by Steven R. Hofmann

Y673 Spring 2002

Institutional Analysis and Development Framework
Professor Michael D. McGinnis
University of Indiana

Presented at the Institutional Analysis and Development
Mini-Conference on April 27th and 29th, 2002

Abstract: How can one understand the divergent paths followed by Somalia (still marked by interclan violence and an ineffective Transitional National Government) and Somaliland (which has seen practically no inter-clan violence over the past five years, and recently ratified a constitution via popular referendum) in the eleven years since the collapse of the Somali state'. Somaliland's success has been due, at least in part, to the decision of its leaders to utilize indigenous forms of self-governance after the collapse of the Somali state in early 1991. Somalia's difficulties, from this standpoint, can be at least partially attributed to the lack of meaningful attempts to do the same. The inability of Somalia's leaders to make use of indigenous institutions of self-governance is, in turn, a manifestation of the variance in the political and societal effects caused by the centralization of political authority experienced by Somalia and Somaliland during the colonial period, the civilian administrations of early post-colonialism (1960-69), and the military government of Siyad Barre (1969-1991). Actions taken by the international community, especially the United Nations, after the collapse of Somalia further aggravated the ability of southern Somalis to effectively utilize traditional institutions of self-governance in an effort to consolidate peace and reestablish effective and just institutions of governance at the level of the state.

In the wake of the January 1991 collapse of the Somali state, the leaders of the Somali National Movement (SNM), a politico-military organization headed by members of the Isaaq clan-family, declared that the 1960 Act of Union that had joined the British Somaliland Protectorate and the Italian-administered UN Trusteeship of Somalia had been revoked by the peoples of Somalia's northwest regions, thereby announcing the secession of `the Republic of Somaliland' from the Somali state. This new de facto state, the borders of which are coterminous with those of the former British Somaliland Protectorate1, is inhabited by 3 to 3.5 million people, the majority of whom are pastoral nomads2, and is characterized by two wet seasons (gu and dayr) and two dry seasons (hagaa and jiilaal). During these cycles of rain drought, the various nomadic clans move across the barren landscape to those areas where the conditions are most favorable for finding water and grazing land. Each nomadic clan remains within specific `zones of movement', and the movements of these clans are set to the rhythm of the seasons.

In the eleven years since Somaliland's declaration of independence, the area has been characterized by relative peace and stability3, especially when compared to southern Somalia, in which inter-clan battles for territory are still waged on a daily basis.

Somaliland has yet to be recognized as an independent state by the international community, but has been able to restore peace, establish a relatively effective central government, and ratify via popular referendum a new constitution based on democratic principles. How can these achievements be explained' Why has Somalia been unable to

In order to explain how the people of Somaliland have been able to solve the problems of governance which it faced shortly after its secession from Somalia, one must first understand the traditional social and political institutions of the pastoral nomads of the region, and how patterns of person-to-person relationships are shaped by these institutions. It will be argued below that Somaliland's success has been due to the decision of its leaders to utilize indigenous forms of self-governance after the collapse of the Somali state in early 1991. Somalia's difficulties, from this standpoint, can be at least partially attributed to the lack of meaningful attempts to do the same. The inability of Somalia's leaders to make use of indigenous institutions of self-governance is, in turn, a manifestation of the variance in the political and societal effects caused by the centralization of political authority experienced by Somalia and Somaliland during the colonial period, the civilian administrations of early post-colonialism (1960-69), and the military government of Siyad Barre (1969-1991). Actions taken by the international community, especially the United Nations, after the collapse of Somalia further aggravated the ability of southern Somalis to effectively utilize traditional institutions of self-governance in an effort to consolidate peace and reestablish effective and just institutions of governance at the level of the state.

