventions. However, if a policy intervention can alter the private costs or benefits of the volunteer activity, even individuals motivated by purely altruistic impulses are likely to undertake more volunteering, as net benefits increase (Handy et al., Reference Handy, Cnaan, Brudney, Ascoli, Meijs and Ranade2000). For example, subsidising transportation to the sports club may attract more volunteers among those vulnerable to transportation costs. Thus, even minor ‘third party’ policy interventions could promote volunteering (Haski-Leventhal et al., Reference Haski-Leventhal, Meijs and Hustinx2010).
Volunteering policies can be implemented at various levels (local, national) and by various actors (government, NGOs, corporations, educational institutions). For instance, Chadderton (Reference Chadderton2016) argued that structural racism often exists in the opportunities to volunteer; therefore, volunteering cannot lead ‘to social cohesion in an ethnically diverse society if racial inequalities are not addressed’ (2016: 233). But, as Handy and Greenspan (Reference Handy and Greenspan2009) argue, if nonprofit organisations actively sought to provide an immigrant-welcoming environment, it would better enable immigrants to take up volunteer opportunities. In other words, there is an interaction between an individual's free will to volunteer and the design of institutional structures, which policies could ameliorate.
In recent years, Germany has emerged as a popular migration destination. In 2015, it had the highest level of net-migration inflow, with nearly 1.3 million refugees applying for asylum (BBM, 2016). Over 15 million people with an immigrant background lived in Germany as of 2014 (Bendel, Reference Bendel2014), rising to 18.6 million in 2016 (Reuters, 2017).
Although Germany was considered a reluctant host country in the past (Triadafilopoulos and Schönwälder, Reference Triadafilopoulos and Schönwälder2006) it has, since 1998, undergone a ‘paradigm shift’ in its policies toward migration and expressed a growing recognition of the importance of immigrant integration (Bendel, Reference Bendel2014). When immigrants are not adequately integrated into their host communities, an anti-immigration public sentiment may arise (Gerhards et al., Reference Gerhards, Hans and Schupp2016), and a concern about increased competition over jobs and housing may trigger ‘defensive engagement’ or social unrest among local residents (Cook et al., Reference Cook, Dwyer and Waite2012). Since 2005, a 600-hour mandatory training course in German language instruction for non-German-speaking immigrants, including 30 hours of German law, history, and culture, has emerged as the flagship immigration policy in Germany (Bendel, Reference Bendel2014). These factors make Germany a viable context in which to examine the role of volunteering in the process of immigrant integration and its potential policy implications.
For immigrants who seek to replenish their social capital lost in the process of migration, volunteering policies could promote integration. If it can be shown that volunteering by immigrants can mitigate some of the negative sentiment around immigrant integration, where immigrants are suspected as welfare abusers or free-riders (Corrigan, Reference Corrigan2010; Osili and Xie, Reference Osili and Xie2009), or seen as a threat to local residents (Cook et al., Reference Cook, Dwyer and Waite2012), then promoting volunteering could not only help with immigrant integration but also with negative public sentiment. Indeed, prior evidence shows that connections with native-born citizens through volunteering can promote labour market integration (Kanas et al., Reference Kanas, Van-Tubergen and Van der Lippe2011).
Research questions: To understand whether immigrants in Germany engage in volunteer opportunities at the same level as native-born citizens, and to address policy-volunteering interactions, we posed three research questions:
(RQ1) Is volunteering an equally common phenomenon among immigrants and native-born Germans?
(RQ2) What are the differences between immigrants and native-born volunteers on various volunteering indicators, such as hours volunteered?
(RQ3) What socio-demographic and contextual factors are associated with the likelihood of immigrants to volunteer?
RQ2 is an important question because it analyses less-frequently addressed between-group volunteering variables, which can point to barriers and possible discrimination against immigrants in terms of volunteer opportunities. RQ3 is an important within-group question allowing us to identify the factors that might explain why some immigrants volunteer while other immigrants do not.
The paper contributes to existing knowledge by examining policy-sensitive contextual and organisational correlates of volunteering by immigrants, relying on a nationally-representative sample drawn from the German Survey on Volunteering (GSV) and using advanced statistical methods in a much-neglected European context (Garkisch et al., Reference Garkisch, Heidingsfelder and Beckmann2017).
2. Volunteering by immigrants
When immigrants arrive in a new country, especially as refugees or asylum-seekers, they face significant emotional, social and economic distress, and must undergo an arduous process of acculturation (Berry, Reference Berry1997). The challenges of relocation normally lead to lower levels of civic participation as compared to a native-born who knows the ‘rules of the game’ (Carabain and Bekkers, Reference Carabain and Bekkers2011; Manatschal, Reference Manatschal2015; Osili and Xie, Reference Osili and Xie2009). It may take years before volunteering – an activity in which ‘time is given freely to benefit another person, group, or organization’ (Wilson, Reference Wilson2012: 215) – becomes a viable option for immigrants.
