He who rides the lion

Authoritarian rule in a plural society:the Republic of Yemen

by Omar Daair

MSc dissertation, School of Oriental and African Studies, London (September 2001)



Theoretical faramework
Migdal’s Approach - Monarchical Presidency - Corporatist Praetorianism and Limited Pluralism

The role of tribes
Background - Role in Social Control - Role in Military Security - Role in the Political System - Do Tribes constitute a ‘Strong’ Element of Society?

The role of Islamic groups
Background - Role in Social Control - Role in Military Security - Role in the Political System - Do Islamists Constitute a ‘Strong’ Element of Society?

Saleh's rule
Military and Security - Politics of Survival





"The unluckiest man in the world is he who rides the lion or rules Yemen" - from an ancient Yemeni poem

Migdal’s conception of strong society and weak state in the third world would seem to find a supportive example in the Republic of Yemen given the existence of traditional power centres outside of the state. The fact that tribalism is a major component of Yemeni society and that conservative Islamism appears to be on the rise should accentuate the lack of control Ali Abdullah Saleh’s state has over this society and leave him in a comparatively less secure position. However, in reality neither tribes nor Islamic groups can simply been seen as a traditional expression of power, both have become part of a modern system and Saleh has used limited pluralism to allow these groups expression yet maintain control over them. Neither grouping is homogenous or truly able to threaten Saleh’s rule. He uses patrimonialism and client-networks in the manner of many third world leaders to reinforce his own monarchical presidency. It is the strongman rule of Saleh that is the most important factor in the current Yemeni political system and not some strong/weak dichotomy between state and society.


Ibn Khaldun’s belief was that, "in the lands which are inhabited by a multitude of tribes it is difficult to establish a state" [1]. This claim can seem to take on added importance when used in relation to the Republic of Yemen, perhaps the most tribal of Arab states. As Sheila Carapico has written, "To the extent that a conventional wisdom exists about Yemen…it is that a combination of Islam and tribalism explains everything" [2]. What this research is concerned with is the nature of the relationship between these apparently strong elements of society and the structure of political power in modern Yemen.

Ali Abdullah Saleh does, indeed, rule with a firm hand, however, it also seems that the dominance of his regime is not as total as that achieved by Saddam Hussein, Hafez al-Asad or even, arguably, Hosni Mubarak. The laws and policies of the state have difficulty penetrating tribal heartlands and Islamist activity appears to be on the rise. One could attempt to explain this situation through the usage of a Migdal approach [3], in that the Yemeni state is weakened by the resilience of society and that the traditional pluralism that tribal and religious groups have created makes it harder for the regime to penetrate this society. In his PhD thesis, Ahmed Abdul Kareem Saif claims that, "The state in Yemen, historically, has had very insubstantial roots in society. The weakness of the state and the relative autonomy of society in Yemen made the state unstable…and obstructed the development of civic culture" [4].

One can question why, if societal elements are so strong, has Saleh survived for so long, considering also the context of his survival. He has ruled through the unification of the two former Yemens, an unmatched openness for an Arab state during the 1990-1994 period and a civil war. Add to this the vast economic problems Yemen has suffered over the past decade, heightened by the mass return of expatriate workers during the Gulf War, and Saleh’s regime maintenance seems an unexpected success. The issue becomes whether he remains in power because tribal and Islamic elements of society allow him to, by providing their support, or whether his regime has managed to exert a greater control over these elements than he is given credit for. The basic argument here is that society is less strong and Saleh more strong than a Migdal approach would allow.

The nature of the power balance between Saleh, the tribes and Islamic elements, including the religious Islah Party [5], will be investigated. This will be done in the context of Migdal’s arguments regarding weak states and strong societies; these theoretical dimensions will be expanded upon in the following chapter. The argument of this research shall be that while Yemeni society is indeed pluralistic, it has been too fragmented to really be ‘strong’, thus not truly fitting with Migdal’s approach. Saleh has been able to control and co-opt both tribal and Islamic elites for his own purposes and the leaders of these groups have allowed this to happen because it furthers their own interests. Neither tribes nor Islamic groups act as homogenous units; leaders have increasingly become distanced from their constituencies and different elements within both groupings have their own rivalries and competitions with which they must deal.

State-society relations are, at heart, a competition for control of resources and Saleh is winning this competition. His control of most of these resources allows him to give handouts in the form of material benefits and access to employment and positions of importance; neither tribal nor Islamic leaders have been able to truly challenge this. While Saleh does not have a totally freehand, he is as Chuck Schmitz has described him, "an astute strategist, and is not to be outdone in the game of coalition building" [6]. Christopher Clapham [7] discusses the difference between two types of leadership in the Third World, the lion and the fox. Ali Abdullah Saleh is a fox. He has not, in the main, sought to wipe out his rivals nor destroy opposition, as some other Arab leaders have done, but prefers negotiation and reconciliation to confrontation. This has stood him in good stead in his relationship to society; he has created his own power and maintains it with less violence than most Arab presidents.

While Saleh does not actually control all tribes or Islamic groups, the argument here is that their autonomy within the political system is limited, the pluralism in Yemen is limited pluralism. Saleh operates within a patronage system using clientelism in the manner of many other Third World leaders; for example, he uses his own tribe, Sanhan, to staff many important posts in the state and military apparatus, yet his regime is not tribal as such. Sheila Carapico argued that, "Yemen may be the one country where a regime can be forced to move incrementally and unwillingly, to incorporate the real pluralism of its society into the practise of statecraft" [8], my agreement with her would be slightly qualified. While Saleh must make concessions to the pluralism of society and must be careful to at least appear consultative, especially given the heavily armed nature of Yemeni society, he has not been forced to accept true power sharing with society. In fact he seems to be extending his power. He was able to increase the number of seats won by his party, the General People’s Congress (GPC) [9], in the 1997 parliamentary elections, in 1999 he was re-elected as President, challenged only by a member of his own party, and in 2001 he succeeded in amending the constitution in the way he desired despite opposition from Islah and the socialists. Tribes and Islamic groups have often acted outside the ‘rules’ of the state but this has not translated into a challenge to the state. One must distinguish between relatively low-level protest, which may be violent, against state behaviour and real opposition to the state and the regime itself that can reduce the power of the president.

The state around Saleh is not ‘strong’ yet neither is society strong. It is the personal nature of his rule that has become the dominant element, above the plural character of society and above the weak institutions of the state. Saleh is limited by the fact that there are sources of loyalty within Yemen other than the state, but these other elites have not successfully used this plurality to reduce his power. The argument here is that the Yemeni system is a result or a variety of factors, including a plural society but also affected by the nature of patrimonial relationships in the country and the nature of political deals that have been made by elites.

This research is divided into five further chapters. It will begin by discussing the theoretical basis of the research followed by an examination of the role played by both tribes and Islamic groups in post-independence Yemen. Further to this is a discussion of the patronage networks of Saleh’s regime and the manner in which he has entrenched his rule at the expense of these other groups in society. Finally, there will be an overall conclusion.


  1. Quoted in Tibi, Bassam, The simultaneity of the unsimultaneous: old tribes and imposed nation-states in the modern Middle East, in Khoury, Philip & Kostiner, Joseph (eds.), Tribes and State Formation in the Middle East, Berkley, University of California Press, 1990, p.128

  2. Carapico, Sheila, Civil Society in Yemen: The Political Economy of Activism in Modern Arabia, Cambridge, Cambridge University Press, 1998, p. 201

  3. See Migdal, Joel, Strong Societies and Weak States, Princeton, Princeton University Press, 1988

  4. Saif, Ahmed Abul Kareem, A Legislature in Transition: The Parliament of the Republic of Yemen 1990-99, PhD Thesis, University of Exeter, April 2000, p. 127

  5. Islah is the second largest political party in the Yemeni parliament and has been so ever since the 1993 elections. It has, in the main, been a partner to Saleh. It is not a simple Islamist party, but rather a coalition of conservative elements. Its role and make-up will be expanded on in chapter 4.

  6. Schmitz, Chuck, Civil war in Yemen: The price of unity?, in Current History, Vol. 94, No. 588, January 1995, pp. 33-36, p. 35

  7. Clapham, Christopher, Third World Politics, London, Routledge, 1985, pp. 72-74. A ‘Lion’ is akin to an autocrat who seeks to totally obliterate his opponents while a ‘Fox’ is closer to Machiavelli’s Prince and seeks to negotiate with, co-opt and control opponents without destroying them.

  8. Carapico, Sheila, Pluralism, polarization and popular politics in Yemen, in Korany, Bahgat et al (eds.), Political Liberalisation and Democratization in the Arab World: Volume 2, London, Lynne Rienner, 1998, p. 264

  9. The GPC was created by Saleh prior to unification as an umbrella organisation replacing all political parties in a similar vein to Gamal Abdel Nasser’s Arab Socialist Union. With the political opening that followed unification the GPC was transformed into a political party and has won the largest number of seats in all subsequent elections.

