States of (in)security: Institutionalized arbitrariness as a mode of governance in northern Uganda

Prepared for the March 2 – 3, 2017 seminar, Theorizing (Dis)Order: Governing in an Uncertain World, organized by the winners of the 2016 – 2017 WPF Student Seminar Competition. 

In northern Uganda, where I have conducted field research on local security initiatives over the past three years, issues related to politics, power, and the state are both good and bad, present and absent, open and closed; fundamentally, they encompass multiple contradictory things simultaneously, such that they cannot be determined without obscuring their very nature. Cecilie Lanken Verma, a political anthropologist of violence and militarization in northern Uganda, describes this as “lakite”—or “somehow”. Such contradictions frequently play a central role in ethnographies of vigilante groups, gangs, self-help groups, and the like: a twilight or boundary nature; the fluidity of individual roles in the group, and group roles in the community; the ambiguity of intent. To understand these phenomena, we must orient our questions around the “how” and “why” of governance. Why is the relationship between violence and governance ambiguous in northern Uganda? What does it tell us about how the state governs the local?

Decades of research on the Ugandan state have produced theories that could explain the presence of multiple, diverse, and incoherent security and governance actors, and the unpredictability that accompanies them, under Uganda’s apparently stable National Resistance Movement (NRM) regime. The NRM, and its political leader, President Yoweri Museveni, have held power since leading a successful armed insurgency in 1986. Scholars have convincingly categorized Museveni and his NRM regime as anywhere from pathologically weak (Reinikka and Svensson 2005) or wilfully absent (Jones 2009), to a semi-authoritarian regime that relies on neo-patrimonialism (Mwenda and Tangri 2005; Freeland 2015), or indirect rule (Mamdani 1996; Finnström 2008). These theories understand the state as consolidated or unconsolidated; and centralized or decentralized. Other authors have tried to account for these apparent contradictions by describing the Ugandan state as “hybrid” (Tripp 2010) or “informal” (Khisa 2013). However, inclusive descriptors tell us little about how various logics can co-exist, and nothing about how the nature of these logics changes when they are combined. Here, I present a theory of governance called “institutionalized arbitrariness” which aims to address both concerns, explaining how the state might appear to be absent (pathologically weak or willfully absent) and present (neo-patrimonial); public (military state) and private (patrimonial), while retaining the meaning of these distinct categories. While the theory is based on the particulars of northern Uganda, it may have broader applicability, particularly in post-colonial contexts where it has been generally assumed that decolonization resulted in a rupture between sovereign violence and governing authority.

I argue that the power of Uganda’s central government, as experienced in the north of the country, relies on a system in which competing and ill-defined authorities unpredictably claim and deny the authority to intervene in matters of concern to civilians (from domestic disputes to theft to murder), thereby creating an atmosphere of jurisdictional uncertainty in the security and justice sectors (also see Tapscott 2017). This arbitrary governance is made possible by four factors, briefly elaborated in this note: the non-institutionalization of the public/private division; the perception of state control of sovereign violence; the potential presence of the state; and numerous security and governance institutions with unclear hierarchies. Together, these four factors produce an environment of the potential for seemingly arbitrary violence that makes the government ever-present in civilian imagination, despite its general material absence in terms of service provision or law enforcement.

All four factors are present in the context of northern Uganda. Although each factor is continuous in the sense that it can be more or less pronounced, we can also think of them as binary, in the sense that there is a minimum level at which they must each be present for institutionalized arbitrariness to function:

Non-institutionalized public/private division: Most theory on state formation and subnational governance focuses on ordering—how rulers organize the ruled to maximize control and extraction. A key theoretical component of ordering is the division of the world into “public” and “private” spaces: the public space is where the state directly extracts and enforces; the private space is the remainder, where it is inefficient for the state to directly do so. These divisions are socially constructed and reinforced through iterative acts of claim-making. A related consequence is the establishment of predictable terms of exchange, which allow citizens to maximize returns under given constraints, as well as mutual claim-making, often called a state-society compact.

However, in northern Uganda, the boundary between “public” and “private” spheres is fluid. As a result, citizens often are unable to develop expectations that would be necessary to accurately predict the terms of exchange between citizens and state authorities. For example, it is not unusual for a police officer to send a complainant home, explaining that the complaint would be better handled by a local authority or within the family. Cases range from petty theft through to criminal offenses like rape and murder. The fluidity of the public/private division facilitates jurisdictional ambiguity—a lack of clarity about which authorities and institutions are responsible for any given infraction. Moreover, it means that any claim could be made radically public or radically private at any moment in time.