Despite these challenges, research has pointed to the importance of volunteering in the integration of immigrants because it enhances their cultural networks and peer support (Howard Ecklund, Reference Howard-Ecklund2005), and it builds up bonding social capital (Putnam, Reference Putnam2000). Volunteering also provides immigrants with opportunities for skill and language development (Handy and Greenspan, Reference Handy and Greenspan2009), access to workforce experience (Baert and Vujic, Reference Baert and Vujic2016), exposure to the host country's cultural norms (Schoeneberg, Reference Schoeneberg1985), and opportunities to co-produce social services for more recent arrivals (Strokosch and Osborne, Reference Strokosch and Osborne2016). Nevertheless, immigrants may also face discriminatory anti-immigrant sentiment from the local community and limited volunteer opportunities in mainstream organisations (Chadderton, Reference Chadderton2016; Tomlinson, Reference Tomlinson2010).
2.1 Correlates of volunteering by immigrants
Gaps still exist in understanding why some immigrants volunteer while others do not. Following work on general volunteer participation (Lee and Brudney, Reference Lee and Brudney2012; Musick and Wilson, Reference Musick and Wilson2008), we first consider socio-demographic characteristics as potential correlates of immigrants’ proclivity to volunteer. But, aside from these individual-level characteristics, contextual factors might be associated with the volunteering of immigrants (Handy and Greenspan, Reference Handy and Greenspan2009). These underexplored constructs contextualize volunteering and take into account social and organisational structures within which the individual is immersed. We refer below to cultural capital, social capital and organisational affiliation as such contextual correlates.
Socio-demographic characteristics
Wilson (Reference Wilson2012) has argued that the field of immigrant volunteering requires more attention since ‘the pattern of association [between migration status and other correlates] is so complex’ (2012: 185). He linked volunteering of immigrants to variables, such as ethnicity, age, age at migration and gender (Wilson, Reference Wilson2012; Musick and Wilson, Reference Musick and Wilson2008). Unlike the general population, immigrant women are less likely to volunteer compared to immigrant men (Couton and Gaudet, Reference Couton and Gaudet2008). Similar to the general population, volunteer rates and age have had curvilinear relationships among immigrants, with rates peaking at middle age (Osili and Xie, Reference Osili and Xie2009). Higher education is positively associated with immigrant volunteering (Manatschal, Reference Manatschal2015), while full-time employees had a lower probability for volunteering because of lack of time (Manatschal, Reference Manatschal2015; Carabain and Bekkers, Reference Carabain and Bekkers2011).
Cultural capital is the long-lasting norms and values, such as ethnicity or religion, that are widely shared and deeply held within a population, and the extent to which individuals hold these shared values (Bourdieu, 1986). A culture of ‘civic voluntarism’, for example, encompasses the norms and values regarding civil society, and can be positive or constraining (Verba et al., Reference Verba, Schlozman and Brady1995). After migration, immigrants need to navigate between the cultural norms of both their home and host countries (Voicu, Reference Voicu2014). The length of time since migration affects immigrants' acculturation in the host society and thus volunteering rates often increase with length of residence (Handy and Greenspan, Reference Handy and Greenspan2009). In Germany, a major factor influencing volunteering by immigrants is their citizenship status and their identity as first- or second-generation immigrants (Simonson et al., Reference Simonson, Ziegelmann, Vogel and Tesch-Römer2016).
Attitudes towards volunteering are not equally positive across all cultures (Voicu, Reference Voicu2014; Wiepking and Handy, Reference Wiepking and Handy2015). In some cultures, volunteering is perceived as unattractive, unpaid work, and is regulated or discouraged (e.g., Xu and Ngai, Reference Xu and Ngai2009, in China; Akboga, Reference Akboga, Butcher and Einolf2017, in Turkey). Hence, not all immigrants arrive with similar readiness to undertake volunteer opportunities. They are influenced, for example, by patterns of volunteer participation in their countries of origin, and by their age when migrating (Voicu, Reference Voicu2014). It is thus important to consider immigrants’ country of origin and time since migration as proxies for cultural capital; we expect differences in volunteering proclivity depending on these measures.
