Introduction

Political representation is a tenet of democracy. As a standard of “fairness,” it is fundamentally married to ideas of legitimacy in governance. The vision of an ideal representative democracy that gives voice to its constituents allowing them “to be present” in policy making (Pitkin 1967) is challenged by the diversity of preferences that seek representation. The growing presence of resident non-citizens lends further complexity to this issue (Hammer 1999).

Political participation is the sine qua non of representative democracy. It has been defined as “voluntary activities by ordinary people directed toward influencing directly or indirectly political outcomes at various levels of the political system” (Verba et al. 1995: 38–39). At the heart of debates that surround political participation, it is the question: Who can legitimately make claims on the political system? In responding to this question, Irene Bloemraad (2000; Bloemraad et al. 2008) flips the logic of participation through membership on its head by categorizing participation as a criterion of political membership. Documented events of successful claims-making through immigrant political participation (McNevin 2009) lend support to this perspective. This view supplies a critical space where the immigrant, acting as her own “agent,” is understood as both subjected to and purposefully seeking to influence the social arrangements that bear upon the quality of her integration into host society (Bevelander and Pendakur 2011; De Rooij 2011; Bloemraad 2000). Within this schemata, politically salient attributes that distinguish the politically active may hold consequences for policy outcomes and for the quality of democratic representation (Verba et al. 1995).

Previous studies have explained variation in political participation by means of the heterogeneity in each individual’s personal attributes including resources, gender, race, and age (Verba et al. 1995; Verba and Nie 1972; Jennings and Niem 1981). When extended to immigrant political participation, these studies have found that factors specific to migration, such as language proficiency and length of residence in host country, advantage some immigrants over others (Lien 2004; Brady et al. 1995; Tam Cho 1999). The approach that has been employed to study immigrant political participation situates action at the individual level. While informative, these attribute-based approaches underplay the role of intermediate social structures, like communities or social networks, in determining political participation.

Indeed, there is a wealth of evidence to support the idea that relationship structures are critical for determining who participates in which political acts (Vertovec 2003; Van Heelsum 2002; Fennema and Tillie 1999, 2001; McAdam et al. 1996). Early work by Klandermans and Oegema (1987) demonstrates that individuals advantageously embedded in social networks closely linked to like-minded individuals tend to participate more frequently in protest. More recently, Chaeyoon Lim (2008) shows that recruitment efforts by a joint member of one’s association has more impact compared to recruitment efforts by outsiders. One of the most influential works on group-level influences on political participation was carried out by Meindert Fennema and Jean Tillie (1999, 2001). They employ a social network analysis of migrant organizations in Amsterdam to reveal that national groups with dense, horizontally connected associations—what they call civic community—exhibit higher levels of political participation. These authors theorize that civic community fosters trust or social capital, which is then translated into collective action. Therefore, social capital may stand as an appropriate and accessible way to integrate relationship structures into the aforementioned attribute-based models.

Fennema and Tillie’s work is one of many studies that postulate a critical link between social capital and participatory acts (Leighley 1995; Putnam 1993, 2000; La Due Lake and Huckfeldt 1998). Social capital is the value entrenched in our relationships that can be transformed into tangible assets for the pursuit of individual and collective interests. It has been proposed to foster political participation by (1) facilitating the dissemination of politically relevant knowledge, know-how, and ideas (Jacobs and Tillie 2004; La Due Lake and Huckfeldt 1998), (2) promoting norms of political participation (Leighley 1995; Putnam 2000), and (3) establishing and maintaining relationships that enhance participation (Aldrich and Crook 2008). In sum, social capital improves a community’s capacity to organize and, therefore, to mobilize.

Tests of Fennama and Tillie’s civic community model, however, have generated conflicting findings. For example, Jacobs et al. 2004 find that in Brussels, the expected relationship between associational membership, civic community, and political participation does not hold. Similarly, Togeby (2004) finds that while associational membership seems to promote the political participation of Turkish immigrants in Denmark, these effects are less pronounced for other ethnic groups. As a final example, Berger et al. (2004) demonstrates that in Berlin, it is membership in German associations and not in ethnic associations that facilitates the participation of immigrants. These discrepancies between theory and empirical observations seem to suggest that the link between social capital and political participation is tenuous, at best, and relies heavily on facilitative institutional settings (Jacobs et al. 2004; Togeby 2004).