The paper will proceed as follows: Part I will describe the Somali kinship system and the concept of xeer (social contract), which has traditionally patterned person-toperson relationships among the pastoral nomads of Somaliland and Somalia; Part II will examine the effects of political centralization in these two regions, from the era of colonialism through the Barre regime; Part III will focus on the widely divergent paths to conflict-resolution followed by Somalia and Somaliland during the 1990s; and the conclusion will summarize the effects of centralization and international intervention in Somalia and Somaliland, and propose suggestions for future research.

I. Traditional Institutions of Self-Governance

In 1961, I.M. Lewis, the foremost Western anthropologist of the Somali people, described the traditional political and social structure of the pastoral nomads living in the British Somaliland Protectorate as a "pastoral democracy". Much like the Nuer of southern Sudan, whose covenantal way of life and acephalous social structure are examined by Duany (1992), the social structure of the northern pastoral Somalis is characterized by little social stratification and no centralized government. Lewis states that: "Few societies can so conspicuously lack those judicial, administrative, and political procedures which lie at the heart of the western conception of government..Yet, although they.lack to a remarkable degree all the machinery of centralized government, they are not without government or political institutions (Lewis 1961, 1).

This egalitarian societal structure is based on kinship; one's patriarchal descent determines his political affiliation. Each individual's place in society and their precise connections with others in society are determined by his lineage. These patterns of interaction are contextual; one level of lineage may be more salient than others in varying situations. Agnation is viewed as a binding and absolute tie that cannot be severed; however, female links can be `cut' if necessary (ibid., 137).

Lineage is traced, at the highest level, to the clan-family. The Somali people are divided into six clan families, viz. Dir, Isaaq, Hawiye, Daarood, Digil, Rahanwiin. The latter two of these six lineage groups are primarily agriculturalists living between the Shebelle and Juba Rivers in southern Somalia, while the first four listed above are generally pastoral nomads. Somaliland is comprised primarily of members of the Isaaq, Dir, and Daarood clan-families, with the Isaaq being the numerically predominant lineage group. The Daarood, Hawiye, Digil, and Rahanwiin clan families inhabit southern Somalia.

Because clan-families are generally quite large and often geographically dispersed, the level of lineage one step below the clan-family, viz. the clan, marks highest lineage level able to act as a corporate political unit; the clan may or may not be lead by a clan-head. However, even if a clan-head (often called a `Sultan') exists, there is no centralized administration or government structure within the clan itself (ibid., 5). Below this level of lineage is the sub-clan, the main importance of which is in those cases where the clan also too large to act a political unit, when this is the case, the sub-clan takes on the functions generally ascribed to the clan.

The next level of agnatic lineage system is the primary lineage group. When telling others to what lineage he belongs, this is generally the level to which one refers (ibid., 6). At the base of the lineage system is the dia-paying group, which is the group within which one most often acts as a member, and the most stable of the lineage-based corporate political units. Ties created through marriage are utilized as a method to `bridge gaps' between alliances of dia-paying groups in which no patriarchal connection exists (ibid., 141).

It is also at this level of agnatic kinship where the concept of xeer most often defines the interaction of person-to-person relationships. Xeer is described by Lewis as a form of egalitarian social contract that explicitly formulates the obligations, rights, and duties of those parties that have entered into the contract, thereby regulating the relations between lineage groups (ibid., 161). Xeer regulates various forms of interaction between different clans as well as relations within the same clan; the regulation of access to water sources and grazing land is generally governed by xeer due to the fundamental importance of these common-pool resources to the lives of nomadic pastoralists, especially during the dry seasons. These contracts can be renegotiated or dissolved when the various exigencies of nomadic pastoral life call for such a change.