Social capital refers to individuals’ social ties and their ability ‘to secure benefits by virtue of membership in social networks or other structures’ (Portes, Reference Portes1998: 6). Strong social networks are positively associated with volunteering among immigrants (Wang and Handy, Reference Wang and Handy2014). Social networks increase knowledge about volunteer opportunities and the probability of being asked to volunteer (Wilson, Reference Wilson2012; Lee and Brudney, Reference Lee and Brudney2012). The loss of social networks due to migration motivates immigrants to rebuild their social ties in the host society (Facchini et al., Reference Facchini, Patacchini and Steinhardt2015). Indeed, one of the frequently-reported motivations to volunteer among the general population is the desire to make new friends (Clary and Snyder, Reference Clary and Snyder1999). By forging social ties and social capital through volunteering, immigrants can gain access to the job market (Baert and Vujic, Reference Baert and Vujic2016; Lee and Moon, Reference Lee and Moon2011; Manatschal, Reference Manatschal2015), learn the host society's social norms of reciprocity and connectivity, and build trust in social and political institutions.
Putnam (Reference Putnam2000) suggests that social capital consists of ‘bonding’ and ‘bridging’ social capital. The former refers to social ties among members of similar groups, and the latter to social ties with members of other groups (Putnam, Reference Putnam2000). Volunteering in ethnically-homogenous or religious organisations produces bonding social capital, while engagement in mainstream organisations where immigrants interact with peers from other groups produces bridging social capital (Putnam, Reference Putnam2000). If immigrants’ sole participation is in organisations of their own ethnic origin, they may only develop ‘particularized trust’ (Uslaner and Conley, Reference Uslaner and Conley2003), which can result in ‘a lack of social integration’ (Schoeneberg, Reference Schoeneberg1985: 416). Schoeneberg (Reference Schoeneberg1985) found that the participation of immigrants in ethnic associations in West Germany helped or hindered integration, depending on the orientation of the association. However, North American studies found that immigrants who volunteered within their own ethnic congregations were simultaneously involved with mainstream organisations, thereby building both bridging and bonding social capital (e.g., Lee and Moon, Reference Lee and Moon2011; Sinha et al., Reference Sinha, Greenspan and Handy2011).
Organisational membership refers to the formal ties of individuals to organisations, and the role of these ties in mobilizing participation. Organisational membership is oftentimes obtained following a recruitment effort; it is positively correlated with, but not equal to, volunteering (Lee and Brudney, Reference Lee and Brudney2012). Several mechanisms can explain how organisational membership mobilises volunteering. First, organisational leaders can reach out to members, thus raising their awareness and stressing the importance of participation (Beyerlein and Hipp, Reference Beyerlein and Hipp2006).
Second, membership encourages participation through the interpersonal and organisational networks that recruit and incentivise others to participate (Verba et al., Reference Verba, Schlozman and Brady1995).Footnote 1 In particular, a person might have the resources and psychological willingness to participate but often remains inactive unless recruited by peer members. Indeed, in the Netherlands, being asked to volunteer partially explains the difference in volunteering between immigrants and native-born citizens (Carabain and Bekkers, Reference Carabain and Bekkers2011). Third, organisations offer their members opportunities to participate in a range of activities within the organisation, with potential transferability to external volunteering in the community (Becker and Dhingra, Reference Becker and Dhingra2001; Verba et al., Reference Verba, Schlozman and Brady1995). Lastly, members have the opportunity to volunteer in the wider community by participating in the provision of outreach services planned by the organisation. A typical example of the latter is faith-based social services (Cnaan, Reference Cnaan2002).
In a 53-country study that included Germany, Ruiter and De Graaf (Reference Ruiter and De Graaf2006) found that members of religious organisations were more likely to volunteer for both religious and secular organisations due to a strong ‘spillover effect’. This suggests that religious people are also more likely to volunteer for secular organisations. They also observed that economic development (measured by GDP) had a positive association with membership but not with volunteering, suggesting that these two constructs differ. Native-born Dutch were more likely to volunteer in secular organisations while immigrants were more likely to do so in religious organisations (Carabain and Bekkers, Reference Carabain and Bekkers2011). In examining below correlates of organisational membership to immigrant volunteering, we expect similar effects: membership in secular and religious organisations will have a positive association with immigrant volunteering.
2.2 The policy context: volunteering and immigrant integration in Germany
To contextualize our study, we first clarify the term ‘volunteering’ in Germany. There is no single term in German capturing the single English term ‘volunteering’; rather the terms Ehrenamt and Freiwilliges Engagement are both proxies for volunteering (Alscher et al., Reference Alscher, Dathe, Priller and Speth2009). These terms describe two types of engagement. Ehrenamt Footnote 2 refers to traditional volunteer roles (such as board members or sports coaches) that take place mainly in the social welfare and health systems, churches, sports clubs and fire brigades (Schürmann, Reference Schürmann2013). Freiwilliges Engagement (also referred to as ‘new Ehrenamt’) means free-will engagement, and is a broader and more flexible type of volunteering, taking place in civic-expressive organisations, like self-help groups, grassroots organisations, social movement organisations and political campaigns (Schürmann, Reference Schürmann2013).
