ASSESSMENT FRAMEWORK 1: DILEMMAS OF SECURITY?

In the tripartite framework of analysis that we have adopted in this symposium, one path from transition to violence occurs when state security forces significantly weaken or collapse alongside regime change. In the absence of state-sponsored security, local groups may choose to take up arms in self-defense, giving rise to a domestic security dilemma, a spiral of mistrust, and, ultimately, large-scale violence. Armed groups may also engage in opportunistic violence, such as looting or mafia-style exploitation. Both of these dynamics have some merit in explaining and predicting transitional violence in Yemen. However, as analytic tools their utility is tempered by the country's particular historical legacy of a weak state security apparatus, multiple nonstate armed actors, and strong informal mechanisms of conflict management.

Since the 2011 uprising, when the already-weak military began to fracture, a variety of armed nonstate actors have stepped in to fill the expanding security void. Those actors include the primarily northern-based Huthi movement, which expanded its territorial control over Saada and parts of the al-Jawf, Hajjah, and Amran governorates; various tribal groupings (some affiliated with parties), which augmented their influence in the northern highlands while also gaining traction in areas known for greater degrees of state control, such as Taiz governorate; and small parts of the southern Hiraak, primarily located in the Dalia and Lahj governorates (see figure 1).Footnote ⁴ Two new, armed entities also emerged: Al-Qaeda affiliated militias known as Ansar al-Sharia and a range of “popular committees,” which formed to oppose Ansar.Footnote ⁵ In the context of a weakened and fractured security apparatus, there have been recurrent clashes between a number of these armed groups.

Figure 1

Administrative Map of Yemen

From The World Factbook 2013–14, Washington, DC: Central Intelligence Agency, 2013 . The Factbook is in the public domain. (Color online.) https://www.cia.gov/library/publications/the-world-factbook/index.html (Color online.)

Many instances of violence have been products of opportunistic behavior as groups compete for popular support, economic resources, and territorial control in the post-Saleh era. For example, as the Saleh regime focused on controlling Sanaa in 2011, the Huthis consolidated and expanded control over northern provinces. In the process, they clashed with Islah-affiliated militias in al-Jawf governorate, which opposed their political and military expansion (Winter Reference Winter2012). Now the Huthis, whose bargaining position vis-à-vis others is augmented by their military prowess and territorial control, are a powerful force in the national dialogue and the domestic political balance of power more generally.

As state control contracted in the southern governorate of Abyan, violence there had an element of both a domestic security dilemma and opportunism. In Abyan, Ansar al-Sharia capitalized on the security vacuum in May 2011 and, in a matter of weeks, took control of a significant swathe of territory. When the state failed to contain the rise of extremists, local communities formed popular committees in self-defense (ICG 2011b, 26–28). These committees later worked in coordination with government forces to oust al-Qaeda affiliates from urban centers in June 2012. The government now directly supports these committees, and they remain the cornerstone of security in Abyan today.Footnote ⁶ Nevertheless, al-Qaeda affiliates are still active, and clashes continue between Ansar and community militias in Abyan and surrounding governorates.

Although the structural conditions for a security dilemma and an associated spiral of violence to civil war appear to be present in Abyan and beyond, the model falls short of fully explaining and predicting violence in the Yemeni case, more broadly. In practice, the immediate likelihood of spiraling violence is actually low, for two main reasons.

First, dissolution of the Saleh regime precipitated a fracturing—rather than wholesale collapse—of state security services. Today, the security services are clearly weaker and more divided than they were prior to 2011. However, they retain some basic security provision capacity, in part because the negotiated settlement supported gradual reform and changes to patronage networks, rather than wholesale “de-salehfication.”

Second, and more importantly, in many areas of the country, no functional state structures were in place at the point of regime change, and so the impact of transition on (the lack of) state security provision has been negligible in those areas. Under Saleh, state forces were weak, corrupt, and enjoyed only limited writ outside of cities. Security, especially in the northern highlands, often rested on a patchwork of state forces and armed nonstate actors. Both sets of actors occasionally contributed to security provision, but both also undermined security on occasion. Given the limited reach of the state under Saleh (and, indeed, before his rule), tribally based militias have long been part of the security landscape, especially in the north. Over time, those armed groups have established informal rules that govern their interactions and mitigate intercommunal violence (see Caton Reference Caton2005; Dresch Reference Dresch1993; Weir Reference Weir2007). Those informal arrangements, along with the population's conflict fatigue after a year of armed clashes, provide important counterweights to the potential for violence to spiral toward unbound conflict.Footnote ⁷

Going forward, the outcome of the national dialogue is likely to have a significant impact on the nature of relations among armed groups. Ideally, a settlement will ensure that all major players—including the Huthis and the Hiraak (each of which has an armed component, although the Huthis to a much greater degree)—feel secure enough to fully engage the political process and eventually forgo the threat of armed violence. Such an inclusive compromise would also provide a further check on the potential for violence among militias by reducing the likelihood that political stakeholders such as the Salehs, Mohsen, or the al-Ahmar clan might seek to use armed groups as spoilers to the transition or as tools in political negotiations. If negotiations fail or are significantly stalled, however, an ongoing danger of violent clashes will persist between the country's numerous and diverse armed factions.