Theoretical framework

Migdal’s Approach

As mentioned in the introduction, the main theory within which the relationship between tribes and Islamic groups, as part of society and Saleh’s state will be examined is that of Joel Migdal. Migdal has examined state-society relationships in the third world extensively and has argued that, in the main, they exhibit weak states and strong societies [1]. This is an interesting test to put to Yemen. The fact that Yemeni society contains various centres of loyalty, such as the tribes, should mean that it fits well with Migdal’s argument and Saleh’s regime should be weakened by its inability to penetrate society or control it due to these other power centres.

According to Migdal, strong states are those with high capacities to penetrate society, regulate societal relationships, extract resources and appropriate these resources in a determined way. He believes that the fragmented control on society makes it strong and this forces the state to behave in a particular manner and forces leaders to act in specific ways if they wish to survive [2]. Through this approach both tribes and Islamic groups would be seen as competing with the state in Yemen. One can note the existence of heavily armed tribesmen who operate under distinctly tribal codes of conduct or the actions of groups of Islamist veterans of the Afghanistan war in attacking the Yemeni Socialist Party (YSP) prior to the civil war; both these examples provide a basis to claim that society in Yemen operates, to a large extent, independently of the state. One could argue that Saleh’s regime is only one among many groups challenging for citizens’ loyalty and that it has been unable to achieve sovereignty in the sense in that it has neither a monopoly on the use of violence or on creating the ‘rules of the game’.

For Migdal, the strength of a state should be taken as the degree to which it can expect voluntary compliance with its rules. A purely coercive state is not a strong state because a strong state needs to be able to alter the desires and goals of its people and this cannot be done through force alone. Ayubi [3] has dealt with this by distinguishing between ‘strong’, ‘hard’ and ‘fierce’ states. In his view Arab states do not qualify as strong but rather are fierce states, in that they are unsuccessful in generating legitimacy and acceptance for their policies amongst citizens, thus he writes, "the Arab state is therefore often violent because it is weak" [4]. This view of state strength encompasses what Michael Mann calls ‘infrastructural power’ (as opposed to despotic power), which is, "the capacity of the state to penetrate civil society and to implement logistically political decisions throughout the realm" [5]. In Migdal’s view, Arab states do not possess this and, therefore, they are weak. In Yemen Migdal would argue that the autonomy of tribes and religious groups proves that society is stronger than the state.

Migdal writes, "State leaders in weak states have taken to pulverising the very arms of the state that could achieve their goal of mobilisation. Their purpose has been to prevent leading officials in important agencies from using their own mobilisational capabilities against the state leadership" [6]. This highlights one of the reasons why strongman leaders have sought to weaken the state, the fact that even if politics can be removed from society, it may still be present amongst state agencies, and therefore could come to challenge the regime. For a leader such as Ali Abdullah Saleh, the potential challenges to his regime could come from within his ruling clan, from the military or from traditional groups such as the tribes or Islamists. For this reason much energy must be expended on survival and this makes it very difficult to develop hegemony. It is not Migdal’s argument that authoritarian leaders cannot survive in a situation of strong society and weak state, but he believes they are forced to pursue ‘politics of survival’. The politics of survival involve the big shuffle, non-merit appointments and dirty tricks [7] and this, for Migdal further weakens the state. To judge increased social control he sets out three criteria for examination, compliance, participation and legtimation. One can discuss how well Saleh has used force, semi-democratisation and negotiation with tribes and religious groups to solidify his regime in this respect.

Monarchical Presidency

If hegemony has been difficult to achieve and, thus, Saleh cannot totally dominate over all elements of society, the longevity of his rule must be explained. One must also question whether ‘politics of survival’ do fully explain the nature Saleh’s rule, especially in relation to tribes and Islah. One can also ask whether simply because the state is weak society must be strong or whether the fragmentation of both leaves both weak. It is within this structure of weak state and weak society that the type of ‘monarchical presidency’ that exists in much of the Middle East becomes important.

Though they do not look at any Arab countries, Chehabi and Linz’s conception of ‘Sultanistic regimes’ [8] can be applied; a situation where the lines between the regime and the state or even the leader and the state become blurred. In most Arab republics the president reigns almost untouchably supreme, dominating all other institutions within the state; to quote Ayubi, "The Boss is extremely crucial and is usually a ‘presidential monarch’ in the sense that he enjoys constitutional or de facto life tenure of office" [9]. As already mentioned, Saleh has not achieved the dominance over the state that some of his contemporaries in the Arab world have and thus, once again we may be led to see this as the resilience of Yemeni society due to the greater traditional autonomy of groups within it.

However, one could also ask whether Saleh has actually been much weaker than other Arab leaders. Arguably he has not. Much of what has taken place in Yemen over the last decade has been as focussed on regime-maintenance as any other Arab leader. Society has not really acted against Saleh; in fact tribal leaders and Islamic groups have helped to protect his regime. He himself has been successful in playing various groups off against each other to ensure that his image becomes completely associated with the image of the Yemeni state and that to think of the state without Saleh becomes increasingly difficult. Brian Whitaker has written, "Even some of Saleh’s opponents grudgingly concede he may be the only driver capable of keeping the truck on the road" [10]. It appears that Migdal’s politics of survival are only part of the story and that, in fact, Saleh has gone further, not just surviving but building up the strength of his rule at the expense of both state and society organisms.

Corporatist Praetorianism and Limited Pluralism

Perlmutter in his study on modern authoritarianism [11] described many Arab regimes as ‘oligarchic praetorianism’; Saleh’s would seem to fit with this. He writes, "Oligarchy is at the heart of exclusionary corporatist and praetorian politics. The regimes substitute patrimonial, oligarchic, bureaucratic, or corporate arrangements for an otherwise institutionalised party-state or state-dominated authoritarianism" [12]. This can be seen as fitting with the Migdal approach in that this type of regime fails to institutionalise the structures of its dominance and thus must rely on support from corporate coalitions such as the military, the ‘church’, industrialists and so on, on a patrimonial and clientalistic basis. Saleh’s regime does depend heavily on the military and his elite Republican Guard for security but it has not been as ‘fierce’ as numerous other Arab regimes and it can be said that tribes and Islamic groups fall into the corporatist praetorian system in Yemen.

The fact that patrimonialism is such an important part of regime survival in the region cannot be underestimated and it is this patrimonialism that overarches any simple strong-weak dichotomy in terms of state and society. The thread of clientelism that runs through Yemeni and Arab society ends up limiting the autonomy and effectiveness of both societal and state machinations, thus, the leader’s rule quest for regime-maintenance becomes distinct from the two. It is also significant that the pluralism of Yemeni society is only limited because this creates space for Saleh to manoeuvre.

It is useful here to quote Perlmutter at length in discussing his scepticism of ‘limited pluralism’. "To speak of limited pluralism is to confuse elite circulation and co-operation with a more widespread collective political behaviour. Limited pluralism, if it is a valid and verifiable theory, must refer to the existence of autonomous and independent political groups of counterelites…It must demonstrate in however limited a manner, that the system represents a multiplicity of interests and that there is a political process which revolves around conflicting sets of cross-cutting alliances, where bargaining and group accommodation occur" [13]. Yemen does fulfil many of these criteria. There exist elites, such as tribal leaders and Islamist figures, who do attract the loyalty of large groups of people, and within both ‘tribes’ and ‘Islamists’ there are crosscutting cleavages that prevent them from being monolithic blocks. The pluralism of society and the political system is limited in that Saleh remains the strongman whose client networks secure his power, yet at the same time he does have to bargain with the elites growing out of the traditional plural nature of Yemeni society. It is because this limited pluralism exists that tribal and religious leaders can be co-opted into the polity and away from their constituency and they become a part of a larger client network, rather than representatives of societal interests.

In this situation there is no single set of ‘tribal’ or ‘Islamic’ core concerns with which Saleh’s regime can be challenged. Therefore, it can be argued that while Migdal may be right in seeing the state as weak because of its inability to penetrate society, the leader may actually be able to use this to strengthen his rule and use this fragmentation of power to his own benefit to a greater degree than simple survival. In light of this, Migdal’s approach must be adapted and developed to fit the Yemeni case and this shall be discussed in the following chapters.


  1. See Migdal, Joel, Strong Societies and Weak State, Princeton, Princeton University Press, 1988

  2. Migdal argues that due to the resilience of society, the state is limited in what policies it can pursue without totally weakening itself. The state’s actions are dictated by the need to survive in the face of this societal resilience.

