Critical African Studies ISSN: 2168-1392 (Print) 2040-7211 (Online) Journal homepage: www.tandfonline.com/journals/rcaf20 Decentralization and territorial politics: the dilemma of constructing and managing identities in Uganda Morris Adam Nsamba To cite this article: Morris Adam Nsamba (2013) Decentralization and territorial politics: the dilemma of constructing and managing identities in Uganda, Critical African Studies, 5:1, 48-60, DOI: 10.1080/21681392.2013.774835 To link to this article: https://doi.org/10.1080/21681392.2013.774835 Published online: 11 Mar 2013. Submit your article to this journal Article views: 335 View related articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rcaf20 https://www.tandfonline.com/journals/rcaf20?src=pdf https://www.tandfonline.com/action/showCitFormats?doi=10.1080/21681392.2013.774835 https://doi.org/10.1080/21681392.2013.774835 https://www.tandfonline.com/action/authorSubmission?journalCode=rcaf20&show=instructions&src=pdf https://www.tandfonline.com/action/authorSubmission?journalCode=rcaf20&show=instructions&src=pdf https://www.tandfonline.com/doi/mlt/10.1080/21681392.2013.774835?src=pdf https://www.tandfonline.com/doi/mlt/10.1080/21681392.2013.774835?src=pdf https://www.tandfonline.com/action/journalInformation?journalCode=rcaf20 Decentralization and territorial politics: the dilemma of constructing and managing identities in Uganda Morris Adam Nsamba* African Research and Resource Forum (ARRF), Nairobi, Kenya (Received 24 October 2012; final version received 10 December 2012) By examining the process of internal boundary adjustment in the Karamoja region of Uganda, this paper argues that internal boundary adjustments in Uganda have been pursued along ethnic lines, leading to a territorial kind of citizenship and an ethnic ‘containerization’ of communities. Based on an examination of the 2001 and 2011 presidential elections in Uganda’s Acholi and Lango regions, the paper further argues that internal administrative boundary adjustments in Uganda have a political usage and meaning that goes beyond the official rhetoric of administrative efficiency used to justify these reforms. Internal boundary adjustment does not, however, always lead to the political ends intended by the ruling party. Both the National Resistance Army/Movement (NRA/M) government and opposition parties have attempted, and benefitted from, internal boundary adjustments. The empirical data presented are based on official reports and documents from the Government of Uganda, Electoral Commission election reports, and a review of published books, journals and newspapers articles. Keywords: boundary adjustment; political power; ethnicity; National Resistance Army (NRA); National Resistance Movement (NRM); Uganda Introduction Although internal boundary adjustment under the policy of decentralization in Uganda is justified on the basis of democracy and administrative efficiency, it has resulted in the creation of both ethnic districts and a territorial regime of citizenship and rights. Internal boundary adjustment has also offered structures for political mobilization at local levels especially, particularly where the district structures are captured by political elites. This paper examines the process of internal boundary adjustment in Uganda. It uses the Karamoja region of north-eastern Uganda as an example to illustrate the extent to which internal boundary adjustment has created ethnic districts, furthering ‘divide and rule’ techniques for the benefit of existing political elites. It also examines the 2001 and 2011 presidential elections in the Acholi and Lango regions to demon- strate how internal boundary adjustment has been used, beyond the official rhetoric of adminis- trative efficiency. The process of extending the reach of the state apparatus through decentralization has gone hand in hand with increasing calls1 (Carbone 2001) for separation and recognition by different ethnic groups in their attempts to address a legacy of socio-economic © 2013 Centre of African Studies, University of Edinburgh *Email: mnsamba@arrforum.org; morrisnsamba@gmail.com This article was originally published with errors. This version has been updated. Please see Corrigendum (http://dx.doi.org/10.1080/ 21681392.2013.786908) Critical African Studies, 2013 Vol. 5, No. 1, 48–60, http://dx.doi.org/10.1080/21681392.2013.774835 injustices and marginalization. These demands and pressures reflect an organic process of social mobilization, amidst politics to which the state responds with two forms of justification: first, by justifying and explaining the demands as a sign of democratic growth; and second, in terms of delivering development. Using these two reinforcing justifications the state has managed to present this ‘balkanization’ as part of the democratic and developmental agenda, effectively derailing internal and external scrutiny of the entire internal boundary adjustment enterprise. In 1993, as part of structural adjustment reforms recommended by The World Bank, and internal political desires to address a legacy of socio-economic imbalance and a history of politi- cal turmoil and injustices, Uganda initiated a rigorous decentralization policy. Its stated aims were: (1) transferring real power to local government in order to reduce the workload on remote and under-resourced central officials; (2) bringing political and administrative control over services so that they are actually delivered, and thereby promoting a degree of popular par- ticipation; (3) improving financial accountability and responsibility by establishing a link between the payment of taxes and