Tanzania under Magufuli: the personalization of a party-based regime

ABSTRACT

Contemporary research has shown that authoritarian regimes are not static. At the same time, gradual changes are often difficult to detect and the literature has not yet developed convincing tools to identify autocracy-to-autocracy transitions outside the visible ruptures of coups, power transfers, and opposition victories. Building on fieldwork in Tanzania, we show that patterns of rule shifted significantly under Magufuli. Once the model case of a party-based system in Africa, we argue that Tanzania should be reclassified as a party-personalist regime for the time of his presidency. The basis for his success lies in the increasing factional tensions within the CCM which gave him the power to act as the arbiter and to manipulate party institutions and nominations to his favour. Beyond providing a thick description of a single case, we address the theoretical and empirical challenges of correctly classifying authoritarian regimes.

KEYWORDS:

• Africa
• Tanzania
• electoral authoritarianism
• authoritarian institutions
• party-based regimes
• personalism

1. Introduction

John Magufuli, President of Tanzania from 2015 to 2021, died on 17 March 2021. His time of rule was characterized by an extent of personalization previously unprecedented in Tanzanian history. This is in itself puzzling: how and why does a process of personalization occur in an otherwise strongly institutionalized party-based regime with a well-entrenched ruling party like the Chama Cha Mapinduzi (CCM)?

Our article makes two important contributions. The first one is descriptive. We use Geddes’ indicators of personalism to provide case-based evidence for. In this article, we provide evidence for the qualitative change of leadership styles and patterns of rule under Magufuli. We show that authoritarian regimes are not static, and that personalization can be a response to adaptation pressures. In short, we argue that this occurs when elite cohesion is challenged by factionalism. While the ruling party CCM has clearly dominated politics since independence, its model of party institutionalization came under pressure due to a combination of internal and external factors. First, the advent of democratization meant more competition from opposition parties and, for the first time, the danger of elite defections.¹ Secondly, economic liberalization transformed patron-client relations, which resulted in less centralized structures harder to control by the party leadership.² Although party cohesion as such remained largely intact, elite rivalries and factional competition within the CCM became so intense that they created space for an arbiter or moderator who not only re-centralized party discipline and patronage, but who also accumulated personal power in this process. We therefore argue that Magufuli’s time in office marks a qualitative shift to personalism and should be seen as an autocracy-to-autocracy transition.

This article is based on rich and original evidence from field interviews and adds important nuances to the conventional narrative of Tanzania as the role model of an African party-based regime. We argue that Tanzania should be reclassified as a party-personalist regime for the time of Magufuli’s presidency. Moreover, we make a causal argument by suggesting a general theoretical mechanism for autocracy-to-autocracy transitions. A process of personalization can occur in party-based regimes when the institutionalized structures of elite management come under pressure from external and internal rivalries. While high levels of institutionalization make parties less vulnerable to splits and elite defections, they also become less flexible in their response to factionalism and deadlock. Exactly in those situations the emergence of a strong leader can be functional for regime survival.

Our work has wider implications for the research on authoritarianism. We need to question track record classifications more carefully and continuously gather and re-assess data on authoritarian processes of decision-making. While Geddes’ typology acknowledges the possibility of shifts between the types, the literature has not yet developed convincing theoretical and empirical tools to identify autocracy-to-autocracy transitions outside the visible ruptures of coups, power transfers, and opposition victories. Paying more attention to the slow and sometimes hardly visible evolutionary patterns will not only help to improve our regime classifications but possibly also help to improve the predictive power of models that are based on these classifications. We limit our analysis to Magufuli's presidency to provide clear-cut and empirically based arguments. An examination of the developments under his successor Samia Suluhu Hassan, is beyond the scope of this article and will be subject to future work.

The article is organized as follows: Section 2 provides a literature overview and theoretical discussion. Section 3 explains data collection and empirical strategy. Section 4 gives a historical background to our arguments and analyses the process of personalization under Magufuli. Section 5 proposes a re-classification of the period under Magufuli as a party-personalist regime.

2. Literature review and theory

2.1. Autocracy to autocracy transitions – what do we know?

Geddes’ threefold regime typology had an enormous influence on the study of authoritarian regimes.³ One of the innovative aspects lies in the fact that Geddes does not measure the “democraticness” of a polity but uses decision-making power as the defining criterion. A regime is defined as a “set of formal and informal rules for choosing leaders and policies”.⁴ These rules determine “what interests are represented in the authoritarian leadership group and whether these interests can constrain the dictator.”⁵ Using this definition, Geddes classifies authoritarian regimes as military, single-party, or personalist regimes: In military regimes, a group of military officers decides who rules the country and influences policy decisions.⁶ In single-party regimes, one party exercises control over access to political office and state resources, and in personalist regimes the access to political office depends on personal discretion of an individual leader.⁷ These different regime types also have different likelihoods of surviving or democratizing.

So far, the literature often argues that the origin of a regime explains duration and survival. Rent access and organized opposition,⁸ social revolutions,⁹ or institutional factors¹⁰ have been found to explain the absence or presence of a strong single party and original patterns of power sharing. Our article brings in a dynamic element. Although the original constellation remains important, regimes change and adapt to new challenges and changed environments over time. An emerging literature is showing that not all cases of regime change or breakdown lead to democratization. In a number of cases, a transition finally results in another authoritarian regime.¹¹ These autocracy-to-autocracy transitions have only recently gained scholarly attention and remain poorly understood. Panta defines them as “moves between two non-democratic regimes marked by singular, characteristic events”.¹² Nearly all autocracy-to-autocracy transitions have been caused by events like coups, popular uprisings, or downfall due to rebellion or foreign invasions.¹³

While such events can be clear markers for the onset of a transition, other indicators are less straightforward. Leadership change for example can simply represent the transfer of power within a regime.¹⁴ Given the fact that research has put so much emphasis on authoritarian survival strategies there is however little reason to believe that regimes remain static outside big events. Political leaders tend to implement changes gradually in order to hide the fact that the rules of the game are being changed, thereby preventing other actors from interfering.¹⁵ The adaptation of decision-making procedures is a survival strategy with which rulers react to a shifting environment to prevent a crisis.¹⁶ The “coup proofing” literature is a case in point. This strand of research is seeking to understand authoritarian stability by exploring coalition-building, (ethnic) balancing, and elite appointments.¹⁷ In this perspective, events like coups, rebellions, and uprisings are only the tip of an iceberg. They mark the failure of gradual adaptation by exposing the ruler’s inability to accommodate elites and or quell popular unrest. Gradual adaptation is extremely difficult to study since its very success actually depends on its opaque nature. The coup proofing and ethnic balancing literature has exploited information on cabinet appointments to approximated shifts in authoritarian ruling coalitions. Although this has proven to be extremely insightful, cabinets are only one of many aspects of strategic change.

