Sage Journals: Discover world-class research

Abstract

Scholarship at the intersection of migration studies and international relations has been increasingly interested in diasporas in world politics, yet analysis integrating relevant literature in systematic ways through rigorous empirical methods has been largely missing. How can mixed-methods techniques contribute to the evolution of a relatively new research program, such as the study of diaspora mobilizations in conflict processes? How can these techniques identify data patterns that could help establish meaningful analytical categories and factor in multi-sited complexity? Based on a large-scale European Research Council migration project "Diasporas and Contested Sovereignty," this Research Note offers a pathway for future scholarship on how to systematically do so. It presents a rich theoretically informed coding procedure that systematizes the analysis of four researchers working across conflict-generated diasporas in multiple host-countries and linked to conflicts in multiple countries of origin. Correspondence and cluster analyses are used to isolate profiles of diaspora entrepreneurs. Procedures used early in the project's life-cycle helped later to design a unique survey and inform a deeper comparative causal analysis, thus creating a coherent conversation among different research products.

Keywords

diaspora conflict multi-methods

Introduction

Scholarship at the intersection of migration studies and international relations has been increasingly interested in the transnational activity of diasporas in world politics (Collier and Hoeffler 2000; Østergaard-Nielsen 2001; Shain and Barth 2003; Kapur 2005; Adamson and Demetriou 2007; Orjuela 2008; Koinova 2009; Brinkerhoff 2016; Marinova 2017). Yet analysis that integrates relevant literature in systematic ways through the rigorous coding of large-scale migration data is still largely missing. This paper revolves around two major questions: How can mixed-methods techniques contribute to the evolution of a relatively new research program, such as the study of diaspora mobilizations in conflict processes? How can these techniques identify data patterns that could help establish meaningful analytical categories and factor in multi-sited complexity?

This Research Note presents a rich theoretically informed coding procedure that systematizes the analyses of four researchers working on a large-scale research project on conflict-generated diasporas in multiple European host-countries linked to conflicts in multiple countries of origin. Initially facing scattered literatures that could not adequately cover the complexity of diaspora mobilizations in breath and variability, we introduced a comprehensive coding of qualitative data followed by quantitative correspondence and cluster analyses, and isolated the profiles of diaspora entrepreneurs as an important analytical category. Such profiles allowed us to not confine the notion of diaspora to that of a group, but to highlight individual agency, which is much needed in scholarship and practice but difficult to arrive at without access to large-scale data.

This paper briefly reviews migration and international relations scholarship relevant to diaspora mobilizations and crucial to our integrative analysis. Further, we discuss how we conducted semi-structured interviews with diaspora political entrepreneurs in Germany, Sweden, and the Netherlands in 2012–2016. Interviews were systematically coded on the basis of 11 codebook sections, subjected to inter-coder procedures, and analyzed through cluster and correspondence analyses. We conclude by reviewing the impact of these early procedures on subsequent research products—a cross-national survey and a comparative book—that advanced the research program in a coherent way.

Diasporas in World Politics

At the outset of our investigations, several lines of research provided varying insights into diaspora activities in conflict, post-conflict, and development. A large-N World Bank study argued that if an intra-state conflict draws resources from an affluent US-based diaspora, conflict is perpetuated (Collier and Hoeffler 2000). These results turned contradictory, as a new quantitative study showed that large diasporas can significantly reduce conflict risks (Collier, Hoeffler and Söderbom 2008). Case studies of Armenian, Chechen, Kosovar, Sikh, Somali, and Tamil intra-state conflicts found that conflict resolution was resisted when diasporas sponsored rebel factions (Lyons 2006; Adamson and Demetriou 2007). Diasporas can also affect the purchase and smuggle of weapons, provide sanctuary for rebels, and draft soldiers from diaspora circles (Byman et al. 2001). Other scholars have argued that diasporas can have both positive and negative influences on conflict dynamics with their remittances, philanthropy, human capital, and policy influences (Smith and Stares 2007; Orjuela 2008; Koinova 2011; Brinkerhoff 2011). It eventually became clear that research needed to better understand the conditions and context-shaping behaviors of diasporas. More researchers started calling for the use of comparative and quantitative methods.

For their part, development studies focused on diasporas more generally, and beyond the conflict-generated ones discussed here. Diasporas send much-cherished remittances back home, constituting 10–26 percent of a fragile state's GDP (World Bank 2017). Remittances maintain households during warfare and rebuild lives, housing, and infrastructure in the aftermath. Diasporas also invest in various enterprises (Smart and Hsu 2004), buy diaspora bonds (Leblang 2010), and make philanthropic contributions (Sidel 2003; Brinkerhoff 2008). Diasporas act upon political opportunities and constraints by mobilizing structures available to them locally and globally (Østergaard-Nielsen 2001; Smith and Stares 2007).

Diaspora entrepreneurs are important in such movements. We acknowledge the growing scholarship on them from a business perspective (Zapata-Barrero and Rezaei 2020), yet focus on its political dimensions, especially on mobilizations in conflict processes. We define diaspora entrepreneurs as “formal and informal leaders in a diaspora community, associated with migrant, religious, and other identity-based institutions, … acting autonomously as activists, businessmen, politicians, … who actively make public claims with a homeland-oriented goal. These are political and social entrepreneurs, even if some of them may have a business background. What unites them is a strong commitment to a cause related to their original homeland” (Koinova 2021: 11).

Diaspora entrepreneurs can bridge structural gaps through “brokerage” (Adamson 2013), “in-between advantage” (Brinkerhoff 2016), and their socio-spatial positionality in different global contexts (Koinova 2017). They can mobilize more or less contentiously through: (a) Different channels, preferring state-based or transnational ones to organize their activities and (b) different levels of intensity ranging from weak, medium-strong, to strong. Discussing diaspora entrepreneurs as political actors, Brinkerhoff sheds light on diaspora individual agency from a leadership perspective, drawing evidence from the US (2016: 29–32). In contrast, this Research Note identifies how profiles of diaspora entrepreneurs are contextually embedded in Europe and maintain connections to their original homelands.

