Autocratisation, electoral politics and the limits of incumbency in African democracies

ABSTRACT

The world is experiencing a new wave of autocratisation, characterised by a global democratic reversal. From 2010 to 2020, the share of the world population living in autocracies increased from 48 to 68%. Electoral autocracies are now the world's most common regime type, and along with closed autocracies they number 87 of the world's 195 states. Even during the height of the third wave of democratisation, elections in Africa rarely led to an alternation of power. Thirty years after the third wave, this special issue introduction takes stock of how many transfers of power occurred in the three crucial decades between 1991 and 2021. In this special issue, we focus on Zambia to understand some of the factors that contributed to an electoral turnover, notwithstanding the many benefits of incumbency that were enjoyed by the ruling Patriotic Front (PF) led by President Edgar Lungu. We show that the outcome of Zambia's August 2021 election demonstrates the limits of incumbency. We suggest that voters and opposition parties in countries with previous experiences of peaceful transfers of power might rely on a ‘democratic muscle-memory’, to dislodge autocrats and call for more research on when and why incumbents lose.

KEYWORDS:

• Alternation
• incumbency advantage
• democracy
• turnover
• autocratisation
• Zambia

The world is experiencing a new wave of autocratisation, characterised by a global democratic reversal. The Varieties of Democracy Project's (V-Dem) 2021 report notes that from 2010 to 2020, the share of the world population living in autocracies has increased from 48% to 68%.¹ Electoral autocracies are now the world's most common regime type, and along with closed autocracies they number 87 of the world's 195 states. V-Dem argues that this constitutes a ‘third wave of autocratisation’, rolling back many of the gains made since the 1990s.² The number of democratising countries, on the other hand, has halved in the last decade to just 16 – home to just 4% of the world's population. This pessimism is matched by analyses from Freedom House who also argue that in the last 15 years, the global balance has shifted from democratisation towards tyranny, with nearly 75% of the global population living in countries whose democratic credentials declined in 2020.³ Sub-Saharan Africa has noted a substantial decline in the quality of its democracy, with an increase in the number of military seizures of power experienced since 2019. In 2021 alone, coups occurred in Sudan, Guinea, Mali, and Niger.

Even during the height of the third wave of democratisation, elections in Africa rarely led to an alternation of power. As far back as 2010, Cheeseman argued that incumbents won 88% of elections,⁴ and when writing a decade later Bleck and van de Walle counted just 21 cases from 184 elections between 1990 and 2015 where a presidential incumbent was defeated.⁵ Despite the difference of a decade, both find that presidential incumbents win about 88% of all elections in which they run. But given that nine of those 21 cases were in the founding elections of the early 1990s, the rate of incumbent loss in Africa since the mid-1990s is much lower. Between 2000 and 2015, just nine sitting presidents were removed from power in Africa. What accounts for the particularly high re-election rates of African incumbents? Of course, much has been written about the incumbency advantage in African elections. Incumbency advantage effects are central to the ability of presidents to retain power, with the capacity of presidents – especially in super-presidential systems – to control key institutions, write preferential electoral rules, suppress dissent, and tilt the electoral playing field in their favour.⁶ These effects can be the result of the benefits conferred on the incumbent in democratic systems by virtue of their position at the top of the state – such as greater visibility, or the ability to claim policy successes – or as a result of using more illiberal means of shifting the playing field in their favour.⁷

In this special issue, we focus on Zambia to understand some of the factors that contributed to an electoral turnover, notwithstanding the many benefits of incumbency that were enjoyed by the ruling Patriotic Front (PF) led by President Edgar Lungu. As well as demonstrating the limits of incumbency, the outcome of Zambia's August 2021 election has the potential to generate new perspectives on some of the ongoing academic debates on elections and political change in Africa's democracies. These include the question of what motivates voters in Africa beyond patronage and ethnic clientelism, the influence of the military in democratic transitions, the growing power of social media on party campaigns, and the role of the electoral commission, the judiciary and election observation in shaping political outcomes. We also examine the impact of previous transitions of power on the likelihood of future transfers and what might lead to such turnovers, even in contexts of growing authoritarianism. The special issue comprises seven articles divided into three broad categories: campaigns and media, voters, and institutions.

Democratic turnovers in Africa: 1991–2021

Fairly common in the 1990s, the defeat of sitting presidents in Africa's multi-party democracies became rare during the 2000s as incumbents – enjoying greater advantage over their political opponents – almost always won re-election.⁸ This seemingly invulnerable power of incumbency prompted researchers of African politics to shift their attention from studying party alternation to exploring the relationship between elections and democratisation.⁹ Much of the literature that examines this “second wave” of turnovers argues that open-seat elections – that is electoral contests where the incumbent, constrained by term limits, is not seeking re-election – are more likely to lead to political change. This position was expressed most notably by Cheeseman¹⁰ who argued that open-seat polls are particularly likely to result in opposition victories in sub-Saharan Africa because of the challenges that they pose for ruling parties and because they are often more transparent and fairer.

Most recently, Arriola et al. have argued that African incumbents have been able to forestall further democratisation using lawfare and international alliances, leading to what they refer to as stagnant or anaemic democratisation in Africa since the 1990s.¹¹ However, their continent-wide analysis undertaken using V-Dem data obscures wide variation in democratic backsliding – through autocratisation and coups – and democratic breakthrough at the ballot box in unexpected places such as The Gambia. They outline the ways that incumbent advantage limits further democratisation, but they do not theorise about how these advantages might be overcome. So following unexpected victories by opposition candidates against incumbents in presidential elections in Cote d’Ivoire (2010), Zambia (2011), Senegal (2012), Nigeria (2015), Ghana (2016), The Gambia (2016), Sierra Leone (2018), and, more recently, in Malawi (2020) Seychelles (2020) and Zambia (2021), it is perhaps time to theorise more carefully about the possible limits of incumbency, or under what conditions incumbents are less likely to succeed at the ballot box.

In the last three years, there has been clear evidence of a general decline in Africa's democracy. Since 2020, there have been attempted coups in the Central African Republic, Guinea Bissau, The Gambia, Sao Tome and Principe and Sudan. There have also been two successful coups in Burkina Faso and Mali and one each in Guinea and Sudan. Meanwhile, leaders in Cote d’Ivoire and Guinea won unconstitutional third terms and both did so amidst the COVID-19 pandemic. During this period, Malawi, Seychelles and Zambia held elections which resulted in the defeat of an incumbent.¹² What is remarkable about the latest turnovers in Africa is that most of them occurred in elections where the ruling party was running an incumbent candidate and under authoritarian conditions – worsened by the COVID-19 pandemic – that were extremely unfavourable to the opposition. Like other countries on the continent, Malawi and Zambia have had their fair share of challenges with democracy in the years leading to their most recent national elections. Yet these countries still managed to overcome an entrenched incumbent to enable a transfer of power to the opposition. How common is this for the rest of Sub-Saharan Africa?