In this article, I propose that a previously unaccounted for disharmony between individual-level and group-level social capital may supply a feasible explanation with which Fennema and Tillie’s theoretical work and the empirical observations described above can be reconciled. An important assumption that underlies the putative link between social capital and political participation is that relationships influence actions. Or as Granovetter puts it, social outcomes “are affected by actors’ dyadic (pairwise) relations and by the structure of the overall network of relationships” (Granovetter 1992: 3). Granovetter calls this tendency “embeddedness.” Thus, embeddedness suggests that individual-level influences may impact political participation alongside group-level influences. A thorough accounting for of these distinctions is missing in the literature. In previous work, individual-level and group-level social capital have either (1) received little consideration, (2) been treated as harmonious, or (3) used interchangeably. Instead, as is discussed in the next section, not only are the effects of individual-level social capital and group-level social capital distinct, these effects may even work against each other.

This article investigates the connection between social capital and the political participation of immigrants in Rome (Italy). Its main finding is that social capital facilitates the participation of immigrants in Rome in voting but not in protest. It argues that these results are best explained vis-à-vis the previously overlooked interplay between group level—and individual level—bonding and bridging social capital. In the following section, the literature on group-level and individual-level social capital as it pertains to political participation is reviewed, drawing heavily on insights developed in the field of sociology. Following this review, a novel theory on the impact that these distinctive influences may have on political participation is then laid out. Before detailing the materials and methods and proceeding to the results, the case study—which draws on survey data, ethnographic records, and interviews collected through 8 months of fieldwork in Rome, Italy—is described.

Considering Groups and the Individuals Nested Within Groups

Group-level social capital binds individuals together in such a way that encourages them to privilege members of their in-group over others (Zhou 2004; Portes and Sensenbrenner 1993; Coleman 1988). This is the explanation Fenemma and Tillie (1999, 2001) summon to account for their observed association between participation and dense networks of associations (Fennema and Tillie 1999). Their claim is that trust in ethnic leadership travels through the links between ethnic associations (Fennema and Tillie 2001). This trust, however, represents a particularized (focused toward one’s in-group) rather than generalized (focused toward society, in general) trust, which some scholars have claimed inhibits rather than enables participation (Uslaner and Conley 2003; Putnam 1993). Moreover, individuals are diversely embedded within networks, and these differences may hold consequences for their participation. For example, Huckfeldt (1979) demonstrated that status conflicts impose negative pressure on participation. Specifically, he shows that when actors find themselves embedded in a neighborhood populated primarily by non-peers, observing their neighbor’s political participation dampens their own. This suggests that individual-level impacts are distinct from group-level influences.

In the literature on immigrant political participation, consideration for group-level and individual-level social capital has been framed in terms of possible variation among individuals’ abilities to access to group-level social capital. For example, Tillie (2004) accounts for the distinction between individual-level social capital and group-level social capital by using participation in groups or having a personal relationship with someone who is a member of a group, to measure individual-level social capital. In his theorizing, individuals’ social capital rests on their ability to access to civic community (see also Tillie and Slijper 2006). This theorizing, however, assumes that the influence of group-level social capital is felt through direct links rather than more amorphous yet equally powerful instruments of social capital such as traditions, norms, rituals, or belonging (Fukuyama 2001; Portes and Sensenbrenner 1993) that may actually hinder rather than help “x” “to do what she otherwise would not be able to do” (Fennema and Tillie 2001: 29). This conceptualization of individual social capital, which firmly situates it within associations, provides a limited window into its effects on political participation. Moreover, it overlooks potential disharmony between levels of social capital.

Speaking directly to the conflict between levels of social capital, Portes (2000) asks us to consider categories of social capital including group-centered social capital (what he calls bounded solidarity and enforceable trust) and individual-centered social capital (what he calls reciprocated exchanges) (1998, 2000). These forms of social capital may be thought of as different levers on behavior. Bounded solidarity is encoded in shared rituals, norms, and other social artifacts that tie individuals to one another. It engenders a particularized trust, which encourages benevolence toward members of one’s in-group and facilitates exchanges among them. It constitutes the “carrot” of group-level social capital. Enforceable trust, group-level social capital’s “stick,” is expected to flourish where there is deep connectivity between individuals so that monitoring and information exchange is straightforward (Habyarimana et al. 2007). Both types of group-level social capital are expected to be favored by the migratory process. It must be noted that of these three forms of social capital, only reciprocated ties requires direct contacts between individuals for the cultivation of social capital (Browning et al. 2004). In other words, individuals may be able to access (and may be subject to) group-level social capital by sole virtue of shared traits or histories.