Delving deeper, we aimed at combining five relevant aspects of diaspora mobilization discussed separately by existing scholarship: (1) Diaspora group characteristics; (2) host-state migration integration regimes; (3) host-state foreign policies towards the home-state; (4) critical events in the homeland; and (5) homeland policies towards diasporas abroad. Considering these domains through common methodological procedures allowed us to extract from semi-structured interviews with diaspora entrepreneurs a variety of identity-based, contextual, and behavioral “traits” that shape their “profiles.” We discuss these domains below and in more detail in the codebook.

For diaspora group characteristics, size, organization, and spatial concentration stand out. The literature struggled to ascertain diaspora size, as diasporas consist of migrants and refugees, their descendants, and different generations, rendering unclear who belongs within the group. Diasporas are also not counted equally in censuses. Larger diasporas are deemed more likely to mobilize than smaller ones (Cohen 1997). A diaspora's organizational level often depends on age and size, with old, large diasporas usually better organized than new ones. Organizationally strong (Rubenzer 2008), unified (Haney and Vanderbush 1999), and spatially concentrated diasporas are also more likely to mobilize (DeWind and Segura 2014).

In host-state migration incorporation regimes, how citizenship is acquired holds a central role. Compared to the unrestricted jus soli (law of the soil) principle in the US and Canada, some European limitations exist, even in the most open cases, such as the UK and France. Germany's citizenship law has been historically given on the jus sanguinis (law of blood) principle, although it was amended to more open methods after 2000. In Sweden and the Netherlands, citizenship is also conferred on the jus sanguinis principle, with possibilities for naturalization with residence over time. We also considered other institutional arrangements, such as multicultural versus assimilation policies (Banting and Kymlica 2006), isolation of migrants in host-states (Ireland 1994; Koopmans and Statham 2001), incorporation regimes considered with migrants’ collective identities, and homeland influences.

Regarding host-state foreign policies towards the home-state, the most striking characteristics of diaspora groups that are actively and successfully lobbying are those that are partially assimilated, organizationally strong, and allied with other groups (Rubenzer 2008; Shain and Barth 2003; Sheffer 2003). Powerful ethnic lobbies can be instrumental in the “policy capture” of host-state foreign policies and in fostering interventions in conflicts abroad (Mearsheimer and Walt 2007). Until recently scholarship on diaspora lobbying in foreign policy remained largely state-centric, yet emerging discussion has considered a transnational dimension (Koinova 2017; Marinova 2017). Building on their geographical “comparative advantage,” rebel groups strategize on how best to engage diasporas (Lyons and Mandaville 2010), including how to contribute to a homeland cause, education, and media activism (Adamson and Demetriou 2007).

Regarding critical events in the homeland, we have identified major stimuli for diaspora mobilization. Critical junctures and transformative events are especially impactful if violent (Sökefeld 2006: 273), leading to an expansion in diaspora activism (Demmers 2007) and changes in statehood (Koinova 2018a). Regime change can also be important (Kuropas 1991). Such critical events can affect diasporas if they are durably linked to kin in the homelands, as in the 2011 Arab uprisings.

Homeland policies toward diasporas abroad can occur for a variety of reasons: As a resource for remittances, investments, purchasing bonds, philanthropy, and tourism. Sending-states seek to construct or maintain diaspora identities through schools and curricula in the mother-tongue, commemorations of national holidays, and retainment of dual citizenship (Ragazzi 2009; Waterbury 2010). Sending-states govern their diasporas directly or indirectly, fostering their self-reliance and encouraging homeland-bound contributions (Delano and Gamlen 2014).

None of these five theoretical domains could single-handedly or coherently explain diaspora mobilizations during conflict. Hence, one needs to examine the domains concurrently, consider how mobilizations occur in context, and eventually isolate profiles of diaspora entrepreneurs shaped by these contextual dimensions.

Collecting and Coding Diaspora Interviews

Given the growing number of large migration projects involving multiple individual researchers working on a common theme,¹ this Research Note demonstrates how a specially designed coding system can help analyze interview transcripts through a mixture of qualitative and quantitative methods to identify data patterns across researchers’ sub-projects. In our case, the emerging patterns revealed “profiles” of diaspora entrepreneurs. We went through three phases, as indicated in Table 1.

Table 1.

Investigation and Coding Process.

Preparation phase	Wave 1 of coding	Wave 2 of coding
February 2012–June 2014	June 2014–May 2016	May 2016–December 2016
1. Scattered theories inform the launching of a comparative study of diaspora mobilizations across Europe and the Balkans, Middle East, and Caucasus regions.	1. A team of six coders is introduced to a first version of the questionnaire, elaborated from 10 sections and 125 codes derived from the literature.	1. A second, distinct subsample of 40 interviews is extracted, stratified among three diaspora groups (Albanians, Armenians, and Palestinians) and a subset of three countries (Germany, the Netherlands, and Sweden) and decomposed into units of analysis (unitized).
2. A team of four interviewers conducts 300 interviews among an overall sample of six diaspora groups linked to weak and de facto states in the EU neighborhood (Albanians, Armenians, Bosnians, Kurds, Iraqis, and Palestinians) in five EU countries (France, Germany, the Netherlands, Sweden, and UK).	2. A first sub-sample of 24 interviews are selected at random from the 300, with interviews from the six diaspora groups and five states.	2. Two independent pairs of coders test the enhanced codebook against a pilot sample of six unitized interviews, then 20, out of the 40. They propose and discuss more amendments to the codebook, finalized at 11 sections and 138 codes.
	3. Interviews are transcribed, then used to revise the codes through grounded coding analysis.	3. The final coding of the 40 unitized interviews is taking place.

