Kenya’s 2017 elections: winner-takes-all politics as usual?

ABSTRACT

This article asks what Kenya’s 2017 general elections tell us about the capacity of a new constitution to reduce the stakes of political competition and prospects of political instability. Three constitutional changes are particularly important: the adoption of a 50% + 1 threshold for the presidential election; the devolution of power to 47 county governments; and the introduction of a Supreme Court with the right to hear presidential electoral petitions. We find that the impact of the 2010 constitution has been mixed. The 50% plus 1 threshold encouraged coalition formation, but this dynamic has long been evident. Devolution has given a wider set of Kenyans a stake in the system, but has also created new structures that can be used to channel dissent against the state. The Supreme Court demonstrated its capacity to act as an independent institution, but did little to sustain electoral legitimacy. Indeed, while the 2010 constitution has clearly reshaped the political landscape, it was a personal deal that ended the post-election impasse. The elections therefore demonstrate how formal institutions alone cannot change political logics and reveal the continued significance of individual politicians and informal institutions that may compete with or complement their formal counterparts.

KEYWORDS:

• Elections
• constitutional reform
• devolution
• presidentialism
• political culture
• accountability
• ethnic politics
• Kenya

The Kenyan general elections of 2017 were the second to take place under a new constitution introduced in 2010 that, among other things, sought to devolve power away from the president by creating 47 new county governments. They were also the first to take place in a context in which politicians and voters had practical experience of the powers of the new devolved elected posts and operations of the new political dispensation. As a result, the 2017 elections represent the first real opportunity to take stock of whether the 2010 constitution has effectively reduced the stakes of political competition and thus the prospects for political instability. In the wake of the 2007 post-election violence – in which over 1,000 people lost their lives and almost 700,000 were displaced – peace negotiators, international donors, political leaders, civil society groups, academics and ordinary citizens turned their attention to the question of how such events could be prevented from ever happening again. One of the most common analyses to evolve out of this process of political soul searching was that political violence in Kenya was rooted in communal narratives of injustice and the country’s ‘top-heavy’ constitution,¹ which together conferred great power on the president and encouraged a divisive form of ‘winner-takes-all politics’.² This, combined with a history of political corruption and election-related violence, weak institutions, and close elections, was said to have created a ‘perfect storm’.³ The implication of this analysis was clear: only by changing the political system could Kenya prevent something similar from happening again.

To many commentators’ surprise, the 2010 constitution included meaningful reforms to address these concerns.⁴ Many of the appointment powers previously enjoyed by the president were made subject to parliamentary review, political rights and civil liberties were protected by a new bill of rights, and the government was mandated to facilitate public participation in political decision making. In order to address the impact of the 2010 constitution and frame this special issue, we focus on three particularly important and high profile constitutional changes that appeared to have the greatest potential to reshape the nature of political competition: the adoption of a 50% + 1 threshold for winning the presidential election, with the potential for a second round run-off; the system of devolution, complete with 47 new counties each with a directly elected governor, senator, women’s representative and assembly; and, the introduction of a Supreme Court with the right to hear presidential electoral petitions. The latter was designed to encourage aggrieved parties to take a petition before the Court, rather than to the streets as the opposition did in 2007; with a more independent judiciary also envisaged as able to validate results and thus boost public confidence in the electoral process. Taken together, these reforms were designed to weaken the presidency and thus reduce the stakes of national elections, and to simultaneously encourage political leaders to form broader multi-ethnic alliances and adopt more inclusive campaigns.⁵

Our analysis of the effectiveness of these reforms is based on long periods of fieldwork conducted by all five authors. This includes research on the 2017 elections, which began a year before the elections and ended six months afterwards, as well as more long-standing research on Kenyan history and politics. It also draws upon a series of nationally representative surveys and a careful reading of the media as well as of the other articles included in this special issue. Perhaps unsurprisingly given the partial way in which the formal rules of the game have been implemented in Africa’s new democracies,⁶ we find that the impact of the 2010 constitution on political dynamics has been mixed. Following a hotly contested presidential election in which there were only two serious contenders – President Uhuru Kenyatta of the Jubilee Party and opposition leader Raila Odinga of the National Super Alliance (NASA) – the election rapidly descended in to acrimony. After a tense and disputed counting and tallying process, Kenyatta was declared the winner with 54.2% of the vote. Odinga challenged Kenyatta’s victory in court and for the first time in Kenya, and indeed in Africa, the election of a sitting president was nullified on 1 September 2017 when a majority of Supreme Court judges ruled that the presidential election was illegal, null and void, and ordered a repeat poll.⁷ This ‘fresh’ election was held on 26 October but was boycotted by Odinga and NASA who argued that – with the same officials and many of the same procedures in place – the repeat election would be stolen from them once again.⁸ In the resulting poll, Kenyatta was declared the winner with 98.3% of the vote on an official turnout of just 39%. NASA once again alleged that the poll was flawed, which – together with low or no voting across the opposition strongholds – undermined the credibility of yet another election in the eyes of many Kenyans.

The complex legacy of the 2010 constitution is also revealed if we take a more fine-grained approach to the elections and look at the impact of the three high-profile reforms outlined above. The 50% + 1 threshold has provided further incentive for coalition formation, but this pattern was clearly evident in Kenya well before the 2010 reforms and has been driven as much by what Cheeseman and Tendi have referred to as a ‘politics of collusion’ between the country’s political elite as any specific feature of the electoral process.⁹ In turn, devolution provided a wider set of Kenyans a stake in the system because opposition leaders won many governor and senator races. Leaders who then had strong motivations to guard against violence in their own areas.¹⁰ However, in 2017 the victory of the ruling party in many lower level elections, including the capital Nairobi, meant that the rebalancing effect of county elections was less pronounced than it had been in 2013. Moreover, the aftermath of the polls demonstrated that devolution has also generated new political structures that can be used to channel dissent against the state, most notably when some opposition governors raised the prospect of their counties seceding from Kenya.¹¹ Finally, the Supreme Court demonstrated its capacity to act as an independent institution to defend the quality of democracy when it became only the third court ever to annul the election of a sitting president. However, this ultimately did little to sustain the legitimacy of the election because the court immediately faced criticism from the government and was unable to ensure the implementation of much needed reforms ahead of the ‘fresh’ poll.

Indeed, while it is clear that the 2010 constitution has reshaped the political landscape in profound ways, it was not changes to the formal – i.e. written and codified – rules that ended the political impasse, but a personal deal struck between President Kenyatta and Odinga. In other words, the resolution of the 2017 electoral crisis was rooted not so much in the niceties of constitutional reform, but in a long history of elite-level pacts, a classic informal – i.e. unwritten and uncodified – institution. As a result, Kenyan politics is less ‘winner-takes-all’ than it may at first seem, but this is not simply a result of the introduction of a new political dispensation. Instead, the 2017 elections in Kenya serve as an important reminder of how formal institutions alone cannot change political dynamics.