Robert Putnam (2000) distinguishes between reciprocated relationships that bind individuals to their communities (bonding social capital) and those that link them to members of their out-group (bridging social capital). The idea of bridging ties is closely related to Ronald Burt’s structural hole theory (2002, 2009). Burt uses the term “structural holes” to describe the gap between groups that is filled when individuals build bridging ties. It is commonly thought that these “weak ties” are more valuable in creating economic and social opportunities (Granovetter 1973; Lancee 2010; Putnam 2000) than bonding ties. The true value of structural holes is that they enable individuals to circumvent restrictive aspects of group-level social capital (Burt 2002). When extended to migrant political participation, structural hole theory implies that relationships outside the ethnic community, particularly with nationals, may open new opportunities for participation. Gidengil and Stolle (2009) investigate this question by looking into whether relationships with Canadian nationals positively influence the political knowledge of immigrant women in Canada. They find that the effects of these structural holes on political knowledge are negligible. Looking into more active forms of political membership, Berger et al. 2004 find that bridging social capital (membership in German groups) supports immigrant political participation more significantly than bonding social capital (membership in ethnic groups). Their work provides some support for structural hole theory in political participation studies.

If Portes’ theorizing is taken into account, then a view of social capital as always being generated at the group level and accessed solely through associational membership (or links to other members) is too limited. Extending his theory to migrant political participation implies that between-community variation in participation hinges on bounded solidarity and enforceable trust (which may be measured vis-à-vis civic community) while within-community variation hinges on the heterogeneity in individuals’ embeddedness in ethnic communities (bridging versus bonding ties).

Creating Space for Conflict in Social Capital Theorizing

As a point of departure, let us consider immigrants’ tendency toward non-participation which, this article posits, results from their relatively disadvantaged social positions. To be specific, host societies have a long history of employing societal and institutional mechanisms to discourage migrant political participation (Martiniello 2005; Koopmans 1999; Ramakrishnan and Espenshade 2001). These effects are proposed to persist long after policies have shifted toward inclusivity. Recent push back against immigrants that accompanied the onset of the economic crisis and the international war on terror served to cement their proclivity for abstinence from politics. In other words, abstinence from political participation is the steady-state scenario. Participation in all acts from voting to protesting, then, requires the mobilization of economic and social resources; in sum, it requires social capital.

For the most active immigrants, immigrant leaders, participation in politics provides numerous symbolic and instrumental benefits. Political participation may increase an ethnic group’s visibility or contribute to the sensibilization of nationals, and it may also be an effective instrument for claims-making. Immigrant leaders may also wish to capture immediate economic and social benefits from political participation. Specifically, immigrant leaders’ capability in mobilizing co-ethnics is demonstrative of their standing within their ethnic community. This fosters partnerships with host institutional actors who wish to reach out immigrants (Bloemraad and Vermeulen 2014). The resulting improved standing with the state, in turn, brings reputational gains and increases migrant leader’s stature among co-ethnics. Consistent with this theorizing, immigrant leaders have greater motivation to leverage group-level social capital in state-mandated activities like elections compared to informal activities like protest (as depicted below in Fig. 1 below).

Fig. 1
figure 1

Social capital as a source of restriction or autonomy

Full size image

Elections are sanctioned and organized by the state and its institutions. In contrast, protest is usually aimed at challenging dominant institutions (van Zomeren and Iyer 2009; Martiniello 2005; Wright et al. 1990). Eisinger (1973: 5) defines protest as “a host of types of collective manifestations—disruptive in nature—designed to provide ‘relatively powerless people’ with bargaining leverage in the political process.” What separates it from other forms of political action is the implicit threat of violence and disruption carried within the action (Lichbach and Gurr 1981; Eisinger 1973). Klandermans et al. (2008:993) observe that immigrants “are expected to assimilate into the host culture, and failure to do so is considered a sign of lack of loyalty. Immigrants’ loyalty to the country of residence is placed under even more doubt.” Thus, immigrants may not feel entitled to stake claims through protest (Hasselberg 2014; Bilodeau 2008).

Under the scenario depicted above (in Fig. 1), immigrant leaders should use appeals to bounded solidarity to encourage co-ethnics to vote (for co-ethnic candidates sponsored by ethnic leaders) while they should use mechanisms of enforceable trust to discourage protest. The extent to which immigrants heed these pressures depends on their accumulation of bridging and bonding social capital resources. Participation in informal activities requires special types of bridging ties. Specifically, it requires ties to politically active Italian groups or individuals who have the resources necessary to help immigrants overcome barriers to participation. Groups that provide assistance with issues related to immigration can be expected to possess these resources. This hypothesis is consistent with previous research which finds that migrant’s participation in informal activities tends to be more oriented to host-centric issues compared to their participation in formal activities, which tends to be centered on particularistic, ethno-centric claims (Phalet and Swyndegouw 2003).