During the Preparation Phase (Table 1), large-scale fieldwork was conducted among six diaspora groups in five EU countries, targeting diaspora entrepreneurs. The groups were selected on their linkages to different types of weak and fragile states.² Albanians, Armenians, and Palestinians are linked to the de facto states of Kosovo, Nagorno-Karabakh, and Palestine, polities lacking international legal sovereignty. Iraqis and Bosnians are linked to Iraq and Bosnia-Herzegovina, respectively, weak states with international recognition but challenged in their domestic sovereignty by deep ethno-nationalist or sectarian divisions. The Kurdish diaspora is stateless and linked to autonomist and secessionist movements in the Middle East. The five European countries were selected based on their different types of migrant incorporation regimes (UK and France traditionally considered “liberal,” Germany as more “restrictive,” and Sweden and the Netherlands as taking a middle ground; based on Howard 2009). The investigation considered over 300 open-ended and semi-structured interviews, focus groups, and participant observations during diaspora events.

Interviewees were selected through purposive and snowball sampling to represent active diaspora members who are involved in claim-making and organizing of others about homeland-based political and social projects. Respondents were sampled from different networks, to avoid selection bias. Some of the interviewees belonged to diaspora organizations, while others were wealthy individuals who did not need diaspora organizations to endorse them. Still, others held jobs that brought them in frequent contact with the diaspora community, such as shopkeepers, restaurant owners, and hairdressers. This open definition of diaspora entrepreneurship captured a multitude of people who made claims about their original homelands, or mobilized resources for specific projects.

In the Wave 1 coding phase, following Saldana's method for qualitative coding (2013), topics were grouped into ten sections: life in the homeland prior to migration, migration experience, host-country integration, citizenship, relationship to host-land foreign policy and homeland government, substantive homeland-oriented goals and claims, prospects for return, and modes of mobilization. Codes were elaborated from a random sample of interviews from the six diaspora groups (Albanian, Armenian, Bosnian, Iraqi, Kurdish, and Palestinian). Group-level comparisons revealed commonalities, including that individual interviewees (a) do not speak about specific host-land policies, but about their individual perceptions and how they act on them; and (b) are very aware of how supportive host-state foreign policy is regarding homeland-oriented goals, but rarely know specific details. They act on such general perceptions. This first sample of interviews was set at twenty-five, which enabled us to test the range of theoretical categories without exhausting our resources.

The Wave 2 phase consisted in selecting forty interviews from three diaspora groups (Albanian, Armenian, and Palestinian) and three host-countries (Germany, the Netherlands, and Sweden), which enabled us to cover a range of themes relevant to diaspora mobilizations, and guaranteed a degree of diversity among socio-demographic profiles. Every eighth interview was randomly selected from the existing pool of transcribed interviews. Our goal was an in-depth analysis of these forty interviews using quantitative methods to reveal certain patterns across the cases.

This selection elucidates relationships concerning the de facto states mentioned earlier. Our goal was to generalize cautiously about the wide range of diaspora mobilizations from these fragile states. Narrowing this probe to mobilizations in three host-states—Germany, the Netherlands, and Sweden—allowed us to reduce the number of contexts in which to interpret the results. Interviews were then divided into small units, called “conversational moves,”³ or building blocks “for conversational game structure, that reflect the goal structure of the dialogue” (Carletta et al. 1997: 22). Given the multiple themes covered in the different literatures, using the conversational moves enabled us to effectively classify and code smaller segments of text to answer a single question, stated or implied, or multiple but closely related questions dealing with a single, self-contained topic.

The codebook and coding protocols were applied progressively to the forty interviews and improved accordingly. Coders were initially instructed to assign up to two codes to each conversational unit. Intercoder discussions during a pilot phase revealed that these two codes should not be hierarchized (into master and secondary codes), as this made it more difficult to code in subsequent treatments. Eventually, a “negotiated agreement” approach was adopted, so that the two coders could “code a transcript [independently], compare codings, and then discuss their disagreements in an effort to reconcile them and arrive at a final version in which as many discrepancies as possible have been resolved” (Campbell et al. 2013: 305). The final codebook (see Appendix 1) underwent twelve drafts to refine the existing categories.

Analyzing Interview Data

We used successively correspondence analysis (CA) and cluster analysis. We did not formulate and test hypotheses deductively, due to the scattered state of the literature. Nor did we arrive at our conclusions purely inductively, as we started with some theoretical insights from the earlier discussed scholarly domains. We evaluated the coding propositions and amended them sequentially, making the final findings more robust.

CA, our primary (quantitative) analytical method, is a type of factorial analysis focused on categorical variables (Benzecri 1992; Le Roux and Rouanet 2004; Greenacre and Blasius 2006). CA can be counterintuitive to scholars used to testing hypotheses but is helpful at establishing systematic empirical patterns (here diaspora entrepreneurs’ profiles) within a relatively new research program dominated by case studies and some comparative studies, where causal relationships have not been tested yet. CA can cluster personal characteristics, mobilization pathways, and entrepreneurs’ experiences, and establish specific patterns of contextual entanglement.

CA centers on individuals with attributes (e.g., “being a woman,” “having migration experience of political violence,” “mobilizing for one's homeland independence”). CA does not capture ties between individuals connected in a network as social network analysis does. We did not investigate such ties. The merit of CA is different: To capture how the above-mentioned attributes are tied to individuals with similar attributes, within various groups and contexts in the same social field, and to calculate the mathematical distance between cases (here diaspora individuals in Europe linked to conflict and postconflict polities in its neighborhood). CA also has the specificity of progressing inductively and acquiring robustness through ex post statistics.

CA has proven fruitful through its application to multifactorial phenomena, such as the social construction of tastes (Bourdieu 1984), social reproduction in education (Bourdieu and Passeron 1990), social structures and political hierarchies (Hjellbrekke et al. 2007), structure and stability of political elites (Bühlmann, David and Mach 2012), and evolution of electoral constituencies (Blanchard 2007). Ragazzi's (2009) work is an exception in having used CA to study aspects of diaspora politics, most notably how sending-states reach out to diasporas abroad.

We followed CA's three usual steps: (1) Calculate (hidden) factors that summarize optimally the information from the original dataset; (2) assess and interpret factors by means of maps that determine distance between cases and codes, and joint statistical outputs, where each factor is described through a contrast between its extremes; and (3) use the calculated factors to elaborate a classification of interviewees characterized by cross-tabulations and prototypical individuals.