  3. Ayubi, Nazih, Overstating the Arab State, London, IB Tauris, 1995

  4. ibid., p. 450

  5. Mann quoted in Migdal, Joel, et al (eds.), State Power and Social Forces, Cambridge, Cambridge University Press, 1994, p. 12

  6. Migdal, Joel, Op. Cit., p. 264

  7. These will be expanded upon below.

  8. See Chehabi, H.E & Linz, Juan (eds.), Sultanistic Regimes, Baltimore, Johns Hopkins University Press, 1998

  9. Ayubi, Nazih, Op. Cit., p. 205

  10. Whitaker, Brian, Yemen’s Decade of Unity, in Middle East International, No. 625, 19/05/2000, p. 19

  11. See Perlmutter, Amos, Modern Authoritarianism: A Comparative Institutional Analysis, New Haven, Yale University Press, 1981

  12. ibid., pp. 147-148

  13. ibid., p. 169

The role of tribes


During an interview in 1986 Ali Abdullah Saleh was asked to what extent Yemen had succeeded in moving on from the stage of tribalism to that of the state, he replied, "The state is part of the tribes and our people is a collection of tribes" [1]. In contrast, former Prime Minister Abdul Karim al-Iriyani said in a lecture at the University of Exeter that, "it is my view that tribalism is a rural institution. It is being weakened by education, modernization and urbanization" [2]. Clearly, tribalism remains an important part of Yemeni society and politics, and this would be drawn upon by Migdal to highlight that state leaders are not the only ones to exert pressure on the behaviour and loyalty of the people. Social control, he argues, is distributed between various groups and these various social organisations may be in conflict as to who has the right to create the rules of the game. Saleh’s regime must deal with this element of the Yemeni polity if it wishes to survive; the local tribal leaders cannot simply be ignored.

Tribalism’s heartland is in the former North Yemen, it is here that their presence and influence has been most salient. The tribes of the north are organised into three main groupings or ‘confederations’, these being Hashid, Bakil and Madhij, and of those the former two are of greater power. It is Hashid that is often seen as the dominant tribal grouping of the state, given that Saleh’s tribe, the Sanhan, are part of this confederation and Abdullah al-Ahmar, its leading Sheikh, has become increasingly politically important. However, as we shall see, it is not tribal rule but the rule of a smaller circle within it. As Dresch explains, even with tribes which seem to be powerful, it is an elite and not the tribe as a whole which exerts this power, "most Hashidis were as alienated from the holders of power as most Bakilis or people from Madhaj" [3].

The south has been less tribal than the north; this is partly tradition, partly to do with the Marxist rulers of the former People’s Democratic Republic of Yemen (PDRY). They condemned tribalism as akin to feudalism and worked to break up any tribal system and the power of tribal leaders. As a result, tribes in the south tended to be structurally more divided and had less open involvement with national issues than their northern counterparts. This is not to say that tribes in the south have been eradicated, indeed, Saleh’s northern regime sought to use them as a constituency in his competition for power with the YSP prior to the civil war. After unification the northern leaders had worked to rekindle the power of southern tribes such as the Abyan and during the civil war itself southern tribes such as the Awlaqi of Shabwa were important in aiding Saleh’s cause.

Tribes have certainly played an important role in Yemen and one can investigate both whether they contribute to a Migdalian strong society and to limited pluralism. In this context it is important to distinguish between tribes and tribal leaders as to how much power or influence they wield and to what extent Saleh must bend to their wishes.

Role in Social Control

After unification it was Saleh’s aim to continue ruling Yemen in the same manner he had ruled the former Yemen Arab Republic (YAR); this was based on patrimonialism stemming from his inner circle of relatives and members of the Sanhan tribe. Kostiner writes, "although Salih did not fully control all the Northern tribal groups, such as Bakil, tribal values, articulated through an atmosphere of political bargaining and patrimonial appointments, characterised the Northern rule" [4] and he explains that at the time of unification, northern tribes functioned as corporate groups. While they did not always act together in public life, they did seek to disseminate their traditional ideas through the national arena.

It is within this framework that tribes may be seen as contributing to the strength of society in a way that clan may do in Senegal or religious sect in Lebanon. Particularly in the north one could make the argument that tribes operated under a system in which government authority ruled the towns and tribal authority ruled in the countryside. Martha Mundy agrees that the villages see a clash of loyalties, "The village community stands within two hierarchies of power: the shaikhly and the state" [5], indeed, it is not surprising that traditional structures of power remain stronger in the rural, less educated areas. It is, however, deeper than a tradition that has yet to be swept away by the modern world. One can consider the ‘primordial civic realm’ that Carapico has discussed. It is the idea that both religion and tribalism provide services for people and, thus, their importance does not stem simply from tradition but from the reality that they may provide services to citizens which the state is failing to and so they attract the loyalty of these citizens.

The idea that societal institutions, such as tribes, may perform certain governmental tasks better than the state fits with Migdal’s argument. It becomes harder for the state to enforce their laws and policies in tribal regions that feel they benefit little from the central state. In an interview with the Yemen Times, Sheikh Abdullah Mohammed Taoiman, Chief of the Gahm tribe, the biggest in Mareb, made the point that much tribal anger stems from a feeling of lack of investment in the development of tribal areas. He said, "Ex-governors…showed no enthusiasm about providing services for the city", he goes on to criticise the government’s heavy deployment of troops in the area, "Had money spent on those forces been spent on services projects in the governorate, there would have been no acts of sabotage" [6]. Clearly, the Saleh’s regime must create the belief that the state does have something to offer tribal areas.

This is something Saleh’s regime was attempting even before unification. As Dunbar writes, "The keys to success were the construction of schools and hospitals in tribal areas and, in particular, the discovery of oil in the Marib Province, in the heart of the eastern tribal country" [7]. This gave the government an increased ability to dispense services and jobs to tribal members and, through this, gain their increased acceptance of a growing official presence. This contributes to Migdal’s criteria of increased social control in all three categories: compliance, participation and legitimation. Through an increased presence the government has greater ability to control and organise, at least to some extent, citizen’s behaviour and through their consumption of government provided services, tribesmen give a type of tacit consent to Saleh’s regime. Dunbar further points out that this has had an impact on the tribal leaders because, "These men, who were previously the predominant sources of power and patronage in the regions found themselves in the position of vying for favours that the government was increasingly in a position to give, and thus being co-opted into what was becoming a Yemeni Establishment" [8]. The regime increased this distancing of tribal leaders from their tribes through giving them lucrative business opportunities; meaning that tribal leaders became more devoted to their business and less to maintaining their position and influence within their tribes.

Role in Military Security

One of the main reasons for seeing tribes as a strong entity with which Saleh must deal is the ready availability of arms in Yemen and the fact that tribesmen do hold large stores of weaponry. The Yemen Times recently ran an article concerning tribal clashes with the army in Obeidah over the piercing of an oil pipeline, under the headline "They are Challenging the State!" [9], but this is exaggerating the situation. It is, in fact, rare for clashes to develop to a serious level between tribes and the state, in most cases it is lower level banditry such as the piercing of pipelines or kidnappings which are resolved relatively peacefully, sometimes with direct intervention from Saleh himself.

The northern tribes have enjoyed Saudi Arabian financial support for a long period, given this and given Saudi dislike and distrust of Saleh [10] one would expect more direct challenges from tribal leaders to the regime. Indeed, one can consider the civil war as an ideal time when tribes could have reasserted themselves or demanded more from Saleh in return for support but this was not the case. Dresch states that, "Put briefly, there was no tribal factor in the 1994 Yemeni crisis…none of the major tribes joined the fighting as a tribe" [11]; while it is true that the large tribes did not officially get involved it can be taken further. Kostiner points out that the northern forces did benefit from the addition of thousands of men from northern tribal militias, particularly from Hashid, and Halliday [12] argues that there was significant support or at least acquiescence for the northern advance by key southern tribes.

This is significant in terms of the strength of tribes as societal organisms. First, it indicates that tribes do not present a strong single block with which Saleh must deal. They are divided amongst themselves and seek to further their own interests, not some greater tribal agenda; this allows Saleh to deal with them separately and also encourage competition amongst them (see below). The fact that tribes did not become overly involved in the civil war suggests that they have not sought to truly challenge Saleh’s rule, it would have been an ideal opportunity to strike, but in reality leaders saw the chaos could have resulted had the north failed to win. Tribal leaders accepted that Saleh’s was the best system on offer to serve their interests. To quote Taoiman, "I, once again, stress that tribes have become more aware of consequences of conflicts. What they need is more security and stability. When they feel that they can travel to any place in Yemen with no fears about their safety, they will be the first to disarm themselves" [13]. The more Saleh’s image becomes dominant the greater his sway with tribes and his ability to avert armed conflict.

Role in the Political System

The Yemeni political system experienced unmatched openness during the 1990-94 period and though there has been a rolling back of some of the liberalisation process, Yemen remains a more developed political system than most Arab states. One can question what role tribes have actually taken in the political system. One significant factor here is the role of the Islah party, the party in many ways straddles the two elements of this research in that it is partly tribal and partly Islamic, however, it will be dealt with more extensively in the next chapter.

During the more open period tribal groupings, like many others, became increasingly vocal, as Carapico has written, "Some seven other tribe-based but civic mass conferences in 1992 each issued written demands for the rule of law, pluralism, economic development and local autonomy" [14]. Hashid, Bakil and Madhij all held an increasing number of tribal conferences, but they did not represent a cohesive block; they created no real agenda to force upon Saleh. Tribal figures have been prominent in all elections since unification and while this should be taken into account, it should not be confused with a tribal take-over of the political system.

During the 1993 elections, Carapico notes, Islah ran two types of candidates. In the Hashid strongholds they ran people chosen more by tribal or kinship links with the al-Ahmar family, while outside they ran those more broadly identified with the Islamists. Indeed, Renaud Detalle argues that the GPC did exactly the same thing, "In putting together their slate, the GPC looked for persons well-rooted in their communities, with party affiliation taking second place. Many tribal leaders, of course, but also big merchants and high officials…" [15]. One cannot be surprised that in a developing democratic system patrimonial or local concerns play a large part, this is not something unique to Yemen. One need only look at the far more developed democracy of India to see that these types of concerns are still of a high level of significance in politics.