Building on the database developed by Bleck and van de Walle,¹³ we have extended their analysis up to the end of 2021 and have thus included all transfers of power that have occurred in the thirty years since the beginning of the “third wave”. But we have also expanded the table to include a consideration of whether the turnover is the country's first or whether a turnover had occurred previously. Countries that had not previously experienced a turnover during this period are coded in grey (Table 1).

Table 1. Turnovers of power in presidential elections in sub-Saharan Africa, 1991–2021^a.

Country	Election year	Open seat	Winner	Voting system	Previous turnover
Benin	1991	N	Nicéphore Soglo	TRS	N
Cape Verde	1991	N	António Mascarenhas Monteiro	TRS	N
Zambia	1991	N	Frederick Chiluba	FPTP	N
Burundi	1993	N	Melchiore Ndadaye	TRS	N
Central African Republic	1993	N	Ange-Félix Patassé	TRS	N
Madagascar	1993	N	Albert Zafy	TRS	N
Niger	1993	Y	Mahamane Ousmane	TRS	N
Malawi	1994	N	Bakili Muluzi	FPTP	N
South Africa	1994	N	Nelson Mandela		N
Mauritius	1995	N	Navin Ramgoolam		N
Sierra Leone	1996	Y	Ahmad Tejan Kabbah	TRS	N
Benin	1996	N	Mathieu Kérékou	TRS	Y (1991)
Madagascar	1996	N	Didier Ratsiraka	TRS	Y (1993)
Guinea-Bissau	1999	N	Kumba Yala	TRS	N
Cote d’Ivoire	2000	Y	Laurent Gbagbo	TRS	N
Ghana	2000	Y	John Kofuor	TRS	N
Mauritius	2000	N	Sir Anerood Jugnauth		N
Senegal	2000	N	Abdoulaye Wade	TRS	N
Madagascar	2001	N	Marc Ravalomanana	TRS	Y (1993, 1996)
Cape Verde	2001	Y	Pedro Pires	TRS	Y (1991)
Kenya	2002	Y	Mwai Kibaki	FPTP	N
Mali	2002	Y	Amadou Toumani Touré	TRS	N
Guinea-Bissau	2005	N	Nino Veira	TRS	Y (1999)
Liberia	2005	Y	Ellen Johnson Sirleaf	TRS	N
Mauritius	2005	N	Navin Ramgoolam		Y (1995)
Benin	2006	Y	Yayi Boni	TRS	Y (1991, 1996)
Sierra Leone	2007	Y	Ernst Bai Koroma	TRS	Y (1996)
Ghana	2008	Y	John Atta Mills	TRS	Y (2000)
Guinea	2010	Y	Alpha Condé	TRS	N
Cote d’Ivoire	2010	N	Alassane Ouattara	TRS	Y (2000)
Sao Tomé and Principe	2011	N	Manuel Pinto da Costa	TRS	N
Cape Verde	2011	Y	Jorge Carlos Fonseca	TRS	Y (1991, 2001, 2011)
Zambia	2011	N	Michael Sata	FPTP	Y (1991)
Niger	2011	Y	Mahamadou Issoufou	TRS	Y (1993)
Senegal	2012	N	Macky Sall	TRS	Y (2000)
Mali	2013	Y	Ibrahim Boubacar Keita	FPTP	Y (2002)
Kenya	2013	Y	Uhuru Kenyatta	FPTP	Y (2002)
Nigeria	2015	N	Muhammadu Buhari	TRS	N
Benin	2016	Y	Patrice Talon	TRS	Y (1991, 1996, 2006)
Comoros	2016	Y	Azali Assoumani	TRS	N
Ghana	2016	N	Nana Akufo-Addo	TRS	Y (2000, 2008)
The Gambia	2016	N	Adama Barrow	FPTP	N
Liberia	2017	Y	George Weah	TRS	Y (2005)
Madagascar	2017	N	Andry Rajoelina	TRS	Y (1993, 1996, 2001)
Sierra Leone	2018	Y	Julius Maada Bio	TRS	Y (1996, 2007)
Guinea-Bissau	2019	N	Umaro Sissoco Embaló	TRS	Y (1999, 2005)
Malawi	2020	N	Lazarus Chakwera	TRS	Y (1994)
Seychelles	2020	N	Wavel Ramkalawan	TRS	N
Cape Verde	2021	Y	José Maria Neves	TRS	Y (1991, 2001, 2011)
Zambia	2021	N	Hakainde Hichilema	TRS	Y (1991, 2011)

^a Data for the period 1991–2013 from Bleck and van de Walle, pp. 75–76. Data from 2013 to 2021 collected by the authors. Malawi's 2014 election is excluded from the table because there was no clear transfer of power from the ruling Democratic Progressive Party to Joyce Banda's People's Party in 2012, Elections in Burkina Faso (2015) and DRC (2018) were excluded from the sample because of the uncertainty after the coup and the rigging in Tshisekedi by the incumbent respectively, making it difficult to classify both as a ‘turnover’.

In the 46 countries of Sub-Saharan Africa, there were 50 turnovers of power in 25 countries in the 30 years between 1991 and 2021. This means that 21 countries in SSA are yet to experience a peaceful transfer of power from the ruling party to the opposition. From 2000 to 2021, 64% (23) of all turnovers occurred in countries which had previously experienced a transfer of power, while 36% (13) of turnovers occurred for the first time in the twenty-first century.

There were 22 transfers of power between 2010 and 2021, in a context of declining democracy. But only six (27%) occurred in countries that had not previously experienced a turnover. During this decade, Guinea (2010), Sao Tomé and Principe (2011), Nigeria (2015), Comoros (2016), The Gambia (2016), and the Seychelles (2020) were the only countries to undergo their first democratic transition via the ballot box, after opposition challengers defeated incumbent presidents. By comparison, turnovers happened in eleven countries where there had previously been a transition. In total, there were sixteen elections in these eleven countries that resulted in turnovers. Of the sixteen elections, incumbents were defeated eight times when they ran against challengers. In eight cases, challengers won in an open-seat election.

Of all 50 transitions, 21 (42%) occurred in the context of open-seat elections where there was no sitting incumbent. But if we discount transitions that occurred in the first decade of democratisation, the period from 2000 to 2021 saw challengers win seventeen open-seat elections, accounting for nearly half of all 36 transitions. Correspondingly, in nearly 50% of all transfers of power since 2000, incumbents have lost elections. So what does this mean for thinking about the limits of incumbency? To be clear, electoral turnovers in Africa do not guarantee democratisation or democratic consolidation and most countries on the continent can best be described as hybrid regimes. In many African countries where alternation of power has occurred – as in Zambia in 2021 – opposition parties won elections notwithstanding the presence of authoritarian institutions. In power, newly elected parties have little incentive to democratise the political system and level the playing field, because doing so would disadvantage their own electoral prospects.¹⁴

It is also evident that countries that experienced turnovers between 1990 and 2009 were twice as likely to experience turnovers in the last decade than countries that never experienced alternation. This allows us to theorise about countries with a history of electoral alternation. Are citizens in these countries more likely to build a “democratic muscle-memory”, which makes them more likely to sanction their leaders in periods of severe economic decline, than countries that lack such a memory? It does appear that while alternations in the 1990s and 2000s did not guarantee democratic consolidation, they appear to have “weakened” the extent of authoritarianism. Several factors explain this apparent waning of authoritarianism and decline in incumbency. These include the weak institutionalisation of political systems that often manifests itself in the fragmentation of incumbent parties often due to succession disputes. The failure for incumbent parties to substantially improve the living standards of their citizens has often led to widespread discontent that has proved difficult to mediate using authoritarian tactics. Moreover, the changing media landscape has provided alternative – and more effective – platforms for citizens, the opposition, and civil society, to hold incumbents accountable.