Our analysis is based on the transcripts of forty interviews, stratified by diaspora group and host-country. CA is appropriate for our analysis, as statistical representativity is less important than the richness of cases, the structure of similarities between them, and how all cases give shape to the field of diaspora mobilizations. On average, each interview was divided into thirty-nine conversational moves and attributed thirty-two unique codes, chosen from a list of 138 provided in Appendix 1. Codes refer to values of variables, mostly categorical, often binary. A few exceptions are ordinal, such as age range, or public opinion about immigration, with three nominal values (Negative, Divided, or Positive). Due to the predominance of binary codes in the dataset, we use simple CA instead of the more popular multiple CA (Le Roux and Rouanet 2004: 2061). Only codes used as “active variables” contribute to building the model. Respondents’ host-countries (Germany, the Netherlands, and Sweden) and diaspora groups (Albanians, Armenians, and Palestinians) do not: they are used as “illustrative variables.” This formal homogeneity of active variables (mostly binary) keeps the analysis statistically consistent and sociologically meaningful, i.e., consistent with our ability to interpret variables from the perspective of theory.

Figure 1 shows twenty extracted factors with a fairly smooth decrease in explanatory power (variance). Such a vast and diverse corpus is fully multifactorial. Axes 1, 2, and 3 account for 21 percent of the explained variance, and appear to structure the experience of diaspora entrepreneurs. This relatively low variability is the rule more than the exception in CA. This reflects the fact that many aspects of the multivariate combinations of codes are not summarized by the clustering. Nevertheless, CA extracts the largest contrasts within the space of codes produced by the coded corpus.

Figure 1.

Distribution of factors extracted by means of correspondence analysis.

Axis 1 contrasts those diaspora entrepreneurs who are reasonably well “segmentally assimilated” (Portes and Zhou 1993) into the host-society with those who make homeland-oriented claims based on political grievances. This does not mean that a person who is segmentally assimilated cannot potentially make homeland-oriented claims based on political grievances, but that the opposition between these two characteristics is structuring axis 1. This is the first and largest axis, yet it exhausts only part of the total explained variance and may not be structuring for other axes. Axis 2 contrasts individuals shaped mostly by experiences in the original homeland with those engaged in host-state institutions or transnationally. Axis 3 (Appendix 2) contrasts individuals who are predominantly self-employed and engaged mostly in communal activities with those based in migrant-dominated areas and more actively involved in their host society. Overall, the main three axes contrast personal experiences of migration and settlement with balancing identity between home and host-country. This means that some respondents engage better than others with the host-land, and diaspora mobilizations are constrained in some cases, or more enabled in others.

Exposing Clusters and Profiles of Diaspora Entrepreneurs

An examination of the axes reveals the main differences between diaspora entrepreneurs. However, these remain abstract from the reality of the respondents. To present respondents’ specific characteristics more concretely, we further cluster them. The concept of statistical clustering is similar to other approaches to build typologies but differs insofar as it relies on precise calculations (its algorithm). Clustering aims to maximize the similarities between cases belonging to the same cluster and to minimize them between those belonging to distinct clusters.

Here cluster analysis is based on the three main axes detailed above, plus some minor ones, grouping sets of interviews with similar combinations of properties over the range of 138 codes. All interviews are examined in pairs, with the most similar clustered first. All remaining interviews and clusters are examined in the same way, and those most similar are clustered again. Eventually, each small branch of the “clustering tree” merges into a bigger branch. The process moves from the forty distinct interviews (left side of Figure 2), towards the whole population (right side).⁴

Figure 2.

Clustering tree. The 40 interviews on the left-hand side are aggregated progressively from left to right, from pairs with the most similar codes to most dissimilar. The horizontal axis materializes dissimilarity between interviews: The more similar two interviews, the shorter the fork that pairs them.

Labeling the clusters associates human characteristics with the statistical modeling results, and creates diaspora entrepreneur profiles based on a combination of extracted “traits,” mobilization modes, and contextual embeddedness. A “constrained” diaspora entrepreneur has relatively restricted access to host-institutions, yet still lobbies them through media, protests, and boycotts. A “contented” diaspora entrepreneur has a more positive outlook of host-country institutions, yet is more limited in mobilizing. An “enabled” diaspora entrepreneur has soft powers based on economic resources and relies highly on education to engage with their original homeland, whereas a “discontented” diaspora entrepreneur believes conditions in the host-country and its public opinion to be negative or divided and demonstrates strong transnational activism.

Cluster 1, or the constrained diaspora entrepreneur, is the largest: It contains nineteen interviews, gathering 44 percent of the codes. It occupies the lowest third of the clustering tree (Figure 2), and is over-represented in the Netherlands and among Armenians and Palestinians (Table 2). This cluster derives codes nearly exclusively from six codebook sections (perception of host-countries, access to host-institutions, participation in host-country, types of claims, activism modes, and events triggering mobilization). This is a striking result, as no statistical rule explains why characteristic codes for a cluster would be so selective regarding the code sections.

Table 2.

Diaspora Entrepreneurs and Their Mobilizations.