Records from the German Socioeconomic Panel (1985-1999) showed that, on average, 23 per cent of Germans engaged in volunteering, with higher rates among men (28 per cent) than women (20 per cent) (Meier and Stutzer, Reference Meier and Stutzer2008). In the last two decades, rates of volunteering have risen to 34 per cent in 1999 and 36 per cent in 2009 (DG-EAC, 2010).Footnote 3 While volunteering is more prevalent in civic-expressive fields than in welfare and health service fields, this recent growth has occurred predominantly in the latter (Gensicke and Geiss, Reference Gensicke and Geiss2010). This distinction between civic-expressive and welfare fields is explained by the corporatist structure of the welfare regime in Germany (Anheier and Salomon, Reference Anheier and Salamon1999; Esping-Andersen, Reference Esping-Andersen1990), in which welfare services are funded by the government but provided by six large nonprofit welfare associations (Walk et al., Reference Walk, Schinnenburg and Handy2014a). Under this arrangement, known as the subsidiary principle, government funding is guaranteed by law to a selective group of nonprofit welfare organisations, who are given priority over other service providers (Zimmer et al., Reference Zimmer, Gärtner and Priller2004). These nonprofits provide services using professionals, making the role of volunteers more limited.
Since the late 1990s, the importance of volunteering to social cohesion was increasingly realised by both newly-founded civil society organisations and the traditional welfare associations. New policies and consequent public debate, supported by the German Parliament's Enquete Commission on the Future of Civic Engagement (1999-2002), were launched to increase volunteer participation (Alscher et al., Reference Alscher, Dathe, Priller and Speth2009; GHK, 2010). The Network on Civic Engagement (Bundesnetzwerk Bürgerschaftliches Engagement, BBE), founded in 2002, is a policy-oriented forum bringing together government, civil society and business actors to disseminate knowledge about, and to promote, civic engagement. Other examples of government policies to foster volunteering include the coordination of projects at national, state and municipal levels; partnerships with the business and nonprofit sectors on educational programmes and campaigns; and infrastructure to facilitate volunteering through volunteer agencies (Freiwilligenagenturen, BMFSFJ, 2016). The objective is to create incentives to foster individual engagement and strengthen voluntary action that complements government welfare provision (Alscher et al., Reference Alscher, Dathe, Priller and Speth2009). Although immigrants are not the focus of these new volunteering policies, both the BBE and the Federal Office for Migration and Refugees support these and other programmes promoting immigrant integration through volunteering. With the recent refugee inflows, renewed emphasis has been placed on policies to engage newcomers in volunteering. While several countries, for example Australia, require welfare recipients, immigrants and native-born individuals alike to engage in community service (Warburton and Smith, Reference Warburton and Smith2003), Germany does not (Bendel, Reference Bendel2014).
In summary, we lack a comprehensive understanding of not only immigrant volunteering, as compared to native-born volunteering, but also of the underlying structural influences on immigrants’ decisions to volunteer. Using large-scale survey data and advanced statistical methodology, we now turn to our investigation of the research questions outlined earlier through an analysis of the differences and determinants of immigrant and native-born volunteering and subsequent policy implications.
3. Methodology
We use the 2009 wave of the German Survey on Volunteering (GSV), a representative sample of the German-speaking population in Germany aged 14 and older (Schmiade et al., Reference Schmiade, Müller, Kausmann, Vogel, Ziegelmann and Simonson2014). GSV is funded by the Federal Ministry for Family Affairs, Senior Citizens, Women and Youth (BMFSFJ) as part of a government effort to promote civic engagement. The sample was drawn through random-digit-dialling phone interviews and stratified by state size. To study adult volunteering, we limit the analysis to individuals aged 18 or older (N=19,172).