the provision of services; and (4) improving the capacity of local gov- ernments to plan, finance and manage the delivery of services to users (Kauzya 2007). The policy of decentralization thus articulated administrative rather than the political objectives which have since become visible in its application. These political objectives can be identified through careful scrutiny of the political rationale and process of internal boundary adjustment. Post-independence governments in Uganda have adjusted internal boundaries, although the magnitude of these efforts and the reasons given for them have differed from one government to another. They include, but are not limited to, the correction of colonial internal cultural and boundary distortions (The Daily Monitor June 1, 2009), and the promotion of democracy. When the current decentralization policy was launched the country was divided into 38 districts. Since then the number has increased to over 100. One thread has been shared across each phase of district creation. Districts have been created whenever the country is preparing for an election, or after an election specifically to fulfil the campaign pledges. Although the National Resistance Army/Movement (NRA/M) has over time exhibited great enthusiasm in creating districts – justi- fied as vehicles for social service provision and participatory democracy – it has hardly managed to tame the thriving elite capture and corruption that has occurred within these structures (Golola 2001; Government of Uganda 2006; Muhumuza 2008). The post-independence boundary alterations and their implications for ethnic identity and political power The principle of local administration in colonial Uganda was based on the philosophy that prestige and influence of the indigenous authorities could best be upheld by letting peasants see that the government itself treated them as an integral part of the machinery of administration, based on their ‘tradition’ and ‘custom’ (Gore 1935). District councils were given corporate powers and responsibilities in those areas where no agreements had been signed (Karugire 1980), although recognized indigenous authorities in districts had to act within the parameters of accepted policy (Mamdani 1996; Richard and Sheridan 1975). Acting within accepted policy meant that districts were to be ‘pure’ in ‘culture, custom and tradition’. The 1949 ordinance spelt out clearly that dis- tricts were to be composed of one ‘tribe’. As such, in the process of determining district boundaries, much care was taken to include, wherever possible and practical, one ethnic group in each district (Karugire 1980). To post-colonial governments this situation was unattainable in some places because it marginalized many communities, especially the numerically small ones. Table 1 indicates the extent to which each government has adjusted the internal boundaries of Uganda by region. Nevertheless, post-colonial boundary adjustments did not really seek to reform the internal colonial boundaries. The new subdivisions that have happened have taken place within the Critical African Studies 49 Table 1. Number of new districts created by the regime (per region and in total) in Uganda since 1962. Region/ province Number of independence subdivisions 1966–1971 (first Obote government; first Republican Constitution) 1972–1979 (Amin government and UNLF government) 1986–1996 (first NRA/M government under the Republican Constitution of 1967) 1996–2001 (first NRA/M government under the 1995 Constitution) 2001–2006 (Second NRA/M government under the 1995 Constitution) 2006–2011 (First NRM-O government under a multi-party political system) Total Acholi 1 1 1 1 3 7 Ankole Kingdom 1 1 1 3 4 10 Buganda 1 3 2 3 4 3 7 23 Bugisu 1 1 1 2 1 6 Bukedea 1 1 1 2 1 6 Bunyoro Kingdom 1 1 1 1 1 5 Busoga 1 2 1 2 2 3 11 Karamoja 1 1 1 2 2 7 Kigezi 1 1 2 4 Lango 1 3 3 7 Madi 1 1 Sebei 1 1 1 3 Teso 1 1 2 2 2 8 Toro Kingdom 1 1 2 2 2 8 West Nile 1 1 1 2 1 6 Total 15 5 9 10 18 24 31 112 Source: Republic of Uganda (1990). Note: UNLF, Uganda National Liberation Front; NRA/M, National Resistance Army/Movement; NRM-O, National Resistance Movement-Organisation. 50 M .A . N sam ba bigger colonial boundaries. For example, there is no division – district – created out of territory annexed from two different colonial subdivisions. As Table 1 indicates, Buganda as a region has witnessed the highest number of subdivisions. This region has been central to the political history ofUganda, being the seat of government, home to its largest ethnic community and recipient of the majority of its economic development. Each post-colonial government has had to court this region, or at least struggled to win its support. From Table 1 it is clear that under Yoweri Museveni (President since 1986), the total number of districts created has increasedwith each successive elec- toral cycle. Obote 1, the Idi Amin regime and the politics of boundaries It is not the intention of this paper to review exhaustively all the internal boundary changes made by the Milton Obote and Idi Amin governments, or even to discuss their politics, since it is well discussed in the literature. They will only be discussed briefly for the purposes of illustrating the extent to which internal boundary adjustment was used to further the political interests of these two first post-independence governments. The first alterations of internal boundaries within Uganda took place under the Obote government, which divided Buganda into four districts. The division, an outcome of the starting point for confrontation between the Buganda government and the central government