Volunteering for Refugees and Asylum Seekers in Greece

Since 2015 the massive refugee influx has challenged European countries in terms of tackling one of the largest movements of displaced people through European borders since the World War II. Greece became the major entry point by sea receiving massive inflows of refugees and asylum seekers arriving in the Greek shores. While European leaders debated on how to deal with the refugee crisis and traditional donors delayed funding, the role of volunteers has been instrumental in covering basic and urgent needs. Based on a hybrid approach of sociological and political determinants, the study enriches the scarce empirical evidence on volunteering specifically for refugees and asylum seekers, by portraying the profiles of volunteers providing solidarity to refugees and asylum seekers fleeing repression and conflict zones to seek safety in Europe.

refugees/asylum seekers in Greece reached the record figure of 821,000. The bulk of the flow was directed towards the Greek islands bordering Turkey (IOM 2015). 3 The large-scale arrival of refugees/asylum seekers and the resulting transformation of the migrant landscape in the country have challenged Greeks to cope with a dual crisis: the current refugee crisis as well as the economic depression which has severely affected the country over the last six years.
While the European response has been characterized by confusion and lack of universal policy (Tramountanis 2017) and traditional donors delayed funding, thousands of ordinary people have joined efforts to provide services and support to refugees/asylum seekers arriving to Greek shores. The role of volunteers in responding to the refugee crisis has been remarkable. Volunteers have provided a plethora of solidarity activities including food supplies, collecting and sorting clothes, providing medical aid, legal and financial support, rescuing people from the sea, cooking, setting up laundries, building shelters and so on (Evangelinidis 2016;Gkionakis 2016;Latimir 2016). Several media reports emphasize that despite the acute economic crisis, volunteers in Greece have stepped into covering for the gap left by the Greek state and EU leaders to support for refugees' humanitarian needs. 4,5 Previous research has consistently underpinned the lower levels of volunteering in Greece (e.g. European Commission 2007 along with a weaker civil society (Mouzelis 1995;Lyrintzis 2002) compared to other European countries. Despite such arguments, other scholars emphasize that there is a vibrant, informal, non-institutionalized and often non-registered Greek civil society sector which does not fall into the normative definitions (Karamichas 2007;Rozakou 2011). This informal civil society usually tends to be distant from the state and primarily aims to protect vested interests in specific local areas or volunteer to help people in need (Sotiropoulos 2004).
The main rationale of the chapter is to explore volunteering for refugees/asylum seekers which is defined in the present study as active membership in an organization (volunteering in an organization) to support the rights of refugees/asylum seekers in Greece. Greece becomes an interesting case of investigating volunteering for the specific vulnerable group as in the context of the recent refugee crisis the country has experienced an unprecedented influx of refugees/asylum seekers entering its territory en route to wealthier countries. Moreover, as shown in Table 7.1, volunteering for refugees/asylum seekers is higher in Greece (8.1%) compared to other countries participating in TransSOL project. However, despite its higher prevalence crossnationally, in Greece fewer individuals volunteer for refugees/asylum seekers than for other vulnerable groups such as the unemployed (9.3%) and disabled (10.3%).
Although past international research has produced numerous and valuable insights into volunteering, the domain of volunteering specifically for refugees/asylum seekers has been little explored. Exceptions involve Erickson's study (Erickson 2012) in Fargo, North Dakota, during 2007-2008, which investigates how volunteers embrace and contest hegemonic forms of "worthy" citizenship. A study conducted in Hungary shows that the current refugee crisis has a strong mobilizing effect for almost 3% of the Hungarian population; some volunteers have altruistic motivations, whilst others are mainly driven as a response to the political situation (Toth and Kertesz 2016). In Germany, recent empirical evidence underpins that since 2015 volunteering for refugees/asylum seekers has become a widespread phenomenon with thousands of people donating money and distributing food, medicines, clothing and other essentials Kleist 2015, 2016;Rose 2016).