This survey was conducted in German; thus, it only includes individuals able to communicate in this language. While the level of language competence is not measured in the survey, we expected respondents to have a minimum level of competence although interviewers could clarify questions if needed. Understandably, one might argue that language requirements may limit the representativeness of the immigrants in the sample.Footnote 4 Indeed, much of the literature on immigrant integration has indicated that language poses a barrier not easily surmounted. We acknowledge this limitation but also note its strength, because it removes the ‘language barrier’ (Carabain and Bekkers, Reference Carabain and Bekkers2011) from our possible determinants of volunteering. If past studies have suggested that language proficiency is a key barrier for immigrant integration and volunteering (Sundeen et al., Reference Sundeen, Garcia and Wang2007; Walk et al., Reference Walk, Greenspan, Crossley and Handy2014b), our data allow exploration of barriers other than language. In this respect, we expect that the differences we find might be understated. If differences are found, it reinforces the fact that even immigrants without language barriers (like postcolonial immigrants from Suriname in the Dutch case; see Carabain and Bekkers, Reference Carabain and Bekkers2011) might still face structural, organisational or cultural barriers to volunteering and social integration.
Two additional limitations are: first, the question on volunteering in the GSV was not time-dependent; people could volunteer one, a few, or many times and still respond positively to the question. Second, the data captured formal volunteering in organisations but not informal volunteering. During the current European refugee crisis, for example, much assistance to refugees was organised via social media, outside of traditional organisations (Simsa et al., Reference Simsa, Auf and Bratke2016). Thus, the GSV provides a conservative estimate of engagement in volunteering.
3.1 Measurements
Dependent Variable
‘Volunteer proclivity’ is a binary variable (1=volunteered; 0=not volunteered) measuring whether or not a respondent has volunteered in at least one organisation (Schmiade et al., Reference Schmiade, Müller, Kausmann, Vogel, Ziegelmann and Simonson2014). The variable was constructed based on aggregation of 14 binary questions on areas of volunteering. The survey question emphasised both types of formal volunteering, stating: ‘freiwillig oder ehrenamtlich’.
Independent Variables
Migration status measures whether an individual is native-born (=0) or an immigrant (=1). First-generation immigrants are those who ‘were not born in Germany, and/or do not hold German citizenship, or hold German citizenship but not since birth’ (Schmiade et al., Reference Schmiade, Müller, Kausmann, Vogel, Ziegelmann and Simonson2014: 26).
Socio-demographic variables include:
• ‘Gender’ (binary, female=0; male=1);
• ‘Age’ and ‘Age-squared’ (continuous, in years);
• ‘Education’ (originally six categories, which were dichotomized as ‘less than college education’=0; ‘college education or more’=1);
• ‘Employment’ (not employed=0; employed=1. Not employed includes: student, housewife/man, retired, army service, civil service, and other).
Cultural capital was measured as
• ‘Time since migration to Germany’ (continuous, in years);
• ‘Place of origin’ (five aggregated geographical regions: Eastern Europe, Rest of Europe, Commonwealth of Independent States (CIS), Asia, and all others).Footnote 5
Social capital was operationalised as:
• ‘Interest in politics and public life’ is a categorical variable in response to the question: ‘How strongly are you interested in politics and in public life?’ with response options: strongly=3, neutral/medium=2, little=1.
• ‘Social Network’ is a categorical variable in response to the question: ‘How big is your circle of friends and acquaintances?’ with response options: very large=3, medium=2, rather small=1).Footnote 6
Organisational membership was operationalised as:
• ‘Membership in secular organisations’ (binary, no membership=0; membership in at least one of 10 organisations=1),Footnote 7
• ‘Membership in religious organisations’ (binary, no religious membership=0, membership in religious organisation=1).
3.2 Analytical procedure
Between-group analysis
We first test the differences between immigrants and native-born individuals in terms of the overall ‘volunteer proclivity’ using the full sample (Table 1), and then the differences between several indicators of volunteer behaviour (age when first volunteered, volunteer hours, areas of volunteering, volunteer tasks, volunteer labelling, and volunteer initiative) on the sample of those who indicated that they had volunteered (Table 2). In Table 1, we also compare volunteering with a measure of the charitable giving of the population, as these behaviours are often correlated (Osili and Xie, Reference Osili and Xie2009).Footnote 8
TABLE 1. Differences in volunteering and giving between native-born and immigrants
Note: mean values, SD in parentheses. *p<.05, **p<.01, ***p<.001, ****p<.0001.
TABLE 2. Volunteering indicators of native-born and immigrant volunteers
Note: mean values, SD in parentheses. *p<.05, **p<.01, ***p<.001
The group differences were analysed in two ways. First, chi-square tests (categorical variables) and t-tests (continuous variables) were used to test for differences in the full sample and among volunteers only (N=6,747). Second, given the relatively small proportion of immigrants in the full sample (8 per cent), we matched a paired sample of the immigrants and native-born populations using the propensity score matching method (Austin, Reference Austin2011). Propensity score matching accounts for systematic differences between groups, thus reducing confounding. In this method, every individual in the treatment group (immigrants) is matched with the most compatible individual in the comparison group (native-born) following a calculation of propensity score based on pre-determined individual-level characteristics. We matched on gender, age, education, employment status and religious membership. This procedure reduces the selection effect because it ensures that similar immigrants and native-born individuals are compared (Caliendo and Kopeinig, Reference Caliendo and Kopeinig2008). It has been applied in past research to immigrant populations (Thomsen et al., Reference Thomsen, Walter and Aldashev2013) and volunteering (Hong and Morrow-Howell, Reference Hong and Morrow-Howell2010). Appendix 1 (online) presents the full 5-step propensity score procedure.
