Titel scriptie: Identity Politics and Political Trust in Nigeria: A study on ethnic affinities and trust in presidents in Nigeria
Naam: Daan Hendrik te Riele
Studentnummer: 10526420
E-mailadres: Daan@posteo.de
Opleiding: Politicologie
Namen beoordelaars: dr. M. (Mike) Medeiros (begeleider) & dr. R.S.B. (Roderik) Rekker (tweede lezer).
Identity Politics and Political
Trust in Nigeria
A study on ethnic affinities and trust in presidents in Nigeria
Name: Daan te Riele (10526420)
Subject: Bachelor thesis / Politics of Identity Supervisor: dr. M. (Mike) Medeiros
Second reader: dr. R.S.B. (Roderik) Rekker Date: 06 July 2018
Word count: 9331
Abstract:
Although the role of ethnic identities has been center stage in much research on politics in sub-Saharan African countries, few quantitative studies have researched the role of ethnic identities and individual political trust in a direct way. This is remarkable as political trust is regarded as central to effective governance and democracy. Qualitative research has
identified that in many countries on the African continent, trust in the political center is lacking. Quantitative studies trying to investigate these relations in depth and country-specific are however scarce. This research will answer the question: does ethnic affinity to the
president affect trust in the president? It will do so by focusing on the case of Nigeria since this is a country where ethnic identities (including religion and region) are believed to play an important role in politics and thus allow us to investigate claims on the complex dynamics of ethnicity and political trust. Using the two survey rounds of the Afrobarometer covering two different presidents from different ethnic, religious and regional backgrounds allow to get a bottom-up perspective whether ethnicity matters in political trust. This research finds that ethnic affinity does matter but fails to consistently explain levels of trust in both presidents (Yar’Adua and Goodluck Jonathan). This finding has important implications for our
theoretical understanding of (ethnic) identities in the political sphere but also give insight on how to increase political trust in the political center, and thereby its legitimacy.
Table of Contents
Introduction ... 1
Conceptualization ... 3
Literature Review... 5
Data and Methodology ... 9
Results ... 15
Table I ... 21
Table II ... 22
Conclusion and Discussion ... 24
1
Introduction
“Nigeria is not a nation, it is a mere geographical expression. There are no “Nigerians” in the same sense as there are “English” or “Welsh” or “French”. The word Nigeria is merely a distinctive appellation to distinguish those who live within the boundaries of Nigeria from those who do not” (Awolowo 1947).
This quote is from Obafemi Awolowo (1909 – 1987) who played a central role in Nigeria’s independence movement. It reflects the historical roots of the tensions and
challenges the nation-building process of Nigeria has witnessed. Nigeria is a country where identity-politics is a recurrent theme in its political history (Falola & Heaton 2008; Lewis 2007). But Nigeria is not a case on its own, since many countries in sub-Saharan Africa and beyond have seen identity-politics and questions of ‘’belonging’’ as a central political theme (Davidson 1992; Schulte Nordholt 2008). In this thesis the focus is on the relationship between identities (ethnic, religious and regional) and trust in the political center in Nigeria.
When studying politics in African countries, there seems to be little specific attention to the study of political trust in a quantitative way (but see Hutchison & Johnson 2011; 2013; Linke 2013). Especially rare are those studies looking at individual political trust and ethnic identities using survey data. This seems to confirm the observation by Karakoç (2013) that the relationship between ethnicity and political trust is an understudied topic. The relative absence of the study of political trust from a bottom-up perspective is remarkable, since many studies have noted that the absence of political trust can have negative consequences on the effective functioning of the state and its institutions (Karakoç 2013: 93; Hetherington 2006; Marien & Hooghe 2011). By looking at the relation of ethnic affinities and trust in the political center, this research tries to fill this gap and add to a better understanding of the ways ethnic identities influence perceived trust and legitimacy of presidents in Nigeria. Nigeria is a country where politicized ethnic identities are deemed to be central (Lewis 2007: iv; Falola & Heaton 2008). Moreover, it enables to see whether the common conceptions of African politics as ethnically motivated are still relevant (Bratton et al. 2012). To study this, the case of Nigeria will be used to answer the question: does ethnic affinity increase trust in the president in Nigeria?
A lot of research focusing on ethnic identities and political trust in sub-Saharan Africa and Nigeria is qualitative (see for example: Berman 2004; 2013; Ekeh 1975; Nnoli 1995). By
2
conducting a quantitative study on Nigeria, this research can complement this set of literature by looking whether some of the observations by these authors are still valid. Most
quantitative literature studying ethnicity in a more general fashion have taken a cross-national approach. Although these studies give valuable insights on a wide-variety of cases, and I certainly do not wish to reject the validity of large-N studies, it is also problematic. The reason being that it fails to shed light on the deeper dynamics and complexities of individual countries because its research design loses a lot of context. Studying these deeper dynamics does not only need to be through qualitative accounts. Using a case study based on surveys -as this thesis does- will also allow to do this. Next to this, most research, especially in economics and conflict studies, uses the variable of ‘ethnic fractionalization’ to measure the effects of ethnic diversity on societies (Alesina et al. 1999; Collier 2001; Collier & Hoeffler 2004; Easterly & Levine 1997). This does not tell us much on these identities and its role in specific societies as it merely uses ethnic and cultural diversity as an independent variable. Rather, it runs the risk of leading to simplified and essentialist conclusions on the effect of ethnic diversity in Africa. By adopting a case-study approach relying on survey-data, it helps to prevent falling in these simplified generalized characterizations. The use of the
Afrobarometer surveys give this opportunity to study ethnic identities and political trust from a more bottom-up perspective.
Besides adding to a better theoretical understanding of ethnic identities and trust, it can also give valuable insights in addressing the complex developmental and state-formation challenges in Nigeria, and perhaps beyond Nigeria. As already mentioned, political trust is central to the legitimacy of institutions and effective governance. Studying its determinants might illuminate some of the ways plural societies, like Nigeria, can create more legitimacy supporting effective governance.
This thesis is structured as follows. First, to better understand the arguments in this research, the central concepts: political trust, identity and ethnicity will be conceptualized. This will be followed by a discussion of the existing literature on (ethnic) identity and political trust in general and on sub-Saharan Africa and Nigeria in particular. Based on this literature review and Nigeria’s political history, this thesis formulates a set of hypotheses expecting that ethnic affinity to the president in Nigeria will increase trust in the president. With sub-hypotheses for the different affinities: ‘tribe’, religion and region. Third, the research method and variables used to test the hypotheses will be discussed. Fourth, an ordinal logistic regression will be conducted followed by an interpretation of the results. Last,
3
the main conclusions of this thesis and its implications will be discussed. Here some of the limitations of this thesis and possibilities for further research will also be highlighted.