The fact that parties recruit candidates because they already have a high societal standing could show that the state is weak in the face of society because parties cannot depend on the loyalty of their MP’s. Ahmed Abdul Kareem Saif has made this argument, claiming that the fact that social figures are recruited by parties because of the influence they already exert means that MP’s "ask what their party can do for them, not what they can do for their party" [16]. Even if this is the case, I would argue that this does not weaken Saleh’s regime substantially. Saif is right in claiming that even GPC MP’s do not feel obligated to support the government in parliament because they are rarely consulted in policy-making, but the importance of this in Saleh’s system is limited. Achieving a large majority for the GPC is the point of the parliament, what happens within that parliament is less important for the President. By creating a somewhat open political system, he increases people’s participation in that system, thus increasing its legitimacy, two of Migdal’s criteria for increased control. The fact that tribal leaders are brought into the polity gives them a stake in it, without truly giving them control over Saleh’s executive. One of Migdal’s politics of survival is non-merit appointment and Saleh uses this extensively to solidify his power base by drawing prominent figures closer into his orbit and making them a force for the status quo rather than reform.

Do Tribes constitute a ‘Strong’ Element of Society?

If we return to the concept of limited pluralism then tribes do appear to be part of this; their elites are influential and they do involve themselves in politics. It is limited because they do not take it upon themselves to challenge the system that Saleh has created and is entrenching, his negotiation and co-option has brought him out on top. As Schmitz wrote, "The regime is led by Ali Abdulla Saleh and his fellow clansmen from Sanhan…The major posts in the military and the security apparatus are stocked with ‘relatives’ of the President…Manipulation of tribal loyalties creates a favoured minority loyal to the regime and reinforces tribal cleavages in Yemeni society" [17]. The fact that it is a favoured minority is important. Saleh uses the existence of a tribal system and tribal loyalties to develop his network of patronage and clientelism in the same way Asad used the Alawi network and Saddam uses the Takriti clan. The tribal system creates a space in which Saleh can co-opt the ‘right’ people and draw them into the polity in order to reduce the likelihood of any challenge to his rule.

In Migdal’s assessment of state-society relationships he argues that the state must take notice of local strongmen as the sum total of chiefs could stand up to the state; however, the point is that the sum total do not work together. While Saleh does not attempt to destroy the power of tribal leaders he does detach them from the larger tribe. Ayubi is correct in writing that, "The incorporation of most tribal leadership became more extensive in the 1970’s and 1980’s, and this became part of a much more elaborate corporate formula under President Ali Abdullah Salih" [18]. It fits with Perlmutter’s conception of oligarchic praetorianism; tribal leaders have become part of Saleh’s corporate coalition.

According to Kostiner, Saleh’s Sanhan clansmen occupied 48% of the top political posts and 70% of the public administrative posts in Yemen [19]; clearly he depends on an inner circle of tribesmen for power rather than on the tribal system as a whole. In such a system, which is prevalent throughout the Arab world, the power of others in the polity stems from their proximity to the leader. To ensure this Saleh has had to work for a distancing of major Sheikhs from their followers, a process known as tabaud [20]. By drawing Sheikhs into business and economic arrangements that affect them personally, the regime has been able to reduce their concentration on promoting tribal issues and to sow seeds of competition and discontent amongst tribes and tribesmen. It also heightens the commitment of these tribal leaders to the status quo because it is a situation that benefits them materially. Al-Ahmar is an extreme example, while his high profile, including the position as Speaker of Parliament, should give the Hashid confederation greater benefits and influence, it in reality does not. Al-Ahmar is part of a small circle within Hashid that holds any influence. The political power of the tribes as a whole has been reducing vis-à-vis the government over the past fifteen years and its elite has been pulled further into Saleh’s polity.

Tribes have become part of the controlled corporatist system and there are crosscutting cleavages between and within tribes. There is not a ‘party of the tribes’ or individual parties that represent individual tribes or confederations, even with al-Ahmar’s involvement in Islah. As Carapico writes, "parties did not represent tribes nor did party loyalty rest on tribal affiliation. Rather, within each locality (and some families) were many parties and within each party were people of different tribal (and non-tribal) origins" [21]. Once again one can see that the limited political opening allowed under Saleh’s system has here worked to his benefit because it divides these societal organisms which Migdal would claim are strong.

Within the limited pluralism of the Yemeni political system tribesmen have become increasingly involved in taking part in elections or attempting vote-rigging and so on, and this further develops their participation in the system. The fact that they take part and even attempt to cheat gives Saleh’s system an edge of legitimacy that increases his hold on society. This is then combined with the clientelism that tribal leaders have been drawn into and which permeates all the way down through state and society to enhance Saleh’s personal position. The system does represent the ‘multiplicity of interests’ of which Perlmutter writes, but it does so in a way controlled by Saleh’s regime. Thus, while tribes and tribal autonomy have been a source of weakness for the state in extending its monopoly on rulemaking, they have not been a source of real societal power and are decreasingly so. It is Saleh’s presidency that now dictates the direction and pace of Yemen’s development, albeit in a more negotiated form than several of his contemporaries. The tribes, therefore, do not, as a whole, find themselves in much more a powerful position than corporate groups in other Arab states.


  1. Quoted in Dresch, Paul, Tribes, Government and History in Yemen, Oxford, Oxford University Press, 1989, p. 7

  2. Lecture delivered by Dr al-Iriyani to the Centre for Arab Gulf Studies at the University of Exeter during a conference on ‘Yemen: The Challenge of Social, Economic and Democratic Development’, 1-4 April 1998. Dr Iriyani was Foreign Minister at the time. Transcript can be read at, www.al-bab.com/yemen/gov/iryani.htm

  3. Dresch, Paul, The tribal factor in the Yemeni crisis, in al-Suwaidi, Jamal (ed.), The Yemeni war of 1994: Causes and Consequences, London, Saqi Books, 1995, p. 41

  4. Kostiner, Joseph, Yemen: The Tortuous Quest for Unity, 1990-94, London, Royal Institute for International Affairs, 1996, p. 28

  5. Mundy, Martha, Domestic Government: Kinship, Community and Polity in North Yemen, London, IB Tauris, 1995, p. 201

  6. Interview with Sheikh Abdullah Mohammed Taoiman published in the Yemen Times, July 31st-August 7th 2000

  7. Dunbar, Charles, The unification of Yemen: process, politics and prospects, in Middle East Journal, Vol. 46, No. 3, Summer 1992, pp. 456-476, p.467. See also Dunbar, Charles, Internal politics in Yemen: recovery or regression, in al-Suwaidi, Jamal, Op. Cit., for further discussion on the issue

  8. ibid., p. 468

  9. Yemen Times, July 2-8, 2001

  10. For example Kostiner writes, "In meeting with the US, GCC and Egyptian official [during the civil war], the Saudi leaders…depicted Salih as a ‘mini-Saddam’…portraying Salih as the villan who must be stopped". Kostiner, Joseph, Op. Cit., p.99

  11. Dresch, Paul, Tribes, Government and History in Yemen, Oxford, Oxford University Press, 1989, p. 33

  12. See Halliday, Fred, The third inter-Yemeni War, in Asian Affairs, Vol. 26, No. 2, 1995, pp. 131-140

  13. Interview with Sheikh Abdullah Taoiman, Op. Cit.

  14. Carapico, Sheila, Elections and mass politics in Yemen, in Middle East Report, Vol. 23, No. 6, November-December 1993, pp. 2-7, p. 4

  15. Detalle, Renaud, The Yemeni elections up close, in ibid., pp. 8-12, p. 8

  16. Saif, Ahmed Abdul Kareem, Yemeni electoral and party systems, in Al-Masar Journal, Vol. 1, No. 1, Winter 2000, pp. 15-32, p. 18

  17. Schmitz, Chuck, Civil war in Yemen: The price of unity?, in Current History, Vol. 94, No. 588, January 1995, pp. 33-36, p.35

  18. Ayubi, Nazih, Overstating the Arab State, London, IB Tauris, 1995

  19. Kostiner, Joseph, Op. Cit., p. 86

  20. See Dresch, Paul, Op. Cit.

  21. Carapico, Sheila, Civil Society in Yemen: The Political Economy of Activism in Modern Arabia, Cambridge, Cambridge University Press, 1998, p. 166

The role of Islamic groups


Yemen, particularly the former YAR, is a deeply religious society and since unification has seen a growth in so-called Islamist activity. The Islah party with its prominent position in the political system is often taken as evidence of this growth of traditionalist fundamentalism in the country; this is, however, a simplification. As Carapico writes, "Islah…is a thoroughly modern party, critical of many Yemeni religious and folk traditions, but a conservative, anticommunist party valourizing private property, family values, capital punishment…[etc]" [1]. Clearly the role of Islah as a religious party requires further discussion, but one should first clarify the main religious groupings in the country.