Within-group analysis
For testing the differences between immigrant volunteers and non-volunteers, we conducted multivariate analysis using logistic regression predicting the ‘volunteer proclivity’ of immigrants in the full sample (Table 3). We separately tested socio-demographics (Model 1), cultural capital (Model 2), and social capital and organisational membership (Model 3). We tested the variables for multicollinearity; all Variance Inflation Factors were below 2.5 with the exception of age and age-squared, as expected. Descriptive statistics with bivariate correlation matrix are available in Appendix 2 (online).
TABLE 3. Ordered logistic regression predicting immigrants’ proclivity to volunteer
Notes: +<.10, *p<.05, **p<.01, ***p<.001, ****p<.0001. a entries are odds ratios; standard errors in parentheses. b reference category: Eastern Europe
4. Findings
Table 1 shows that rates of volunteering among immigrants in the full sample are significantly lower compared to the native-born population (25 per cent and 36 per cent respectively). The two groups also differ in their giving behaviour: only 51 per cent of the immigrants donated in the last 12 months as compared to 59 per cent among the native-born population. However, in the matched sample, the difference in volunteering remains high and statistically significant, while the difference in giving loses its level of significance.
We examined differences between immigrant and native-born volunteers on several indicators. Table 2 shows that, for most measures, the difference in the total sample of volunteers was not statistically significant. In the matched sample, we find a few significant differences between native-born and immigrant volunteers in: age when first volunteered (immigrants begin volunteering later than native-born volunteers), areas of volunteering (native-born individuals volunteer more in sports and recreation while immigrants volunteer more in cultural organisations), and volunteer tasks (immigrant volunteers offer personal assistance more often than native-born). Interestingly, there are also significant differences in the choice of words used to refer to volunteering: more native-born volunteers than immigrant volunteers use the term Ehrenamt. Despite these few differences, it appears that immigrants who do volunteer undertake volunteering behaviours quite similar to native-born volunteers.
We conducted within-group analysis of the correlates of volunteer behaviour among immigrants using logistic regression (Table 3). The results demonstrate that in Model 1 only the level of education was significant; immigrants with college education were 2.04 times more likely to volunteer as compared to those without (OR=2.04, p<.0001). This effect carries over into Models 2 and 3 even when contextual factors were added.
In Model 2, in which cultural capital is added, immigrants from Eastern European countries are significantly more likely to volunteer compared to immigrants from Asia and the CIS. Time since migration is also found to be significant, suggesting that the longer immigrants live in Germany, the more likely they are to volunteer. The effect size is noteworthy, as every additional year in the host country adds three per cent to the odds of volunteering (OR=1.03, p<.0001). However, when other contextual variables are added (Model 3), time since migration loses some of its power.
Model 3 adds social capital and organisational membership variables, all of which are found to have a significant positive effect on immigrants’ proclivity to volunteer. Most notable is the large effect size of membership in secular organisations. Immigrants who are members of secular organisations are about six times more likely to volunteer compared with non-members (OR= 6.01, p<.0001). Several robustness checks confirmed these findings (see Appendix 3 online).
5. Discussion and policy implications
Our study provides a more nuanced understanding of the differences between immigrant and non-immigrant volunteering behaviour, and of the differences within the immigrant population between volunteers and non-volunteers. It contributes to the discussion on the possibility of developing policy interventions to promote volunteering as a tool for immigrant integration.
Similar to findings in other countries (e.g., Carabain and Bekkers, Reference Carabain and Bekkers2011 in the Netherlands; Osili and Xie, Reference Osili and Xie2009, in the US; Qvist, Reference Qvist2017, in Denmark; Voicu, Reference Voicu2014), our study first confirmed that immigrants in Germany engage less in volunteering than native-born individuals. This finding is particularly robust in our analysis given the use of a propensity score-matched sample. The reported levels of volunteer proclivity are strikingly similar to records in other German surveys, where 38.1 per cent of native-born Germans reported volunteering, compared to 17.6 per cent of first-generation immigrants and 25.9 per cent of second-generation immigrants (Facchini et al., Reference Facchini, Patacchini and Steinhardt2015). Furthermore, while gaps between immigrants and native-born individuals on levels of charitable giving disappeared in the matched sample, the differences in levels of volunteering remained significant.Footnote 9 This makes a case for promoting specific policies to encourage volunteering among immigrants.
