Engaging with elections: Ethno-regional mobilization, demands for federalism, and electoral politics in central Uganda

How do ethno-regional movements in electoral autocracies engage with electoral politics? This article argues that such engagement follows the general logic of social movement mobilization but differs from it in degree because of difficult political conditions. As movements in such contexts often face restricted political opportunity structures and command limited tangible resources, they need to compensate by drawing on intellectual and communicative resources to promote their demands in ways that overcome the challenges of sensitivity and particularism associated with their interests and identities. This study examines how Buganda Kingdom in Uganda has promoted its demands for federalism around three general elections and analyses how the Kingdom has addressed the challenges of sensitivity and particularism. The article concludes that the Kingdom’s efforts to overcome the sensitivity of their claims were not supported by an ability to transcend the particularism of their demands, and thus could not be sustained.


Introduction
How do ethno-regional movements in electoral autocracies engage with electoral politics? While this is an important issue in many regions of the world, it is particularly pressing in sub-Saharan Africa, where many societies are shaped by a combination of electoral autocratic rule and intense territorial politics. The latter includes pronounced ethno-regional grievances that revolve around group-based inequalities and marginalization at the level of national politics (Boone 2007). Such dissatisfaction is occasionally translated into demands for more influence at the centre, different kinds of self-rule, or, most commonly, a combination of the two. Sometimes, ethno-regional with the National Resistance Movement (NRM) government of President Yoweri Museveni through most of the 1990s following the restoration of the Kingdom in 1993, relations deteriorated towards the turn of the millennium (Goodfellow and Lindemann 2013). Ever since the reintroduction of multi-party politics in Uganda in 2005, the position of Buganda has been closely observed by all political actors ahead of every subsequent election.
Drawing on social movement theory, this article analyses how Buganda Kingdom acted and pursued electoral strategies marked by engagement or withdrawal during the three general elections of 2006, 2011, and 2016. To investigate this topic empirically, the article interrogates whether and how actors representing the Kingdom promoted grievances and claims around the Kingdom's most important issue and long-standing demand, federalism, and how the Kingdom has sought to reclaim the federal status it enjoyed for a few years after independence in 1962. In particular, the article examines the Kingdom's promotion of federalism in relation to challenges of sensitivity and particularism by analysing whether the Kingdom sought to raise or reduce the electoral salience of federalism and to anchor its demands in political parties, electoral coalitions, or regions beyond Buganda. The article concludes that the Kingdom's efforts to overcome the sensitive nature of their claims were not complemented by an ability to transcend the particularism of their demands, and thus could not be sustained.
Previous research on ethno-regional mobilization around electoral politics Recent overviews of research related to the topic show a dearth of studies connecting the fields of social movements and electoral politics (Hutter, Kriesi, and Lorenzini 2019, 322), linking social movements to ethno-regional mobilization (Muro 2015, 185-186), and (Europe and Latin America aside) examining how ethno-regional demands are connected to electoral politics (Muro 2015, 186). This article contributes to these overlapping fields. Research on social movements and electoral politics under authoritarian rule has drawn upon insights from the literature on movement mobilization in democratic contexts. Explanations of why and how movements engage with electoral politics centre on how the combination of opportunities, threats, resources, and framing contribute to movements' capacities to set their agendas by raising the salience of issues or cleavages, mobilizing mass constituencies, and promoting electoral coalitions by supporting or joining them (McAdam and Tarrow 2019).
Elections generally create strong incentives for collective mobilization. The incentive structure, however, is slightly different under electoral authoritarianism, as the calculus of protest becomes more sensitive (Schedler 2013). The introduction of elections in authoritarian settings intensifies both electoral and non-institutionalized politics. New opportunities increase the potential benefits of engagement, but also deepen the risks (Trejo 2014). While there are examples of successful mobilization by electoral coalitions of movements and opposition parties (Trejo 2014), forceful alliances of that kind are difficult to build under electoral authoritarian conditions. Typically, both opposition parties and social movements lack the resources and organizational capacities to establish long-term credible commitment in harsh and unpredictable institutional contexts (Hutter, Kriesi, and Lorenzini 2019, 331-332).