The chief sectarian split in Yemen is between Zaydism, found mainly in the northern highlands of the former YAR, and Shafi’ism, found in most of the rest of the country; the former is a branch of Shi’i Islam while the latter is part of Sunnism. While sects have not tended to be as antagonistic towards one another, particularly in recent years, as is often the case in the Middle East, one can note that most of the ruling class are Zaydis. Saleh, for example, is a Zaydi and as is Sheikh al-Ahmar. Islah, which al-Ahmar heads, has come to represent something of a neo-Wahhabism, drawing on Saudi influence, while al-Haqq, the second largest religious party, is a more traditional Zaydi party seeking a return to the rule of Zaydi sayyids. The different status of the two parties can be seen by the number of the seats won by each in the 1993 and 1997 parliamentary elections. Islah gained 63 and 53 respectively, while al-Haqq managed only 2 and then none at all [2].

It is also worth noting the difference between the former north and south Yemens in their attitude towards religion. The north was always more conservative and the YSP was often accused of being atheist and damaging Islam. As Gerd Nonneman has written, "Islamist opposition [to unification] was indicative of the tensions between northern and southern social mores- traditional and markedly religious in the north versus more liberal and secular ways in the south, especially Aden" [3]. Indeed, the PDRY did tend to be more socially liberal; women, for example, enjoyed greater access to employment and political positions in the south and were subject to more liberal personal status laws [4]. The northern victory in the civil war and the influence of Islah saw conservatism spread further through the country.

Outside of political parties, one can also discuss the role of Islamic groups such as the Muslim Brotherhood in Yemen, founded by Sheikh al-Zindani who is al-Ahmar’s partner in Islah, or even the grouping who were nicknamed Yemen’s Islamic Jihad, run by Sheikh al-Fadhli [5]. The role these all play in state-society relations shall be examined and one can ask if Islamic groups, like tribes, have also become part of Saleh’s corporatist system. Abdul Rahman Ali al-Jifri, the exiled leader of the Sons of Yemen League said, with noted personal bias, "Ali Abdullah Saleh presents himself to the world, and especially to Egypt, as someone who is capable of resisting extremism, whilst domestically he has been encouraging this very extremism" [6]. Clearly, one must establish to what degree Saleh is attempting to, and how successfully he does, co-opt Islamic groups or parties into the patronage network and into the limited pluralism over which he dominates.

Role in Social Control

Islam does, certainly, play a major role in the lives of Yemeni citizens, but one can question whether Islamic groups operate any sort of effective social control. In line with al-Jifri’s statement, Carapico has claimed that Saleh encouraged the Islamists in the late 1980’s as a counter against leftist elements in the north who sympathised with the PDRY [7]. This was something he continued in the united Yemen, "Salih did set the religious right against both the northern leftists and the Yemeni Socialist Party, helping to propel Islamist ideologue Abd al-Majid al-Zindani…into one of five seats on the ruling presidential council" [8]. One can see the growth of social conservatism in Yemen and, indeed, the rivalry between Islah and the YSP can be viewed partially in this light. This must also consider Islah’s links with more extreme fundamentalist groups such as the Islamic Jihad organisation.

Islah represents a type of coalition between moderate Islamists, radical Islamists, tribal leaders and some conservative businessmen; this collection of interests has prevented it from becoming a real ‘fundamentalist’ party. Indeed, its two main leaders al-Ahmar and al-Zindani are by no means the same. Ahmar is not an Islamist but rather a social conservative who has made a political alliance with Islamists such as Zindani and it is an alliance that has served him personally very well.

One of the first disputes between the YSP and Islah in the united Republic was whether Islam should be the sole, rather than the main, source of legislation. Islah and Islamist groups were angered that Article 3 of the constitution did open the way for other sources of legislation, but this was something they managed to change after their enhanced position following the civil war (see below). A further dispute was over the so-called ‘scientific organisations’ that were operated by religious groups; they were presented by Islamists as educational institutes, however, the socialists saw them as a front for the training of militants. In 1992 the YSP introduced legislation to bring an end to these institutions and this met with a great deal of opposition from Islah and there were increased attacks on YSP members. One should note that at this time the YSP still enjoyed a high standing in the polity due to the almost 50-50 split of political roles between themselves and the GPC after unification; it was only after Islah’s success at the 1993 elections and the defeat of the YSP in the civil war that Islah could push a more conservative agenda.

In this early period of unity, Islamist groups lacked institutional reach in society, they did not have the structured framework of the tribes and so often tribal and Islamist actions were linked. This can be highlighted by the alliance of Ahmar, Zindani and Fadhli. As Watkins writes, "The full history of the relationship between the al-Islah party and the Islamic Jihad Organisation has yet to be written. But, in a broad outline, they appear to have been partly overt and partly covert partners joined by the common aim of undermining the former rulers of the Marxist south" [9]. In his argument they were linked not just by Islam but by tribalism as well in that Zindani and Ahmar (but not Fadhli) had tribal links through Hashid; these tribal links could also be extended to President Saleh. He claims, "It thus appears…that Shaykh Zindani and his followers were duped into providing an Islamist front- and thus popular support- for what was essentially tribal opposition to the YSP" [10]. He is overstating the case somewhat, it was not simply that the Hashid confederation was opposed ideologically to the socialists and their leaders expressed this; it was the case of the top elite of the north seeking to destabilise their rivals for power. Most Hashidi tribesmen were as divorced from this decision as anybody.

Once again, more fundamentalist Islamists such as Zindani and Fadhli were allowed to express their views and gain the loyalty of their followers within the limited pluralism that exists in Yemen but Saleh ensures the leaders are co-opted in a safe way. The GPC adopts some of Islah’s policies, such as conservatism on polygamy or the sale of alcohol, in order to strip them of their ability to use the Islamic card on a larger political scale, as they had done against the YSP. Saleh has also used al-Haqq in order to underline his control, for example after the 1997 elections, the Awqaf (Islamic endowments) portfolio was given to al-Haqq rather than Islah, something of a slap in the face for his partners.

Petrodollars, particularly money coming from Saudi Arabia, have boosted neo-Islamic organisations in the Republic. Most of this money has gone to Islah and much of it used on social services, especially in the south, as Saif writes, "[Islah] took the initiative to help people in the South. It filled gaps in the social services: health care, emergency relief, post-secondary vocational training, religious education…These projects…reached many thousands of lower income families" [11]. While this may seem to be deepening Islamist penetration of society it can also benefit the regime, as Carapico writes, "In cushioning society from the ravages of corrupt economics, they also helped cushion the government form popular discontent" [12]. Once again Saleh’s regime has allowed enough activity to bring benefits and legitimation to his regime, yet he has ensured that the Islamists own patronage links run back to leaders who are within his own elite or client-network.

Role in Military Security

Islamic groups have not been armed to the extent of the tribes, but in a society where weapons are so widely available there is clearly a risk that this could occur; indeed, as Hassan Abu Taleb has written, "The civil war was a golden opportunity for fundamentalist groups to acquire a de facto legitimate position across the whole of the country and seize military stocks" [13]. Al-Fadhli’s organisation especially, was violently mobilised against the YSP. Prior to the civil war his Jihad Organisation made up to a large extent of veterans of the Afghan war carried out terrorism against the socialists, including assassinations of party members and relatives of Ali Salem al-Baydh [14]. However, Zindani and even Ahmar made emotive calls for Jihad against the YSP prior to 1994 and Watkins believes this, in part, encouraged the civil war.

While the tribes did not officially take part in the war, Islamic elements were much more vocal and this was due in large part to the feeling of differences of culture and beliefs with the YSP. This is interesting in the context of Saudi Arabia being a patron to Islamic groups because they had given much support to al-Baydh in order to encourage separatism and a weakening of Saleh’s regime. What this highlights is that various Islamists have different aims, views and leanings, much like any other group and are not the homogenous unit the western press often paints them as. Some Yemeni Islamists look more towards Saudi Arabia, others to Turabi in Sudan or to other Muslim Brotherhood organisations in the Arab world.

After the civil war, Fadhli declared that the destruction of the YSP as a political power meant that the Jihad Organisation had achieved its aim and no longer had reason to exist. Islah, however, felt that its support for Saleh during the war, combined with its performance in the 1993 elections should give it a greater say in the political system, particularly in social legislation.

Role in the Political System

Ever since the first Parliamentary elections in 1993, the Islah party has been second only to the GPC in terms of seats won and given that the YSP lost almost all its power following the civil war, one would expect to see a growth of power that could restrict Saleh. As already mentioned, Saleh did encourage the Islamists as a counter to the socialism of the YSP and indeed, the GPC and Islah co-operated a great deal; Dresch and Haykel claim that, "During the 1993 elections, it was said, no less than seventy Islah candidates withdrew in favour of GPC candidates, while thirty people elected in the name of GPC were in fact Islah supporters" [15].

