CHARACTERISTICS OF CLANS AND CLAN-BASED SOCIETIES

Clan identity is only one of various identities that individuals may privilege in a given situation (Barfield Reference Barfield2010, 21; Coburn Reference Coburn2011, 2–4). As Beck explains, “Tribes were not static entities, however, but were historically and situationally dynamic” (Beck Reference Beck, Khoury and Kostiner1990, 190). Rather than being “conceptually stable over time,” these affiliations are instead “on-going objects of affiliation” that are “rearticulated” encounter by encounter (Miller Reference Miller2007, 189). Although we use the term “clan” to explicate our thesis and explain our data, we recognize that we are reducing a great deal of diversity and nuance into a more simplified concept—an analytic—for the purposes of operationalization and empirical testing.

Native or local actors may conceptualize clan identity in a variety of ways, but generally focus on the importance of group descent from a common ancestor, whether putative or actual (Asfura-Heim Reference Asfura-Heim and Isser2011, 242–46; Hart Reference Hart and Sweet1970, 12; Lindholm Reference Lindholm1982, 70–74). However, scholars recognize that common descent may well be fictive instead of blood related (Barfield Reference Barfield2010, 18–22; Eickelman Reference Eickelman2002, 116–17), and “it is common to find subgroups that have become effective kin by ritual rather than by blood” (Gellner Reference Gellner, Khoury and Kostiner1990, 110).Footnote ³ Assertion of common descent serves to justify loyalty to the group, which in turn guarantees security and well-being for individuals who “make strategic decisions on how to define themselves” (Barfield Reference Barfield2010, 21).

The vast majority of lineage-based groups trace descent through the patriline, practice patrilocality,Footnote ⁴ and inherit land and property through the patriline (Barfield Reference Barfield2010, 22; Eickelman Reference Eickelman2002, 75; Lancaster Reference Lancaster1997, 36–42).Footnote ⁵ Even in the rare matrilineal societies, power, land, and resources are still held by male kin, in the form of brothers from a particular mother. Patrilineality permits groups of male relatives bound by blood ties to become politically powerful and band together in allegiances when conflict arises, and hence scholars have noted the deep importance attached by the clan to biological replication (Barfield Reference Barfield2010, 20). Charrad observes, “The socially meaningful ties unifying the network thus bind men together and bypass women” (Reference Barfield2001, 53). The early forms of private property were held not by individuals but by patrilines: Kinship relations founded on agnaticFootnote ⁶ lineage allow both property and labor, including reproductive labor, to remain within the clan under male control (UNDP 2005, 164).

Clans offer significant benefits to members. Although the prevailing ideology of clans is one of kinship morality, in clans that ideology is tangibly married to “shared interest, advantage, and service” (Tapper Reference Tapper, Khoury and Kostiner1990, 58) or, as Kostiner puts it, it is both “blood ties and booty” (or, now, subsidies) that keep clans both an affective and a rational investment (Reference Tapper, Khoury and Kostiner1990, 230). As Collins explains (Reference Collins2006, 332 and 335),

Inevitably, then “clans emerged as distinctly political, by becoming ensnared with questions of distribution and exchange” (Schatz Reference Schatz2004, 97). Barfield even suggests that “[t]ribes and states . . . did not function as two separate, opposing systems. They represented alternative polities, each creating and solving political problems for the other” (Reference Barfield, Khoury and Kostiner1990, 191). There may be a cyclical aspect to the relationship between tribe and state: As Khoury and Kostiner explain, “from the perspective of the state, modernization and rapid social change were often counterproductive in that tribes were given a renewed role through their participation in the conflicts these changes inevitably produced” (Reference Barfield, Khoury and Kostiner1990, 16). Patricia Crone (Reference Crone and Hall1986) explains this is so because “tribes and states are not sequential stages but alternative answers to the problem of security” (emphasis added, quoted in Dresch, Reference Dresch, Khoury and Kostiner1990, 281). Rather than leave the individual defenseless against the power of the state, clans are “authority structures capable of countering other authority structures,” (Beck Reference Beck, Khoury and Kostiner1990, 192) available to many, even most, within the society simply by fact of birth.