Political Membership as It Bears out in Rome (Italy)

Little is known about the political participation of immigrants in Italy’s urban centers. Given its recent shift from a country of out-migration to in-migration, Italy’s approach to immigration and integration is still in its nascent stages. Its weak colonial history and the diversity in provenance of its foreign-born resident population further distinguish Italy from many of its European neighbors. What is most striking, however, is the exponential increase of in-migration into Italy and the rapid evolution of discourse on the subject (Sciortino and Colombo 2004). Notwithstanding, the scholarship on immigration in Italy is sparse, more so is the scholarship on the political participation of its immigrants. As a signatory of “Convention on the Participation of Foreigners in Public Life at the Local Level,” the country has pledged to expand migrant’s access to electoral participation. This pledge has yet to be honored. This study—and others in this vein—speak to the potentialities that may accompany the actualization of these political promises. As the nation’s capital and seat of its parliament, Rome rests at the apex of these developments.

In Italy, a significant portion of rule-making vis-à-vis immigration has been devolved to cities (Meli and Enwereuzor 2003). Local governments may make up for the shortcomings of national policies by instituting initiatives aimed at inclusivity and integration (Alexander 2003). For example, Law 5/2004 adopted in Emilia Romagna and Tuscany’s Regional Law 29/2000 and Regional Law 33/2005 establish programs to facilitate integration in multicultural communities (Campani 2007). Alternatively, local governments may exclude and marginalize immigrants (see Ambrosini 2013). This points to an important difficulty in studying political participation: the perceived and actual exclusion from politics because of immigrants’ status as non-citizens. In this respect, Rome makes a good case for the study of immigrant political participation because in 2004, the former mayor of Rome, Walter Veltroni, created elected positions for foreign residents in city government. In addition, his administration also founded a consultative body to communicate migrant policy needs. With this action, Veltroni effectively opened the opportunity structure in Rome. His initiative provides an opportunity for research that may be instructive not only in the case of Rome but also for other European cities especially urban centers in Europe’s southern states.

Materials and Methods

This article uses a mixed method approach involving ethnographies, interviews, and surveys. Altheide (1987: p.66) defines ethnography as “the description of people and their culture.” Immigrants from different national groups and ethnicities were observed in their day-to-day lives, at their place of work, at their homes, and also as they were participating in community events such as festivals, embassy meetings, and worshipping at mosques or churches. This ethnography includes over 14 celebrations, half a dozen meetings and assemblies at the City of Rome, work at a service desk of a large trade union, and visits to stalls at five markets with varied levels of ethnic business. It also includes volunteer work for an ethnic association and teaching English to the Ambassador and staff at one of the embassies. In addition, 85 qualitative interviews were conducted beginning with a series of interviews with experts on Italian immigration (such as researchers at Iniziative e Studi sulla Multietnicita (ISMU) and Dossier Statistico Immigrazione (IDOS), journalists, and service providers (administrators of the neighborhood sportelli, embassies, and NGO groups) and later with immigrant leaders and activists. Experts were identified by means of consultation with faculty at the University La Sapienza and Roma Tre during a preliminary research period in 2010. Together, these qualitative methods were used to study migrant routines and identify times and locations for sampling, improve the survey instrument, and to validate and explain survey results.

The survey instrument (a mix of closed and open-ended questions) was used to estimate the proportion of immigrants involved in political acts. The data on immigrants collected by the Italian census (ISTAT) is reputed to suffer from inconsistencies in data collection and inadequacies in reporting (Mudu 2006). Therefore, the absence of reliable data about the target population made it impossible to make the necessary calculations to achieve scientific sampling requirements and complicated data collection. To compensate for this, a time-location sampling (TLS) frame was employed. Taking advantage of immigrants’ tendency to cluster together (Karon 2005), research sites and times were uncovered by means of ethnography (Magnani et al. 2005). These times and locations were then validated against interview records (as in Watters and Biernacki 1989). Surveys were administered face to face during the daytime (between the hours of 10:00 to 18:00) at three different sites frequented by immigrants for work, leisure, or residence: the covered market in the neighborhood Esquilino, at a popular immigrant helpdesk on Via Buonarotti, and at the central train station, Termini (Mudu 2006; Knights 1996). TLS may introduce bias when not all sites are identified (Magnani et al. 2005) or when certain subpopulations are overrepresented or underrepresented at selected sites selected (Karon 2005). Care was taken to account for this: The process of site selection involved two preliminary site visits to Rome in 2010 as well as two and a half months of ethnographies and interviews.