Islah competed in the 1993 elections with the slogan ‘The Quran and the Sunna Supersede the Constitution and Law’ clearly placing itself opposed to the more radical change to the system the YSP wanted. As mentioned in the previous chapter, Islah did run both religious and tribal candidates and al-Haqq also ran sixty-seven candidates who Detalle describes as a "veritable who’s who of sayyid families" [16]. The success of Islah was notable in that they achieved more seats than the YSP, 63 as compared to 56, however, they were a completely northern party, failing to take even a single seat in the former PDRY. Again the limited pluralism of the system worked to Saleh’s advantage because the GPC, though mainly northern, came closest to being a national party and the relatively high level of participation in the elections help legitimate his rule [17].

Islah did increase its power, it was given increased representation in the executive and Zindani gained a seat on the five man Presidential Council. After the civil war one would expect this to have been built upon but this was not the case, "Islamist aspirations were soon dashed and, indeed, it seems that those aspirations had been deliberately unleashed only to serve the interests of the Saleh regime" [18]. What became clear was that Saleh had used the Islamists and Islah for a specific purpose and while he would not attempt to push Islah out of all influence, which would have damaged his corporatist system, his actions underlined his own dominance.

Article Three was changed so Islam was made the sole source of legislation and a new article (article fifty-nine) was created to make defending religion a sacred duty. However, rather than really moving power towards Islamists, Saleh was merely placating them and recognising, as he usually does, his need to negotiate. He acted very quickly to rehabilitate the YSP, minus its chief leaders, because he did not wish the right to gain too much strength. Indeed, he dissolved the Presidential Council, thus pushing Zindani out of his influential position. The fact that Islamist elements are not unified in ideological orientation or external alliances made it harder for them to put pressure on Saleh’s regime; Islah has attempted to gloss over the differences between Wahhabis, Salafis, the Muslim Brotherhood and so on, that exist within the party. The Islamic groups did not act to create strong societal organisations against the state. They have been allowed to operate basically within the framework of Saleh’s patronage network and as Carapico writes, "The re-Islamicization of a deeply religious, homogenously Muslim culture is also government policy…like Riyadh, Sana’a uses religion to justify repression, arbitrary justice, and summary executions" [19].

Do Islamists Constitute a ‘Strong’ Element of Society?

One can argue that, in fact, Islamist forces have been encouraged in order to strengthen tribal leaders’ and, ultimately, Saleh’s position and power throughout the country. Islah are part of the corporatist system and both Zindani and, especially, Ahmar are part of Saleh’s network. Dresch and Haykel are correct in writing, "the style of politics in which Islah’s leaders took part was that associated with the President’s GPC, a style of patronage and connections rather than of ideology or of activists as vanguard of the masses" [20]. Thus, leaders of Islamist groups, like tribal leaders become somewhat disassociated from their larger constituency and more concerned with their own proximity to power or business deals. Islah’s business and tribal elements are strong forces for conservatism and preserving the status quo, stopping the party from becoming too radical. They do not represent societal resilience in a Migdalian manner because of this defence of the status quo, which Saleh too is defending.

It is not to say that Saleh has not been wary of their potential for increased power in society. Drawing on Saif’s work once more, one can note that the President did work to reduce the growing power of the fundamentalist wing of Islah. He co-opted some of their more moderate leaders into the GPC using non-merit appointments and also attempted to draw their scientific institutions into the state education system in order to reduce their influence. Similarly, militias were banned and Saleh worked to improve relations with Saudi Arabia to try to reduce support for fundamentalist groups. It would appear that Saleh has been relatively successful in containing any threat they may pose to his rule. As with tribes, action seems constricted to random acts of terrorism such as the bombing of the USS Cole or the British embassy; acts that are not a real challenge to Saleh himself.

The YSP boycotted the 1997 elections because of a rumour that the GPC and Islah had made a deal to fix the results so as to win 160 and 80 seats respectively. In actuality, the GPC won 188 and Islah only 53, less than it had achieved in 1993. In the wake of this no Islah members were given cabinet posts and al-Haqq took the Awqaf portfolio. Clearly, Saleh is secure enough to keep Islah out of direct power without fearing them mobilising against him violently. The constitutional reforms of 2001 did bring Islah into more vocal opposition in that they campaigned, unsuccessfully, for a ‘no’ vote to the reforms. However, in the main they have been a type of co-opted opposition, this is something which occurs in much of the Arab world. As Zartman writes, "the stability of contemporary Arab regimes can be partly explained by a complementarity of roles, expectations and activities between government and opposition which provides support for the polity" [21]. Indeed after the elections Saleh had commented on the apparent falling out with Islah saying, "This had only occurred because of election fever, we will be back together and our coalition with Islah will continue" [22], it is interesting though that he also stressed there would be no coalition cabinet. A ‘coalition’ in Saleh’s definition is very much a controlled one.


  1. Carapico, Sheila, Pluralism, polarization and popular politics in Yemen, in Korany, Bahgat, et al (eds.), Political Liberalisation and Democratisation in the Arab World. Volume 2, London, Lynne Rienner, 1998, p. 247

  2. Figures taken from Saif, Ahmed Abdul Kareem, A Legislature in Transition: The Parliament of the Republic of Yemen, 1990-99, PhD Thesis, University of Exeter, April 2000, p. 182

  3. Nonneman, Gerd, The Yemen Republic: from unification to civil war and beyond, in Jawad, Haifaa, The Middle East in the New World Order, London, Macmillan, 1997, p. 72

  4. For example, at unification the PDRY had 100 female judges while the YAR had none. There were women on the YSP Central Committee while the GPC placed no women in prominent positions.

  5. Sheikh al-Fadhli is a tribal leader from the former PDRY who was greatly opposed to the ‘atheist’ YSP.

  6. Quoted in Watkins, Eric, Islamism and tribalism in Yemen, in Sidahmed, Abdel Salam & Anoushiravan, Ehteshami (eds.), Islamic Fundamentalism, Oxford, Westview Press, 1996, p. 225

  7. See Carapico, Sheila, Op. Cit.

  8. Carapico, Sheila, From ballot box to battlefield: the war of the two Alis, in Middle East Report, Vol. 24, No. 5, September-October 1994, pp. 24-27, p. 25

  9. Watkins, Eric, Op. Cit., p. 212

  10. ibid., p. 220

  11. Saif, Ahmed Abdul Kareem, The Politics of Survival and the Structure of Control in the Unified Yemen, 1990-97, MA dissertation, University of Exeter, September 1997

  12. Carapico, Sheila, Civil Society in Yemen: The Political Economy of Activism in Modern Arabia, Cambridge, Cambridge University Press, 1998, p. 205

  13. Abu Taleb, Hassan, The future of Yemen after the civil war: three scenarios, in Joffe, E.G, et al (eds.), Yemen Today: Crisis and Solutions, London, Caravel Press, 1995, p. 59

  14. In the three years after unification some 150 YSP party members were murdered and the socialists laid the blame with Islah

  15. Dresch, Paul, & Haykel, Bernard, Stereotypes and political styles: Islamists and tribesfolk in Yemen, in International Journal of Middle East Studies, Vol. 27, 1995, pp 405-431, p. 407. See also Katz, Mark, Election day in Aden, in Middle East Policy, Vol. 5, No. 3, September 1997, pp. 40-50 for similar claims regarding the 1997 elections.

  16. Detalle, Renaud, The Yemeni elections up close, in Middle East Report, Vol. 23, No. 6, pp. 8-12, p.8

  17. In 1993 84% of registered voters actually voted. In 1997 this figure fell to 61%, though the percentage of those eligible to vote who were registered rose from 43% to 66%.

  18. Watkins, Eric, Op. Cit., p. 221

  19. Carapico, Sheila, Civil Society…, p. 204

  20. Dresch, Paul & Haykel, Bernard, Op. Cit., p. 406

  21. Zartman, I. William, Opposition as support of the state, in Dawisha, Adeed & Zartman, I. William (eds.), Beyond Coercion: The Durability of the Arab State, London, Croom Helm, 1988, p. 84

  22. Saleh quoted in Yemen Times, July 16-22 2001

Saleh's rule

Having discussed the nature of the relations between Saleh’s regime and the societal elements of tribes and Islamic groups, it is useful to look more closely at the actual structure of his rule. The fact that he has remained in power for almost twenty-five years means one must consider what is at the root of this survival. Here one can consider the military/security apparatus, Migdal’s politics of survival and the patronage networks already discussed. One can also discuss to what degree he has achieved legitimacy or popularity amongst citizens; according to Brian Whitaker in October 1999, "there is no doubt that Saleh enjoys wide support and that politically he is probably stronger now than at any time since he came to power in 1978" [1].

Military and Security

The military has been an important component of monarchical presidencies throughout the Arab world; Saleh, like Nasser, Asad, Sadat, Gaddafi, emerged from the military. As Hudson has written of him, "The President was first and foremost a military man whose métier was security" [2], he continues on to say, "while not as tyrannical, cruel or isolated as certain other Middle East leaders, one can imagine that by now he had acquired the habit of ruling with a firm hand" [3]. Given the lack of an institutionalised system or a real mobilisational basis, the military becomes highly important. Between 1981 and 1996 52% of the total budget was spent on the army and security services [4] and it is not purely the military that is important but also the elite Republican Guard, which is in the model of the Iraqi version and some 30000 men strong [5]. The military might that Saleh can exert does act as a deterrent to rebellion even if he is not as violent as some of his contemporaries; Saleh needs to be in control of a loyal military in order to deal with any tribal violence or activity by Islamist militias. Again, the fact that Yemen is such a heavily armed society makes this of even greater importance.