When clans are strong, law focuses on the larger family or clan group as the principal unit (status societies); in contrast, contract societies recognize the importance of individual as the fundamental agent under the law (Maine Reference Maine1986, 163–65; Weiner Reference Weiner2013, 10–13). As societies modernize, they generally transform from the “status” to the “contract” type, typically freeing individuals from subordination to their extended families and granting them both rights and opportunities to develop on their own.

In status societies, such individual development is sharply curtailed. Each member of the clan is born into a certain status within the kin group, and to preserve and augment that status, individuals generally conform to goals determined by patrilineal clan leaders. This hierarchy reaches down to the household level; for example, the Romans saw each of these agnatic lineages as an “imperium in imperio” under the Patria Potestas, establishing the paterfamilias as emperor within the domestic realm (Maine Reference Maine1986, 144). This logic helps explain why clan influence is never associated with democratization (Collins Reference Collins2004, Reference Collins2006).

The delicate task of maintaining equilibrium within and between clans usually leads to the development of communal law, called here “clan law.” The necessity of maintaining such equilibrium is paramount because the spiral of clan-against-clan warfare can quickly become annihilative, or as Caton puts it, “blood descent leads[s] to bloody dissent” (Reference Caton, Khoury and Kostiner1990, 91). Clan leaders thus attempt to resolve problems quickly and decisively through established precedent in clan law regulating the resolution of murders, feuds, and so on, before they can escalate uncontrollably. The key to this “harmony” relies on an intangible—an honorable reputation—that serves to strengthen the group and its individual members. Good reputation is mandatory for leaders of the clan, because the respect paid them by clan members makes possible the enforcement of their judgments. For men, honor includes the vital characteristics of dignity, generosity, virility, courage, military prowess, hospitality, honesty, and the ability to protect not only one's family but also the vulnerable. Jacobson notes, “This notion of honor cuts across different kinds of religions, but it takes a particular form in tribal patriarchies—one that is martial, male and based on face-saving” (Reference Jacobson2013, 140).

Although clans serve to bind individuals together, it is important to realize that clans may use this unity to pursue conflict with other clans, as expressed by the Bedouin epigraph to this chapter (Hart Reference Hart and Sweet1970, 71–73; Lindholm Reference Lindholm1982, 66ff). Thus, procedures designed to maintain harmony among clans and not just within clans are also essential to clan governance societies, and their use often results in a long-lived shaky truce that cannot really be called peace. As Fukuyama notes, “In tribal societies, justice between individuals is a bit like contemporary international relations, based on the self-help of rival groups in a world where there is no higher third-party enforcer of rules” (Reference Fukuyama2011, 254). Although peaceful coexistence and the pursuit of well-being are nominally the goals, the means by which clans pursue these goals result in a system of “disequilibrium in equilibrium” (Hart Reference Hart and Sweet1970, 74). As Jacobson (Reference Jacobson2013, 60) explains,

Clans are capable of using their influence in a strategic fashion to create temporary peace and stability in place of continual violence (Caton Reference Caton2010; Coburn Reference Coburn2011, 5). But despite their best efforts, certain behavioral predispositions of clans predictably and chronically undermine the stability and security of their societies.

BEHAVIORAL PREDISPOSITIONS OF CLAN-BASED GOVERNANCE AND THEIR EFFECTS ON STATE STABILITY AND SECURITY: A THEORETICAL FRAMEWORK

State security and state stability are conceptually intertwined. State/national security can be defined as “a capacity to control those domestic and foreign conditions that the public opinion of a given community believes necessary to enjoy its own self-determination or autonomy, prosperity and wellbeing” (Maier Reference Maier1990, as quoted in Romm Reference Romm1993, 5), but that “capacity” needs to be further defined.