Conceptualization
Political trust in this thesis is understood to be citizens’ confidence in political institutions (Turper & Aarts 2017: 417). The degree of political trust is an important indicator of the amount of political legitimacy a specific actor or institution in question enjoys (ibid.). When one measures political trust, it can be understood to measure the degree to which people perceive their government producing outcomes consistent with their expectations that in turn influence their confidence (Hetheringhton 2006; Karakoç 2013: 93). The basis for these expectations that result in increased trust are numerous and vary including social identities, policy performance and evaluations of economic performance (Uslaner 2018). In this research, the relation between ethnic identities and political trust is central.
A political (group) identity is ‘’a person’s sense of belonging to a group if (it) influences his [or her] political behavior’’ (Erickson 1968: 57). Political behavior can be actual action (such as voting or protest) but can also be in the form of attitudes and trust (ibid.). To study political identities and trust, this thesis works from a theoretical
understanding of political identities on what I term: a ‘’soft-instrumentalist’’ position. It is instrumentalist in its assumption that political identities serve as a mean to gain access to the centers of political and economic power (Jega 2000: 15; Muro 2015: 6). However, I would call it ‘’soft’’ because it allows for constructivist accounts of identity that see identities shaped by social processes and stress the importance of context influencing the resonating force of certain social identities (Muro 2015). Which identities inform political behavior involve the identification of people and their own perceived functioning in the political arena (bottom-up), as well as the active role of elite’s in actively manipulating and using these identities within the polity (top-down). This understanding means that political identity is fluid and can changes over time (Muro 2015: 6). This notion of fluidity is important because this thesis looks at the influence of ethnic affinity influencing trust in the president and whether it changes among different presidents. Adopting this position allows for a more open-ended approach to the functioning of social identities over adhering to only one
theoretical view. By conducting this research, it can give us valuable theoretical insights as it can show the degree of fluidity of social identities in determining political trust in Nigeria. Moreover, doing this can also shed light on the primordial understanding that sees ethnic
4
categories as fixed over time and consistent in determining its influence, thereby being more sensitive to the historical embeddedness of politicized identities (idem: 5; Nnoli 1995).
A third important concept used in this thesis is that of ‘ethnicity’ (or ‘ethnic identity’). Within social science literature the meaning of this concept is debated (for an overview of this debate see: Hale 2004). This thesis works from a broad conception of ethnicity consisting of people’s personal points of reference, as people use various social categorizations to understand complex situations in society (idem: 473). Belonging to an ethnic group means that one ascribes him or herself (or is ascribed by others) to a particular group on the basis of their distinct differences that the members of the group and other ethnic groups see as
significant to their identity (Adetiba & Rahim 2012: 658). Such an understanding is useful as this paper relies on self-identified measures of ethnic identities where people ascribe
themselves to different categorizations. Ethnic group categories involve the active
maintenance of boundaries between in and out-groups and are therefore actively constructed and malleable (ibid.). These boundaries are not purely biological but reflect a socio-political identity that can be based on social identities such as religion and region. Therefore, this thesis understands ‘ethnic identities’ to also encompass religion and region when they are considered important points of reference for group formation. In the case of Nigeria, these two identities are said to be important socio-political identities that influence political behavior (Adetiba & Rahim 2012; Falola & Heaton 2008).
Why then study political trust and ethnic identity? As has already been said, the importance of investigating the relation between ethnic affinity and trust is first and foremost because political trust is a crucial element when it comes to the legitimacy and effectiveness of states. Authors have identified trust to be important in fostering regime stability, effective governance and support for democracy (Karakoç 2013: 93; Hetherington 2006; Marien & Hooghe 2011). Especially in plural societies (Nigeria has more than 200 different ethno-linguistic groups, and three major ethnic groups, of which neither holds an outright majority) we can argue that having a more generalized trust would contribute to a better functioning system rather than only a particularized trust, where people have only a strong feeling towards their own ethnic or religious group. More generalized political trust enables collective action and leads to more acceptance of political opponents once they hold power (Berman 2010: 4).
5
Literature Review
Concentrating on an African context, it has been argued that many African nation-states lack a certain degree of trust and attachment resulting in: “a civic public in Africa [that] is amoral and lacks the generalized moral imperatives operative in the private realm and in the
primordial public.’’ (Ekeh 1975: 92). The common people in many sub-Saharan African countries, Ekeh argues, lack trust in the central state which we might observe in other societies such as in the West (ibid.). This lack of moral imperative is however not static, as the word ‘’primordial’’ might imply. It is a direct result of colonial rule and its institutions, as well as the post-colonial African elites that uncritically adopted and perpetuated the imposed colonial structures (e.g., the nation-state) and institutions (ibid.).
Davidson, in agreement with Ekeh, argues that the adoption of the nation-state in many post-colonial sub-Saharan countries has led to an increased alienation on the part of wide sections of the populace (1992). This led to the emergence of a form of politics where patronage networks are central and identities are actively politicized by elites to secure resources in the political center (Berman 2010; 2013; Eifert et al. 2010; Eze 2010; Green 2017; Posner 2007). As a result, Berman (2004) concludes that what is missing in many African states is a form of collaborative trust, a civic trust that transcends ethnicity and patron-client relations. The political center is only seen as a ‘cake’ from which one should eat and feed its own group (ibid.).
The salience of ethnic identities, according to Berman, is believed to be an important element in trust in the state (ibid.). Ethnically based patron-client networks have been extended to the very center of the state apparatus, where the sate-society relationship is opportunistic such as in Kenya or Nigeria (idem: 47). This seems to confirm an
instrumentalist account of identity in African politics as ethnicities are used in these patron-client relationships. Other research studying political behavior seems to support these characterizations above. For example, it has been shown that having a co-ethnic president in power increases the access to resources for specific ethnic groups (Ahlerup & Isaksson 2015; Kramon & Posner 2016). Focusing on voting behavior and party systems, more structural theories of political behavior claim that in ethnically segmented societies, ethnic identities such as ‘tribe’ or religion form a core basis of political action (Horowitz 1985; Norris & Mattes 2003).