Given this, one of the most important lessons to take from the recent polls is the extent to which the effects of formal institutions are shaped by their informal counterparts.¹² As Cheeseman has recently argued, ‘[e]fforts to understand political institutions and democratisation in Africa will be dangerously incomplete unless they address the informal foundations of formal institutions.’¹³ To date, the literature on African studies has typically focussed on the way in which informal institutions such as patrimonialism may undermine their formal counterparts,¹⁴ leading to weak parties and parliaments. However, this is not inevitable. Instead, informal institutions may supplement and complement formal institutions. As Douglas North has argued, no formal institution is truly consolidated until it is underpinned by a set of supportive informal norms and practices.¹⁵ In the Kenyan case, informal institutions such as clientelism and patrimonial politics have undermined the effectiveness of many of the reforms introduced in 2010. However, the same informal institutions also underpin the high degree of elite cohesiveness and informal deal making in Kenyan politics, which – as with previous elite pacts, such as that forged between President Kibaki and Odinga following the post-election violence of 2007/8 – is what enabled Kenyatta and Odinga to reach a political compromise and avert a deeper political crisis. In this way, informal institutions both compete with, but also help to shape and compensate for, their formal counterparts.

But while the elite deal struck between Kenyatta and Odinga boosted short-term political order, it had very different implications for other democratic virtues such as accountability. Press coverage suggests widespread support for the way that the ‘handshake’ between Kenyatta and Odinga returned the country to something like normality, but also reveals wider concern that it has done little to address underlying issues and has also undermined the evolution of a more effective and credible opposition.¹⁶

The 2017 elections: a summary

Before analysing the meaning and impact of the 2017 elections it is important to briefly set out some of the main events and controversies of the polls. Kenya’s 2017 election had eight presidential candidates but only two mattered. The incumbent president, Uhuru Kenyatta – with his deputy, William Ruto, as his running mate – stood on the ticket of the Jubilee Party. This party, the successor to the Jubilee Alliance of 2013, brought together several allied parties, including Kenyatta’s The National Alliance (TNA) and Ruto’s United Republican Party (URP). The second candidate was Raila Odinga – the flagbearer for NASA. This coalition, formed in January 2017, expanded the Coalition for Reform and Democracy (CORD) on whose ticket Odinga had run in 2013. Thus, while CORD had included Odinga’s Orange Democratic Movement (ODM), Kalonzo Musyoka’s Wiper Democratic Movement (WDM) and Moses Wetang’ula’s Forum for the Restoration of Democracy-Kenya (FORD-Kenya), NASA also included the Amani National Congress (ANC) headed by Musalia Mudavadi and Chama Cha Mashinani (CCM) headed by Isaac Ruto.

Kenyatta, and many lower-level Jubilee candidates, benefited from the powers of incumbency in multiple ways. This included the use of state resources and officials and interference with media freedoms.¹⁷ Odinga and his supporters also accused Kenyatta and Jubilee of manipulating the polls in multiple other ways, both during the campaigns and after.¹⁸ These accusations were informed by bitter memories of the 2007 and 2013 elections, which many opposition supporters believed that Odinga had won;¹⁹ by a longer history of malpractice that stretches back to the single-party era;²⁰ and by ongoing developments. Following a sustained campaign by Odinga and his supporters, which reached a peak in mid-2016, the previous commissioners of the electoral management body – the Independent Electoral and Boundaries Commission (IEBC) – had been forced to resign, with new commissioners only coming into office in January 2017. However, the opposition continued to challenge every aspect of the process, including the updating of the electoral register, the printing of ballot papers, and the relationship between paper and digital results.²¹ Some inexplicably bad decisions on procurement and chronically poor communication from the IEBC heightened suspicions further.²² So too did the abduction and murder of the IEBC’s acting head of information technology, Chris Msando, less than a week before the polls – a crime which remains unsolved.²³

On election day itself, 8 August 2017, the process seemed at first to go very well. The new biometric voter identification kits worked smoothly in the vast majority of cases, while initial comments from electoral observation teams were positive, even buoyant.²⁴ However, in the hours after the polls closed, things began to go wrong. Polling stations had been expected to transmit the results of the presidential poll electronically, and to follow this up by sending a digital image of the paper results form, which would then be taken physically to the tallying centre for verification. Many results did come in electronically, and Kenyatta took an early lead. But the number of rejected ballots seemed unrealistically high, and Odinga gave a press conference at which he alleged that the electronic figures were bogus and that the IEBC servers had been hacked.²⁵ Worse still, some polling stations could not transmit the results and/or were unable to send the digital image. Three days of accusations and rumour followed; the electronic tally crept up, but around a quarter of the digital scans remained unavailable. Amidst growing uncertainty and confusion, NASA produced what they claimed was evidence of hacking, though independent scrutiny did not confirm this.²⁶ NASA also released what they claimed to be the real results, giving Odinga victory, but these generated no more confidence than the official tally.²⁷

To make matters worse, the IEBC gave inconsistent explanations of the problem with the results transmission, creating uncertainty as to whether they had all the paper forms. Then, on the evening of 11 August, the IEBC declared Kenyatta the winner of the presidential election, despite the fact that not all of the forms had been accounted for and the official timeline meant that they had two more days in which to iron out any problems.²⁸ NASA immediately denounced the outcome and – after a tense few days of uncertainty and intermittent demonstrations that were violently suppressed with significant loss of life²⁹ – lodged a petition against the result.

In court, NASA’s case rested on the allegation of digital fraud; the presidential results, they said, had been altered through interference with the IEBC’s servers, and the paper record subsequently adjusted to fit these. Out of court, they continued to insist that they were in possession of the real results, and that Odinga had won. Since NASA, due to problems with their own system of party agents, did not have reliable copies of all polling station results, and since copies of all forms were not available on the IEBC website, their claims were hard to either prove or disprove. The IEBC denied NASA’s allegations. However, senior electoral officials struggled to explain the problems and discrepancies in the process and failed to give access to the servers to allow the allegations to be thoroughly investigated. To the great surprise of most commentators, the Supreme Court – by a majority verdict of four-to-two – made the bold decision to annul the presidential election on the basis of multiple irregularities in process and the failure of the IEBC to allay its concerns.³⁰ The judgment offered no opinion as to whether those irregularities had changed the outcome and ordered that a fresh presidential election be held within sixty days.