Despite such difficult circumstances, however, movements do engage. Understandably, most research on how social movements relate to autocratic elections has examined protest mobilization in reaction to charges of electoral fraud (Beaulieu 2014;Brancati 2016;Shirah 2016). Less is known about preelection engagement that does not target electoral regulation directly, especially when the engagement revolves around ethno-regional issues. Most existing work on ethno-regional mobilization and electoral politics covers either ethno-regional parties in culturally pluralistic states in Europe (Hepburn 2009;Zuber 2011;Gillespie 2015), India (Huber and Suryanarayan 2016) or the entry of indigenous movements into party politics in Latin America (Yashar 2005;Vogt 2016). This body of research supports the argument of this article that while all forms of social mobilization face problems with organizing collective action, ethno-regional demands that challenge state structures and national identities make mobilization more complicated, even in democratic settings.
In autocratic contexts characterized by restricted opportunities and more intimidating threats, such mobilization is predictably even more difficult. In addition, since the 1990s, the identity basis of ethnic politics in sub-Saharan Africa has differed from that in Latin American countries, where politicized indigenous movements have demanded rights and representation. Ethnoregional politics in sub-Saharan Africa has its roots in colonial divide-andrule policies (Mamdani 1996) and the subsequent utilization by post-independence governments of the ethno-regional segmentation and stratification it generated (Boone 2003(Boone , 2007. Identity politics in the region is typically articulated as fractured competition for state power between shifting electoral coalitions of various ethno-regional components. While earlier literature described electoral politics in sub-Saharan Africa as ethnic census exercises (Horowitz 1985), subsequent research has shown that relations between politics and ethnicity are more complicated. Ethnic and regional cleavages are certainly important in shaping politics on the continent, but these typically develop in combination with other identities and interests and in indirect ways. While in some countries the territorial concentration of ethnic groups and regional socio-economic differences have generated a more pronounced tendency towards ethnic bloc voting and collective ethno-regional mobilization through a political party, this is not the pattern in sub-Saharan Africa as a whole (Basedau et al. 2011;Elischer 2013). In the absence of parties as vehicles for ethno-regional interests, discrepancies between cleavages at the societal level and party structures therefore sometimes result in the mobilization of ethno-regional movements. This is particularly likely to be the case when the ethno-regional concerns involve calls for some form of self-rule, which is different from being an equal partner in a multi-ethnic coalition, as observed in regions in Nigeria, the Democratic Republic of Congo, and Kenya (Verweijen and Vlassenroot 2015;Sjögren and Angerbrandt 2019).
Theoretical argument, analytical framework, and methodology Social movements seek to translate social conditions into politicized cleavages and grievances and to mobilize around the latter to create collective demands and action. There is relative scholarly consensus on the key dimensions that shape social movement mobilization: opportunities and threats, resources, and framing (McAdam and Tarrow 2019; Hutter, Kriesi, and Lorenzini 2019). Opportunities and threats create incentives to engage in or refrain from collective action by shaping expectations of rewards and punishments and indicating the urgency and necessity of action. The scope for mobilization is also shaped by institutionalized access, including the variety, significance, and accessibility of power centres and arenas for contestation, the nature of alliances, and the capacity of the contender (McAdam and Tarrow 2019, 21).
To utilize opportunities, movements need financial, organizational, intellectual, and communicative resources, all of which facilitate the expansion and coordination of movement coalitions. Movement expansion extends pools of resources and repertoires of action, and multiplies opportunities for engagement by creating access to a greater number of channels and arenas. Movement coordination contributes towards minimizing internal cleavages based on identity and interests. Finally, movements need communicative skills to respond to and recreate opportunities, threats, and incentives by framing the narratives of their grievances and demands (Schedler 2013, 313).
This article sets out from the argument that overall, ethno-regional mobilization in electoral autocracies corresponds to the logic of social movement mobilization outlined above but differs from it in emphasis because of the difficult political contexts in which it occurs. Opportunities are restricted: institutionalized access to power centres and arenas for contestation is more limited, and threats by governments are more serious and frequent. Movements are also more likely to be hampered by limited financial and organizational resources. In addition, ethno-regional movements face collective action problems when attempting to galvanize support beyond their particular identity group and are vulnerable to divide-and-rule strategies by opponents. Thus, the article further argues that such movements need to address the two challenges of sensitivity and particularism and need to overcome both to achieve sustainable success. When seeking to engage in electoral politics, ethno-regional movements need to navigate carefully and compensate for their relative shortage of tangible resources by drawing on intellectual and symbolic resources, such as strategic and communicative skills, and use these to craft coalitions and effectively frame grievances and demands (Snow, Vliegenthart, and Ketelaars 2019).