However, despite the benefits of a strong military, it may also pose a threat to an authoritarian ruler, challenge may come from within this circle. Thus, a leader such as Saleh must make sure that the military’s loyalty is to him and that it is not controlled by other elites that exist within the limited pluralism of Yemen. As Rita Brooks writes, "Ensuring political control over the military entails depriving it of both the means and the motives to challenge the regime. Leaders use a combination of inducements and safeguards to give the armed forces a vested interest in the status quo" [6]. Saleh again deals with this through patronage and kinship links. The top military posts are given to either relatives of the President or close kinsmen from Sanhan; this is the pattern across the Arab world. In the case of Yemen it develops what Dresch has called a ‘tribal-military-commercial complex’. He writes, "The state became a family business. Around the family there developed…a military-commercial complex…high-ranking army officers and a few great merchant families all had their hands in each other’s pockets" [7]. The bond between them incorporates tribal linkages and influential men from each strain are drawn into Saleh’s orbit through dispensing material favours and political sinecures. Thus, the army and security services were used to entrench the corporatist system, clearly falling into Perlmutter’s idea of corporate praetorianism.

Politics of Survival

Migdal writes, "when successfully practised, the politics of survival can lead to longevity for both regimes and particular leaders. Political stability has resulted even in the absence of what Huntington felt was the prime requisite for such stability, political institutionalisation" [8]; Saleh has used politics of survival as part of his regime-maintenance. He has used the ‘big shuffle’, which is a kind of pre-emptive act to stop others within the elite from attracting too much loyalty. This can be seen in the rearranging of his cabinet, particularly with the removal of Islah after 1997 and the handing of the Awqaf portfolio to al-Haqq. After the local elections earlier this year there was a major shuffle, as Whitaker writes, "In Yemen’s biggest government clear-out for many years, President Saleh replaced his Prime Minister and more than half the cabinet in early April" [9]. He uses this kind of shuffle to balance groups, including tribes, religious groups as well as political groups like the socialists, against each other. For example, the new Prime Minister is a southerner and could be seen as a move against growing Islamist power.

Non-merit appointments are also highly important and have been discussed above. Positions are given out to elite members of corporate groups in order to co-opt them. What this highlights is that it is not a case of the military, or tribes or Islamic groups becoming powerful as a group, but rather it is only their leaders who actually hold any influence. This system is used to draw the elites away from their constituency and ensure the patrimonial links eventually lead back to Saleh. If one is to return to Perlmutter’s criteria for limited pluralism we can see that in the Yemeni system there are crosscutting alliances and bargaining between groups, but it all occurs within an environment that is, to a large degree, controlled. In 1999 Saleh appointed a new 59 man Consultative Council and as Carapico writes, "The appointments were very wisely made to include a wide spectrum of prominent personalities" [10]. Again this highlights that Saleh is careful to include people who are high up members of society and who may not be in total agreement with him but who are more easily controlled within his polity than outside it. To quote Saif, "The Consultative Council, therefore, works both as a cushion that absorbs the frustrations of different influential groups and individuals…and as a tool to incorporate and co-opt rivals" [11]. This would cover both tribal and Islamic groupings and is part of the clientelism that has been discussed throughout this research.

‘Dirty tricks’ have also been used by the regime but these occurred more in relation to the YSP prior to 1994, though propaganda and such have been used against other groups. For example, the Yemen Times reported recently on a "Barrage of media criticism between Islah and GPC" which involved accusing one another of going soft on Israel [12]. Again the personal nature of the Yemeni political system becomes important because it is normally particular personalities that are attacked. The exact power balances that exist within Saleh’s system are hard to accurately establish due to this highly personalised nature. However, one can say that he has been highly skilled at using the plural nature of Yemeni society to develop a corpratist coalition of elites that reinforce his rule. A simple consideration of the environment in Yemen, its economics and societal make-up, would lead one to expect Saleh to be far less secure than his contemporaries but, in fact, this is not the case at all.


  1. Whitaker, Brian, New resolve?, in Middle East International, No. 611, 29/10/1999, p17

  2. Hudson, Michael, Bipolarity, rational calculation and war in Yemen, in al-Suwaidi, Jamal (ed.), The Yemeni War of 1994: Causes and Consequences, London, Saqi Books, 1995, p.28

  3. ibid., p.29

  4. see Saif, Ahmed Abdul Kareem, A Legislature in Transition: The Parliament of the Republic of Yemen 1990-99, PhD Thesis, University of Exeter, April 2000, p. 140

  5. Figures taken from Halliday, Fred, The third inter-Yemeni war, in Asian Affairs, Vol. 26, No. 2, 1995, pp. 131-140, p. 133

  6. Brooks, Rita, Political-Military Relations and the Stability of Arab Regimes, Oxford, Oxford University Press, 1998, p. 19

  7. Dresch, Paul, The tribal factor in the Yemeni crisis, in al-Suwaidi, Jamal (ed.), Op. Cit., p. 34

  8. Migdal, Joel, Strong Societies and Weak States, Princeton, Princeton University Press, 1988, pp. 225-226

  9. Whitaker, Brian, Clear-out at the top, in Middle East International, No. 648, 20/04/01, p.16

  10. Carapico, Sheila, Pluralism, polarization and popular politics in Yemen, in Korany, Bahgat, et al (eds.), Political Liberalisation and Democratization in the Arab World. Volume 2, London, Lynne Rienner, 1998, p. 262

  11. Saif, Ahmed Abdul Kareem, Op. Cit., p. 105

  12. See Yemen Times July 16-22 2001


For someone who is riding a lion Ali Abdullah Saleh is surprisingly secure. This is not only to disagree with the ancient poet but also to disagree with conceptions about a resilient and powerful society that, because of its traditional pluralism, is hard to control. What this research has sought to argue is that while Islamic conservatism and tribal loyalties are a large part of Yemeni culture and they do create diverse identities amongst Yemenis they do not represent a traditionalism that weakens the regime. In fact, tribes and Islamic groups have become part of a modern corporatist system that is headed by Saleh. Business interests, personal quests for power and internal disunity have all served to transform counter-elites from something that could challenge the state to something that works within it and, thus, reinforces it.

It is true that the Yemeni government has been ineffective in dealing with many of the country’s economic and developmental problems, but this is not simply about a weak state with weak institutional capacity. The fact that politics in the Arab world, as in much of the third world, is so related to access to resources means that effectiveness becomes a lower priority. Candidates receive support due to their proximity to the centre and, therefore, to resources; authoritarian leaders such as Saleh then use patrimonialism and client-networks to ensure that this system reinforces their own rule and they become limited in how radical or effective they can be. What this results in is not a situation where Saleh is beholden to society or to societal groups but rather that he must maintain the loyalty of specific influential people and by using them against each other and shuffling positions of prominence, he prevents the build-up of power that could challenge him.

This is to argue that the notion of a strong society contrasting with a weak state is not a real explanation of the Yemeni system. The type of monarchical presidency that is in operation has lead to a fragmented and weak society as well as an ineffective and weak state. Saleh the strongman is the key element due to the patrimonial political system. If one returns to Carapico’s point mentioned in the introduction that conventional wisdom on Yemen is that Islam and tribalism explains everything then we can see she is correct to be dismissive of this idea. Tribes and Islamic groups are not somehow unique or distinct from societal groups in other countries. They are not a traditional expression of power that the state or leader can do nothing about, they have been co-opted into the corporate coalition in the same way that sects, classes or ideological groups are in other states. It is much more useful to see these supposedly traditional elements in this light because it offers a greater explanation as to how Saleh has been able to maintain his authoritarian rule despite the expectation of a strong society.

This is not to say that Saleh does not need to play his balancing game between elements such as Islamists and tribes but rather that the idea of a strong society is overemphasised. The limited pluralism of Yemen has meant that different interests are represented and crosscutting cleavages do exist yet Saleh has been able to manipulate this to make certain that he remains secure. The strong/weak dichotomy of Migdal fails to fully interpret the Yemeni situation. True, different centres of loyalty exist outside of the state and Saleh cannot exercise total dominance over them but he does not need to. What he needs to do to survive is co-opt particular elite members of societal organisms and this is what he has done in such a way as to maintain much autonomy for his state. He allows the likes of Islah to be vocal and influential yet he ensures that they do so only in a way that does not explicitly affect him; indeed, when Islah is critical of the government, its propaganda is directed against other members of the GPC and not Saleh himself. Similarly, tribal leaders are not attacked by the state in the way that the YSP had done in the south; Saleh neither could nor would want to destroy tribal loyalties. Tribalism is again something he has brought into his corporate system in such a way as to allow it some autonomy but prevent it destabilising his regime. The distancing of elites from their constituencies has deepened the fragmentation of society in Yemen and further prevented it from being strong. The way Saleh deals with other power centres is not much different to the paths taken by, for example, Mubarak or Asad. Yemen is not an exceptional example.