In its working papers, the Princeton Project on National Security suggests ten key functions of the state that gauge its capacity to ensure state security: a legitimate monopoly on the means of violence, administrative control, sound management of public finances, investment in human capital, the creation of citizenship right and duties, the provision of infrastructure, the regulation of the market, management of state assets, effective management of international relations, and maintenance of the rule of law (Ghani, Carnahan, and Lockhart Reference Ghani, Carnahan and Lockhart2006). Lockhart suggests that “a state is fragile . . . where one or more of these functions have really eroded or are not being performed properly” (IMF Survey 2013). Pertinent to our discussion of clan-based polities, Buzan comments, “Note that national security in this sense presupposes a strong state (where government and society enjoy a high degree of consensual integration). The concept of national security is difficult to apply to weak states (low integration between government and society and high levels of coercion) because feuding parts will make their own security claims against each other” (Buzan Reference Buzan2007).

Stable states, then, are states that have moved beyond insecurity by achieving capacity in the provision of state functions, as listed earlier, and that are not significantly regressing on these measures. A stable state, accordingly, is relatively secure, so that when we speak of state stability, inferences about state security can be made as well. The World Bank uses this linkage, operationalizing several Worldwide Governance Indicators that capture state capacity and functioning, such as government effectiveness, rule of law, and control of corruption.

In what ways, then, might greater clan influence in a society affect state security and stability? Those researching patrilineal clan-based societies describe behavioral dispositions that appear remarkably similar across societies, comprising what might be considered a “syndrome” of clan governance sequelae apparent even in societies with established state structures and institutions. Lapidus observes that “we are dealing with a system involving two types of political and cultural entities [on] the same territory, competing for power and legitimacy” (Reference Lapidus, Khoury and Kostiner1990, 28). The tension between the goals of the clans and the state lead many to conclude that “tribes have proved an unstable basis on which to build the future of [a] country” (Tapper Reference Tapper, Khoury and Kostiner1990, 52).

The extent to which clan networks degrade state stability is dependent on the degree to which clans have “captured” state institutions and capabilities. Such capture becomes easier under certain conditions, such as endemic economic shortages. Other capture mechanisms include clan-based voting, kin-based patronage, the stripping of state assets (such as oil) by clans in power, and what Collins (Reference Collins2004) calls “crowding out” participation of nonclan members in formal institutions such as political parties, the media, or unions by clan mobilization:

Underlying these macro-level observations linking state instability and clan influence are several micro-level processes at work. First, such societies are often ill prepared to pursue meritocracy. Individuals are provided opportunities simply because of their place in the clan hierarchy; “family, relatives, and extended clan fill the power ministries” (Collins Reference Collins2004, 333). The placement of kin (in some cases, female kin) in powerful government positions generally solidifies the power of the clan. Unless competition within and between clans is limited by such distribution of powerful positions, violence may well ensue. Clan interests, and the need for some measure of clan balancing to keep the peace, drive political appointments and policy decisions, rendering the political process irrational and incapable of pursuing national interests should they diverge from clan interests (Schatz Reference Schatz2004).

Our second proposition is that this type of nepotism tends to breed a culture of endemic corruption. Weiner suggests, “What we tend loosely to call corruption is often the distribution of favors along clan lines” (Reference Weiner2013, 81). The linkage between corruption and state instability is widely acknowledged:

In clan-based societies, government positions are generally sought and used to obtain additional assets for the clan and its members, and not to build up the broader society or the institutions of the state. The resources of the government are often viewed as potential booty, which clans feel should be preemptively stripped lest they fall to the control of other clans. For example, in Kenya, there is even a colloquialism for this asset stripping by the clan in power—“eating.” As a new clan comes to power, members crow, “It's our turn to eat!” (Wrong Reference Wrong2010). But it is important to understand that clan-based corruption is a horse of a different color from run-of-the-mill corruption. As we have noted, “blood and booty” hold clans together: Booty, or asset stripping, is the only possible foundation of the economic system where clans are powerful.

