%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
INTRODUCTION
Maritime piracy in Somalia exploded between 2007 and 2012, but was not evenly distributed throughout the country, nor did it appear in what would seem to be the most likely culprits: the most ‘failed’ parts of Somalia (Coggins Reference Coggins2011). In southern Somalia, which has experienced the worst ravages of conflict and humanitarian catastrophe, piracy was minimal, suggesting that some level of infrastructure and centralised (if often informal) governance capacity is necessary for piracy syndicates to survive and prosper (Hastings Reference Hastings2009; Coggins Reference Coggins2011; Dua & Menkhaus Reference Dua and Menkhaus2012; Percy & Shortland Reference Percy and Shortland2013; World Bank Group 2013). Somaliland, the self-declared (though internationally unrecognised) independent ‘republic’ in the north-western corner of Somalia also saw very few pirate attacks originating from its shores. Instead, with a few exceptions, piracy in its heyday emanated from Puntland, the autonomous region in the north-eastern corner of Somalia, and the area immediately to the south of Puntland, Galmudug (United Nations Security Council 2008, 2010, 2011; World Bank Group 2013). Why was Somaliland almost entirely left out of the piracy networks that permeated Puntland and central Somalia for years? We argue that Somaliland's much publicised ‘piratelessness’ has implications not only for our understanding of the conditions that lead to piracy, but also for the foundations of political order, quite apart from formal state capacity.
Specifically, we use the relative lack of piracy in Somaliland to show how some territories in Africa come to experience political order in the absence of effective state institutions, and avoid being caught in the cycles of violence and insecurity so widely predicted in the state capacity and ‘state failure’ literatures (Collier Reference Collier2008; Ghani & Lockhart Reference Ghani and Lockhart2008; Braithwaite Reference Braithwaite2010; DeRouen et al. Reference DeRouen, Ferguson, Norton, Park, Lea and Streat-Bartlett2010; Gleditsch & Ruggeri Reference Gleditsch and Ruggeri2010; World Bank Group 2011; Walter Reference Walter2015). State ‘failure’ has long been assumed to lead to an increased incidence of crime, terrorism and ambient violence (Rotberg Reference Rotberg2002, Reference Rotberg2003; Takeyh & Gvosdev Reference Takeyh and Gvosdev2002; Piazza Reference Piazza2008), with the intuition being that state institutions may be either too weak to prevent violence, or fragmented enough to be ignored/co-opted by violent actors. Given the interconnectedness of the global economy, and the ease with which modern technology allows communications and mobility, the internal problems of states with weak central governance are framed within the ‘state failure’ literature as readily exportable, and thus dangerous for the rest of the world (Rabasa et al. Reference Rabasa, Boraz, Chalk, Cragin, Karasik, Moroney, O'Brien and Peters2007; Clunan & Trinkunas Reference Clunan and Trinkunas2010).
On the other hand, there is a rich literature within the Area Studies and Postcolonial fields that challenges the notion of state ‘failure’ (see for example, Hill Reference Hill2005; Hagmann & Höhne Reference Hagmann and Höhne2009; Nurazzaman Reference Nuruzzaman2009). One prevalent line of critique contends that ‘failure’ is an ideological position that ‘Others’ the global South and thereby reinforces prevailing power asymmetries between north and south. Another is that governance does not require a formal government to provide a reasonable level of security, and can instead be performed at the local level by informal (sometimes referred to as ‘hybrid’) actors (see for example, Menkhaus Reference Menkhaus2006/7; Raeymaekers et al. Reference Raeymaekers, Menkhaus and Vlassenroot2008; Boege et al. Reference Boege, Brown, Clements and Nolan2009; Kabamba Reference Kabamba2010). We diverge from the focus of both by instead focusing on the discursive (as opposed to the rule-based, though informal, institutional) mechanisms that can also provide order.
The solution posited within the ‘state failure’ literature for states experiencing terrorism, piracy or other types of organised violence is to improve state capacity, usually with external assistance (Rotberg Reference Rotberg2002, Reference Rotberg2003; Crocker Reference Crocker2003; Collier Reference Collier2009; World Bank Group 2011). This literature assumes, however, that the stronger a state's institutions are, the greater the level of political order they engender, and the weaker they are, the greater the level of disorder they permit. While this is a probabilistic claim (suggesting that greater weakness tends to be associated with an increased likelihood of disorder, such as piracy), explanation of the underlying mechanisms that produce civil order is still required. We argue that while issues of institutional legitimacy are less observable than the blunt measures of state capacity so often used in the literature, exploring how legitimacy is articulated locally offers greater insight into the drivers of civil order than does the study of institutional forms.
In this paper, we explore these deeper mechanisms by looking at micro-level dynamics of maritime piracy (and its absence) in Somalia. Piracy has plagued a number of countries on the continent, notably those in the Gulf of Guinea (Nigeria, Togo and Benin) and Horn of Africa (particularly Somalia) (Murphy Reference Murphy2007, Reference Murphy2008, Reference Murphy2010, Reference Murphy2013; Chalk Reference Chalk1998, Reference Chalk2008; Kaplan 2009; Daxecker & Prins Reference Daxecker and Prins2013, Reference Daxecker and Prins2015a ; Hastings & Phillips Reference Hastings and Phillips2015), and we argue that exploring the local nuances of this phenomenon offers a window into the violence and criminality often associated in the literature with African state institutions. We exploit in-country variation in piracy incidence in different regions of Somalia – widely framed as the ‘world's most comprehensively failed state’ (The Economist 2008) – to investigate how some territories with low state capacity nonetheless deter piracy and maintain relative order.
