Methodological Appendix
This book is based on research that compares six cases – three national groups (Libyan, Syrian, and Yemeni) across two host-countries (the United States and Britain) – using original, comparative evidence on the Arab Spring abroad. The unit of analysis is diaspora movement by national group (Libyan, Syrian, and Yemeni) and host-country (the United States and Britain); the units of observation are individual diaspora activists and movement groups. Initially, I justified this research design as providing what social scientists call a comparison based on “most similar” cases, since the Arab Spring gave new life to anti-regime activism, protest movements, and social movement organizations in each diaspora group. However, while the Arab Spring signified a new wave of mobilizations against authoritarian regimes, I came to discover that their collective actions varied in notable ways. This presented me with an ideal opportunity to explain this variation among the three national groups residing across two host-countries.
Before and upon arrival in each locale where interviews took place, which included Los Angeles, DC, New York, London, and Manchester, I used what social scientists call snowball and purposive sampling techniques in each community. These methods, which are common to case-based social movement research, allowed me to contact activists via referrals from their peers and to request interviews from specific organizations and groups, including women- and youth-run initiatives, in order to make the study as inclusive as possible. Snowball sampling provides access to relatively “hidden” populations and their “interactional units”; it also draws on the insider knowledge of those who know the relevant participants in a given “strategic action field” (Biernacki and Waldorf Reference Biernacki and Waldorf1981: 141; Fligstein and McAdam Reference Fligstein and McAdam2012). This was especially important in my case since many organizations were dominated by older male elites (see also Chapters 2 and 4 on this point). I additionally used social movement websites, public Facebook pages, and media reports to identify participants and avoid sampling within insular networks. I made efforts to reach out to activists who had founded or led prominent diasporic pro-revolution organizations, as well as those who were identified by their peers as having contributed in a meaningful way to rebellion and relief efforts. What a “meaningful” contribution or being an “activist” in these contexts meant was intentionally kept open-ended so as not to prematurely assume or limit the kinds of activities that were undertaken and considered important by members. Interviewees often provided me with information and referrals across networks because they recognized that I did not just want to hear “one side” or about one type of experience. They often posited this as an invitation – talk to them and they will tell you! – as a way to affirm the veracity of their accounts, even if they did not agree with or get along personally with the referred-to individuals.
As described in the Introduction, this study is based on primary data that include interviews with 239 individuals (of which 231 were members of the three national groups studied); 30 ethnographic participant observations of diaspora-sponsored events, including fundraisers, concerts, and picnics; and secondary data such as electronic media (e.g., online newspapers, streaming news services, blogs, and social media such as Facebook and Twitter), think tank reports, documentary films, and scholarly books and memoirs published on the Arab Spring revolutions. Further information about the characteristics of my interviewees is listed in Table A.1. The ensuing dataset includes approximately three hundred hours of digital recordings and two thousand pages of single-spaced transcriptions and field notes. I sorted the data according to the principles of process tracing (George and Bennett Reference George and Bennett2004) and grounded methods (Charmaz Reference Charmaz2006; Glaser Reference Glaser1965; Strauss and Corbin Reference Strauss and Corbin1990), using a sample of the interviews to derive open codes denoting activists’ experiences, such as “delivered aid to refugee camps” and “protested on the National Mall.” I then grouped and refined the codes into focused categories, such as “volunteering on the front lines” and “broadcasting,” using NVivo software, which helps qualitative researchers to lump and sort their data according to empirical patterns and emergent conceptual categories (Lofland et al. Reference Lofland, Snow, Anderson and Lofland2006[1971]).
Table A.1 Characteristics of respondents by country of origin
| Descriptors | Libya | Syria | Yemen | |
|---|---|---|---|---|
| Sex | ||||
| Male | 43 (67.2%) | 59 (77.6%) | 65 (71.4%) | |
| Female | 21 (32.8%) | 17 (22.4%) | 26 (28.6%) | |
| Immigrant generation | ||||
| Firsta | 46 (71.9%) | 58 (76.3%) | 68 (74.7%) | |
| Second | 18 (28.1%) | 17 (22.4%) | 22 (24.2%) | |
| Third | 0 (0.0%) | 1 (1.3%) | 1 (1.1%) | |
| Age at revolution’s onset | ||||
| 15–24b | 12 (18.8%) | 20 (26.3%) | 22 (24.2%) | |
| 25–34 | 24 (37.5%) | 20 (26.3%) | 37 (40.6%) | |
| 35–44 | 14 (21.9%) | 21 (27.7%) | 18 (19.8%) | |
| 45–54 | 13 (20.3%) | 14 (18.4%) | 10 (11.0%) | |
| 55+ | 1 (1.5%) | 1 (1.3%) | 4 (4.4%) | |
| Minority status | ||||
| Ethnic minority | 4 (6.3%) | 7 (9.2%) | 0 (0.0%) | |
| Religious minority | 0 (0.0%) | 3 (3.9%) | 0 (0.0%) | |
| South Yemenic | – | – | 25 (27.5%) | |
| Total | 4 (6.3%) | 10 (13.1%) | 25 (27.5%) | |
| Host-country | ||||
| United States | 37 (57.8%) | 49 (64.5%) | 34 (37.4%) | |
| Britain | 27 (42.2%) | 25 (32.9%) | 57 (62.6%) | |
| Both | 0 (0.0%) | 2 (2.6%) | 0 (0.0%) | |
| Active in collective efforts against regime before 2011 | 16 (25.0%) | 16 (21.1%) | 14 (15.4%) | |
| Self/family forced to emigrate due to repression before 2011 | 32 (50.0%) | 24 (31.6%) | 7 (7.7%) | |
| TOTAL | 64 (100.0%) | 76 (100.0%) | 91 (100.0%) | |
a Not all first-generation participants emigrated from Libya, Syria, or Yemen directly.