This verdict appeared to be a victory for Odinga, and he and his supporters celebrated it as confirmation that the election in August – as well as those in 2007 and 2013 – had indeed been rigged. By contrast, Kenyatta officially accepted the judgment but also made no attempt to hide his anger at the Supreme Court, while a number of Jubilee politicians openly called for the Court’s powers to be reduced and/or its members replaced.³¹ But while NASA supporters celebrated the outcome, the opposition was left with a major challenge. Jubilee had outperformed NASA in the other electoral races, securing a majority in both the National Assembly and the Senate, and gaining control of a majority of counties. NASA claimed that this was also the result of ‘computer-generated results,’ but while challenges to the results were heard in court, Jubilee could call on more elected politicians to help campaign for Kenyatta in the fresh polls. With fewer elected politicians, more limited resources, serious challenges of internal cohesion – as some who had lost in the August elections switched their support to Kenyatta – and the same electoral management body and rules in place, NASA was in a difficult position. As a result, when the second campaign began it was more uneven than the first.

The quality of the ‘re-run’ was also called into question by the government’s decision to force through revisions to electoral legislation that were explicitly intended to prevent the annulment of another election on procedural grounds. This represented a significant break with precedent; the changes pushed through parliament during the campaign against protests by NASA, civil society activists and some members of the international community.³² Amidst these developments, Odinga abruptly announced his withdrawal from the election in early October on the grounds that the minimum reforms to the electoral process that he had demanded – an extensive list, which could not practically have been managed in the time limit set by the court – had not been met.³³ This decision was also informed and justified with reference to the very public turmoil within the IEBC, whose commissioners were deeply divided. One commissioner resigned and fled the country a week before the fresh polls citing security concerns, after which the chair made an extraordinary speech in which he seemed to teeter on the brink of resignation.³⁴

Against a backdrop of growing intimidation by government supporters, which included the use of excessive force against periodic protests, the Supreme Court called a session to consider a challenge to the timing and organisation of the fresh election. However, five of its seven members failed to appear – including one whose driver had been mysteriously shot at the day before. As a result, the session was inquorate and the election went ahead according to the timetable favoured by the government.³⁵ When the fresh election was finally held on 26 October it thus took place in an atmosphere of confusion and controversy. Given that Kenyatta was the only real candidate, the official turnout of 39% was far from the clear legitimation that Jubilee had sought, but was not insignificant. Ultimately, Kenyatta officially secured 7.5 million votes against 8.2 million in August, although both figures continue to be disputed. The geographical distribution of the vote and fact that polling stations across four counties in Luo Nyanza were unable to open due to protests by opposition supporters also revealed a deeply divided country.³⁶

Winner-takes-all politics and the 2010 constitution

The controversy surrounding the outcome of the 2017 elections demonstrates that political competition in Kenya remains high stakes. Does this suggest that the 2010 constitution has failed in its bid to make Kenyan politics more inclusive, address past grievances, change political culture, and hence render electoral contests less divisive? Perhaps the strongest evidence for such an interpretation is that – in addition to the violence and disputes that surrounded the polls – the government introduced legislation into the National Assembly that sought to amend the country’s electoral regulations to prevent the Court from invalidating an election on procedural grounds alone by requiring judges to demonstrate that any electoral irregularities were large enough to have changed the result.³⁷ Although the Supreme Court retained the right to rule the legislation unconstitutional, the episode highlighted the precarity of Kenya’s democratic gains. In this sense, the contestation over the election not only demonstrated the limited impact of some areas of the constitution, but also its vulnerability. However, while there are many reasons to be cautious about the transformative impact of constitutional reform, to conclude from this that the reforms introduced in 2010 have simply failed would be to oversimplify the complex impact of the multifaceted constitution. There is instead a need for careful counter-factual analysis to demonstrate the effect of new institutional arrangements; for example, it could be that in the absence of a new constitution the political violence and instability surrounding the 2017 elections would have been substantially worse.

The 50% + 1 clause and the presidency

Prior to 2010, what critics called Kenya’s ‘imperial presidency’ gave the incumbent a potent bundle of powers, combining formal control over an enduring administrative framework (known as the provincial administration) with the ability to make appointments to multiple public and parastatal offices and extensive (often informal) influence over a wide range of resources – from government procurement contracts to the distribution of development projects.³⁸ Control over such resources made the president Kenya’s apex patron – the point of convergence for all networks of clientelism.

The 2010 constitution sought to remedy this in multiple ways. In addition to devolving considerable power away from the presidency, the country’s electoral framework was changed. Since 1992, the presidential election had been decided by a plurality, though with a requirement that the winning candidate must also win at least 25% of the vote in at least half of Kenya’s provinces, failing which there would be a run-off. In the 1992 and 1997 elections, Daniel arap Moi of the ruling Kenya African National Union (KANU) had won the presidential vote with less than half of the total votes cast – 37% and 40%, respectively. In contrast, the 2010 constitution demanded an absolute majority to secure victory, with a run-off between the two candidates with the largest popular vote to be held if no one secured over 50% + 1 vote.

This new requirement was part of a wider architecture of laws aimed at encouraging the building of national political parties, which, collectively, have not been very successful. Since 2010, political parties have continued to be mostly ephemeral and linked to particular politicians who are seen, more or less explicitly, as ethnic patrons. The partial exceptions have been ODM and Jubilee, each of which has been the focus of some efforts to create a wider national organisation and membership, though they still rely heavily on key ethnic spokesmen to help mobilise support and remain rooted in particular areas and ethnic groups. The requirement for an absolute majority has however encouraged a tendency, already apparent before 2010, that makes presidential elections more than ‘winner-takes-all’ events. The presidency has emerged as the focus of a prolonged process of negotiation, in which elections appear not as decisive moments that settle the distribution of power, but rather as markers that confirm the terms of one set of elite pacts, while immediately opening a whole new phase of deal-making over the next election cycle.

However, it is not clear that this aspect of Kenyan politics can solely be attributed to the formal political changes made in 2010, in large part because it has been an occasional feature of Kenya’s presidential politics since the single-party era. Moi’s succession to the presidency on the death of Jomo Kenyatta was the product of an elite pact made with an eye on ethnic constituencies.³⁹ During these years, informal backroom negotiations were even more apparent in lower-level races where competition could be intense. The unsuccessful challenger to an established regional ‘Big Man’ could catch the eye of a patron if they secured enough votes: a losing campaign could launch a political career, leading perhaps to appointed office and then back into electoral politics. In presidential politics, both Jomo Kenyatta and Moi, while they had cabinets that were skewed towards their own co-ethnics, also ensured that they had ethnically diverse cabinets that brought in key point-men from across the country.⁴⁰ In turn, in the face of increasingly competitive multi-party politics in the 1990s and a small parliamentary majority, Moi brought various opposition leaders into KANU; most notably, Raila Odinga and his National Development Party (NDP) in the wake of the 1997 elections. Such negotiations became even more significant from 2002, when the presidential election in December was preceded by an intense period of deal-making that produced the broad National Rainbow Coalition (NaRC) behind Mwai Kibaki. The results of the election were – uniquely in recent Kenyan history – accepted by all major players; but the immediate consequence was a renewed period of negotiation amongst an elite whose eyes were set on the next election, and on their need to reassure their support base (in every case largely an ethno-regional one) that their patrons were looking after their interests. In short, the 2002 election results provided a bargaining tool – evidence of a leader’s ability to mobilise ethnic voters in future and a sign of their ability to make the government legitimate in the eyes of their distinctive constituency.