Authorities manipulate this fluidity through a dynamic process of both claiming and denying jurisdictional authority, thereby strategically divesting themselves of responsibility to intervene. Relatedly, the ruling regime maintains its power over other claimants to public authority such as the local security actors by unpredictably shifting the boundary between public and private, and enforcing these seemingly arbitrary edicts with a very real threat of violent force. Together, these two phenomena prevent civilians from developing expectations of public authorities, thereby destabilizing and fragmenting citizen claims on the state. It also prevents the creation of alternative public authorities. Both elements help to explain the non-emergence of a Tillyean state-society contract. If those who have superior access to material and symbolic power can continually redefine their responsibilities and use their access to violence and resources to reify that assertion, there is very little that ordinary citizens can do to stop them. Moreover, if those assertions and denials of authority are irregular, citizens cannot easily organize around a given set of expectations, and this disperses the threat of resistance or rebellion (also see Tapscott 2016).

Perception of sovereign violence: The second key factor enabling institutionalized arbitrariness is the perception that the governing authority possesses sovereign violence. Drawing on Schmitt (2005), sovereign violence refers to the state’s capacity to suspend the law and use extra-legal violence to define states of exception. This is done with impunity and a will to govern. Where for Schmitt, sovereign violence is characterized by the power to define the exception, here I focus on the ability to continually redefine the exception, such that sovereign violence does not merely underpin notions of public and private, but rather continually redraws them. Thus, institutionalized arbitrariness rests on citizens’ beliefs that any moment can potentially be made exceptional at any time. This is essential: the state’s ongoing violent interventions make threats meaningful; memories of the use of violence make society cognizant of this.

For Gulu’s population, state violence is proximate and real—today’s young adults grew up during the twenty-year-long conflict between the Lord’s Resistance Army and the Government of Uganda, witnessing extreme acts of violence by combatants on both sides. Over 90% of the population was forcibly relocated to displacement camps, where they faced dire conditions and military command dictated daily life. Moreover, many northerners believe the war benefited the NRM regime, increasing donor support and helping consolidate the NRM’s power in the rest of the country. The current “post-conflict” status of the north allows the government to justify intervention as necessary to keep the peace. The state’s militarization, institutionalization of violence, and history of violent intervention against civilians, reinforces the image of the state as a force that, while generally absent, has potential to intervene with overwhelming force to define whether, when, how, and for whom law will apply. Thus, there is a commonly held perception that the regime has the capacity to create extra-legal spaces, which it can pacify through intervention or ignore.

Potential presence of the state: For institutionalized arbitrariness to function, it is necessary that citizens believe the state could be present at any moment. Otherwise, it would be possible to establish alternative sources of authority independent from the central state. One way potential state presence is achieved in Uganda is through citizens’ perceptions of state surveillance, which is bolstered by the common perception of an effective, efficient, and omnipresent state security apparatus, as well as the regime’s legacy as a no-party state. Respondents describe a sprawling security apparatus, with secret operatives “deep down” in the villages reporting directly to the President. This has also been documented by numerous scholars (Branch 2011; Finnström 2008; Jones 2009; Mamdani 1995; Lanken Verma 2012; Zeller 2013). The fears of an extensive spy network may be warranted: in 2015, Privacy International published a series of documents confirming that the GoU had purchased an intrusion malware in 2011 to facilitate spying on opposition politicians, media, and establishment insiders in hotels, key government institutions, people’s homes, and other meeting spaces with the explicit goal of blackmailing targets and members of the political opposition (Privacy International 2015).

The perception that the state is potentially always present contributes to feeling among citizens that the state’s actions (and non-actions) are intentional. After all, if state authorities are informed and have the technical capacity to act, then inaction can be interpreted as equally purposeful to action. Relatedly, potential state presence means that any place, person, or event can be made “public” at any time. Equally so, if state authorities opt to intervene, the perception of potential state presence leads citizens to attribute a logic to the intervention (see also Bozzini 2015). Thus, potential state presence contributes to an environment where citizens reasonably view state actions as intentional, targeted, and potentially imminent, thereby incentivizing citizens to modify their behavior to approximate what they believe state authorities want.