A dia-paying group bound by xeer is collectively responsible for those actions committed by its members that go against the obligations set forth in the contract. The term dia itself refers to blood compensation; if a member of one lineage group kills a member of another group with which xeer has been entered into, the group of the transgressor is collectively responsible for paying the aggrieved dia-paying group 100 camels (fifty camels if the victim is female) (ibid., 163).4

The specific stipulations of xeer are deliberated by the shir, an informal council of all adult men of the affected lineage groups at which each can discuss his opinions and concerns. The eight basic principles of institutional resource management articulated by Elinor Ostrom (1990) are found within the pastoral nomadic traditions of the Somalis, and are taken into account when xeer concerning access to common pool resources are being deliberated (Shivakumar 1998, 8). After the conditions of xeer are agreed upon by Xeer, then, although technically a social contract (which stipulates explicit terms and agreements between parties), can also be viewed as a covenantal relationship much like that described by Tocqueville (1990 [1835]) in his analysis of the origins and workings of American democracy. Explicit terms and agreements are obviously contained within specific xeer, but the concept of xeer essentially acts as agreement concerning the norms that will govern future conduct. Under such a covenantal relationship, "sovereignty conceived as the right to make laws reside[s] with the people in diverse communities of relationships" (Ostrom 1999b, 397).

However, this traditional, indigenous polycentric system of conflict-resolution and governance was increasingly undermined by the centralization policies of colonialists, the civilian administration established after independence (1960-69), and the military regime lead by Siyad Barre (1969-91). The next section will examine these policies and their effects on the kinship system and xeer.

II. Centralization and its Effects

Vincent Ostrom notes that If one assumes.[that] continuities [of social life and human organization] are desirable and ought to be maintained, then the integrity of the basic institutions within a society which structure and transmit them need to be recognized and maintained. If those institutions are seriously disrupted, the continuity of the society itself may be disrupted (V. Ostrom 1990, 234).

Somalia serves as an example of a society disrupted by policies of centralization, which served to weaken Somali society's indigenous institutions of self-governance that had regulated patterns of person-to-person interaction for centuries. However, while centralization had an impact on both Somaliland and southern Somalia, its effects were far greater in southern Somalia than in Somaliland, which has contributed to the various problems that currently plague southern Somalia.

The era of colonialism

During the colonial period, the country known today as Somalia was two separate entities: the British Somaliland Protectorate in the north and Italian Somaliland in the south.5 In both the British Somaliland Protectorate and Italian Somaliland, the stability of dia-paying groups was weakened by the appointment of chiefs for each clan by the British and Italian colonial authorities in an effort to ease the running of their administrations (Issa-Salwe 1996, 5). This practice, known as the Akils system in British Somaliland, vested, in a single person, the powers of assisting the administration in maintaining law and order within the clan, enforcing government orders and regulations, and bringing persons guilty of crimes within the clan to justice, which diminished the bonds of contractual solidarity within the dia-paying group (Lewis 1961, 201).

Furthermore, because these appointed chiefs were provided with a government stipend (and in Italian Somaliland, a group of armed men to police his clan), competition was created among clan elders to be appointed to this position. The system was used by both colonial powers as a means to damage the integrity of the clan and the office of the clan elders; in essence, it was an attempt to undermine the traditional source of leadership within the traditional system of self-governance (Issa-Salwe 1996, 5).

However, the Italians went much further in their efforts to weaken the kinship system and xeer. During the era of Italian fascism, land was appropriated from Somali owners, who were then forced to work the land essentially as sharecroppers, reducing Somalis to a source of cheap labor. Trade and commerce within Italian Somaliland were controlled by governmental monopolies; participation was denied to Somalis in any sector of the economy that was deemed essential to Italian interests. Many discriminatory laws antithetical to xeer were passed that further degraded the status of Somalis, who also were subject to "severe and arbitrary punishments for trivial offenses" by the Italian legal system. However, Lewis notes that the Italians, whatever their motives and tactics, were able to inculcate a "more modern attitude towards centralized government" in the southern Somali people, and did more to modernize the economic system of the south than did the British in Somaliland (Lewis 1980, 112-113). Because of the modernization efforts of the Italians, many Somalis began moving from the countryside into the cities during this time.