We argue that autocracy-to-autocracy transitions can also take place gradually, leading to a shift from one authoritarian regime type to another. While there is no clear theorization of these transitions – for example the onset, completion, causal process, or empirical indicators – there is evidence of their existence. Slater's study on Mahathir’s rule in Malaysia is instructive in this regard.¹⁸ From his case-based empirical observation, he identifies three tools of personalization – packing, rigging and circumventing – which can transform a single-party regime into something which resembles a more personalized authoritarian rule.¹⁹ He concludes that ambitious autocrats need well-established regime institutions in order to execute their instructions. Moreover, Slater's study underlines that autocratic regimes can become more personalized without being less resilient.²⁰ Other case-based evidence shows how different regime leaders such as Recep Tayyip Erdoğan in Turkey,²¹ Vladimir Putin in Russia,²² Xi Jinping in China²³ and Hun Sen in Cambodia²⁴ have implemented reforms and laws which allow for personalized power and effectively repress opponents.

Not all regime types are easily transferable into each other. For example, a gradual shift to military rule seems rather unlikely. Before we address problems of gradual classification, we provide theoretical arguments about the possible causal mechanism of shifts from party-based to personalist systems.

2.2. Pathways from party-based systems to personalist rule

Personalism is characterized by relatively unconstrained decision-making power of the dictator in the political system.²⁵ The regime leader exerts personal control over personnel, policy and distribution of resources and has the capacity to appoint, promote and discharge government members, officials and high-ranking military officers.²⁶ In order to manage elite coalitions, a personalist dictator relies on a divide-and-rule approach.²⁷ The regime leader needs support, but he has the ability to select members of his ruling coalition from competing factions.²⁸ The lack of binding limits and institutional controls on the dictator gives him an incentive to abuse his followers in ways they could not foresee before.²⁹

Party-based systems, by contrast, are marked by a higher degree of predictability and institutionalized decision-making. Party institutionalization is understood as “the process by which parties reproduce consistent patterns of mass mobilisation and internal organisation.”³⁰ Ruling parties in authoritarian countries are believed to enhance regime survival by organizing elite accommodation and regulating the access to spoils.³¹ They also offer regime elites predictable career patterns, reward their loyalty and mediate between competing factions.³² Although ruling parties are good power sharing instruments, they are not immune to factional tensions and intra-party conflicts that can pose a threat to regime survival.³³

While the benefits of authoritarian parties for regime survival are often emphasized, party-based systems also vary in their durability.³⁴ The most successful ones are those that not only provide patronage but also a degree of non-material cohesiveness, for example a common memory of an armed revolutionary struggle or liberation movement.³⁵ While the common memory helps to hold up elite cohesion, post-nationalist parties are not spared from internal power struggles.³⁶ In addition, the effects are not permanent and degrade over time.³⁷

Ruling parties must continuously invest in strategies to manage factionalism and avoid elite defections. Some regimes have been relatively successful in this,³⁸ others have not been able to prevent party erosion, especially when defecting elites join sides with street protest.³⁹

Obviously, then, incumbent dominant parties are not monolithic blocs. Rather they are coalitions of rival opportunists.⁴⁰ The number of posts and positions in a political party is always limited, and this leads to intense internal struggles. In dominant party systems, the cost of defection is high, elites will therefore rather seek to build a support base within to influence the intra-party distribution of spoils. This can lead to factional confrontations. A ruling party’s ability to mitigate the potential threats that arise from factional competition will therefore be an important determinant of its survival.⁴¹

The most obvious strategy to incorporate increasing demands would be an expansion of patronage. The ability to do so, however, depends on the particular way in which state and party are fused and on the margins of electoral victories. Over the last electoral cycles, the CCM for example found itself in a situation of increased competition from opposition parties, leading to a decrease in parliamentary seats as possible reservoirs of clientelistic redistribution. A loss of popularity of incumbent parties – even if it does not lead to an opposition victory – puts the intra-party elite accommodation system under stress. Already existing factional struggles intensify, eventually leading to deadlocks and dangerous disintegration tendencies. Under such circumstances, a strong leader can emerge and stabilize the party by strengthening legitimation and offering a fresh focus of identification.⁴² This prevents the acceleration of a crisis by enhancing decision-making and mediating between the factions. In the longer run, power consolidates by playing competing elites against each other and building a new power base based on personal loyalty to the leader. This can alter decision-making to such an extent that it represents a shift from a party-based to a personalist system in the sense of Geddes, where the leader enjoys relatively unconstrained power of appointments and other crucial political areas.

In short, we argue that an autocracy-to-autocracy transition can be an adaptation of authoritarian rule to a changing environment. While it might seem uncontroversial that such adaptation processes are taking place in authoritarian states, the implications for regime classifications are more complex. We discuss some of them in the next section.

2.3. Personalism and party institutionalization: conflicting or compatible?

Research on democracy and autocracy has intensely debated conceptual and measurement issues. In more recent times, there is a growing literature on the importance of authoritarian institutions for regime survival,⁴³ and on personalization as a feature of democratic backsliding.⁴⁴ Since our work touches upon both problems, this section attempts to clarify some theoretical issues.

For a long time, especially the Africanist literature has conflated personal rule, (neo)-patrimonialism, and authoritarianism.⁴⁵ Newer work is arguing that the process of personalization can occur in different regimes, including democracies, but that it is often an indicator for democratic backsliding.⁴⁶ Since Magufuli’s rule is associated with increasing authoritarian tendencies, Tanzania is a case in point. This should not obscure, however, two other facts: on the one hand, there have always been strong personalities emerging from the CCM, and secondly, CCM rule has always had authoritarian tendencies before Magufuli.⁴⁷ We therefore argue that personalism and authoritarianism are not the same process and should not be conflated, although they are often related and mutually reinforcing. In this article, we are only interested in measuring the level of personalization under Magufuli, not democratic backsliding. Geddes’ approach is a good solution here, because it measures personalism independently from the level of democracy.

The second issue is the role of authoritarian institutions. A large literature is arguing that institutions, especially parties and parliaments, enhance regime survival by regulating the distribution of spoils.⁴⁸ Especially parties play an important role as institutions of elite management. They make career patterns predictable, provide clear avenues for elite mobility, and a shared framework for conflict management and patronage distribution.⁴⁹ These procedures have worked extremely well in the CCM for the most part of its history. Even leadership succession has usually been smooth, pointing to the acceptance of party hierarchies and shared norms on candidate selection. But authoritarian institutions are not static, and at times they need to respond to incremental changes as in Tanzania, or to external shocks. To give just an example: the CCM changed the mode of candidate selection in party primaries in an attempt to manage factionalism, but the outcome was actually a deepening of factionalism.⁵⁰ In line with Pepinsky’s critique of the institutional turn in comparative authoritarianism, we argue that institutions do not always do what their creators want.⁵¹ While a strong party is a solid foundation for regime stability, it is no guarantee for survival. Given the high level of institutionalization, it might even be less flexible than a weakly institutionalized party, where ad hoc decisions create immediate room for manoeuvre. These are the conditions that create the pressure for change and the opportunity for personalist leadership. A tension between institutionalization and personalization is not new in the CCM. Section 4.1. will provide a historical analysis of this tension and highlight the new elements of Magufuli's term.

On the conceptual front, we are thus interested in the distribution of power in a political regime, and most importantly in the interplay between institutions and actors. While some authors have successfully analysed questions of this type within the framework of political settlements,⁵² this framework is too loose for our purpose. Our focus is more on the measurement of the shift towards personalization, and we therefore see Geddes’ framework as the most appropriate way of structuring our empirical material. In the next section, we explain our approach to data collection and analysis.

