Lewis, A 2013 Violence in Yemen: Thinking About Violence in Fragile States Beyond the Confines of Conflict and Terrorism.Stability: stability International Journal of Security & Development, 2(1): 13, pp. 1-22, DOI: http://dx.doi.org/10.5334/sta.az

ARTICLE Violence in Yemen: Thinking About Violence in Fragile States Beyond the Confines of Conflict and Terrorism Dr Alexandra Lewis*

This article examines the different forms of criminal violence that affect fragile states, with special reference to Yemen. The article is particularly interested in analysing the relationship between violent offending with no clear political motive, underdevelopment and conflict. It does so by conducting an in-depth evaluation of conflict and crime in Yemen, using publically accessible data to suggest new ways of understanding violent criminal behaviour in Yemen and elsewhere. This article is written in response to a prioritisation of political violence, insurgency and terrorism in international development and stabilisation strategies, which has emerged along- side the broad securitisation of international aid. Common forms of criminal violence have been overlooked in a number of fragile contexts, as they have been in Yemen. In light of rising levels of insecurity, resulting from poor relationships between the state and its citizens, there is a need to re-evaluate this unstated omission if the new Yemeni Government is to gain increased legitimacy by being seen to prioritise the protection of its citizens.

Introduction suited to ‘21st century violence’. Due to ‘suc- It is argued in the 2011 World Development cesses in reducing interstate war, the remain- Report that the ‘global system’ of interna- ing forms of conflict and violence do not fit tional aid is centred on a ‘paradigm of con- neatly either into “war” or “peace,” or into flict’, whereby both national and interna- “criminal violence” or “political violence”’ tional actors – in development, diplomacy (World Bank 2011: 2). In particular, a prior- and humanitarianism – have become chiefly itisation of political violence, insurgency and concerned with providing states with the terrorism in international development and means and capabilities to overcome civil war stabilisation strategies has meant that more through increased prosperity and dispute common forms of criminal violence have resolution (World Bank 2011: 2). This view, been overlooked in a number of fragile con- writes the World Bank, prioritises violence texts, as they have been in Yemen. and conflict as an inherent cause of under- In light of Yemen’s on-going political cri- development, but does not take into account ses, the prioritisation of counter-terrorism the reality that such an approach is no longer and counter-insurgency programming is not surprising. Yemen is a fragile state, in accord- ance with Graziella Bertocchi and Andrea * Research Fellow, Post-war Reconstruction and Development Unit, University of York, UK Guerzoni’s definition of the term (2011: 2), [email protected] where fragile states are categorised by an Art. 13, page 2 of 22 Lewis: Violence in Yemen