By the time the British Somaliland Protectorate and the Italian Somaliland6 gained their independence in 1960, the two entities were marked by widely divergent levels of centralization. The Italians had decided that the best way to prepare Somalia for independence was to create governing institutions based on the centralized Western political model. In 1950, a Territorial Council was created by Italian trusteeship administration, which was a central consultative body responsible for all government activities other than foreign policy and defense. Twenty-eight of the thirty-five seats on the Council were reserved for Somalis, and were chosen by a mixed election system; a direct vote was allowed to Somalis living in municipal districts, whereas pastoral nomads and those living in smaller villages chose their representative through the shir. This lead to a number of clan leaders joining more than one shir within the same clan, which was an attempt to misrepresent the overall size of their clans in order obtain more seats on the Council (Tripodi 1999, 80). Conversely, in the British Somaliland Protectorate, the only centralized body was the Protectorate Advisory Council, which meant only twice a year and did not have actual decision or policy-making capabilities. Due to the lack of any real political power vested in the Council, there was no incentive for clans to misrepresent their size in order to gain increased representation within the body.

The Civilian Government of Post-Colonial Somalia (1960-69)

The British Somaliland Protectorate became independent on 26 June 1960, and Italian Somaliland followed suit five days later, gaining its independence on 1 July 1960. This latter date also marked the merger of the two territories and the creation of a unified Somali Republic. In the north, nomadic pastoralists still made up almost ninety percent of the population, whereas less than half of the inhabitants of the south were involved in pastoralism at the time of independence. (ibid., 83).

It must be noted that the unification of the two territories was established, as Paolo Tripodi succinctly puts it, on a "misunderstanding" (1999, 107). Although the legislative bodies of both territories7 unanimously approved unification, no treaty was ever signed, and no agreement was ever reached as to the relative political powers of the two territories after unification and independence. The southern regions dominated the new government, controlling the posts of Commander of the Police Force and Commander of the National Army in the state military apparatus, and the posts of President, Prime Minister, Minister of Defense, Minister of the Interior, and Minister of

A number of political parties had proliferated during the previous decade as the two territories were preparing for independence; these parties were dominated by urban Somalis, many of whom had served the British and Italian administrations, and whose interests did not match those of the nomadic pastoralists whose social and political structure was still characterized by the importance of kinship and xeer. "Clanism" was viewed as divisive by these urban elites, most of whom referred to their "ex-clan" rather than their clan when asked about their patriarchal lineage by others, preferring to be considered simply Somali rather than a member of a specific lineage (Lewis 1980, 168).8

After independence and unification, these urban elite politicians gained control of a highly centralized government as a result of the masses' belief that elite interests matched their own (A. Samatar 1988, 48). Somalia's new constitution, based on the Italian political system, created a unicameral legislative body (the National Assembly) consisting of 123 directly-elected members; the National Assembly would elect a President (who would serve as head of state) every six years; the President was vested with the power to nominate the Prime Minister from the leading parliamentary party, to elect deputies every four years, and to appoint five justices to the new Supreme Court. At the regional level, eight administrative regions and thirty-six districts were created; regional governors and district administrators were appointed by the central Foreign Affairs in the central government. As the southern city of Mogadishu had already been selected as Somalia's capital, northern politicians believed they would be better represented in the government as a compromise, but this was obviously not to be the case.

During its nine years of civilian government, Somalia increasingly began to take on the features of a predatory state. Predatory states are characterized by the presence of "state functionaries [that] prey upon their own subjects, whose autonomous cultural infrastructures are threatened with destruction without their opportunities for selfgovernance being increased" (Ostrom 1999a, 176). The few achievements of the three civilian administrations in the area of social development primarily benefited urban areas; for example, a small number of schools were constructed primarily in southern cities, a hospital was built in Mogadishu, and a paved road was constructed linking Mogadishu with Afgoi (Ahmed I. Samatar 1988, 65). However, the rural population was essentially ignored by the central government; rather than attempting to create a dialogue with the nomadic pastoralists, the first civilian regime, led by Prime Minister Abdul-Rasheed C. Sharmaarke, only managed to anger them by increasing the taxing of the nomads.