Neither grievances nor threats and opportunities are given or stable. They are constructed, and in difficult contexts, such construction is critical. Actors who face challenging conditions need to 'manipulate the parameters of collective action' (Schedler 2013, 304) by reframing threats and opportunities and rephrasing grievances and demands to better activate mobilization (McAdam and Tarrow 2019, 21). Movements that advocate for federalism or other territorial issues pursue their claims for self-rule directly with the central government and indirectly with opposition parties, other regions, and identity groups. They thus need to carefully balance, target, and articulate their demands.
The article examines ethno-regional movements' strategic electoral engagement. In view of the challenges that such movements face when seeking to promote sensitive and particularistic issues, the article examines whether and how their strategies addresses those obstacles. For each election, the study first examines strategies to address the challenge of sensitivity by establishing whether and how groups related to Buganda Kingdom sought to raise or reduce the electoral salience of the topic. Options range from trying to minimize the risks associated with elections by not participating to seeking to shape elections by amplifying grievances and demands in direct or indirect ways.
From there, the analysis turns to strategies to address the challenge of particularism and examines whether and how the Kingdom has sought to reinforce or downplay the potentially particularistic nature of their claims. Specifically, the study examines whether federalism is promoted as a demand specific to Buganda or as a broader concern for all regions, for example by linking it to particular political parties or electoral coalitions. Historically, federalism has been closely associated with Buganda, and recent research demonstrates that support for federal arrangements is still strongest in the central region (Ricart-Huguet and Green 2018, 77). Ugandan political discourse has long harboured notions of the privileged position of Buganda. The Kingdom's demands for federalism have therefore been accused of seeking to attain a special status similar to that it enjoyed after independence, which would necessitate careful articulation of the intended scope and content of federalism.
The article presents a sequential analysis of Buganda Kingdom's engagement with electoral politics through the promotion of federalism over three electoral cycles. It does so by examining strategies to address the challenges of sensitivity and particularism. It examines each electoral cycle separately but interprets the strategies of each subsequent cycle as a product of the previous one(s), with expectations of risks and rewards, and the constructions of grievances, opportunities, threats, and demands shaped through experiences.
While Buganda Kingdom may not conform to the definition of a prototypical social movementit is more accurately described as an entity straddling the labels of cultural institution, social movement, and pressure groupthe logic of its social mobilization is appropriate for the purposes of this article. References to Buganda Kingdom in the singular overshadow a range of competing perspectives and internal dissent (For a rich analysis of how this shaped the restoration and subsequent politics of the Kingdom, see Kasfir 2019). This article uses this shorthand, however, and focuses on centrally placed actors in the establishment of the Kingdom (or Mengo 1 ), because its structures are relatively hierarchical and such actors offer authoritative and influential expressions of the Kingdom's demands of the central government. These actors include the Kabaka (King of Buganda), the Katikkiro (Prime Minister of Buganda Kingdom), members of the Katikkiro's government, and the Lukiiko (the Kingdom parliament). 2 Through extensive fieldwork carried out over many years, this study included 30 semi-structured interviews with both political participants and analysts. The categories of interviewees ranged from Mengo officials with various degrees of seniority, members of parliament representing different political parties in Buganda constituencies, academics, civil society activists, and journalists. Most of the interviews were done in 2017, but a few were conducted as early as 2010. The interviews normally took about one hour and centred on various aspects of the Buganda Kingdom's political engagement. They were conducted in English, but an interpreter was always present to ensure that conversations could switch to Luganda whenever respondents so wished. The article also examines media material and written documentation, including petitions and election manifestos. For the media material, a systematic search was conducted in Factiva, a global news database covering local media and international news wires, for reports beginning in 2003, retrieved using relevant keywords. 3

Buganda Kingdom and the quest for federalism
Uganda, Buganda, and federalism: A background Territorial politics, including the relations between central rulers and subnational centres of power, has been at the heart of structuring the political order in Uganda from colonialism onwards and remains deeply contested since independence (Golooba-Mutebi 2008). Through policies that regulated patterns of production, trade, and education, colonial rule deepened separation between and established hierarchies among regions, which were governed by administratively and legally separate institutions (Mamdani 1996). Before independence, these policies led to a series of cleavages and conflicts around territorial politics. One of the most important conflicts concerned the position of Buganda Kingdom, the country's biggest, most populous, and most economically and politically important region. 4 A significant constituency in Buganda advanced aspirations for special status or even self-rule, and the Protestant elite in Mengo formed a monarchist political party, Kabaka Yekka (KY, meaning 'Kabaka only') to pursue Buganda's interests at the level of national politics through the framework of kingship. KY won elections to the Lukiiko and struck a controversial alliance with the republican Uganda People's Congress (UPC) to form the first independence government (Kasfir 2019, 524).