Control Variables

The survey included a battery of questions intended to collect data on attributes that have previously been associated with participation such as education, gender, age, and occupation (Verba et al. 1995; Brady et al. 1995; Jennings and Markus 1988). It also asks respondents to report on so-called migration-related factors which have been linked to participation such as language proficiency (Tam Cho 1999; Brady et al. 1995; Uhlaner et al. 1989), period of stay (de Rooij 2011). In addition, work by Lien (2004) and Bilodeau (2008) suggests that immigrants from countries with political systems similar to Italy should have higher rates of participation compared to others (de Rooij 2011; Lien 2004; Bueker 2005). Similarities in political systems are accounted for in this analysis using their Polity IV scores (Marshall et al. 2011).

Associational Membership

Italy guarantees migrants the right to assembly and to association. This right was initially designated in 1990 through the Martelli Act, the nation’s first law on in-migration. Following previous work (Berger et al. 2004; Jacobs et al. 2004; Tillie 2004), this article uses membership in a civic association as a measure of social capital. As in these works, it is used here as a proxy for direct access to civic community.

Ethnic Embeddedness

Ethnic communities tend to hold a wide range of resources not all of which are accessible through associations. Language, documentation, and employment are the three most common and pressing needs for the immigrants in Rome. The measure of ethnic embeddedness used in this study is based on an index which weights each on survey respondents’ answers to the following open-ended questions:

23. If someone wanted to learn Italian, where can he/she go?

24. If someone needed help filling out his/her permesso di soggiorno where can he/she go?

25. If someone needed help finding work, where can he/she go?

Answers to the survey questions reflect the extent of the respondents’ dependence on their ethnic community for meeting these needs. This dependence is used as a proxy for bonding and bridging social capital. Individuals with low embeddedness scores are assumed to have greater access to outside resources (through bridging social capital) than individuals with higher embeddedness scores (higher bonding social capital ties). Responses were coded using the following rules: When a respondent named a non-ethnic resource, the answer was coded as “0.” If, instead, the respondent named an ethnic resource, then the answer was coded as “1.” A weighted index was then calculated with these responses as its base.

Social Capital and Participation in Rome, as Viewed from Many Angles

Surveys were refused in 37 of the cases (the response rate was 82.9 %). In the end, 180 subjects agreed to a survey interview. Of these, 170 were employed for this analysis (ten surveys were excluded because of incomplete answers). Immigrants from 19 countries responded to the survey. Of the respondents, 55.9 % were female, the majority of whom worked as domestic helpers. Of the respondents, 57 % fall in two age brackets: 35–44 (28 % of survey respondents) and 45–54 (29 % of survey respondents. These are age groups that should have higher rates of participation compared to others (Jennings and Niem 1981; Jennings and Markus 1988). The minimum period of stay in the country was less than 1 year while the maximum was 41, a range which accounts for the high standard deviation of 8.4 years. Most of the respondents had lived in Italy for an extended period of time (60 % were eligible to participate in the elections, which requires uninterrupted legal residency of at least 5 years), and the average period of stay was 12 years.

Overall, respondents reported low participation in all of the political activities. Of those surveyed who were eligible to vote, 36.6 % did while 20.6 % of respondents reported having participated in protest. Our analysis returned low correlation coefficients for voting and protest (R = 0.115). This suggests that there was little overlap between the individuals who participated in protest and those who voted.

This article employs a log-likelihood model to examine the role that resources, migration-related factors, and social capital play in determining immigrants’ political participation in voting and protest. Voting and protesting were measured as binary values (“1” for participate and “0” for not) calling for a logistic regression model. Following Berger et al. (2004), these relationships are analyzed using a stepwise model. The base model, model 1, is used to account for skills, engagement, and social capital. In model 1, participation was regressed against years of schooling, gender, religion, associational membership, and interest in Italian politics. In model 2, migration-related factors (years in Rome, ideological similarity with Italians, and language spoken at work) were integrated into the model. Finally, embeddedness scores were added in model 3. Hosmer Lemeshow tests indicate that the goodness of fit is improved in model 3 compared to models 2 and 1 especially in the case of voting. The resulting estimates given in Tables 1 and 2 show how much a one unit increase of each predictor is likely to increase the log-odds that a respondent will participate in the given activity.