Douglass North and his colleagues describe this system as one in which “the political system [is used] to create economic rents; the rents order social relations, control violence, and establish social cooperation” (North, Wallis, and Weingast Reference North, Wallis and Weingast2009, xii). Rather than being perceived as an impediment to governance, corruption is often seen as the means of governance and the primary leverage with which to control violence. Economic reform is typically a nonstarter in these states because it is largely through rent creation and allocation that “elites create credible incentives to cooperate rather than fight among themselves . . . rents then secure political order” (18). Gellner (Reference Gellner, Khoury and Kostiner1990, 116) elaborates,

Modern examples of these “Games of Clan Thrones” abound. For example, Chinese personal and familial relationships have aided lineage groups in gaining access to business opportunities (Weiner Reference Weiner2013, 81). In response to numerous scandals, President Xi Jinping launched a major anticorruption campaign on taking office, though it is unclear whether even the power of a one-party state can offset that of powerful patrilines (Anderlini Reference Anderlini2013). In Saudi Arabia, “The Saudi polity tributarises other clan groups, no longer nomadic, and ties them . . . to the redistribution of Saudi wealth; for plunder is substituted by subsidy and the privilege of citizenship, such as the legal sponsorship of foreign business is akin in many ways to the exaction of protection money. Thus tribalism becomes ascendant” (Aziz al-Azmeh, quoted in Kostiner Reference Kostiner, Khoury and Kostiner1990, 245).

Third, the culture of relational corruption acts to undermine any sense of common ties through citizenship. Members of the clan may not see themselves primarily as the citizens of their nation-state, but rather as clan members. As such, the notion of common sacrifice for the good of the country is attenuated because kin-based patronage networks are the dominant resource allocation mechanisms in society. Note what Weiner says about Indian jatis:

Fourth, clan members generally have an instrumental view of formal government: In cases where the formal government supports the clan's interests—which may be ensured by placing its own members in governmental positions—the clan can strengthen the central government. However, where the formal government opposes clan interests, the clan may seek to undermine it. Seen in this light, many authoritarian regimes are actually not authoritarian at all: The “supreme leader” may lack the real power to govern society and be highly dependent on clan acquiescence, as we see in several post-Soviet Central Asian nations (Collins Reference Collins2006).

Fifth, clan governance lessens a sense of personal accountability under state law, which is viewed as less important than clan law, thus undermining the rule of law. As North and his colleagues put it, “The law cannot enforce individual rights if . . . every relationship between two individuals depends uniquely on their identity within the [group]” (Reference North, Wallis and Weingast2009, 63). Such a system may openly sanction impunity for the guilty and punishment for the innocent, undermining the very concept of justice being based on accountability for individual actions before the law. Such states “may have elections, but they do not have extensive systems of rights of rule of law for most citizens” (139).

Sixth, when clans engage in conflict, swift and dangerous escalation often ensues. In fact, this escalation is one of the most striking behavioral predispositions of societies governed by clans. As Weiner describes it, “Without a larger force to intervene to bring the feud to a close, the exchange of reciprocal violence can be ratcheted up until a community implodes in bloodshed. This outcome is especially likely when the ancient legal tool of feud is conducted with weaponry built for warfare by modern nation states” (Reference Weiner2013, 181). He calls this escalation of conflict the “Achilles heel of clan-based government” (37). North and colleagues note that in these societies, “the threat of violence permeates society . . . everyone must be prepared to be violent, and the military resources of the community are widely dispersed throughout the population” (Reference North, Wallis and Weingast2009, 173). As one shopkeeper in western Afghanistan commented, “You must have a gun to stay alive in Ghor. It is more important to have a gun at home than food” (Ali Reference Ali2013).

To summarize, then, clan-based governance tends to produce a dysfunctional, unstable, insecure, and frequently bloody society. Because clan-based governance can exist in failed states, weak states, democratic states, and authoritarian states—virtually any state—the ability to identify the underlying clan structure of a society would seem an important avenue of inquiry for scholars interested in state security and stability.