Non-hierarchical and fragmented governance institutions (security sector): Uganda’s security and justice sectors have long been fragmented, even more so under Museveni’s leadership. In practice, these agencies have similar roles and responsibilities and their personnel are frequently reshuffled. One local politician who worked closely with the community policing program in Gulu explained, “The government has brought all these [auxiliary forces] to confuse the community. Sometimes these people [crime preventers, police, etc.] work in parallel. You don’t understand who is who” (Interview, 16.09.15). Similarly, under the NRM regime, Uganda’s administration has been increasingly fragmented through decentralization. Since Museveni took power in 1986, the number of districts in Uganda has tripled from 33 to 111 in 2016. Redistricting has strengthened the regime’s patronage system, allowing distribution of new government jobs. It has also allowed fragmentation of alternative units of political organization, dividing potential opposition to the NRM regime by bolstering ethnic divisions and local conflict, with smaller ethnic groups making requests for their own districts.

These public authorities lack clearly defined mandates, as they are all subject to the constantly shifting public/private divide. Thus, they must compete among each other for limited resources, striving to define their role and build legitimacy with the community, while constantly keeping in mind the active possibility that the central state will violently intervene in their nascent zone of authority. This strategy is extremely effective to preclude the establishment of shared goals and expectations, thereby preventing opposition and civilian organization. It similarly limits the costs of the government, which does not have to expend resources necessary for direct rule (for example, through military occupation), nor cede power as in indirect rule (for example, to subnational authorities).

Second, assume that the logic at play is in fact neo-patrimonialism, and that the explanation for what I have described as “unpredictability” is actually various authorities doling out resources to their constituents and sidelining others who are not among their supporters. In such a scenario, abiding by rules of the patrimonial system results in predictable rewards; similarly, breaking rules of the system must result in predictable costs. However, this is not the case—while the system looks similar to a neo-patrimonial regime in the rules, it does not abide by this logic in the exception, and thus, we can deduce that something else is (also) going on. For example, in northern Uganda, the police often fail to intervene in issues of security—indeed, youth security groups frequently make arrests within the village and take the suspect to the police post. At that time, as is often reported, exchanges take place—bribes are paid, favors called in, deals cut. However, at times, the police do intervene, arriving on the scene to arrest, detain, and bring criminal charges against community members, auxiliary forces, and the like. These interventions are experienced as arbitrary and unpredictable.

Third, civilians purport that they also have difficulty navigating the system. A common refrain among my respondents was “government things have to be complicated” or “with government policy, nothing should be straight forward.” One NRM district level politician explained that the NRM had infiltrated opposition parties: “If young get three [members of an opposition party], one is NRM. He’s just confusing them.” Another NRM political operative reflecting on a community policing program described it as “confused.” One participant further described a training in which he and others were instructed to confuse voters on Election Day:

“When someone has come [to the polling station] and maybe you don’t get his name, send him to another polling station—[tell him] ‘you first go and check there in a good manner, don’t be rude. Maybe your name is not here, because there is some mistake from the Headquarter. For us, we don’t know. We are just guarding.’ That one is going to make a lot of confusion. People will be very many. You will be confused. They are going to stop [voting] at 4pm exactly. They will confuse you till four. And we [security] will say ‘sorry, sorry, sorry, sorry.’” (Interview, 09.02.16).

If, as many have suggested, a logic of neo-patrimonialism truly determined interactions, confusion should not be such a common experience; instead, uncertainty and confusion appear fundamental to citizens’ experiences of the NRM state. Alternative logics (such as neo-patrimonialism) may determine individual experiences or even dictate most interactions with a given authority—however, the nature of this logic is fundamentally altered given that it can be overturned unexpectedly at any time.

A theory of institutionalized arbitrariness contributes to existing literature in that it describes a relationship among the “hybrid” aspects of governance, and shows how disorder can function as an efficient mode of governance. Thus, the NRM state is in neither centralized nor decentralized, and rules neither directly nor indirectly—instead, it has maintained control of the peripheral areas of the country primarily by injecting layers of uncertainty into civilian interactions with state authorities. This is achieved through destabilizing the public/private boundary, thereby allowing the state to constantly re-define actors, actions, institutions, and claims as public or private; as well as fluidity between states of normalcy and exception, which allows the threat of exceptional violence to loom behind all interactions with state authorities. This unpredictable and violent state intervention causes citizens to perceive the state as potentially always present, despite its general material absence, and thus constitutes an efficient mode of governance that allows for control without requiring physical occupation of territory or provision of public goods.