3. Data collection and empirical strategy

The article draws on fieldwork in Tanzania and remote interviews. We want to be as transparent as possible about the research process to make clear what the limitations of our article are. Qualitative interviewing in authoritarian regimes is challenging anyhow because it requires the researcher to be extremely sensitive to the country context, to build trust with respondents, collect their informed consent, and respect their “red lines” as well. Not all information is complete and not all questions can be asked. These challenges grow if field access is restricted due to travel bans and health concerns in relation to the Covid-19 crisis.

Originally, the research framework was more tailored towards exploring the influence of factionalism on authoritarian strategies in a party-based system. There has been little systematic work on this topic, but there was reason to believe that factions mattered in the CCM. In addition, it was known that the country had become increasingly authoritarian under Magufuli – hence, change in personalization and regime characteristics was already underway. As such, Tanzania seemed to be the obvious choice of a typical case. Since qualitative research is based on an interaction with the subject that is studied, the approach was rather explorative with a semi-standardized questionnaire that would ensure flexibility and encourage respondents to present their views, experiences, and stories about factional struggles. Fieldwork began in 2020 but was interrupted by the Covid-19 crisis. With the field being closed up, the original empirical strategy became unfeasible as there was no more access to interview partners from the different CCM factions. Yet, the first interviews and field notes revealed that there was a much deeper shift in authoritarian strategies than initially expected. It became evident that factional struggles gave rise to a personalization of power. The research strategy hence began to make a virtue of necessity, analysing more closely the personalist transformation that respondents had described. Later, online and phone interviews were conducted that followed up on this information.

The gathered information was analysed along Geddes’ checklist and structured along these items in the empirical part.⁵³ The eight were collapsed into six to fit the Tanzania case study.⁵⁴ We keep the content of the indicators (right column in Table 1), but rename them slightly to make them more precise in the country context (left column in Table 1).⁵⁵

Table 1. Indicators used to measure personalization in Tanzania.

Indicators observed in Tanzania	Indicator Geddes et al. “A Measure of Personalism in Dictatorships”
1. Overseeing appointments to high office	Does access to high office depend on personal loyalty to the regime leader?
2. Control over membership in the Central Committee and National Executive Committee of the CCM	Does the regime leader control appointments to the party executive committee?
3. Exercising control over decision-making processes in the Central Committee and National Executive Committee	Is the party executive committee absent or simply a rubber stamp for the regime leader's decisions?
4. Personal Control over the security apparatus	Does the regime leader personally control the security apparatus? Does the regime leader create paramilitary forces, a president's guard, or new security force loyal to himself?
5. Appointments of officials along ethnic, regional and family ties	Does the regime leader promote officers loyal to himself or from his ethnic, tribal, regional, or partisan group, or are there widespread forced retirement of officers from other groups?
6. Marginalization of senior party elites and silencing of critics	Does the regime leader imprison/kill officers from groups other than his own without a reasonably fair trial?

Indicator 1 evaluates the control of the dictator over appointments to important positions in the bureaucratic apparatus, military, government and ruling party.⁵⁶ Indicator 2 captures the relationship between the dictator and the leadership of the ruling party. Hereby Geddes et al. are interested if the dictator can choose or veto members of the party executive committee. A concentration of power is observed if the regime leader can select top party leaders instead of party leaders choosing him.⁵⁷ It is possible that a dictator was initially chosen by party members but changes the composition of the party organ in order to gain full control over the composition of the party executive committee.⁵⁸ Indicator 3 variable assesses to what extent power grabs of the regime leader can be observed in the executive committee of the ruling party. If there are no political discussions or debates in higher party organs, it is assumed that the regime leader has already concentrated power over political decision-making in his hands.⁵⁹ Indicator 4 measures if the regime leader can personally control the security apparatus. Personal control over the security forces gives the dictator an information advantage vis-à-vis other members of the political elite, but also the ability to use force against them, e.g. order security personnel to arrest them.⁶⁰ A concentration of power can be detected if the regime leader directly appoints the head of the security service, creates a new security agency or appoints a family member or a friend as head of the security service.⁶¹ Indicator 5 assesses if the dictator promotes those officers loyal to him or from his ethnic, tribal, regional, or partisan group to the military. Moreover, it is measured whether or not he possesses sufficient power to marginalize officers from other groups than his own.⁶² A concentration of power becomes evident if the regime leader appoints a family member or friend as head of this group and if members are mainly recruited from his tribe, region, or clan. Indicator 6 captures if the regime leader is able to imprison or execute officers without fair trials, thereby manipulating decision-making bodies of the military apparatus and eliminating regime critics.⁶³

There is a potential overlap between these indicators since dictators most likely apply different strategies to personalize their power and some events or pieces of information fit in more than one category.⁶⁴ For example, the appointment of a relative as head of the security forces can be seen as evidence for personal control over the security apparatus but also as personal control over appointments to high office.⁶⁵

This new measure of personalism is different from Geddes’ former categorization approach of authoritarian regimes.⁶⁶ It can be applied to analyse the degree of personalism across all authoritarian regime types to detect variations between regimes, differences between leaders in the same regime as well as over time during any dictator’s term in office.⁶⁷

Furthermore, both Geddes et al⁶⁸ and Kendall-Taylor et al⁶⁹ add to our understanding of the ex-ante characteristics of personalization, that weaker, fragmented inner circles and factionalism tend to give rise to a personalization of power. Divided elites are unable to make credible threats to overthrow the dictator if he refuses to share power.⁷⁰ As soon as a dictator takes measures to concentrate power in his hands, his power and resources can be used to eliminate competitors who might challenge him in the future, even if he is initially weak following a rise to power.⁷¹ By appointing members who are relying on the dictator due to a lack of independent supporters, the possibilities of restricting the dictator by members of the inner circle are limited in the long run.⁷²

Most of the information is taken directly from the field interviews. Interviews are numbered as I-1 to I-19. In line with ethical guidelines on protecting respondents in non-democratic settings, the names of interview partners are kept confidential due to the sensitivity of the information.