However, the most far-reaching deal-making followed the 2007 elections, which pitted the incumbent President, Mwai Kibaki, now in the Party of National Unity (PNU), against Raila Odinga and ODM. While Kibaki was announced the winner by a narrow margin, multiple and egregious malpractices made a majority of Kenyans and analysts believe that Odinga had actually won.⁴¹ Combined with the dramatic violence of the post-election period, this crisis of confidence prompted a negotiated settlement and the formation of a Government of National Unity (GNU) in February 2008 that ushered in an uneasy peace. When that settlement fractured, the subsequent manoeuvring again looked both back and forward. Voting patterns from 2007, new dynamics introduced by the International Criminal Court’s intervention, and considerations of who would get to be the presidential candidate in future polls, all formed the basis for alliance-building ahead of the next elections in 2013.⁴² Against that background, the exclusion of Odinga and his allies in the wake of the 2013 elections might be the exception rather than the rule – an exception based on the confidence of Kenyatta and Ruto that they had a deal that would endure for another electoral cycle, and perhaps made bearable for Odinga by the extent of the CORD coalition’s wins at lower-level elections.

Viewed in this context, the ‘handshake’ between Odinga and Kenyatta that brought an abrupt end to months of political brinksmanship in March 2018 seems to emerge out of a long-term pattern of elite deal making. While this has been encouraged by the 50% + 1 rule, it has its roots elsewhere and is likely to continue to be a major feature of Kenyan politics into the future. Term limits mean that Kenyatta cannot stand again; so the 2017 election results immediately became the basis for negotiations to put together an alliance that can deliver a majority in 2022. The immediate rewards to Odinga (and subsequently to Odinga’s ally, Kalonzo Musyoka, when he endorsed the deal) were much less significant than those of 2008. The new constitution has made it difficult to give much more than some relatively minor appointments, and Odinga and his allies have struggled to insist that his appointment as an African Union special envoy on infrastructure development is a significant post.⁴³ Yet perhaps more important is the evident sense that Odinga is somehow back in the networks of influence and patronage, appearing with the president at some events, and with some access to power, and that he may be supported by Kenyatta and others in 2022. As a result, the declared winners of the election have not quite taken all – because they have good reason to reach out to at least some losers.

The handshake was also important in a broader sense, because it brought an end to public demonstrations against the result and to the multiple other activities through which Odinga and his allies had sought to bring the legitimacy of the government – and indeed, of the state – into question. As discussion of secession and the setting up of a parallel government were dropped, the focus of politics switched immediately to the 2022 presidential election. At the time of writing, in November 2018, preparations for that election have once again taken the form of elite negotiations. Kenya’s ability to come back from the brink of another electoral crisis perhaps had less to do with the formal constitutional changes introduced in 2010, and more with an established set of informal institutions through which elites have managed, and to an extent shared, power since independence.

Devolution and making national losers local winners

For devolution to reduce the stakes of political competition by giving marginalised groups a stake in the political system (at least) two conditions need to hold. The first is that devolution generates a robust system of government that is not simply an extension of central control and therefore represents a credible opportunity for a broader range of leaders and communities to play a role in the political system. The second is that these positions are genuinely competitive and that opposition parties are able to win significant representation. Only when both conditions are met will parties and communities that lose national level elections feel that they still have a stake in the political system as a result of their representation at the county level.

Despite the controversies that have surrounded the introduction of devolution in Kenya, there are good reasons to think that the first condition holds, at least to a significant extent. Immediately after devolution was introduced in 2010, opposition leaders, governors and journalists fretted that the government – known for resisting efforts to devolve power – would find ways to stifle and ultimately suffocate the new political system. There has been some evidence of that. Just before the new constitution came into force, Kibaki pushed through legislation to re-structure the provincial administration as a parallel administrative structure under the Office of the President, vitiating what some saw as a major aim of devolution.⁴⁴ Devolved powers over the allocation, use and registration of land – again, a core and highly charged issue in the debate over devolution – remain limited and contested.⁴⁵

However, while important, central government encroachment on county level politics has been far less effective than Kenya’s history of decentralisation would have suggested:⁴⁶ governors have shown ‘both the motivation and the capacity to resist capture by central government’.⁴⁷ As a result, the proportion of government revenue given to the counties has consistently exceeded the constitutionally mandated minimum of 15%, hovering around 21% of the annual budget.⁴⁸ This is also a product of broad support. The 2010 constitution was passed by public referendum with a sizeable majority of the vote (69% to 31%), while, according to a national survey conducted in August 2014, 70% of Kenyans still supported the principle of devolution, with this increasing to almost 80% in opposition strongholds.⁴⁹

One significant indicator of the vibrancy of devolution, and the meaning of the devolved positions within the wider body politic, is that those who hold these posts enjoy control over significant resources. In this respect, the impact of devolution has been largely felt through two layers of elected officials: governors and MCAs. Governors enjoy a budget that far exceeds that of MPs and are able to construct their own patronage networks through their control of contracts and appointments.⁵⁰ They are also located at the local level in the counties and are therefore not ‘distant’ from voters as is the case with the national executive. With local presence, they have emerged as prominent regional spokespeople willing to protect and promote local interests to maintain support.⁵¹ They are capable of airing regional grievances on behalf of a far greater number of people than MPs, whose constituencies are much smaller.⁵² Moreover, unlike the councillors who represented the wards in local government under the previous constitution, most MCAs have a ward office, and some have control of a ward development fund despite uncertainty over the legality of the same.⁵³ Precisely because they control resources and are more available than other elected politicians, MCAs have become an increasingly important cog in the Kenyan political machine.⁵⁴

While this has not let to a dramatic change in the way that campaigns are conducted, or the capacity of women to get elected,⁵⁵ it has led to a genuine shift in how Kenyan voters view the importance of different political positions. While the attention of aspirants and voters had tended to focus on presidential and legislative races in the past, the contests for governors and MCAs have become more prominent since 2013. This is reflected in the increased number of candidates contesting for MCA positions in 2017 – a significant increase on 2013 (see Table 1).