Diaspora entrepreneur profile	Percentage of sample codes in context	Host-land related aspects to mobilizations	Homeland related aspects to mobilizations	Issues of mobilizations	Modes of mobilizations
Constrained - 44 percent of overall sample	Host-country:GER 26 percentNL 37 percentSWE 37 percentDiaspora group:ALB 5 percentARM 47 percentPAL 47 percent	Considering themselves having limited access to host-land institutions but better access to political parties; concerns about rival lobby groups; host-land foreign policy considered adverse for pursuing homeland-oriented political goals	Events triggering mobilizations: anniversaries and warfare; limited participation in homeland institutions	Religion, genocide recognition, less frequently human and gender rights.	Lobbying, protests, boycotts
Contented - 21 percent of overall sample.	Host-country:GER 25 percentNL 25 percentSWE 50 percentDiaspora group:ALB 38 percentARM 25 percentPAL 47 percent	Positive view of host-land institutions and society; some membership in host-land associations	Leaving home due to violence and/or economic reasons; some have homeland-based education	Views that the host-land needs to recognize independence of the homeland	Minimal mobilizations
Enabled - 6 percent of overall sample	Host-country:GER 75 percentNL O percentSWE 25 percentDiaspora group:ALB 50 percentARM 20 percentPAL 20 percent	Well settled in the host-country usually through the mainstream of society; soft powers and education	Importance to build connections between host-land and homeland; homeland government seeking ties with diaspora abroad	Partnerships, trade, education,	Lobbying, advocacy
Discontented - 29 percent of overall sample	Host-country:GER 44 percentNL 33 percentSWE 22 percentDiaspora group:ALB 100 percentARM 0 percentPAL 0 percent	Views conditions and public opinion in host-land as negative and divided, concerns about discrimination and limited job and other opportunities	Well linked to a homeland, also through political parties; more likely to wish to return; reiterate that diaspora sends remittances	Calling for host-country to refrain from discrimination, for homeland to introduce profound political reforms.	Strong transnational activism

Cluster 1, the constrained diaspora entrepreneur wishes they had more access to host-institutions. They advance claims associated with religion, genocide recognition, and less frequently gender and human rights. They define themselves by their ability to mobilize, struggling to do so at times because of moderate host-country support or limited personal involvement in host-country institutions. They are not strongly linked with their homeland, as if marooned between it and their host-country, and can't access either's valuable resources.

Cluster 2, the contented diaspora entrepreneur, is the third-largest. It contains eight interviews, gathering 21 percent of the codes (middle block, Figure 2), and is slightly over-represented in Sweden and among Palestinians (Table 2). In contrast to Cluster 1, it is mostly defined by two sections of the codebook: Personal attributes and experience with migration and integration. Dominant codes are, in decreasing order of magnitude, having transited through the host-country (and/or possibly a third country), leaving home due to violence or economic reasons, being a woman, and having attended university, mainly in the homeland. These individuals have close friends among co-ethnics and host-country nationals, have a good command of both homeland and host-country languages, and want to integrate themselves in the host-country. Contented diaspora entrepreneurs share quite positive views on many aspects of integration, including knowledge of the host-country, social and institutional resources that can enable them to act and mobilize, and support from local institutions. However, they don't mention much about the kinds and modes of mobilization they favor. They are less politically active.

Cluster 3, the enabled diaspora entrepreneur, is the smallest in our dataset, grouping four interviews at the top of Figure 2, and gathering 6 percent of all codes. It is over-represented in Germany (Table 2). An enabled diaspora entrepreneur is well-settled in the host-country, mostly through the host-land mainstream. Education and other soft means of influence are crucial to their role as diaspora leaders. They talk much more than other respondents about diaspora educational activities aimed at the community (language classes, job-training schemes) and specific youth programs in the diaspora. They establish and maintain links between citizens of the host-country, home-country, and other locations around trade, partnerships, educational initiatives, and other activities, and are often engaged with homeland advocacy through social media. They describe tight relationships with the homeland, which seems to translate to their own successful involvement. Yet they also report disagreements with homeland institutions, a sign that they cannot reach all their objectives as diaspora leaders.

Cluster 4, of the discontented diaspora entrepreneur, is the second-largest in our dataset, with nine interviews and 29 percent of all codes (Figure 2). It is over-represented in Germany and among Albanians (9/9 respondents) (Table 2). This cluster centers on homeland relations with the diaspora. These are seen as weak when the homeland is not consistently engaged with its diaspora, but also as strong when homeland officials frequently consult diaspora leaders, appoint them to government positions, or seek their investments. Respondents indicate strong relations with homeland parties and extraterritorial electoral campaigns, as well as committed to remittances to support their families. A discontented diaspora entrepreneur views conditions and public opinion in the host-country as negative and divided, with the host-country discriminating against the diaspora and migrants more generally. Respondents stress calling upon the homeland to politically change, reform, and democratize its institutions and are discontent about the inability to temporarily or permanently return to political and economic conditions there.

Table 2 demonstrates how our methodology integrated aspects of various literatures, listing some elements of these and how they relate to specific diaspora entrepreneur profiles in context.

From Interview Profiles to Survey Design and Comparative Causal Analysis

The early methodological work in this large-scale migration project established a common codebook informed by data in individual research projects, and later provided the foundations for a cross-national survey and a book featuring comparative causal analysis. Therefore, the above-mentioned procedures should not be seen as seeking to create causal inferences, but as using qualitative and quantitative methodologies to essentially conduct a rigorous and systematic qualitative analysis. They touched on generalizability only minimally by identifying different diaspora entrepreneur profiles depending on context, and categories that emerged from data patterns that researchers would not have been able to identify otherwise. In-depth work on generalizability followed.

The cross-national survey was conducted among 3,000 respondents among the Palestinian, Iraqi, and Kurdish diasporas in the UK, Sweden, and Germany. Country groups from the original pool were modified toward conflict-generated diasporas specifically linked to the Middle East. The survey questionnaire drew heavily from the eleven codebook sections, yet it covered only aspects highlighted by the analysis as important for diaspora mobilization. For example, while there were many personal attributes in the original codebook, the survey questionnaire focused more on factors essential to the above-mentioned profiles, such as decisions to leave, experiences in transit, education status, and access to host-land institutions.

The survey respondents were “regular” diaspora members, not diaspora entrepreneurs who organize others; therefore, direct operationalization of profiles was neither intended nor possible. Yet, by using questions relevant to diaspora entrepreneurs, we were able to compare different categories of individuals within the same country-group and across groups. The sound empirical grounds enabled us to consider many factors already identified as having an impact on diaspora mobilization. Inspired by the early findings, we asked in the survey how diaspora members are being simultaneously embedded in different contexts and about the resulting political effects. Our survey explored how, for example, diaspora members are more or less integrated in their host-countries, while maintaining translocal ties with their places of origin, engaging politically in host-land and homeland elections, responding to critical events in the homeland or another country, campaigning for genocide recognition, or supporting independence movements and military interventions.