The recent explosion of refugees/asylum seekers fleeing conflict and persecution and the pivotal role of volunteers to tackle the refugee crisis Table 7.1 Volunteering for refugees/asylum seekers, unemployed and disabled in countries participating to TransSOL project has led to a drastic increase of scientific interest in the field. Although the other chapters in the book explore different forms of civic engagement, activism or "solidarity" practices to different vulnerable groups, 6 the present chapter focuses on volunteering specifically for refugees/asylum seekers in Greece. Based on a hybrid approach which combines the sociological and political approaches to volunteering, the analysis is guided by the research question of "Who volunteers for refugees/asylum seekers in Greece?" The findings shed some light on our research question by portraying the profiles of volunteers helping refugees/asylum seekers arriving in the country.

theoretIcal Background and hypotheses
Volunteering embraces different functions and explanations in different disciplines (Musick and Wilson 2008;Hustinx et al. 2010). For instance, sociological approaches emphasize different forms of capital or resources, such as human, social and cultural capital in explaining volunteering (Wilson and Musick 1997a;Wilson 2000). Psychologists focus on key traits of personality such as extraversion, agreeableness and resilience that impact on individuals' predisposition to volunteer (Bekkers 2005;Matsuba et al. 2007). For political scientists, volunteering acts as a critical form of civic engagement and an expression of democratic values (Theiss-Morse and Hibbing 2005), underlying the critical impact of citizens' political engagement on volunteering (Bekkers 2005). Economic scientists adopt a rational-based approach, viewing volunteering as a form of unpaid labour where volunteers undertake activity depending on the consuming resources and the rewards they may gain (Wilson 2000;Musick and Wilson 2008;Hustinx et al. 2010;Wilson 2012).
The present paper adopts a hybrid approach. It explores the profiles of individuals volunteering for refugees/asylum seekers in Greece based on their demographic attributes, their human, social and cultural capital developed in sociological approaches and their political conventional and unconventional behaviours developed in political approaches.
Literature has shown that, generally, people with different demographic characteristics vary in their propensities to volunteer (Wilson 2000;Musick and Wilson 2008;Hustinx et al. 2010). With respect to gender, previous research shows different rates and patterns of volunteering (Wilson 2012). Gaskin and Smith (1997) suggest there is no clear pattern of gender differences in volunteering across European countries. However, other scholars suggest that gender does make a difference in specific domains of volunteering, since women tend to have higher rates in informal volunteering activities associated with more caring tasks and lower rates in political activities (Thompson 1993;Schlozman et al. 1994;Cnaan et al. 1996;Rochester et al. 2010). This pattern appears quite consistent across different age groups and countries (Wuthnow 1995). Gender ideologies, as well as the gendered division of labour, partly explain why women tend to volunteer more in activities associated with caring tasks (Wilson 2000).
Age-related variables are also important in determining volunteering. Some scholars underpin that voluntary participation varies by age or lifecycle stage associated with the different adult roles (e.g. with work, family obligations) taken throughout the life cycle (Wilson 2000;Musick and Wilson 2008;Smith and Wang 2016). The empirical evidence shows that volunteering is generally higher among middle-aged citizens compared to the elderly and youth (Wymer 1998;Curtis et al. 2001;Pho 2008). However, Wilson (2000) argues that high-risk volunteering activities primarily attract younger people compared to older people. Moreover, he suggests that different types of volunteering activities become more or less attractive in different life-cycle stages. For instance, younger citizens mainly volunteer in organizations related to self-and career-oriented activism, middle-aged volunteers primarily engage in community-oriented work, whereas older volunteers participate to "service organizations, recreational clubs and agencies to help the elderly" (Wilson 2000, p. 227).
A plethora of scholars emphasize that, at least in advanced industrial societies, education is often the most consistent predictor of volunteering Nie et al. 1996;Wilson 2000Wilson , 2012Musick and Wilson 2008;Huang et al. 2009;Rochester et al. 2010;Van Ingen and Dekker 2011). The critical impact of education on volunteering is associated with the outcomes of educational processes that expose individuals to norms and values favourable to volunteering as well as to civic skills, advanced awareness of problems and stronger feelings of efficacy.