We find that the usual explanation – state ‘failure’ in Somalia, compared with a reasonably functional government in Somaliland – is based on assumptions that do not withstand contextual scrutiny. Indeed, neither Puntland nor Somaliland had the formal institutional capacity to defeat pirates from 2007 to 2012, and other commonly discussed factors – geographic, economic and human development levels – did not clearly make Somaliland structurally less prone to piracy than Puntland. If the lack of piracy in Somaliland is not primarily attributable to the ability of its formal government apparatus to contain it, then what are the more salient (non-state) processes, structures and ideas guiding this outcome?
More boldly, what might these processes and structures suggest about the relationship between institutions and political order regardless of the degree to which external, rule-based institutions have been formally codified? In places where the state does not command (and, perhaps, does not necessarily seek to command) the monopoly on legitimate violence that is so widely assumed to be a universal attribute of statehood, it can be useful to conceptualise political order in ‘post-Weberian’ terms (Migdal & Schlichte Reference Migdal, Schlichte and Schlichte2005). That is, rather than thinking of state institutions as the most likely means of providing political order, we must look more broadly at the suite of mechanisms that renders the use of violence unlikely in situations of conflict. For such mechanisms to be analytically salient, they must function with a reasonable degree of predictability, and have popular legitimacy such that there is an understanding that conflict will be dealt with, in the main, through non-violent means.Footnote 1 The value of thinking more broadly about the mechanisms that engender political order is that it allows us to conceive of order without there necessarily being an entity that asserts, or seeks to assert, a monopoly on legitimate force. The key point becomes whether there are widely accepted mechanisms for resolving conflict that do not usually require the application of physical force. Shifting the focus in this way means that scholars must be more attuned to the discursive and ideational means by which these mechanisms are upheld, rather than simply to their external institutional manifestations. It is to these discursive foundations that this piece now turns.
Focusing on popular ideas within much of Somaliland about its ‘legitimate’ claim of independence from the rest of Somalia, we argue that the near lack of piracy in Somaliland was more attributable to its dominant discourse about the supposed incompatibility of the Somaliland identity with violence and criminality than it was to effective state (or state-like) institutions. This discourse, referred to throughout this paper as the Independence discourse, frames the country's ability to maintain peace and political order as the cornerstone of its case for international recognition of its sovereignty. It frames Somalilanders as unified in their desire to avoid repeating the violence of Somalia's past which, it purports, both differentiates Somalilanders from other Somalis, and justifies their permanent legal separation from the Republic of Somalia. The discourse began to emerge during the late 1980s, as Somalilanders were subjected to increasing brutality by the regime of Siyad Barre. That brutality, which culminated in the collapse of the regime and Somaliland's declaration of independence in 1991, was so overwhelming that it established a widespread belief that the security of Isaaq clan members (the majority clan in the north-west) could only be guaranteed by a separation from Somalia. The discourse builds a case for independence by emphasising the supposed distinctiveness of the Somaliland identity from that of Other Somalis (see Höhne Reference Höhne2009: 264).
The discourse constructs a binary opposition between ‘peaceful’ and ‘law-abiding’ Somalilanders and a ‘violent’ or otherwise ‘criminal’ Somali Other to justify its secession from Somalia, and holds that ‘all the pirates are from the South; [they are] not Somalilanders’.Footnote 2 While there are empirical problems with this assertion that are discussed below, all discourses exclude contradictory information in order to frame their subjects in ways that limit the range of actions that are considered possible (Epstein Reference Epstein2008: 2; see also Milliken Reference Milliken1999; Hansen Reference Hansen2006). When a discourse is dominant, it is widely accepted as presenting a self-evident or common-sense understanding of the way that something ‘naturally’ is. It does so, however, not because it presents the only logical understanding of that thing but by actively excluding alternative ways of understanding it (Epstein Reference Epstein2008: 9), rendering them inappropriate or simply wrong. In the present case, Somaliland's Independence discourse asserts that the country has not experienced piracy because piracy is unacceptable to genuine Somalilanders. With that caveat of authenticity, the small number of piracy incidents that have occurred within Somaliland's claimed territory may be glossed over as having been committed by non-genuine Somalilanders.
This paper employs a Discursive Institutional approach (Schmidt Reference Schmidt2008, Reference Schmidt2011; Blyth Reference Blyth, Gofas and Hay2010; Carstensen Reference Carstensen2011) to bring together its two key arguments, that is that there are other ways to provide order besides external rule-based institutions, and that sub-national variation in the prevalence of piracy offers scholars analytical purchase on these mechanisms. The core proposition of Discursive Institutionalism is that the ‘interactive process of conveying ideas’ (discourse) both constrains and incentivises behaviour (Schmidt Reference Schmidt2008: 303). By regulating practice, therefore, discourses function like institutions, though they do so in ways that are less visible than external state institutions.
Methodologically, the paper utilises a range of sources that narrate Somaliland's proclaimed independence from Somalia, including interviews with local actors in Somaliland, government publications and local media stories.Footnote 3 It is sensitive to the fact that one of the core goals of Somaliland's Independence discourse is to convince an international audience of the viability and legitimacy of its proclaimed independence, and that the discourse is therefore likely to be particularly pronounced when articulated to an external audience. It attempts to balance the distortions that this may bring by analysing United Nations economic data in both Somaliland and Puntland. These data allow us to roughly compare Somaliland and Puntland's institutional capacity and, we argue, support our contention that the variation in piracy incidence cannot be explained by variations in economic performance or the capacity of external rule-based institutions (whether formal or clan-based) alone.