‘inability to provide basic services and meet and to trace their connections to conflict and vital needs, unstable and weak governance, insecurity. The article also aims to improve persistent and extreme poverty, lack of terri- the public availability of information on torial control, and high propensity to conflict crime patterns in Yemen in order to encour- and civil war’. Without effective peace-build- age a debate on international engagement in ing policies and an effective peace-keeping fragile states that looks beyond the confines system, Yemen, in the aftermath of the Arab of conflict and terrorism. Spring, will be at heightened risk of state Due to the pervasive nature of violence in failure. This could have potentially dras- fragile states2 (which are often war-affected), tic repercussions for the region, keeping there is a tendency amongst analysts not to in mind that the country is now home to a delve too deeply into criminological analy- resurgent Al Qaeda in the Arab Penninsula ses in contexts where crime is viewed as a (AQAP) movement. ‘natural’ side-effect of underdevelopment. The 2011 Arab Spring and the 2012 Rev- The relationship between crime and conflict olution of Institutions have pushed the appears, superficially, to be straightforward, Yemeni Government and the international with crime emerging as a bi-product of the community to begin planning for a severe social chaos generated by political instabil- restructuring of military, police and judicial ity. However, the logical extension of this services in the country. However, there is argument – that the international commu- a danger that a lack of understanding of – nity needs therefore to intervene to restore and information about – more mainstream political stability in order to solve the prob- criminal violence in Yemen could lead to the lem of crime – is fundamentally flawed. One side-lining of spending on conventional law has only to look at cases like Iraq, El Salva- enforcement needs. The British Government dor, Tajikistan and Nicaragua to see that – which is taking the lead in police train- serious offending patterns not dealt with ing in Yemen as part of a broader European at peak points of conflict will continue to Union commitment to post-2013 security have a costly impact upon public safety long sector reform – asserts that restructured after peace-building strategies have been police services in the country will likely take initiated, with crime (especially organised on a ‘paramilitary’ identity and that Yemeni crime) often overtaking conflict as a lead- police will be tasked with reinstating terri- ing cause of insecurity. A lack of compara- torial control of their state.1 They will likely tive and case-by-case investigations means remain a highly politicised institution, oper- that the challenge of violent crime in fragile ating according to a ‘statist’ or ‘state-centric’, states remains poorly understood, leading to rather than a ‘protectionist’ or ‘rights-based’, increased difficulties for policy-makers work- narrative of security, in which criminality is ing in security sector reform: law enforce- secondary to other strategic concerns. It is ment is side-lined by militarised policing unclear what the implications of this are for strategies, and the impact of crime on secu- the potential escalation of criminal violence, rity is overlooked. or how resulting criminality might feed back Meanwhile, low rates of crime detection in into political insecurity. fragile states indicate that the inflated per- In light of Yemen’s current transition pro- centages of violent offending made visible cess and upcoming national dialogue, this by official records3 are likely to be masking paper’s primary purpose is to highlight the an even greater volume of undetected vio- importance of crime in security by analysing lent offences. While this is often also true the relationship between criminal violence of stable countries, low rates of birth regis- and political violence. The article’s main tration, lack of access to basic services, low function is to point readers towards forms school attendance, high unemployment and of violence that are commonly overlooked mass displacement further impact the abil- Lewis: Violence in Yemen Art. 13, page 3 of 22 ity of fragile states to monitor their citizens. Conflict and Security in Yemen This leads to a lack of detection of murders Yemen is currently experiencing no less than and disappearances, not to mention other four distinct security crises, meaning that crimes. Thus, the Organisation for Economic the Yemeni Government has little effective Co-operation and Development (OECD) control over the vast majority of its territory. notes that ‘homicidal violence and violent The main political challenges that the state crime’ as well as ‘armed violence dispropor- faces today are: an on-going conflict against tionately [affect] low … income countries’ localised combatant groups in the far North; (2009: 28–9). This is a reality that needs to rising separatist sentiment in the South; a be seen to be addressed, if fragile states are growing terrorist insurgency from within to be perceived by constituents as demon- Yemen’s borders; and threats to the state’s strating concern for the safety and welfare very survival in the wake of the 2011 Arab of their citizens: but a tendency by fragile Spring. War and conflict are products and states and donors to neglect criminality in causes of continuing fragility. They contrib- favour of stabilisation has failed to lead to ute to Yemen’s patterns of criminality by: increased legitimacy in many contexts, par- disrupting the rule of law; increasing social ticularly in Yemen where the state’s security chaos; and, diverting security services into priorities are not seen as relevant to the counter-insurgency and counter-terrorism needs of communities4. strategies. Due to the reciprocal relationship The biggest challenge in studying these between crime and conflict, it is likely that issues is definitional: it is extremely diffi- this side-lining of law enforcement in Yemen cult, even in countries that are not affected is unhelpful to restoring security and stabil- by significant obstacles preventing effective ity in the country. data collection, to separate out incidents Hostilities in the far north of Yemen can of criminality from incidents of conflict- be traced to economic competition between related violence. North and South Yemen before and after the The definitional issue is addressed here in formal unification of the country in 1990 part through the selection of a fragile state (which culminated in the 1994 Civil War case study in which administrative processes and was followed immediately by a need to are already in place within the existing judi- consolidate a new government leadership). cial system to sort violent actors according For a prolonged space in Yemen’s history, to criminal or politically motivated offences. peripheral governorates were marginalised Yemeni courts disaggregate criminal, politi- by state development programmes in favour cal and morality offences in statistical data5. of building a strong capital city with a solid This renders statistical analyses of Yemeni administrative base in Sana’a (Boucek 2010). crime rates easier by listing casualties of war The resulting situation of chronic underde- or terrorism as victims of political violence. velopment that emerged in the far North, As evidenced in the next section, it has long particularly in the areas that border the King- been to the benefit of the Yemeni Govern- dom of Saudi Arabia, contributed to overall ment to link as many deaths as possible to increases in levels of hostility by local com- political actions in order to justify its inter- munities towards President Saleh’s Govern- nal military policies and bids for foreign ment (disbanded in 2011). When combined military aid (especially from Western powers with historic grievances, the formation of and Saudi Arabia). Analysts are therefore left new identities and increased population with the reasonable assumption that remain- movement, all of which came to the fore in ing records of violent crime offer an accurate the late 1990s, hostilities sparked mass pro- portrayal of offending behaviour, even if that tests, arbitrary arrests in 2004, and a wave of portrayal may be severely limited by a lack of military clashes resulting from a refusal by crime detection capacity. protest leaders to accompany state escorts to Art. 13, page 4 of 22 Lewis: Violence in Yemen prison (International Crisis Group 2009). In Southern resources – especially oil and gas 2004, emerging groups of non-state combat- – were being siphoned by Sana’a to promote ants included the Houthis and the Believing Northern interests. The gradual reduction of Youth. Since then, six consecutive wars have the ruling Yemeni Socialist Party’s author- enveloped Sa’ada and its surrounding gover- ity and influence across the country (which norates – Hajja, Amran and Al Jawf. Accord- favoured the South) and growing dissatisfac- ing to Christopher Boucek, over 250,000 tion with the new status quo, eventually led people have been displaced by the Sa’ada to the 1994 Civil War for separatism. South- wars and, while ‘[t]here is no good data on ern forces were quickly overwhelmed by the casualties, … estimates of the number killed North in a brutal campaign that devastated range from several hundred to several thou- their infrastructure (Johnsen 2013: 45). sand’ (2010: 1). The aftermath of that war has not yet been In 2013, the main implication of the fully analysed. Following the unquestionable Sa’ada wars for security is that they are as victory of the North of Yemen over the South, yet unresolved and risk being reignited. President Saleh’s regime set about a process Resistance to arrest and outside interven- of purging Southern leaders from key posi- tion, which sparked hostilities in 2004, tions in the South and replacing them with has far-reaching implications for policing, Northern officials (Johnsen 2013: 43). While which has traditionally been unwelcome President Saleh succeeded – by integrat- in Yemen’s largely inaccessible peripheries. ing the Southern and Northern armies - in Conceptions of crime and justice also differ crushing the ability of the former Southern between the centre and the peripheries, so leadership to mobilise any significant forces that official attempts at justice administra- against the Sana’a Government, he failed tion are prone to igniting violence between to effectively address Southern grievances tribes and police services6. Meanwhile, cases (Human Rights Watch 2009). This led to pro- in which citizens legitimately need state pro- longed and continuing hostility towards his tection from abusive local tribal leaders are regime after 1994. Southern Yemenis have easily overlooked7. found few outlets in mainstream politics The 2011 Arab Spring allowed the North- for voicing their discontent, pushing them ern peripheries to establish an almost fed- to protest frequently against the Yemeni eral self-governance system. Concerns for state. Ever fearful of budding revolutions, the future of the region emerge from the the Government of Yemen has responded to as yet undetermined position of President these protests with brutal reprisals, inadvert- Saleh’s successor, President Hadi, towards ently sowing the seeds of what has become a the Northern governorates and from doubts flourishing secessionist movement (Human as to whether or not future governments will Rights Watch 2009). Struggles between the allow the region to maintain its new sense of two continue to this day. autonomy. Other dangers stem from the real- In 2011, the Southern Mobility Movement, ity that the severe underdevelopment of the also known as al-Hirak al-Janubi, put aside far North has yet to be reversed. their secessionist narrative in order to unite The 1990 unification did not bring much with other Arab Spring protesters in favour material benefit to former South Yemeni of removing President Saleh from power. territories either. In the aftermath of their Though peaceful overall, al-Hirak al-Janubi democratisation, Southern Socialist-Marxist is now increasingly – and perhaps unfairly – government structures were slowly disman- linked to a growing trend towards radicalisa- tled, services were privatised and subsidised tion (Phillips 2010). This trend is associated healthcare and education were brought with focused counter-terrorism strategies to an end. These cuts reinforced rumours in the South, which have angered the local spreading among the local population that population, putting peaceful communities Lewis: Violence in Yemen Art. 13, page 5 of 22 in harm’s way through shelling and other latter are characterised by a prevalence of military campaigns. Crucially, there is now under-reporting and retributive violence – the potential for Southern communities to phenomena that can be associated with law turn violent if the post-Saleh transition pro- enforcement and justice deficits that are a cess fails to grant the South increased politi- direct result of the institutional marginalisa- cal representation in the long term. Rising tion of the issue of crime. insecurity and unrest in the South in 2013 Due to its pre-existing conflicts, Yemen are commonly linked to a lack of resolution has been significantly affected by the Arab of ‘the Southern issue’.8 Spring. Protests began in Yemen between From a law enforcement perspective, there January and February 2011, when thou- is a need to tread carefully in the South with sands of demonstrators took to the streets, regards to police reformation and security demanding a change of leadership; the sector reform. The precarious position of removal of President Saleh as the head of the policing services in Yemeni society is influ- Yemeni Government was a central and non- enced by its relatively brief history, whereby negotiable demand for most participants protracted colonization meant that North (Lewis 2012). This goal was achieved only in and South Yemen established their own 2012, but not without first fracturing Yem- security systems after the independence of en’s army and leaving the new Government both countries in the 1960s. These police in a state of crisis. systems differed remarkably between the The protests, which pitted citizens against two countries: ‘The Aden Police’ maintained law enforcement personnel, have had a a clear division between the three services negative impact on perceptions of police of Armed, Riot and Security Police in the legitimacy. Many police and military person- South; and the Northern police, according nel and their families were under threat of to Nathan Brown et al, emerged as a much random attacks and street violence in 2012, more passive, ‘less centralized’ and ‘less particularly by young people and emerg- pervasive’ institution that aimed to strike ing gang structures.9 Independent surveys a balance between state and tribal author- conducted by the author in 2011 and 2012 ity (2010: 6). Thus, a far stronger tradition further revealed that citizens in Sana’a of tribal autonomy and self-administration and elsewhere witnessed a rise in property evolved in the North than in the South of damage (particularly against public build- the country, where a prioritization of tribal ings), hooliganism and clashes between affiliations was categorized by the state as law enforcement services and communities backwards. However, a long tradition of throughout this period. A recent survey by police brutality has meant that Southern the Yemen Polling Centre (2013) indicates communities are quick to voice suspicions that the situation seems to be improving in of police involvement in their affairs. Brown 2013 and that Yemenis are now open to the et al note that, when competing policing idea of increased policing at a national level. institutions and values were merged under However, much work remains to be done if the management of the Ministry of Interior Yemeni communities are to come to trust upon Yemen’s unification in 1990, ‘South- and rely upon their police and security ser- erners … complained about this new force, vices in future. as well as the army, which they held directly Finally, as a buffer between the Middle East responsible for increased banditry’ (2010: and the Horn of Africa, Yemen is affected by 6). Varying perceptions of the legitimacy of a number of strategic issues, including an law enforcement services in different gover- on-going regional struggle between Shia and norates have fed directly into both the way Sunni domination (which can be considered in which conflicts have evolved over time in more of a politically-motivated struggle than Yemen and into patterns of criminality. The a religious one) (ICG 2010), increased Islamic Art. 13, page 6 of 22 Lewis: Violence in Yemen radicalisation and terrorism, and rising levels the country’s severe lack of capacity and of organised crime and piracy. Most famously, identity as a fragile state. Mark Shaw notes Yemen is the home of the late Osama bin that sovereign states have historically been Laden, founder of the Al Qaeda terrorist defined as enclosed or ‘bordered’ political movement, and the new operational centre organisations with exclusive ownership over of the Al Qaeda in the Arab Peninsula (AQAP) legitimate (or simply internally legalised) movement – a core component of the overall violence within their territories (1997: 500). organisation (Johnsen 2013: xiv-xv). However, as Graham Brown and Frances In practice, the threat of AQAP and its Stewart note, fragile states are defined, con- affiliate, Ansar al-Sharia, has acted as one of versely, as suffering at once from failures in the main determining factors for continued authority, services and legitimacy10 (2010: American and Saudi Arabian support for the 10). As a fragile state, Yemen falls short of Yemeni government. This support proved Shaw’s governance criteria, whereby author- crucial for former President Saleh’s regime ity and legitimacy failures inhibit its Govern- in determining domestic military policy, par- ment from fully and effectively maintaining ticularly in relation to the war against the its territory in accordance with a bordered Houthis, and is currently proving equally conception of statehood. The legitimacy and important for President Hadi. legality of violence committed by the Yemeni America has now been intervening in Government within its territory is questioned Yemen for over a decade. In November 2002, by its citizens and the international commu- an American missile strike that resulted in nity (Human Rights Watch 2010). Insurgents the death of Abu Ali al Harithi, then-leader of and foreign powers continue to use violence Al Qaeda’s Yemen offshoot, effectively crip- to bid for control of governorates, with and pled the movement. However, after a prison without the Government’s approval (John- break in February of 2006 allowed Nasser al sen 2013: 49). Ultimately, Yemen’s main Wahayshi and Qasim al Raymi to reform the security challenges are not enclosed within organisation, AQAP once again grew to con- its territory; piracy, terrorism and organised stitute a significant threat in Yemen (Phillips crime are cross-border phenomena and tribal 2010). These on-going security challenges divisions also fail to recognise official state or have led to the continuous diversion of governorate lines (Dresch 1993). Government funding to military needs, and Yemen’s unique socio-political reality the substitution of traditional community reflects its unusual evolution. The United policing roles with new counter-insurgency Nations Development Project (2010b: 11) ones. These strategies are likely to continue writes that: to dominate security sector reform packages in 2014 and beyond, leading to the side- Overlaying a modern state upon Yem- lining of conventional law enforcement pro- en’s traditional governance system grammes, which, as it will be argued, stems has proved difficult, and both state from a lack of appreciation for the simulta- formation and nation building re- neous security threat that crime poses in the main works-in-progress. Meanwhile, country. The international community is not to ensure its survival, the govern- currently interested in funding the develop- ment has created informal political ment of law enforcement in Yemen that is alliances with traditional shaiks, reli- geared specially towards the control of apo- gious leaders, and powerful interest litical violence and crime. groups through intensive patronage The either/or attitude to addressing con- networks outside the formal state flict and crime that results implicitly from structures – which has given rise to policy interventions in Yemen emerges from the so-called ‘parallel state’. Lewis: Violence in Yemen Art. 13, page 7 of 22