Understanding the declining power of incumbency advantage

Thirty years since the democratic wave of the early 1990s, the idea that regular multiparty elections would lead to democratisation has not proved true. Incumbents still benefit from disparities in access to financial resources and the media, and from controlling state institutions to the disadvantage of opposition parties. What this demonstrates is that incumbency has its limits. There are at least four factors that explain why more countries have experienced turnovers since 2010, but not in the preceding decades, despite the recent decline in democracy.

The first is that the growth of the party system in the multiparty era in Africa did not take off in several countries until nearly a decade and a half after the transitions to democracy. The broad-based nature of the alliances that were forged to topple one-party regimes across the continent meant that following the transitions to multiparty politics in the1990s, opposition to the new parties in power was concentrated within such dominant parties rather than outside them. The fact that incumbents, most of whom were elected on a popular mandate, ran in the second round of multiparty elections in the mid-1990s meant that budding successors shelved their ambitions to a later date but chose to remain in the ruling party where they attempted to build power bases. It was not until the early 2000s that competitive politics returned to the fore in several African democracies. A key reason for this change in the direction of political life was poorly managed successions that led to the exodus of many political elites from ruling parties. Such elites went on to form their own opposition parties, but a combination of both internal and external factors meant that it was not until the 2010s that they were able to pose a significant threat to, or indeed wrestle power from, incumbents.

The second factor that undermined competitive politics and alternation before the 2010s was the state of the economies that had been inherited from the one-party state. At the time of the transition to multiparty democracy, the economies of many African countries, most of which had been run on a command model, were in a precarious state. The transition from a state-run economy to free enterprise and the constant appeals for patience by ruling elites attempting to repair the economy meant that the consequences of this turn to neoliberalism on organised labour did not appear until the late 1990s and early 2000s when several industries collapsed and tens of thousands of jobs in the civil service were lost.¹⁵ This changing economic context created vast numbers of urbanites with shared social demands who became the political base of populist politicians, many of them with roots in the ruling parties, seeking to wrestle power from incumbents.¹⁶

Added to this, most urbanites were new young voters, born in the 1990s, whose globalised sense of how young people live and consume in other contexts (due to increased usage of social media), left them more likely to prioritise the economy. Even rural areas, mainly agrarian ones, were greatly affected by the state's inability to provide farming inputs on a large scale which was characteristic of the one-party regimes. Moreover, the poor state of the economy even under the multiparty era reduced the capacity of political elites to buy off voters, many of whom became increasingly suspicious of short-term financial fixes in place of their longstanding concerns. The result was the migration of vast constituencies of support that had previously served as the backbone of the governing parties to the opposition. It was these bases that helped the new opposition parties to defeat incumbents, starting in the 2010s.

The third factor is the changing media landscape. Across the continent, the state's monopoly on media ownership and control was entrenched at the start of the 1990s. As part of expressing their commitment to democratic reforms, the new political leaders who succeeded the authoritarian ones opened the media industry to private players over the course of the 1990s. Many of the independent newspapers and community radio stations that were established collapsed while the few that survived took time before they could reach a broader audience. By the early 2000s, the media landscape in much of the continent was still dominated by state-run publications and broadcasting stations. This picture began to change in the late 2000s with the rise of social media, which complemented the developing private media. As Lynch and Gadjanova have shown in this special issue, both platforms have served as important spaces for opposition voices, increased the consumption of alternative information and undermined the power that incumbent parties had enjoyed in electoral contests.¹⁷

The final factor is the role of civil society. Until the second decade of multiparty democracy, civic organisations in much of Africa were primarily concerned with holding the executive branch of government to account, especially on matters of public accountability, constitutional reform, and respect for presidential term limits.¹⁸ Starting in the 2000s, civil society expanded its scope of work to include working closely with electoral bodies to build trust in the electoral process. As well as raising awareness on the importance of voting, civic bodies have employed robust measures to monitor competitive elections and prevent rigging or manipulation of election results. The importance of civil society to alternation was demonstrated in the recent turnovers in Malawi and Zambia. In both countries, non-state actors took vote protection seriously and conducted parallel vote tabulation that captured the election results at polling station level, ensuring that any manipulation would be exposed.¹⁹ Elsewhere, especially in countries like Kenya, Malawi and Zambia, civic groups also initiated several court cases against the abuse of state power, the persistent attacks on the erosion of the rule of law and human rights. Though these cases do not always result in victories, they help increase voter awareness, draw attention to the erosion of democracy, and delegitimise authoritarian ruling parties. All this demonstrates just how important civil society groups can be, and why it is essential to support them through hard times.

Zambia as a case study in the limits of incumbency advantage

Zambia has long been seen as a bell-weather of democratisation in Africa. Having been frequently cited as one of the earliest democratisers in 1991, with the election of the Movement for Multiparty Democracy (MMD) that peacefully retired the United National Independence Party (UNIP) and President Kenneth Kaunda, Zambia embodied the hopes of the “third wave” optimism. By 2020, however, the country was counted by V-Dem as one of the world's fastest de-democratising states.²⁰ Its democratic status has long been contested, but Zambia has seen a steep decline in the quality of elections and an increase in polarisation since the 2015 presidential by-election and the 2016 general election.²¹ Placed at number 64 of 129 countries on the Bertelsmann Transformation Index in 2018, Zambia had dropped to number 101 of 137 countries assessed for the 2022 report. This is significant, given that the period from 2018 to 2022 saw generally declining trends of governance in most countries across the index. This made Zambia one of the fastest-declining countries assessed. From the 2016 closure of independent media, the 2017 treason charges against the main opposition leader, to the increasing arrests of opposition leaders and citizens for ‘insulting the president’, Zambia's democratic status was widely seen as compromised. As the 2021 polls approached, the incumbent was able to use patronage, COVID restrictions, control of public media, stifling of dissent and state institutions such as the police, army and courts to shift the playing field in his own favour. Incumbency advantage – long theorised to be the primary reason for a lack of electoral alternation in Africa – was expected to win out, with the Economist newspaper predicting a win for the incumbent president.