6
Based on these characterizations, I expect that ethnic affinity will thus also significantly influence trust in the president. This leads to the hypothesis:
H1: Ethnic affinity to the president will significantly increase trust in the president.
The identified importance of ethnic identities in an African context makes it worthwhile to investigate the links between ethnic identities and trust in the political center. Especially plural societies that have been said to be characterized by strong cleavages and contestations between different groups, such as in Nigeria. In the following, the case of Nigeria will be presented from which I will propose three hypotheses that serve as the basis for the research in this thesis.
Nigeria
There seems to be a common agreement that ethnic identities also matter when explaining Nigerian politics and its history as a nation-state (Falola & Heaton 2008; Jega 2000; Eze 2010; Paden 2005; Nnoli 1995; van Peteghem & Limpens 2011). Authors have described Nigerian politics as a form ‘winner-takes-all politics’ where the political center is regarded as a locus of domination and inter-group relations are characterized by a ‘fear of domination’ (Falola & Heaton 2008: 142; Eze 2010; Ekeh 1975; Lewis 2007; Nnoli 1995; Sklar et al. 2006). This fear of domination has been most prominent among the three major ethnic groups: the Hausa-Fulani, Yoruba and Igbo making up 29%, 21% and 18% percent of the total population respectively (CIA World Factbook 2018). But also, the smaller ethnic groups such as the Ijaw (10%) and Kanuri (4%) (ibid.). In total, Nigeria is believed to consist of over more than 200 different ethno-linguistic groups (ibid.).
The roots of these intergroup-relations are said to be fueled mainly by elite
competition that find its roots in Nigeria’s colonial era (Ekeh 1975; Eze 2010). The Nigerian polity and nation-state that we know today is not a result of an organic process but an
artificial colonial creation by the British. In the year 1914, the so-called Northern and Southern ‘protectorate’ were forcibly combined into creating modern-day Nigeria (Falola & Heaton 2008: 110), combining very different cultural life-worlds. Next to this, the colonial period was characterized by a divide and rule policy under the banner of ’indirect rule’ (ibid.). This translated in the active politicization of ethnic groups and the creation of ‘tribes’ where every ethnic community was believed (and made) to be governed by a ‘chief’, making
7
the community lines rigid and dogmatic (Eze 2010:74). This colonial politicization of ethnic identities is important because it formed the basis of the many struggles of nationalism in Nigeria as lines between ethnic communities became rigid and came to play an important role in the tensions of fostering a national identity in Nigeria between different groups (ibid.).
The active manipulation of these identities could however not have ‘’succeeded’’ with the active involvement of the indigenous Nigerian elite that increasingly used these rigid identities as a way to gain and sustain power (idem: 74). The uncritical continuation of the colonial nation-state structures at independence in 1960 meant that there was an active tendency to centralize the Nigerian state and push for a single national identity. This seemed to work as long as the different communities in Nigeria could unite in opposition to their colonial rulers. But with the departure of the British, ethno-regional and religious fissures came to the surface and the national question became (and still is): what is the Nigerian nation? (Falola & Heaton 2008; Jega 2000).
So, the post-colonial history of an invented Nigeria was based on a strong distrust and competition between elites. The distrust between the different elites in Nigeria, especially between the three major ethnic groups: Hausa, Yoruba and Igbo, resulted in the active usage of ethnic identities to gain power to the political center. The first pan-national movement in Nigeria: the Nigerian Youth Movement (NYM), founded in 1934, split apart due to the suspicion of ethnic domination between the Igbo and Yoruba in particular (Sklar 2004). This distrust resulted in that the Nigerian state around 1960 could be described as: ‘’a state without a nation’’, with people feeling almost no allegiance to the political center (Falola & Heaton: 159). This all meant that the political center (especially national) in Nigeria is as a locus of contestation and rupture. Two examples illustrate this. First, to the most extreme, it led to the Nigerian civil war (or Biafran war) in 1967. This was a direct consequence of the regional and ethnic fissures (Falola & Heaton 2008; Nnoli 1995). After the first military coup in 1966, the Igbo actively sought to secede from the Nigerian state out of fear of domination by other ethnicities and groups. This ultimately resulted in a bloody war that took more than one million lives. Here the fear of domination and distrust took a major price and left a big mark on the distrust between different ethnicities. A second example that shows the fear of distrust and domination is the creation of the many federal states in Nigeria. At independence,
Nigeria was divided in three regions: (Yoruba) West, (Hausa-Fulani) North and (Igbo) East. Anno 2018, Nigeria has thirty-six different states that have partly been created as a reaction to the fears of smaller ethnicities on the domination of the three major ones (Sklar et al.
8
2006). These historical paths show that trust has been undermined between the different ethnicities.
The salience of religion (between a largely Muslim north and Christian South) as a line of contestation has been identified as increasing in intensity in Nigerian politics (Paden 2005; Eze 2010). The most recent elections in 2015 between Muhammadu Buhari and Goodluck Jonathan have been framed by many people as a contestation between Muslims and Christians, represented by Buhari and Goodluck respectively. Paden concludes that: “ethnic identity may turn out to be less salient than religious in identity in the Fourth Republic1” (Paden 2005: 209).
Some authors argue that when it comes to the salience of ethnicity in the political scene there is something of a path-dependency, or ‘’a threshold of irreversibility’’ (Nnoli 1995: 21). Claiming that there are critical moments in the history of states that are
determinant of the future of inter-ethnic tensions and can lead to the ‘’freezing’’ of ethnic boundaries (idem: 31; 46). This ‘primordialist’ reasoning leads to suspect that ethnic affinity will influence the amount of trust people have in the president. In combination with the lack of generalized trust among elites in Nigeria, it leads to suspect that ethnic affinity is an important measure determining political trust of the political center has, as measured by the president. Because the president in Nigeria is the most powerful political executive position. By focusing on self-reported ethnic identifications, it can be detected whether the ethnic identities indeed do play such a pronounced role as argued by earlier research. In addition, it can also add to the debate whether the identity politics argument on the relevance of ethnicity and religion is overstated, as some authors claim it is (Lewis 2007; Orji 2015).