Figure 1: Map of Total Reported Offences in 2009, Extrapolated from Yemen’s Security and Justice Statistics

The existence of the Yemeni parallel state ture necessary to detect and process offences has worked at once to secure the Govern- through official channels. Associated dis- ment’s position of authority prior to 2011 crepancies in data trends are evident from and to severely limit the extent of its power. the fact that certain peripheral governorates While the parallel state multiplies the num- in Yemen, such as Al Jawf, have consistently ber of informal institutions that might be shown a significantly lower rate of report- called upon to enhance the Government’s ing than their neighbours. However, while decision-implementing capacity, it actu- parts of Yemen (particularly in the capital, ally serves to undermine official structures. where reporting is substantially higher than Many manifestations of Yemen’s fragility are elsewhere) may be said to provide a more thus intrinsically connected to weak govern- detailed survey area than Al Jawf, data avail- ance, whereby the state remains incapable ability in Yemen is relatively high considering of monopolising control over legitimate vio- its severe fragility. lence and lacks the institutional capacity to Government statistics may provide une- exercise significant positive change in the ven coverage of crime rates across gover- lives of its citizens. These contribute directly norates, but they nonetheless provide a to both crime and conflict, but only conflict continuous and consistent portrait of each has been analysed thoroughly be policy mak- governorate over time, allowing for the ers and researchers investigating insecurity extrapolation of country-wide and area-spe- in the country. cific trends. These and other statistics11 have been used in this analysis. Crime in Yemen Statistical records in Yemen are unreliable, Criminal Violence with large areas of the country isolated from Yemen offers a fairly typical example of the state by conflict or natural disaster, and crime and violent offending within a fragile other rural settings lacking the infrastruc- state context. Its statistically recorded rate Art. 13, page 8 of 22 Lewis: Violence in Yemen of total crime is very low, but its homicide statistics made available by the Central Sta- rate is high, as is its rate of violent offending. tistical Organisation and the Ministry of Jus- Yemen registered a rate of 0.0336276 homi- tice do not provide a breakdown of victim cides per 1,000 people in 2004, placing it at demographics and are not very helpful in the 27th position out of 62 countries with determining the extent of gender-based vio- available data (UNODC 2005). In 2008, this lence in the country. The extremely low rate number increased to approximately 0.04, of social mobility among women and girls in according to the Human Development Index, Yemen means that those who are exposed indicating a steady and consistent increase to violence or abuse generally do not seek (UNDP 2010a). assistance from the Government or its law Yemen now exhibits a high rate of crimi- enforcement services. Official rape statistics nal violence, particularly against women and therefore show sexual offences to be virtu- children. Uprooted communities seem to ally non-existent in the country, with only be especially vulnerable; 67.5% of displaced 0.004 offences committed per 1000 people people under the age of 18 show evidence of (UNODC 2005). being subjected to domestic violence (UNCT Actual numbers are likely to be much 2010: 5). These rates of abuse are likely con- higher. Violence against female minors is sistent with elevated levels of family stress also reported to be on the increase. 1256 caused by displacement, uncertainty about girls under the age of 18 had been recorded the future, lack of physical stability and as falling victim to gender-based violence in increased vulnerability due to loss of employ- 2008, as compared to only 167 girls in 2007 ment and the break-up of pre-existing com- (Women’s National Committee 2008: 60). Of munity structures.12 the violent crimes committed against young Violence against children in schools and girls, battery, murder and attempted rape on the streets in Yemen has additionally were listed as the most prominent. been proven to be common, where aggra- vating factors have included: a lack of safe Weapons Proliferation spaces for young people (EDC 2008); poor Yemen is considered one of the most heav- community capacities for the monitoring of ily armed countries in the world, both a children; and, poverty. result of the important cultural symbolism Adult men are found to be at risk of griev- of weapons within Yemeni society and of the ous bodily harm and fatality, often as a result country’s very low border control capacity, of revenge killings or violent disagreements which allows for the relatively easy traffick- over natural resources. The Small Arms Sur- ing of small arms into and out of its borders. vey (2010) estimates that arguments relat- Curved blades called janbiya that are worn ing specifically to water and land currently on the belt have traditionally symbolised account for approximately 4,000 deaths per male honour and tribal affiliation, especially year in rural Yemen alone. Poor reliance on in the North. However, these are increasingly official state justice mechanisms, particularly being accompanied by illegal firearms. in the peripheries, means that people tend In 2003, Derek B. Miller wrote that there to take justice into their own hands, turning were approximately 6–9 million small arms personal confrontations into family, tribal in the country – contrary to common esti- or religious ones that affect a much larger mates, which stated that the number was 50 populous. These confrontations carry clear million (10). Even if Miller’s lower estimate implications for continuing conflicts, as well is to be adopted, dangers to the civilian pop- as for escalating criminal violence, but are ulation stem from the reality that many of neglected by statist conceptions of security. these weapons have yet to be accounted for Women in Yemen are equally, if not addi- by the Government. Due to the importance tionally, vulnerable. Governmental crime of weapons in Yemeni customs and cultural Lewis: Violence in Yemen Art. 13, page 9 of 22 practices, any official efforts to enforce con- upon in Yemeni society, making it more dif- trol over weapons distribution and posses- ficult for the authorities to arrest women, par- sion have in the past been met with great ticularly in relation to political or other kinds hostility in certain governorates; those poli- of activism’ (2012: 17). cies that are passed successfully are made There is also a higher general tendency redundant by a lack of implementation towards female offending among young capacity. These trends are aggravated by on- offenders in Yemen than among the general going conflicts and political crises, which population13. This trend could be associ- ensure the continued demand for weapons ated with improved social mobility among and ammunitions trafficking. younger generations of women in Yemen. Young offending among women is predomi- Offender Demographics nantly linked to adultery-related offences, Within this context of widespread crimi- violations of immigration laws and cases of nal violence and the escalation of violence light assault. This leaves boys and young men through the availability of weapons, there as the main body of violent offenders in the has emerged a growing trend towards violent under 18 population. behaviour among children and young peo- ple under the age of eighteen, attributed to Young Offending the socialisation of children to violence. The Young offenders comprise a very small per- United Nations Country Team argues that centage of statistically recorded crime rates the use of child combatants in the Northern in Yemen. However, the number of docu- conflicts of Yemen is creating a generation of mented offences committed by young peo- young people ‘who perceive violent oppor- ple under the age of 18 has steadily increased tunism to be a normal aspect of life’ (2010: since 2003, both in quantity and as a per- 25). The same link can be made for those centage of total offences (CSO 2004 - 2010). children who are now growing up in areas of Although it is not possible to fully document high criminal violence and who have not been the extent of the young offending problem exposed to war. The visible effect of this grow- in Yemen, it is possible to extrapolate a pic- ing culture of violence are an increased occur- ture of offending patterns among children rence of young offending and a growing rate and young people. Detected and recognised of violent offending among young people. young offenders can be taken as a represent- Most offenders in Yemen are identified and ative sample group of the greater undetected processed as male adults, above the age of 18. or misclassified whole. Yet, many of these offenders include young There is a strong tendency among young males who are misclassified due to a lack of offenders towards violent crime (such as appropriate documentation to prove their physical offences against the person), which entitlement to processing under specialised represented 53.81% of offences commit- youth justice systems. Overall, the demo- ted by young people in 2009. Violent crime graphic composition of offenders is extremely accounted for 48.40% of all offences com- gender weighted: Yemeni men represent mitted in 2009, indicating a disproportional 93.64% of all offenders (Walmsey 2007). trend towards violent behaviour among Women, who are hugely under-represented those under the age of 18 (CSO 2009). Fig- in crime statistics, are most commonly arrested ures relating to murder, attempted murder for adultery-related offences and public inde- and manslaughter, which accounted for cency. Statistics may be skewed, however, by a 10.08% of offences committed by young reported tendency for men and young boys to offenders (CSO 2009), are staggering, par- go to prison in place of their female relatives. ticularly in light of the Yemeni Government’s Amnesty International writes that: ‘Generally, separation of criminal and so-termed ‘politi- arrest and detention of women is frowned cal’ violence in official data. A significant por- Art. 13, page 10 of 22 Lewis: Violence in Yemen