Table 1 Binary logistic regression for voting
Full size table
Table 2 Binary logistic regression for protesting
Full size table

Resources and Participation

As has been found in previous studies of political participation, higher human capital endowments (measured here as the number of years of schooling) is positively related to participation. While the regression coefficients are consistently significant and positive, the strength of the associations estimated here is markedly less pronounced compared to those found in studies of participation by naturalized immigrants (Lien 2004; Brady et al. 1995). In the case of voting, these effects diminish when controls for migration-related factors and embeddedness scores are introduced.

This pattern may be explained by the disruption of the link between educational attainment and labor market outcomes, which occurs during the process of immigration (Borjas 1994). Most jobs open to immigrants have famously been characterized by sociologist Maurizio Ambrosini (2001) using the five Ps: pesanti, precari, pericolosi, poco pagati e penalizzati socialmente (heavy, unstable, dangerous, low paid, and socially penalized). Informant 769, an employee at an NGO in Rome, Casa per Diritti Sociali, observed, “the majority of immigrants are here to work. And most of them find themselves in work conditions which are comparable to slavery.” (2011, Translated from Italian by the Author). The “ethnic penalties” that befall immigrants in receiving country labor markets (Phalet and Swyndegouw 2003) are likely to negatively affect their connectivity to the people, groups, and institutions that encourage political participation.

Migration-Related Factors and Participation

As expected, the relationship between participation and the migration-related factors (ideological distance, length of stay, and language proficiency) is positive and significant. However, their influence is negligible when compared to two other characteristics that gain political salience through the process of migration: religion and gender.

Immigration is a transformational experience which may renew, awaken, or activate a new political identity. It is difficult to touch on questions of Islam without touching on the question of immigration because the majority of Europe’s Muslims are immigrants or of immigrant origin. Increasingly, the debate was about protecting European values by trying to bring existing minorities into line.” Being Muslim is expected to be politically salient for participation (Klandermans et al. 2008; Soysal 1997). These expectations bear out in this analysis throughout the process of stepwise regression. Religion, or being Muslim, is the strongest predictor of voting. Although the effect is much less pronounced for protest, the regression coefficients are still positive and significant. And, after gender, being Muslim is the most significant predictor of protest.

It may be that these regression estimates capture organizational capacity rather than religious identity (Verba et al. 1995; Jones-Correa and Leal 2001). To test for this, the survey includes a question about regular attendance of religious services or religiosity. The bivariate correlation between religious identity and religiosity is positive but not highly significant (R = 0.333). In addition, the correlation coefficients between religiosity and voting (R = 0.024) as well as religiosity and protest (R = −0.038) indicate that it is not a good predictor of participation.

Gender is also hypothesized to be an identity salient to migrant political participation. Specifically, male immigrants are expected to be more likely to participate in politics (Klandermans et al. 2008; Berger et al. 2004). These expectations bear out in the findings reported in this article. As discussed above, gender, or being male, is the most significant predictor of participation in protest, and the regression coefficients returned for voting are both positive and significant.

Social Capital and Participation

The analysis also suggests a negligible role for group membership in facilitating political participation. The models returned low estimated values for voting and protest. However, while the relationship between membership and participation improved slightly for voting after moving from model 1 to model 3, it decreased slightly for protest. The low estimates suggest that if associationism does indeed contribute to social capital, it can be felt through non-direct forms like bounded solidarity rather than direct forms like reciprocated exchange. In contrast, as expected, the analysis returned significant and positive relationship between voting (an institutionally sanctioned activity) and ethnic embeddedness. This results signal strong support for these activities within ethnic communities. Reliance on co-ethnics is the second most important predictor of voting. Also in line with our expectations, this relationship does not hold for protest (an informal activity). Ethnographic work supports the idea that some ethnic communities may actively seek to discourage protest. For example, the leader of an important Filipino association in Rome discussed his group’s policy to explicitly discourage participation in protest.

Discussion

The literature on social capital bids us to consider how relationships and the ways in which they are nested in wider social systems inform behaviors and decision-making (Granovetter 1992). For immigrant political participation, embeddedness in the form of bonding social capital implies that individuals are more likely to participate in those activities which are approved of by their reference group and/or in activities which their immediate contacts (friends, family, co-workers) also participate. The literature on political participation provides numerous empirical examples supporting these expectations (Klandermans et al. 2008; Putnam 2000). This exploratory work leverages Alejandro Portes’ theoretical work on social capital to examine the political participation patterns of Rome’s immigrants. It finds that not only does social capital impact political participation but it also determines the type of activity in which immigrants participate.