From a sociological perspective, individuals' decision to volunteer is influenced by various types of resources or capital, such as human capital, cultural capital and social capital (Wilson and Musick 1997a;Wilson and Musick 1998;Wilson 2000;Musick and Wilson 2008). Human capital primarily involves income, occupational class and employment status (Wilson and Musick 1997a;Wilson 2000). Most empirical evidence across different countries reports that low-income earners are less likely to volunteer than higher earners (Vaillancourt 1994;Freeman 1997;Hurley et al. 2008). For instance, Pho (2008) explored volunteering in the United States from 2002 to 2005 and found that low-to medium-wage earners are less likely to volunteer than high-wage earners.
Whether or not someone is employed and the nature of their employment can influence volunteering in several ways. Employment is a prime determinant of social status, it provides opportunities to integrate into society and develop those adequate civic skills that increase the likelihood of volunteering. The relation between employment status and volunteering has been explored by various scholars, underlying that part-time employees are more likely to volunteer than either full-time employees or individuals who are not in the labour force (Johnson et al. 2004;Lasby 2004;Low et al. 2007;Hurley et al. 2008;Einolf 2011). Meanwhile, unemployment status is usually associated with lower levels of volunteering (Pho 2008;Wilson 2012).
In Wilson's (2000, p. 221) words: "As occupational status increases so does the likelihood of volunteering." Occupational status has been shown to play a critical role in volunteering (Wilson and Musick 1997b;Hodgkinson 2003;Rotolo and Wilson 2007). For instance, Reed and Selbee (2001) found that individuals in Canada with jobs high in occupational prestige, higher income and higher educational attainment are more likely to volunteer. Similarly, Rotolo and Wilson (2007) show that even after controlling for family traits, women with professional and managerial occupations exhibit greater tendencies to volunteer than women in lower occupational jobs. The association between volunteering and high occupational prestige is related to the fact that top managers or professionals are more likely to be asked to volunteer as well as to be socially active as part of their job role (Wilson 2000;Wilson and Musick 1997b).
For Putnam (2000, p. 19), social capital refers to "social networks and the norms of reciprocity and trustworthiness that arise from them". The key resources that form "social capital" involve social networks or social ties, including friendship networks as well as trust in others, that is, elements which tend to foster collective action (Wilson and Musick 1997a). Much research has been conducted on social capital in recent years, in particular measured as individuals' friendship networks, informal social interactions and social trust, as correlates of volunteering. For instance, Wilson and Musick (1997a) found a positive association between formal volunteering and informal social interactions measured as frequent conversations and meetings with friends and acquaintances. Brown and Ferris (2007) found that individuals' associational networks, their trust in others and in their community are important determinants of giving and volunteering. Cross-national surveys underpin that social trust is positively associated with volunteering regardless of socio-economic differences (Anheier and Kendall 2002). It should be noted that some scholars underline that social trust is associated with specific types of volunteering activities which primarily target to provide services to individuals in need. On the contrary, trusting people are "less likely to volunteer in activities that involve confrontation with authorities or working to change the system" (Musick and Wilson 2008, p. 46). In line with such arguments, Greenberg (2001) supports that politically oriented volunteering associated with government-related activities, among others, is motivated by lack of social trust, whereas service-oriented volunteering including non-governmental activism is motivated by trust in others.
In Wilson and Musick's (1997a) integrated theory of volunteer work, religiosity is an indicator of cultural capital which is positively associated with formal volunteering. A cultural capital perspective posits that religiosity provides an ethic of caring which reinforces the decision to volunteer (Wuthnow 2004). As most religions encourage altruistic values, highly religious individuals are more concerned with the welfare of others (Dillon et al. 2003); therefore their value preferences are more compatible with volunteering. Previous research underpins that more religious individuals are more likely to be involved in volunteering than their secular counterparts (Wilson and Janoski 1995;Becker and Dhingra 2001;Musick and Wilson 2008;McNamara and Gonzales 2011).