Thus based on the literature review above, the three major ethnic cleavages within Nigerian politics can be said to consist of the following: 1) ethnic groups 2) religion 3) region. Based on these three cleavages and our central question, this thesis comes up with three hypothesis that will be the central focus of this thesis:
H1: Ethnic affinity with the president will lead to greater trust in the president. H1.1: Identifying with the same ethnic group (‘tribe’) as the president will increase political trust in the president
1 The Fourth Republic referring to the period of multiparty democracy that was restored in 1999 after the end of
9
H1.2: Identifying with the same religion as the president will increase political trust in the president
H1.3: Coming from the same region as the president will increase political trust in the president
Based on these hypotheses and the existing literature this research expects to find a set of inverted relationships for all three sub-hypotheses. This thesis expects to find a lower trust among people who identify with another ethnic group or religion compared to those who share an affinity in terms of tribe, religion or region with the president in power. How this will be studied exactly will be the focus of the next data and methodology section.
Data and methodology
To test the hypotheses in this research an ordinal logistic regression will be conducted using SPPS. The data used are two rounds of the Afrobarometer survey2. The benefit of using the
Afrobarometer survey is that it allows to study political phenomena from a bottom-up perspective as it asks individual respondents their opinions and evaluations. This makes the use of surveys an advantage over exogenous categorizations that have a stronger risk of bias.
This research will use the Afrobarometer rounds 4 and 5, conducted in 2008 and 2012/2013 respectively. The reason for choosing these rounds is because they reflect two different presidents in office in Nigeria3. President Umaru Musa Yar'Adua who was in office from 2007 until 2010 and president Goodluck Ebele Azikiwe Jonathan (Goodluck Jonathan), in office from 2010 until 2015. The fact that these presidents are from different ethnic, regional and religious backgrounds gives an opportunity to answer the central question whether ethnic affinity influences trust in the president.
President Yar’Adua (round 4) belongs to the ethnic Fulani ‘tribe’, mainly
concentrated in the North and is a Muslim. President Goodluck Jonathan (round 5), belonging to the ethnic Ijaw group concentrated in the South-South region ( ‘Niger-delta’), is a
2 The Afrobarometer is: “a pan-African, non-partisan research network that conducts public attitude surveys on
democracy, governance, economic conditions, and related issues in more than 35 countries in Africa”. In the case of Nigeria, it has conducted surveys in the years: 2000, 2001, 2003, 2005, 2007, 2008, 2012 and 2017. (Afrobarometer.org 2018)
3 The most recent Afrobarometer survey (round 6) was published in 2016 and conducted in 2014 whilst
Goodluck Jonathan was still president in Nigeria. Therefore, comparison with the latest president - Muhammadu Buhari- is not possible. In addition, round 6 shows more differences in variables compared to earlier rounds complicating comparisons with previous rounds.
10
Christian. Thus, both presidents differ in the major cleavages said to be relevant to Nigerian politics: ethnic ‘tribe’, religion and region. The fact that they both belonged to the same People’s Democratic Party (PDP) seems to exclude a partisan element. Comparing both presidents will allow to see if people who self-identify with a particular ethnic identity change in their degree of trust when another president is in office that does not share those affinities.
This thesis only focuses on the case of Nigeria. The reason to do this has already been explained in the introduction. But to add, the choice of using a single-case study is beneficial in that it can give us a more contextualized and deeper understanding at the relations at play when compared to larger-N comparison (Landman 2008: 29). Many single-case studies have provided new concepts and ways of understanding important political phenomena (ibid.). The next section will introduce the set of independent variables, as well as the control variables, to study the relations between ethnic affinities and trust in the political center.
Variables
Dependent variable
One dependent variable will be used to test the hypotheses: Trust in the president. In order to get a sense of trust in the political center in Nigeria, this thesis will use the variable of trust in the president. The reason why trust in the president is chosen as a measure of trust in the political center is because the Nigerian president holds most executive power in the country and is the most powerful political figure in the state (Falola & Heaton 2008; Lewis 2007). The fact that this office is so powerful means that elections regarding the presidential office are periods of a lot of tensions. Not incidentally the election periods see large scale riots between different groups supporting the opposing presidents4. Furthermore, national elections (including parliamentary) are almost always framed by proxy of the president reflected in the attention of Nigerian media outlets and social media (Omotola & Adebayo 2007). Therefore, it seems justified to treat the office of the presidency as a measure of political trust in the center, at least for the case of Nigeria. Furthermore, because power is so concentrated in the office of the president, the political center becomes as a locus of
‘domination’ as already shown in the theoretical framework.
4 For example, in the heavy riots of 2011 among Hausa youth in the North expressing their anger over the loss of
11
The specific question used in the Afrobarometer dataset for this variable is: “How much do you trust each of the following, or haven’t you heard enough about them to say: The President?” (Afrobarometer codebook 2015). Respondents had three options in answering: 0=Not at all, 1=Just a little, 2=Somewhat, 3=A lot. As this dependent variable is measured on an ordinal scale, using an ordinal logistics regression is most suitable.
Independent variables
In order to test the relation between ethnic affinity and trust in the political center, this research will run three separate models each reflecting the major identity cleavages (ethnic group or ‘tribe’, religion and region). Because the risk of multicollinearity, as these identities can overlap significantly, these three cleavages will be included separately in three models. This allows to isolate and study some of the identities determining trust. The three
independent variables are:
1. Ethnic group/ tribe. In both rounds of the Afrobarometer survey people have been asked to identify themselves with a particular ethnic community. Asking them: “What is your tribe? You know, your ethnic or cultural group.” To which they could place themselves in a particular group5.
The variable has been recoded in such a way that it will aid the interpretation of the
regression models. Instead of including all 39 ethnic groups, the variable is recoded into: 1= Igbo, 2= Yoruba, 3= Rest, 4= Hausa-Fulani6. The reason why the Hausa-Fulani group are
grouped together is because these groups are closely affiliated within Nigeria, both socially (e.g., in religion) and politically (Falola & Heaton 2008: 159; Suleiman & Maiangwa 2017). Thus, grouping them seems justified.
For round 5, it is unfortunately not possible to measure the self-identification of ‘tribe’ and trust in the president (Goodluck Jonathan) based on the Afrobarometer survey rounds. The sample size of the ethnic Ijaw group to which the president belongs is too small to
5 Respondents had an option to also choose for: “National identity only, or "doesn't think of self in those terms”.
The fact that almost no one chose these last options indicates that ethnic identities play an important role in people’s lives. The distribution in both rounds seems to correspond to the actual distribution of ethnic identities increasing the validity of the sampling of both datasets.