Independence negotiations created a hybrid constitutional order that formally stratified the status of various regions. The independence constitution gave federalism to Buganda and semi-federal status to the other kingdoms in the south and incorporated the rest of the country directly under the central government. Tensions between the central government and Buganda, however, led in 1967 to the abolishment of kingdoms and traditional authorities and the creation of a centralized political order that lasted for two decades under subsequent one-party, military, and multiparty political regimes (Green 2006;Golooba-Mutebi 2008).
The NRM, which came to power in 1986, developed different approaches to sub-national identities and centres of influence. Political parties and traditional authorities were condemned for having contributed to sectarianism: political divisions along ethnic, regional, and religious lines (Kasfir 2019). Parties remained banned until 2005, but traditional authorities, including kingdoms, were restored in 1993 on the condition that they would not engage in 'political' activities. The restoration is generally regarded as having been granted to Buganda in exchange for the region's continued acceptance of restrictions on political party activity (Oloka-Onyango 1997). The informal arrangement between the government and Buganda Kingdom appeared to work well for both sides for the first few years. However, when it gradually became clear to Kingdom politicians that the autonomy of the Kingdom would remain restricted and that its demands for federal status, among other things, would not be granted, relations with the central government grew colder (Englebert 2002).
Following the restoration of the monarchy, the Kingdom repeatedly voiced demands for federalism, or federo as it is called in the vernacular. Mengo raised the salience of the issue and ignored its sensitivity but did not concern itself with addressing the challenge of particularism. The first round of contestation erupted in 1994 in the Constituent Assembly deliberations on making a new constitution. In the debate over state structures and governing systems, the national government proposed a unitary state with decentralized sub-units. Buganda Kingdom instead demanded the restoration of federalism, including a political role for the Kabaka (Kayunga 2000). When the Kingdom's position was backed by groups and individuals associated with the informal opposition, the government reached out to Buganda. However, after having secured the Kingdom's support for a continued ban on political pluralism, the government made an about-turn and withdrew its promise of federalism (Oloka-Onyango 1997, 183). The Kingdom viewed the compromise solution of allowing cooperation between two or more districts on matters of culture and development as a betrayal and set up its own parallel administrative structures. These entities, however, lacked financial underpinning and legal powers (Englebert 2002).
Against this background, it is unsurprising that opposition politicians drew on federalism to make inroads into the Buganda electorate. During the 1996 presidential campaigns, the opposition candidate Paul Ssemogerere enjoyed the backing of members of the royal family and Buganda Kingdom ministers, and he promised to revisit the federalism issue. In the 2001 presidential election campaigns, Kizza Besigye of the opposition did the same, also with perceived support from influential individuals in the Kingdom (Gay 2014, 254-255); however, Museveni was declared the winner in both the 1996 and 2001 elections, receiving the majority of the votes in Buganda 5 , and the quest for federo seemed elusive.

The 2006 elections: Implicit influence
Between 2001 and 2006, Uganda's political system was overhauled and conditions for mobilization changed dramatically. Following splits within the NRM and pressure from opposition groups and civil society actors, the government turned around to propose the return to multi-party politics. The reintroduction of political pluralism was linked, however, to the lifting of presidential term limits, enabling President Museveni to stand for office indefinitely. While these proposals were not made law until mid-2005, they were accommodated into political calculations from their first mention in 2003. This section first chronicles the key steps of Buganda's demands for federalism up to and including the 2006 elections, interwoven with an analysis of the Kingdom's strategies to address the sensitivity challenge by influencing the salience of the issue, and then turns to the challenges of particularism.
Following the 2001 elections, the Constitutional Review Commission (CRC) presented yet another opportunity to promote federalism. Among the Kingdom's proposals to the CRC, federalism was the most important (Kingdom of Buganda 2003). The request was rejected: the government's proposed constitutional amendments in 2004 instead contained a slightly extended version of the regional tier system based on district cooperation (Green 2006, 383). Buganda's demands for federalism were thus again discarded, but not abandoned. With elections approaching, both the government and the Kingdom were keen to find a solution. Mengo was aware of its electoral importance, but also of its delicate predicament. While the government was likely to be willing to accommodate some of Buganda's concerns for electoral purposes, federalism was declared to be out of reach. Immediately following the news of this rejection, the Katikkiro called for a day of mourning, and the Kingdom made use of its media channels to raise public awareness (Kato 2004). More significant in the long run, however, was that the Kingdom opted for closed and direct negotiations with the government. At this point in time, the electoral arena was only a future possibility: political pluralism was still formally outlawed, and parties were untested entities.