The political approach to volunteering highlights its role as a form of civic engagement and expression of democratic values. Putnam argues that "volunteering is part of the syndrome of good citizenship and political involvement" (Putnam 2000, p. 132). Different scholars report that volunteers tend to be more politically active compared to non-volunteers Dekker and Van den Broek 1998;Hodgkinson 2003;Musick and Wilson 2008). The grounds of the association between volunteering and political engagement involve, among others, the opportunity to develop specific civic skills (such as the ability to organize a meeting), sharing information and fostering general trust Stolle 1998). Hodgkinson (2003), in her study using EVS/WVS 1999-2002 data, found that volunteers are more likely to be politically engaged (in terms of discussing politics and signing petitions) than non-volunteers in the vast majority of the countries under study. 7 Dekker and Van den Broek (1996), using data from five countries (the United States, Great Britain, West Germany, Italy and Mexico), found that active volunteers compared to passive volunteers 8 are more likely to be politically engaged in conventional and unconventional political acts (such as contributing time to political organizations, participation to protests/demonstrations, etc.). 9 Bekkers' (2005) study in the Netherlands shows that individuals with a greater interest in politics and post-materialistic value orientations are more likely to be volunteers-also, voting preferences are important since non-voters are less likely to volunteer than voters who prefer leftist or Christian political parties. Similarly, Knoke (1990) found that active volunteering goes along with being active in local politics, including among others, voting in local elections.
Drawing on the theoretical arguments and the empirical evidence discussed, we hypothesize the following: Hypothesis 1 Individuals with specific demographic attributes, that is, women, middle-aged and higher educated individuals, are more likely to volunteer for refugees/asylum seekers.
Hypothesis 2 Individuals' human capital in terms of higher income and occupational class is positively associated with volunteering for refugees/ asylum seekers. Moreover, part-time employees are more likely to volunteer than either full-time employees or individuals who are not in the labour force.
Hypothesis 3 Individuals' social capital, in terms of social trust and informal social interactions with friends, is positively associated with volunteering for refugees/asylum seekers.

Hypothesis 4
Individuals' cultural capital, in terms of religiosity, is positively associated with volunteering for refugees/asylum seekers.
Hypothesis 5 Individuals' political engagement in conventional and unconventional political behaviours is positively associated with volunteering for refugees/asylum seekers.

data and Methods
The analysis draws on an original dataset of n = 2061 respondents (aged 18+) in Greece matched for age, gender, region and education level quotas to national population statistics. To explore the profiles of individuals volunteering for refugees/asylum seekers, specific items from the project's questionnaire are used. The dependent variable, that is, volunteering for refugees/asylum seekers, is measured with one item asking respondents, among others, 10 their active membership in an organization (volunteering in an organization) to support the rights of refugees/asylum seekers. The dichotomous variable is used to capture volunteering and non-volunteering for refugees/asylum seekers in Greece.
The independent variables involve a set of items capturing demographic characteristics, including gender, age and education, measures of human, cultural and social capital as well as individuals' political behaviour. 11 Age is measured with an ordinal scale including three broad age groups, that is, 18-34, 35-54 and over 55 years old. Educational attainment is measured with three responses capturing individuals with higher education (i.e. university and above), intermediate education (i.e. upper secondary and post-secondary non-tertiary education) and lower education (i.e. less than primary and lower secondary education).
Human capital is measured with indicators capturing respondents' income, employment status and occupational class. Income is measured with an item asking respondents for their household monthly net income, after tax and compulsory deductions, from all sources. The recoded variable includes three responses measuring lower (i.e. less than 775 euro), middle (i.e. between 776 and 1425 euro) and higher income earners (i.e. more than 1426 euro). Respondents' occupational class is measured with a recoded variable including three responses: higher occupational class (professional/managerial workers), middle (clerical/sales or services/foreman or supervisor of other workers) and low (skilled/semi-skilled or unskilled manual workers) occupational class. Respondents' employment status is measured with a recoded variable including four responses: full-time employee, part-time employee, other employment status (such as permanently sick or disabled, retired, community or military service, doing housework, looking after children or other persons) and unemployed.