4. Empirical analysis

4.1. Overview

The sub-Saharan state Tanzania is widely seen as the typical example for a party-based regime.⁷³ Since the introduction of the multi-party system in 1992, the ruling party Chama Cha Mapinduzi (CCM) has never lost its absolute majority and is the longest ruling party in Africa.⁷⁴

The CCM receives its legitimation from its association with Julius Nyerere, who was not only the first president of Tanzania but is also seen as the “Father of the Nation”.⁷⁵ By building strong party institutions and civil-society associations while eradicating societal structures which existed alongside the Tanganyika African National Union (TANU), like chiefdoms, Nyerere strengthened civil-party and civil–military relations.⁷⁶ Despite the institutional rules structuring the relation between the party and the president, personal relations clearly mattered as well.⁷⁷ The political atmosphere can best be described as a “culture of consensus” in which the president was seen as the “architect” and decision-making was achieved through participation and open debate in the higher party organs.⁷⁸ The introduction of Swahili as the official language and Nyerere’s socialist ideology were seen as important steps for the construction of a Tanzanian identity.⁷⁹

However, Nyerere’s socialist project failed in economic terms and he was unwilling to accept the strict conditionalities of the donors.⁸⁰ As a result, different factions emerged in the CCM with reformers who advocated to accept the conditions and socialists on the other side who rejected the reform package.⁸¹

Under Nyerere’s successor Ali Hassan Mwinyi the CCM fundamentally changed its political approach,, which was officially enacted in a NEC meeting in 1990.⁸² As a result of the economic liberalization, an entrepreneurial party elite emerged. These individuals were largely politicians, bureaucrats or ex-bureaucrats who were responsible for certain economic sectors in the government or in the civil administration (I-1). When these sectors were liberalized, they set up their own private businesses in those sectors or acted as commercial middle men between investors and the emerging industries (I-1).⁸³ This was quite a novelty as there was no African bourgeoisie under Nyerere at all (I-1). However, when Mwinyi’s term was about to end, his economic laissez-faire approach was widely accused of having given rise to corruption, land grabbing and lawlessness.⁸⁴ Mwinyi’s government was not only facing a legitimacy crisis, but the intra-party factions of entrepreneurs and socialists began to undermine the institutional apparatus of the party.⁸⁵

In terms of regime survival, the so-called state-party fusion which was preserved after the introduction of the multi-party system led to an “uneven playing field” with huge advantages for the CCM.⁸⁶ After 1995 CCM’s ideological conflict turned also into a generational cleavage with young, dynamic CCM aspirants criticizing the traditional world views of senior CCM members.⁸⁷ Among these younger CCM aspirants were Kikwete and Lowassa, who both competed in the presidential election of 1995 and served as cabinet ministers under Mkapa.⁸⁸ The friendship between Lowassa and Kikwete is a good example for the early formation of networks in CCM around the National Service and the University of Dar es Salaam (I-8, I-9, I-10, I-13). Following the 1995 election, Kikwete, supported by Lowassa and Rostam Aziz, established a faction, which was then known as mtandao to prepare for the presidential elections in 2005.⁸⁹

Besides the activities of mtandao, local and regional interests shaped the factional landscape within CCM even if this was less obvious (I-8, I-14, I-15).⁹⁰ National politicians gathered themselves into loose regional power groups in order to gain attention for their regional activities from the government.⁹¹

Among these competing factions, Nyerere stood out as an authority figure who preserved the original values of the CCM and continued to have extensive influence in the party even after his retirement.⁹² His death in 1999 left an open gap which could not be filled in terms of integrity by members of the party leadership (I-9). An attempt to include traditional CCM values in the presidential primaries was the advisory body of the CCM Elders who counselled the CC and the NEC.⁹³

Kikwete’s election as president in 2005 was a mere formality, due to the rising influence of mtandao and his popularity among younger voters.⁹⁴ After Kikwete’s election as president sharing the fruits of office among influential mtandao members presented itself as challenging.⁹⁵ Factional conflicts took place in parliament, where disappointed CCM members allied with members of opposition parties to defame CCM MP’s or ministers suspected of involvement in corruption (I-1, I-9, I-13).⁹⁶

In the aftermath of the Richmond scandal, CCM members surrounding Sitta associated themselves as CCM safi, a “clean”, anti-corrupt CCM, while Lowassa supporters were presented as CCM mafisadi, a “dirty”, corrupt CCM (I-11, I-15).⁹⁷ Opposition parties, especially CHADEMA, had gained in organizational strength and civic support as the grand corruption scandals outlined factional divisions and excessive abuse of state resources by the CCM elite.⁹⁸

Concerning the relationship between Kikwete and the CCM, his first term revealed that he managed the party in a different way than Mkapa. Being supported by mtandao in the nomination process in 2005, Kikwete was not an impartial leader and rewarded his supporters by distributing patronage (I-4, I-7, I-9).⁹⁹ However, like Mkapa, Kikwete was an experienced diplomat having served as Foreign Minister for 10 years; he also understood and respected the internal rules of the party as he was a long-time NEC member (I-3, I-9, I-10, I-17). Thus, Kikwete was aware of the heterogenous nature of the CCM factions and tried to balance them by outplaying the different groups and distributing the spoils of office (I-1, I-9, I-11, I-14).

After CCM’s relatively poor result of the 2010 election, Kikwete’s laissez-faire leadership style was seen in a more critical light with having given rise to factional tensions which also created a presidential successor problem (I-9, I-10, I-11, I-12, I-13, I-16). Firstly, the different factions divided the party, so that it was not to be expected that they would agree on one joint candidate. Secondly, the nomination of one of the frontrunners of the major factions might have caused further disintegration of the party and a continuation of the conflicts after 2015. In light of Kikwete’s diminishing control over the factional dynamics and therefore also over the presidential successor, the presidential nomination process was adapted in 2012.¹⁰⁰ Observers understood the newly established Ethics Committee as a tactical move of Kikwete and other party leaders to exercise greater control over the nomination process (I-1, I-15, I-16).

During CCM’s internal nomination for the presidential candidate in 2015, it was speculated that Kikwete had promised many candidates his support in order to hide the underlying conflict between Sitta’s safi and Lowassa’s mafisadi faction (I-11, I-13). In fact, the Ethics Committee recommended not only to drop Lowassa but also Sitta and other senior CCM candidates, so that Lowassa’s and Sitta’s rejection appeared less exceptional (I-11, I-16). The three final candidates were Minister of Works Magufuli, Minister of Constitutional and Legal Affairs Asha-Rose Migiro and Representative of the African Union to the United States Amina Salum Ali, none of whom had a factional association (I-10, I-13).¹⁰¹ Magufuli was then elected by the National Congress.¹⁰²

After the election of Magufuli as the 5th president of Tanzania in 2015, a major shift in Tanzanian politics has been observed.¹⁰³ Magufuli’s political agenda displayed an outstanding authoritarian turn for Tanzania, characterized by repression, restriction of human rights and resource nationalism.¹⁰⁴

Magufuli did not refrain from implementing changes in his own party either: reforms, restructuring of CCM party organs and dismissals of so-called “phantom officials” have affected thousands of party members who had profited from the party until then.¹⁰⁵ Those who dared to criticize him openly were either excluded from the party or had to apologize publicly.¹⁰⁶

Nevertheless, researchers were still attempting to analyse the recent political developments in Tanzania under the existing paradigm of a strong, institutionalized regime-supporting party which can control the dictator.¹⁰⁷ Paget¹⁰⁸ recognized that the political discourse was changing, however, in his analysis he decided to leave the rising tensions between CCM members and Magufuli untouched. Due to the discrepancy between theory and empirical observations,

Magufuli’s rule should be investigated under the notion of an autocracy-to-autocracy transition. Increasing factional tendencies in the CCM gave him the opportunity to act as the moderator between competing elites, thereby accumulating personal power. From the system’s perspective, this was functional for regime stability as it prevented power struggles from breaking out more openly.

The next sections will give a detailed account of the increase in personalism along the indicators developed in Section 3.