6 The reason why the Hausa-Fulani group is chosen as the fourth category is because SPSS automatically takes
the last category as reference category when using an ordinal regression. Thus, it allows to interpret the results with the baseline being the Hausa/Fulani ‘tribe’. The Hausa-Fulani group is the ethnic group of the president Yar’Adua in 2008.
12
include in a model7 and doesn’t allow for a meaningful interpretation as the small N will
cause confidence intervals to be too wide. Thus, this research only able to measure the identification of ‘tribe’ and trust in president for round 4 (model 1 in table I).
2. Religion: Religion is a central feature is Nigerian politics and is believed to have increased in salience in political conflicts. Especially the cleavage between Muslims and Christians is thought to be of prime importance. Therefore, this research has used the nominal variable on religious identification of respondents and recoded this to create a binary variable between Christians and Muslims8. Because SPSS
automatically uses the last category of a variable as reference category when
conducting an ordinal regression, the reference category is Muslim. This allows us to see whether both religious groups have a significant change trust a other president of a different religion is in office. Based on the hypothesis H1.2 it is expected that we find an opposed relationship with Christians being less trusting compared to Muslims in table I, and more trusting in table II.
3. Region: The third independent variable used is that of region. As already mentioned, geography is also said to play important role in the discourse in Nigerian politics (Falola & Heaton 2008; Paden 2005). This is partly due to the fact that different ethnic groups (‘tribes’ and religious) are concentrated in different regions. The distinction between north and south and the fear of being dominated by either is a recurrent feature in Nigeria, going back to the tension since gaining independence (ibid.). The nominal variable for region in the Afrobarometer survey is composed of 36 individual states. For this thesis, the variable for region has been recoded into six geopolitical zones: South-South (0), South-East (1), South-West (2), North-Central (Middle Belt) (3), North-East (4), North-West (5). These six geopolitical zones have been created in the 1990s in Nigeria as part of an active centralizing policy and ease ethnic tensions in Nigeria (Sklar et al. 2006). The geopolitical zones are an actual classification in Nigerian politics. By recoding the 36 states into these six geopolitical zones it offers a good way to measure regional cleavages and come closest to
investigate whether the ‘’north-south discourse’’ can be observed in terms of trust.
7 N=48, making up only 2% of total number of respondents which are: N=2400
8 Strong diversity exists within these groups but based on the literature in Nigeria it seems justified to reduce the
categories to a binary variable: Muslim or Christian, especially when it comes to national politics (van Peteghem & Limpens 2011)
13
Based on the hypothesis H1.3, we would expect that in round 5, the other regions (especially the south-south) would show a positive relationship compared to the reference region North-West because of president Goodluck Jonathan from the south-south becoming president.
Control Variables
:
To understand if ethnic affinity determines trust in the political center, several control variables will be included that could potentially influence the relationship between the independent and dependent variables. But they are also used to test for alternative explanations.
Demographic indicators. This research uses a set of common demographic
indicators: age and gender in measuring political trust (Bratton et al. 2012; Eifert et al. 2010). Furthermore, geographical dispersion is controlled for by including a binary variable measuring whether the respondent is located in a urban or rural area.
Socio-economic indicators. In order to control for socio-economic determinants and
measure social-class two ordinal variables are used, education-level and employment
status. Using these variables allows us to see whether class distinctions play an important
role in determining trust in the president. For employment status it allows to control for economic dissatisfaction among the jobless.
To measure the degree of self-perceived poverty influencing trust in the president, this research uses the ordinal control variable: Your present living conditions. This variable is on a 5-point scale ranging from ‘’very bad’’ to ‘’very good’’.
Performance, economic rationale and freeness of elections. Performance of state
institutions is believed to be a major determinant of political trust (Uslaner 2018; Hutchison & Johnson 2011). Including the variable: performance of the president is important. This thesis expects that the perceived performance of the president will significantly influence political trust in the president. Performance of the president is measured on a 4-points scale ranging from negative to positive.
An obvious variable thought to influence political trust in the president is the degree
of corruption of the president. Corruption can logically be understood to erode political
legitimacy of institutions and offices (Uslaner 2017; Anderson & Tverdova 2003). Because of this, this research expects perceptions of corruption to be a strong determinant of trust in the presidential office. Therefore, this thesis includes a control variable which measures the
14
evaluation of respondent on the corruption of the president. This variable is measured on a 4-point scale ranging from saying the president is not corrupt to saying the president is totally corrupt.
As earlier research has argued that evaluations of economic policy performance are an important determinant of voting behavior in African countries (Bratton et al. 2012) a control variable measuring evaluation of the government’s (president) handling of the economy is included as a control variable to see in how far consideration vis-à-vis the economy matter. This variable is measured on a 4-point scale ranging from negative to positive on the president handling of the economy
A variable measuring the perceived free and fairness of last national elections is
also included because it can be believed that the perceived freeness of an election can influence the degree of legitimacy a president holds. If a respondent does not believe that elections were conducted in a reasonably just manner, this would translate in lower amounts of political trust in the president and thereby its legitimacy. This variable is also measured on a 4-point scale from: completely not free (negative) and fair until completely fair and free (positive).
Ethnic group treated unfairly. A control variable measuring whether the respondent
perceived their ethnic group to be treated unfairly is included. The reason is that we might assume that those persons who feel their ethnic group to treated unfairly treated will have a lower trust in the president. This variable is measured on a 4-point scale ranging from never to always.
Ethnic versus national identity. This control variable is worth highlighting because
we may assume that if someone feels more attached to national identity, then their ethnic or religious affinity might matter less in trust for president. A higher degree of national identity is believed to translate in more generalized trust and acceptance of political opponents and can so translate in political trust (Berman 2004; Berg & Hjerm 2010). Thus, this is a variable worth including as a control variable in this research because we can test whether feeling more attached to a national identity will have an impact or trust in the president compared to ethnicity. This variable is measured on a 5-point scale ranging from ethnic ID only to national ID only.
15
Results
Tables I and II provide the results for the ordinal regression analyses. In order to test the hypotheses for both survey rounds (corresponding with the two different presidents). All three models have been included in both tables9. Model 1 focuses on ethnic group/’tribe’, model 2 on religion and model 3 on region.
