Authoritarian clientelism: the case of the president’s ‘creatures’ in Cameroon

Within this article, we aim at exploring the topic of clientelism in Cameroon as a species of a wider phenomenon affecting Central and Western Francophone Africa. Our argument is that, in spite of the repeated efforts of reinforcing the local power structures, we have witnessed a process of centralisation of clientelism: the new networks are shaped around the ‘Creatures’, who are the President Paul Biya’s formal or informal appointees and play the role of nodal elements relying the rest of the chain to the central command. This happened on the expenses of the locally dispersed and more autonomous clientelistic groups that were either included in or smashed by the pyramidal Creatures’ structured. In order to test our assumption, we analysed a specific body of literature on the theorization of clientelism and on its African and Cameroonian specificity and organized four focus groups with the actual and former members of the clientelistic chains at different levels (central, regional and local). If our main presupposition proved to be generally correct, one of the sub-arguments was only partially validated through this empirical component of our research.


Introduction
Literature on political clientelism is already abundant and diverse. Various theoretical aspects, such as defining clientelism (Kitschelt and Wilkinson 2007), exploring its historical roots (Piatoni 2001) or designing and amending different models of clientelism (Gherghina and Volintiru 2017) have been deeply and convincingly approached by numerous relevant authors. Among the key dimensions explored in literature, some relevant analyses focused on the cultural differentiations within the worldwide phenomenon of clientelism (Briquet and Sawicki 1998), leading to local, regional, national or 'civilizational' specificities of the patron-client networks (Gans- Morse et al. 2014).
Within this latter set of scientific concerns, the Sub-Saharan African societal and political systems appear as being among the most fertile grounds for studying clientelism. This is mainly because the ethno-tribal relations are dominant in almost all the countries of the region and party politics developed within a pre-structured framework where the traditional bonds of affinity and allegiance, which were mediated via complex systems of ritual offer and informal taxation, had to be extended and institutionalized to the newly created 'modern' national layers (Posner 2005, pp. 1-20;Falola and Mbah 2014, pp. 5-8). Consequently, as various authors observed, the traditional practices of patronage were the fishbone structure for the emergence of the wider political formations (Olivier de Sardan 1996;Basedau et al. 2011). These practices lost their immediacy because of the scale-effect of the new national frameworks and created increasingly diffuse, sophisticated and difficultly readable wider patron-client systems. Moreover, these practices are not only contributed to the initial emergence of the Sub-Saharan African political systems but also remained dominant even several decades after decolonization (Cheeseman and Larmer 2013;Kramon 2017, pp. 3-31).
At the same time, studying clientelism and other related practices in Sub-Saharan Africa proves to be revealing for the general understanding of this phenomenon. Firstly, because of the richness and variety of the clientelistic practices and of the various forms of hybridization between clientelism and other related phenomena in the region, such as, for instance, neopatrimonialism (Bach and Gazibo 2013). And then because the Western, Eastern and Austral African countries are the mirror of the radically uneven economic growth that the contemporary phase of globalization produced, with extremely thin strata of wealthy and cosmopolitan elites and with the quasi-totality of the others confronted with poverty and severe marginalization (Sembene 2015). Consequently, the pre-existing state of material dependence emerging during the post-colonial decades accentuated accross the last twenty years and allowed for the further consolidation and diversification of the clientelistic structures (Harrison 2006). But such processes of polarization seem to affect also an increasing number of countries in Latin America (e.g. Brazil and Venezuela), Southeast Asia (e.g. Vietnam and the Philippines) or the former Soviet space (e.g. Ukraine and Kazakhstan) (Chakravarty 2010;Berenshot and Aspinall 2020). The Sub-Saharan case studies for clientelism could thus enhance our capacity to understand its complexity in economically and socially increasingly polarized societies.
Our research aims at bringing a contribution to the understanding of clientelism by producing an analysis of the new particular manners in which, in quasi-authoritarian regimes, clientelism becomes centralized and rearticulated around nodal points. What is the specificity of these nodal points within these clientelistic systems? What are their functions and how do the "inferior" layers of the clientelistic networks perceive their status and influence within the political systems, especially in terms of mobilization and resource distribution? As we will show, these nodal points are embodied by key actors known as 'Creatures' who have become essential for both the stability and the permanent adaptation of the political systems, insuring in this way their continuity. Concentrating on Cameroon as a case study, we will analyse the contribution of the Creatures-defined as political-administrative elites who owe their existence to a presidential formal or informal decision of nomination-to the perpetuation of the incumbent president in power via their role within the clientelistic system.
We organize our article in four parts: (1) In the first, we will briefly study the theoretical framework and will present our main arguments; (2) in the second, we will motivate our choice of Cameroon as case study and will present the research design; (3) in the third, we will shortly present the Cameroonian political context, insisting on aspects related to clientelism; (4) and in the fourth, we will test our arguments using and discussing the results of the four focus groups we conducted with former members of Creatures' staffs and with party militants in three different key-regions of the country.