The paper proceeds initially by examining several alternative explanations for why Somaliland did not see the piracy that Puntland did through a paired comparison of the two regions. While there are a number of plausible political, economic and geographic explanations, we find them all limited when looking at the actual conditions in Somaliland and Puntland both before and during the peak piracy period. Next, we draw from interviews with stakeholders and observers in Somaliland to argue that, first, Somaliland did not have the state capacity to tackle piracy through its coercive institutions. Second, we propose that rather than being able to eliminate (or prevent the rise of) piracy through state institutions, Somalilanders have instead collectively defined – and purposefully articulated – piracy out of their Independence discourse, essentially shaming it out of existence as something that is incongruent with the Somaliland identity. Third, we argue that Somalilanders who support independence tend to define territory within Somaliland that did experience piracy as not part of the ‘real’ Somaliland, thereby maintaining the discursively constructed link between the Somaliland identity and orderly, lawful conduct. We conclude with a discussion of the implications of our argument for the study of state institutions and crime in Africa, where states do not necessarily seek a monopoly on the use of legitimate force.
COMPARING PUNTLAND AND SOMALILAND
Several alternative explanations can be found in the literature for why Somaliland did not see much piracy while Puntland (and central Somalia) did. All are limited in their ability to explain the variation between the regions, and lead us to look to discourse as a crucial element in explaining the lack of piracy.
Economic factors
First, there may be an economic logic to the difference in the incidence of piracy in Somaliland and Puntland: if non-piracy-related economic opportunities are lacking in Puntland, this may lead to an increase in pirate attacks. Conversely, if the economy in Somaliland is doing well, this may lead to fewer pirate attacks as people find piracy (and the risks associated with it) to be less viable than other ways of making money. Daxecker & Prins (Reference Daxecker and Prins2013) argue that maritime piracy incidence increases with a decline in fisheries stocks (and thus a loss of economic opportunity for fishermen, who become pirates). Jablonski & Oliver (Reference Jablonski and Oliver2013) find a similar pattern, with changes in labour and capital-intensive commodity prices having a significant, consistent effect on pirate attacks in countries’ territorial waters; that is, when commodity prices indicate there is likely to be money in working in the relevant industries, the attractiveness of piracy goes down, and so do the number of attacks.
Yet, a fisheries-based economic explanation for piracy incidence may work for explaining piracy incidence in all of Somalia, but it does not provide an explanation for within-country variation. A drop in fisheries stocks is likely to affect both Somaliland and Puntland equally, given that people from both areas are fishing in the same waters. For such an explanation to work in the Puntland and Somaliland cases, the economy in Somaliland should have been significantly better immediately before and during the piracy period than the economy in Puntland. While robust economic figures are hard to come by – even for UN officials working on the ground for this purposeFootnote 4 – there are several reasons making it unlikely that there would have been vast differences between their economies. Both regions have a roughly commensurate population size (though Somaliland is probably larger – (UNFPA 2014)),Footnote 5 with economies that are predominantly centred on pastoralism, the exchange of services, and the receipt of remittances from members of the diaspora (Hastings & Phillips Reference Hastings and Phillips2015). Being largely pastoral, the economies of both Puntland and Somaliland would have been fairly equally affected by the nine-year ban that was imposed by Saudi Arabia on Somali livestock exports until late 2009. When the ban was finally lifted, which was seen as an economic boon in both Puntland and Somaliland, piracy emanating from Puntland (and central Somalia) did not diminish, as this argument would suggest that it should have. In fact, it increased until it peaked in 2010 and 2011 (see Figure 1 – the numbers listed for Puntland/Galmudug are the minimum, and are likely to be have been higher).Footnote 6
Known successful hijacking attacks in Somalia (with Puntland/Galmudug and Somaliland) (2005–2012).
From strictly economic indicators, it is not obvious that Puntland should have been the main piracy haven, while Somaliland was destined to be largely free of piracy. One way of measuring the two regions’ relative economic vitality is through exports of key commodities – which throughout Somalia means cattle – from each of their main ports. While Somaliland generally has higher absolute exports of cattle than Puntland (Figure 2), the period of time when piracy was at its height (2007–2011) was a period in which Somaliland and Puntland actually exported approximately the same number of cattle. Somaliland did not begin pulling away from Puntland in cattle exports until 2009, when piracy was already nearing its height. Moreover, Somaliland's cattle exports, while higher than Puntland's, are also more volatile than Puntland's, which were relatively stable throughout the entire period under study. Similarly, camel exports in Puntland and Somaliland were both approximately the same in the lead-up to the rise of piracy in 2007 to 2009 (Figure 3), and Somaliland began to pull away from Puntland in exports at the height of piracy in Puntland, suggesting that the rise of piracy preceded problems in Puntland's export sector. In fact, recent analysis has argued that the injection of ransom money from piracy hijackings into Puntland's economy actually led to Dutch Disease-like symptoms, with inflation in the prices of consumables, and a relative decrease in productive exports like cattle (notably, Puntland exports of cattle stayed relatively flat across the whole time period, while Somaliland exports increased from 2009 onward) (Oliver et al. Reference Oliver, Jablonski and Hastings2017). Thus, any deficiencies in Puntland's economy relative to Somaliland's may have been caused by the rise of piracy, not the other way around. This suggests relative economic strength in Somaliland is unlikely to be the primary cause of the lack of piracy emanating from the region.
Exports of cattle from Somaliland (Berbera) and Puntland (Bosaso) (1996–2015).
Exports of camels from Somaliland (Berbera) and Puntland (Bosaso) (1996–2015).