Against widespread predictions, and in a deeply unfair electoral environment, Zambia's main opposition candidate Hakainde Hichilema of the United Party for National Development (UPND) defeated President Lungu (PF) and fourteen other candidates to win the country's 12 August 2021 general election. Hichilema obtained 2.8 million votes, or 59% of the valid votes cast – 1 million votes more than Lungu. With a winning margin of 21%, this was the largest electoral victory in 25 years. The UPND also won a majority in parliament, obtaining 82 of 156 seats. Lungu's defeated PF secured 60 seats while thirteen candidates won election as independent MPs. One seat went to a smaller opposition party.

Hichilema's resounding victory was the third peaceful transfer of power since Zambia's transition to multiparty democracy in 1991 and was particularly surprising because it came amidst deepening autocratisation and the efforts of an incumbent who did everything possible to secure re-election including manipulating state resources, controlling the media, judiciary, and security forces, and enacting various anti-media laws to intimidate the opposition and suppress voting. For the first time in Zambia's democratic history, the military were deployed to the streets ten days before the election to cow the electorate. The Zambian case runs counter to much of what we think we know about the behaviour of voters, and how incumbents retain power. It is therefore a central case study in the limits of incumbency advantage.

Incumbency and control over media

Incumbents across Africa have substantial control over the media, which allows them to present the best possible face of the administration to the public. Most of Africa's key and most trusted media outlets remain state-owned and this has significant implications for the difference between the opposition and incumbent's access to positive media coverage. The influence exerted over the media is often deployed through ownership of key outlets with strong editorial lines, regulation and legislation governing media operations. Other means include the deployment of government advertising budgets, maintaining close ties with “friendly” journalists, and abuse in the awarding of broadcasting licenses. In more illiberal contexts, governments may resort to threats and intimidation, arrests of critical voices, shutting down independent media and the criminalisation of criticism in both traditional and social media. Such illiberal management of the flow of information has been critical to the maintenance of control over the state in places such as Djibouti, Egypt, Ethiopia, and Zimbabwe.

While Zambia has at least four independent newspapers, over 120 privately owned radio stations and 42 privately owned Television stations, the media's operating environment has proven to be tightly controlled. State media, which includes the Zambian National Broadcasting Corporation (ZNBC), and two state-run newspaper outlets – the Zambia Daily Mail and the Times of Zambia – are still some of the most widely circulated media. Most of the 120 privately operated radio stations were owned by local elites sympathetic to the former ruling party, while critical voices struggled to access broadcasting licences. Between 2016 and 2021, the media space in Zambia shrank significantly, with notable low points being the 2016 closure of the influential The Post newspaper²² and the 2020 withdrawal of independent Prime Television's broadcast license.²³ There were significant and substantial threats against the media during these years, which included the arrest of a newspaper editor who spent nearly a year in prison for contempt of court, and the withdrawal of licenses of several radio and TV stations in 2019.²⁴ Many independent media practitioners admitted to self-censorship during the Lungu administration, to avoid litigation, attacks by ruling party members, or the potential of being shut down. This led some scholars to argue that press freedom has been ‘quickly regressing over the last decade more than at any other time in the country's 57 years of independence’.²⁵

With increased access to the internet and the rapid growth of social media outlets, African opposition parties have found opportunities to circumvent the restrictions associated with press freedom. Nonetheless, autocratic regimes constantly seek to manipulate social media access to disadvantage the opposition. Several mechanisms are used to achieve such ends, including the spread of “fake news” to discredit opponents of the regime.²⁶ While spreading fake news is a weapon that is also at the disposal of opposition parties, incumbents have many other avenues to control or censor the internet and social media platforms. For example, many incumbents have the capacity to hack into opponents’ networks and create an army of fake social media accounts to dominate the digital space. Incumbents also use overt methods to restrict social media access including through the growing use of internet shutdowns. Such shutdowns have increasingly become the norm during elections as a means of censoring the flow of information. In 2020 alone, internet access was cut off for several days during elections in Burundi, Guinea, Tanzania, and Togo, which were all won by governing parties.²⁷

The period from January 2021 saw increasing attacks on independent radio stations by ruling-party affiliated activists and the passing of the repressive Cyber Security and Cyber Crimes Act No. 2 of 2021 which imposes unjustifiable limitations on free speech.²⁸ During the election campaign itself, the European Union (EU) election observer team noted that the media environment was highly monetised, with the government and ruling party using their substantial financial advantage to monopolise prime time slots on both public and private media.²⁹ As is so often the case during election campaigns, state media gave disproportionate coverage to the president during prime-time viewing. During the period in which media was monitored by the EU, the most-watched ZNBC TV 1 – over 7 million prime-time viewers every day – allocated 86% of its news coverage to the President, the PF and the government. By contrast, the opposition UPND received just 6% of news coverage and was consistently featured negatively.³⁰

The first article in this special issue by Lynch and Gadjanova focuses on the ways in which the incumbent advantage in terms of media control was subverted by the role of social media during the election campaign period.³¹ The authors demonstrate that traditional media outlets controlled by the state were heavily biased in favour of the governing PF throughout the campaign period. This had important implications on the electoral landscape given the closure of The Post newspaper and Prime Television. Marja Hinfelaar, Lise Rakner, Sishuwa Sishuwa, and Nicolas van de Walle discuss the closure of these outlets in more detail, in their contribution to this issue that focuses on the extent to which the state under President Lungu deployed state institutions to weaken democracy.³² But it is precisely the incomplete capture of state institutions that allowed for turnover, as the army, electoral commission and judiciary – despite expectations to the contrary – acted to respect the will of Zambian voters. This suggests that despite efforts to bend institutions to the will of the executive, prior experience of party turnover might make them more resilient to pressure from incumbents.

In their article, Lynch and Gadjanova show how social media platforms such as Facebook and WhatsApp became effective tools for the opposition UPND to communicate its message to voters. Given an already polarised environment, the opposition's campaign efforts were further undermined by COVID-19 restrictions that effectively barred in-person campaign events such as rallies. According to the EU observer mission's final report, the ‘lack of pluralistic debate in traditional media compounded by the COVID-19 pandemic elevated Facebook to the prime discussion forum’.³³ Although PF and UPND both benefitted from the use of social media, the PF still had access to traditional media outlets and used its incumbency to defy COVID-19 restrictions and hold in-person campaign events, without any legal consequences. The steady growth of Zambian citizens accessing the internet, primarily through their mobile phones, allowed the UPND to reach a much broader audience than they would have been able to in the absence of social media platforms. The fact that opposition supporters who were afraid of being victimised for their political views by the state or ruling party cadres were able to engage with UPND activists and fellow citizens using much safer social media spaces, demonstrates how advancements in technology and increased access to the internet limited the extent to which the state could muzzle opposition campaigns.