Defense and civil administration were the largest items of government expenditure, while spending in the agricultural and especially the pastoral sector remained miniscule (ibid., 65).

Eventually, the bureaucratic elite came to realize that personal wealth could be accumulated quickly and easily in the upper levels of the state apparatus, which lead to an explosion of the number of candidates vying for a seat in the National Assembly during the March 1969 election campaign. More than 1,000 candidates from over sixty parties attempted to gain access to this lucrative body, some spending up to $30,000 (US) in a country with an annual budget of approximately $35 million (US). The Somali Youth League won 73 of the 123 seats (thanks in large part to modifications of electoral laws that favored the ruling SYL, and its raiding of the state treasury for its members' campaign funding); when the new National Assembly met for the first time, all of the non-SYL members (with the exception of former Prime Minister C. H. Hussein) "crossed the floor" and joined the SYL in order to recoup their campaign expenses by gaining access to government portfolios. By increasing the number cabinet positions from twelve to twenty-one and creating ten vice-ministerial positions, Prime Minister Muhammed Ibrahim Egal showed his appreciation to these new SYL members. Somalia had become, in essence, a one-party state.

These actions made it blatantly obvious to most observers that the government's "emphasis was on party politics and personal power rather than on mobilization for national development" (ibid., 71). However, because of the under-representation of northern politicians in the central government9, they made up a very small proportion of the new predatory `state class'. Furthermore, the nomads in the countryside remained relatively calm during this period; according to Ahmed I. Samatar, this was due primarily to the ideology of kinship and xeer that still prevailed in the `hinterlands', despite the creation of the centralized Somali state (1988, 70). Even though Somalia took on many of the characteristics of a predatory state during the reign of the civilian government, it had made no real attempt to destroy the system of traditional governance practiced by the nomads. This would change after Siyad Barre came to power in 1969 via a military coup d'etat.

On 15 October 1969, President Sharmaarke was killed by a member of his own police force; six days later, in the face of indecision by the Egal administration as to how to appropriately handle this situation, a military coup d'etat enabled General Siyad Barre to take over control of the state. Although many Somalis initially hailed the coup, viewing it as an attempt to eradicate the rampant corruption that had plagued the civilian government, it soon became clear that the previous predatory regime had simply been replaced with another, more virulent, one.

James S. Wunsch (1990) notes that centralist policies are found to varying degrees in every post-colonial African state. These policies include creation of one-party or no-party regimes, economic development strategies based on national planning and the creation of a bloated bureaucracy in order to carry them out, the concentration of legitimate authority in the hands of a single executive, attempts to abolish indigenous local-level governance structures (which are viewed as potential competitors to the centralized regime), and financial centralization (Wunsch 1990, 47-68). After coming to power, the Barre regime immediately began to implement every one of these centralist policies.

Within forty-eight hours of the coup, Barre's Socialist Revolutionary Council (SRC) banned all political activity, including membership in political parties, thereby creating a no-party regime. All of the institutions of the civilian government were also abolished by the SRC, which took on the previous functions of the President, Council of Ministers, National Assembly, and Supreme Court. At the local and regional levels, military or SRC functionaries directly accountable to the central government were appointed as district and provincial officials. In May 1970, the SRC began to nationalize oil distributing companies, all foreign banks, the Italo-Somali Electric Society, and SNAI (the jointly owned Italian-Somali sugar industry) (Ahmed I. Samatar 1988, 87-88).

Shortly after Barre announced on 21 October 1970 that "scientific socialism" was the ideology guiding the SRC, it was announced that more nationalization would take place.10 Barre believed that scientific socialism would be able "to turn this `nation of nomads' into a modern socialist state, to which people could look for leadership, security and welfare instead of the clan" (Bradbury 1997, 6). This meant, in essence, that the state was attempting to replace the system of kinship and xeer that had regulated person-toperson relationships in Somalia for centuries.