Income from remittances, another potential source of economic well-being, is difficult to track between regions with any accuracy because most available estimates cover all Somali territories (that is, Somaliland, Puntland and southern Somalia) together. One study that did disaggregate between Somaliland and Puntland suggested that Somalilanders received more on average: ‘in Somaliland, 66 percent [of recipients surveyed] received between $1000 and $6000 [annually]. In Puntland, 39 percent received between $1000 and $6000; the remainder of Puntland respondents received less than $1000’ (FSNAU 2013: 15). At the same time, however, the study found that the majority of transfers made are to urban households, rather than rural ones, suggesting that this money is less likely to be sent to the areas of the coast (whether in Somaliland or Puntland) that have been most affected by piracy. Moreover, the sample of rural villages surveyed did not include any coastal towns, further limiting the amount that one can confidently assume about income from remittances in piracy-affected areas (FSNAU 2013: 5).
Geography and structural factors
Second, geography and structural factors could help to explain the low level of piracy emanating from Somaliland: its coastal areas are less populated than Puntland's, and offer less mountainous terrain for refuge; the coastline is also considerably shorter than Puntland's is, and is closer to the state of Djibouti, which has taken a role in managing piracy. Coastline length and mountainous terrain have both been used as control variables in quantitative studies of conflict incidence (Fearon Reference Fearon1999; Fearon & Laitin Reference Fearon and Laitin2003; Daxecker & Prins Reference Daxecker and Prins2015a , Reference Daxecker and Prins2015b ). Yet, pirates themselves do not usually hide in mountainous terrain, given that they generally anchor their ships on the coast. Since the ships themselves are at known locations and are highly visible during negotiations, it is the hostages that prevent pirates from being overrun by international navies and law enforcement, not hiding per se. Since the pirates operate reasonably openly from known locations in Somalia, it is again not clear that the length of the coastline is particularly important. They are not using the entire coastline, but specific points along it. The entirety of Somaliland's coastline is also closer to international shipping lanes than is Puntland's coastline. It is therefore plausible that Somaliland's proximity to Djibouti (with its foreign military bases) may have played a role in the limited instances of piracy, but when looking strictly at the geography of piracy, there is no obvious reason why Puntland should have been more physically hospitable to pirate groups than Somaliland.
In fact, the vast majority of the initial attacks in the great wave of Somali piracy that began in 2005 (before the arrival of foreign naval escorts on the scene) were in the Gulf of Aden, with many of the attacks occurring closer to the Somaliland coastline than to the Puntland coastline. The pirate bases themselves – the ports from which the pirates departed to stage attacks, and the areas where they took the ships after they were hijacked – were mostly along the eastern coast of Puntland, and down into central Somalia (Galmudug) rather than along Puntland's northern coast directly on the Gulf of Aden (World Bank Group 2013: 145–6). While these patterns changed as naval patrols and protected convoys in the Gulf of Aden pushed the attacks out into the Arabian Sea and Indian Ocean, Somaliland continued to see very few pirates launching from territory that it claimed (World Bank Group 2013: 144–5). In other words, at least in the initial years of Somali piracy, it would have been logistically easier for pirates to operate from Somaliland, or even northern Puntland, but instead it was the east coast of Puntland that saw the majority of the attacks.
State capacity
Third, it could be that Somaliland had higher institutional capacity – coastguard, police, intelligence apparatus, court system, etc. – than Puntland, and thus was able to prevent maritime pirates from gaining a foothold in its territory. There is certainly an upper limit on state capacity beyond which it is difficult for pirates to operate: Western Europe, North America and North-east Asia see little piracy, for instance, and the vast majority of sophisticated piracy attacks take place in countries below the 60th percentile in World Bank governance indicators (Hastings Reference Hastings2009). Weak state institutions that can either be co-opted by criminal elements, or that are too weak to stop criminal elements from operating, have been associated with increased criminality and other social ills: drug trafficking, for instance, increases in states with weaker institutions (Rabasa & Chalk Reference Rabasa and Chalk2001), while some studies have found that such states produce more terrorism (Piazza Reference Piazza2008). Daxecker & Prins (Reference Daxecker and Prins2013, Reference Daxecker and Prins2015b ) find specifically that lower state capacity is associated with an increase in piracy incidence, positing that state institutions that lack effective control over their territory or sea provide markets for pirates to use, sanctuaries from which pirates can operate, unemployed workers who can be recruited for piracy, and underpaid and corrupt state officials who may conspire with pirates. They are also unable to stop foreign poaching, which can drive local fishermen into penury (and thus, perhaps, piracy).
Yet there may also be a lower limit as well, at least for sophisticated pirate attacks. Coggins (Reference Coggins2011) argues that Puntland in fact was one of the more stable, less conflict-prone parts of Somalia, which provided relative safety. The World Bank Group (2013: 142–4) and Percy & Shortland (Reference Percy and Shortland2013) both suggest that there is a ‘sweet spot’ for pirates in terms of elite fragmentation (which suggests weak or compromised state institutions): if elites are unified against piracy, then there are far fewer opportunities for pirates to co-opt, and elites likely have the capacity to crack down on piracy. If elites are too fragmented, there are fewer elites worth co-opting by pirates: any elite supportive of pirates would be unlikely to be able to deliver security to the pirates over the long term due to the constant fighting (World Bank Group 2013). Instead, elites need to be fragmented enough that they can be bought (or fragmented enough that the central state is unable to move against the pirates), but not so fragmented that co-opted elites cannot deliver at least some security. In support of this idea, Coggins (Reference Coggins2016) finds that states with governance levels approaching anarchy do indeed produce more piracy, although not sophisticated piracy. That is, states with a semblance of political stability and corruptible elites tend to produce a larger sophisticated piracy industry, while states with weak institutions and less-corruptible elites are more likely to stave off piracy.