4.2. Overseeing of appointments to high office

As discussed already, Tanzania’s constitution provides the president with great power to appoint and dismiss officials in the bureaucracy, military, judiciary as well as over cabinet ministers. Magufuli’s personal control over high office has manifested itself through reshuffles, dismissals and structural changes.

Firstly, in his inauguration speech at the Tanzanian parliament on 20 November 2015, Magufuli announced that fundamental changes would take place under his leadership and emphasized that he would do anything to fight corruption, to identify corrupt politicians or officials and to fire those people.¹⁰⁹ Hence, under the guise of fighting corruption Magufuli fired not only more than 10,000 “ghost-workers” in the bureaucracy but also various board members such as the heads of the Tanzania Investment Centre, Tanzania Revenue Authority, the Tanzania Ports Authority and the Director of the Prevention and Combatting of Corruption Bureau.¹¹⁰ In fact, this wave of dismissals, combined with the ban on holding more than one political position per person, affected a large proportion of employees and their families economically, as a whole family often benefited and lived from only one position (I-1, I-12, I-19).

Furthermore, Magufuli reduced the cabinet size from 29 to 19 ministers; eight ministers of Kikwete’s cabinet were not given posts in Magufuli’s cabinet.¹¹¹ In particular, politicians who were associated with the mtandao or mafisadi faction and seen as allies of Lowassa were not appointed, like Sophia Simba (I-12). In later reshuffles Mwigulu Nchemba, January Makamba, Sospeter Muhongo, Charles Tizeba, Charles Mwijage and Kangi Lugola were sacked over alleged underperformance.¹¹² Especially, the dismissal of January Makamba was seen as an attempt to eliminate a potential competitor for the presidency ahead of the 2020 elections (I-3).

Moreover, Magufuli made further structural adaptations by moving the Regional Administration and Local Government (TAMISEMI) from the Prime Minister’s Office under the authority of the President’s Office.¹¹³ Lowassa's influence as prime minister was probably also based on his authority over TAMISEMI, where he was able to expand his network in the regions of Tanzania (I-3, I-9). Thus, as a precautionary measure, Magufuli might have aimed to prevent a future conflict between the president and the prime minister and at the same time centralized more power in the presidency over the premiership.

Observers accused Magufuli of a “hire-and-fire mentality” (I-1, I-3, I-5, I-12, I-18). Magufuli was seen as a president who showed a tendency to appoint those loyal to him to important positions. Personal loyalty was paraphrased as “obeying”, “praising” Magufuli or “saying always yes” to his projects and legislation (I-1, I-3, I-5, I-18). The consequences of disagreements with Magufuli or any doubts about his leadership style often resulted in the loss of a position or even the disappearance or death of the critic (I-8, I-9). In addition, it was argued that Magufuli’s reshuffles were an attempt to dismantle existing networks while positioning his close allies in key positions (I-8, I-15, I-18).

Magufuli’s “hire-and-fire mentality”, his personal control over and dealing with his appointees was showcased by a series of events. First of all, Magufuli replaced the majority of regional commissioners in 2016, while the appointment of Paul Makonda as regional commissioner of Dar es Salaam turned out to be the most controversial.¹¹⁴ On 19 March 2017, Makonda, accompanied by armed security officers, stormed into a radio station of the Clouds Media Group and demanded the broadcasting of a report which would have discredited Bishop Josephat Gwajima who had openly criticized Makonda.¹¹⁵ Various opposition members and Information Minister Nnauye requested an investigation of the incident and legal consequences for Makonda’s behaviour. Somehow unexpectedly, Magufuli did not remove Makonda but sacked Information Minister Nnauye. ¹¹⁶

Meanwhile, Magufuli emphasized that he alone had the mandate to make personnel decisions and that he did not tolerate any interference by the media or other politicians.¹¹⁷ On 20 March 2017 in a closed meeting with CCM MP’s, Magufuli read out parts of the “Riot Act” which allowed him to dissolve the parliament in case CCM MP’s wanted to ally with opposition MP’s and table a vote of no confidence against the prime minister.¹¹⁸

Moreover, two incidents occurred at the Ministry of Finance. Shortly after Controller and Auditor General Juma Mussa Assad presented his budgetary report of 2016/17 and raised questions about irregularities, Magufuli replaced him with Charles Kicherere in 2019.¹¹⁹ There were claims that this replacement of the Controller and Auditor General was an unconstitutional process (I-5). In addition, Leopold Lwabaje, who administered the European Development Funds (EDF), was found dead under mysterious circumstances following a report to police in which he claimed to be abducted.¹²⁰ It was rumoured that Lwabaje had denied the Paymaster General, Magufuli’s nephew, financial resources from the EDF, so that Lwabaje had courted Magufuli’s resentment (I-5, I-8).

One of Magufuli’s demonstrations of power, which also caused international sensation, concerned the dismissal of staff of the National Health Laboratory during the COVID-19 pandemic on 3 May 2020.¹²¹ Magufuli accused the National Health Laboratory of having wrongly tested samples of a goat and a papaya as corona positive. Afterwards he fired the Director of the National Health Laboratory Nyambura Moremi and the Quality and Assurance Director Jacob Lusekelo. Moreover, Magufuli appointed a new Deputy Health Minister, Chief Medical Officer and Permanent Secretary at the Ministry of Health.¹²² Furthermore, he installed a committee to investigate the activities of the laboratory.

4.3. Control over membership of the Central Committee and National Executive Committee

With regard to his own party, Magufuli revealed his ability to exert control over membership in the highest decision-making bodies. With the amendment of the CCM constitution in 2017 Magufuli centralized more power in the role of the chairman. Among other things, this seemed to be one part of his strategy in changing the modus operandi of the party.

First of all, Magufuli drastically reduced the number of seats of the CC and the NEC. Before 2015 the CC had 34 members but Magufuli reduced it to only 24 members.¹²³ The majority of CC members were there by virtue of their position, they were appointees of the CCM chairperson.¹²⁴ Three seats are each reserved for members from Zanzibar and the Mainland, elected by the NEC. The NEC was downsized from 388 members to 158 members.¹²⁵ Even though there are more elected positions in the NEC than in the CC, a vetting process allows the party chairman to screen out candidates “he doesn’t like” before someone can be elected (I-19). Moreover, Magufuli abandoned the so-called kofia mbili possibility, allowing CCM members to hold more than one position; only one position per person was allowed (I-19). Therefore, even though the party chairman cannot appoint every position in the CC and NEC, these adaptations clearly implied a centralization of power, as fewer people were given a voice in the decision-making processes of the CCM.

Secondly, Magufuli appointed “new” or “unknown” CCM members to the CC and NEC (I-3, I-5, I-15, I-19). Some members of Magufuli’s CC and NEC had never held a position in the party, which was quite unusual: before the election of Magufuli, an informal rule or consensus existed that one could only attain a seat in the CC or NEC if one had a large intra-party network, often based on patronage, and was sufficiently trained and “groomed” within the party (I-4, I-15, I-19). Former chairmen were more likely to take the experience and knowledge of party processes and previously held positions of potential candidates into account (I-19). Magufuli’s new people were often academics who had criticized CCM’s practices before their appointments (I-3, I-5, I-9, I-19). By appointing these people, Magufuli demonstrated not only a technique to absorb critics but also his attempt to restore CCM’s public image as academics are considered CCM safi-members (I-15). Due to CCM’s social control and outreach, one cannot easily refuse offers from the party chairman without facing consequences (I-8, I-15).