Theory and argument
While literature extensively concentrates on classical cases, such as Western Europe (Piattoni 2001), North America and Southeast Asia (from Scott 1972 to Hicken 2011) and more recently on Eastern Europe (Volintiru 2016) and Latin America (Hilgers 2013), African political systems' clientelisms have been to some extent more tardily approached (Banywesize 2015;Ruiz de Elvira et al. 2019, pp. 5-22). Moreover, among the existing studies on clientelism in Africa, the ones on former French Equatorial Africa are even less numerous. Why is that? First, because the studies about clientelism follow the iron laws of progressiveness in science: as the African states are among the latest states to be created, the understanding processes of the political phenomena taking place in those states has lacked for a very long time the necessary knowledge infrastructures in terms of capacity of understanding. And then, because of the legacy of the French colonial model itself. As opposed to the British one, the French model imposed a Jacobin pattern to its colonies, with no recognition of the local chieftains and other traditional leaders, and with a homogeneous and centralized administrative management of the territories (Sy 2017). This led to an almost total ignorance with regard to the mechanisms of functioning of the 'indigenous' society and consequently, later on, of the processes of networking, political aggregation, decision-making or policy implementation within the local and the regional layers of the post-colonial states.
Early literature discusses the incipient development of the electoral behaviour in Francophone Western and Central Africa and stresses its specificity: the feeble nature of political-electoral competitiveness and the abundance of the single-party or one-dominant-party systems (Debbach 1966;Jackson and Rosberg 1982;Aly Dieng 2012); the overwhelmingly ethno-tribal vote (Bogaards 2004, pp. 59-60;Carbone 2006); and the full dependence of the electorates of the elites' will to discretionarily distribute the whole range of basic resources and services, including food, water, electricity, vaccination and minimal security. This is what remained as la politique du ventre (the 'politics of the stomach'), a famous formula through which Jean-François Bayart (Bayart 1989) described the penchant of the African communities to promise their electoral support to those who promise in return the fulfilment of these basic needs (which does not imply that either of the two would keep their promise as such). Client-patron relationships would thus be from the outset the very way of functioning of African politics (Médard 2007;Lindberg 2003, p. 140). The democratic wave which followed the fall of Communism in the USSR and in Eastern Europe (1989)(1990)(1991) and which enhanced pluralism or, in other cases, engendered it for the first time could only reinforce clientelism, as more competitors meant more opportunities to restructure the existing networks and to create new ones (Bratton and Van der Walle 1994, pp. 453-6). The increasingly competitive electoral systems of the Central and especially of the Western former French colonial African countries made the patron-client relations system more complex and less easily readable, especially for the non-local observers (Morice 2000;Van der Walle 2014, pp. 230-243).
Within the existing literature on clientelism in the Sub-Saharan Francophone countries, most contributions have focused on the states where pluralism has produced either regular alternation in power, as it has been the case in Benin and Senegal, where some incumbent presidents and their parliamentary majorities occasionally lost the elections (Wantchékon 2003;Beck 2008;Mişcoiu 2012), or on the cases of at least one change of majority (like in Guinea, Niger, Mali and the Ivory Coast). With some remarkable exceptions (such as Lemarchand 1972), fewer papers concentrate on clientelism in the semi-authoritarian regimes, where real alternatives to power have never happened since the 1960s independence, as it is the case in Gabon, Congo-Brazzaville, Togo and Cameroon. The main focus of these writings is on the clientelistic systems of influence established between France via its industrial holdings and its former African colonies, particularly salient in the cases of the oil exporting Gabon and Congo-Brazzaville (Obiang 2007, pp. 9-27). On a contrary, there are even fewer analyses concerning the domestic systems of clientelism proper to the semi-authoritarian regimes in former French Central Africa (Vicente and Wantchékon 2009;Lust-Okar 2009).
Moreover, one of the most salient theoretical delineations-the one between organizational clientelism and electoral clientelism (Kopeczy et al. 2012) -seems to be only partly pertinent when studying the semi-authoritarian Central African systems. On one hand, the circuits between the local party leaders and the other local elites are fluid and the party organizational structures and rules are at most formal, as their existence and configuration permanently depends of the authoritarian decisions taken at the central level (De Walle 2003;Diop 2006, pp. 11-35). Thus, who belongs to a political party in a particular moment and who does what within that party can be changed by decisions taken by the top-hierarchy: potent local businessmen and religious or traditional leaders are permanently co-opted in the parties, often invited to take over local or regional branches and might be removed with the same haste (Carbone 2006;Souaré 2017). On the other hand, local and regional party leaders engage to a lesser extent with regular voters in relations of direct electoral clientelism (Driscoll 2018). Instead, because of the complexity of the still traditional societal configurations, they have to persuade the relevant chieftains who control at their turn the votes of the traditional community leaders and the heads of wider families (Lemarchand 1972, pp. 72-76). Some chieftains themselves may directly play the roles of local party leaders or may wish to claim such functions, reversing in this way the patron-client relations and thus blurring even more the distinctions between organizational and electoral clientelism (Blundo et al. 2007, pp. 102-112;Lindberg and Morrison 2008, pp. 108-115).