Results on the correlates of volunteering among the immigrant population (Table 3) suggest that contextual factors are more influential in driving immigrant volunteering. With the exception of education, socio-demographic characteristics were not associated with immigrants’ proclivity to volunteer, while contextual factors – especially social networks and membership of organisations –were associated with this outcome. This is an important finding because it suggests that greater emphasis should be given in policy design and interventions to examining contextual factors of immigrant volunteering at the organisational level.
5.1 Organisational adaptations
We find a positive association between membership of secular and religious organisations and volunteering. These results are good news for policymakers since they indicate that membership of organisations is a salient avenue for immigrant volunteering and, we believe, organisational membership makes individuals more amenable to organisational interventions. Organisations can target immigrants in their volunteer recruitment efforts by emphasising positive integration benefits, such as social ties, language skill development, work experience and well-being. If these benefits are labour-market related or have to do with children's social integration, they are more likely to motivate newcomers to volunteer.
One can envision, for example, that sports organisations, which are dominant in Germany, can offer membership incentives and volunteering opportunities to younger immigrants (Cuskelly, Reference Cuskelly2004). Another example is the grassroots community organisation, like community gardens, whose communal nature might be attractive for immigrants who come from collectivist cultures. To ensure that such opportunities exist, government policies that work in tandem with existing membership-based nonprofits would need to be designed, especially around cultural, sports, and community activities, where language barriers are less of a concern.
These examples notwithstanding, one has to remember that in many membership organisations, only a minority of the members volunteer. Therefore, without active recruitment of immigrants by organisational leadership (for example, through the four mechanisms described above), immigrants will still experience a lower likelihood of being asked to volunteer, a lower likelihood of being accepted to volunteer, and a higher likelihood of dropping out.
Diverging from past research, a larger effect size of membership is found in secular organisations compared to religious organisations.Footnote 10 A supply-side explanation can possibly account for this divergence: There might not be as many religious organisational opportunities for (non-Christian) immigrants to volunteer given that religious congregations in Germany do not follow the North American Congregationalist model, where congregations function as community centres, social clubs and convenient venues for volunteering (Cnaan, Reference Cnaan2002; Howard-Ecklund, Reference Howard-Ecklund2005). An alternative explanation is the lower level of religiosity in Germany, even among immigrants (Eicken and Schmitz-Veltin, Reference Eicken and Schmitz-Veltin2010).
The centrality of contextual factors suggest that immigrants’ lower levels of participation are grounded in lack of equal opportunities and discriminatory, or at least inequitable, practices more than personal barriers. For instance, as Chadderton (Reference Chadderton2016) has demonstrated with regards to the Technical Relief Service, a Federally-created volunteer-oriented agency, the German organisational system is embedded in ‘racialized structures’ (2016: 239) and assumptions of immigrant's cultural deficit, resulting in attitudes and norms that can contribute to the exclusion of immigrants from membership organisations and volunteer opportunities. Thus, overcoming structural barriers to increase levels of immigrant membership and volunteering will not be an easy task. A concerted effort to address institutional barriers within nonprofit organisations, such as language, cross-cultural adaptations and racial and ethnic biases, is required to help recruit immigrants first as members and subsequently as volunteers (Simonson et al., Reference Simonson, Ziegelmann, Vogel and Tesch-Römer2016).
5.2 Policy interventions
Given that time since migration was found to positively influence immigrant volunteering, and given that immigrants are responsive to recruitment efforts (more than half of them reported volunteering after ‘being asked’), it is advisable to incentivise long-term immigrants to volunteer with newcomers, for example in language translation needs. With the recent refugee influx, a new trend is emerging, in which high numbers of volunteers supporting newcomers are veteran immigrants themselves (Karakayali and Kleist, Reference Karakayali and Kleist2015; Simsa et al., Reference Simsa, Auf and Bratke2016), thereby giving new opportunities for volunteering. This serves three purposes: first, it exposes current immigrants to the local culture of volunteering; second, it becomes a high-potential mentoring programme; and third, it provides role models for future volunteering as newcomers get settled and feel ready to reciprocate. Such an arrangement is a good example of co-production of services by volunteers and service recipients (Strokosch and Osborne, Reference Strokosch and Osborne2016).