Social capital is measured with indicators associated with respondents' social trust and informal social interactions. The former is captured with one item measuring on a scale from 0 to 10 respondents' level of trustfulness of people, where higher scores indicate higher levels of social trust. The intensity of informal social interactions is measured with one item asking respondents how often, in the past month, they met socially with friends not living in their household. The recoded variable is a dichotomous measure including "Once or twice this month or less" and "Every week or almost every day". The former captures respondents' low intensity of informal social interactions and the latter high intensity of informal social interactions.
Cultural capital is measured with one item capturing religiosity. Specifically, respondents are asked to report on a scale from 0 to 10 how religious they are, with higher scores indicating stronger religiosity.
Political engagement is measured with items capturing involvement in conventional and unconventional political behaviours. The former is measured with a question asking respondents if they voted or not in the last Greek national election (on 20 September 2015). The recoded dichotomous variable captures respondents' engagement or non-engagement in conventional behaviours. Unconventional political behaviour is measured with an additive score based on responses on five items measuring participation in the past 12 months in (a) signing petitions, (b) boycotting products for political/ethical/environmental reasons, (c) attending a demonstration, march or rally, (d) joining a strike and (e) joining an occupation, sit-in or blockade. 12 In the composite index, higher scores indicate higher levels of respondents' involvement in unconventional political behaviour.
The analysis uses exploratory and explanatory analyses to investigate volunteers' profiles supporting refugees/asylum seekers in Greece. With respect to the former, the Chi-Square test of Independence and independent sample t-test are used to determine associations and differences between volunteers and non-volunteers in relation to the variables under study. Explanatory analysis involves the application of logistic regression to predict volunteering for refugees/asylum seekers (compared to nonvolunteering) based on the variables measuring respondents' demographic traits, human, social and cultural capital as well as political conventional and unconventional behaviours. The variables are entered into five blocks; the first includes items associated with demographics, the second with human capital, the third with social capital, the fourth with cultural capital and the last with political behaviours. results Tables 7.2 and 7.3 present Chi-Square test of Independence and independent sample t-test results, respectively. As shown in Table 7.2, volunteering is significantly associated with gender, age, occupational class, employment status and respondents' informal interactions with friends. Specifically, more women (10.3%) than men (5.7%) volunteer for refugees/asylum seekers. A higher prevalence of volunteering is found for older age groups (i.e. more than 55 years old) (9.9%) and younger age groups (18-34 years old) (8.3%) compared to middleaged ones (6.2%). Moreover, individuals with higher education (14.1%) are more likely to volunteer for refugees/asylum seekers compared to individuals with intermediate (6.7%) or lower education (6.6%). With respect to income, middle-income earners (i.e. 776-1425 euro) have the highest prevalence of volunteering (9.2%), whereas low-income earners (i.e. less than 775 euro) the lowest one (5.9%). Volunteering is more popular among individuals of higher occupational class (i.e. in professional or managerial positions) (13.6%) than individuals of middle (7.9%) or lower occupational class (7.2%).
Part-time employees (10.2%) and individuals with other employment status (e.g. retired, housewives) (10.4%) have higher rates of volunteering compared to full-time employees (6.2%) or unemployed individuals (6.2%). Additionally, more frequent informal interactions with friends (9.5%) are positively associated with volunteering. Although volunteering is higher among individuals with specific conventional political behaviours such as voting (8.9%), the reported association is non-significant.