The theory put forward and tested in this article tries to reconcile disparate findings with regards to social capital and political participation. It borrows heavily from Alejandro Portes’ theorizing on social capital and makes one of the first tests of its kind in the context of political participation. Although the scale of this project is relatively modest and several improvements could be made, provided the availability of sufficient resources, its findings are provocative.

First, bridging ties are important. In the case studied here, bridging ties between associational leadership and Roman institutions favored voting. In fact, the Veltroni administration launched a campaign to raise awareness about the elections. This campaign was quite targeted, leaning heavily on the administration’s familiarity with migrant social networks. To be specific, it hired 19 immigrant leaders from its consultative body to promote the elections. The uneven participation by national groups was clearly a reflection of this. At the district level, Filipinos (36.8 %) and Bangladeshis (36.8 %) won more than two thirds of the positions in play. This is particularly remarkable when one considers the diversity of the migrant community in Rome and the fact that neither Filipinos nor Bangladeshi are amongst the most populous national groups in the city. The absence of bridging ties is also important. When group members have few bridging ties to outside entities, appeals to bounded solidarity or the imposition of enforceable trust are more likely to hold sway.

Bridging ties are hypothesized to facilitate informal activities. Since the late 1970s, unions, Catholic groups, and Italian activists have supplied immigrants with the resources and connections necessary for political organization. One interesting and recent case involves a community of Sikh Indians working under exploitative conditions in Latina, a small agricultural city on Rome’s outskirts. A large trade union courted this community and encouraged it to unionize and to organize. Together, and through protest, they brought visibility to the illegal and exploitative work conditions that these Sikh’s face and to demand improved protections from the state. This case is illustrative of the instrumentality of bridging relationships for accessing material resources and for learning about one’s rights.

The diversity of activism facilitated by bridging ties is important if one considers that political acts vary in their capacity to secure effective political representation. Voting, for example, is reputedly serves as a “blunt instrument” for political communication (Verba et al. 1995) compared to protest. As an important Bangladeshi community leader opined,

“the institution only knows one language. This is the power of the public square. They don’t understand Italian, Bangladeshi, or Pakistani. Nothing. Nothing. If we are many in the square, in movement, they understand. That is, this is a language that they are familiar with.” (2011, Translated from Italian by the Author).

Second, ethnic networks matter for political integration. Political integration has been defined as the “process of becoming part of mainstream political debates, practices and decision-making, without privileging any individual assimilation or group-based incorporationist view” (Bloemraad 2006: 6–7). In essence, integration implies full political membership. It has been argued that the involvement of immigrants in the social and political life of their ethnic communities may eventually segue into full political membership in host society (Bloemraad 2007). The qualitative fieldwork which is part of this article provides some support for this hypothesized relationship. Immigrant activism is critical to bring to bear the true depth and diversity of the migrant-lived experience. The participation of marginalized or underserved groups in mainstream politics provides critical input to societies faced with policy problems for which these specific populations may offer appropriate solutions (Mansbridge 2003). It follows that involving immigrants in politics requires the involvement of their networks and leaders. When policies to promote participation are conceived without consideration for how mobilization unfolds, national and ethnic groups that are unable to leverage ethnic networks are silenced and marginalized. This is exactly what occurred in the 2006 communal elections in Rome. Informant 768 reflects,

“for instance notwithstanding the African continent we have a very reasonable number of immigrants that are residents but in the experiences that we have had it appears that Africa had the lowest number of participants… . I think that community organization then information sensibilization is part of it. The more a community is cohesive, the easier it is for them to understand the issues.”

The question of participation and integration touches upon a related question regarding the extent to which immigrants can realistically be expected to integrate. It has been suggested that low levels of participation signals immigrant apathy toward host political processes (for a review, see Bloemraad and Vermeulen 2014). Instead, this paper posits that institutional and societal barriers may be prohibitive for participation. Barriers which bridging ties help overcome. Indeed, immigrant political leaders and activists frame the conversation on turnout in terms of the costs of participation. As informant 795 succinctly expressed, “participating in politics is a luxury.” Informant 850, an employee of one of the trade union’s local service desk, provided greater detail saying,

“[immigrants] are always worried, I think, since there are many other immigrants they believe that if they miss one day of work they may be fired [and easily replaced]. And since they have come to Italy to work and to earn money to send home, this is their main priority” (2011, Translated from Italian by the Author).