Our findings, that immigrants and native-born volunteers were similar on most volunteering indicators, suggests that the main barrier for immigrants’ volunteering lies in their transition from non-participation to participation. In other words, if the participation threshold is crossed by immigrants, the differences in other indicators of volunteering are reduced (Haski-Leventhal et al., Reference Haski-Leventhal, Meijs, Lockstone‐Binney, Holmes and Oppenheimer2017). The challenge for immigrant integration through volunteering is not about what volunteer activities immigrants undertake, but more about whether or not they are able to cross the participation threshold. This insight makes it imperative to design policies that encourage immigrants’ initial participation in volunteering and remove systemic obstacles (Chadderton, Reference Chadderton2016).
While the data predate the current refugee crisis, one policy intervention might include incentives for organisations that are able to draft more immigrants as volunteers, with the intent of benefitting both the organisation and the immigrants. Another intervention could be the strengthening of the 500 existing volunteer match organisations (‘Freiwilligenagenturen’ in German, BAGFA, 2018). These semi-governmental agencies link potential volunteers with nonprofits but have not, so far, given specific attention to recruiting immigrants as volunteers. If done with cultural-sensitivity, the potential for success is high, especially since many of these organisations establish projects targeting immigrants (Backhaus-Maul and Speck, Reference Backhaus-Maul and Speck2011).
Policies targeting immigrants’ first-time volunteering may have a greater pay-off than other types of assistance. Such policies include training offered to immigrants shortly after arrival about existing volunteering opportunities and the expected benefits from such participation. A norm of reciprocity can be particularly appealing to immigrants (Manatschal and Freitag, Reference Manatschal and Freitag2014). This norm can motivate specific groups to engage in volunteering ‘such as people with low levels of education or immigrants, two groups that would otherwise have a low propensity to volunteer’ (Manatschal and Freitag, Reference Manatschal and Freitag2014: 226–227). It is likely that due to a desire to reciprocate, long-term immigrants will be motivated to volunteer alongside newer immigrants at the organisations from which they received help. Such an opportunity could be beneficial to the newcomer, to the established immigrant volunteer, and to the hosting organisation.
An alternative policy direction is to target negative public sentiment toward immigrants by utilising educational programmes and campaigns that portray the positive contributions of immigrants to civic participation. Yet, government advancement of volunteering should be taken with care, given that excessive involvement may raise concerns that the nature of volunteering is too aligned with government agendas, or seen as mandated service and thus not purely voluntary (Haski-Leventhal et al., Reference Haski-Leventhal, Meijs and Hustinx2010).
6. Conclusions
Immigrants have been largely overlooked as volunteers in German civil society, and volunteering is still a neglected tool in integration policy and practice. This is unfortunate, since volunteering by immigrants is a win-win situation for immigrants and their host countries. On the one hand, immigrants are an untapped resource in civil society and can potentially contribute to its strength, diversity and multicultural nature (Bendel, Reference Bendel2014). On the other hand, volunteering has implications for immigrant and refugee integration outcomes and thus for social policy because volunteering offers multiple cultural, economic and personal benefits for the immigrants.
A key finding from this study is the importance of the contextual factors in facilitating immigrant volunteer opportunities, particularly organisational membership in organisations other than congregations. This finding is perhaps somewhat encouraging, given that decision-makers and leadership in both nonprofit organisations and government are more likely to influence change in organisational settings than in the values, attitudes and behaviour of individuals.
In order to gain a better understanding of the value of volunteering in specific fields of immigrant activity, future research could disaggregate the secular membership variable into its 14 sub-fields. Sports, for example, is a popular area of volunteering in Germany, and could yield interesting results with regard to immigrant volunteering (BMFSFJ, 2016). Furthermore, a classification of citizenship, beyond the binary categorization of native and immigrant, is also likely to yield important observations that the native/foreign-born dichotomy cannot capture. Likewise, future studies could examine the effects of generational status on immigrants’ propensity to volunteer and ability to utilise volunteering as a means for social integration. Finally, better specification of what constitutes membership (attending meetings, paying dues, receiving organisational updates) could offer a more-nuanced analysis.
Acknowledgements
We are thankful to the German Centre of Gerontology for providing us with access to the German Survey on Volunteering. We also thank Anne Stefanie Ruiz Sportmann for her invaluable research assistantship. We are grateful for constructive comments on previous versions received from two anonymous reviewers, the Journal editors, and participants at the Philanthropy Research Workshop at the Lilly Family School of Philanthropy at IUPUI and the Emerging Scholar Roundtable at ARNOVA. Partial funding for this project was made available through the International Partnership Development Fund from the Office of International Affairs at IUPUI.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0047279418000211