Table 1. Number of aspirants for different elected positions in 2013 and 2014.

Post	2013	2017	% change
	No. Contesting	No. Contesting
President	8	8	0.0
Governor	239	211	−11.7
Senator	246	256	4.1
MP	2,133	1,892	−11.3
Women’s rep	155	298	92.3
MCA	9,910	11,857	19.6

Source: Compiled by authors from Independent Electoral and Boundaries Commission (IEBC) gazettes.

While the number contesting gubernatorial seats fell slightly, public attitudes reveal the importance of that position, as well as that of MCA. Significantly, while the MCA position is officially perhaps the least prominent and significant of the roles discussed here, it is seen by many citizens to be one of the most important. In a nationally representative survey conducted by three of the authors in April 2017 (see Figure 1), 70% of respondents identified the presidency as the most important post for them, demonstrating the continued significance of the executive, as already discussed. But tellingly 14% identified MCAs as the most important – the second highest figure overall – while 11% selected governors. MPs, once the critical cog in the Kenyan political machine, polled just 3%. Moreover, the most historically marginalised part of the country, North Eastern Kenya, actually rated the MCA as the most important post. It is therefore clear that devolution has established a robust tier of government that Kenyan politicians and voters are deeply invested in.

Figure 1. What is the most important elected post?

Source: Authors’ survey, April 2017.

The second criteria was also clearly met in 2013, and again in 2017, although to a lesser extent. In the first election to be held under the new devolved system, Odinga’s CORD alliance won around half of all governorships, including the capital city, Nairobi, and the strategically important county of Mombasa, home to the country’s main port. Indeed, in 2013, candidates from Odinga’s own ODM (one component of CORD) won 26 governorships, double that of Kenyatta’s TNA (one component of Jubilee), and elected 377 MCAs – more than any other single party. It was therefore clear that in many parts of the country, Odinga supporters were national losers but local winners and that this was one of several reasons that the controversy surrounding the 2013 elections did not generate a greater degree of unrest and instability.⁵⁶

However, in 2017, the consolatory power of local wins was significantly diminished for two reasons. The first, was a sense that the presidency remained the ultimate prize and one that had once again been stolen from the opposition. The second was the weaker performance of opposition candidates in these lower level races. Many of these outcomes were disputed by the opposition, but in contrast to the presidential election relatively few of the results have so far been overturned. If the results are largely legitimate, they are most likely explained by two factors. First, Kenyatta’s incumbency ensured that Jubilee had effective control over state resources and officials, and was therefore able to outspend the opposition at both the presidential and the county level. Second, in between 2013 and 2017 Jubilee transformed itself from a coalition into the Jubilee Party, which, as Elena Gadjanvoa shows, ‘was able to solve intra-party infighting and put forward single candidates.’⁵⁷ For example, by being able to compensate individuals that lost out in the party primary elections held to select candidates – for example, by offering alternative positions or forms of employment to those not selected – the Jubilee Party suffered relatively few damaging defections. In turn, this led to a more cohesive campaign: whereas in 2013 candidates from different parties within the Jubilee Alliance had competed against each other, dividing the vote in some parts of the country, this did not happen in 2017. Even though Jubilee candidates often still faced competition from ‘friendly’ parties, which supported Kenyatta’s re-election, or from Independents.

By contrast, while Odinga was able to secure a greater degree of unity among NASA leaders than CORD had done in 2013, the opposition alliance was not able to agree on a common slate of candidates. As a result, candidates from different components of NASA competed against one another in many sub-national elections.⁵⁸ Along with the fact that opposition candidates tended to have less funding and fewer opportunities to co-opt state resources and officials, this helps to explain why Jubilee won a greater proportion of sub-national contests than it had done previously. Most notable in this regard was the victory of Jubilee’s Mike Sonko in the symbolically important gubernatorial race in Nairobi. This diminished opposition success in lower level races reduced the extent to which devolution was able to sustain public support for the wider political system. However, the victory of opposition candidates in NASA strongholds such as Nyanza and the Coast also meant that they still controlled considerable resources in their own back yard – and so continue to benefit from devolution. It is therefore unsurprising that opposition leaders continue to call for devolution to be strengthened rather than abandoned.

However, while there is evidence that the introduction of county government has softened the blow of losing national elections,⁵⁹ and thus contributed to political stability, the aftermath of the 2017 elections demonstrates that the consequences of constitutional change are not always predictable or desirable. In the weeks before and after the re-run of the presidential election in October, it became clear that decentralisation may also reinforce patterns of ethnic politics and facilitate challenges to the prevailing constitutional order in at least two ways: first, through the ability of disgruntled leaders to use the county administration to fund efforts to subvert the political system; and second, by fostering sub-national identities that may ultimately undermine efforts to build a cohesive national polity. For example, when Odinga announced the formation of a ‘People’s Assembly’ to represent opposition viewpoints and contest Jubilee’s victory in November 2017 – effectively calling into question the legitimacy of the government and challenging the primacy of the National Assembly – he turned to the counties to make his plan operational.⁶⁰ Several opposition-controlled county assemblies passed bills in support of the formation of assemblies, and also helped to raise funds for opposition activities – for example, by helping to ferry supporters to Nairobi for Odinga’s unofficial swearing-in in January 2018.⁶¹ This county support, together with contributions from the diaspora and wealthy individuals, enabled the cash-strapped opposition to remain active.⁶²

County-level structures were also used to threaten the possible break-up of Kenya as a national unit. For example, at the Coast – which has voted for Odinga in the last three elections by a ratio of about 2 to 1 – it was Mombasa Governor Ali Hassan Joho who made headlines by stating that, along with a number of other elected politicians in the region, he would campaign for secession from Kenya.⁶³ In the process, Joho apparently – if only briefly – offered endorsement to a secessionist movement that has been intermittently active since independence.⁶⁴ The implication – as Hannah Waddilove argues in her contribution to this volume – is that a political framework that forces governors to be responsive to the hopes and fears of their supporters may generate centrifugal pressures as well as integrative ones when popular opinion turns against the political system.⁶⁵

The talk of secession and alternative governments elicited a strong reaction from the government, with dramatic threats of legal clampdowns, and the withdrawal of funding from counties.⁶⁶ These were not followed through, nor did parts of Kenya actually secede. Instead, the rhetoric of both secession and recentralization subsided, as governors and central government stepped back from confrontation in the wake of the ‘handshake’ between Kenyatta and Odinga. Given this, the 2017 election suggests that although the new political structures introduced under devolution have the potential to moderate ‘winner-takes-all’ politics – the continued prevalence of presidentialism notwithstanding – they may also exacerbate sub-national identities in a way that generates new challenges for the political system. While the decision to divide Kenya into 47 smaller counties – as opposed to, say, 7 larger provinces – has reduced the capacity of any one county or group of counties to break away, it is also important to note that the situation remains fluid, and further opposition losses at the sub-national level in future elections could undermine the stabilising effects of devolution.