As shown in Table 7.3, the t-test analysis indicates significant differences in social trust and unconventional political behaviour between volunteers and non-volunteers. Specifically, volunteers for refugees/asylum report higher scores in social trust (M = 4.62, SD = 2.75) and in unconventional political behaviour (M = 2.44, SD = 1.19) than non-volunteers. Moreover, the mean score of religiosity is lower among volunteers (M = 5.46, SD = 3.18) than non-volunteers (M = 5.71, SD = 3.10), however the reported difference is non-significant.  Table 7.4 presents the results from binary logistic regression models for predicting volunteering for refugees/asylum seekers in Greece. To test the hypothesis associated with demographic attributes (see Hypothesis 1), the first model includes gender, age and education, all of them significantly contribute on predicting volunteering. Volunteers, in line with our hypothesis, are more likely to be women. Similar results are reported in previous studies exploring volunteering for refugees/asylum seekers. For instance, research conducted in Germany shows that volunteers for refugees are predominantly female Kleist 2015, 2016). Likewise, in Erickson's study (Erickson 2012), the majority of volunteers for refugees in Fargo are women.
The analysis shows that, that in line with our expectations and previous research (Wilson 2000(Wilson , 2012Musick and Wilson 2008;Rochester et al. 2010;Van Ingen and Dekker 2011), educational attainment does play a critical role in volunteering as higher educated individuals in Greece are more likely to engage in volunteering for refugees/asylum seekers. Contradicting our hypothesis, volunteers are more likely to belong to younger age groups (i.e. 18-34-year-olds).  study finds that volunteers for refugees are more likely to be either in their 20s or over 60, indicating that past empirical evidence supporting that volunteering is more prevalent among middle-aged citizens (Wymer 1998;Curtis et al. 2001;Pho 2008) might not hold for the specific domain of volunteering.
To examine the hypothesis associated with human capital (see Hypothesis 2) income, occupational class and employment status are  Gender ( included in the second model. The analysis shows that controlling for the demographic attributes under study, income and employment status do not significantly contribute on predicting volunteering. Only occupational class is associated with volunteering indicating that middle occupational class individuals are less likely to volunteer compared to those from the lowest occupational class (i.e. manual workers). Such findings contradict our hypothesis as well as previous research underlining that volunteering is more strongly supported among individuals with higher human capital, in terms of higher income and occupational class (Rotolo and Wilson 2007;Hurley et al. 2008;Pho 2008).
To test the hypothesis associated with social capital (see Hypothesis 3), the indicators of social trust and informal social interactions with friends are included in the third model. In agreement with our hypothesis, social  Table presents logistic regression coefficients B with standard errors in parentheses +p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001. Data weighted capital and specifically social trust plays a critical role in volunteering since individuals' social trust is positively associated with volunteering for refugees/asylum seekers (Wilson and Musick 1997a;Brown and Ferries 2007). Respondents' intense informal social interactions (i.e. every week or almost every day) positively contribute to volunteering; however the reported association is significant at p < 0.10. In the fourth model the indicator of cultural capital, that is, religiosity, is added. Contradicting our expectations (see Hypothesis 4), the analysis shows that religiosity is not associated with volunteering for refugees/ asylum seekers in Greece. Whilst non-significant, the negative sign of the religiosity coefficient provides some preliminary evidence of the negative association between religiosity and volunteering.
To examine the hypothesis associated with political behaviours (see Hypothesis 5), unconventional and conventional political behaviours are added in the final model. The former significantly contributes on predicting volunteering, as unconventional political behaviour is associated with volunteering for refugees/asylum seekers Dekker and Van den Broek 1998;Hodgkinson 2003;Bekkers 2005). Similar results are reported for the conventional political behaviour of voting; however, the reported association is non-significant.

conclusIons
Since 2015, the influx of refugees to Europe-primarily from North Africa in the aftermath of the Arab Spring and from the Middle East due to the civil war in Syria-has challenged Europe to tackle one of the largest movements of displaced people through European borders since World War II (UNHCR 2016). According to Eurostat (2016), in 2015 a record number of over 1.2 million first-time asylum seekers registered in EU member-states. Almost one out of three first-time asylum seekers originate from Syria, while many are also Afghans and Iraqis.