Diverse academic researches indicate that several aspects related to the constitution, the componence and the functioning of the clientelistic structures are both important and problematic. Firstly, the very fishbone of the clientelistic structures and especially the identity of the actors who fulfil the role of the structures' articulatory vertebrae. Several authors consider that these structures depend of who are the nodal points of their articulation, both in terms of identity, origin or social, economic and cultural status, and in terms of mechanisms of selection (election, nomination, self-empowerment etc.) (Abente-Brun 2014; Szwarcberg 2015, pp. 1-32). Secondly, attention has been payed to the role distribution within the clientelistic structures and to the capacity of its different elements to mobilize voters, either directly or via some more local influencers whom they co-opt and then control (such as local chieftains, heads of tribes, councillors etc.) (Roniger 1994, pp. 208-210;Osei 2012, pp. 256-259). And finally, other authors investigated the aspects related to effective capacity of the structures' nodal elements to contribute to the preservation of the loyalty of the group's members and to the fulfilment of their targets in terms of the number of voters and/or in terms of insuring the loyalty of the local factors of influence who mobilize at their turn the constituents (Beck 2008, pp. 222-226;Siegel 2017).
In formulating our main argument, we kept in mind these series of aspects that have recurrently appeared as being relevant for the understanding of the clientelistic practices under semi-authoritarian regimes. Our argument is that the particular type of clientelism established in the Central African semi-authoritarian regimes is meant to insure the re-election of the incumbent President through the extensive use of the influence of the 'Creatures', who play a crucial triple role: (1) they mobilize the local and regional networks for the elections, (2) they insure the party discipline and allegiance to the decisions of the President, and (3) they act as negotiators between the political command centre that dispatches the resources and the local stakeholders who directly provide the electoral support.
The origin of the notion of 'Creature' is traceable in a 2010 speech of the former Cameroonian Minister of Higher Education, Jacques Fame Ndongo that became rapidly viral and determined a vivid public debate about the establishment's degree of responsibility and reliability: We are all the creatures or the creations of President Paul Biya, it is him who is entitled to the whole glory of whatever we do. None of us is important, we are nothing more but his servants or, even better, his slaves. 1 Borrowing this already publicly entrenched term, we define as Creature any senior political-administrative elite that owes his/her career to a presidential decision (Malaquais 2002). This includes several categories of appointed or elected members of the political-administrative (Pigeaud 2011). In the following section, after explaining the reasons for choosing our case study, we will define, explore and discuss in detail the notion of Creature.

Research design
Choosing the case study: why Cameroon?
Among the Sub-Saharan Francophone ex-colonies, Cameroon is a particularly interesting case for clientelism for at least two major reasons, the first concerning the constitutional structure and the historical evolution of the country and the second being related to length of uninterrupted continuity in power of the actual President. Following the guidelines formulated by John Gerring (Garring 2016, pp. 39-55), these two reasons should be sufficient to justify our option for Cameroon as case study for analysing the above-mentioned aspects of clientelism. First, because it is a state founded as a result of the merger of two territories, one having been a part of the French colonial empire (with a special status), the other (a smaller South-Western and North-Western part) being before a segment of a wider British colony. Consequently, after the Independence, Cameroon became initially a federation, with two widely autonomous pillars (Johnson 1970, pp. 26-40;Ndi 2014). The progressive reinforcement of the Capital's power under the increasingly dictatorial rule of the first President, Amadou Ahidjo, allowed for the transformation of the country into a unitary state in 1972. This process required the penetration of the Anglophone areas by the pro-Yaoundé networks of influence (Bayart 1985). Such an operation could not have taken place in a direct and non-mediated way, but only via the construction of some regional English-speaking networks and some national bilingual networks, all attached to the central government via the customary practice according to which while the President was naturally Francophone, the Prime Minister should be a representative of the Anglophone community. The formation and the effectiveness of this clientelistic system, with semi-autonomous Anglophone regional and national leaders, organized in the Social Democratic Front/SDF, prompting (sometimes conditionally) the Francophone ruling party, the Cameroon People's Democratic Movement/RDPC, made it a model for the entire country, so for the main Francophone territories also (Banock 1992;Etogo 2012).