The Supreme Court and electoral fairness

The final constitutional innovation for discussion is the creation of a Supreme Court with the exclusive mandate to hear and determine presidential petitions. The 2007 election descended into widespread violence, in part, because the opposition did not trust the judiciary – which had been packed with the president’s allies – to make a fair ruling and thus took their dispute to the streets.⁶⁷ Together with long-standing claims of judicial bias and corruption, this ensured that judicial reform was of central concern for the architects of the 2010 constitution, which, among other things, established an independent Judicial Service Commission to recommend judges to the president and established a new Supreme Court with exclusive jurisdiction to hear and determine presidential election disputes. The constitution also ‘set a strict time line of 14 days during which the Supreme Court should hear and rule on a presidential election dispute’ and ‘did away with the requirement for personal service in election petition matters’.⁶⁸

These developments, together with the public vetting of judges and magistrates between the 2007 and 2013 elections, were seen to play a positive role in 2013 when Odinga again rejected the presidential result, but opted to take an election petition to the Supreme Court. However, while the petition helped to bring various irregularities to light – and thus shed further doubt on Kenyatta’s slim first-round victory with 50.07% of the vote – the Court ruled that the petitioners had not shown that the results had changed the outcome and that Kenyatta’s victory should thus hold. Odinga begrudgingly accepted this decision, but continued to insist that the election had been stolen. The decision also attracted broader criticism for having been swayed by a desire to maintain stability and order; disallowed evidence on petty technical grounds to the detriment of the public interest; and thus reinforced ‘the powers of the executive and the model of a unitary state beyond the reach of the law’.⁶⁹

With the memories of 2013 in mind, NASA suggested ahead of the 8 August polls that they would not bother to bring a presidential petition before the Court in 2017. But just before the deadline, NASA filed a petition. Leaders argued that they had been forced to do so by Jubilee who ‘had started a campaign of muzzling civil society and the media’ and that they wanted to give the Court ‘a chance to redeem itself’.⁷⁰ However, once the petition began, it became clear that NASA had likely been preparing to go to Court for some time. Partly because Odinga kept his cards so close to his chest, and partly because of the decision of the Court in 2013, Jubilee leaders appeared to be complacent, assuming that Kenyatta’s victory – which was much clearer than it had been in 2013 – would be upheld. However, in a landmark ruling on 1 September 2017, the Supreme Court declared in a majority decision of 4 to 2 that, while the impact on the final results was unknown, the IEBC had not conducted the presidential election of 8 August in accordance with the constitution and applicable laws and that the election should be re-run.

However, while this landmark decision was widely acclaimed at home and abroad as a sign of judicial independence and ‘democratic maturity’, it prompted a backlash from President Kenyatta and his allies as the president castigated the judges as ‘criminals’ who he would ‘fix’ for having ‘overturned’ the will of the people.⁷¹ At the same time, and as discussed above, Jubilee pushed through new legislation that would make a future petition more difficult, whilst public infighting within the IEBC brought the electoral commission into further disrepute. As Karuti Kanyinga and Collins Odote outline in this volume, Jubilee’s response formed part of a broader pattern of the politicisation of the judiciary over the course of the electoral cycle as politicians from across the political divide sought to use the courts to advance their own agendas. As part of this process, judicial decisions were consistently criticised by the losing side.

The experience of the judiciary in 2017 points to both its strengths and limitations: the courts displayed a willingness to make decisions independent of, and against, both the government and opposition; but proved unable to ensure that those decisions were accepted by the losing side, or to ensure that the electoral process itself enjoyed broad credibility. This is for two key reasons. First, like most judiciaries around the world, the Supreme Court lacks the power of the purse or the sword. In the absence of financial resources, enforcement capacity, and legislative power, the impact of the Court’s rulings thus largely depends on how others respond to them.⁷² This was evident in the government’s response to the Court’s nullification of the first presidential election, which included a vilification of the judiciary, a refusal to discuss electoral reform, the implementation of new laws that would make a future petition more difficult, and a clear lack of concern as to the IEBC’s crisis of legitimacy. It was also evident in NASA’s decision to boycott the ‘fresh polls’ that were ordered by the Supreme Court, to refuse to bring a second petition before the Court, and in its rejection of the Court’s ruling in this second petition as the result of political pressure.

Second, the courts operate within a wider political system in which their operations are not grounded in a set of supportive informal norms – instead, their independence is consistently undermined by persistent patron-client politics and breaches of the rule of law.⁷³ This is critical as it helps to explain how and why politicians were able to respond in the ways that they did and not lose the support of the majority of their constituents. The outcome is a judiciary that can – and did – make important decisions that influenced the electoral process, but which could not determine the course of political debate or ensure the credibility of the polls.

Institutions and the process of democratisation

As in all countries, political stability and democratisation in Kenya depend on the interaction of formal and informal institutions.⁷⁴ As Helmke and Levitsky have argued, formal and informal institutions may compete, with the latter undermining the former, but customs and norms may also support strong democratic institutions and compensate for weak ones.⁷⁵ It is easy, given Kenya’s history of corruption and electoral controversies, to assume that the relationship between formal rules and informal norms is always problematic. Indeed, many of the formal provisions in the country’s new constitution and electoral regulations, such as rules regarding the regulation of political parties, have been undermined by contradictory informal processes that are often described as ‘patrimonial’, such as patron-client leadership and handouts. This dynamic is particularly challenging as a result of its interactive nature: as we have seen, the introduction of new political structures designed to devolve power has had the unintended consequence of breathing fresh life into a divisive form of ethnic politics at the county level.

This tension between formal and informal institutions is evident in each of the submissions to this volume. The contributions show how devolution has created important new positions that have changed political dynamics in Kenya, without fundamentally changing the norms, values and expectations that animate electoral competition. Politicians still mobilise support through patronage politics and along ethnic lines by seeking to present themselves as those best positioned to protect and promote the interests of their constituencies in ways that disadvantage women aspirants.⁷⁶ Incumbents seek to maintain control of key sources of patronage such as land and still benefit from skewed powers and resources.⁷⁷ Important police reforms between 2007 and 2013 were undermined by a failure to address an institutional culture that supported the use of excessive force.⁷⁸ A better prepared judiciary was able to deal more efficiently with both the pre-election cases and post-election petitions that were brought before it, yet was not immune to politicians’ efforts to use the courts for their own ends. At the same time, legal decisions revealed ideological and political divisions within the judiciary, while the Supreme Court was able to act as an important election management body, but not to ensure the polls’ credibility.⁷⁹

However, the contributions to this volume also reveal how formal and informal institutions do not always compete, and are sometimes complementary. In this vein, patronage politics in the context of devolution is shown to help guard against ethnic violence as politicians seek to ensure stability within their own counties.⁸⁰ Moreover, as this Introduction has demonstrated, the political stability achieved in Kenya through the ‘handshake’ was only possible because of the existence of strong and historically rooted informal institutions of patron-client ties and elite collusion, which enabled once bitter rivals to negotiate their way out of a political crisis. In other words, informal institutions help to explain both why Kenyan elections so often take the country to the brink of disaster, and why it has yet to fall into the abyss.