Figure 2: Graph of Recorded Juvenile Offending Cases, Extrapolated from Yemen’s Security and Justice statistics for 2009 (CSO 2010) tion of the 363 murder, attempted murder stability and development’ as ‘illicit traf- and manslaughter cases detected in 2009 ficking (particularly of drugs)’ and ‘criminal are not considered to be related to on-going networks (including terrorism and its financ- conflicts in the country (CSO 2009). With ing)’. However, many of Yemen’s weapons are much violent crime being situated around left-overs of previous conflicts. Its primary conflict-affected governorates, and especially source of narcotics consumption – qat – is around the Northern peripheries, violent also legal. Therefore, much of the evidence young offending might be read as a legacy linking organised crime to mainstream crim- of prolonged periods of warfare in Yemen. inogenic commodities in Yemen is circum- It feeds continuously into political violence, stantial. Organised crime might be linked to with severe implications for security. the trafficking of specialised weaponry (such as valuable antique guns) and alcohol. While Organised Crime and Corruption some alcohol, including Baladi (a vodka-like Yemen also suffers from organised crime, a liqueur), is brewed in Yemen, most is smug- bi-product of nation-wide underdevelop- gled into the country from Ethiopia and Dji- ment and emerging networks of patron- bouti (Murphy, 2010). age. The UNODC (2010) identifies ‘the two Other forms of organised crime include predominant threats to Yemen’s security, banditry and kidnapping (CSO 2009), Lewis: Violence in Yemen Art. 13, page 11 of 22

whereby a relatively small group of peo- The Relationship Between Yemeni Crime ple, or even an entire community, is able to Statistics, Underdevelopment and Conflict win quick financial profits by harassing and Statistics made available by the CSO cover intimidating other communities and indi- the period of 2003 to 2010. In this same viduals. With long roads, large desert and period, the Government of Yemen has made mountain terrains, and low levels of police significant efforts to increase its own opera- and law enforcement outside the major cit- tional capacity (MOPIC 2010). However, fol- ies, Yemen lends itself well to such criminal lowing the initiation of hostilities in the opportunism. Organised crime in Yemen is north of Yemen in 2004, and the breakdown mainly linked to human smuggling and traf- of the security situation across the country, ficking14 and to the smuggling of agricultural the Government’s capacity for law enforce- and fishing produce15 – including qat. Diffi- ment has diminished in many governorates cult mountainous terrain in the north is an and much of its operational budget has been obstacle to Government intervention and siphoned towards military priorities. has encouraged local combatants and insur- Despite a decline in crime detection capac- gents familiar with the landscape to adopt ities, Yemeni statistics have shown a consist- guerrilla tactics, so that conflict and banditry ent increase in rates of offending, which rose overlap in difficult-to-traverse areas. from 24,406 offences in 2003 to 40,090 in The challenges of organised crime and 2010 (CSO 2004, 2010). Violent non-political conflict in Yemen are connected to the pre- offences comprised 42.08% of all crime in dominance of corruption16. For many Yem- 2003, and 48.40% in 2010, showing a paral- eni citizens, corruption is a daily reality (TI lel increase in severity of registered offences. 2010) that has resulted in the emergence of Overall decreases in budgeting capacity a highly contested parallel state, based on and security indicate that this rising crime tribal affiliations and the entrenchment of rate is connected to broader social condi- patronage networks. This has acted as an tions, rather than to increased reporting, of important impediment for reform and is a which there has been no evidence17. Dennis leading cause of bureaucratic stalling in the Jay Kenney summarises that ‘Yemenis are at dispersal of bilateral aid and the delivery of somewhat greater risk of most crimes – espe- development projects. One explanation for cially crimes of violence – than are citizens of the parallel state is that the Government’s many other countries throughout the world’ limited ability to exercise its authority has (2008: 10). From 2011 to 2012, widespread traditionally fostered reliance upon patron- protests further destabilised security and, age systems and tribal allegiances (Econo- in 2013, deteriorating levels of public safety mist Intelligence Unit 2010). Another reason have been linked to uncertainties about the is that Yemen’s very poor law enforcement future, in light of Yemen’s political transi- system allows for criminal opportunism, tion18. Violent crime against foreigners is also while the poor salaries of low-level public on the increase, with Westerners particularly officials encourage bribery. Those commu- at risk of targeted violence19. nities unable to access benefits awarded Crime is distributed across the country through Government patronage are partic- in urban, rather than rural, areas20 – as is ularly likely to reject state authority, either to be reasonably expected given popula- falling into mainstream patterns of offend- tion density and police capacities in differ- ing or contributing to on-going conflicts. ent regions. Yemen’s largest cities – Sana’a, Corruption facilitates the formation of the Aden, Taizz, Hudaydah, Mukalla, Ibb and give-and-take relationships that are neces- Dhamar – are located in Sana’a, Aden, Al sary for organised crime to flourish (Mar- Hudaydah, Hadramaut, Ibb and Dhamar tínez 2007). governorates, respectively; these gover- Art. 13, page 12 of 22 Lewis: Violence in Yemen