Control over the economy and resource distribution

In almost all political systems worldwide, incumbents are at a significant advantage over challengers and have superior access to resources that are unavailable to opposition leaders. By virtue of being incumbents, they have access to staff and financial support to travel around the country and meet with constituents. They also have greater capacities for fundraising due to reliance on other state actors, the private sector and multinational finance which sees investment in ruling parties as a safer bet.³⁴ Invariably, ruling parties outspend the opposition at enormous rates, and this is enabled by either grey zone financing from backers, or from the use of state resources for partisan ends. In some instances, incumbent parties have openly sabotaged businesses that financed the opposition by either blacklisting them or denying them access to state contracts.³⁵ More commonly perhaps, control over budgets and state resources provides opportunities for patronage and pork barrel politics which occurs with relative frequency even in the most established Western democracies.³⁶

Ruling parties and the governments they control are also able to award contracts and allocate state resources in ways that may be politically advantageous at election time. It is common for ruling party officials to enrich themselves, their business partners and clients using discretionary access to public funds, state contracts, and private sector finance including through legal avenues³⁷ During Lungu's presidency, several prominent government officials were the subject of corruption allegations, including the ministers of Health and of Infrastructure Development. It is also believed that money that the government borrowed to build roads and bridges was mostly lost to corruption which benefitted ruling party elites.³⁸

Incumbents are also more likely to benefit from an electoral bounce during periods of higher economic growth or increased commodity prices which leads to improved capacity of government to deliver social services. In periods of economic decline, they may resort to discretionary spending by introducing programmes that provide benefits to ruling party supporters or voting blocs that are deemed crucial for their re-election efforts. Throughout Lungu's presidency that began in January 2015, Zambia's economy was in significant decline. The economy was effectively in crisis partly due to the collapse of global copper prices but also because of poor economic management.³⁹ During Lungu's tenure, the Zambian economy was characterised by severe electricity shortages, exchange rate volatility, rising inflation, high levels of international and domestic indebtedness, and slow (and even negative) economic growth.⁴⁰ These factors co-existed with high levels of unemployment.

A poorly performing economy has the potential to hurt incumbents, and this provides incentives for governing parties to resort to vote buying and the implementation of programmes that distribute small-scale ‘club’ goods to targeted communities or voting blocs.⁴¹ Ahead of the 2021 election, the PF increased the nominal budget for the Farm Input Subsidy Programme (FISP). The programme, which provides subsidised farming inputs to more than a million farmers, was increased fourfold between 2020 and 2021. Historically, programmes such as FISP have been used to shore up support among farmers in Zambia, who constitute the largest employment category in rural areas.⁴² The PF government also introduced a 470-million-kwacha (about US$ 28 million) multi-sectoral youth empowerment fund in August 2020, that targeted 150,000 beneficiaries, ostensibly to cushion the impact of COVID-19 on small and medium enterprises.⁴³ The empowerment funds were mostly targeted at urban informal economy workers who form the largest employment group in urban areas and also constituted part of the PF's core support base. Widespread but anecdotal evidence suggests that these resources were made available only to known ruling party supporters ahead of the polls.⁴⁴

The extent to which a bad economy or the measures that incumbents use to target crucial voting blocs vary across elections and countries. Afrobarometer survey data of 16 African countries which include Ghana, Kenya, Malawi, South Africa and Zambia, show that popular evaluations of government economic performance are more likely to inform voting behaviour among Africans than hitherto understood.⁴⁵ Yet, African voters may also make strategic calculations in search of patronage or for fear of being sanctioned, particularly in cases where incumbents are at the helm of a dominant party system with a high likelihood of re-election. This helps to explain why African incumbents are routinely re-elected even when economic conditions are bad. However, the study of African elections in the last decade reveals that clientelist inducements and economic evaluations, are not mutually exclusive. Evidence from Ghana⁴⁶ and Zambia⁴⁷ shows that voters can accept clientelist goods from politicians, but vote based on economic factors. This accords with Cheeseman, Lynch & Willis’ findings in Ghana where they propose the rise of evaluative clientelism in which voters’ rising expectations of the state interact with continued clientelistic practices, but lead to a greater propensity to prioritise collective goods (such as healthcare and education provision) over private goods such as money for votes.⁴⁸

While social media was crucial for spreading opposition campaign messaging, the lived experiences of citizens also contributed to their voting intentions. A deepening economic crisis had dramatically impacted the prices of basic foodstuffs, forcing as many as 40% of Zambians to reduce the number of meals that they ate per day.⁴⁹ Afrobarometer survey data showed that declared support for the PF, both in terms of partisan affiliation and intended vote choice, had eroded significantly between April 2017 and December 2020.⁵⁰ In historically high refusal rates, over half of Afrobarometer survey respondents refused to declare which party candidate they would support in the 2021 election. While this did not show proof of support for UPND, it suggested that the increase in undeclared voting intensions was at the expense of the PF. Much of the erosion in support was attributed to increased dissatisfaction in the PF's economic performance.

In this issue, survey data is used to understand why the PF lost support in some of its own provincial strongholds. Jeremy Seekings uses a combination of Afrobarometer survey data and the Zambia Elections Panel Survey⁵¹ to understand factors that informed voting attitudes in the Eastern Province of Zambia.⁵² The Eastern Province, which is predominantly rural, was previously a stronghold for the former governing MMD but was won by PF in the 2015 presidential by-election. While Sata had lost Eastern Province to the MMD's Banda in 2011, Lungu won the province in 2015. This was partly because Lungu's family origins were in the province and due to an alliance with Banda who was also from Eastern Province. Yet, Seekings shows that while Lungu won the most votes in the province in 2021, his support dropped to 58% from 79% in 2016. At the same time, Hichilema's support increased from 16% in 2016 to 40% in 2021.

Seekings’ contribution to this issue highlights at least two factors that help to explain the decline in support for the PF in the province. Firstly, the PF in Eastern Province was an opportunistic coalition of local barons that lacked organisational strength. This created factionalism in the province and resulted in parallel slates of parliamentary candidates, with some running as PF candidates and others as independents. Popular candidates who ran and won as independents, were able to shift support away from the PF at parliamentary level. Secondly, the regression analysis provided by Seekings also shows that voters in Eastern Province were much more dissatisfied with the PF's management of the economy in 2020 and 2021, than they were in earlier years. Thus, they were also more willing to vote for the opposition. This analysis contributes to the issue by demonstrating that incumbent advantages and co-ethnic voting are not sufficient to sustain support for a ruling party when it lacks organisational strength and fails to manage an economy to the satisfaction of even its core supporters.

Similarly, Hangala Siachiwena's contribution to this issue shows that the PF lost its urban support to the UPND in 2021.⁵³ The PF became the largest opposition party in 2006 after winning control of urban constituencies and councils in the country's two predominantly urban provinces.⁵⁴ Under Sata, the PF had used populist mobilisation strategies to win the support of networks of poor and informal economy workers, and unionised miners.⁵⁵ In the 2015 and 2016 elections, Lungu won majorities in both Lusaka and Copperbelt but lost the two provinces to Hichilema in the 2021 election. What explains the urban loss? Wahman and Boone found that across Africa, urban voters are more likely to swing their support away from ruling parties than their counterparts in rural areas because they are more likely to sanction leaders if their economic conditions do not improve.⁵⁶ Like Seekings, Siachiwena also uses Afrobarometer survey data to show that urbanites were dissatisfied with the PF's economic performance. Urban residents were particularly concerned about jobs and the management of the economy.