Looking at the results, what is most notable is that for the period when president Yar'adua10 was president, the tribal, religious and regional affinities all determined trust in a significant way (see table I). Based on this, it offers strong support for the hypotheses. However, when looking at the presidency of Goodluck Jonathan11 (table II), the ethnic
affinities do not seem to play a significant role anymore. In the following a deeper analysis of these results will be made by focusing on each ethnic affinity. After interpreting the results, the next section (conclusion and discussion) will elaborate on the implications of the findings.
Ethnic group / ‘tribe’
From table I we see that those tribes that do not share an ethnic affinity with the president are less trusting of the president when compared to the Hausa/Fulani group (to which the president belongs). This seems to support hypothesis H1.1 that belonging to the same ethnic group (‘tribe’) will increase trust in the president. The relationship is strongest for the Igbo and Yoruba group, who are the other two major dominant ethnic groups in Nigeria that historically have been in competition with each other. Identifying with these groups lead to significant lower levels of trust in the president. For the ‘rest’ group, which contains all the smaller minority groups, this relationship is weakly significant. This could be due to different factors: their small size making them unable to dominate other ethnicities or the heterogeneity and geographical dispersion across the country. The fact that we see a lower trust of Igbo and Yoruba compared to the Hausa-Fulani give reason to support the idea that there is a distrust between these ethnic groups, as earlier research has argued (Falola & Heaton 2008; Nnoli 1995; Eze 2010). The relation between the Igbo and the Hausa-Fulani reflected in table I indicates a sense of distrust. This in accordance with a sense of alienation among some Igbo’s that still strive for their own independent state of Biafra (the Vanguard 2018b). For the Yoruba’s, they are said to have a strong feeling of cultural separateness
9 With the exception of model 1 looking at the influence of ethnic affiliation by tribe in table II. The reason is
that the sample of the Ijaw group is too small making interpretation problematic.
10 A Muslim president coming from the North-Western zone and belonging tot the Fulani ethnic group. 11 A Christian president coming from the South-South zone and belonging to the ijaw ethnic group.
16
which has from time to time resulted in manifestations of ethno-nationalism (Ukeje &
Adebanwi 2008). The fact that we see a significant relation indicates that ethnic affinity does play a role at least for trust in president Yar’Adua. However, this research has been unable to measure whether this holds true for the presidency of Goodluck Jonathan (table II), who belongs to the ethnic Ijaw ethnic group making this interpretation limited to only being able to say something on round 4 of the data. Therefore, we cannot completely accept hypothesis H1.1.
A control variable included asked respondents whether they saw their ethnic group to be treated unfairly. From analyzing both tables we can see that this perception does not have a significant effect. Saying that your ethnic group has been treated unfairly does not result in a significant lower trust level.This might be due to the fact that the model also asks about the presidents performance and handling of the economy that seem stronger predictors than thinking your own group is treated unfairly.
When it comes to feeling of having a more ethnic or national, both tables show that this has no impact on trust levels of the president. Both tables do not show a consistent significant correlation that feeling ‘’less national’’ or ‘’more ethnic’’ results in less trust in either of the two presidents. This is an interesting result as some authors have argued that a more ethnic oriented identity will lead to less trust (Berman 2004). This seems to have a positive implication in that it might indicate that feeling more ‘’ethnic’’ should not necessarily inhibit more political trust in the president and political center questioning the negative connotation ‘’feeling ethnic’’ sometimes has in the literature (Green 2017). People might combine their ethnic identity with a more generalized trust.
Religion
Model 2 focuses on religion, coded as a binary category between Christians and Muslims. With Muslims being the reference category in both tables. Based on our hypothesis H1.2 this thesis expects that in table I we will see Christians to be significantly less trusty than Muslims and in table II: Christians to be significantly more trusting of the president than the Muslim reference group.
If we look at table I, Christians are indeed significantly more distrusting of the
president (Yar'Adua) compared to Muslims. This seems to correspond with literature arguing that there is a (increasing) distrust between Muslims and Christians within Nigeria (Paden
17
2005; Eze 2010). For table II we would then expect to see an opposite relationship, where Christians would be more trusting since they have a religious affiliation in the political center through the new president. This is not the case. This research finds that Christians, despite having a Christian president in office, will not lead to increased trust. This finding leads us to question the ‘fixedness’ of this social identity within the Nigerian political sphere. It seems to give support to a more constructivist account of identity in that the religious identity does play a role but that the salience in determining trust is dependent on possible other contextual factors such as political discourses or active manipulation by political entrepreneurs. It also contradicts more primordial accounts that argue that identity is fixed and relatively stable over time (Nnoli 1995). In the conclusion and discussion section will be a more extensive treatment of the theoretical implication these results have. At the same time, for both table I and II perceptions of performance remain strong predictors of trust in the president meaning that these evaluations can be said to be consistent for influencing trust in both presidents. So, although religious affinity does tell us something when it comes to table I, the fact that we do not see religious affinity to influence trust in both tables for each Muslim and Christian president leads to reject hypothesis H1.2.
Region
The third model focuses on how regional affinities with the president influence levels of trust in the president. In the same vein as the two other models, this thesis expects that regional affinity to the president will lead to more trust in the president (hypothesis H1.3). This expectation is on various authors that have described Nigerian politics to be characterized by a ‘’fear of domination’’ among different regions over each other, especially between the ‘’North’’ versus ‘’South’’12 (Falola & Heaton 2008; Sklar et al. 2006; Eze 2010: 69-87).
To measure the role of regions, a variable was created reflecting the official six geopolitical zones in Nigeria (see data & methodology). The north-western category is used as a reference category, as this was the zone president Yar’Adua is from. For table II we would expect that the region where Goodluck Jonathan is from (South-South zone) to show a significant positive trust compared to the North-Western region than in table I.
The results in table I point towards an existence of a dichotomy between the north and the south influencing trust in the president. Compared to the North-West zone in Nigeria,
12 These discourses on ‘north’ and the ‘south’ overlap with the various ethnic identities such as religion, where
18
people from the South-South, but also the South-East and South-West, are less trusting of the president (see Table I). The northern regions show a significant positive relationship. Again, based on these results from table I only we would conclude that regional affinity does play a role. But, as with religious affinity, this does not translate in an opposite relationship for table II when there is a president from the South-South in office. The results in table II do not show that there is a significant increase in trust for people from the South-South when compared to the North-Western region. This seems to go against the notion of a ‘fixed’ dichotomy
between the north-south. Other dynamics could be at play that determine the levels of trust that this research has not been able to focus on. The fact that people from the south-south are not more trusting when a president of their own regional affinity is in power shows that more performance-based elements could be more important, as they remain a significant predictor of trust in model 3 in table II. It leads to suspect that having the president ‘’deliver’’ seems more important. Moreover, a possible additional explanation could be that people from the South-South (which covers the Delta-region) are in general less trusting of the president and institutions in general (regardless of ethnicity) as they feel marginalized by the political center in Nigeria13. This might offer a possible explanation but remains speculation as this research has not been able to research this.