SEEING THE CONCEALED: LOOK TO THE WOMEN

Clan-based governance underlying state structures may be obscured from view, often purposefully. The state may formally prohibit clan identification, or as in the former USSR, it may create new “ethnicities” that cross clan lines in order to dilute their power. Clan members themselves may conceal clan identification to mitigate resentment. Schatz asserts that “the most important inherent quality of clan divisions [is] their concealability [though] we should not conflate the lack of a public presence with the lack of importance” (Reference Schatz2004, xxii, 48) and suggests that clan politics is “like watching bulldogs fight under a carpet” (95). Clans are “among those political phenomena that are hard to quantify. They are concealable, often invisible, private, and hotly contested. They are not election results or interest rates; they lend themselves poorly to quantification” (109).

To approach an empirical analysis of clan influence on state security and stability, it is necessary to find some way to gauge the relative power of clans in society. In our review of the literature, we found one possible lever to reveal the degree of influence: Clans can only reproduce themselves in a relatively exclusionary manner. Consider the following assertions by clan scholars:

• • “The most powerful metaphor of exclusion was the clan” (Schatz Reference Schatz2004, 128).

• • “[Tribes] invoked genealogies to define and maintain their exclusivity, and their lineages were often highly endogamous” (Beck Reference Beck, Khoury and Kostiner1990, 194).

• • “The clan is the basis of a strong, but narrow and exclusivist social organization” (Collins Reference Collins2004, 232).

We argue that clan exclusivity can only be reproduced by restricting the agency of clan women in marriage. Arguably the most vulnerable family members in status societies are the women whose role is to reproduce the patriline, because the subordination of female interests, reproductive or otherwise, is how patrilineal clans are formed in the first place. Schatz is correct when he states, “If we can identify the mechanisms of identity reproduction, we gain exceptional purchase on both identity persistence and identity construction. . . . Whether clan divisions persist or not hinges on identifiable mechanisms of identity reproduction” (Reference Schatz2004, 13). Female subordination, specifically in marriage, plays that role for agnatic clans.

Patrilineal authority must define the parameters of marriage choice for women in order to strengthen the male kinship bond, or else the clan as a demarcated group will disappear. Goody explains that the supremacy of the agnatic lineage has been maintained because “the conjugal pair was incorporated into the larger unit, male authority was supported by the kin group and a woman's independent role in the household was minimal” (Reference Goody1983, 30). As Weiner notes, “The anti-individualism of the rule of the clan burdens each and every member of a clan society, but most of all it burdens women. The fate of women lays bare the basic values of the rule of the clan, and as outsiders, citizens of liberal states often find their own values clarified when they confront the lives clans afford their female members” (Reference Weiner2013, 64).

It is critical to understand that patrilineal clans cannot exist without the subordination of female interests to the goals of the male members of the clan. According to the Arab Human Development Report, “Patriarchal authority enshrines the hereditary nature of the economic structure and positions of leadership and sovereignty at the expense of women” (UNDP 2005, 167). As Fukuyama notes, “In agnatic societies, women achieve legal personhood only by virtue of their marriage to and mothering of a male in the lineage;” that is, women only “exist” in these societies as they create the patriline because patrilines cannot exist without women creating them (Reference Fukuyama2011, 233). He goes on to note, “We should not underestimate the importance of sex and access to women as a driver of political organization, particularly in segmentary or tribal societies that routinely use women as a medium of exchange” (76).

The fierceness and the sensitivity with which the subordinate status of women in clan-based societies is guarded by the men of these societies testify to Fukuyama's proposition. Charrad observes, “Women represent a potential source of rupture in the web uniting the men of the patrilineage” (Reference Charrad2001, 55). Men—and not women—must therefore control the clan's assets, whether these are children or land or cattle, or the power of the clan will dissipate. For example, Fukuyama observes, “While widows and unmarried daughters may have certain inheritance rights, they are usually required to keep the lineage's property within the agnatic line” (Reference Fukuyama2011, 233). The marriage of a daughter thus becomes a type of disinheritance: Early legal systems, such as Roman law, included the maxim, “Mulier est finis familiae . . . none of the descendants of a female are included in the . . . family relationship” (Maine Reference Maine1986, 143).