The analysis of voters in Eastern Province by Seekings, and in urban areas by Siachiwena, shows that incumbents who prevail over poorly performing economies are at great risk of being sanctioned by citizens. This also demonstrates that in the absence of sound economic management and strong party organisational strength, ruling parties are not guaranteed to sustain support in their strongholds even when ethnic considerations and repression favour the incumbent. Previous turnovers in 1991 and 2011 were also attributed to economic conditions. Both turnovers also occurred notwithstanding incumbent advantages, with the 1991 elections being held by an entrenched authoritarian regime. Most Zambians voting in 2021 were aware of this history and would likely have had confidence that another turnover could occur amidst similar economic decline and autocratic conditions.

Control over state institutions

Across much of Africa, formal political institutions account for much of the low rate of electoral alternation.⁵⁷ Institutions such as courts, electoral commissions, parliaments, and security agencies, are often subverted by incumbents.⁵⁸ While multiparty elections may be held regularly, electoral rules and term limits can be changed to suit the interests of the incumbent.⁵⁹ Security agencies such as the police are often deployed to harass, intimidate, and even arrest opposition leaders and their supporters.⁶⁰ Moreover, opposition parties are frequently under surveillance and occasionally subjected to violence.⁶¹ In extreme cases, political institutions are used to disqualify opposition candidates from contesting elections, to restrict opposition campaign programmes, and manipulate the outcome of the vote through mechanisms such as ballot-box stuffing and inflating vote tallies in favour of the incumbent.⁶²

The fourth and fifth articles in this volume consider how the state used Zambian institutions to stifle democracy in the years leading to the 2021 elections while also demonstrating limits of autocratisation. Nicole Beardsworth and Kalonde Mutuna show that one way through which Lungu undermined democracy was his neglect for ‘tribal balancing’ and regional representation.⁶³ Since Zambia's independence in 1964, successive presidents have ensured that their appointments to Cabinet, heads of security wings, and senior public service positions, reflected the ethnic (or tribal) and regional diversity of the country. Yet, garnering little support in the south, west and north-west of the country in 2016, Lungu and his PF effectively adopted an exclusionary election strategy. The strategy involved ostracising members of ethnic groups or regions that were perceived to be loyal to the opposition including through retirements in the “national interest”. This had the effect of creating a state apparatus that consisted of individuals who came mostly from the Bemba and Nyanja language groups, which are the most spoken languages in the regions PF won in 2016. According to Beardsworth and Mutuna, Lungu systematically excluded large sections of society from government based on ethno-linguistic identities to consolidate his own power and secure re-election.

Lungu's ethnic strategy is one avenue through which institutions can be weakened. Cheeseman observed that across Africa, incumbents who wish to maintain their grip on power weaken and undermine state institutions by establishing informal patronage networks.⁶⁴ The appointment of party loyalists or the presidents’ co-ethnics to lead political institutions contributes to the establishment of such networks and weakens the checks and balances that exist on the executive, thereby constraining the prospects for democracy. However, Beardsworth and Mutuna show that Lungu's strategy created perceptions of tribal marginalisation by the affected regions which weakened rather than strengthened Lungu's hold on power. Lungu's strategy also ran counter to the idea of multi-ethnic coalition building which has political implications in electorally competitive and multi-ethnic societies such as Kenya and Zambia.⁶⁵

The analysis by Beardsworth and Mutuna is complemented by Hinfelaar et al., who demonstrate other measures that Lungu used to systematically erode governance in Zambia. They show that state institutions routinely undermined aspects of democracy in Zambia, especially those that stood in the way of Lungu's re-election efforts. For example, the Zambia Revenue Authority (ZRA) was used to shut down The Post in 2016 because of tax arrears, while state-owned newspapers owed the ZRA even more money than The Post. The newspaper had been very critical of Lungu and was a trusted source of information especially for pro-democracy supporters. Law enforcement agencies were also used routinely to harass, intimidate and arrest prominent civil society leaders and even Hichilema who spent four months in a maximum-security prison in 2017. Hinfelaar et al. demonstrate that the PF used (or abused) various pieces of legislation to achieve its ends. However, Lungu's ‘autocratisation by lawfare’ had its limits because of a resurgent civil society coalition that mobilised against regressive legislation⁶⁶ and because the then-opposition UPND had enough numbers in parliament to block PF-supported constitutional reforms. Constitutional reforms aimed at consolidating power and weakening democracy are a common feature in autocratising countries.⁶⁷

The sixth paper in this special issue focuses on institutions that are crucial for conducting free and fair elections and providing legitimacy to the process. Autocratic countries generally lack independent electoral commissions because it allows incumbents to manipulate electoral processes.⁶⁸ Even when electoral bodies have some degree of independence, autocratic leaders still seek to exert control of them through measures such as appointing ruling party loyalists to head the commissions, manipulating voter registers, and influencing the procurement of election technology that can be manipulated by the state.⁶⁹ Partly due to weaknesses with electoral commissions, the role of election monitors and international observers has grown in importance. In Kenya's problematic 2007 elections for example, the EU observer mission provided evidence showing that the electoral commission had inflated the vote tally for the incumbent Mwai Kibaki in some constituencies.⁷⁰ The role of observers does not guarantee that an election outcome will reflect the democratic will of the people (as Kenya's 2007 election would suggest), but it does contribute to a ‘learning process’, whereby lessons learnt from previous elections are used to improve the management and observation of subsequent contests.⁷¹

Despite a pre-election environment that favoured the ruling party at the expense of the opposition, election observation played an important role in protecting the democratic will of the Zambian people. O’Brien Kaaba, Marja Hinfelaar, and Koffi Sawyer contrast the roles of international election observers in the 2016 and 2021 elections.⁷² Their article shows that while international observers deployed a much smaller contingent in 2021 than they did in 2016 due to COVID-19 related restrictions, they had much better coordination amongst themselves and with domestic observers and civil society this time around. This contributed to more robust vote monitoring and Parallel Voter Tabulation (PVT). Kaaba et al. show that international observers learnt lessons from the 2016 election which had a controversial outcome. Moreover, two former African heads of state who headed observer missions, Ernest Bai Koroma from Sierra Leone, and Jakaya Kikwete from Tanzania, played important mediation roles together with Zambia's Rupiah Banda. Their mediation was particularly crucial during the vote counting process when it appeared that Hichilema would win the vote. The three ex-presidents brought Lungu and Hichilema together for closed-door talks, before final vote tallies were announced by the Electoral Commission of Zambia (ECZ). The talks are widely believed to have influenced Lungu's decision to concede defeat, but this could not have happened if the ECZ had bowed to pressure from the incumbent to shift tallies or invalidate the results in three opposition-supporting provinces. This unexpected institutional resilience may be in part a result of previous turnovers which have shown that state institutions and bureaucrats can withstand pressure from unpopular incumbents.