in 1966, also mirrors the manner in which the central government responded to what has become known as ‘the Buganda question’. Through the constitutional crisis of 1966 and the subsequent promulgation of the 1967 constitution, the Obote government viewed Buganda as agitating for secession. Obote and the Kabaka (or ‘King’ of Buganda) had personal differences and interests. Each of them wanted to be Uganda’s president, yet neither fully understood how a parliamentary system operated. These misunderstandings, disagreements and conflicting interests led to the criminalization and demonization of Buganda as a region by Obote’s government, and dividing it into four districts was part of a process of reducing Bugan- da’s influence and power in the body politics of Uganda. In 1971 Amin justified his military coup, in part, on the basis of how Obote dealt with the Buganda question. To some sections in Buganda, he was viewed as a saviour in the initial years. Although Amin subdivided and recognized the internal boundaries, he did not substantially change the outlook of the internal boundaries. If it had been the colonial state that had given birth to a politics of territorial identity within Uganda, then both Obote and Amin reconstructed that poli- tics in a manner that radiated distrust in ethnic politics, but did not do away with it. During Obote’s government, Buganda was branded an enemy of the state. Although Amin mostly favoured people fromWest Nile for recruitment into the military, he did not subject any region to demonization. His demonization was individual not communal (Kasozi 1999; Mamdani 1996). Both governments ‘entered into marriage’ with individuals from different ethnic and political groups as a means of securing and maintaining political power. None of the groups felt bound to resist the terrorizing arm of the state, unless that terror happened within its territory or targeted its own group. These techniques of divide and rule and the highly territorialized form of citizenship it engendered were so deep rooted that no ethnic group imagined that terror as a form of rule could overflow to everyone at the same time. Both Obote and Amin’s governments construed the threat to their political power as one of ethnic identity, the solution therefore called for the isolation and containment of specific ethnic groups, with the help of other groups. Museveni’s districts: carving out space for a little known force After capturing state power in 1986, the NRA/M was faced with the problem that it lacked legiti- macy in most parts of the north and east of Uganda because it was perceived to be a southern Critical African Studies 51 military government. It is true that the political elites of the previous governments had come from the north and east. The colonial state had created a north–south divide that too was reinforced by all post-independence leaders, including those of the NRA, in the initial years of their rule. The core of the rank and file of the NRAwere mostly people from the south, yet previous governments and their militaries had mostly recruited from the north – Lango and Acholi – and West Nile. But the composition of the NRA guerrilla movement was more a reflection of its historical base of the NRA than its ideological disposition. Although it mostly comprised of communities from the south, it was not the policy of the NRA to exclude on the basis of ethnicity. Indeed, it responded to the link between ethnicity and political participation by introducing reforms that sought to dis- entangle territory from political rights – in particular voting rights. It pegged voting to residence rather than ethnic territory, including banning political party activities if those parties were con- sidered ethnically based. While this offered an array of hope to the Ugandans, it was short lived. The NRA considered itself as a movement to which every person in Uganda belonged. However, like any political movement, when it was exposed to competitive politics it became increasingly difficult to maintain its ideological principles. Its rhetoric was that of a movement, but increasingly the manner in which it functioned exhibited more the tendencies of a political party (Mamdani 1995; Mugaju and Olukoshi 2004). During the 1993 constituent assembly elec- tions the NRA/M was faced with a real political challenge to its power from the older political parties, in particular the Democratic Party (DP) and Uganda People’s Congress (UPC). The NRA/M responded with a solution which not only listed 56 indigenous communities in Uganda, but also territorialized those communities just as the colonial state had done. Under the tutelage of The World Bank and the International Monetary Fund (IMF), as well as internal press- ures, the NRA/M devolved power to the districts – local administrative units. This reduced resources at the disposal of the central government, limiting its ability to keep up the existing patron–client leadership model. But the new reforms suffered a theoretical deficiency. The World Bank and IMF argued that decentralization of power would increase citizen participation in politics. But politics was defined to mean elections, without interrogating the decision-making dynamics inherent to decentralization (Muhumuza 2008). In addition, local governments were treated as though they were politically free and totally detached from central government politics. Hence, over time the political elites in Uganda have utilized the political opportunities and incen- tives within the structures and policy of decentralization to create and increase their political capital, utilizing