This understanding that clans can only be formed through the subordination of female interests, especially in marriage, suggests that mechanisms of control over female interests in sex, marriage, and reproduction may in fact be the most visible indicators of clan governance; thus, they should give us some purchase on the salience and influence of clans within the society. We therefore argue that the most effective way of seeing clan-based governance structures is by examining the usually very visible situation of women within society and particularly in marriage; as Jacobson puts it, “Gender is the hinge . . . the role of gender in tribes and tribalism is foundational” (Reference Jacobson2013, 115–16).

More general indicators of gender inequality within a society are therefore not sufficient for our purposes. The wage gap between men and women and the representation of women in parliament, for example, do not tap into the reproduction of clan exclusivity. But the prevalence of child marriage for women, the prevalence of patrilocal marriage, and other indicators of such control can inform us whether the reproduction of exclusivity is taking place. Charrad notes, “The central questions concern procedures for marriage, rights and obligations of each spouse, polygamy, conditions for divorce, custody of children, and inheritance” (Reference Charrad2001, 5).

We therefore propose that eight mechanisms of this particular type of female subordination serve as indicators of the degree to which clans are salient to governance in a given society: (1) patrilocality, (2) inequitable family law favoring males, (3) prevalence of polygyny, (4) early/child marriage for females, (5) marriage of women within the patriline (cousin marriage), (6) sanction for honor-based violence against women, (7) high levels of violence against women, and (8) lack of property rights for women in practice. Each is discussed below.

VARIABLES AND METHODS

Our overarching hypothesis is that nation-states scoring highly on these eight measures of subordination of women in marriage will exhibit greater political instability and violence than nations that do not, because this particular syndrome of subordination marks the heightened reproduction of clan exclusivity and thus the relatively greater importance of clan-based networks in state governance.

Patrilocality of marriage. Patrilocality makes the formation of agnatic clans a fairly straightforward task and thus is universally favored by patrilineal groups. In this system, most males in a particular area are kin, which forms a natural foundation for male alliances in patrilineal clans. Such societies, in which land and resources belong exclusively to men, effect the complete economic dependence of females, resulting in a profoundly subordinate status for women.

Inequitable family law that discriminates against females. Patrilineal clans make family law in the image of their own reproductive interests, as shown by inequitable family laws that favor men across time and space (Smuts Reference Smuts1992). The purpose of such inequality between male rights and female rights in marriage is clearly the subordination of female interests, especially reproductive interests, to male reproductive interests. Thus we see that adultery is a greater crime for women than for men, divorce may be easy for men to obtain but difficult for women, child custody may default to the father's kin, and so forth. This discrimination is most clearly seen in clan law, but can also be reproduced in formal state law where significant clan governance is found. It is important to examine practice as well, given that states with equitable family laws on the books might not enforce these laws because of clan-based tradition or practices (Hudson, Bowen, and Nielsen Reference Hudson, Bowen and Nielsen2011).

Polygyny. Prevalent polygyny goes hand in hand with a patrilineal clan-based social structure, making polygyny an excellent indicator of the presence of agnatic clans. Women in societies with prevalent polygyny are viewed as assets accumulated for economic benefit, for the status/honor such accumulation affords within the culture, or both. Furthermore, clan leaders often use polygyny to create inter-clan alliances or at least mitigate inter-clan rivalry; for example, Barfield notes, “The emphasis placed on reciprocal marriage ties created patterns of alliances that crosscut the seemingly rigid set of patrilineal relationships within a conical clan. For this reason, polygynous marriages by tribal rulers were common” (Reference Barfield, Khoury and Kostiner1990, 163). A broad literature already exists linking polygyny to societal governance and state instability (see, for example, Alexander Reference Alexander1979; Betzig Reference Betzig1986; Hudson, Bowen, and Nielsen Reference Hudson, Bowen and Nielsen2011; McDermott and Cowden, forthcoming; Wright Reference Wright1995).