However, when supplied with adequate resources that to overcome barriers to participation, immigrants have and will participate. For example, migrant leaders worked to give a face to the issue of migration being deliberated prior to the enactment of the Martelli Act. Informant 714, a Filipino who works for one of Italy’s largest labor unions, describes the activism of that period:

“In the 90s we began to make more concrete demands regarding migrant rights. In those days, our work contract was different from those given to Italians. So we pushed the labor unions to support us in asking for equal contracts for migrant domestic workers. This was a success not only for the Filipino community, but also [for] three others. From this, we established a migrant federation composed of different migrant associations including students, like an Egyptian Student Association [and] Palestinians. This expanded our concerns to the political realm, beyond labor. These were the rights we helped establish through Martelli, like the right to convene in federations and migrant initiatives. Or to petition for family members to join us in Italy, which before was not there. These initiatives that go beyond culture were made possible because the government could see that they had support within the migrant community.” (2011, Translated from Tagalog by the Author)

Third, the findings of this study have implications for the scholarship on political participation by all members of a polity not just by migrants. Indeed, as they suggest that previous research that underplays or disregards the relationship between social capital and participation may overestimate or underestimate the influence of other relevant factors—such as education or gender on participation. In other words, these models were misspecified. This opens a call for improved tools for measuring social capital and then incorporating these measures in studies of participation.

A, final, related point that bears mentioning is that one thing that is clearly missing in studies that followed Fennama and Tillie’s is a comparable measure of civic community that can be easily integrated tests of their model. In Rome unlike in Amsterdam, for example, there is no reliable list of associations nor is there a way to study their links exhaustively without conducting extensive social network analysis using firsthand data collection. Today, the true number of immigrant associations in Italy is unknown but estimated to be more than a thousand. Many associations operate informally—they are not registered with the Italian government. Also, the high “churn rate” complicates the task of keeping track of migrant associations. Indeed, it is not uncommon for subgroups to break off from existing established associations or for disagreements to cause associations to disband. More work needs to be done to increase the comparability of these studies. An important first step is the acknowledgement of how social capital may operate differently at distinct levels.

It must also be noted that this project does not address the intersectionality between migration and other phenotypic attributes such as race or gender as well as the intersectionality between migration and ethno-cultural attributes such as religion. Intersectionalities are expected to influence the way in which embeddedness affects participation in politics. Gender, for example, is expected to affect political participation, but the directionality of this influence may hinge on the culture or national origin of the woman (Tillie and Slijper 2006). Theories of the conjugal contract suggest that as women’s economic contribution to the family increases, these women can renegotiate their position in the household (Sen 1999). In other words, a shift in gendered roles favoring female migrants may translate to their empowerment (Jones-Correa 1998; Verba et al. 1995). The ethnographic record shows that a number of national groups in Italy are headed by female migratory chains. Within these communities, women often hold important positions of leadership and responsibility. Conversely, women that come at the tail end of a male-headed migratory chain often find themselves suddenly thrust into the role of “symbolic markers of ethnicity who are responsible for securing and maintaining the boundaries between their ethno-cultural community” and consequently experience more acute forms of subjugation (Spitzer et al. 2003). This article does not test for interaction effects between gender and culture. Further theorizing and instrument developing needs to be undertaken to address these limitations.

Conclusion

With very few exceptions (for an example, see Eisenberg 2013), most of the scholarship on immigrant political participation seems to favor the enfranchisement of immigrants (Justwan 2014; Lenard 2014; Hayduk 2003; Jacobs 1999). At the heart of these arguments are questions of social justice. First, the unequal distribution of voting rights among Europe’s non-citizen resident population creates a layered system of belonging. For example, the enfranchisement of European migrants through 1997’s Amsterdam Treaty establishes a system of preferential treatment, which only considers country of origin and ignores tenure and other investments in the receiving country. These and other rules that govern migrant electoral participation have led to the establishment of diverse regimes of inclusion, exclusion, and discrimination (Earnest 2014; Shaw 2006). Second, a main argument regards migrant rights to self-determination. Specifically, it has been posited that electoral participation is necessary for immigrants to defend their economic and political rights and, in essence, to be able to fulfill themselves (Lenard 2014). This argument, however, leads to an oversimplification of the way in which we view immigrants. The diversity of immigrants, the challenges they face, their needs, and political preferences cannot be understated. Thus, where we observe systematic differences across ethno-cultural groups in their participation, we should also expect the political representation of immigrants to be skewed in favor of some groups and not others. To the extent that social capital in its varied forms contributes to this skewness in the representation of immigrants, further investigation of this aspect of participation should be interesting to policy scholars and policy makers alike. The systematic overrepresentation of some groups relative to others may hold consequences for democratic governance and calls attention to the need for a scholarly examination of these processes.