A second reason why Cameroon is an interesting case for clientelism is that, while there are other countries in the area with limited or no real alternative in power (Mişcoiu 2018), Cameroon is the only one where there has been the same party in power since the Independence (1960) and where the same President has been leading with no interruption since the era of the single state-party to today's formal pluralism (with more than 300 registered parties). The fact that Paul Biya has been continuously in power since 1982, ruling the country through the same single and then dominant party has allowed him to appoint, control and handle several entire generations of political and administrative leaders at all levels (Kakdeu 2015). For instance, up to April 2019, the President of the Republic of Cameroon had formed 36 governments (with 315 ministers), an average of one government a year since he took power. This qualifies Cameroon as a special case of centralized linguistically and regionally heterogeneous clientelism.

Identifying the creatures
In the semi-authoritarian regime of Cameroon, all major appointed or elected public officials are designated by the same person who is at the same time the national president of the party (RDPC) and the President of the Republic. Thus, not only the appointed officials but also the candidates of the party in power to parliamentary and local elections are nominated by the same country's and party's head, Paul Biya. In this sense, he is effectively their "Creator" and consequently they are his "Creatures". Who are these Creatures? Based on a prior systematic research (Mişcoiu and Kakdeu 2019), we identified several categories of Creatures.
First, major officials in the central administration. The acting government appointed in 2019 counts 63 members including a Prime Minister, 4 ministers of State, 31 ministers, 12 deputy ministers, 5 ministers in charge of missions and 10 secretaries of state. According to the 2011 presidential decree reorganizing the Presidency of the Republic, 6 employees of the President of the Republic are also members of the government (the secretary general of the presidency and his two deputies, the director of the civil cabinet and his deputy, and the general delegate for national security). According to the organization of the Prime Minister's Services, 3 prime ministerial staff also have the rank and prerogatives of minister (the Secretary General of the Prime Minister's Service and his deputy, and the Director of the Prime Minister's Office). So, we have a total of 71 Creatures who are members of the Government.
At the same central level, in the ministerial departments, Cameroon has 37 secretaries general of ministries, 364 delegated or entrusted governmental commissionaires and 55 directors of public establishments. This makes a total of 456 other Creatures in the second ranks of the ministries.
Then, majority MPs in the National Assembly and the Senate. Out of the 180 deputies of the National Assembly elected in February 2020, 152 belong to the RDPC. To these deputies we should add the 3 members of the Assemblies' bureau appointed by presidential decree (including the secretary general and 2 deputies). This makes a total of 155 Creatures in the National Assembly. As for the Senate, out of the 100 senators designated in 2018 (30 appointed by the President of the Republic and 70 elected by the municipal councillors), the presidential majority counts 93. If we add the 3 members appointed by presidential decree in this chamber's bureau also, we reach a total of 96 Creatures in the Senate.
According to the 2008 and 2019 administrative organization laws, there are 10 governors of region, 58 divisional officers, 360 sub-divisional officers, 360 mayors of municipalities, and 14 mayors of cities. Following the February 2020 elections, the RDPC obtained 316 municipal mayors and all the city 14 mayors. This makes a total of 758 local and regional administration officials "created" by the President of the Republic.
All in all, there are currently 1536 Creatures. More than three quarters are men and more than three fifths are entrenched outside the capital city, Yaoundé, where they spend most of their time. Only about 90 of them are not RDPC registered members, mainly because they continue to be active in the military or in the justice system, in spite of the formal but rarely enforced interdiction to do so.

Methods
Once again, our argument is that the Creatures are the articulatory nodes of the Cameroonian clientelistic structures and play a crucial triple role: (1) they mobilize the local and regional networks for the elections, (2) they insure the party discipline and allegiance to the central decisions, and (3) they act as negotiators between the centre and the local stakeholders.
In order to check the validity of our arguments, we organized four focus groups. The first (FG1) was made of former members of nine Creatures' staffs and was meant to allow us to penetrate into the functioning details of the Creatures as nodal points between the territorial networks and the President's inner circle. We mainly focused on the mechanisms of transmission of the President's mobilizing message to the local and regional actors within the vertical clientelistic structures which they belong to. The second (FG2), the third (FG3) and the forth (FG4) focus groups assembled regional and local activists of the RDPC from three different regions: his stronghold of the South (the Beti ethnic region, where Biya made 92% in the 2018 presidential elections)-FG2, with nine participants; an overall pro-Biya region (the West) but with a strong pro-opposition movement (the Bamileke ethnic region, where he made 62%)-FG3 (eight members); and the opposition stronghold of the Littoral (a mix ethnic region, where he came second with 35% after the opposition's leader Maurice Kamto)-FG4 (seven participants). The purpose of these three FGs was to test the perceived relevance of the Creatures' influence on the electoral mobilization of the local structures and to evaluate the perception of the regular militants with regard to the Creatures' roles as discipline-makers and President's interests' representatives.