When juxtaposed with other articles in this issue, the contribution made by Kaaba et al. shows that the effective use of social media campaigns and the overwhelming rejection of the PF at the polls, may not have sufficed to end PF's rule in the absence of pre-and post-election observation and mediation. The article also demonstrates that the ‘learning process’ from previous elections contributes to strengthening the capacity of election institutions such as electoral commissions and election observers to conduct credible elections. The legitimacy that election observers give to an electoral process also has implications for the extent to which an incumbent can subvert the democratic will of the electorate, especially in a ‘weak’ autocratic state with a history of prior electoral turnovers.

The final article by Sishuwa demonstrates the importance of institutions such as the military and the electoral commission.⁷³ The strength of these institutions has implications for the willingness of incumbents to hand over power after being defeated by an opposition challenger. While both institutions appeared generally compromised heading into the election, they ultimately fell into line for different reasons. Apart from having a long history of being apolitical, the army did not have great reason to support the incumbent president because they generally considered him responsible for their institutional and material deprivation. As Sishuwa's article shows in greater detail, the decisive vote by the rank and file of the military in support of the opposition candidate placed significant constraints on the capacity of their leadership to support the incumbent's attempts to alter the outcome and prolong his stay in power. Unlike in neighbouring Zimbabwe, the military in Zambia is not insulated by patronage from the pressures affecting the rest of society⁷⁴ but is embedded within society and soldiers were thus more likely to vote against the incumbent. Like international observers, the army was an important factor in the short but crucial period between voting day and the declaration of the winner. As for the ECZ, the article shows that the capacity of officials to resist an incumbent's manoeuvres to alter the result is influenced not so much by the leadership of such bodies but by the desire of the rank and file to do the right thing both to secure their careers and protect institutional integrity. Sishuwa's article shows that after 30 years of multiparty democracy, political institutions in Zambia were not fully captured. Prior turnovers, weak authoritarianism, and poor economic conditions combined to ensure some degree of institutional independence that limited the advantages of the incumbent.

Conclusion

Of the 22 transfers of power that occurred on the continent between 2010 and 2021, more than half (12) unseated incumbent presidents. This shows that incumbency advantage is no longer a guarantee of electoral success today as it was before the early 2000s. The Zambia 2021 case study allows us to explore some mechanisms through which incumbency advantages can be countered. The presence of a strong opposition that represents the competitive nature of politics, the state of the economy, the increasing importance of social and private media, and the resilience of civil society – were at play in the Zambian election. Although democratic transitions are neither universal, easily transferable nor replicable in other areas, the presence of the factors discussed here and in other essays in this special issue increased the likelihood of incumbency failure and a peaceful transfer of power.

While the lessons learnt from Zambia may not travel easily to other countries on the continent, we theorise that those countries that experienced at least one turnover between 1990 and 2009 were much more likely to experience alternation in the 2010s than those where turnovers had never occurred. Prior turnovers did not guarantee democratisation, but they contributed to weakening the extent of authoritarianism. We suggest that a popular memory of prior transitions – a “democratic muscle memory” – may make future electoral turnovers more likely as institutions and bureaucrats are more resilient against autocratic incumbents. Turnovers, we argue, disrupt (often only temporarily) autocratisation and give citizens confidence that unpopular leaders can be removed through the ballot box. The co-existence of a democratic muscle-memory with weak authoritarianism limits the benefits of incumbency. This introduction is a call for more research on why we see repeated turnovers of power in countries that previously experienced turnovers, through micro-level evidence of voter behaviour, institutional histories and interviews with bureaucrats, judges and electoral officials amongst others.

As for Zambia, two years since the election of Hichilema, the implications of the recent turnover for democratic consolidation are mixed. There has been a considerable decline in reports of political violence, while some private media houses that were previously closed had their licenses reinstated. Importantly, a law that allowed police to arrest citizens for insulting the president was abolished. But the UPND has not yet enacted crucial legislation including the Access to Information Bill, or provided a clear roadmap for the constitutional and institutional reforms that they promised while in opposition. It is evident that the 2021 transfer of power has not guaranteed the institutional redress that would ensure the strengthening of institutions and consolidation of democracy. On the economic front, the UPND has expanded social spending, including introducing free primary education, reinstating bursaries for students at public universities, and increasing resources allocated to constituency development that benefits local (and mostly rural) communities. In June 2023, the Zambian government concluded a deal to restructure $6.3 billion in debt owed to external creditors. While these reforms are positive for the economy, some Zambian citizens are dissatisfied with the lack of jobs, the slow pace of reforms to the mining sector, and the high cost of fuel and maize meal. Critics of the new government have frequently taken to both private and social media outlets where they have threatened to retire the new government at the next election if these pressing economic and social conditions are not addressed. Most Zambians today have experienced at least one electoral turnover and this “democratic muscle memory” might condition both voters and state institution officeholders to believe that turnovers can occur in future electoral contests, especially if the country is faced with poor economic conditions and autocratic rule.

Acknowledgements

This issue draws on the expertise of scholars affiliated with the Zambia Elections Research Network (ZERN), which was established through a GCRF grant from the Institute of Advanced Studies (IAS) at the University of Warwick, in collaboration with Warwick, the Southern African Institute for Policy and Research (SAIPAR) in Zambia, and the Universities of Witwatersrand and Cape Town in South Africa. ZERN benefitted greatly from support from the British Institute in Eastern Africa (BIEA). The network comprises Zambian-based and international scholars focusing on various aspects of elections and democracy in Zambia.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Hellemeier et al., “State of the World 2020.”

2 Boese, Lindberg, and Lührmann, “Waves of Autocratization and Democratization.”

3 Repucci, Sarah, and Amy Slipowitz. “Democracy under Siege.” Freedom House, 2021. https://freedomhouse.org/report/freedom-world/2021/democracy-under-siege.

4 Cheeseman, “African Elections as Vehicles for Change.”

5 Bleck and van de Walle, Electoral Politics in Africa since 1990, 56.

6 Ibid.

7 Cheeseman and Klaas, How to Rig an Election; Schedler, “The Menu of Manipulation”; Arriola, Rakner, and van de Walle, Democratic Backsliding in Africa?

8 Bleck and van de Walle, Electoral Politics in Africa since 1990; and Cheeseman, “African Elections as Vehicles for Change.”