b All participants were interviewed in accordance with Institutional Review Board protocols.
c South Yemeni is an important minority regional identity in Yemen.
Interviewees’ preferences dictated where our conversations took place, which included Yemeni Community Associations, Syrian organization offices, cafes, restaurants, and in respondents’ homes. In many cases, I was privy to what historian Sarah Gualtieri (Reference Gualtieri2020: 14) calls “archival transactions”; during these transactions, the researcher is “shown material from family papers, given videos and photographs,” and other material that supplements interviewees’ oral histories and testimonials. In numerous cases, I was shown photographs of family members who had been imprisoned by regimes, grainy cell phone videos and photographs of anti-regime events, handmade flyers advertising events, typed catalogs and informational materials produced by activists on regime atrocities, PowerPoint presentations used to communicate claims to policymakers and the media, and emails with respondents’ contacts in the US government and NATO. While I was not “given” these materials to publish or own, these transactions provided important supplementary evidence of members’ activities and illustrated the character of their mobilization dynamics and interactions in full color, so to speak.
This project illustrates the methodological promise of using a grounded approach (Charmaz Reference Charmaz2006) to understand and analyze collective action dynamics among social groups. This is particularly important for the study of populations that remain largely invisible in social science research due to an iterative cycle of theoretical neglect and a lack of existing historical and survey data (see also my point on under-counted groups in the Conclusion). In this way, this study highlights the importance of substantiating community dynamics that are often subsumed under aggregated characteristics (such as “Arab,” which excludes ethnic categories like Syrian Kurdish and Libyan Amazigh) through interview and ethnographic data-collection methods. The findings also emphasize the importance of taking respondents’ accounts seriously. Such accounts can reveal perceptions and experiences that remain unaccounted for in other types of data sources, as I find in the case of transnational repression and conflict transmission. Furthermore, in contrast to predominant analytical tendencies in movement studies to focus narrowly on the emergence and frequency of protest events visible in major newspapers (Earl et al. Reference Earl, Martin, McCarthy and Soule2004), the grounded approach used here sheds light on activists’ fuller transnational tactical repertoires, including the ways they worked behind the scenes to facilitate rebellion and relief.
At the same time, this approach does not mean that investigators must rely solely on respondents’ testimonials to understand their collective dynamics. Accounts should instead be used to formulate exploratory questions that can be investigated in depth and tested across cases. These accounts should also be triangulated with external data sources whenever possible, as I have done here, and grounded in comparisons that point to causal factors and processes (such as geopolitical support in the form of military and humanitarian intervention) (McAdam et al. Reference McAdam, Tarrow and Tilly2001). This further emphasizes the usefulness of comparative case-study research in revealing patterns across cases in ways that defy prediction and insiders’ beliefs (Bloemraad Reference Bloemraad2013; Ragin Reference Ragin2000, Reference Ragin2008; Ragin et al. Reference Ragin, Nagel and White2004; Yin Reference Yin2008). In its earliest inception, this study was designed as a binational comparison because reports by Yemeni activists attested that diaspora activism during the revolution had varied significantly between the US and British contexts in ways that warranted investigation. Through extensive, comparative fieldwork across the two countries, I discovered that in contrast to what many Yemeni respondents believed – that other diaspora communities had done a better job at mobilizing to help the revolution than they had – Yemeni activists across local and national contexts were instead impeded by a common set of challenges. Although the orienting assumptions that shaped the initial research design turned out to be unsupported by the data, the binational case comparison nevertheless revealed how the mobilization dynamics of diaspora groups across communities are shaped by a similar set of conditions. It also suggested the importance of expanding the comparison to other national groups, which proceeded according to my description in the Introduction. In this way, the comparison revealed important information about similarities in the mobilization patterns of conationals across contexts, differences between the three national groups, and the mechanisms producing these findings.
Open access