The duration of each FG was between 2h1/2 and 3 h and each FG was co-led by the two co-authors of this research paper. The randomized selection of the FG participants was variably problematic. For the FG1, we tried to keep the highest degree of diversity, in terms of gender, language, ethnicity and seniority in service both in what concerns the members of the FG and the profiles of the Creatures that they served before. For the other FGs, we tried to balance the imperatives of diversity in representation with the need to proportionally reflect as much as possible the regional specificity of the RDPC's sociology. For instance, we introduced more urban militants and more women in the FG3 than in the FG2. One problematic aspect was related to the discrepancy between the willingness of the Southern region RDPC activists to participate to the FG2 and the hesitations of those of the Littoral to join the FG4, which were obviously explainable by the important differences in the degree of popularity of Biya and of his party in the two provinces, as Biya belongs to the Beti ethnic group who is dominant in the South and with whom the Bamileke, more numerous in the West, hold a historical rivalry (Onana 2005). Almost all the participants insisted on the need to remain anonymous and some of them asked us to change the initial letters of their names in the research reports so as they could not be retrieved by the party's watchers. Thus, we changed all the initials of the names of the FG participants and in this way all the participants accepted to be audio-recorded and to be quoted. In terms of compensation, we offered symbolic gifts and one lunch/dinner to all the participants.
We used a two-part scenario: the first part was common to all the three FGs, while the second was specific to FG1 and, respectively to the last three FGs, targeting relevant issues, such as the particular understandings of the Creatures' status and the real impact on determining the electoral behaviour of the local electorates or the capacity of the FG participants to mobilize resources and voters. For the FG results' analysis, we adapted the methodology proposed by John M. Creswell (Creswell 2014). We resorted to a two-layered analysis-that of the individual views, observations and motivations, on one hand; and that of the groups' dynamics and its influence on the individuals' opinions on the issues in discussion, on the other hand. The research focused on issues such as the strength of individual beliefs and opinions on the Cameroonian political developments (e.g. the new candidature of the incumbent president), the reasons for correlations between analysis and behaviours (e.g. backing the President's campaign), the articulation mechanisms of different political topics or events and the propensity for changing or reinforcing the participants' views on the main topic of the research (e.g. the reinforcement of the opposition's platforms and its incapacity to gather in a single counter-candidature against Biya). The adaptation we made consisted in a stronger insistence on the individual level, which is more relevant for depicting the reasons behind the members' own understandings of the political situation and their participation to Biya's campaigns, including in some occasions active lobbying and direct conditioning of the future financial support for certain field agents and elected officials. And, unlike Creswell's methods, we took less into account the focus groups' dynamics, except for those situations in which the participants lacked a precise opinion on certain relevant aspects and the FG experiment made them form some views as a result of their interactions with the other FG participants.

The Cameroonian general context in brief
In addition to the reasons that justify picking Cameroon as a case study, we will briefly expose here some relevant elements about this country's societal and political framework. First, there is an overwhelming influence of the precolonial traditional society, structured around chiefdoms which play a quasi-official role of administrative assisting bodies. There are 80 first-level chiefdoms in the country, 862s-level chiefdoms and innumerable third-level chiefdoms which break down into neighbourhood chiefdoms. The chiefdom has the authority to render traditional justice especially for land and civil affairs, including inheritance (Hamani 2005). It influences political practice, especially with regard to the perception of power. Indeed, the idea of the verticality of power and presidentialism is not only a colonial legacy (Kakdeu 2015). In ancestral society, a powerful leader had to show manliness at the risk of being overthrown; he was the custodian of all powers: judicial, military, legislative, executive and mystical. During his enthronement, people used to collect the various mystical forces of the kingdom and offer them to the chief (Owona 2015). Thus, after taking power in 1982, Paul Biya had toured the country to be initiated and ended up by being named "Nnom Ngui" [Chief of Chiefs] in his region of origin (Kpwang 2011, pp. 51-68).
Secondly, Cameroon has a centralized and hierarchical administrative organization. It goes from the traditional chiefdoms to the regions passing by the subdivisions and the divisions, respectively. The country has 360 subdivisions, 58 divisions and 10 regions. Regions are under the authority of governors appointed by the President of the Republic who also appoints sub-divisional and divisional officers. The designation of a Chief is done under the supervision of administrative authorities. Therefore, all administrative officials respond to the President of the Republic despite the existence of an official hierarchy among themselves. This act of appointment gives to the President the power of 'creation'. At the same time, the process of decentralisation, initiated in 1996, has been particularly slow and inconsistent (Kakdeu 2015).