The negotiation began in July 2004. Mengo's main strategy during this period can be analysed as seeking to deal with the sensitivity of federalism by reducing the electoral salience of the issue and removing it from contentious politics. It did so by negotiating as a professional interest group for a specific aim rather than mobilizing as a social movement around more farreaching demands with uncertain outcomes. To complement this strategy, whenever talks threatened to stall, the Kingdom raised the salience of federo in relation to the Baganda public through media campaigns and other messaging (interview, Mengo official, March 3, 2010). When the negotiations concluded in February 2005 with an agreement on a regional government system approved by the Katikkiro's cabinet and the Lukiiko, this twotrack strategy seemed to have been successful.
The situation, however, proved to be more complicated. In the new polarized political landscape, internal discontent came to the fore more easily. From the late 1980s, there was renewed internal disagreement in Buganda about how to conceive and articulate the relations between culture, politics, and kingship. The group that emerged in the early 1990s as the prominent voice of the Kingdom consisted of younger urban professionals who held more pragmatic views than their older, culture-oriented, clan-based competitors (Kasfir 2019). Now, the latter groups reappeared. Encouraged by the views of some opposition politicians and opposition-leaning Mengo leaders (interview, Mengo official, March 3, 2010), a group of clan heads pleaded with the Kabaka to reject the agreement as a betrayal of true federo. The regional government proposal was district-based, separated Kampala from Buganda, and suggested the direct election of the Katikkiro and the Lukiiko, all of which were declared unacceptable by opponents to the proposal. The ground shifted, and in December 2005 Katikkiro Ssemwogerere resigned and was replaced by the outspoken Dan Muliika, widely regarded as leaning towards the opposition. In February 2006, the Lukiiko formally rejected the regional government agreement and instead demanded full federalism (Mwanje and Nalugo 2006).
Whether or not it was the result of a deliberate strategy to overcome the challenge of sensitivity by throwing caution to the wind, Mengo's aboutturn fed into the February 2006 election campaigns. Following the restoration of multi-party politics, the political actors and arenas had multiplied, and appeals to ethno-regional bloc votes became more urgent and complicated. Following exiled opposition leader Kizza Besigye's return to Uganda in November 2005 and his immediate arrest for treason, the country was deeply polarized and tense, and Buganda was regarded as a particularly important potential voting bloc. During hearings of the CRC, opposition parties had already supported the introduction of a federal system of government (Kayunga 2000), and before the 2006 election they reached out to the Kingdom and promised to introduce federalism if voted into power. After the rejection of the regional tier agreement and the change of Katikkiro and his cabinet, Mengo was treated by the government as a hostile force. However, Museveni and the NRM were declared electoral winners both across the country and in Buganda.
The Kingdom's strategies in relation to the challenge of particularism during this period were inconsistent, again, partly due to the many voices calling for federo. In its submission to the CRC (Kingdom of Buganda 2003), federalism was described and justified at length, and its scope, content, and purpose were placed within a general, civic, and developmental framework. The submission emphasized the nationalist scope of the demands and the allegedly broad support that federalism enjoyed throughout the country. It stressed that federalism was not about tribalism or monarchism, nor was it only for Buganda. The submission proposed that federalism be introduced throughout the country and presented it as the best way to share and utilize power and resources. However, the submission also contained monarchic components by way of proposing a non-political Kabaka as the constitutional monarch of Buganda.
After the demands to the CRC had been rejected, Buganda sought intermittent backing from the other restored Kingdoms of Busoga, Bunyoro, and Tooro, but received only occasional lukewarm support from Busoga (Namungalu 2004). 6 On the whole, however, notions of federo as a particularistic demand lingered on. In its negotiations with the government, Buganda Kingdom represented only itself. This was a pragmatic choice: to negotiate for federalism under an umbrella of many regions would have risked watering down the content of its demands. According to one negotiator, a benefit of the regional tier solution would have been to overcome one form of particularism by having 'Mutebi recognized as Kabaka of Buganda, not, as is now the case, Kabaka of the Baganda' (interview, Mengo official, March 3, 2010). This was not to be.