To this end, the Barre regime took extensive efforts to eradicate all vestiges of "tribalism". In 1970-71, a national campaign against tribalism was undertaken, which included demonstrations and the burning or burying of effigies that symbolized tribalism, corruption, misrule, and corruption. The death penalty was introduced in an effort to replace the traditional dia system. This was a fundamental challenge to the concept of xeer; the `death penalty' did not exist under the xeer system, instead, the harshest penalty that could be imposed upon a clan member bound by xeer was banishment from the clan. Individuals could no longer refer to their `ex-clan' as they had at the end of colonial rule and during the civilian regime. Furthermore, the term `cousin', the traditional way to address a member of the same clan, was outlawed and replaced with the term `comrade'.

Marriages, which were the traditional way to create a connection between loosely-bound lineage groups, were ordered to be held at newly-built orientation centers, and were stripped of any clan significance. The Akils system was manipulated once again as the title of Akil was changed to `peace-keeper', and the position became part of the state apparatus. Finally, regions were renamed to exclude any reference to clan lineage.

In addition to these direct efforts to strip the country of kinship ties, economic policies were introduced to further disrupt the life of pastoralists. The Land Registration Act of 1975 brought grazing lands under state control, and the National Range Agency was created in 1976 in order to settle pastoralists, as "a sedentary population is easier to control and tax" (ibid., 8). The Livestock Development Agency was also formed in the mid-1970s, with the goal of controlling livestock exports and `protecting' nomadic producers. Despite the creation of these programs and agencies, the relative share of state development funds spent on the pastoralist sector of the economy actually dropped from 8.6% during the 1971-73 development plan to 5.3% in the 1974-78 plan. In fact, the policies and programs that attempted to bring pastoralists under state control were never fully realized. This was at least partly due to the rising livestock prices during the mid to late-1970s, and the regime's realization that the state could still reap the benefits of these higher prices without bringing the pastoralist sector under full state control.

These economic policies, crafted in an effort to bring nomads more securely into the fold of the regime, were not the only failure. The more direct efforts of the regime to eradicate kinship ties and the tradition of xeer in the countryside were also met with resistance. Although pastoralists went through the motions of holding weddings at orientation centers, and referring to each other as comrade (when a regime functionary or sympathizer was present), these forced practices had little influence on the bonds of kinship and xeer that were so important to the nomadic pastoral way of life. As Lewis notes, "It would be unrealistic to expect these measures to have mad much immediate impact amongst nomads" (Lewis 1980, 214).

Barre's tactics, however, did serve to speed the decline of xeer in the southern regions, while at the same time strengthening clan distinctions. Barre built his base of power around three clan groups (his own Mareehaan clan, the Ogaden clan of his mother, and his son-in-law's Dhulbahante clan) and rewarded these clans (referred to as the MOD Alliance), as well as other clans and sub-clans that showed their loyalty to his regime, with lucrative positions in the state apparatus. Clan groups who were viewed as disloyal to the regime, on the other hand, were brutally repressed. Starting in 1978, these persecuted clans began organizing armed opposition movements.11 The first of these politico-military groups was the Somali Salvation Democratic Front (SSDF), formed in 1980 by members of the Majeerteen clan; this was followed by the creation in 1980 of the Somali National Movement (SNM), comprised of members of the Isaaq clan-family; the Ogaadeen-based Somali Patriotic Movement (SPM) formed in 198912; the United Somali Congress (USC), made up of members of the Hawiye clan-family, created in 1989; and the Rahanweyne-based Somali Democratic Movement (SDM) also formed in 1989.

By the beginning of the 1980s, Somalia had become almost entirely dependent on foreign aid in order to keep the state functioning due to the failure of the government's centrally planned economic policies.