The appointment of Dr. Bashiru Ally as Secretary General of the party in May 2018 illustrated Magufuli’s strategic decisions. Before his appointment, Ally taught as a senior lecturer at the College of Social Sciences at the University of Dar es Salaam.¹²⁶ On the one hand, he was known for criticizing the practices of CCM (I-3, I-5, I-9, I-15). On the other hand, Ally was associated with socialist values and Nyerere’s policies partly because Issa Shivji has supervised his doctoral thesis (I-3, I-8, I-9, I-15). Nevertheless, a person like Ally who has never held any position in the party would have never been considered for such an outstanding position as the Secretary General before 2015 (I-19). In December 2017, Magufuli appointed Ally as head of a committee to investigate CCM’s assets (Kolumbia 2018). Shortly after, Ally was appointed as the new Secretary General. Thus, Ally’s appointment clearly demonstrated Magufuli’s strategy to make CCM “cleaner” but also displayed Magufuli’s mistrust of party elites (I-15). 

4.4. Exercising control over decision-making processes in the Central Committee and National Executive Committee

Another aspect of Magufuli’s personalism was his management of decision-making processes in the highest party committees. In his role as party chairman Magufuli chaired both committees. To recall, the CCM has always been a heterogenous party, which, especially during Kikwete’s leadership, has been characterized by severe factional differences (I-1, I-9, I-13, I-17). But a culture of consensus, with the search for compromise through discussion and open debate, had structured the party's decision-making processes since the rule of Nyerere, regardless of who the party chairman was (I-1, I-9, I-17, I-18).

However, it was mentioned that almost no debates took place in the CC and NEC under Magufuli’s chairmanship (I-5, I-18, I-19). There was hardly any discourse nor were there opinions other than those of the party chairman raised by members of the decision-making bodies. In contrast to the pre-2015 dynamics, CCM’s various factions seemed to have disappeared under Magufuli (I-1, I-10, I-13). Magufuli himself was willing to make decisions and implement changes (I-13, I-17). In July 2020, Magufuli was unanimously elected as CCM’s presidential candidate by the National Congress.¹²⁷

Various attempts to explain the absence of debates in the CC and NEC were raised during the interviews. For instance, it was assumed that CC and NEC members were still in the “sit-and-wait phase” (I-17). Thus, they analysed the current situation, the changes Magufuli had induced and tried to understand in which direction the party was heading; a reason for the lack of debate (I-17).

More convincing seemed to be the explanation that certain conflicts and cleavages existed by that time, but they were suppressed or “not dealt with” (I-12, I-14). An atmosphere of fear and brutality characterized Magufuli’s chairmanship (I-5, I-14). It was reported that members of the larger committee NEC were openly threatened in internal meetings by Magufuli not to raise their voice against him and not to dare to challenge him in the presidential election in 2020 (I-5, I-9). What was perceived as one potential bargaining chip of Magufuli and his Secretary General Ally vis-à-vis influential CCM members was Ally’s report over the CCM assets (I-7, I-8, I-12, I-4, I-15). Although it was publicly announced that Ally’s report included a list of about 100 CCM members who were involved in irregularities and corruption, the report was not published.¹²⁸ Therefore, it was speculated that the report was shared only among Magufuli’s inner circle and might has been used to pressurize CC and NEC members in case of non-compliance (I-12, I-14, I-15).

Another incident, which suggested that there were animosities at the party leadership, was the poisoning of CCM National Vice Chairman Mangula during a CC meeting in the presence of Magufuli.¹²⁹ Interestingly, this was the same CC meeting in which the CC members discussed consequences for the behaviour of Kinana, Yusuf Makamba and Membe.¹³⁰

The absence of dialogue was also observed in the way decisions were made and framed by the Tanzanian government. For instance, Magufuli showed a tendency to govern through orders, directives and edicts instead of following official procedures.¹³¹ Furthermore, official statements often included the sentence “Tunaunga mikono juhudi za Raisi wetu John Pombe Magufuli” meaning “We support the efforts of our President John Pombe Magufuli” which could be interpreted as an order to comply to Magufuli’s rule unconditionally (I-13).

4.5. Appointments of officials along ethnic, regional and family ties

A frequently pronounced feature of Magufuli’s rule was his tendency to appoint members of his Sukuma ethnic group from the Lake Zone and from his family to government and party positions (I-1, I-2, I-5, I-8, I-9, I-13, I-14, I-18). In brief, Magufuli came from Chato, a town in Geita region located in North-Western Tanzania close to Lake Victoria.¹³² Strictly speaking Magufuli was a member of the ethnic group Zinza, a sub-group of the larger ethnic group Sukuma, but typically these smaller sub-groups identify themselves as Sukuma (I-8). The Sukuma are the largest ethnic group in Tanzania and they traditionally live in the Lake Zone in the regions Geita, Mwanza, Shinyanga, Simiyu, Mara, Kagera and also in Tabora in Western Tanzania.¹³³ Together with the Nyamwezi who also live in these regions, the Sukuma account for approximately 19% of the Tanzanian population. Thus, importantly, Magufuli was the first president of Tanzania to come from such a large ethnic group (I-5, I-9).

In CCM, Magufuli has filled the most influential and strategically important positions with allies from his region (I-17). His Secretary General Ally, the national chairperson of CCM’s Women’s Wing Gaudensia Kabaka were both from the Lake Zone, while the national chairperson of CCM’s Youth League Kheri James was Sukuma (I-8). The Secretary General of the CCM has great control over the daily business of the party as the national chairpersons of the Youth League and Women’s Wing can exert influence onto CCM’s youth and female members (I-17).

More Sukuma and people from the Lake Zone were appointed by Magufuli to positions in the government, bureaucracy and parastatals. The Minister of Industry, Trade and Investment Innocent Lugha Bashungwa, Minister of Natural Resources and Tourism Hamisi Kigwangalla, Minister of Minerals Mashaka Biteko and Minister of Energy Medard Kalemani were from the Lake Zone (I-8). Kigwangalla was Nyamwezi, while Biteko and Kalemani were Sukuma (I-8). Kalemani was the MP for Magufuli’s electoral constituency Chato.

In the judiciary and security sector appointments reflected his ethnic and regional ties as well. The Attorney General Adelardus Kilangi, Deputy Attorney General Evarist Longopa, Solicitor General Gabriel Paschal Malata and Deputy Solicitor General Boniface Luhende were Sukuma (I-8). The Chief Justice Ibrahim Juma, the Inspector General of Police Simon Sirro, the Chief of the Defence Forces Venance Mabeyo, the Chief of National Service Charles Mbuge and Commissoner General of Prison Phaustine Martin Kasike, all appointed by Magufuli, came from the Lake Zone (I-5, I-8). Furthermore, the Chief Medical Officer Prof. Abel Makubi and the Executive Secretary of the Tanzania Investment Center Dr. Maduhu Isaac Charles were Sukuma as well (I-8).