The 2011 elections: Different strands of open engagement
After 2006, relations between the government and Buganda Kingdom deteriorated rapidly. The Kingdom was drawn deeper into the national political contest between Museveni and Besigye, 'partly in ways we didn't ask for, partly of our own making […] this probably explains the heavy-handed methods used against both us and the opposition' (interview, Mengo official, Kampala, August 28, 2017). A series of legislative proposals worsened the antagonism (Goodfellow 2014). In 2006, the government announced its intention to place management of the capital Kampala under government control and to extend its boundaries. Mengo reacted angrily to what it regarded as a tool to undercut its demands for federalism (Gore and Muwanga 2014). In 2007, the government tabled a bill to revise the contentious Land Act with specific references to the need to protect tenants against evictions; the Bill met fierce resistance from Buganda Kingdom. After two years of intense debate, the Land (Amendment) Bill was passed by parliament in November 2009.
In this polarized context, Mengo disregarded the sensitivity of federalism and continued to raise the salience of the topic. When in July 2009 the government declared that there would be no more negotiations over federalism, the Lukiiko responded with a strongly worded resolution in which it promised to continue to pursue its demands and push for federo (Nganda 2009). Two months later, as the Kabaka set out to visit Bugerere county in Kayunga district that hosted a small ethnic community that had declared itself independent from Buganda, his envoys were blocked by police. The news of this led to several days of pro-monarchic riots in Kampala during which 27 people lost their lives, many more were injured, and great swaths of property were destroyed. In the aftermath, the government cracked down on the Kingdom. Four of its radio stations were closed for more than a year, and another series of legislative changes were (re-)introduced. In December 2009, the government tabled the Institution of Traditional or Cultural Leaders Bill aimed to restrict any traditional or cultural leader from taking part in politics. The bill was shelved, but was reintroduced and passed shortly before the 2011 elections. The Regional Tier Bill had also been reintroduced in 2009. Finally, the government introduced the Public Order Management Bill, which would severely restrict public gatherings (Goodfellow 2014). In March 2010, the Kasubi Tombs, the burial grounds of some Buganda kings, caught fire, and the circumstances directed suspicion towards the government. Nothing could be proven, but the incident deepened the resentment many Baganda held against the government (Sjögren 2015).
Because of its broken relations with the government after the Kayunga riots, it was widely expected that Buganda would be lean towards the opposition in the 2011 elections (Brisset-Foucault 2013a, 510;Tangri and Mwenda 2010, 43). One important indication of this was the mid-2010 launch of the political pressure group Suubi ('Hope'), led by prominent Mengo personalities, including former Katikkiro Ssemwogerere, set up with the stated aim to support candidates who would campaign for the interests of Buganda, including federo (Baral 2014). Suubi could not be an open and direct extension of Mengo demandsthe incumbent Katikkiro disassociated himself from its activitiesbut the presence of a former Katikkiro and other Mengo officials among the Suubi team inevitably created the impression that the initiative had the blessing of at least some sections of the Buganda establishment, which lent it prestige. The experiences of 2009 had led to different views within Mengo on whether to withdraw from or to engage in national politics. Suubi represented the position that Mengo's demands could only be advanced through a strengthened presence in national electoral politics.
Suubi signed a memorandum of understanding with the opposition Inter-Party Coalition (IPC) and agreed to operate within it (Brisset-Foucault 2013a; Baral 2014, 324-325). In its election manifesto, the IPC committed to promoting federalism, which it presented as a way to give power to the people and to share resources equally. In the manifesto, however, the policy for countrywide federalism was linked to promises to defend Buganda issues (Forum for Democratic Change 2011, 26-27). Suubi thus both reflected and reinforced the heightened salience of Buganda Kingdom demands ahead of the 2011 elections. This strategy of engaging electoral politics and raising salience, of which Suubi was the main expression, was constrained in two ways. Not only were the demands of Mengo necessarily indirect, the electoral strategy was also deployed under difficult circumstances. The shutdown of Kingdom media and the close monitoring of content when radio stations were allowed to operate again restricted Suubi's opportunities to communicate its message through these otherwise benevolent channels (Brisset-Foucault 2013b, 79-85).
While political demands associated with the Kingdom and issues with monarchist connotations were highly salient campaign issues in Buganda in 2011, the amplification of these issues made it difficult to communicate notions of federalism beyond its particularistic connotations. The salience of Buganda Kingdom concerns in the 2011 elections allowed different perspectives on Buganda's true interests to arise and created dynamics of both infighting and appropriation. Suubi defined itself as a political mobilization group. This gave its members leeway to choose different political vehicles, largely to protect the Kingdom from accusations of direct partisan involvement. Most of Suubi's members came from the Democratic Party (DP), some were from Besigye's Forum for Democratic Change (FDC), and one or two ran as independent candidates.