both state and non-state resources and structures. One of those opportunities is the creation of districts under the rubric of local democracy. In 2005, during the launch of Nakaseke district, Museveni eloquently outlined the challenges and opportunities of this local democracy. He observed that ‘although a new district takes a lot of money, this is the democracy we fought for. People must ask for what they want and get’ (The Daily Monitor August 8, 2005). By imprinting it as democracy, local and national political elites have been able to justify and legitimize their actions of redrawing boundaries and managing identities within a particular ter- ritory. It is, however, imperative to distinguish between state policy and the interests of the indi- vidual political elites. State policy – mostly represented by the bureaucrats in the ministries and districts – views district creation in an administrative manner, whereas individual political elites mostly view the created structures in terms of the political opportunities they create. The present author’s interest is the actions and inactions of the political rather than the bureaucratic elites. Internal boundary adjustment has afforded political elites the opportunity to maintain patron– client relations, albeit using government funding and structures, with limited scrutiny from the development partners and the citizens. Where internal boundary demarcation has been based on ethnicity, it has led to divisions. First is a division between the rural and urban in a manner prescribing culture for the former and politics for the latter (Mamdani 1995). The second division has happened within the territories created at the local level, separating those considered 52 M.A. Nsamba ‘indigenous’ and those considered non-indigenous settlers (Nsamba 2009). Indigeneity in this regard is considered culturally permanent, turning cultural identities into permanent political iden- tities. While cultural identities are adaptive to change, political identities are not (Berman 1998). Despite the argument that politicizing cultural identities in Uganda is intended to keep cultures ‘pure’ (The Daily Monitor June 1, 2009), the real political purpose is to disrupt social movements that challenge central government by dividing up ethnic groups. Karamoja region in north-eastern Uganda offers a clear picture of how boundary changes have created ethnic ‘containers’ in Uganda. Karamoja region: ethnic compartmentalization Internal administrative divisions in Karamoja are punctuated by the desire to create ethnic and sub-ethnic compartments, in part justified on the basis of security and development. Karamoja has long experienced cattle rustling as a particular cultural form associated with pastoral societies, which over time has attained a more economic motive. This has changed the methods and objec- tives of rustling. While in the past it was seasonal, planned and constrained by clearly known rules of engagement/war, in the recent past the introduction of the AK-47 guns has transformed rustling into economic raiding. This has had impacts upon and limited socio-economic development in the region, as most of the time the population is in fear. With a history of colonial marginalization the socio-economic situation in Karamoja went from bad to worse. Both colonial and post-colonial governments have long underrated pastoralism as a form of livelihood. Part of the process of changing and contributing to socio-economic development in the region has included sedentari- zation of pastoralists through district creation and livelihood zoning. For present purposes the paper restrict itself to district creation not livelihood zoning. Table 2 shows boundary changes in the region. Table 2 indicates a trend common to districts creation in Karamoja; they have all been created when Uganda is preparing for an election. But as indicated, district creation has resulted in ethnic compartmentalization in the region, with each dominant ethnic community getting its own district. Table 3 shows the dominant ethnic communities by district in Karamoja region. This internal ‘balkanization’ of Karamoja has succeeded in creating ethnic and sub-ethnic containers for each community in the region. Each district and often sub-counties now represent a dominant ethnic community. In 2006, Nakapiripirit district council passed a resolution for the creation of Amudat district and Figure 1 summarizes the key justifications as presented to parliament, which approved the creation of Amudat district on the same basis. If ethnic ‘purity’ was one of the key reasons for internal boundary adjustment during the colo- nial period, the NRA/M government used the same rationale to create Amudat district in Karamajo. Compared with all 18 districts considered by the parliamentary social services commit- tee and local government, it was only Amudat district that was justified partly on the social Table 2. Karamoja internal divisions by year and parent district. Parent district 1971 2000 2005 2006 2009 2010 Karamoja South Karamoja Moroto (South Karamoja) Nakapiripirit Napak Kotido (North Karamoja) Kaabong Abim Nakapiripirit Amudat Source: Republic of Uganda “Ministry of Local Government” (2010). Critical African Studies 53 features of its people (Republic of Uganda 2009). While in other parts of the country ‘balkaniza- tion’ has taken several trajectories sometime devoid of ethnic or social features, in Karamoja there is a strong emphasis on ethnicity. Social features and ethnic identity have been sharpened to justify ‘balkanization’ whose essence is to pin each ethnic country to a particular