Daxecker & Prins (Reference Daxecker and Prins2015b ) likewise find some evidence that pirates tend to attack farther away from centres of state power as state institutions become stronger, and are comfortable attacking closer to central power in states with weaker formal institutions. Analysts studying terrorism have come to similar conclusions: states that are incapable of exercising public authority may not be particularly useful to international terrorists precisely because the necessary infrastructure is so poor that terrorists and civilians alike are caught up in the ambient violence and have trouble resourcing their operation, as al-Qaeda found during its foray into Somalia in the early 1990s, when it had to fend off attacks and faced high costs in transportation and communication. Moreover, the fragmentation of elites makes it almost impossible for a violent actor to co-opt enough groups to establish a foothold and fend off attacks from other groups (Harmony Project 2007).
Taken together, the literature suggests that at least two conditions help pirates involved in kidnappings for ransom thrive. First, a relative lack of conflict provides some measure of safety and stability for piracy operations (and ensures that elites are not so fragmented as to be unable to provide political cover for pirates). Second, state institutions may be too weak to stop pirates from operating in their territory, either because of a lack of coercive capacity or (relatedly) because of a fragmentation of elites that allows pirates to co-opt some elites to operate.
For the period under investigation (2005–2012), southern Somalia (centred around Mogadishu) had weak state institutions and fragmented elites, but also had a level of conflict that was high enough to make it dangerous for pirates to engage in sophisticated, long-term operations (World Bank Group 2013: 143–4). Galmudug, Puntland and Somaliland all had relatively low levels of conflict and also weak central state institutions incapable of stopping piracy by themselves, but local elites that were strong enough to provide cover to piracy (if co-opted). They should theoretically all have produced piracy, but since only Somaliland did not produce significant piracy, further explanation is required. In the face of a lack of state institutions that are sufficiently strong to stop pirates, a dominant anti-piracy narrative adhered to by both elites and the general population can make the environment intolerant of piracy. Unity of elites does not necessarily equate to strong, functioning state institutions; it can also be associated with a narrative that strongly condemns piracy, even in the face of weak institutions. But only Somaliland had a narrative of otherness from the rest of Somalia (and specifically from Puntland) that meant the general population and local elites were largely unified in their intolerance of maritime piracy in their midst.
SOMALILAND'S INSTITUTIONAL CAPACITY
Taken to its logical conclusion, the institutional capacity argument would seem to suggest that the lack of piracy emanating from within the territory claimed by the Republic of Somaliland is because its government has higher institutional capacity than Puntland, with the ability to crack down on pirates and the inability of the pirates to subvert fragmented elites to buy themselves political security (Bahadur Reference Bahadur2011; World Bank Group 2013: 142–8). Stig Jarle Hansen notes:
Somaliland's achievements in the struggle against piracy are amazing … Despite having a very weak coast guard service, pirate attacks in their part of the Somali maritime economic zone number less than one every two years over the last ten years. Somaliland reacts fast against rumoured pirate groups, catching pirates when they are in the process of organizing themselves. (Hansen Reference Hansen2009: 30)
Likewise, Jay Bahadur writes:
The difference [between Puntland and Somaliland] is due to Somaliland's greater political stability … Its central government can exert control over its territory in a way that Puntland's leaders, who must navigate a much more fractured clan landscape, cannot. In the south, in short, the pirates had to fear other criminals; in Somaliland, the danger came from a more traditional source: the police. (Bahadur Reference Bahadur2011: 39)
Piracy originating from Puntland does appear to have increased during the Puntland government budget crisis of April 2008, when the Puntland government essentially ran out of money, and the region's coastguard was put out of work. This led to both a lack of enforcement from Puntland as well as a move by some newly unemployed coastguards into piracy (Hansen Reference Hansen2009, Reference Hansen2011; Jablonski & Oliver Reference Jablonski and Oliver2013).Footnote 7 Pirates have themselves also sometimes justified their actions in terms of Somalia's institutional weakness, arguing that its lack of a coastguard allowed foreign fishing vessels to encroach on Somali territorial waters and plunder the country's once lucrative fishery stocks. Pirates have routinely reasoned that they were justified in attacking foreign vessels as a way of recovering some of the lost income from fishing, and of defending Somalia's territory (Bueger Reference Bueger2013; Dua Reference Dua2013; Hastings & Phillips Reference Hastings and Phillips2015).
Yet it is not clear that, whatever the problems of Puntland (and Galmudug), Somaliland's government exercises anything approaching the ability to coercively prevent piracy through its formal apparatus, even in the estimation of its own officials. The Executive Director of Somaliland's Counter-Piracy Coordination Office stated in an interview in 2015, for example, that ‘by the end of 2016 the Somaliland coastguard should have an operational capacity to police our waters … We don't yet have the capacity to do this.’Footnote 8 That is, Somaliland did not have this capacity when piracy was at its peak either. While Somaliland does have a police force, its capacity is also limited, with a high number of ghost employees, members being required to use their own weapons in the line of duty, and private citizens reporting to have greater trust in clan and community elders on matters of security than in the police or the government's statutory courts (Observatory of Conflict and Violence Prevention 2015: 30–3). If the government of Somaliland did not in fact have significantly greater coercive control against pirates than Puntland did, this creates a problem for theories of piracy incidence (or non-state violence more generally) that rely on state capacity for their basic explanation.