Thirdly, Cameroon is affected by the economic precariousness that drives citizens to compromise. Given the underemployment rate of people aged 15 and over at 69.6% (INS 2016, 147) and a poverty rate of 37.5% in 2014 (INS 2016, 66), the battle of social climbing is tough. After an initial period of prosperity based on the high raw material prices (1965 to 1985), the economic situation has sharply deteriorated. Nowadays, the economy is dominated by small farms which have grown by exporting crops since colonial times. The average income of a small family farm is 300,000 FCFA (about 458 euros) per year. In this context, it is common for citizens to exchange their votes or their activism for money or a promise of social climbing (Vicente and Wantchékon 2009).
Finally, opposition parties are weak and political transhumance is high. In 1990, most African Francophone countries started their democratization process to sustain their relation with the North (Lynch and Crawford 2011, pp. 275-283). In Cameroon, despite the emergence of pluralism, President Paul Biya and his party succeeded to continuously remain in power, securing and even consolidating his majority in Parliament after the sole tight election that took place in 1992 (Nkainfon Pefura 1996). The public administration is politicized and openly campaigns for the RDPC, while the entire electoral apparatus is under its control. Biya appoints the members of the electoral commission and the members of the Constitutional Council in charge of the settlement of electoral disputes and the proclamation of the results. In 2018, after an unfair vote (Jeune Afrique 2018), he was re-elected with 71% and few weeks later, the main opposition candidate, Maurice Kamto, and his party's staff have been accused of civic insubordination and arrested.

The focus group results
The FG research allowed us to test the main arguments and then to draw some relevant conclusions concerning the conditions of its validity. Once again, our idea was that the FG groups would reveal the triple role of the Creatures: (1) to mobilize the local and regional structures for the elections, (2) to insure party discipline and allegiance to the decisions of the President, and (3) to act as negotiators between the political centre and the local stakeholders who mobilize the voters.

What is a creature?
To begin with, it is important to assess the FG participants' understandings of the notion of Creature. In fact, there was a high degree of consensus over the existence of the Creatures, about the empirical validity of this notion and a relatively high consensus about the Creatures' characteristics. The members of FG1 were naturally much more convinced that the label 'Creature' was fit to symbolically describe the status and the features of Paul Biya's key-appointees. Αs one of the FG members observed, "not all the prominent leaders of the regime are Creatures" (Y.N., former minister's chief of staff). Three criteria seemed essential: to have a "certain seniority, but not comparable to the one of Biya himself", "to be directly appointed or sanctioned by Biya" (S.N., former adjunct press officer) and to "notoriously depend of the President's will to be kept in office" (Y.N., former minister's chief of staff). According to the FG1 participants, these characteristics are generally known not only by the party members, but also by those who pay attention to politics (journalists, officers, academia etc.), while the notion of Creature itself is "well known and spread, even among the public opinion" (J.-L.K., former councillor of a state undersecretary).
On the other hand, some members of FG2 and FG3 observed that the notion of Creature "should not necessarily have a negative connotation" (G.B., party activist in the South). As the President is the Father of the Nation, being his Creature is, on a contrary, a "sacred unction", which consequently imposes to its bearer a "holly duty" (A.A., party local vice-leader in the West). Because of the imperfections of representative democracy and because of the "unpreparedness of the People to make rational and valuable choices" (P.N., regional leadership member in the South), Cameroon needs a system of Creatures and this "system shows its efficiency".
Being also fully aware of the notoriety of the concept of Creature, some of the FG4 participants, while admitting that they militate in a rather politically hostile environment (especially in Douala, the second largest city in Cameroon and an opposition stronghold), believe that the proliferation of the idea that the RDPC is led by some "presidential super-appointees" (the Creatures), while the opposition leaders are widely elected by the citizens of the region, "hinders the capacity of the local party organizations to locally entrench the President's discourse" (Y.S., party local official in Littoral). Consequently, while not denying the existence of the Creatures, they would prefer them to "vanish" from the regional public conscience, as the "effects of their presence are sometimes contrary to their intentions" (M.E., RDPC regional communication advisor in Littoral).

The creatures as agents of structures' articulation
We had the opportunity to analyse and check the role of the Creatures within the clientelistic structures as sources of electoral mobilization through a series of specific questions and also through the debates taking place within the focus groups. A trend towards the inflation of the role and influence of the Creatures could be observed in the case of most FG1 and FG2 members. This could be explained as there is a correlation between the FG members' own importance (as former collaborators of the Creatures or as their local/regional counterparts) and the Creatures' personal importance and relevance within the establishment. On a contrary, most FG3 and especially most FG4 participants were inclined to underestimate the influence of the Creatures and to augment in this way their own degree of autonomy and their own proximity to the local decision-makers and voters.