9 Van Ham and Lindberg, “From Sticks to Carrots.”

10 Cheeseman, “African Elections as Vehicles for Change.”

11 Arriola, Rakner, and van de Walle, Democratic Backsliding in Africa?

12 Cape Verde also experienced an electoral turnover, but it was in an open-seat election.

13 Bleck and van de Walle, Electoral Politics in Africa since 1990.

14 Arriola, Rakner, and van de Walle, Democratic Backsliding in Africa?

15 Fraser and Larmer, Zambia, Mining, and Neoliberalism.

16 Resnick, Urban Poverty and Party Populism.

17 Lynch and Gadjanova, “Overcoming Incumbency Advantage.”

18 Sishuwa, “Surviving on Borrowed Power.”

19 Sishuwa Sishuwa and Nic Cheeseman, “Three Lessons for Africa from Zambia's landslide opposition victory.” African Arguments, 22 August 2021. https://africanarguments.org/2021/08/three-lessons-for-africa-from-zambia-landslide-opposition-victory/ (accessed 3 July 2023).

20 Lührmann et al., “Autocratization Surges – Resistance Grows.”

21 Goldring and Wahman, “Democracy in Reverse.”

22 Matthew Hill, “U.S. Urges Zambia to Let Post Newspaper Reopen as Vote Nears.” Bloomberg, 23 June 2016. http://www.bloomberg.com/europe. (accessed 30 June 2022).

23 Zambia Reports, “IBA Cancels Prime TV Licence.” 9 April 2020. https://zambiareports.com/2020/04/09/iba-cancels-prime-tv-licence/ (accessed 27 May 2022).

24 Reporters Sans Frontieres, “Zambia 2022 Report.” 2022. https://rsf.org/en/country/zambia (accessed 13 January 2022).

25 Ndawana, Knowles, and Vaughan, “The Historicity of Media Regulation in Zambia,” 61.

26 Cheeseman and Klaas, How to Rig an Election.

27 Julie Owono. “Internet Shutdowns in Africa Threaten Democracy,” The Mail & Guardian, 16 June 2021. https://mg.co.za/africa/2021-06-16-internet-shutdowns-in-africa-threaten-democracy/ (accessed 28 June 2022).

28 Kamufisa Manchishi. “State of the Media Report, Zambia Q1/2021.” Media Institute of Southern Africa, 30 April 2021. https://zambia.fes.de/e/state-of-the-media-in-zambia-report-q1-2021 (accessed 4 June 2022).

29 European Union Election Observation Mission to Zambia, “EU EOM Zambia 2021 Final Report.”

30 Ibid., 31.

31 Lynch and Gadjanova, “Overcoming Incumbency Advantage.”

32 Hinfelaar et al., “Legal autocratisation.”

33 European Union Election Observation Mission to Zambia, “EU EOM Zambia 2021 Final Report,” 5.

34 Bleck and van de Walle, Electoral Politics in Africa since 1990, 80–2.

35 Levitsky and Way, Competitive Authoritarianism.

36 Lancaster and Patterson, “Comparative Pork Barrel Politics.”

37 Hill, Democratisation in the Maghreb, 18.

38 Sishuwa Sishuwa. “In a Fair Election, Lungu Can't Win. In an Unfair One, He Can't Lose.” 8 August 2021, African Arguments. https://africanarguments.org/2021/08/zambia-in-a-fair-election-lungu-cant-win-in-an-unfair-one-he-cant-lose/ (accessed 3 July 2023).

39 Hinfelaar and Sichone, “The Challenge of Sustaining.”

40 Zambia Institute for Policy Analysis and Research. “Huge Debt, Rising Inflation, Low Productivity, and Youth Unemployment Represent Some of the Immediate and Pressing Challenges of the New Government.” 30 August 2021. https://www.zipar.org.zm/huge-debt-rising-inflation-low-productivity-and-youth-unemployment-represent-some-of-the-immediate-and-pressing-challenges-of-the-new-government/ (accessed 20 June 2022).

41 Lindberg, “Have the Cake and Eat It.”

42 Kim, “Party Strategy in Multidimensional Competition in Africa.”

43 Lusaka Times. “President Lungu's Full Speech at the Launch of K470 million Youth Fund.” 16 August 2020, https://www.lusakatimes.com/2020/08/16/president-lungus-full-speech-at-the-launch-of-k470-million-youth-fund/ (accessed 20 January 2022).

44 Interviews in Lusaka, July 2021.

45 Bratton, Bhavnani, and Chen, “Voting Intentions in Africa,” 28.

46 Lindberg and Morrison, “Are African Voters Really Ethnic or Clientelistic?”

47 Resnick, “Urban Governance and Service Delivery in African Cities.”

48 Cheeseman, Lynch, and Willis, “Ghana,” 95.

49 Nicole Beardsworth, O’Brien Kaaba, and Nic Cheeseman. “Ahead of Zambia's Elections, Fears Rise of a Political Crisis.” World Politics Review (blog), 10 August 2021 (accessed 10 August 2021) https://www.worldpoliticsreview.com/articles/29872/ahead-of-zambia-elections-fears-rise-of-a-political-crisis.

50 Seekings and Siachiwena, “Voting Preferences among Zambian Voters.”

51 The Zambian Election Panel Survey (ZEPS). https://gld.gu.se/en/projects/the-zambian-election-panel-survey-zeps/ (accessed 30 June 2021).

52 Seekings, “Incumbent Disadvantage in a Swing Province.”

53 Siachiwena, “The Urban Vote.”

54 Larmer and Fraser, “Of Cabbages and King Cobra.”

55 Resnick, Urban Poverty and Party Populism; Sishuwa, ‘“I Am Zambia's Redeemer.”

56 Wahman and Boone, “Captured Countryside?”

57 Bleck and van de Walle, Electoral Politics in Africa since 1990, 64.

58 Bogaards and Elischer, Democratization and Competitive Authoritarianism in Africa, 7.

59 Bleck and van de Walle, Electoral Politics in Africa since 1990, 64.

60 Ibid., 63.

61 Levitsky and Way, “Beyond Patronage,” 12.

62 Cheeseman and Klaas, How to Rig an Election.

63 Beardsworth and Mutuna, “Tribal Balancing.”

64 Cheeseman, Democracy in Africa.

65 Arriola, Multi-Ethnic Coalitions in Africa.

66 Rakner, “Don't Touch My Constitution.”

67 Bleck and van de Walle, Electoral Politics in Africa since 1990; Arriola, Rakner, and van de Walle, Democratic Backsliding in Africa?

68 Bleck and van de Walle, Electoral Politics in Africa since 1990, 89.

69 Cheeseman and Klaas, How to Rig an Election.

70 Ibid.

71 Bleck and van de Walle, Electoral Politics in Africa since 1990, 65.

72 Kaaba et al., “A Comparison of the Role.”

73 Sishuwa, “The Outcome of a Historical Process.”

74 MacDonald Dzirutwe. “Zimbabwe Plans Subsidised Shops for Security Forces.” 26 February 2020. https://www.nasdaq.com/articles/zimbabwe-plans-subsidised-shops-for-security-forces-2020-02-26 (accessed 28 June 2023); Noyes, “A New Zimbabwe?”