In the case of Somaliland and Puntland specifically, Bridget Coggins argues that Puntland had corruptible elites, while Somaliland did not. Nor did the general population of Somaliland support piracy. Taken together, this may partly explain why Somaliland did not experience the level of piracy that Puntland did. She notes: ‘Somaliland's stability did not engender a local pirate industry because its authorities were not easily corruptible and popular sentiments did not favor attacks on foreigners’ (Coggins Reference Coggins2016: 262). The suggestion that Somaliland's authorities are not readily corruptible is debated in the literature (Human Rights Watch 2009; Harper Reference Harper2012: 137), and discontent over government corruption was widely expressed by interviewees throughout the field research within Somaliland for this project. Coggins (Reference Coggins2016: 262) writes: ‘The Somaliland government was not physically strong enough to deter or counter much piracy, but it has apparently not needed to. Instead, Somalilanders’ goal to secede and join the international community took precedence over any potential short-term gains from piracy.’ This foreshadows the pertinence of ordering mechanisms other than those provided by the government, and it is to this puzzle that our argument now turns. We ask why Somaliland's elites were less co-optable than Puntland's – and why ordinary Somalilanders were less attracted by piracy – given that neither territory had sufficient institutional capacity to prevent pirates from operating, and neither was apparently much stronger economically than the other.
DEFINING PIRACY OUT OF SOMALILAND
We argue that Somalilanders have created, and continue to actively produce, a discourse about Somaliland's legitimate independence from Somalia, throughout which runs a narrative of ‘exceptionalism’ (Dua Reference Dua2011). Here, Somaliland is framed as being separate from, and more orderly than, other Somali territories. Crucially, however, that narrative of exceptionalism is preserved through its ambiguity surrounding the territory that constitutes Somaliland. The part of Somaliland that has produced a limited amount of piracy is located on the eastern coast of Somaliland (in the Sanaag region), which is also claimed by Puntland. As discussed in the following section, these incidents of piracy are excluded from the Independence discourse, essentially solving the problem of piracy emanating from within Somaliland by either ignoring it or attributing it to ‘inauthentic’ Somalilanders – that is, to Puntlanders.
While currently unrecognised, there is broad acceptance that Somaliland meets many – some say most – of the criteria for international recognition (Bryden Reference Bryden2003: 341; The Economist 2015). Somaliland's government argues that its legal case for independence is atypical because an independent Somaliland actually had sovereign recognition in June 1960 but the country voluntarily united with Italian Somalia just five days later to form the Republic of Somalia. During these five days, the Republic of Somaliland was recognised by 34 UN member states, including all five permanent members of the Security Council (Ismail Reference Ismail2006). In this view, recognition would merely reinstate Somaliland's pre-existing sovereignty without setting a legal precedent for other secessionist entities within Africa.Footnote 9
Within Somaliland, or at least the parts of it that support its declaration of independence – that is, neither the inhabitants of much of the east, nor parts of the far west – the difference between Somaliland and the rest of Somalia is often expressed through a colonial narrative. In that narrative, the divergence between Somalilanders and other Somalis is explained as a result of Britain leaving Somaliland's indigenous social and political structures relatively intact, while the Italian administration in the rest of Somalia decimated them. Jatin Dua argues that this way of framing the perceived difference between the two groups takes liberties with the historical record but that its purpose is to cast Somalilanders as holding a ‘British’ sense of respect for law and order while casting other Somalis as prone to mafia-like violence and criminality, of which piracy is a prime example (Dua Reference Dua2011). As Dua points out, the fact that piracy has not been a problem off Somaliland's coastline is used as further evidence of the exceptionalism that Somalilanders who support independence argue exists between themselves and other Somalis. In the analysis of one senior official in the Ministry of Foreign Affairs, for example, Somaliland is both exceptional for not succumbing to the temptations of piracy but also for remaining steadfast against it in the face of strong incentives to do otherwise: ‘90 per cent of the pirates are from Puntland and yet Puntland has gained about 95 per cent of the international resources available … The more peace you make, the less resources you receive.’Footnote 10 This framing seems intended to beg the question of why Somalilanders would bother refraining from such an obviously rewarding enterprise if they were not truly exceptional from other Somalis (Phillips Reference Phillips2013). Likewise, a former minister responded publicly to a comparison of peace in Somaliland and the conflict in other Somali territories with explicit reference to the apparent divergence in pirate activity:
Imagine a nightmare. Imagine that Somalia had access to the Red Sea like Somaliland has. Imagine what would have happened. Imagine where the world would have been. Somaliland has maintained their coast – that gateway to the Red Sea, that trade route – as best as it could because we wish it to be safe. We wish it to be stable. So I hope that the world will at least recognise and reward Somaliland and Somalilanders for having maintained the international waterways, free of disturbances.Footnote 11
As this quote illustrates, the narrative of exceptionalism is a core thread of the discourse on Somaliland's claim for international recognition of its independence from Somalia. Somaliland is eager to be engaged in international efforts against piracy and thereby demonstrate that it is a legitimate member of the international community. Reuters quotes the speaker of Somaliland's House of Representatives, Abdirahman Abdillahi, as highlighting the connection between Somaliland's actions against piracy and the legitimacy of its claims to statehood: ‘The passing of these [anti-piracy] laws proves that we are willing to cooperate with the international community’ (Anderson Reference Anderson2012). Referring to an agreement to transfer convicted pirates from the Seychelles to prisons in Somaliland, senior officials in the Ministry of Justice emphasised Somaliland's ‘hope’ of receiving further prisoners in the future so that they would have the opportunity to demonstrate Somaliland's reliability.Footnote 12 Such a hope underlines the way that much of Somaliland's engagement on piracy is about showing itself able to take responsibility for the by-products of the dysfunction seen to pervade the rest of Somalia. It is intended to reinforce the notion that Somaliland and Somalilanders are distinct from that dysfunction. The Foreign Minister stated in an interview in 2012 that:
None of those captured are Somalilanders – Somaliland's interest is to work with the international community to fight something that is damaging its economic interests … The pirates are mainly from Puntland and southern Somalia. We will keep them [in prison] while southern Somalia gets their institutions functioning and then we will hand them back … Somaliland is increasingly showing itself as a credible partner so that the international community will see Somaliland as a state actor that can enter into formal contracts … and become a normal member of the international community.Footnote 13
The Executive Director of Somaliland's Counter-Piracy Coordination Office described the difference between Puntland and Somaliland as follows: ‘Piracy starts on land; they need a base. Somaliland authorities, with the cooperation of the local communities are refusing pirates to have bases alongside the Somaliland coast.’Footnote 14 He suggested that the success of counter-piracy in Somaliland was due mainly to local communities informing the police of potential operations, which was done to both protect the income earned through trade at Berbera Port and as part of ‘a desire to practise [being] a state, not to allow criminals into their space and also to be part of the international community’. This was echoed by a local analyst working for an international non-governmental organisation: ‘the key thing for the [local] community is recognition and anything that will get in the way of recognition, whether it's piracy or extremist ideology, anything, I think the population is alert to what the international community needs to see to recognise Somaliland. They want to protect that.’Footnote 15 This discourse does not just reflect common beliefs about, or imaginings of, a shared past but also influences people's behaviour and shapes their perceptions of what is politically desirable and possible.