The structural role of the Creatures in the articulation and in the functioning of the clientelistic structures did not make the object of a consensus. Three of the FG1 members observed that the Creatures were able to use their key positions to "promise a specific allocation of resources in the immediately post-electoral periods" (G.T., former advisor of a ministerial director). Were such promises realistic? The FG1 semi-contradictory debates finally led to the consensual analysis according to which there were at least three situations: (a) the Creatures had (some) real influence over resource distribution and tried to keep their promises by exerting influence in favour of their local and/or regional clientele; (b) the Creatures presumably had some real influence but deceived their clientele and did not honour the promises made in exchange of electoral support; (c) the Creatures had no influence on resource allocation and their promises proved to be vain. Nevertheless, these postures were far from being clearly defined. As one of the FG1 participants said, "it was sometimes difficult even for us to understand in which of the three situations the Creatures (and implicitly us) were" (Y.N., former minister's chief of staff). Regardless of their intentions and capacities, numerous participants of all the FGs concluded that the Creatures were the only high-ranking officials capable to fuel and to manipulate the patron-client relations in Cameroonian politics.
At the same time, seen from the party militants' perspective, the situation seems to be equally complicated. As a local regular activist observed, "While you generally know that the chieftain of a local community controls the votes of his people, you can't be sure that he would keep his word" (M.D., militant in the West region). In what respect could he not keep his word? The FG3 members observe that the main strategy of 'deception' does not consist in promising 'his' votes to the RDPC and in fact giving them to some other candidate, "as this would be very easy to notice and to later punish" (O.O., deputy local leader in the West region). Instead, to promise a favourable vote for Paul Biya or for the party and then to stay home the day of the elections is a much safer and more subtle strategy. So, if fear of repercussions seems indeed to be high, local communities' credulity is overrated: the response to the sometimes limitless promises of the Creatures and of their local party militants is rather a conditioned kind of loyalty practiced by some local communities.
On an even more pragmatic note, a FG4 participant quoted the former Prime Minister Achidi Achu when saying that "politics na ndjangui [politics is a tontine]" (W.M., member of a local party committee in the Littoral). In the logic of the tontine, only those who contribute are entitled to profit (Tcheuyap 2014, p. 219). This is the rationale of the give-and-take practice of clientelism: the voters have an interest in voting for the President in power as he has more means to reward their loyalty than the opposition candidates. But the tontine could also work the other way around: if not satisfied, if chronically deceived in their expectations, the voters would rather boycott the election, an attitude that is difficultly punishable by the governmental party officials, as " […] in Cameroon, numerous reasons can always be invoked for not going to vote, from floods to superstitions" (B.P., former staff member of a state secretary). In this way, the local chieftains, the heads of family and the regular citizens inflict to the regime the only loss they can afford to inflict without retribution.

Overarching guardians and/or direct negotiators?
There was no common view among the FG participants on the answer to this question, but rather a fruitful debate resulting into some more or less consensual conclusions. During the discussions, there was a permanent tension between, on one hand, the need of some participants to underline the 'greatness' of the Creatures, as a direct emanation of President's will, and, on the other, to demonstrate their ability to talk with local influencers and voters and to persuade the latter to support the President.
The first image of the Creatures as distant masters of the puppets, too highly placed to involve in the 'vulgar' business of negotiation with basic political agents and voters, is spread among numerous FG2 participants. As an early RDPC member observed, "[…] even if they don't belong to the intimate inner circle of the President, they are to some extent his Chosen Ones" (T.K., local party committee member in the South). The Creature would be like "a Cardinal, spreading the President's message to us, checking the state of the affairs in the provinces and in the villages, giving us highly valuable hints on what to do and how to speak to the people" (A.A., party local vice-leader in the West). The Creature could be a negotiator only when he/she would need to "appease the local or regional conflicts" and to impose the "rule of law" over some "occasionally dissident factions". Some FG members brought several examples of Creatures (ministers, state secretaries, MPs, heads of the national police or general directors) who intervened in the name of President Biya to "restore the order" in several party local or departmental organizations where the clashes between the clientelistic local structures for the access to resources "was damaging the RDPC". In all these cases, the Creatures rather acted as guardians of the presidential rule over the country's territories and populations and as keepers of the system's reproduction via classical exchange mechanisms (votes for money) at the macro and median levels.
There is nevertheless a second image of the Creatures: they are top but "normal" political appointees who bargain over how much the government should spend " […] to mobilize the local and regional voters' owners in its favour" (D.L., former chief of staff of a state undersecretary). Some RDPC activists admit at their turn that the main expectation of the local party structures is to use the Creatures as assets to put the tribal chieftains, the neighbourhood clan leaders and the other local influencers "at work, to gather our voters" (K.K., local party treasurer, Littoral region). Given their experience, the Creatures "know that nothing is for free" and consequently "present what Yaoundé can offer in exchange of the political support" (T.P., regional vice-leader, West).