ASSESSMENT FRAMEWORK 2: DILEMMAS OF GOVERNANCE?

The second pathway from regime change to violence occurs when intense political competition develops between groups jostling for control over the levers of government or process of transition. This path has already given rise to several episodes of violence in Yemen, and a danger remains that such episodes may spiral into large-scale conflict—particularly if the national dialogue fails to produce a durable power-sharing arrangement. With the exception of youth and civil society groups, all political actors have the ability to resort to violence as a form of “politics by other means,” and they have a history of doing so.

Intense competition between the (mostly Saleh-aligned) General People's Congress (GPC) and the JMP over the spoils of the state, and intense disagreement between those factions about details of the transition settlement, have already paralyzed the government. That said, to effectively understand the dynamics of political competition—and the potential for that competition to turn violent—it is necessary to look beyond the GPC-JMP divide because the political field is multipolar, highly fractured, and fluid. On the GPC side of the equation, there is a split between those who are calling for Saleh to retain his position as party leader and those who are pushing for his resignation. The latter are generally aligned with President Hadi, whom they would like to see take the reins as GPC party president (ICG 2012, 14). The situation is more complicated on the side of the former revolutionary coalition. Although the alliance was initially bound together by a common opposition to Saleh, that bond weakened following his resignation. Since then, independent youth and civil society activists have regularly objected to the continued presence of established JMP party elites in the “new” Yemen and also to the ongoing prominence of former regime supporters, such as Ali Mohsen and the al-Ahmars.Footnote ⁸ Open divisions are also increasing within the JMP, which is a coalition that was formed in 2002 with the explicit goal of organizing opposition to the GPC and Saleh. Specifically, latent tensions have surfaced between the most powerful coalition partner, Islah, and smaller constituent parties, which often complain that Islah wants to monopolize the political scene.Footnote ⁹

Divisions among the ranks of the revolutionary coalition are further complicated by sectarian and regional cleavages. In the north, relations between Islah and the Huthis turned sour, with Huthi supporters accusing Islah of betraying the revolution and trying to monopolize power.Footnote ¹⁰ As recently as July 2013, Islah and Huthi supporters violently clashed over control of mosques during the holy month of Ramadan. In the south, Hiraak activists who initially sympathized with the national uprising have now distanced themselves from the revolutionary movement, having become frustrated with the lack of attention given to southern demands (ICG 2011b). That frustration has led to increasing calls for secession and fueled tensions between Islah and Hiraak activists—as evidenced on February 21, 2013, when an Islah-organized rally celebrating Hadi's election prompted counter-demonstrations from Hiraak supporters. In the melee that ensued, at least six Hiraak demonstrators were killed by Yemeni security services. Those killings sparked several days of violence in southern governorates (AFP 2013).

Of all the political divisions, the issue of “the south” is arguably the most divisive and, currently, the most intractable. Southerners, in general, have long been unhappy with the political status quo, which they view as privileging the north and exploiting southern resources. Among the ranks of the Hiraak, however, there is disagreement over how to act on their grievances; while one bloc supports immediate separation, another is open to (con)federation in two parts, north and south (ICG 2011b, 12–15). Until now, the transitional government has done little to address security, employment, and land grievances in the south—steps that could open up political space for a compromise. The current absence of such measures provides political fodder to those who are committed to immediate separation, rather than negotiation.

Whereas the complex political divisions described are evidently problematic in terms of reaching a political settlement, they may have the unintended benefit of reducing the likelihood of large-scale political violence in the immediate term. That is, the political field currently is so factionalized that no one political group can realistically aspire to unilateral control over the levers of government, and, even if it did, no one faction currently has the necessary level of military backing or international support that would allow it to use significant force in support of such an endeavor. Consequently, although a spiral of violence toward sustained, large-scale conflict appears unlikely, turf battles between political factions that are jostling for localized power may continue to produce finite episodes of violence, especially in the south.

Sustained and widespread clashes could transpire, however, if the national dialogue fails to produce any sort of agreement, or if an agreement is reached but that deal is seen to favor the agenda of one set of actors (particularly power elites from the old regime) to the exclusion of other stakeholders. At a minimum, any agreement must allow for a significant devolution of power. In the absence of an acceptable compromise on the question of devolution, elections tentatively scheduled for February 2014 could trigger significant regionally based violence. Specifically, if a large percentage of southerners continue to reject a compromise within unity, elections may provide an opportunity for nationalist out-bidding on the part of separatists and prounity supporters alike. This scenario could result in armed clashes between opposing camps.

ASSESSMENT FRAMEWORK 3: SECURITY SECTOR DILEMMA?