Electoral competition, however, made it difficult to establish cohesion. In addition, apart from Suubi candidates, many other political contenders sought to be associated with Mengo, including candidates representing the Justice Forum (JEEMA), the faction of DP that did not associate with Suubi, the Conservative Party, and the recently established Uganda Federal Alliance (UFA). The multiple uses of alleged connections to Mengo created a campaign context characterized by divergent messages and fragmented mobilization in relation to issues promoted by the Kingdom (Brisset-Foucault 2013a, 523-524). Kingdom demands were thus inserted into electoral politics within Buganda. Extending and transforming the frame of federalism to appeal to actors outside the region proved more challenging. Attempts were made to transform perceptions of federo from a Buganda-specific concern into a way to promote democracy and development for the whole of Uganda (Brisset-Foucault 2013a, 514-515). The support of other kingdoms and regions in late 2009 for the regional tier proposal, however, made such efforts difficult. The electoral results of the Suubi effort were mixed. All of Suubi's main candidates for the parliamentary elections won, as did Erias Lukwago in the race for Kampala Mayor. Nevertheless, Suubi and the IPC failed to create a Buganda bloc vote for the opposition. Instead, the NRM won the majority of votes and seats in Buganda.
The 2016 elections: Dodging the ballot and negotiating for the future After the 2011 elections, Mengo found itself in a difficult position. Its escalating confrontation with the government had resulted in a severe crackdown after the 2009 riots. The early 2011 enactment of the Institution of Traditional or Cultural Leaders Bill, which made it an offence for such a leader to join or participate in partisan politics (as carefully defined), imposed serious constraints on Mengo's ability to communicate directly their concerns or demands. Although Kingdom radio stations were allowed to re-open ahead of the 2011 elections, the scope of what could be debated or promoted on those stations was much restricted (Brisset-Foucault 2013b, 82-85). The 2011 polls had resulted in the election of a few vocal members of parliament loyal to Mengo's cause, but by and large the NRM had consolidated its grip on power. Most arenas for public engagement were thus closed down, leading to a period of withdrawal, reflection, and rethinking for Mengo.
The content of this re-strategizing was made evident to the public on 1 August 2013, when a signed agreement between the government and the Kabaka was announced (Butagira 2013). In the agreement, the government promised to return Kingdom properties and to allow the Kabaka to travel without restrictions within Buganda, and both sides agreed to 'totally refrain from engaging in hostile propaganda against each other' (Republic of Uganda 2013). The agreement did not, however, mention either of the two main issues under contention: the regulation of access to land and Buganda's demands for federalism.
The agreement marked the beginning of a new period of relations between the Kingdom and the government characterized by acceptance and moderation. It also coincided with, and was to some extent dependent on, a new Katikkiro, Charles Peter Mayiga, who over a few years had made a political journey from firebrand Mengo activist to pragmatic negotiator. Mayiga's approach was communicated clearly: there would be no politics in Mengo. The critics of the new approach found it difficult to express their concerns within the structures of the Kingdom. Critical voices were asked to moderate their views on radio talk shows or were not invited to them, many of the pressure groups active since the late 1980s (Brisset-Foucault 2013a, 512) were deactivated (interview opposition Member of Parliament, Kampala, August 27, 2017), and the annual Buganda conference, Ttabamiruka, was met with more hesitation by Mengo (Lumu and Kaaya 2013). The shift from the earlier period was distinct. It was a turn from vocal and sometimes radical contestation with popular back-up to the pursuit of long-term demands through cautious elite-confined negotiations and from engagement with electoral politics to withdrawal: 'the Katikkiro understands the limits to confronting the government' (interview, Mengo official, Kampala, August 28, 2017).
Despite federalism being the second item on Mayiga's list of priorities, the issue was rarely addressed. A number of interviewees gave the same explanation for this: it was simply too sensitive. 'Federalism is about sharing power, and sharing power with Museveni is difficult' (interview, Mengo official, Kampala, August 28, 2017). If Mengo were to pursue demands for federalism, it would be regarded by the government as promoting the kind of hostile propaganda the agreement disallowed (interview, journalist, Kampala, August 20, 2017;interview, journalist, Kampala, August 25, 2017, interview, opposition Member of Parliament, Kampala, August 27, 2017. A different perspective held that the cautious approach constituted a change of means, not of aims: 'We are now building the conditions for federalism rather than shouting about it' (interview, Mengo official, Kampala, August 29, 2017).