territory. These ethnic cubicles in the name of districts have territorialized citizenship, political rights and the freedom of movement. For example, movement of animals between districts is adminis- tratively controlled through the office of the chief administrative officer and the veterinary officer. These two officers have to approve the movement of animals inside and outside the district boundaries. Persons wishing to graze their animals beyond the boundaries of their district have to seek permission to move outside. It is not just people who have been sedentarized, livestock have equally been pinned to territory. A common notion now is that one is Jie because one is from Kotido and someone is from Kotido because they are Jie. While this may be perceived as a right – at least to those interested in sedentary livelihoods – it is also a restructuring of identities and citizenship that is framing con- flicts in the region. A district senior official observed that: ‘districts define and permanently create conflictual tendencies. When people here raid the Pian the politicians here would like to protect Table 3. Dominant ethnic communities in Karamoja by district. District Date created Dominant ethnic/sub-ethnic community Moroto 1971 Matheniko Kotido 1971 Jie Nakapiripirit 2000 Chekwi Kaabong 2005 Dodoth Abim 2006 Acholi-Labwor Amudat 2009 Pokot Napak 2010 Bokora Source: Republic of Uganda “Ministry of Local Government” (2010). Figure 1. Justifications for the creation of Amudat district. Source: Republic of Uganda (2009). 54 M.A. Nsamba people whereas the other would want to recover animals.’2 The desire to respect district bound- aries is placing strain on watering and grazing lands, which have to be used without fallow. It has created a false perception that justifies raiding outside ‘your’ district – which now represents a birthplace. It is very common for the Jie to raid the Matheniko, or the Dodoth to raid the Bokora. The raiders now have no remorse, because victims of their actions are enemies, not broth- ers or sisters, justifying the use of force, killing and arson in the process of raiding. Furthermore, within these geographical ‘containers’ there are other small and sub-ethnic communities, either because the divisions left them on the borderlines or because they are numerically too weak to stand up to the dominant ethnic communities. Such ethnic communities include, but are not limited to, the Ike and Tepth who are increasingly threatened by this dominant notion of territor- ialized citizenship. Although recognized in the constitution, the process of internal boundary changes is yet to grant such small ethnic communities an ethnic home, in effect turning them into non-indigenous ‘settlers’. There is limited questioning of the processes of district creation, decentralization and the reconstruction of identities it has involved, because it has happened within the current framework of democracy and good governance. Thus, decentralization feeds into existing notions of ethno- territorialism and the contested politics of belonging, prevalent in Uganda (Espeland 2007; Kulumba 2004; Schelnberger 2005). District creation has entrenched popular as well as political perceptions of the indigeneity of certain ethnic communities, while excluding others. Sub- sequently, regional origin are background matters in terms of political rights. Similarly, the recon- struction of identities in Uganda is masked by the notion of service provision, back grounding ethnicity. If the need for taxation, security and cultivation to support the colonial economy under- pinned how the colonial state approached sedentarization in Karamajo, the current government’s policy of district creation and sedentarization is buttressed for similar reasons. The increasing availability of weapons in the region in 1970s changed the security equation making it more dif- ficult for the government to maintain law and order (Mirzeler and Young 2000). The government approached the security problem in the region in the same way as the colonial state had, by encouraging sedentarization, but also, most importantly, by restricting inter-ethnic movement. It is clear from the above analysis that internal boundary adjustment in Karamajo has sought to pin each ethnic community to a particular territory. This has had ramifications for the region, in particular in terms of inter-ethnic relations, citizenship and rights – the freedom of movement as well as political rights. It has redefined the politics of belonging, that is, who is considered indi- genous and who is not. It has changed the way in which cattle raiding takes place and is justified, introducing the notion that it is okay to raid a community outside of your own increasingly ter- ritorialized community, district or ethnic group. Whereas the government might have pursued the policy with the intention of resolving security challenges in the region, and justified it on the basis of administrative efficiency, it is clear that the consequences have not always matched the pol- icies’ intentions or justifications. The following section uses the example of Lango and Acholi regions to show the relationship between district creation and political power retention in Uganda. District creation and electoral politics: comparative analysis of the 2001 and 2011 presidential election results in Lango and Acholi regions If the colonial state had sometimes used ‘tradition’ to ethnicize and ‘contain’ Africans, in the era of liberalization and people-centred government, service delivery became the new catchword. But services had to be provided by districts as new