This theory helps to explain observations made by a host of academics and practitioners of Uganda, who have described the Ugandan state as using “arbitrariness and unpredictability” to restrict media workers (Tripp 2004, 12); as striking a “delicate balance between formal and informal” by fragmenting “political power at the centre…[,] ceding […power] to unviable local units, [and] the creation of occasional fear and insecurity” (Khisa 221); as akin to “the dry season rains—something occasional and potentially destructive” (Jones 2009, 3); as creating “seemingly deliberate confusion” around civil militias (Janmyr 212); as “produc[ing] ‘security’ and ‘insecurity’ simultaneously in a constant aporetic relationship” and as having “ambiguity or double-faced meaning of things” (Lanken Verma 2012, 57); as well as the president’s “tendency to keep things around him as disorganized as possible to avoid the formation of any ordered arrangement that might possibly be turned against his own, personal raw power” (Carbone 2008, 29; see also Bayart 1993). Despite the emphasis these scholars place on fragmentation, arbitrary and unpredictable intervention, and resultant uncertainty, they stop short of understanding these realities as part of a broader system of governance.

Works Cited

Bayart, Jean-François. 1993. The State in Africa: The Politics of the Belly. London: Longman.

Bozzini, David. 2015. “The Fines and the Spies: Fears of State Surveillance in Eritrea and in the Diaspora.” Social Analysis 59 (4): 32–49.

Branch, Adam. 2011. Displacing Human Rights: War and Intervention in Northern Uganda. Oxford: Oxford University Press.

Carbone, Giovanni. 2008. No-Party Democracy? Ugandan Politics in Comparative Perspective. Boulder: Lynne Rienner Publishers.

Finnström, Sverker. 2008. Living with Bad Surroundings: War, History, and Everyday Moments in Northern Uganda. Duke University Press.

Freeland, Valerie. 2015. “Rebranding the State: Uganda’s Strategic Use of the International Criminal Court.” Development and Change, 1–27.

Jones, Ben. 2009. Beyond the State in Rural Africa. Edinburgh: Edinburgh University Press.

Khisa, Moses. 2013. “The Making of the ‘Informal State’ in Uganda.” Africa Development 38 (1&2): 191–226.

Lanken Verma, Cecilie. 2012. “Guns and Tricks: State Becoming and Political Subjectivity in War-Torn Northern Uganda.” Department of Anthropology, Denmark: University of Copenhagen.

Mamdani, Mahmood. 1995. “And Fire Does Not Always Beget Ash: Critical Reflections on the NRM.” Monitor.

———. 1996. Citizen and Subject – Contemporary Africa and the Legacy of Late Colonialism. Princeton: Princeton University Press.

Mwenda, Andrew, and Roger Tangri. 2005. “Patronage Politics, Donor Reforms, and Regime Consolidation in Uganda.” African Affairs 104 (416): 449–67.

Privacy International. 2015. “For God and My President: State Surveillance In Uganda.”

Reinikka, Ritva, and Jakob Svensson. 2005. “Fighting Corruption to Improve Schooling: Evidence from a Newspaper Campaign in Uganda.” Journal of the European Economic Association 3 (2–3): 259–67.

Schmitt, Carl. 2005. Political Theology: Four Chapters on the Concept of Sovereignty. Translated by George Schwab. Chicago: University of Chicago Press.

Tapscott, Rebecca. 2016. “Local Security and the (Un)making of Public Authority in Gulu, Northern Uganda.” African Affairs 116 (462): 39–59. doi:

———. 2017. “The Government Has Long Hands: Institutionalized Arbitrariness and Local Security Initiatives in Northern Uganda.” Development and Change 48 (2): 263–285.

Tripp, Aili Mari. 2010. Museveni’s Uganda: Paradoxes of Power in a Hybrid Regime. Boulder, CO: Lynne Rienner Publishers.

Zeller, Wolfgang. 2013. “Get It While You Can: Governance between Wars in the Uganda-South Sudan Borderland.” In Violence on the Margins: States, Conflict and Borderlands, edited by Benedict Korf and Timothy Raeymaekers. Palgrave Macmillan.

[i] In contrast to Chabal and Daloz, disorder here connotes uncertainty, unpredictability, and arbitrariness, rather than a latent order based on personal relations that is simply unintelligible to western eyes.

Stay Connected