The second item that drew on this early methodological analysis is a comparative analysis, published in a book, "Diaspora Entrepreneurs and Contested States" (Koinova 2021). The book developed a typology of diaspora entrepreneurs and a typological theory of how they interact with external factors—host-state foreign policies, extraterritorial outreach of homeland governments and non-state actors, and critical events—to produce different levels of contention during mobilization. Informed by 300 interviews, this work unpacked major elements originally structuring the data-driven profiles, such as personal characteristics, contextual factors, and ways of mobilization, and studied them through a comparative causal analysis using the following procedures.

First, insights about the data-driven profiles were refined by identifying and prioritizing a diaspora entrepreneur's linkages to different global contexts crucial to the socio-spatial theory developed in that book. The diaspora entrepreneur types of the Broker, Local, Distant, and Reserved emerged. While these types broadly correspond to the contended, constrained, discontented, and enabled diaspora entrepreneurs profiles, respectively, the types were already conceptually derived and based on a single theoretical dimension: the relative strength/weakness of one's linkages to the host-land versus the homeland and other global locations. The Broker has strong socio-spatial linkages to both host-land and homeland (and other global locations); the Local has stronger linkages to the host-land and relatively weaker ones to the homeland (and other global locations); the Distant has relatively weaker linkages to the host-land but stronger ones to the homeland (and other global locations), and the Reserved has weak linkages to all these contexts. Hence, the types became an independent variable (IV) in the typological theory.

Furthermore, by refining insights from the original profiles, other contextual factors were singled out, such as host-land foreign policies, critical events, and other homeland-based influences. These became further IVs in the typological theory. Since the profiles also signified that diaspora entrepreneurs can act more or less contentiously, the category “contention,” important theoretically, was singled out as being analytically important and a dependent variable (DV) for the typological theory. Conceptually, this category was further unpacked into three nominal values: contentious, non-contentious, and dual-pronged mobilizations. IVs and DV were then introduced in a thorough process-tracing analysis, resulting in nine causal pathways that repeated themselves across different cases of conflict-generated diasporas linked to contested states. In contrast to the profiles, which captured data-driven patterns of multi-sited embeddedness, the typological theory took the analysis further. It used causal pathways to show when and how diaspora entrepreneurs act more or less contentiously, when interacting with specific factors from their political environment, spanning different global locations.

Concluding Remarks

This Research Note presents a novel contribution to scholarship at the intersection of migration studies and international relations in several ways. First, it presents a possible blueprint for conducting integrative analysis of dispersed theoretical streams in a relatively new research program, especially relevant for large-scale projects that must process the output of several researchers and their sub-projects simultaneously. As Horvath and Latcheva (2019) rightly observe, while migration research has used different methods—interviews, ethnographic analysis, text analyses, and surveys—very few research projects have integrated qualitative and quantitative methods as a coherent whole. Challenges emerge from migration being a transnational phenomenon, reflected in growing critiques of methodological nationalism (Wimmer and Glick-Schiller 2003), essentialism, and the need to consider multi-sited approaches (Beauchemin 2014; DAMR 2021; Fauser 2018).

This Research Note speaks directly to these concerns from the vantage point of one large research project. It seeks to provide a way to de-essentialize diasporas by analyzing transcripts featuring individual behaviors, and to use theoretically informed grounded coding in conjunction with quantitative methods to extract data patterns and produce meaningful analytical categories. Here diaspora entrepreneur profiles factor in multisited embeddedness, providing insights for a later generalization through cross-sectional survey research and comparative causal analysis. Other publications from this large-scale project have also benefitted from the initial systematic coding procedures. Three journal special issues have featured the importance of diaspora connectivities to different global contexts, on topics of conflict and post-conflict reconstruction, sending-states’ diaspora engagement, and transitional justice (Koinova 2018b, Koinova and Tsourapas 2018, Koinova and Karabegovic 2019). The early coding procedures were helpful in integrating the analysis of diaspora mobilizations across different global contexts and in selecting which factors to potentially prioritize for the analysis. Once the sound fundamentals of a new research program are established, the analytical endeavor could then move on to new data gathering, including longitudinal data, and creating causal models that factor in how contexts condition behaviors over time (see Kupchik, Highberger and Bear 2022).

Second, we have presented an under-utilized CA method. Coupled with cluster analysis it helped to identify meaningful analytical categories. Sometimes wrongly regarded as mere exploratory and descriptive, these tools are strong in the ways they spot structuring relationships between multiple attributes, here derived from thousands of lines of interviews, and analyzed in deep detail through a rich and structured coding grid, closely informed by theory. This comprehensive set of codes is intended to make best usage of information conveyed by respondents. CA also identifies factors that could be relevant for hypothesis testing. Without such careful analytical work, further quantitative analyses could suffer from omitted variable bias, feature variables without potential causal impact, and identify wrong hypotheses to test.

On a final note, our paper could be useful for policy-makers. This paper paints a much more nuanced picture than thinking of diaspora entrepreneurs as conflict mongers or peace-makers: individual diaspora entrepreneurs are preoccupied with different aspects of homeland-oriented activism. Some are content and enabled to influence transnational politics related to their original homelands; others feel more discontent or constrained. Policy-makers need to start thinking about the variety of diaspora entrepreneurs and their specific inclinations to engage in homeland-oriented activism.