Based on these two tables together, it forces to reject hypothesis (H1.3) that regional affinity to the president will lead to more trust in the president. As with the earlier models, it does not mean that this affinity does not play any role (it does) but it is not fixed and consistent over time as it fails to significantly correlate with trust levels in both presidents.
So far, the ethnic affinities seem to only play a role in determining trust in the results of table I. This leads to partly reject all three hypotheses regarding ethnic affinities and trust in the president. We can only say that ethnic affinity plays a significant role for the
presidency of Yar’Adua. At a minimum, it leads to weaken the expectation on the role ethnic affinities play in trust in the political center in Nigeria. The implications of rejecting these three hypothesis has on notions of Nigerian politics as well as the theoretical understanding of identities will be the focus of the conclusion and discussion section. Before discussing those implications, both tables included a set of control variables in order to control for spurious correlations but also to investigate alternative explanations. To these control variables we will now turn.
13 The South-South zone is home to the vast oil reserves in Nigeria. People from these areas have felt that they
19
Results for control variables
For all three models in both tables a set of control variables have been included (see section data & methodology for a full overview). For the variable of age, this only has a significant correlation with trust in the president in table II, which could indicate that young people have lower levels of trust in the president Goodluck Jonathan. For gender, we so no meaningful significant relationship. For urban/rural distinction, this shows that urban people are less trusting than rural people only in table II.
Based on the literature14, this thesis expected variables related to performance of the president to be influential predictors of trust in the president. From analyzing both tables, having a more positive perception of the president’s performance significantly correlates with higher levels of trust. This relationship remains significant among all three models in both survey rounds. The strength of this relationship tells us that performance driven evaluation is influential in determining trust in the president. It also gives a reason to believe that the possibilities of fostering more trust in the president in Nigeria lies in performance related solutions. Other independent variables such as evaluations of corruption, freeness of elections and perceptions on government handling the economy also indicate this.
In both tables there seems to be no correlation between levels of education and trust in the president. Lower categories of schooling do not result in lower levels of trust. In
combination with the rather weak relationship for employment it might indicate that class-identities are less salient in determining political trust in Nigeria (Nnoli 1995; Berman 2004). The non-significant relationship of living conditions also points in this direction.
Evaluations on the levels of corruption of the presidency significantly influence the amount of trust people have in the president. People who think the president is more corrupt will have lower level of trust. This is in accordance with literature on the effects of corruption on political trust (Anderson & Tverdova 2003). It is also logical because seeing a president as corrupt can be believed to undermine evaluation on his legitimacy and capacity. Corruption is also a recurrent political theme in popular discourses in Nigeria (e.g., Daily Post 2018; the Vanguard 2018b).
In both tables I and II it can be seen that the negative perceptions of the free- and fairness of elections significantly influence the amount of trust a president holds. For respondents who
20
are more negative about the latest national elections we see a lower degree of trust compared to the group thought the last national elections were completely free and fair. This is
understandable because when someone thinks that the way a president was elected and came to power was not free and fair, it would influence amount of trust and legitimacy he holds in the eye of the people. The fact that we see a stronger relationship in table I than in table II can be because the 2007 elections were marred by a lot of irregularities and regarded as flawed (Omotola & Adebayo 2007).
Next to these general perceptions related to performance of the president and election, the models also controlled for perceptions on ‘’handling of the economy’’. This variable was used to test whether economic evaluations correlate with levels of trust. Earlier research has shown that economic issues are an important for determining voting behavior in African politics (Bratton et al. 2012). From the results, we indeed see that more positive perceptions on the presidents’ handling of the economy result in significantly higher levels of trust. This relationship remains significant in all models.
Another economically related variable is that for employment status. For those people who have no job (but looking) we see that they are significantly less less trusting of the president, compared to the group who has a full-time job (reference category). This can be because this group is more economically dissatisfied. We can expect that when you are jobless and on looking for a job, you are more sensitive to economic issues. As a jobless person you might attribute your inability to find a stable income to the most powerful political office in the center. Thus, economic evaluations are an influential determinant in perceptions of trust. This further substantiates that trust in the president in Nigeria seems more consistently driven by (economic) performance evaluations rather than ‘’primordial hatreds’’ and functions a strong alternative explanation for trust in the president in Nigeria. This is in accordance with the literature that stresses that political trust is driven by
21
Table I
. Summary for Ordinal Logistic Regression results for round 4 of
Afrobarometer survey. Dependent variable: trust in the president.
Model 1: ethnic group/ ‘tribe’
Model 2: Religion
Model 3: region Independent variables: Estimates: Estimates: Estimates: Ethnic group (‘tribe’):
Igbo Yoruba Rest Hausa/Fulani (reference)15 -.713*** (.151) -.587*** (.141) -.203* (.125) 0 - - - - - - - Religion: Christian Muslim (reference) - - -.390*** (.099) 0 - - Region: South-South South-East South-West North-Central North-East North-West (reference) - - - - - - - - - - - - -.429*** (.150) -.649*** (.173) -.530*** (.140) .206 (.156) .398** (.157) 0 Control variables: Age .004 (.004) .002 (.004) .005 (.004) Gender: Male Female (reference) .012 (.090) 0 -.005 (.091) 0 -.001 (.090) 0 Urban/Rural: Urban Rural (reference) -.206** (.094) 0 -.345*** (.091) 0 -.213** (.094) 0 Education: No formal schooling Primary schooling Secondary schooling
Post-secondary schooling (reference)
-.070 (.171) .006 (.148) -.115 (.106) 0 -.087 (.172) -.036 (.152) -.129 (.107) 0 -.140 (.168) -.071 (.150) -.171 (.106) 0 Employment status: No (not looking) No (looking) Yes, part time
Yes, full time (reference)
-.146 (.129) -.316*** (.129) -.189 (.121) 0 -.099 (.128) -.275** (.130) -.196 (.122) 0 -.084 (.128) -.242** (.130) -.115 (.122) 0
Your present living condition: -.003 (.039) -.002 (.039) .010 (.039)
15 SPSS automatically takes the last category of a variable as the reference category. Therefore, the relationships
22
Performance of the president: .495*** (.058) .517*** (.059) .524*** (.059) Corruption: office of the presidency -.486*** (.057) -.477*** (.058) -.480*** (.057) Last national elections free and fair: .258*** (.048) .251*** (.048) .232*** (.048)
Handling of the economy: .575*** (.055) .585*** (.056) .533*** (.056)
Ethnic group treated unfairly: -.067 (.050) -.063 (.050) -.061 (.051) Ethnic versus national identity: -.041 (.041) -.037 (.041) -0.52 (.041)
N = 1885 1832 1885
Nagelkerke16 .282 .288 .293
Note:
- Dependent variable: trust in the president with categories 0=Not at all, 1=Just a little, 2=Somewhat, 3=A lot. - Round 4 of the survey coincides with the presidential term (2007-2010) of Umaru Musa Yar'Adua.