To look at Migdal’s criteria of increased social control, compliance, participation and legitimation, would lead one to conclude that the state is gradually increasing its power vis-à-vis society. International observers [1] have given generally favourable reports on the parliamentary elections in Yemen and the gradual development of its political infrastructure. Citizens have participated in the elections giving Saleh more of an air of legitimacy and in general his state is achieving greater compliance. Again the difference between low-level banditry and real opposition to the state is important. Despite the seeming basis for a stronger society in Yemen due to the traditional pluralism, this has not translated into a weak or insecure position for Saleh and his government. There appears little likelihood of Saleh being challenged in anything like the foreseeable future and this is probably best for Yemen in its current situation, as Carapico writes, "the current state is a toddler, unsteady on its feet" [2].

This argument is not attempting to paint Yemen as a strong state, but if it is to achieve more infrastructural power, more institutional capacity, then it must first have stability. Saleh is not the man to bring full democracy to Yemen, his regime is too deeply intertwined with patrimonial networks and corporatist praetorianism, but he can lay the grounding for further development. The fact that political participation in Yemen is comparatively high and civil society comparatively more developed than many Arab countries does mean that the limited pluralism may be extended in the future. The fact that Islamist and tribal elites are involved in the polity creates a broader spectrum within that polity. It is unlikely that the successor to Saleh will be able to dominate over this delicate system in as complete a manner and this is where change may come. This returns one to the point made earlier that what is so significant in Yemen is the position and regime-maintenance of Saleh himself and not a competition between state on the one hand and society on the other. Saleh’s is a softer form of authoritarianism than many, but authoritarian it certainly is. He uses his inner circle, based on kinship and patronage, to permeate down through state and society in the manner of many other third world leaders. In this environment tribal and Islamic leaders have been drawn into a position where to challenge Saleh would in reality damage their own interests as well; leaders have made a broader alliance which maintains the status quo, diverging only on relatively less significant points.

In this light the monarchical presidency of Ali Abdullah Saleh is not under threat from society. Ayubi [3] has argued that Arab states have basically co-opted their societies from the outside rather than truly penetrating them and this is an accurate description of Saleh’s rule. What will become interesting is what is left of the regime after Saleh and whether he will eventually step down or take up life tenure. I was told the story of a Yemeni traveller in Syria shortly after Asad’s death asking a shopkeeper how he felt about his new leader, he was greeted with a smile and the wry comment, "Does your leader have a son?" [4]. Indeed he does. Whether Saleh has created a strong enough system that could tolerate this kind of dynasty is perhaps unlikely, however, one cannot rule it out completely. It is clientelism and a praetorian corporatism that keeps Saleh in power; his current rule is larger than state and larger than society. How well his successor can ride this lion depends on how well he can slip into position as head of these patrimonial networks.


  1. See reports by organisations such as the National Democratic Institute (NDI) and the Inter-parliamentary Union (IPU)

  2. Carapico, Sheila, Civil Society in Yemen: The Political Economy of Activism in Modern Arabia, Cambridge, Cambridge University Press, 1998, p. 206

  3. See Ayubi, Nazih, Overstating the Arab State, London, IB Tauris, 1995

  4. Anecdote relayed to the author in Sana’a August 2000


I would like to thank Dr Charles Tripp for all the advice and knowledge he has imparted, not just during the writing of this dissertation but throughout the year. A more general thanks to both SOAS and the LSE for having been lively, stimulating and chaotic places to study.

A deep personal thanks to my Aunt Samira, Uncle Abdulwahed and their family in Sana’a for the wonderful hospitality, kindness and good humour they showed during my stay in Yemen.

Most importantly, my heartfelt gratitude goes to my parents who have, over the years, provided so much support and encouragement that written acknowledgement can do no justice to it. Thank you. Finally, an added thanks to my mother for the usage of her mantra: you can do it Duffy Moon!


Books and Chapters

  • Abu Taleb, Hassan, The Future of Yemen after the Civil War: Three Scenarios, in Joffe, E.G, et al (eds.), Yemen Today: Crisis and Solutions, London, Caravel Press, 1995

  • Al-Suwaidi, Jamal (ed.), The Yemeni War of 1994: Causes and Consequences, London, Saqi Books, 1995

  • Al-Suwaidi, Jamal, Conclusion, in Al-Suwaidi, Jamal (ed.), Op. Cit.

  • Ayubi, Nazih, Overstating the Arab State, London, IB Tauris, 1995

  • Carapico, Sheila, Civil Society in Yemen: The Political Economy of Activism in Modern Arabia, Cambridge, Cambridge University Press, 1998

  • Chehabi, H.E & Linz, Juan (eds.), Sultanistic Regimes, Baltimore, Johns Hopkins University Press, 1998

  • Clapham, Christopher, Third World Politics, London, Routledge, 1995

  • Dawisha, Adeed & Zartman, I. William (eds.), Beyond Coercion: The Durability of the Arab State, London, Croom Helm, 1988

  • Dresch, Paul, The Tribal Factor in the Yemeni Crisis, in Al-Suwaidi, Jamal (ed.), Op. Cit.

  • Dresch, Paul, Tribes, Government and History in Yemen, Oxford, Oxford University Press, 1989

  • Dunbar, Charles, Internal Politics in Yemen, Recovery or Regression, in Al-Suwaidi, Jamal (ed.), Op. Cit.

  • Hudson, Michael, Bipolarity, Rational Calculation and War in Yemen, in Al-Suwaidi, Jamal (ed.), Op. Cit.

  • Jawad, Haifaa, The Middle East in the New World Order, London, Macmillan, 1997

  • Joffe, E.G, et al (eds.), Yemen Today: Crisis and Solutions, London, Caravel Press, 1995

  • Kostiner, Joseph, Yemen: The Tortuous Quest for Unity, 1990-94, London, Royal Institute of International Affairs, 1996

  • Luciani, Giacomo, The Arab State, London, Routledge, 1990

  • Migdal, Joel, Strong Societies, Weak States, Princeton, Princeton University Press, 1988

  • Migdal, Joel, et al (eds.), State Power and Social Forces, Cambridge, Cambridge University Press, 1994

  • Mundy, Martha, Domestic Government: Kinship, Community and Polity in North Yemen, London, IB Tauris, 1995

  • Nonneman, Gerd, The Yemen Republic: From Unification and Liberalism to Civil War and Beyond, in Jawad, Haifaa, Op. Cit.

  • Perlmutter, Amos, Modern Authoritarianism, New Haven, Yale University Press, 1981

  • Sidahmed, Abdel Salam & Ehteshami, Anoushiravan (eds.), Islamic Fundamentalism, Oxford, Westview Press, 1996

  • Watkins, Eric, Islamism and Tribalism in Yemen, in Sidahmed, Abdel Salam & Ehteshami, Anoushiravan, Op. Cit.

  • Whitaker, Brian, National Unity and Democracy in Yemen: A Marriage of Inconvenience, in Joffe, E.G, et al (eds.), Op. Cit.

  • Zartman, I. William, Opposition as support of the state, in Dawisha, Adeed and Zartman, I. William (eds.), Op. Cit.


  • Carapico, Sheila, Elections and Mass Politics in Yemen, in Middle East Report, Vol. 23, No. 185, November-December 1993, pp. 2-7

  • Carapico, Sheila, From Ballot Box to Battlefield: The War of the two Alis, in Middle East Report, Vol. 1994, No. 190, September-October 1994, pp 24-27

  • Carapico, Sheila, The Economic Dimension of Yemeni Unity, in Middle East Report, Vol. 23, No. 184, pp. 9-14

  • Detalle, Renaud, The Yemeni Elections Up Close, in Middle East Report, Vol. 23, No. 185, November-December 1993, pp. 8-12

  • Dresch, Paul & Haykel, Bernard, Stereotypes and Political Styles: Islamists and Tribesfolk in Yemen, in International Journal of Middle East Studies, Vol. 27, 1995, pp. 405-431

  • Dunbar, Charles, The Unification of Yemen: Process, Politics and Prospects, in Middle East Journal, Vol. 46, No. 3, Summer 1992, pp 456-476

  • Halliday, Fred, The Third Inter-Yemeni War, in Asian Affairs, Vol. 26, No. 2, 1995, pp. 131-140

  • Katz, Mark, Election Day in Aden, in Middle East Policy, Vol. 5, No. 3, September 1997, pp. 40-50

  • Saif, Ahmed Abdel Kareem, Yemeni Electoral and Party Systems, in Al- Masar Journal, Vol. 1, No. 1, Winter 2000, pp. 15-32

  • Schmitz, Chuck, Civil War in Yemen: The Price of Unity?, in Current History, No. 94, No. 588, January 1995, pp. 33-36

  • Whitaker, Brian, numerous articles, in Middle East International


  • Saif, Ahmed Abdel Kareem, The Politics of Survival and the Structure of Control in the Unified Yemen, 1990-97, MA Dissertation, University of Exeter, September 1997

  • Saif, Ahmed Abdel Kareem, A Legislature in Transition: The Parliament of the Republic of Yemen 1990-99, PhD Thesis, University of Exeter, April 2000


  • Al-Thowra

  • Gulf News

  • Yemen Times