As a consequence of the 2013 agreement, the role of Buganda Kingdom in the 2016 elections and the place of its demands differed markedly from those in the 2006 and 2011 elections. Mengo representatives kept aloof and silent and were rarely linked, directly or indirectly, to political candidates. Mengo's concerns were largely absent from the campaigns, and with the exception of a promise in the FDC manifesto to facilitate the introduction of federalism (Forum for Democratic Change 2016, 49), the topic was never brought up. The sensitivity challenges defined by the government and embedded in the 2013 agreement restricted the Kingdom's ability to make demands and confined it to bilateral relations with the government around particularistic interests.

Conclusions: The occasionally political kingdom
Over the course of the three electoral cycles examined, Buganda Kingdom's attempts to promote federalism shifted from engagement and confrontation, occasionally translated into electoral political activity, to retreat and silence. Essentially, this withdrawal was a forced response to the limits drawn by state authority. But it also reflected a strategic shift emerging from the Kingdom's inability to extend and transform the conception of federo and the particularistic connotations attached to it.
Even prior to the reintroduction of multi-party politics in 2005, Mengo had begun to pursue federalism more actively. The new electoral arena allowed new possibilities and actors, and most electoral roads to power had to take Buganda Kingdom into account. Both Mengo and other actors contributed to raise the salience of federalism as a political issue. Fuelled by the politicization of other Mengo concerns, the amplification of federalism continued up to the Kayunga riots of 2009. The aftermath of the riots demonstrated the limits to the activist strategy of heightened confrontation; the Suubi project of 2011, reflecting remaining contradictions within Mengo, was the last expression of the approach that emphasized engagement through electoral politics. Following the 2013 agreement with the government, it became very difficult for Mengo to voice federal demands.
The various strategies for dealing with the challenge of sensitivity reflect both the content of federo and the different understandings within the Kingdom of whether a regional tier solution was an acceptable compromise or a betrayal. The scope of federo becomes clearer through analysis of the strategies to deal with the challenge of particularism in efforts to promote federalism. Ever since the independence constitution, Buganda's demands for federo were surrounded by ambiguity and vulnerable to accusations of chauvinism and privilege. The Kingdom's critics were quick to suggest that in reality Mengo sought not federalism for Uganda, but federo for Buganda. This legacy lingered on. Even when Mengo tried to reach out to other regions, its efforts were restricted by its history of particularism and its inability to move decisively beyond its own identities and interests. Mengo's occasional attempts to raise the salience of federalism were not complemented by strategies to craft alliances or reframe the scope and content of federalism to overcome the particularism of their demands, and could hence not be sustained. It is useful to compare this with the early post-independence era, when a monarchist party was successful in promoting Buganda's interests, including federalism. Those gains, however, were short lived: the fragile pact between the UPC and the KY did not address the underlying challenges of sensitivity or particularism of the monarchist demands and broke down after only four years. Indeed, the experiences of this era partly explain the difficulties in advancing demands for federalism today: the 1966 crisis made every successive government wary of ceding too much influence to Mengo.
The experiences of Buganda can be compared with the other restored kingdoms of Bunyoro, Busoga, and Tooro and the unreconstructed kingdom of Ankole, which have been even less successful in promoting ethno-regional demands. The government has made sure to prevent any ethno-regional mobilization from these kingdoms by reinforcing their inherent weaknesses. All the other kingdoms are financially weak, hampered by leadership wrangles, and home to solidly NRM-voting populations, reproducing their deep dependence on the central government. Only Bunyoro Kingdom has attempted to make use of its oil resources to draw concessions from the government, but even then, only intermittently and to very modest effect (Sjögren 2015).
Theoretically, the experiences of Buganda Kingdom underline the significance for movements of extending and transforming their demands to transcend their particular(istic) concerns. In contexts similar to the one examined here such endeavours are bound to be difficult. In structural terms, the prospects of politicizing identity-based concerns by creating coalitions, even if temporary, are likely to be more advantageous where territorial grievances with the central government are more broadly shared and demands for federalism can be anchored among a broader section of subnational regions. Government, Uppsala University, Maria Wendt, and the two anonymous reviewers for helpful comments on previous versions of the article.