The Ogaden had been part of Barre's MOD alliance, but members of this nomadic pastoral clan based in the Ogaden region of Ethiopia became incensed with Barre after his rapprochement with Ethiopia after the Somalia's defeat in the Ogaden War, due to their desire for the Ogaden territory to become part of Somalia. Massive influx of Ogaden refugees from Ethiopia allowed Somalia to receive an estimated $120 million (US) per year in humanitarian assistance; by 1982, it was the third largest recipient among African countries of foreign aid from the United States; by 1985, it ranked as one of the highest recipients of official developmental assistance per capita according to a World Bank report (Bradbury 1997, 9). The formation of armed politicomilitary organizations and their violent clashes with the Barre regime between 1988-91 were viewed as a struggle among competing groups no longer bound by xeer to capture the state in order to gain complete control of the distribution of this aid. As Abdi Ismail Samatar notes, "The shortsighted opportunism of leading members of the opposition led them to become engulfed in struggle over whatever was left of the carcass of the state" (1992, 637).

However, it must be noted that the Isaaq-led SNM was one of the few groups to actually publish its political agenda, which was based primarily on the reestablishment of xeer at the national level. Shortly after its founding in 1980, the SNM published a document that stated:

We propose a new political system built on Somali cultural values of co-operation rather than coercion; a system which elevates the Somali concept of xeer or inter-family social contract in which no man exercised political power over another except according to established law and custom, to the national level (Ahmed I. Samatar 1988, 142).

The SNM also expressed its desire to "reduce hierarchy and bureaucracy to a minimum", to "integrate effectively traditional Somali egalitarianism and the requirements of good central government", and to enable "both the regional and national governments [to] work according to the twin priorities of maximizing democracy and economic development".13 Although the SNM was formed in London by Isaaq intellectuals, businessmen, and mullahs living abroad, the Isaaq that remained in Somalia were still involved mainly in nomadic pastoralism. The SNM received the full support of the Isaaq after the Somali government killed approximately 50,000 Isaaqs between May 1988 and March 1989 in retaliation for the SNM's brief capture of the northern cities of Burco and Hargeisa.

On 6 August 1990 the SNM, USC, and SPM agreed to form an alliance in an effort to topple the Barre regime. The three groups attacked Mogadishu on 3 December 1990; after almost two months of fighting, Barre was driven out of Mogadishu on 26 January 1991. Once this goal had been attained, the alliance of the three groups fell apart. The USC hastily appointed a President and Prime Minister, then began to `cleanse' any Daarood clansmen (viewed as synonymous with the regime by the Hawiye) that they could find. The USC appointed `government' was not recognized by any other clan faction; almost continuous inter-clan warfare took place from December 1991 to March 1992, when a ceasefire was brokered by the United Nations.

III. After the Collapse of the State

Shortly after the collapse of the Somali state, Abdi Ismail Samatar expressed the view held by many scholars when he wrote, "In the absence of an organized indigenous agency which can establish peace and carry out such an agenda in the immediate future, the Somali people must rely on the international community to save them". Samatar and others believed that xeer had become separated from traditional blood-ties due to the manipulation of political elites vying for the control of state resources (ibid., 640), but this view seems to have been wrong on both counts. The international community (i.e., the United Nations) was unable to `save' the Somali people from further destruction, and xeer was not completely dead; it was, in fact, the basis for the relative peace that currently enjoyed by Somaliland.

UN Involvement in Southern Somalia

The UN had never carried out a peace-keeping mission in a country with no existing state apparatus, which lead to problems it had never encountered. Even after the state had essentially collapsed, the UN insisted that the deployment of a peace-keeping force without the consent of the warring parties would constitute a breach of Somalia's sovereignty. When the UN finally became involved in early 1992, it allowed clan-based warlords to play a central role in the numerous reconciliation conferences it organized.14

However, just as international aid can legitimize warlords and their actions (Anderson 1999, 50), so too can their inclusion in UN-sponsored conferences. To the extent that each of the competing warlords dreamed of replacing Siyad Barre as the sole possessor of political power in a new centralized state, it was highly unlikely that these conferences, which generally focused on the creation of a centralized state apparatus, would be successful. To worsen matters, the length of these reconciliation conferences was constrained by cost considerations; none lasted more than two weeks (Ahmed and Green 1999, 124).