Supplemental Material

sj-pdf-1-mrx-10.1177_01979183231172103 - Supplemental material for Mixed Methods in an Evolving Research Program on Diaspora Entrepreneurs and Their Mobilizations

Supplemental material, sj-pdf-1-mrx-10.1177_01979183231172103 for Mixed Methods in an Evolving Research Program on Diaspora Entrepreneurs and Their Mobilizations by Maria Koinova, Philippe Blanchard and Ben Margulies in International Migration Review

Supplemental Material

sj-docx-2-mrx-10.1177_01979183231172103 - Supplemental material for Mixed Methods in an Evolving Research Program on Diaspora Entrepreneurs and Their Mobilizations

Supplemental material, sj-docx-2-mrx-10.1177_01979183231172103 for Mixed Methods in an Evolving Research Program on Diaspora Entrepreneurs and Their Mobilizations by Maria Koinova, Philippe Blanchard and Ben Margulies in International Migration Review

Footnotes

Acknowledgments

Earlier versions of this Research Note were presented at the annual conventions of the American Political Science Association,European Consortium of Political Research,International Studies Association,the PAIS Departmental Conference at Warwick University,and a workshop of the ERC Project “Diasporas and Contested Sovereignty.” The authors would like to express gratitude to colleagues in those venues who provided excellent feedback,as well as to Jessica di Salvatore,Yehonatan Abramson,Charles Lauder,four anonymous reviewers,and the editorial team of International Migration Review,for their thorough and insightful comments and support for the publication of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The author(s) received financial support for the research,authorship,and/or publication of this article: H2020 European Research Council (grant number 294198).

ORCID iD

Maria Koinova

Supplemental Material

Supplemental material for this article is available online.

References

Adamson

2013. “Mechanisms of Diaspora Mobilization and the Transnationalization of Civil War.” In Transnational Dynamics of Civil War, edited by Checkel

, 63–88. Cambridge, UK: Cambridge University Press.

Adamson

Demetriou

. 2007. “Remapping the Boundaries of ‘State’ and ‘National’ Identity.” European Journal of International Relations 13(4): 489–526. https://doi.org/10.1177/1354066107083145

Banting

Kymlicka

, eds. 2006. Multiculturalism and the Welfare State. Oxford: Oxford University Press.

Beauchemin

2014. “A Manifesto for Quantitative Multi-sited Approaches to International Migration.” International Migration Review 48(4): 921–38. https://doi.org/10.1111/imre.12157

Blanchard, P. 2007. “‘Nouvelle’ UDC: nouveaux électeurs? Evolution de 1995 à 2003 [Does the ‘New’ Swiss People's Party Rely on New Voters? Evolution from 1995 to 2003].” In L'Union démocratique de centre. Un parti, son action, ses soutiens, edited by Gottraux P., C. Péchu and O. Mazzoleni, 155–180. Lausanne, Switzerland: Antipodes.

Benzecri

J. P.

1992. Correspondence Analysis Handbook. New York: Dekker.

Bourdieu

1984. Distinction: A Social Critique of the Judgment of Taste. English Translation. London: Routledge.

Bourdieu

Passeron

J. C.

. 1990. Reproduction in Education, Society and Culture. London: Sage.

Brinkerhoff

, ed. 2008. Diasporas and Development. Boulder, CO: Lynne Rienner.

10.

Brinkerhoff

2011. “Diasporas and Conflict Societies.” Conflict, Security, and Development 11(2): 115–43. https://doi.org/10.1080/14678802.2011.572453

11.

Brinkerhoff

2016. Institutional Reform and Diaspora Entrepreneurs. Oxford: Oxford University Press.

12.

Bühlmann

David

Mach

. 2012. “Political and Economic Elites in Switzerland.” European Societies 14(5): 727–54. https://doi.org/10.1080/14616696.2012.709531

13.

Byman

Chalk

Hoffman

Rosenau

Brennan

. 2001. Trends in Outside Support for Insurgent Movements. Santa Monica, CA: Rand.

14.

Campbell

J. L.

Quincy

Osserman

Pedersen

O. K.

. 2013. “Coding In-depth Semistructured Interviews.” Sociological Methods & Research 42(3): 294–320. https://doi.org/10.1177/0049124113500475

15.

Carletta

Isard

Isart

Kowtko

J. C.

Doherty-Sneddon

Anderson

A. H.

. 1997. “The Reliability of a Dialogue Structure Coding Scheme.” Computational Linguistics 23(1): 13–31. http://hdl.handle.net/1842/1227

16.

Cohen

1997. Global Diasporas. London: UCL Press.

17.

Collier

Hoeffler

. 2000. Greed and Grievance in Civil War. Policy paper 2355. Washington, DC: World Bank.

18.

Collier

Hoeffler

Söderbom

. 2008. “Post-conflict Risks.” Journal of Peace Research 45(4): 461–78. https://doi.org/10.1177/0022343308091356

19.

DAMR. 2021. Mixing Methods in Migration Research Conference. https://www.damr.eu/index.php/2-ongecategoriseerd/50-looking-back-mixing-methods-in-migration-research-conference.

20.

Delano

Gamlen

. 2014. “Comparing and Theorizing State–Diaspora Relations.” Political Geography 41: 43–53. https://doi.org/10.1016/j.polgeo.2014.05.005

21.

Demmers

2007. “New Wars and Diasporas.” Journal of Peace, Conflict & Development 11: 1–26.

22.

DeWind

Segura

, eds. 2014. Diaspora Lobbies and the US Government. New York: NYU Press.

23.

Fauser

2018. “Mixeed Methods and Multi-sited Migration Research.” Journal of Multi-methods Research 12(4): 394–412. https://doi.org/10.1177/1558689817702752

24.

Greenacre

Blasius

. 2006. Multiple Correspondence Analysis and Related Methods. New York: Chapman and Hall.

25.

Haney

Vanderbush

. 1999. “The Role of Ethnic Interest Groups in U.S. Foreign Policy.” International Studies Quarterly 43(2): 341–61. https://doi.org/10.1111/0020-8833.00123

26.

Hjellbrekke

Le Roux

Korsnes

Lebaron

Rosenlund

Rouanet

. 2007. “The Norwegian Field of Power Anno 2000.” European Societies 9(2): 245–73. https://doi.org/10.1080/14616690601002749

27.

Horvath

Latcheva

. 2019. “Mixing Methods in the Age of Migration Politics.” Journal of Mixed Methods Research 13(2): 127–31. https://doi.org/10.1177/155868981882244

28.