As we have argued, the fact that political leaders ultimately found a way to avoid further unrest speaks to the relatively high level of elite cohesion in Kenya, which means that despite pronounced ‘winner-takes-all’ tendencies, losers do not always ‘lose-all’. Indeed, it is precisely the willingness of elites to take the interests of some of their rivals into account that has consistently brought Kenya back from the brink of a deeper and more irreparable political crisis. Most notably, it was this capacity to compromise across ethnic lines at critical moments that underpinned a peaceful transfer of power from Kenya’s first president, Jomo Kenyatta, to his vice-president from a different ethnic group, Daniel arap Moi. And it was the same process of negotiation and accommodation between Moi and the leader who replaced him as president, Mwai Kibaki, that facilitated the country’s first peaceful transfer of power via the ballot box in 2002. The creation of a power sharing government to end the post-election violence in 2007 – and indeed the cross-party movement in support of the 2010 constitution – rested on similar foundations. But just as these prior processes have often seen stability achieved at the expense of genuine political and economic transformation, the ‘handshake’ of 2017 also came with costs.

The agreement between Kenyatta and Odinga was a relief to many because it ended a dangerous period of instability. However, it did little to resolve the factors that gave rise to it. Odinga cannot erase his statements about Kenyatta’s lack of legitimacy as president, or NASA’s decision to swear him in as the people’s president. Similarly, Kenyatta cannot walk back his threats to the judiciary, or the way the security forces were used for partisan ends. Instead, these actions have further entrenched existing grievances and divisions, which means that it will be even harder to persuade voters that the next elections will be free and fair.⁸¹ Turning this situation around will require the institutionalisation of the reforms introduced in 2010 – a process that will require much greater political will than has so far been available. However, we should not underestimate the way in which elite pacts have underpinned the emergence of a more dynamic, competitive and democratic political system in Kenya, if only by preventing it from falling apart.

Acknowledgements

We would like to thank Jamal Abdallah, Castrol Baraza, Job Bwonya, Ngala Chome, Denis Galava, Mwongela Kamencu, Hasan Kochore, Nzau Musau, Hassan Mwakimako, Peter Nderitu, Kevin Obware, George Odhiambo, Betty Okero, Naija Robert, George Sayagie, and Veronica Waitimu for their insights and assistance, which has contributed to the overview and analysis laid out. We would also like to thank two anonymous reviewers and the journal editors for their comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This article draws in part from research supported by the UK Economic and Social Research Council under grant ES/L002345/1, ‘The impact of elections in sub-Saharan Africa’.

Notes

1 Branch and Cheeseman, “Democratization, Sequencing, and State Failure in Africa.”

2 Mueller, “The Political Economy of Kenya’s Crisis.”

3 Gĩthĩnji and Holmquist, “Kenya’s Hopes and Impediments.”

4 Cheeseman and Murray, “Power Sharing in Kenya.”

5 Ghai, “Constitutions and Constitutionalism.”

6 Bleck and Van de Walle, Electoral Politics in Africa.

7 Kanyinga and Odote, “Judicialisation of Politics.”

8 Waddilove, “Support or Subvert.”

9 Cheeseman and Tendi, “Power-sharing in Comparative Perspective.”

10 Mutahi and Ruteere, “Violence, Security and Policing”; Waddilove, “Support or Subvert.”

11 Waddilove, “Support or Subvert.”

12 North understands institutions as ‘humanly devised constraints that structure political, economic and social interactions’ (“Institutions,” 97). It is important to note that both formal organisations (such as legislatures and parties) and informal practices (such as established norms and customs) may count as institutions as long as they feature broadly understood rules that individuals cannot break without exposing themselves to some form of sanction.

13 Cheeseman, “Conclusion,” 373.

14 See for example, Chabal and Daloz, Africa Works.

15 North, “Institutions.”

16 Daily Nation, “Famous Uhuru-Raila Handshake Leaves Opposition in Limbo.” May 20, 2018. https://www.nation.co.ke/news/politics/Famous-Uhuru-Raila-handshake-leaves-opposition-in-limbo/1064-4570644-n1itws/index.html.

17 EUEOM, Final Report, 26–27.

18 For some examples see, EUEOM, Final Report.

19 Cheeseman, Lynch, and Willis, “Democracy and its Discontents.”

20 For example see, Throup and Hornsby, Multi-Party Politics in Kenya.

21 Kanyinga and Odote, “Judicialisation of Politics.”

22 The Star, “IEBC Cancels Tender for Poll Equipment, Seeks to Lease.” March 23, 2017. https://www.the-star.co.ke/news/2017/03/23/iebc-cancels-tender-for-poll-equipment-seeks-to-lease_c1529673.

23 Daily Nation, “IEBC’s Christopher Msando was the Face of Polls Technology.” August 1, 2017. https://www.nation.co.ke/news/IEBC-s-Christopher.

24 The East African, “Foreign Observers Urge Patience as Kenya Awaits Poll Results.” August 10, 2017. https://www.theeastafrican.co.ke/news/ea/Foreign-observers-on-Kenya-elections/4552908-4052606-125uxct/index.html.

25 Daily Nation, “Raila Odinga says IEBC Computer System Hacked.” August 9, 2017. https://www.nation.co.ke/news/politics/Raila-Odinga2017--Elections-hacked/1064-4050980-dsmk7r/index.html.

26 Ibid.

27 The Star, “Declare Raila Winner Now, NASA Tells IEBC, Places Tally at 8.04 Million.” August 10, 2017. https://www.the-star.co.ke/news/2017/08/10/declare-raila-winner-now-nasa-tells-iebc-places-tally-at-804-million_c1614417.

28 Cheeseman, Lynch, and Willis, “Were the Kenya Elections Conducted Successfully?” Foreign Affairs, August 15, 2017. https://www.foreignaffairs.com/articles/africa/2017-08-15/were-kenyan-elections-conducted-successfully.

29 Mutahi and Ruteere, “Violence, Security and Policing.”

30 Standard Digital, “Reasons Why Presidential Election was Declared Invalid, Court Gives Full Verdict.” September 20, 2017. https://www.standardmedia.co.ke/article/2001255107/reasons-why-presidential-election-was-declared-invalid-court-gives-full-verdict.