In the third dynamic of transitional violence, violent conflict is a product of competition between distinct branches of the security apparatus for power and resources in the posttransition era. In the short term, this causal process has limited value for understanding the dynamics of transitional violence in Yemen, for two main reasons. First, the locus of allegiance and competition in the Yemeni security apparatus is not tied to institutional affiliation; instead, it is based on a combination of tribal, regional, and personal loyalties. Thus far, violent competition over resources has centered on these cleavages, rather than competition between the military and the police, for example. Second, until now, security sector restructuring has taken the form of personnel and organizational changes that have been aimed at weakening the power centers of Saleh and, to a lesser extent, Mohsen. There has been no strategic level, national debate over the appropriate (re)allocation of resources among distinct branches of the security services and, therefore, no trigger for institutionally based conflict over this potential reallocation.

Under Saleh, the security services were designed to ensure regime longevity through coup-proofing measures that were built on patronage and nepotism. The security apparatus was virtually immune to civilian oversight, and it was racked by corruption. Loyalty was to individual commanders, who mostly hailed from the president's family or tribe. Toward the end of his 33-year rule, Saleh concentrated power and resources in organizations controlled by his son (the Republican Guard) and nephews (who commanded the paramilitary Central Security Forces and the National Security [Intelligence] Bureau). When the regime fractured during the 2011 uprising, the security sector splintered along intraregime battle lines (ICG 2013). Although the Republican Guard, the Central Security Forces, and the National Security Bureau remained largely loyal to the regime, significant portions of the regular army, including the First Armored Division, defected under the leadership of Ali Mohsen (Gordon and Zimmerman Reference Gordon and Zimmerman2012).

Given this fracturing, the transition agreement prioritized unification and restructuring of the security services. In practice, the restructuring has two overarching goals: one is to weaken the grip of Saleh-era elites on branches of the security services in the short term in order to build public confidence in the transition and strengthen Hadi's position. The second goal is long-term security sector reform through professionalization and restructuring, based on threat assessments. Thus far, President Hadi has attended to the first goal through a series of personnel shifts and organizational changes that have removed Saleh loyalists and strengthened his control over the armed services. In August 2012, for example, Hadi built a new Presidential Protection Unit under his direct authority by pulling brigades from Saleh's son, Ahmed Ali Abdullah Saleh, and Ali Mohsen. Then, in December 2012, he announced major organizational changes to the Ministry of Defense (MoD), which abolished the semiautonomous Republican Guards and the First Armored Division and absorbed both organizations into a new, regionally based military structure.Footnote ¹¹ Finally, in April 2013, far-reaching personnel changes removed Ahmed Ali from military service and appointed Ali Mohsen as a military-security presidential advisor.

Initially, the Saleh camp openly resisted reforms, in an effort to maintain its existing control over military power and resources. For example, Saleh's half-brother refused to relinquish his post as air force commander in April 2012, and, by way of protest, he briefly shutdown Sanaa international airport with his troops before eventually being forced to comply under the threat of UN sanction. Tensions flared again in August 2012 when Republican Guard troops stormed the MoD building, killing two civilians and wounding 15, in protest at their reassignment to a regional command. Over time, such overt opposition has ended, partly due to the threat of UN sanctions for interfering with the transition. Still, subtle forms of resistance have persisted (ICG 2013). Since Hadi's April 2013 decrees, for example, more than 40 mutinies have reportedly taken place against commanders (Abulohoom Reference Abulohoom2013). Financial and personal disputes have triggered some of those disputes, but other defections seem to be related to the continuing elite power struggle.

As of the time of writing, Hadi's restructuring efforts have successfully attenuated the power of the Saleh camp, while doing less to reduce Moshen's influence. By contrast, long-term professionalization and accountability to civilian institutions has yet to be adequately addressed. These more fundamental reforms largely hinge on national dialogue outcomes. In the absence of an enduring political pact through dialogue, low-grade competition and conflict over control of security-service resources and influence is likely to continue, while more fundamental change that redefines civilian-military relations and ensures a national, professional security service, will likely remain elusive.

One clear by-product of the transition has been increased competition around partisan and regional lines within the security services. Recruitment into both the MoD and the Ministry of the Interior (MoI), for example, is one of the most contentious transitional issues. Specifically, GPC and Saleh partisans accuse Ali Mohsen and Islah of partisan recruitment into the army and, in particular, the police because the Interior Minister is affiliated with Islah in the coalition government.Footnote ¹² For their part, Islah accuses the Salehs of similar behavior.Footnote ¹³ Meanwhile, on a regional level, widespread concerns have surfaced that Hadi is repeating the mistakes of the past by recruiting personnel from his home governorate of Abyan into strategic military positions.Footnote ¹⁴ Although these dynamics suggest that intense competition exists over security service resources in transitional Yemen, they also suggest that such competition is taking place along elite, partisan, and regional lines rather than along interservice institutional lines. As suggested earlier, the current lack of significant interagency conflict is partly because institutional affiliations are weak, and partly because reform measures not have yet seriously addressed the issue of relative power redistribution between security organizations in the posttransition era. If that issue is eventually tackled, there could be institutionally based competition between the MoD and MoI over the future distribution of resources and authorities. Competition may also develop between the two intelligence services—the National Security Bureau and the Political Security Organization—which have broadly similar mandates. Whether competitive dynamics between those agencies turn violent depends, in part, on the transparency and inclusiveness of the security sector reform process.