Early marriage for women (child brides). Child marriage is by no means rare: In the early twenty-first century, fifteen million girls, roughly one in nine, were married between the ages of ten and fourteen (Hervish and Feldman-Jacobs Reference Hervish and Feldman-Jacobs2011). Marrying off daughters at young ages is a distinguishing characteristic of traditional marriage patterns that serve the interests of patrilineal clans. Girls are seen as an economic burden on their natal family; the investment in their health and welfare can never be repaid in a patrilineal clan-based society (except through payment of a bride price upon marriage). Furthermore, once girls reach puberty, guarding their honor can become a burdensome responsibility, exposing the clan to social calamity. In addition, the older the girl, the more likely she is to challenge an arranged marriage in the patriline's interest. Thus, the earlier the marriage, the better for the agnatic clan for several reasons. Yet patterns of child marriage certainly do not serve the interests of the girls involved in terms of their physical health, opportunities for education, or level of safety within the household. As a UNICEF report summarizes, “Child marriage is a violation of human rights, compromising the development of girls and often resulting in early pregnancy and social isolation. Young married girls face onerous domestic burdens, constrained decision-making and reduced life choices” (2005).

Cousin marriage. Another marriage-related indicator revealing the strength of patrilineal clans is endogamous marriage. Where cousin marriage is prevalent, such marriages are virtually always arranged between cousins within the patriline, which ensures that all tangible resources, including money that would otherwise be used for dower as well as the woman herself and the children she might bear, stay within the patriline. Cousin marriage thus offsets any siphoning of resources or strength from the patriline consequential to the birth of girls who must be one day be married. Although leaders may practice polygyny for strategic reasons, lower ranking clan members favor cousin marriage to conserve resources within the patriline. As a result, one often sees polygyny and cousin marriage within the same clan-based society, even when endogamy taboos are present (Schatz, Reference Schatz2004). Additionally, rivalry between male cousins in clans is often intense, and thus “agnatic rivalry and a mechanism to contain it characterize the predominant marriage patterns between cousins” (Ahmed Reference Ahmed2013, 50).

Honor-based violence against women. In clan-based societies, as we have noted, group honor is the cultural foundation of clan governance (Weiner Reference Weiner2013, 34–36). Blood feuds may be fought to defend a clan's honor. In some locales such as the Swat Valley, individuals may see honor killing as an essential element of the clan's system of kin solidarity and governance (Lindholm Reference Lindholm1982, 126). The sexual conduct of women is a potent means of manipulating the honor of men, often eliciting honor-based violence.

Violence against women. Patrilines find that women—even women born to the patriline—often do not share their interests. Mothers may object to child marriages of their daughters. Sisters may object to a lack of inheritance. Wives may object to male dominance in household decision making, including decisions that bring a new wife into the family. In addition to inequities in family law that keep male interests superordinate—such as arranged child marriage, polygyny, and patrilocality—the patriline is also prepared to use physical force to enforce the dominance of its interests over those of women. Thus we would expect that clan-based societies would tolerate the highest levels of violence against women (even though all societies tolerate a significant level of such violence).

Women's property rights in practice. Keeping resources within the patriline requires both that inheritance and property rights favor men and that major economic resources such as land will remain solely within male hands. Thus women's property rights in practice (as opposed to formal law) are strongly indicative of whether clans play an important role in societal governance. As Fukuyama notes, “[T]he ability of women to own and bequeath property is an indicator of the deterioration of tribal organization and suggests that strict patrilineality [has] disappeared” (Reference Fukuyama2011, 233).

In sum, we claim that this eight-dimensional index of female subordination in marriage should reveal what is often concealed: the degree to which clans can reproduce their exclusivity and thus become a significant power in the larger society.

To investigate the linkage between clan power and state stability and security, we conducted a two-step process of empirical investigation. First, we created and validated a Clan Governance Index based on these eight gender-focused variables. We then compared our index to those of two other researchers who have attempted to capture this phenomenon without recourse to indicators addressing marriage subordination and violence against women. Second, we built a statistical model that relates measures of state stability and security to our Clan Governance Index and then examined the index's complementarity with other explanatory variables such as wealth, level of democracy, and civilizational identity. We analyzed five operationalizations of state instability and insecurity to probe the involved linkages.