Figure 3: Violent Crime Per Governorate in 2009 as a Percentage of Total Recorded Offences in Yemen’s Security and Justice statistics (CSO 2010) norates in turn demonstrate the highest Despite discrepancies, violent crime across annual rates of recorded crime. the country can be further correlated with Violent crime is particularly concen- the aggregate percentage of urban and rural trated in those governorates that have been poverty of each governorate, as the dem- most strongly affected by conflict, includ- onstrated in the table below, extrapolated ing Sa’ada, Al Jawf, Amran and Hajjah. The from Yemen’s CSO (2010) Statistical Year- greatest discrepancy in these findings is book for 2009. that Hadramaut’s recorded rate of violent Heightened levels of violence in poorer crime remains considerably lower than governorates indicate that both crime and those of more central administrative areas, conflict continue to be connected to broader despite containing a large city. This could be themes of underdevelopment and local dis- explained by diminished law enforcement satisfaction with Government development capacities and poorer data coverage. Had- priorities: this has proved particularly prob- ramaut remains one of the poorest gover- lematic in the Sa’ada-Hajja-Amran-Al Jawf norates in the country (CSO 2010), with very area, where a rejection of state legitimacy has low levels of policing. also led to protracted conflict. Other factors should likewise be taken into account: crime and conflict are higher The Interactions Between Crime, in certain areas of the country due to per- Conflict and Security in Yemen sistent disputes over arable land. The perfo- Yemen is affected by a wide range of criminal ration of tribal structures and levels of local activity, much of which is extremely violent. education or social organisation might also Yet patterns of crime tend to be overlooked be relevant, with some tribes adopting more and under-researched by national, regional violent strategies for negotiating their sur- and international powers who prioritise roundings than others, and some education stabilisation, counter-insurgency and coun- and local structures adopting strong coun- ter-terrorism interventions. This attitude is ter-violence programming. unhelpful, undermining the protection of Lewis: Violence in Yemen Art. 13, page 13 of 22

Offences Against the Person as a Per- Aggregate Percentage (%) of Urban District centage (%) of Total Registered Crime and Rural Poor Ibb 50.78 23.086 – 34.691 Abyan 48.91 37.136 – 48.378 Sana’a City 45.07 7.976 – 17.504 Al Baida 80.87 51.431 – 70.395 Taiz 48.86 37.136 – 48.378 Al Jawf 78.23 37.136 – 48.378 Hajjah 61.6 51.431 – 70.395 Al Hudaydah 38.81 37.136 – 48.378 Hadramout 24.64 37.136 – 48.378 Dhamar 55.74 23.086 – 34.691 Shabwah 35.58 51.431 – 70.395 Sa’ada 72.35 – Sana’a 56.2 23.086 – 34.691 Aden 31.48 23.086 – 34.691 Lahij 39.86 37.136 – 48.378 Mareb 53.16 37.136 – 48.378 Al Mahweet 64.71 23.086 – 34.691 Al Maharah 32 7.976 – 17.504 Amran 85.79 51.431 – 70.395 Al Daleh 56.62 23.086 – 34.691 Reymah 65.12 23.086 – 34.691 Table 1: The Distribution by Governorate of Aggregate Urban and Rural Poverty Rates and Offences Against the Person as a Percentage of Total Recorded Crime in 2009.