31 Kanyinga and Odote, “Judicialisation of Politics.”

32 Daily Nation, “MPs Set out to Change Election Law Ahead of Repeat Poll.” September 27, 2017. https://www.nation.co.ke/news/Kenya-MPs-set-out-to-change-election-law/1056-4114514-hntog1/index.html.

33 Daily Nation, “Raila Odinga Quits: NASA’s Full Statement.” October 10, 2017. https://www.nation.co.ke/news/Raila-Odinga-quits-repeat-presidential-election-Kenya-Nasa/1056-4133616-tr6we4z/index.html.

34 The Star, “IEBC Can’t Deliver Credible Poll, Roselyn Akombe Says as she Throws in Towel.” October 18, 2017. https://www.the-star.co.ke/news/2017/10/18/statement-iebc-cant-deliver-credible-poll-roselyn-akombe-says-as-she_c1654668; Daily Nation, “Wafula Chebukati: I Can’t Guarantee Credible Poll on October 26.” October 18, 2017. https://www.nation.co.ke/news/Wafula-Chebukati-on-repeat-presidential-election/1056-4145232-oyj67sz/index.html.

35 Odote and Kanyinga, “Judicialisation of Politics.”

36 Waddilove, “Support or Subvert?”

37 Daily Nation, “Parliament Passes Bill to Amend Laws.” October 11, 2017. https://www.nation.co.ke/news/politics/MPs-set-to-pass-election-laws/1064-4134458-bm3hh8/index.html.

38 Barkan and Chege, “Decentralizing the State.”

39 Karimi and Ochieng, The Kenyatta Succession.

40 Hornsby, Kenya.

41 Kanyinga, Long, and Ndii, “Was It Rigged?”

42 Lynch, “Electing the ‘Alliance of the Accused’.”

43 Daily Nation, “Raila Outlines his Role as African Union’s Special Envoy.” October 24, 2018. https://www.nation.co.ke/news/Raila-outlines-his-new-role-at-AU/1056-4820330-kitygr/index.html.

44 Hassan, “Continuity Despite Change.”

45 D’Arcy and Nitstotskaya, “Local Grievances, National Control.”

46 Barkan and Chege, “Decentralising the State.”

47 Cheeseman, Lynch, and Willis, “Decentralisation in Kenya,” 6.

48 Cheeseman, Lynch, and Willis, “Decentralisation in Kenya,” 6. This figure is based on the country’s annual budget although the constitution demands that calculations are based on audited accounts. If this figure is based on the last audited account, the share is around 30%, but the audited accounts are 2 years behind.

49 Cheeseman, Lynch, and Willis, “Decentralisation in Kenya,” 30.

50 Steeves, “The 2017 Election in Kenya,” 493.

51 Lockwood, “The Buffalo and the Squirrel.”

52 Cheeseman, Lynch, and Willis, “Decentralisation in Kenya.”

53 Standard Digital, “Controller of Budget Deals a Blow to MCA’s Bid for Ward Funds.” February 14, 2018. https://www.standardmedia.co.ke/article/2001269631/controller-of-budget-deals-a-blow-to-mca-s-bid-for-ward-funds.

54 Waddilove, “Support or Subvert?”

55 Bouka, Berry, and Kamuru, “Women’s Political Inclusion.”

56 Cheeseman, Lynch, and Willis, “Democracy and its Discontents.”

57 Gadjanova, “Treacherous Coat-Tails.”

58 Gadjanova, “Treacherous Coat-Tails.”

59 Mutahi and Ruteere, “Violence, Security and Policing.”

60 Daily Nation, “NASA Leaders, Governors Agree to Launch People’s Assembly.” November 2, 2017. https://www.nation.co.ke/news/politics/Nasa-governors-launch-people-assembly/1064-4167384-v57icc/index.html.

61 The Star, “Kisumu, Migori MCAs Debate People’s Assembly bills, Expected to Sail Through.” November 14, 2017. https://www.the-star.co.ke/news/2017/11/14/kisumu-migori-mcas-debate-peoples-assembly-bills-expected-to-sail_c1668720.

62 It is important to note that this support was not unconditional. County-level politicians had to balance such national demands against the need to pay back debts acquired during the election campaigns and deliver on development at the local level. See Waddilove, “Support or Subvert.”

63 Standard Digital, “Coast Leaders Renew Plot for New Political Party after Secession Talk.” February 18, 2018. https://www.standardmedia.co.ke/article/2001270190/coast-leaders-renew-plot-for-new-political-party-after-secession-talk.

64 Willis and Gona, “Pwani C Kenya?”

65 Waddilove, “Support or Subvert.”

66 Daily Nation, “AG Seeks to Block People’s Assemby.” January 22, 2018. https://www.nation.co.ke/news/politics/AG-seeks-to-block-People-s-Assembly/1064-4274148-125qwtsz/index.html; Daily Nation, “MPs Say Jubilee Plans to Deny NASA Counties Funds.” December 10, 2017. https://www.nation.co.ke/counties/homa-bay/MPs-say-Jubilee-plans-to-deny-Nasa-counties-funds/3444904-4222746-ve9ufe/index.html.

67 Branch and Cheeseman, “Democratization, Sequencing, and State Failure in Africa.”

68 Kanyinga and Odote, “Judicialisation of Politics.”

69 Harrington and Manji, “Restoring Leviathan?,” 175; also Kanyinga and Odote, “Judicialisation of Politics.”

70 Mutahi and Ruteere, “Violence, Security and Policing of Kenya’s 2017 Elections.”

71 Kanyinga and Odote, “Judicialisation of Politics.”

72 Cheeseman, Lynch, and Willis. “Kenya’s Supreme Court.” Op Cit.

73 Ibid.

74 Cheeseman, “Conclusion.”

75 Helmke and Levitsky, “Informal Institutions and Comparative Politics.”

76 Bouka, Berry, and Kamuru, “Women’s Political Inclusion”; Lockwood, “The Buffalo and the Squirrel”; Waddilove, “Support or Subvert?”

77 D’Arcy and Nistotskaya, “Intensified Local Grievances”; Gadjanova, “Treacherous Coat-Tails.”

78 Mutahi and Ruteere, “Violence, Security and the Policing of the Elections.”

79 Kanyinga and Odote, “Judicialisation of Politics.”

80 Mutahi and Ruteere, “Violence, Security and Policing of Kenya’s 2017 Elections”; Waddilove, “Support or Subvert?”

81 Kanyinga and Odote, “Judicialisation of Politics”; Mutahi and Ruteere, “Violence, Security and Policing”; Waddilove, “Support or Subvert?”