centres for local participation in governance, not the central government, whose role was to be restricted to that of policy formulation. Over time districts became centres of power, to the extent that district political leaders would use both private and public resources either to support the NRA/M or to denounce it. In the absence of Critical African Studies 55 well-developed political party structures, the district apparatus came in handy. To the NRA/M there has been the added advantage of not only being able to disperse political patronage in the form of increased numbers of Resident District Commissioners (RDCs) appointed by the pre- sident, but also the opportunity strategically to position its cadres in districts even where the NRA/ M does not have political support. In 2008, President Museveni elaborated the importance of districts in politics, but also high- lighted the place of social service provision in that equation. During a parliamentary by-election in Buikwe, the president said: I have found it imperative to grant Buikwe district status as a way of making service delivery to our people… the government would tarmac the main roads in the district if you vote for NRM, beginning with the stretch from Kyetume to Nyenga. (The Daily Monitor April 28, 2008) The emphasis on social service provision is in part intended to create an incentive to the commu- nities to demand for districts if they want services to be provided. However, no empirical study has shown that the increase in the number of districts has resulted in better service provision. The manner in which a district votes can either improve its relations with the government by way of improved funding outside official government transfers or deny it the same. This is partly because there is no set formula the central government uses to distribute project funds to districts, outside of the grants all districts receive. In 2010 while on an anti-poverty tour of the country, President Museveni is reported to have said that ‘areas which voted for the ruling party candidates have easy access to resources and gov- ernment services … the same will apply to areas which will vote NRM aspirants in the coming polls’ (New Vision February 11, 2010). The manner in which a district votes determines not just what it gets from the central government, but also has an impact on the number of ministers, ambassadors and RDCs appointed from the district. This relationship was eloquently elaborated in 2009 by President Museveni while in Arua, a district he considered opposed to his rule. He is quoted to have argued that: I am looking to 2011 elections in naming the cabinet by rewarding areas that voted overwhelmingly for NRM and sidelining those that did not… if you give me 20% of your votes and another area gives me 90% votes then I will consider the region with the higher votes. (Daily Monitor, February 17, 2009) The control and distribution of the central government resources for party politics, electioneering and patronage in particular districts can clearly be effective. During a focus group discussion in Kasese district, a participant observed that ‘for higher positions we vote for Movement [NRA/M] because they offer protection when the district leaders do not want to help.’3 But it can also work in reverse, so that the control of district structures by smaller opposition parties can hinder effec- tive mobilization in a particular district by the larger dominant party controlling central govern- ment. While the NRA/M created districts in other regions of Uganda as early as 1991, it only created the first district – Pader – in Acholi in 2000, and later, in 2005, it created Amolatar in Lango. While this experience is in part explained by the conflict in the two regions, it is also true that the opposition which was in control of the district structures in the two regions ably used them well to deny the NRA/M means for political mobilization. Figure 2 shows how NRA/M and opposition performed during the 2001 and 2011 presidential elections in the two regions. During the 2001 presidential elections the NRA/M experienced losses in Acholi and Lango, getting only 21% of votes in Lango while the opposition garnered 79%. In Acholi the situ- ation was no better: the NRA/M got 14% against an opposition that garnered 86%. Whereas the 56 M.A. Nsamba NRA/M had the state machinery and was more organized compared with the fragmented opposi- tion, this did not help it win in Acholi and Lango regions. The opposition used its control of the district structures – especially in Acholi – in the context of local conflict and instability, to drum up its political support. With the districts under the control of the opposition and weak NRA/M structures in the region, the NRA/M lacked the infrastructure for political mobilization. The 2001 experience taught the NRA/M lessons which were later to inform its approach and political mobil- ization in both Lango and Acholi. Between 2001 and 2011, the NRA/M increased its presence in the region especially its control of the local structures of governance. One of the mechanisms used was increasing the number of districts. Between 2001 and 2011 the number of districts in Acholi increased from three to seven. In Lango the number of districts increased from two to eight. Most of the new districts were two or single county districts, but these were also strategically created in areas in which the NRA/M knew it was easy to garner political support. The NRA/M managed to win the local council elec- tions in most of the new districts in both Lango and Acholi, effectively putting these new local