Steps One and Two: Creation and Validation of Clan Governance Index

We constructed our Clan Governance Index to include indicators of the reproduction of clan exclusivity through the subordination of women in marriage. Because this approach is quite different from other conceptualizations of clan governance, we validated the index by comparing it to two other extant operationalizations: the Tribal Patriarchy Index and Weiner's (Reference Weiner2013) rating of the prevalence of clan rule.

Sociologist David Jacobson developed the Tribal Patriarchy Index using the following formula:

\begin{eqnarray*} && {\rm Jacobson's\, Tribal\, Patriarchy\, Index}\\ &&\quad = {\rm Corruption\, Perception\, Index} + {\rm 0}{\rm .5}\\ &&\quad \times {\rm (Ethno - } {\rm Linguistic\, Fractionalization\, Index)} \\ &&\quad + {\rm 0}{\rm .5} \times {\rm (Indigenous\, Population\, as\, a\, Percentage\, of } \\ && \times\, {\rm Total\, Population\, Data} + {\rm 2} \\ &&\quad\times\, {\rm (World\, Economic\, Forum's\, Gender\, Gap\, Data)} \\ &&\quad+\, {\rm Group\, Grievance\, Data} \end{eqnarray*}

Notice that corruption is a component of Jacobsen's index, whereas it is an outcome variable in our theoretical framework. Notice also the heavy weight given to the Global Gender Gap component, which seems in line with our own approach. However, the Global Gender Gap measure does not include marriage customs or violence against women. Instead, it examines women's participation in the labor force and the government, the educational attainment of women relative to men, sex ratio, and women's life expectancy. As mentioned previously, we believe these variables are epiphenomenal to the subordination of women's interests to men's interests in marriage and men's corresponding use of physical coercion of women. It is female subordination in marriage, specifically, that allows the clan to reproduce in exclusive fashion. We believe our index will correlate fairly strongly with Jacobson’s, but at the same time captures the most theoretically pertinent bases of clan strength.

Legal scholar Mark Weiner (Reference Weiner2013) rated 160 countries according to the prevalence of clan rule, concentrating on the salience of consanguineous relations; the penetration of private, familial interests within government functions; education levels and the extent of personal freedom for women; and tendencies toward intergroup violence, though political instability per se did not factor into the rankings. With regard to the concept of personal freedom for women, Weiner examines “the ability to own and devise property, the ability to enter the workforce and to have a professional career, the ease of divorce, the extent of allowed sexual freedom, the extent to which female family roles restrict their life opportunities, and the relative strength of socio-cultural pressures to conform to mainstream expectations regarding a woman's proper role” (Weiner, email to authors, February 22, 2014).

As with Jacobson's scaling, there are some similarities between the Clan Governance Index and Weiner's rankings, particularly in the areas of cousin marriage, property rights, and divorce. Nevertheless, like Jacobson, Weiner includes nongender-related dimensions in determining the degree of clan governance and excludes other important gendered variables, such as patrilocality. Because Weiner is not working from a dataset as much as from impressions regarding women's status, the rankings are thus not replicable. A high correlation among the three scales—our Clan Governance Index, Jacobson's Tribal Patriarchy Index, and Weiner's clan ratings—will constitute evidence of the predictive validity of the Clan Governance Index. Furthermore, if these scales are highly correlated, that also says something important about the relationship between female subordination in marriage and the degree of clan governance in society.

Step Three: Model Specification and Verification

We used multivariate linear modeling—more specifically a backward elimination procedure—to determine the best fitting model for our data using our five dependent variables as indicators of state stability and security. Model assumptions were checked, and if one or more of these assumptions were untenable, we eliminated problematic scores from the analysis, combined similar categories, or transformed the variable to address concerns with multi-collinearity, non-normality or variance-covariance heterogeneity. We then repeated the analyses using the transformed or reduced data. Significant results were probed using a post hoc investigation involving a set of univariate tests, one conducted for each dependent variable.

SUPPLEMENTARY MATERIALS

To view supplementary material for this article, please visit http://dx.doi.org/10.1017/10.1017/S0003055415000271