Howard

2009. The Politics of Citizenship in Europe. Cambridge, UK: Cambridge University Press.

29.

Ireland

1994. The Policy Challenge of Ethnic Diversity. Cambridge, MA: Harvard University Press.

30.

Kapur

2005. “Remittances: The New Development Mantra?” G-24 Discussion Paper Series, United Nations.

31.

Koinova

2009. “Diasporas and Democratization in the Postcommunist World.” Communist and Postcommunist Studies 42(1): 41–64.

32.

Koinova

2011. “Can Conflict-Generated Diasporas be Moderate Actors during Episodes of Contested Sovereignty? Lebanese and Albanian Diasporas Compared.” Review of International Studies 37(1): 437–462.

33.

Koinova

2017. “Beyond Statist Paradigms: Sociospatial Positionality and Diaspora Mobilization in International Relations.” International Studies Review 19 (4): 597–621.

34.

Koinova

2018a. “Critical Junctures and Transformative Events in Diaspora Mobilization for Kosovo and Palestinian Statehood.” Journal of Ethnic and Migration Studies 44(8): 1289–1308.

35.

Koinova

2018b. “Diaspora Mobilizations for Conflict and Postconflict Reconstruction: Contextual and Comparative Dimensions.” Journal of Ethnic and Migration Studies 44(8): 1251–1259.

36.

Koinova

Tsourapas

. 2018. “How Do Countries of Origin Engage Migrants and Diasporas? Multiple Actors and Comparative Perspectives.” International Political Science Review 39(3): 311–321.

37.

Koinova

Karabevogic

2019. “Causal Mechanisms in Diaspora Mobilizations for Transitional Justice.” Ethnic and Racial Studies 42 (11):1809–1829.

38.

Koinova

2021. Diaspora Entrepreneurs and Contested States. Oxford: Oxford University Press.

39.

Koopmans

Statham

. 2001. “How National Citizenship Shapes Transnationalism.” Revue européenne des migrations internationales 17(2): 63–100.

40.

Krasner

1999. Sovereignty: Organized Hypocrisy. Princeton, NJ: Princeton University Press.

41.

Kupchik

Highberger

Bear

. 2022. “Identifying the Helpfulness of School Climate.” Psychology in the Schools 59(8): 1538–55. https://doi.org/10.1002/pits.22692

42.

Kuropas

1991. The Ukrainian Americans. Toronto: University of Toronto Press.

43.

Leblang

2010. “Familiarity Breeds Investment.” American Political Science Review 104(3): 584–600. https://doi.org/10.1017/S0003055410000201

44.

Le Roux

Rouanet

. 2004. Geometric Data Analysis: From Correspondence Analysis to Structured Data Analysis. Dordrecht: Kluwer Academic.

45.

Lyons

2006. “Diasporas and Homeland Conflict.” In Territoriality and Conflict in an Era of Globalization, edited by Kahler

Walter

, 111–32. Cambridge, UK: Cambridge University Press.

46.

Lyons

Mandaville

. 2010. “Think Locally, Act Globally.” International Political Sociology 4(2): 124–41. https://doi.org/10.1111/j.1749-5687.2010.00096.x

47.

Marinova

2017. Ask What You Can Do for Your (New) Country. New York: Oxford University Press.

48.

Mearsheimer

Walt

. 2007. The Israeli Lobby and US Foreign Policy. New York: Farrar, Strauss.

49.

Orjuela

2008. “Distant Warriors, Distant Peace-workers.” Global Networks 8(4): 436–52. DOI:10.1111/j.1471-0374.2008.00233.x

50.

Østergaard-Nielsen

2001. “Transnational Political Practices and the Receiving State.” Global Networks 1(3): 261–82. https://doi.org/10.1111/j.1747-7379.2003.tb00157.

51.

Portes, A. and M. Zhou. 1993. “The New Second Generation: Segmented Assimilation and Its Variants.” Annals of the American Academy of Political and Social Science 530: 74–96.

52.

Ragazzi

2009. “Governing Diasporas.” International Political Sociology 3(4): 378–97. https://doi.org/10.1111/j.1749-5687.2009.00082.x

53.

Rubenzer

2008. “Ethnic Minority Interest Group Attributes and US Foreign Policy Influence.” Foreign Policy Analysis 4(2): 169–85. https://doi.org/10.1111/j.1743-8594.2007.00063.x

54.

Saldana

2013. The Coding Manual for Qualitative Researchers. London: Sage.

55.

Shain

Barth

. 2003. “Diasporas in International Relations Theory.” International Organization 57(2): 449–79. https://doi.org/10.1017/S0020818303573015

56.

Sheffer

2003. Diaspora Politics. Cambridge, UK: Cambridge University Press.

57.

Sidel

2003. Diaspora Philanthropy to India. Cambridge, MA: Harvard University.

58.

Smart

Hsu

J. Y.

. 2004. “The Chinese Diaspora, Foreign Investment and Economic Development in China.” Review of International Affairs 3(4): 544–66. https://doi.org/10.1080/1475355042000241511

59.

Smith

Stares

. 2007. Diasporas in Conflict. Tokyo: UN University Press.

60.

Sökefeld

2006. “Mobilizing in Transnational Space.” Global Networks 6(3): 265–84. https://doi.org/10.1111/j.1471-0374.2006.00144.x

61.

Waterbury

2010. Between State and Nation. New York: Palgrave Macmillan.

62.

Wimmer, A and N. Glick-Schiller. 2003. “Methodological Nationalism, The Social Sciences and The Study of Migration.” International Migration Review 37(3): 576–610.

63.

World Bank. 2017. Personal remittances, received (% of GDP). https://data.worldbank.org/indicator/bx.trf.pwkr.dt.gd.zs

64.

Zapata-Barrero

Rezaei

. 2020. “Diaspora Governance and Transnational Entrepreneurship: The Rise of an Emerging Social Global Pattern in Migration Studies.” Journal of Ethnic and Migration Studies 46(10): 1959–1973.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.21 MB

0.16 MB