governance structures under its influence and cadres. During the 2011 elections, these local struc- tures and cadres were very instrumental in mobilizing political support for the NRA/M. But like the 2001 experience, the NRA/M did not focus on building party structures, preferring to use the local governance structures for political mobilization. It is also imperative to note that NRA/M managed to increase its influence and political support in Lango and Acholi because of the increased peace and stability of that time. Figure 2 indicates an improvement in the NRA/M per- formance during the 2011 presidential elections, and shows that the opposition’s performance was far from what it had been in 2001. To understand the influence that the creation of these new districts in these two regions had on electoral outcomes, it is useful to disaggregate further the performance of the NRA/M and the opposition in Lango and Acholi in 2011, according to the new and old districts the NRA/M gov- ernment had created. It is worth comparing how the NRA/M and opposition performed in the new and old districts. What percentage of the 56% and 40% the NRA managed to get in Lango and Acholi, respectively, is from the new and old districts? Figure 3 disaggregates NRA/M and opposition performance in Lango and Acholi, highlight- ing how they both performed in old and new districts. The NRA/M got 57% in Lango: 36% Figure 2. Comparison of the National Resistance Army (NRA) and opposition performance during the 2001 and 2011 presidential elections in Lango and Acholi. Sources: Republic of Uganda “Electoral Commis- sion” (2011); Republic of Uganda (2001). Critical African Studies 57 from new districts and 20% from old districts; in Acholi the NRA/M managed to get 20% from new and old districts. In Acholi, despite the increase in the number of districts, the NRA/M still lost to the opposition. This is a clear indication that adjusting internal boundaries does not always achieve its intended political objectives, yet in Lango it was clearly effective. The poor performance of the NRA/M in Acholi can be explained by a number of factors including the fact that there was a presidential candidate – Mr Nobert Mao – who was not only from Acholi, but also had been the chairperson of Gulu district. His personality was very instrumental in denying the NRA/M votes in Acholi. The other factor was the alleged illegal land acquisition by some of the NRA/M cadres in Acholi region. The creation of new districts in Acholi was vehemently opposed by the opposition and some sections of the public as it was seen as an attempt by the NRA/M to access land in the region during the processes of post-conflict recon- struction and the return of former refugees and internally displaced persons. These factors com- bined to deny the NRA/M victory in Acholi, even when it managed to create districts, most of which were controlled by its cadres. Conclusion Although often justified on the basis of administrative efficiency, and the promotion of decentra- lization, democracy and good governance, it appears that internal boundary adjustment in Uganda often serve far more than just these purposes. Some of the not-so-frequently-cited, but clearly apparent, political objectives and advantages of internal boundary adjustments include enabling the expansion of political patron–client relationships and political mobilization into areas of limited electoral support. This has been done both in perceived NRA/M and opposition strong- holds. However, internal boundary adjustment does not always fulfil the political intentions that may lay behind them. Furthermore, both the NRA/M and opposition can benefit from internal boundary adjustment because both struggle to control the local governance structures in lieu of effective local and regional party structures coupled, and their limited appetite to develop them. Part of the process of adjusting internal boundaries has involved appealing to ethnic commu- nities in a manner that creates incentives for them to demand the alteration of internal boundaries and the creation of new districts. Hence, the narrative of boundary alteration is underwritten in both a language of democracy and political inclusion, and of ‘indigeneity’, belonging and exclu- sion. The altering of internal boundaries by political elites is often portrayed as being responsive Figure 3. National Resistance Army/Movement (NRA/M) and opposition performance in new and old dis- tricts in Acholi and Lango regions during the 2011 presidential election. Source: Republic of Uganda “Elec- toral Commission” (2011). 58 M.A. Nsamba to the demands of society, portraying society as the active catalyst in these processes. This narra- tive is important for dismissing claims that boundary changes serve political interests. It is also imperative for responding to the concerns of development partners and sections of the Ugandan population about the financial costs of creating new districts. Whenever questions about efficiency and the cost of administration are raised, the political elites are quick to justify internal boundary alterations as part of the democratization process, arguing that they are responding to popular demands. Although the extension of social services has also been used to justify the creation of new districts, there is little empirical evidence to suggest that the creation of new districts has really achieved this. Boundary adjustment offers some incentives to the local elites who are very impor- tant in keeping the process rolling, and benefits some local communities around the headquarters of the new districts. Sometimes