The worst error in judgment made by the UN, however, may have been the decision not to include clan elders as part of the reconciliation process. Although most of southern Somalia was controlled by clan-based warlords attempting to gain complete dominance of the state, the UN failed to recognize that the warlords' goals did not match those of the clan elders. By excluding clan elders from the reconciliation process, the UN essentially denied any role for the very actors who may have been able to appeal to xeer as a means to bring peace to southern Somalia.

The Arta Conference and Transitional National Government

In October 2000, after a three month conference held in Arta, Djibouti, involving close to 2,000 Somali delegates, a Transitional National Government (TNG) was formed, consisting of a 245 member Transitional National Assembly elected by the delegates, and a transitional president nominated by the Assembly. The TNG was given the mandate to create a new constitution and a permanent national Somali government within three years. This provisional government has been accepted as legitimate by the international community, including the UN (which played no direct role in its creation), as it is seen to represent the true interests of Somalis; clan elders and other segments of Somali society took part in the Arta Conference, while the warlords were excluded from the proceedings.

However, the TNG is not viewed as legitimate by all Somalis. The majority of warlords do not recognize the TNG, and the leaders of Puntland and Somaliland view the TNG to be simply another faction seeking control of the state, pointing to the fact that a number of its members held upper-level positions in the Barre regime. These members of Barre's `predatory state class' have been accused of resuming their old practices, such as the misappropriation of loans from Saudi Arabia (Doornbos 2002, 93). Prominent clan elders who took part in the Arta Conference have recently called for the resignation of transitional president Abdiqassim Salad Hassan to resign, accusing him of corruption and claiming that he has been not carried out the duties of the office, and stating that the elders would consider him to be another warlord if he remains in office (Agence France- Presse 2002). In the absence of broad-based support from Somali society, the TNG controls only parts of Mogadishu and other small pockets of territory in the country. The troubles currently facing the TNG can be viewed as indications of the pervasiveness and resilience of the predatory state mindset among southern Somali politicians.

1 This included the regions of Awdal, Woqooyi Galbeed, Togdheer, Sanaag, and Sool.

2 An estimated 70% of Somalilanders depend, either directly or indirectly, on the production and export of livestock (Frushone 2001, 29).

3 There has been practically no inter-clan violence in Somaliland since 1997. meet these goals, even after the appointment of the Transitional National Government in October 2000

4 This obviously reveals the unequal status of men and women in Somali society. the shir, a council of elders consensually selected by the shir, called the guurti (which also participates in the deliberations), assumes the duties of monitoring and enforcement.

5 The British Somaliland Protectorate was controlled briefly by the Italians during World War II, and the `possession' of Italian Somalia was briefly handed over to the British after the war.

6 Italy was granted the responsibility of administering the trusteeship of Somalia in 1950 by the UN; Italian Somaliland had been controlled by the British during the previous decade (1941-1950).

7 The Protectorate Advisory Council was replaced by a Legislative Council in the spring of 1960, only a few months prior to the merger of the two territories.

8 Although these urban elites often preferred to be thought of as Somalis, the various parties still garnered their support from particular clan-families and lineage-groups.

(9 Prime Minister Egal was one of the few northern politicians to hold a high-level position before the Barre Regime (1969-91)

10 Indeed, a stated goal of the SRC was to "nationalize the commanding heights of the economy" (ibid., p. 89).

11 The influx of easily obtainable weaponry in Somalia is largely due to military aid provided by the Soviets during the Ogaden War (1977-78), and additional military aid provided by the United States to Somalia after the Soviets switched their support to Ethiopia.

13 The stated goals of the SNM contrast sharply with those of the SSDF, which called for the creation of a productive and planned national economy, and the end of `regionalism', among other things (ibid., pp. 142- 143).

14 This obviously reflects an external "top down" approach to peace-building.