Yemeni civilians, allowing crime to spiral between people and the state. These obser- out of control, and paradoxically contribut- vations lead to questions as to how violence ing to the regeneration of the very political and insecurity in fragile states should be violence patterns that militarised security addressed, implying that only a holistic strat- strategies strive to address in the country. egy that is aware of both areas might succeed While crime and conflict are separate issues in addressing either one. and stem from a variety of causes, they are Criminality and conflict can be relatively aggravated by similar risk factors and feed easily correlated with one another through into one another: for example, they can both root causes, and both are affected by limita- be affected by a lack of community cohesion, tions in the state’s capacity and by specifici- by a lack of livelihoods and education oppor- ties of the development context. Both are also tunities, by low policing/security capacities, influenced by the prevalence of weapons, or by a prevalence of retributive violence21. where “weapon characteristics” are “relevant Furthermore, criminal violence is easily to the degree of harm inflicted” by crime and misinterpreted as political violence and can conflict (Roach and Pease 2011). Crime and illicit political responses, leading to continu- conflict are furthermore incredibly difficult to ous exchanges of violence between tribes or investigate individually, because they overlap communities, or sparking low-level recurring and aggravate each other. When they are ana- violence that triggers new rounds of conflict lysed in tandem in Yemen, they are found to Art. 13, page 14 of 22 Lewis: Violence in Yemen intersect at various different points, at which state, though it also incorporates economi- the independence of crime and conflict as cally motivated violence aimed at achieving a individual phenomena is heavily obscured. tribal objective. This can take the form of ban- There are six key points at which the union ditry, kidnapping, racketeering, robbery or of crime and violence is particularly strong. burglary, as long as the economic outputs of These points include: (1) tribal violence (that these crimes are intended to benefit a tribal is, violence committed to achieve a physical unit or sub-unit23. Such actions may not be benefit for a tribe); (2) equalising violence perceived as criminal by Yemeni tribesmen, (including crimes of revenge or street justice, so long as they are carried out with a politi- where violence is used to right a perceived cal purpose24: they may also not be consid- social wrong); (3) economically motivated ered as connected to conflict by perpetrators, violence (where violence is committed in while being considered as a political attack the criminal pursuit of an economic objec- by victims, depending upon the local histori- tive, such as holding somebody to ransom); cal context of tribal relations and their sup- (4) banditry and organised criminal violence port/resistance of the state. In this sense, (which includes violence committed to cover individuals belonging to one tribe could be up corruption); (5) violence against state kidnapped by another in order to resolve a officials that is not committed in pursuit of a land or resource-based dispute between the political objective or as part of a conflict; and two, at which point the victim’s relatives (6) uncategorised violence. might interpret the act at face value, or take These six unions highlight the interde- offence, based on pre-existing attitudes. pendence of crime and conflict, and they help Tribal violence is not legally sanctioned: to illustrate the argument made in this paper it is comprised of criminal offences. These that, while the study of crime in fragile states offences lead to conflict in cases where vio- has been largely under-researched in favour lence between tribes is read by victims as of investigating the restoration of other types having an unjustified political or religious of “political” security, neither can be over- undercurrent. In the case of the Houthi come without a holistic approach that takes conflict in the North, this has repeatedly both into consideration when planning stabi- reignited fighting between pro-Houthi and lising interventions. A failure to treat crime as pro-Government tribes (Loidolt et al 2010). a serious problem in conflict-affected coun- Owing to the importance of political and tries where counter-insurgency is prioritised religious undercurrents in generating con- can allow apolitical violence to dominate, flict and violence – particularly in terms of causing insecurity for decades to come. water and land disputes (Small Arms Survey As a direct point of interaction between 2010) – these tribal clashes need to be care- crime and conflict, tribal violence in Yemen fully considered in the current Yemeni tran- is a reflection of: the Government’s over- sition. This is particularly important in light reliance upon tribal structures22 and patron- of upcoming constitutional reforms in July age networks; the emergence of a virtually 2013, which might result in a redrafting of untouchable parallel state that favours pow- administrative territories and boundaries erful tribal federations (such as the Al Ahmar between communities.25 Tribal relations can family-run Hashid Federation); the limited be central to security and stability in Yemen: reach of law enforcement in hard-to-access they need to be considered in strategies that geographic areas; and, a lack of basic services address either security or law enforcement, that has pushed Yemeni communities, in particularly in terms of the ways in which the North especially, to look to tribal struc- crime, conflict and violence are conceptual- tures as a tool for self-administration. Tribal ised by tribal members. violence in Yemen tends to involve violence Generally triggered by resource competi- between tribes or between tribes and the tion26, tribal violence is a direct product of Lewis: Violence in Yemen Art. 13, page 15 of 22 the state’s lack of service delivery, both in Attempts by law enforcement personnel to terms of basic services and security. A tribe’s intervene in local violence are often per- need to fend for itself, to compete with other ceived as unwelcome and overly intrusive by tribes for political power and representation, the perpetrators of violence, their families and to protect itself from outside threats are and sometimes their neighbours or other instrumental in increasing tribal reliance supporters, depending on the region28. This upon violence. Disputes can also be a prod- development has previously sparked direct uct of corruption, particularly if powerful (and, in the cases of the Sa’ada wars and the individuals are determined to take advantage Southern secessionist protests, prolonged) of protected underground aquifers by drill- confrontations between the Yemeni state ing into water reserves illegally27, at consider- and its citizens. However, at a ground level, able cost for local communities. they have also severely damaged the popu- A lack of law enforcement capacity and larity of armed security forces, including the judicial guarantees has also led to the rise of military and the police29. As a result, state equalising or retributive justice in non-tribal personnel are, in some areas, at a heightened communities. An absence of guarantees that risk of personal attack in isolated incidents perpetrators of crime will be adequately that are not a part of greater conflicts (Ken- punished by the Yemeni Government has ney 2008). led many Yemeni citizens to take justice Even though attacks against state officials into their own hands, sparking blood feuds may be spontaneous or unrelated to politi- (confrontations that can be carried forward cal ideologies at their core, they are defined through generations) and escalations in the by the state as political offences and often severity of street justice (Colburn 2002: 59). incur military responses. As a result, criminal Blood feuds are easily played upon by insur- actions are met with strategies that emerge gencies, leading to increases in both murder from conflict-based policies and result in and war. Meanwhile, emerging systems of conflict-based responses that politicise patronage that have developed as a result of enforcing institutions. Though some citizens the state’s reliance upon tribalism to admin- would now like to see an expansion of Yem- ister its territories have helped to strengthen eni security services to increase public safety grass-roots, institutional and elite corruption (Yemen Polling Centre 2013), most continue necessary for organised crime and banditry to prioritise the reformation of security to take hold (Moghram 2004). This is because services (especially in the South)30. It will minimal levels of institutional cooperation likely be a long time before state-based law can be instrumental in allowing criminal enforcement officers regain national trust networks to take root in a country. Key to in the aftermath of the Arab Spring. As has reversing these trends is a commitment by been recently argued by Dr Abu Bakr Al-Qirbi the Yemeni state to making law enforcement (2013), Yemen’s Minister of Foreign Affairs, and official justice processes more relevant ‘the real issue’ in Government-community and more visible to ordinary citizens by cre- relations ‘is how to build confidence’ and ating services that prioritise their needs and confidence will begin only ‘[w]hen there is their security: these needs cannot be met real dialogue’ demonstrating that citizens’ by the creation of a paramilitary force that concerns for the future ‘are taken seriously’. undervalues the importance of crime as a Independent surveys conducted by the source of insecurity in Yemen. author have documented that tensions A lack of services that are targeted for between law enforcement personnel and their protection has led to increased self- communities were particularly high through- reliance among tribes and communities out 2011, though surveys conducted by the and has resulted in a delegitimisation of the Yemen Polling Centre (YPC) (2013) indicate state and its representatives (Fattah 2011). that relations are improving in 2013, follow- Art. 13, page 16 of 22 Lewis: Violence in Yemen ing President Saleh’s departure. The YPC data ployment and a lack of basic social services shows that there is a demand for higher lev- is also a leading cause of continuing protests els of police community protection, though and hostilities in the south of Yemen, with 19.6% of those surveyed said that they rising levels of aggression erupting in anti- would feel less secure if police presence were social behaviour and causing the Yemeni increased in their areas (2013). Government to intervene repeatedly in what While the Yemeni people want to see has become a priority area for reinstating the change, most currently believe that informal, rule of law. Such incursions have not been tribal or street justice processes provide bet- welcomed by al-Hirak al-Janubi and oth- ter security than state services. 41.1% of peo- ers (who include Southern tribes and Arab ple surveyed by the YPC believed that they Spring protesters without overtly secession- were solely responsible for their own secu- ist aims)32. Therefore dangerous criminal and rity, while 22.9% believed that security was political trends are affected by similar condi- mainly provided by tribal shaykhs (2013). tions, so that integrated response strategies Reversing this attitude is key to increasing could effectively diminish both. security from a conflict and a crime perspec- While individuals who commit violence in tive: helping to take responsibility for both the pursuit of criminal objectives are clearly types of security-provision out of the hands motivated by different factors to those who of tribes is hugely relevant for state stabil- commit violence in the pursuit of political ity and legitimacy. Evidence suggests that a objectives, there seems to be a strong inter- cooperative system that consults tribes about action between their patterns of behaviour security but vests responsibility for security in countries that are affected by state fragil- with the state could be fairly popular in mul- ity and political instability: the Yemeni case tiple tribal regions (Sharqieh 2013: 16). study may not be universally generalizable, Tribal violence, economically-based vio- but its similarity to other conflict-affected lence, equalising violence, violence against tribalised societies especially, including state officials and violence resulting from Afghanistan and Tajikistan, is startling, with organised crime and banditry can be said to additional similarities to other contexts in result from a failure of the state to deliver the Middle East and North Africa region as security in a sufficiently coherent, credible, well. Yemeni governorates affected by con- effective and legitimate way. This shortfall is flict or poverty are more likely to exhibit a product of the state’s extreme fragility and high levels of violent crime as a percentage underdevelopment, which is reflected in the of their total rates of offending. Other similar Government’s inability to maintain authority observations can be made of governorates and a monopoly over the legitimate use of affected by income inequality, poor service force, but it is also the result of donor atti- delivery, or severe instability, with implica- tudes that undervalue the influence of crime tions for fragile states more generally. Mean- on insecurity. while, rising levels of uncategorised crimi- Poor state capacity and severe nationwide nal violence (e.g., social violence, domestic underdevelopment, combined with a short- abuse, hooliganism, or other forms of illicit age of legitimate livelihood opportunities, behaviour) seem to be emerging in commu- is speculated to be contributing to the rate nities located in high-stress, violent environ- of economically motivated violent offending ments in Yemen, especially among children in Yemen31. This economic criminal activity and young people. is evident in statistical records of burglary, Though crime and conflict are independ- theft, robbery, racketeering, fraud, banditry ent factors, they each form part of a complex and a number of other associated offences tapestry of violence and insecurity in Yemen. that have turned violent (CSO 2009). Unem- Analysing one without awareness of the Lewis: Violence in Yemen Art. 13, page 17 of 22 other can lead to serious shortfalls in under- as to why people offend in Yemen and a fail- standings of the country’s fragility and insta- ure to mitigate the causes of crime. This is bility. The six points of intersection between not a problem that is limited to fragile states, crime and conflict highlighted in this section but the overall sense of discouragement and help to showcase specific areas in which the apathy among policymakers towards non- marginalisation of crime in security policy political violence in conflict-affected fragile have impacted upon the (re-)generation of states is no longer excusable, particularly political violence. given its implications for public welfare. While ending violence in fragile states may Conclusion ultimately require a holistic approach, the Yemen’s rates of crime and patterns of severe lack of research on criminal violence conflict are intrinsically connected to state in Yemen likely necessitates that criminal fragility, as this is manifested by a lack of violence and conflict be investigated sepa- territorial control, poor legitimacy, author- rately and addressed in tandem. ity-based weaknesses, and a lack of service delivery, particularly in relation to law- Notes enforcement and security-delivery. A prior- 1 Based on an interview conducted with itisation of conflict over crime has emerged Nicholas Hopton, British Ambassador to naturally from the militarisation of the Yem- Yemen, on July 20th, 2012. eni state and its policing structures. It has 2 The tenth United Nations Survey of Crime also been supported by international and Trends and Operations of Criminal Justice regional donors, who prioritise counter- Systems (UNODC 2005) indicates that, on terrorism and stabilisation strategies that average, economically stable and devel- are intended to generate increased levels oped countries suffer from significantly of public safety and security. However, such higher statistically recorded rates of crime strategies do not acknowledge the complex per capita than so-termed ‘fragile states’. tapestry of violence in Yemen, and there- In 2005, New Zealand (an example of a fore are not designed to provide protec- stable, developed country) had a rate of tion to civilians from other forms of attack. 105,881 offences committed per 1,000 This limitation not only allows offending to people, positioning the country near the escalate unchecked in Yemen, particularly top of the list. Yemen (a fragile state), on through the rise of retributive violence, but the other hand, ranked the lowest, with also enables crime to feed back into and only 1.16109 crimes per 1,000 people. regenerate conflict, most often through These statistics are misleading, offering crime that can be linked to resource compe- a better picture of the comparative effec- tition or tribalism. tiveness of crime detection capacities and While these challenges imply that crime positive police-community relations than and conflict in Yemen should be dealt with comparative criminologies. Countries according to an integrated strategy, there like New Zealand have much higher law is also a clear need for targeted responses enforcement capacities than Yemen, lead- to both of these issues. This is particularly ing to a higher rate of crime reporting, important because crime has not been effec- detection and recording. If we run a com- tively addressed by existing security strate- parison through the same survey data of gies in Yemen, so that it is on the increase global homicide rates, then incidence- and cannot be resolved through negotiation frequency-related trends are reversed. or other peace-building strategies. Yemen moves up the chart to the 27th Greater challenges stem from a lack of out of 60 positions, New Zealand drops understanding, or perhaps a lack of interest, from second place to 52nd. Violent crime Art. 13, page 18 of 22 Lewis: Violence in Yemen