By contrast, the fragmentation and corruption of elites in Puntland and central Somalia certainly points to the lack of any particularly strong narrative that was adhered to by both elites and regular citizens. Puntland government authorities are suspected by outside observers and their own citizens of high levels of corruption (United Nations Security Council 2008; International Crisis Group 2009: 11; Murphy Reference Murphy2010: 159; World Bank Group 2013: 143, 148–9). Clan politics have played a role as well. As Majerteen sub-clans alternated in power in Puntland, piracy reportedly surged in areas controlled by out-of-power Majerteen sub-clans, inasmuch as the central Puntland government was able to crack down on piracy only by negotiating with local elites in its own sub-clan areas, but elites in other areas were (willing to be) co-opted by pirates. Puntland governments have been accused of complicity in piracy ranging from allowing pirates to operate in territory they control, to providing weapons and supplies to the pirates in exchange for a cut of the ransoms (World Bank Group 2013: 148–9). Pirate leaders such as Boyah also received notably light sentences during Puntland's crackdown on piracy, suggesting some level of co-optation of political elites (World Bank Group 2013: 160). Likewise, clan leaders, businesspeople and other elites were reportedly complicit in providing political and economic cover to pirates in central Somalia (World Bank Group 2013: 143), and Galmudug's political environment saw much less conflict than in southern Somalia, but more political fragmentation than either Puntland or Somaliland, which allowed pirates to co-opt whoever was in power at the time, whether it was al-Shabaab or local elites (World Bank Group 2013: 149–50).
DEFINING SOMALILAND OUT OF PIRACY
Despite the prominence of narratives exclaiming Somaliland's piratelessness, piracy is not entirely absent from within the territory claimed by Somaliland. The UN Monitoring Group on Somalia and Eritrea (United Nations Security Council 2010: 37–8) reported that there had been one significant pirate group operating from Las Qoray on the Sanaag coast – a location claimed by both Somaliland and Puntland, although in reality actively administered by neither. Between 2007–2009, the UNODC (United Nations Office on Drugs and Crime) confirmed that there were five instances of piracy coming from within territory claimed by Somaliland, and that there were probably three different anchorage sites (see the Somaliland numbers in Figure 1). Yemeni authorities captured the leader of this particular group, Fou'ad Hanaano, in 2009, thereby disbanding it (World Bank Group 2013: 147). The World Bank report alleges that Hanaano had received protection from members of the Puntland government as a result of his close ties to some local officials through his (Warsengeli) clan (World Bank Group 2013: 157). However, the fact that this network operated within the territory claimed by the government of Somaliland is absent from Somaliland's Independence discourse, with the apparent justification that Las Qoray is not a ‘core’ part of Somaliland, and its clan affiliations were with Puntland, rather than with the majority clan within Somaliland – the Isaaq.
A brief background to the territorial dispute between Somaliland and Puntland is necessary here: Somaliland's territorial claims are made on the basis of the old Anglo-Italian colonial borders, and the borders of ‘British Somaliland’ correspond to those now claimed by the government of Somaliland. The majority of its inhabitants are from the Isaaq clan, though the Isaaq do not form the majority in either the western region of Awdal (which borders Djibouti), or the eastern regions of Sool, Sanaag and southern Togdheer that lie towards Puntland. Puntland's territorial claims, however, rest on a ‘genealogical logic’ that includes territories inhabited by the Dhulbuhante and Warsengeli clans, which are members of the Harti clan family (Höhne Reference Höhne2015: 21). This also includes the areas of Sool, Sanaag and southern Togdheer that Somaliland claims. Puntland is vehemently opposed to Somaliland's claims of independence. Its government exercises regional autonomy as The Puntland State of Somalia but does not seek outright independence from Somalia in the way that Somaliland does.