incentives have included increased employment opportunities for those considered ‘sons and daughters of the district’, but often these same opportunities have been denied to those now considered non-indigenous ‘settlers’. For local political elites the creation of a new district is a way of greasing patronage networks and accessing government tenders. For new communities created there is sometimes a new perception of government being close, and even accessible, and sometimes a few new miles of tarmac roads around the district are delivered. But the creation of new districts has also given communities a new focal point to voice their frus- trations with poor services, and it has not always clearly improved the electoral prospects of those implementing it. Because boundary adjustment has often been cushioned with incentives to local communities as well as national and local political elites, it has sometimes been difficult to develop objective analysis of these processes in Uganda, but it is very clear that not only do these processes have long historical and colonial precedence, but also there is far more than just ‘administrative efficiency’, ‘decentralization’, and the promotion of ‘democracy’ and ‘good governance’ involved. Notes 1. For a detailed insight into who is, and is not, considered indigenous, see the parliamentary debates of 20 July 2005. 2. Interview with a male senior district official in Moroto District, 12 November 2007. 3. Focus group discussion in Kabaka trading centre in Karusandara, 13 April 2008. References Berman, B. 1998. “Ethnicity, Patronage and the African State: The Politics of Uncivil Nationalism.” African Affairs 97 (388): 305–341. Carbone, G. M. 2001. “Constitutional Alternatives for the Regulation of Ethnic Politics? Institution-Building Principles in Uganda’s and South Africa’s Transitions.” Contemporary African Studies 19 (2): 229–252. Espeland, R. H. 2007. “When Neighbours Become Killers: Ethnic Conflict and Communal Violence in Western Uganda.” CMI Working Paper 3. Berghen: Chr Michelsen Institute. Golola, M. L. 2001. “Decentralization, Local Bureaucracies and Service Delivery.” Helsinki: WIDER Discussion Paper No.115. Gore, O. W. 1935. “The Meaning of Indirect Rule: The Principles of Indigenous Administration and Their Application.” Royal African Society 34 (136): 283–286. Government of Uganda. 2006. Local Government Sector Investment Plan (LGSIP), 2006–2016: Decentralisation for Social Economic Transformation. Kampala: Ministry of Local Government. Karugire, S. R. 1980. A Political History of Uganda. Nairobi: Hienemann. Kasozi, A. B. K. 1999. The Social Origins of Violence in Uganda, 1964–1985. Kampala: Fountain Publishers. Critical African Studies 59 Kauzya, J.-M. 2007. “Political Decentralisation in Africa: Experiences of Uganda, Rwanda and South Africa.” Discussion paper for United Nations Department of Economics and Social Affairs. Series 1. UN: DESA. Kulumba, M. 2004. “Ethnocentrism and Movement Politics in Uganda: An Analysis of the Ethnic Conflict during the 2002 District Council elections in Kibaale District-Bunyoro Kingdom.” In Colonial Studies in Uganda (5). Kampala: CBR/ENRECA Occasional Paper Series. Mamdani, M. 1995. “Politics of Democratic Reforms.” In Uganda: Landmarks in Rebuilding a Nation, edited by P. Langseth, J. Katarobo, E. Brett, and J. Munene, 229–239. Kampala: Fountain Publishers. Mamdani, M. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Kampala: Fountain Publishers. Mirzeler, M., and Young, C. 2000. “Pastoral Politics in the Northeast Periphery of Uganda: AK-47 as Change Agent.” Modern African Studies 38 (3): 407–429. Mugaju, J., and Olukoshi, A. 2004. No-Party Democracy in Uganda: Myths and Realities. Kampala: Fountain Publishers. Muhumuza, W. 2008. “Pitfalls of Decentralisation Reforms in Transitional Societies: The Case of Uganda.” Africa Development 33 (4): 59–81. Nsamba, M. A. 2009. Breeding Fragmentation: The Policy and Practice of Decentralisation in Uganda. Kampala: Refugee Law Project. Republic of Uganda. 1990. Report of the Rationalisation of District Boundaries Committee. Kampala: Ministry of Local Governments. Republic of Uganda. 2009. Report of the Sessional Committee on Public Service and Local Government on the creation of Zumbo, Amudat, Otuke, Lamwo, Kyegegwa, Buikwe, Namayingo, Kiyunga, Kiryandongo, Kibuuku, Serere, Alebtong, Gomba, Kalungu, Buvuma, Butambala, Bulambuli, Kyankwanzi. Kampala: Parliament of Uganda. Republic of Uganda. 2010. “Ministry of Local Government.” List of Local Governments. [Online]. Accessed April 20, 2012. http://www.molg.go.ug/local-governments Republic of Uganda. “Electoral Commission, 2001.” Uganda Presidential Elections March 2001, Summary Results by District. [Online]. Accessed May 14, 2012. Available at: http://www.ec.or.ug/Elec_results/ Dist_Sum_2001.pdf[Accessed 14 May 2012]. Republic of Uganda. “Electoral Commission, 2011.” Presidential Elections Results, District Tally Sheet. [Online]. Accessed May 10, 2012. Available at: http://www.ec.or.ug/Elec_results/2011_Pres_dis.pdf [Accessed 10 May 2012]. Richard, R., and Sheridan, J. 1975. “Local and District: Should they be Forgotten?” Journal of Modern African Studies 13 (2): 309–332. Schelnberger, K. 2005. “Decentralisation as a Means of Conflict Management: A Study of Kibaale District in Uganda.” IEE Working Paper S1: 181. Bochum: Institute of Development Research and Development Policy. 60 M.A. Nsamba http://www.molg.go.ug/local-governments http://www.ec.or.ug/Elec_results/Dist_Sum_2001.pdf http://www.ec.or.ug/Elec_results/Dist_Sum_2001.pdf http://www.ec.or.ug/Elec_results/2011_Pres_dis.pdf