in general, including battery, assault and 10 Authority failures occur when a state sexual abuse, tends to represent a larger ‘lacks the authority to protect its citizens percentage of total recorded offences in from violence of various kinds’, service least developed countries than in devel- failures occur when a state fails to ‘ensure oped ones. While records show crime that all citizens have access to basic ser- rates to be lower in fragile states, the se- vices’ and legitimacy failures occur when verity of crimes exhibited is substantially the state enjoys ‘only limited support higher in these states than in others. among the people’ and is ‘typically not 3 As above, see the United Nations Survey democratic’ (2010: 10). of Crime Trends and Operations of Crimi- 11 In order to gather the statistical data nec- nal Justice Systems (UNODC 2005). essary to provide an account of Yemen’s 4 Expert consultation with Dr Ibrahim crime patterns, the author has had to Sharqieh of the Brookings Institution, 22 rely heavily upon national statistics pub- April 2013. lished by the Government of Yemen’s 5 Of course, readers would be correct in Central Statistical Organisation. Primary assuming that the use of Yemeni Govern- data sources used for this purpose have ment statistics in overcoming definitional included Yemen’s Statistical Yearbooks challenges is overly straightforward and of 2005 to 2010 and Security and Jus- that problems of legitimacy and reliabil- tice Statistics of 2003 to 2009. In order ity abound (as will be detailed shortly). to resolve issues of bias arising from an It is important to keep this definitional over-reliance upon Governmental data challenge in mind when reading the fol- sources, the author has strived wherever lowing sections, as well as the difficulty possible to verify trends emerging from of working in contexts with overlapping Government reports with United Nations bodies of violence. data, made available by Undata, as well as 6 Based on stakeholder consultations car- with trends identified by the Global Peace ried out with the Department for Inter- Index and the World Prison Population national Development in 2010, 2011 List. For a more impartial take on the re- and 2012. liability of Government data, the author 7 Expert consultation with Dr Ibrahim has also relied heavily upon Dennis Jay Sharqieh of the Brookings Institution, 22 Kenney’s report on Public Perceptions of April 2013. the Police in Yemen, which deals with re- 8 Jamal Benomar, Special Adviser to the porting and other issues that may have United Nations Secretary General and impacted the ability of the state to gather United Nations Envoy to Yemen, found statistical data in the country. at a recent Chatham House event titled 12 Similar patterns of domestic violence Friends of Yemen: Aid Allocation and Ac- among displaced communities have been countability that Southern leaders have observed among Iraqi refugee families in been proving increasingly reluctant to Jordan and elsewhere (WCRWC 2007). In committing to principles of non-violence, both cases, where displacement is the re- due to the long term lack of resolution of sult of conflict, there is also a risk of the their grievances (March 6th, 2013). socialisation of families to violence. 9 These findings are based upon exten- 13 In 2009, young women accounted for sive telephone interviews with Yemeni 209 of the 4,060 young offenders de- citizens and former leaders of the Arab tected by the state, representing 5.15% Spring movement. Such interviews have of offenders under the age of 18, as op- been carried out by the author as part of posed to the national average of 1.50% her doctoral research in 2011 and 2012. (CSO 2009). Lewis: Violence in Yemen Art. 13, page 19 of 22

14 Based on stakeholder consultations car- ent tribes are not entirely uncommon. In ried out with the International Organisa- particular, tribes that have crossed over tion for Migration, 16 August, 2010. into Saudi Arabian territory in the far 15 Based on stakeholder consultations car- North of the country have been able to ried out with the Food and Agriculture maintain some independence through Organisation, 14 August, 2010. their isolation from central administra- 16 In 2010, Yemen was at 146th place out tion. These tribesmen identify themselves of 178 countries in the Perceptions of primarily as Yemenis, though they are Corruption Index (TI 2010), despite hav- geographically located beyond the reach ing launched several governance reform of the Yemeni state. Nevertheless, North strategies (Ministry of Planning and Inter- Yemen, and particularly the mountainous national Copperation 2010). regions across the Saudi Arabian border 17 These findings are drawn from the au- have historically been dominated by the thor’s doctoral research. Hashid Federation. Here, the Hashid tribe 18 Based on information gathered from ex- has been able to gain sympathy and influ- pert consultations conducted at a recent ence through shared Zaydi belief systems Chatham House event titled Friends of and common grievances. Yemen: Aid Allocation and Accountability 23 Based on stakeholder consultations car- (March 6th, 2013). ried out with the Department for Inter- 19 As above. national Development in 2010, 2011 20 These findings are drawn from the au- and 2012. thor’s doctoral research. 24 Expert consultation with Dr Ibrahim 21 Each of these dimensions was found to Sharqieh of the Brookings Institution, 22 be a prevalent cause or aggravating risk April 2013. factor for crime and conflict in Yemen, ac- 25 Similar concerns were raised by Helen cording to the author’s doctoral research. Lackner, an expert audience member 22 Yemeni tribes have strong internal social at a recent Chatham House event titled structures and have historically proven ex- Friends of Yemen: Aid Allocation and Ac- tremely self-reliant, so that, today, while it countability, who stated that governance is possible for tribes to function effective- boundaries were an issue that would be ly without the state – it is not possible for under negotiation in upcoming national the state to function without the tribes dialogue talks (March 6th, 2013). (Alexander Knysh, 2011). Prior to 2011, 26 Based on stakeholder consultations car- various rulers of Yemen, both before and ried out with the Department for Inter- after unification, had maintained control national Development in 2010, 2011 over their territories through co-opting and 2012. and bargaining with powerful tribal 27 This finding emerged from the author’s leaders. The tribes themselves form en- doctoral research. trenched social networks that have shift- 28 Based on stakeholder consultations car- ed their composition mildly over the past ried out with the Department for Inter- two decades in response to changes in national Development in 2010, 2011 Yemen’s power structures, but that have and 2012. largely remained the same in their inter- 29 This finding emerged from the author’s nal groupings. They are united according doctoral research. to overarching organisational structures, 30 Expert consultation with Atiaf Al Wazir of the most prominent of which have been Woman in Yemen, 5 June, 2012. the Hashid Federation and the Bakil Fed- 31 This finding emerged from the author’s eration respectively, though independ- doctoral research. Art. 13, page 20 of 22 Lewis: Violence in Yemen