Another incident concerned the appointment of Magufuli’s nephew Doto James as Permanent Secretary and Paymaster General of the Ministry of Finance in 2016 (I-5, I-8). Even though James had only worked at the Tanzania National Roads Agency, he was appointed as Deputy Permanent Secretary of the Ministry of Finance by Magufuli in 2015. After a few months only, James was appointed as Permanent Secretary and Paymaster General of the Ministry of Finance, while the long-serving Servicius Likwelile who had been an appointee of Kikwete was dismissed.¹³⁴ Newspaper outlets described James as a “key confidant” of Magufuli.¹³⁵

When the Controller and Auditor General Juma Assad published his report for the fiscal year 2016/2017 in 2018 and outlined a difference of 1.5 trillion TSH between tax revenue and expenditures, Magufuli and James rejected all accusations of their involvement.¹³⁶ However, speculations continued that Magufuli strategically appointed his nephew to this position to have seemingly legal control over foreign aid and government expenditures (I-5, I-8).

5. Conclusion

As described above, Magufuli personalized power in different domains. Because of CCM’s factional tensions ahead of the 2015 elections, party elders paved the way for Magufuli’s nomination. While Magufuli may have initially come from a position of weakness, he showed his ability to personalize power during his presidency. The basis for his success lied in the increasing factional tensions within CCM which gave him the power to act as the arbiter and to manipulate party institutions and nominations to his favour.

Moreover, the analysis has shown that in some areas in which Magufuli has personalized power, he merely amplified political rules and legislations already in place. Concerning the control of appointments to high office, the military and security apparatus, these are areaswhich, according to the constitution, Tanzanian presidents always had great powers over. But unlike Magufuli, his predecessors placed their available power on hold in favour of the party. Through gradual changes like rising appointments of loyalists from his ethnic group, region and family, Magufuli extended his personal control over personnel decisions.

Magufuli’s personal control over decision-making in the CC and NEC while marginalizing senior party elites and silencing of critics signifies a new development. All other presidents including Nyerere, who gave the CCM its legitimacy, have seen themselves as “party figures”. Though, similarly to Magufuli’s rule, their actions to maintain a unified party also included playing the factions off against each other or temporarily withdrawing positions from party members if they posed a threat to the unity of the party. However, they never prohibited party members from speaking out, expelled senior party members from the party or suppressed the party's culture of discussion and debate like Magufuli did. Respect for the party elders and their influence in decision-making seem to have steadily declined under Magufuli’s predecessors. Even though the party elders played an important role in the 2015 presidential nomination, their voice in the party has lost significance increasingly under Magufuli.

Arguably, the “sit-and-wait mentality” of party members has probably also favoured Magufuli's concentration of power, because a large proportion of party members have observed Magufuli’s changes only passively. Even though he came from a position of weakness, he managed to side-line intra-party factions and any other critics who he perceived as threats to the regime.

From the author’s point of view, Magufuli’s personalization of power affects the classification of the Tanzanian authoritarian regime. So far, Tanzania has always been coded as a party-based regime. Magufuli’s presidency clearly showed elements of personalist rule such as his ability to appoint, promote and dismiss high-level officials. Moreover, he has shown his ability to manipulate political actors and decision-making in the CCM.

To conclude, Tanzania under Magufuli is better classified as a party-personalist regime. The mechanism which is proposed here could be applicable to a wider range of African and non-African cases. Our work could therefore stimulate empirical and theoretical investigations of power distribution and institutional variation within authoritarian regimes.

Disclosure statement

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

Additional information

Funding

This work was supported by H2020 European Research Council [grant number ERC Starting Grant 2017 – DLEDA – 759537].

Notes on contributors

Anja Osei

Anja Osei is a professor of comparative political science with a special focus on Africa at Freie Universität Berlin. Her research interest is in the field of democratization, authoritarianism, political elites and political institutions.

Elisabeth Bruhn

Elisabeth Bruhn holds a Master's degree in Politics and Public Administration with a specialization in International Administration and Conflict Management. She works as a Policy Officer at the Federal Statistical Office of Germany.

Notes

1 On elite defections in Tanzania and Cameroon in comparative perspective, see Morse, How Autocrats Compete.

2 Collord, “Wealth, Power and Institutional Change in Tanzania’s Parliament”.

3 Geddes, “What Do We Know About Democratization After Twenty Years?”

4 Geddes, Wright, and Frantz, “Autocratic Breakdown and Regime Transitions.”

5 Ibid.

6 Geddes, “What Do We Know About Democratization After Twenty Years?”.

7 Ibid., 122.

8 Smith, “Life of the Party.”

9 Lachapelle et al., “Social Revolution and Authoritarian Durability.”

10 Huang-Ting, “Does the Constitution Matter?”

11 Panta, “The Stubbornness of Authoritarianism,” 153; Wahman, Teorell, and Hadenius, “Authoritarian Regime Types Revisited”; Geddes, Wright, and Frantz, “Autocratic Breakdown and Regime Transitions.”

12 Panta, “The Stubbornness of Authoritarianism.”

13 Ibid., 153; Geddes et al., “Autocratic Breakdown and Regime Transitions,” 325f.

14 Geddes, Wright, and Frantz, 313.

15 Lührmann and Lindberg, “A Third Wave of Autocratization is Here,” 1108.

16 Grundholm, “Taking It Personal?” 16; Fumagalli, “Social Contention, Authoritarian Resilience, and Political Change,” 1221; Hill, “Authoritarian Resilience and Regime Cohesion.”

17 Francois, Rainer, and Trebbi, “How is Power Shared in Africa?”

18 Slater, “Iron Cage in an Iron Fist.”

19 Ibid., 82.

20 Ibid., 96.

21 Bayulgen et al., “Elite Survival Strategies and Authoritarian Reversal”; Yilmaz et al., “How an Islamist Party.”

22 Baturo and Elkink, “Dynamics of Regime Personalization.”

23 Shirk, “China in Xi’s ‘New Era’.”

24 Morgenbesser, “Misclassification on the Mekong.”

25 Geddes et al., “A Measure of Personalism in Dictatorships.”

26 Ibid.

27 Ibid.

28 Ibid.

29 Morgenbesser, “Misclassification on the Mekong,” 193.

30 Casal Bértoa, “Political Parties or Party Systems?” 410.

31 Magaloni and Kricheli, “Political Order and One-Party Rule”; Reuter and Turovsky, “Dominant Party Rule”.

32 Brownlee, “Hereditary Succession in Modern Autocracies”.

33 Magaloni, Voting for Autocracy.

34 Levitsky and Way, “Beyond Patronage.”

35 Ibid.

36 Masiya and Maringira, “The Use of Heroism.”

37 Levitsky and Way, “Beyond Patronage,” 872.

38 Khisa, “Managing Elite Defection.”

39 Andrews and Honig, “Elite Defection.”

40 Brownlee, Authoritarianism in an Age of Democratization, 33.

41 Boucek, “Rethinking Factionalism.”

42 The work of Loxton and LeBas also points in that direction. While Loxton and Levitsky explore the relationship between charismatic leaders and parties in Latin America, LeBas argues that personalism could stabilize authoritarian successor parties after defeat. Although she does not develop this in detail, personalism might also stabilize parties in crisis or prevent a mounting crisis.