Despite the general popularity of independence among the majority of the population within Somaliland, there is still a substantial minority that oppose it. The inhabitants of around 30% of Somaliland's territory, perhaps some 20% of the total population (Höhne Reference Höhne, Sharamo and Mesfin2011: 323),Footnote 16 reject the notion that Somaliland should be independent from Somalia, and claim that they are marginalised by the political cliques of Hargeisa (that is, the Isaaq clans that inhabit the ‘middle’ section of the country that covers Hargeisa, Gabiley, Berbera, Oodweyne and Burao). Successive government administrations have centralised power and economic development in the capital of Hargeisa and its environs, to the perceived exclusion of the eastern and western periphery. The population in the eastern areas is far less invested in the idea of an independent Somaliland than the predominantly Isaaq population of ‘middle’ Somaliland. The term ‘Somaliland’ is, therefore, not a neutral one, and is used almost exclusively by those who support its call for sovereign recognition – the majority of whom are from the Isaaq clan (Phillips Reference Phillips2016). Therefore, those who use the term Somaliland almost invariably adhere to the notion that Somalilanders are exceptional to other Somalis, and thus dismiss the notion that Somalilanders engage in piracy.
At the time that Hanaano's pirate network was active in Las Qoray, the town was part of the self-declared Maakhir State of Somalia, which existed from July 2007–January 2009, and opposed the territorial claims of both Somaliland and Puntland. Since the dissolution of Maakhir State, the area has been politically closer to Puntland than Somaliland. The political and clan-based proximity of Las Qoray to Puntland thus fits quite neatly within the Independence discourse (wherein Puntlanders are willing pirates but Somalilanders are not). Those who produce the Independence discourse attach any piracy that does originate from territory claimed by Somaliland to Puntland, thereby dismissing pirates as coming from outside the ‘real’ Somaliland – which is ‘real’ precisely because it disavows criminal activities like piracy. There is a clan-based element to this disavowal, with the majority of Las Qoray's population belonging to the Warsangeli clan, rather than to the Isaaq clan that forms the majority in ‘core’ Somaliland.
What is particularly important for the present argument, however, is the fact that these suggestions push against Somaliland's legal claim to independence, which rests substantially on the argument that independence would simply reinstate borders that were once recognised internationally. To be palatable to the international community, these borders cannot exclude citizens on the basis of their clan identity. However, to uphold the notion that Somalilanders do not engage in piracy (and that they are, therefore, exceptionally orderly) these borders are implicitly sacrificed. An anecdote from an interview with a local NGO worker illustrates this discursive border placement in practice: he was describing an incident in which he was personally involved, where a number of people from Hargeisa were kidnapped while they were working in Las Qoray. The kidnapping was an apparent attempt by the kidnappers to compel the release of some of their relatives, who were being held in Hargeisa Prison on allegations of piracy. When asked why people from Las Qoray had been involved in piracy, he responded that this was ‘because [Las Qoray] was infected by Puntland … all of its interactions are with Puntland’.Footnote 17 In other words, it is people's behaviour – not their place of residence – that determines whether they are considered Somalilanders or Puntlanders. In this way, the widely made statement that ‘all the pirates are from the South; [they are] not Somalilanders’Footnote 18 maintains plausibility. Therefore, should a person's behaviour not cohere with the discursive construction of an orderly Somaliland, they can be categorised as a non-Somalilander. In Somaliland's Independence discourse, territorial claims to its contested peripheral areas may be implicitly sacrificed in order to uphold the even more important narrative of Somaliland's inherent propensity for order, in which piracy has no place.
CONCLUSION
The competing claims to the eastern parts of Somaliland (or, alternatively, the western parts of Puntland, or simply the north coast of Somalia) highlight the ambiguity surrounding political authority in the area. This allows Somalilanders to plausibly claim it when convenient, and exclude it when not. This malleability helps to uphold the narrative of Somalilanders’ ‘inherent’ orderliness in the face of evidence to the contrary. However, it also points to the ways in which powerful discourses can mobilise perceptions of order and volatility as a means of limiting criminal activity – and to the fact that they do so in the absence of a strong central government apparatus. This suggests that the management of piracy in Somaliland rested, to a significant degree, on a discourse that communicated that the practice of piracy undermined a core tenet of Somaliland's claim for independence from Somalia, more than it did on the enforcement capacity of the government's institutions. In the case of Somaliland, there existed a widespread sensitivity among both civilians and elites to the ability of piracy to perform two unwanted functions. The first was that piracy was seen as likely to alter political, economic and social life for the worse by endangering trade. The second, and more important factor, was that piracy was perceived to undermine the claim upon which the independence discourse largely rests: that Somalilanders are exceptional to other Somalis and should thus be legally independent from them.
Somaliland's convenient definition of its territory aside, this paper highlights the ability of strong identities and powerful discourses to undermine the ability of violent non-state actors from doing as they please without the conventional state-based mechanisms depicted in the failed states and state capacity literature. The reification of Somaliland as the region of Somalia that kept piracy at bay undermines arguments about the ability of strong state institutions, political stability or economic opportunities to dampen the ability of pirates to operate, because Somaliland did not necessarily have any of those advantages. More generally, this article contributes to the literature on development in areas of hybrid political authority (Menkhaus Reference Menkhaus2006/7; Boege et al. Reference Boege, Brown, Clements and Nolan2009; Renders & Terlinden Reference Renders and Terlinden2010) by suggesting that the violence and crime that are generally taken to plague areas with weak state institutions are not inevitable and cannot, therefore, be assumed to be necessarily causally linked across time and space. This piece has demonstrated that the mechanisms that provide order and limit the emergence of violent actors are far more complex and discursively re/generated than the consolidation of effective rule-based institutions. Institutions that are capable of providing order rest upon layers of meaning being reproduced over time. We have argued that it is this discursive exchange that underpins the popular legitimacy afforded to institutions, and this should, therefore, be at the forefront of efforts to understand political change beyond the formal capacity of the state.