32 Southern distrust of centralised interven- London: Catholic Institute for Interna- tions is connected to the history of con- tional Relations. flict between the North and the South, Dresch, P 1993 Tribes, Government and His- particularly during the 1994 Civil War, tory in Yemen. Oxford: Clarendon Press. in which Northern forces devastated Economist Intelligence Unit 2010 Country Southern cities in order to send a clear Report: Yemen. London: Economist Intel- message about future resistance (John- ligence Unit. sen 2013: 45). Education Development Centre 2008 Yem- en Cross-Sectoral Youth Assessment: Final References Report. Cairo: United States Agency for Amnesty International 2012 Yemen: Hu- International Development. man Rights Agenda for Change. London: Fattah K 2011 Yemen: A Social Initfada in a Amnesty International. Republic of Sheikhs. In Middle East Policy, Bertocchi, G and Guerzoni, A 2011 The 18(30): 79–85. Washington D.C.: Middle Fragile Definition of State Fragility. Mod- East Policy Council. ena: University of Modena and Reggio Human Rights Watch 2010 All Quiet on the Emilia. Northern Front? Uninvestigated Laws of Boucek, C 2010 War In Sa’ada: From Local War Violations in Yemen’s War with Huthi Insurrection to National Challenge. Wash- Rebels. New York: Human Rights Watch. ington D.C: Carnegie Endowment. -- 2009 In the Name of Unity: The Yemeni Brown, G and Stewart, F 2010 Fragile States. Government’s Brutal Response to South- Oxford: Centre for Research on Inequal- ern Movement Protests. New York: Hu- ity, Human Security and Ethnicity. man Rights Watch. Brown, N J; Crystal, J and Nasr, H 2004 International Crisis Group 2009 Yemen: Criminal Justice and Prosecution in the Diffusing the Sa’ada Time Bomb. Brussels: Arab World. Beirut: United Nations De- International Crisis Group. velopment Project Programme on Gov- Johnsen, G D 2013 The Last Refuge: Yemen, ernance in the Arab Region. Al-Qaeda, and America’s War on Arabia. Central Statistical Organisation 2004 Se- London: W.W. Norton & Company. curity and Justice, 2003. Sana’a: Republic Kenney, D J 2008 Public Perceptions of the of Yemen. Police in Yemen: Nationwide Surveys of -- 2010 Security and Justice, 2009. Sana’a: Residents and Police. New York: John Jay Republic of Yemen. College, City University of New York. -- 2004 Statistical Yearbook, 2003. Sana’a: Lewis, A 2012 Changing Seasons: The ‘Arab Republic of Yemen. Spring’ and Its Impact on Yemen’s On-Go- -- 2005 Statistical Yearbook, 2004. Sana’a: ing Development Challenges. York: Univer- Republic of Yemen. sity of York. -- 2006 Statistical Yearbook, 2005. Sana’a: Loidolt, B; Salmoni, B A and Wells, M 2010 Republic of Yemen. Regime and Periphery in Northern Yemen: -- 2007 Statistical Yearbook, 2006. Sana’a: The Houthi Phenomenon. Arlington: RAND Republic of Yemen. National Defence Research Institute. -- 2008 Statistical Yearbook, 2007. Sana’a: Martínez, A D 2007 Organised Crime, the Republic of Yemen. State and Democracy: The Cases of Central -- 2009 Statistical Yearbook, 2008. Sana’a: America and the Caribbean. Madrid: Fun- Republic of Yemen. dación para las Relaciones Internacion- -- 2010 Statistical Yearbook, 2009. Sana’a: ales y el Diálogo Exterior. Republic of Yemen. Miller, D B 2003 Demand, Stockpiles and Colburn, M 2002 The Republic of Yemen: De- Social Controls: Small Arms in Yemen. Ge- velopment Challenges in the 21st Century. neva: Small Arms Survey, Graduate Insti- Lewis: Violence in Yemen Art. 13, page 21 of 22

tute of International and Development ment In Conflict-Affected Governorates in Studies. North Yemen. Sana’a: UNICEF. Ministry of Planning and International United Nations Development Project Cooperation 2010 National Reform 2010a Human Development Index, 2010. Agenda, Phase II, 2009–2010. Sana’a: Re- New York: UNDP. public of Yemen. -- 2010b Re-thinking Approaches to Local Moghram, M A 2004 Political Culture of Cor- Governance Programming in Conflict- ruption and State of Corruption in Yemen. Affected Countries in the Arab Region: Sana’a: Sana’a University. An Exploratory Study (Draft). Cairo: Re- Murphy, M N 2010 Small Boats, Weak States, gional Bureau of Arab States, UNDP. Dirty Money: Piracy and Maritime Ter- United Nations Department of Economic rorism in the modern World. London: C. and Social Affairs, Population Divi- Hurts & Co. sion 2009 World Population Prospects: Organisation for Economic Co-Operation The 2008 Revision. New York: United and Development 2009 Armed Violence Nations Data. Available at: http://data. Reduction: Enabling Development. Paris: un.org/ [accessed 2 December 2010]. OECD. United Nations Office on Drugs and Phillips, S 2010 ‘What Comes Next in Crime 2005 The United Nations Survey Yemen? Al Qaeda, the Tribes, and State on Crime Trends and the Operations of Building’. In Carnegie Endowment 2010 Criminal Justice Systems. Geneva: United Yemen: On the Brink. Washington DC: Car- Nations. negie Endowment. -- 2010 UNODC and Yemen launch five Roach, J and Pease, K 2011 ‘Evolution and year integrated country programme to the Prevention of Violent Crime’. In Sci- tackle organised crime. Geneva: UNO- entific Research 2011 Psychology, no. 2. DC. Available at: http://www.unodc. Irvine: Scientific Research Publishing. org/ [accessed 22 May 2011]. Sharqieh, I 2013 A Lasting Peace? Yemen’s Walmsley, R 2007 World Prison Population Long Journey to National Reconciliation. List. 7th ed. London: International Centre Doha: Brookings Doha Centre. for Prison Studies, Kings College London. Shaw, M 1997 ‘The State of Globalisation: -- 2009 World Prison Population List. 8th Towards a Theory of State Transforma- ed. London: International Centre for tion’. In Review of International Political Prison Studies, Kings College London. Economy, 4(3). Abingdon: Taylor & Francis Women’s Commission for Refugee Wom- Small Arms Survey 2010 Fault Lines: Track- en and Children 2007 Iraqi Refugee ing Armed Violence in Yemen. Geneva: Women and Youth in Jordan: Reproductive Graduate Institute of International and Health Findings. New York: International Development Studies. Rescue Committee. Transparency International 2010 Corrup- World Bank 2011 World Development Re- tion Perceptions Index, 2010. Berlin: Trans- port, 2011: Conflict, Security and Develop- parency International. ment. Washington, D.C: World Bank. United Nations Country Team Child Pro- Yemen Polling Centre 2013 Public Percep- tection Sub-Cluster 2010 Inter-Agency tions of the Security Sector and Police Work Comprehensive Child Protection Assess- in Yemen. Sana’a: Yemen Polling Centre. Art. 13, page 22 of 22 Lewis: Violence in Yemen

How to cite this article: Lewis, A 2013 Violence in Yemen: Thinking About Violence in Fragile States Beyond the Confines of Conflict and Terrorism.Stability: International Journal of Security & Development, 2(1): 13, pp. 1-22, DOI: http://dx.doi.org/10.5334/sta.az

Published: 14 May 2013

Copyright: © 2013 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 Unported License (CC-BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/3.0/.

Stability: International Journal of Security & Development is a OPEN ACCESS peer-reviewed open access journal published by Ubiquity Press