43 See, among others, Gandhi, Political Institutions under Dictatorship.

44 Brunkert and von Soest, “Praising the Leader”.

45 For a critique of this conflation see Pitcher, Moran, and Johnston, “Rethinking Patrimonialism and Neopatrimonialism in Africa”.

46 Grundholm, “Taking It Personal?”; Brunkert and von Soest, “Praising the Leader”.

47 Cheeseman, Matfess, and Amani, “Tanzania.”

48 Gandhi, Noble, and Svolik, “Legislatures and Legislative Politics without Democracy”; Blaydes, Elections and Distributive Politics in Mubarak’s Egypt.

49 Levitsky and Way, “Beyond Patronage”.

50 Sulley, “Democracy within Parties”.

51 Pepinsky, “The Institutional Turn in Comparative Authoritarianism”.

52 Gray, “The Political Economy of Grand Corruption in Tanzania”.

53 Geddes et al., “A Measure of Personalism in Dictatorships,” 2.

54 Two items regarding the security apparatus are collapsed into one. The item whether the ruler created a new political party was dropped. We are looking at the gradual shift from a party-based to a personalist system. Magufuli came to power thanks to the CCM, it is therefore more reasonable to turn the existing party into a more personalized vehicle of power than to form a new party from scratch.

55 While Geddes’ indicators are developed to be applicable for a large range of countries, we rename them to reflect what exactly our data contain. Our approach is thus similar to that of Morgenbesser’s 2018 study. The general content of the indicators is maintained, but we achieve greater precision for our single case study.

56 Geddes et al., “How Dictatorships Work,” 81.

57 Ibid.

58 Ibid., 81.

59 Ibid.

60 Ibid., 80.

61 Ibid.

62 Ibid., 81.

63 Ibid., 82.

64 Ibid.

65 Geddes, Wright, and Frantz, “A Measure of Personalism in Dictatorships,” 2.

66 Geddes et al., “How Dictatorships Work,” 82.

67 Ibid.

68 Ibid., 93.

69 Kendall-Taylor et al., “The Global Rise of Personalized Politics,” 10.

70 Geddes et al., “How Dictatorships Work,” 93.

71 Ibid., 85.

72 Ibid.

73 Geddes, “What Do We Know,” 124.

74 O’Gorman, “Why the CCM Won’t Lose,” 313.

75 Whitehead, “Historical Legacies,” 1087.

76 Bienen, Tanzania: Party Transformation and Economic Development, 71; Morse, How Autocrats Compete, 103.

77 Van Donge and Liviga, “Tanzanian Political Culture and the Cabinet,” 636.

78 Ibid.

79 Thompson, African Democracy, 301.

80 Shivji, “Nationalism and pan-Africanism,” 110.

81 Thompson, African Democracy, 339.

82 Mmuya, Tanzania: Political Reform in Eclipse, 15–16.

83 Collord, The Political Economy of Institutions in Africa, 113.

84 Hyden, “Top-Down Democratization in Tanzania,” 144.

85 Collord, The Political Economy of Institutions in Africa, 115.

86 Makulilo, “Authoritarian Stability across Space,” 177.

87 Tsubura, “Umoja ni Ushindi,” 68.

88 Ibid.

89 Makulilo, Populism and Democracy in Africa, 179.

90 Kelsall, “Shop Windows and Smoke-filled Rooms,” 612–613.

91 Ibid.

92 Tanzania: Political Reform in Eclipse, 71.

93 Tsubura, “Umoja ni Ushindi,” 67.

94 Morse, How Autocrats Compete, 138. Ibid.

95 Collord, The Political Economy of Institutions in Africa, 174.

96 Ibid., 180.

97 Collord, The Political Economy of Institutions in Africa, 177. Tsubura, “Umoja ni Ushindi,” 70.

98 Paget, “The Authoritarian Origins of Well-organized Opposition Parties,” 704.

99 Collord, The Political Economy of Institutions in Africa, 174.

100 Ibid., 126.

101 Tsubura, “Umoja ni Ushindi,” 73.

102 Ibid.

103 Paget, “Tanzania”; Pallotti, “Lost in Transition?”

104 Africa Confidential, “Laying Down the Law”; Paget, “Tanzania,” 156; Jacob and Pedersen, “New Resource Nationalism?” 291.

105 Andreoni, “Anti-Corruption in Tanzania.”

106 Kimboy, “Kinana, Makamba and Membe to be Grilled by CCM Disciplinary Committee”; Yamola, “Membe: What Tanzania Needs for Elections to be Truly Free, Fair”.

107 Morse, “Electoral Authoritarianism and Weak States in Africa”; Collord, “Wealth, Power and Institutional Change.”

108 Paget, “Again, Making Tanzania Great.”

109 Magufuli, “Hotuba ya Rais wa Jamhuri ya Muungano wa Tanzania, Mheshimiwa Dkt. John Pombe Joseph Magufuli, akifungua rasmi Bunge jipya la Jamhuri ya Muungano wa Tanzania, Dodoma”.

110 Andreoni, “Anti-Corruption in Tanzania,” 33.

111 Mtulya, “Magufuli’s Cabinet Leanest in 20yrs, 7 JK Ministers Back.”

112 Paget, “Again, Making Tanzania Great.”

113 Kwayu, “Tanzania’s COVID-19 Response.”

114 El-Noshokaty, “Präsident Magufuli Macht Sich Feinde.”

115 https://www.thecitizen.co.tz/tanzania/news/national/makonda-going-nowhere-jpm-2583824

116 The Citizen Reporter, “Who’s Next in Magufuli’s Crosshairs?”

117 El-Noshokaty, “Präsident Magufuli Macht Sich Feinde.”

118 Ibid.

119 The Citizen Reporter, “Mixed Reactions Greet Appointment of New Tanzania Controller and Auditor General.”

120 The Citzien, “Missing Finance Ministry Director Found Dead.”

121 Africa Confidential, “The Magufuli Experiment.”

122 Ibid.

123 Andreoni, “Anti-Corruption in Tanzania.”

124 Chama Cha Mapinduzi, Katiba Ya Chama Cha Mapinduzi.

125 Ibid.

126 The Citizen, “The Spotlight Now Turns on CCM’s New Strategist.”

127 The Citizen, “CCM National Delegates Conference in Dodoma.”

128 Therkildsen and Bak, Democratisation in Tanzania, 12.

129 The Citizen, “Phillip Mangula Makes First Public Appearance Months After Poisoning Claims.”

130 Ibid.

131 Bertelsmann Transformation Index, Country Report Tanzania, 2020.

132 Masare, “President Magufuli Takes a Breather from Campaigns.”

133 Abrahams, The Peoples of Greater Unyamwezi; Babeiya, “Multiparty Elections and Party Support in Tanzania,” 92.

134 The Citizen Reporter, “JPM Picks Key Confidant Treasury PS.”

135 Ibid.

136 Anonymous Author, “Tanzania Search for Missing Millions Raises Questions Over $1 Billion.”