The FG3 and FG4 members exposed two examples of secret meetings between local chieftains, religious leaders, mayors, councillors, on one hand, and prominent Creatures, on the other hand. In both cases, the Creatures, who were obviously the stars of the meetings, initiated the discussions by trying to emphasize the inner tensions within those communities and their roles of "peacemakers" via the "heavy investments" that they will procure and that "would change the lives of the inhabitants" (W.M., member of a local party committee in the Littoral). In one of these two cases, as the local public seemed to be discontent of the ensemble of the promises, the Creature "[…] targeted each of the various group representatives who attended the meeting and indirectly offered him/her specific bonuses should they boost their voters' participation to the elections" (A.N., local party vice-leader in the West). These discussions included details such as the proofs that the local voters had to produce (generally by taking pictures of their ballots) and the measures of retaliation to be taken if the proofs were not conclusive enough or, worse, if there was some doubt about "cheating by not attending the poles or voting for other candidates" (A.N., local party vice-leader in the West). So, there is a fair amount of evidence to show that the Creatures act in some relevant occasions as direct negotiators within the core of the clientelistic Cameroonian structures.

Conclusions
The focus group research offered the possibility to check our initial arguments, to draw a series of meaningful but partial conclusions over the way clientelism works in Cameroon and to finally make a series of observations about, among other aspects, the would-be directions of future research on the key actors of clientelistic practices.
The first argument concerning the role of the Creatures as agents of clientelistic relationships' articulation and structures' mobilization was widely validated throughout the discussions of all the four FGs. Although aspects such as the opportunity of their interventions, the consistency of their promises or their capacity to satisfy their clients' demands remained to some extent controversial, there was a general consensus that they benefited of a special status ('the Chosen Ones') and that they fulfilled a key-role as nodal points in articulating the clientelistic relations within and around the presidential party.
Further on, our argument that the Creatures are perceived as the guardians of the President's will, imposing party discipline and allegiance to the decision-makers at the top level was only partially checked. As Paul Biya keeps signing decrees of nomination and rather commutes than fires the former holders of the respective positions (Bigombe Logo and Menthong 1996), there has been an inevitable inflation of Creatures and consequently a relative devaluation of their individual roles. But this did not diminish the well-entrenched behaviours of the regional and local party organizations, which are based on fear of control and traditional respect of hierarchies; thus, the Creatures are still perceived as the President's 'Cardinals', being "the last ones you have the interest to upset" (U.I., regional party committee substitute in the South).
Finally, the last assumption was that the Creatures were also negotiators between the central resource providers and the voters' local influencers. The research allowed us to validate this presupposition and to observe in addition the Creatures' chameleonic nature: in function of the needs, they act either as the all-mighty President's satraps, who inforce his will and keep the party's discipline, or as salesmen of the government's promises, being occasionally obliged to directly negotiate the prices of the resources they are willing to trade for a certain number of votes in favour of the President.
From a more general perspective, this research highlights some specific features of clientelism in semi-authoritarian regimes, at least in the Sub-Saharan African context. The uninterrupted continuity in power provides either the Presidents or the presidential dynasties with a variety of tools not only to control the power fluxes but also with a need to periodically refresh their networks of political-electoral influence. This explains the emergence and the development of the Creatures as a particular instrument of control and influence of the President and, as we saw, as agents of structural articulation for clientelism. This latter hypostasis of the Creatures opens for a wider possibility to analyse the clientelistic systems not merely as varieties of patron-client relations based on open negotiations but as complex pre-settled systems of domination where the extraction of the electoral support is rather hierarchic and often authoritative, leaving a limited marge of manoeuvre to the grass-root actors (as also implied by some comprehensive researches, such as the one conducted by Yonatan Morse (Morse 2018).
In the light of our present inquiry, further research could concentrate on two important aspects. It could question the similarities and the differences in the way the clientelistic structures are built in relation with the central (semi-)authoritative power, exploring, for instance, the strategies of the voters' local influencers to gain more autonomy in spite of the rigidity of the centralized control of the party in power (as also partially suggested in Leon 2011). Other investigations could cover the existence, the nature and the functions of the Creatures (or of their respective correspondents) in some other clientelistic systems but the Central African ones. This could enable us to determine the general processes of emergence of the intermediate agents in the clientelistic chains and to better map in this way the concrete types and modes of functioning of clientelism in different cultural and institutional traditions (as drafted in Schaffer 2007). Seen from this perspective, the Cameroonian case of clientelism could be used not as central-pivotal model but rather as an extreme yardstick that delineates the frontier between two kinds of mobilization: the semi-authoritarian but formally pluralist-electoral one and the dictatorial plebiscitelike variant of popular conscription.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.