In the context of the recent refugee crisis, Greece has been marked by a fast-paced transit of high numbers of refugees/asylum seekers entering its territory en route to Northern and Central Europe. The large-scale arrival of refugees/asylum seekers have challenged Greeks to cope with a twofold crisis: the economic crisis as well as the refugee crisis. Concerning the economic crisis, in the last six years Greece has faced the most acute recession in its modern history with devastating socio-economic impacts on individuals' lives echoed in record unemployment and poverty rates (Matsaganis and Leventi 2014;OECD 2014). Since 2015, the country has been strained by both economic depression and the massive migration inflows of hundreds of thousands of refugees/asylum seekers.
Despite economic hardship, volunteers have been instrumental in providing help (such as food supplies, medical aid, legal and financial support, etc.) to refugees/asylum seekers arriving on Greek shoressimultaneously relieving the state of one of its core roles. Therefore, the government has come to partly rely on the contributions of volunteers in order to tackle the refugee crisis. As Evangelinidis (2016, p. 33) argues: Where the state apparatus was absent, or its structures were insufficient, civil society organizations in many different forms (e.g. professional NGOs, volunteers, ad hoc groups and collectives) tried to fill the gap. With the central government unable to properly provide for many of its citizens, let alone refugees or migrants, the humanitarian vacuum has often been filled with solidarity initiatives.
Based on a hybrid approach which combines the sociological and political approaches to volunteering, the explorative analysis provides some preliminary evidence of volunteers' traits, most in line with past empirical research into volunteering. The explanatory analysis sheds some light on volunteers' profiles who are primarily women, young, higher educated, individuals engaged in unconventional political acts and with higher level of social capital. However, contradicting our hypotheses and previous research, human and cultural resources are not associated with volunteering for refugees/asylum seekers in Greece.
The lack of association between human capital and volunteering may reflect the peculiarities of volunteering in Greece. Some scholars support that individuals with less human capital are more likely to engage in informal volunteering rather than formal volunteering (Williams 2002;Hustinx et al. 2010). As argued in the introduction of the chapter, whilst the official statistics show the low prevalence of formal volunteering in Greece compared to other European countries, some researchers underpin that there is a vibrant informal volunteering sector that has been triggered in different emergency periods (Sotiropoulos 2004;Karamichas 2007;Rozakou 2011). We can assume that the main trend of volunteering for refugees/asylum seekers, as it has happened in the past in Greece (Sotiropoulos 2004;Karamichas 2007;Rozakou 2011), has primarily followed the informal path, which is more common among individuals with lower human capital.
With respect to the lack of association between religiosity as an indicator of cultural capital and volunteering, the finding might reflect specific shortcomings of the proxy applied. Scholars reporting strong correlations between religiosity and volunteering usually apply as proxies religious practices such as frequency of church attendance and of religious prayer rather than subjective measures of religiosity and intensity of beliefs (Wilson and Musick 1997a;Musick and Wilson 2008;van Tienen et al. 2011;Paxton et al. 2014). These practices are more likely to proffer values (such as self-sacrifice and compassion), which reinforce the decision to volunteer (Son and Wilson 2011).
In the study's questionnaire, there are no available measures of such religious practices that would allow the refined measurement of religiosity. Additional limitations of the study involve its cross-sectional design where causal imputation is difficult. Hence, we are unable to determine the direction of specific causal relationships examined, for example, between cultural capital, social capital and/or political engagement measures and volunteering. It should be noted that concerns over selection bias have consistently plagued the volunteering empirical research (Wilson 2000).
Nevertheless, the study enriches the scarce empirical research on volunteering specifically for refugees/asylum seekers, by portraying the profiles of volunteers providing solidarity to thousands of refugees/ asylum seekers arriving in Greece. Undoubtedly, volunteers have been key actors in welcoming and helping refugees/asylum seekers contributing to the first step towards newcomers' integration into the new host countries. However, volunteers' contribution should not be treated as substitute to core state obligations towards refugees/asylum seekers. Policy interventions at the Greek and EU level are urgently necessary to manage the refugee crisis effectively and allow the resettlement of refugees/asylum seekers in safe countries where they can have the opportunity to rebuild their lives.