- All regression coefficients are unstandardized. Standard errors are between parentheses - Significant at: *p < .10. **p < .05 ***p < .01
Table II.
Summary for Ordinal Logistic Regression results for round 5 of
Afrobarometer survey. Dependent variable: trust in the president.
Model 1: ethnic group/’tribe’ Model 2: Religion Model 3: region
Independent variables: Estimates: Estimates: Religion: Christian Muslim (reference category)17 -.153 (.096) 0 - - Region South-South South-East South-West North-Central North-East North-West (reference) - - - - - - .181 (.147) .422*** (.173) -.276* (.141) -.023 (.156) .604*** (.158) 0
16 The Nagelkerke measure is considered a “pseudo R squares”, it can be used to give an indication of the
quality of the model (Halperin & Heath 2012: 411). However, it should be interpreted with caution as these measures are do not share the same strength in explaining quality of the models as r-squares in linear regressions.
17SPSS automatically takes the last category of a variable as the reference category. Therefore, the relationships
23 Control variables: Age .007* (.004) .010*** (.004) Gender: Male Female (reference) .136 (.087) 0 -.002* (.090) 0 Urban/Rural: Urban Rural (reference) -.102 (.087) 0 -.034(.094) 0 Education: No formal schooling Primary schooling Secondary schooling
Post-secondary schooling (reference)
.040 (.1157) .080 (.153) .045 (.107) 0 -.078 (.153) .058 (.143) .025 (.110) 0 Employment status: No (not looking) No (looking) Yes, part time
Yes, full time (reference)
.026 (.112) -.286** (.114) -.143 (.127) 0 .041 (.153) -.283** (.116) .076 (.127) 0
Your present living condition: -.027 (.035) -.022 (.035)
Performance of the president: .961*** (.056) .947*** (.059)
Corruption: office of the presidency -.629*** (.063) -.632*** (.063)
Last national elections free and fair: .096** (.048) .176*** (.046)
Handling of the economy: .563*** (.057) .537*** (.058)
Ethnic group treated unfairly: .024 (.043) -.020 (.149)
Ethnic versus national identity: .107* (.050) .073 (.051)
N = 2164 2181
Nagelkerke .334 .345
Note:
- Round 5 of the survey coincides with the presidential term (2010-2015) of Goodluck Jonathan .
- Dependent variable: trust in the president with categories 0=Not at all, 1=Just a little, 2=Somewhat, 3=A lot. - The ethnic Ijaw group to which the president belongs is too small to include in model 1 for ethnic
group/’tribe’. Including it would not allow for a meaningful interpretation as the small N will cause confidence intervals to be too wide. Thus, this research only able to measure the identification of ‘tribe’ and trust in president for round 4
- All regression coefficients are unstandardized. Standard errors are between parentheses - Significant at: *p < .10. **p < .05 ***p < .01
24
Conclusion and Discussion
Based on the findings of this thesis, it can be concluded that ethnic affinity (tribal, religious and regional) does play a role in determining trust in the president in Nigeria. But the salience of these identities is not fixed and stable over time but fluid as this research shows that ethnic affinity does not consistently matter in determining levels of trust in the president in Nigeria. Using two subsequent Afrobarometer survey rounds for two presidents of different ethnic, religious and regional backgrounds, this thesis shows that ethnic affinity only seems to play a significant role in determining trust levels for president Yar’Adua and not for Goodluck Jonathan. Performance driven evaluations and evaluations related to elections and the economy seem to be more consistent in explaining trust levels in the president. These findings have several important implications.
First, the findings seem to support a constructivist notion of social identities in Nigeria as opposed to more primordial or structural accounts of ethnicity (Nnoli 1995; Horowitz 1985). The conclusion of this thesis supports the claim that “ethnicity is not integral, uniform, or consistent, but rather mutable and contingent” (Lewis 2007: 26). The fact that religious and regional affinity fails to significantly correlate with trust levels for president Goodluck Jonathan point towards the possible importance of contextual factors in explaining the salience of these identities in determining trust in the president. It gives reasons to believe that the role of elites is important in manipulating and politicizing
identities, as argued in much qualitative studies on identities in Nigeria and other sub-Saharan African countries (e.g., Eze 2010; Davidson 1992; Ekeh 1975). The reason that we find ethnic affinities to be influential in table I can be because that there is a sense of distrust between certain ethnic groups in Nigeria but that contextual factors make these salient. During the highly controversial elections bringing Yar’Adua to power, many politicians were playing ethnic and religious ‘cards’ to win votes, increasing the ‘’fear of domination’’
between groups (Omotola & Adebayo 2007). So, the quantitative nature of this research offers as an important contribution to our theoretical understanding of identities and opens avenues for further research to study these contextual factors, that might be studied in a more qualitative way. This highlights that quantitative and qualitative work can have
complementary contributions and should not be exclusionary.
Unfortunately, due to the small sample size of the ethnic Ijaw group in the Afrobarometer surveys, this thesis was unable to measure the tribal affinity of Goodluck
25
Jonathan. Future research would be welcome that studies perceptions of the smaller ethnicities in Nigeria. Especially more research is welcome to see whether people in
particular regions, such as the delta region in the south-south, are generally more distrusting of national politics in Nigeria due to their perceived marginalization. This would require a focus on the intersectionality of certain identities and attitudes.
Because this research has been able to use survey-data means that the findings in this research have an added value in that they rely on the subjective responses of respondents. This thesis has been limited in that it has only been able to study two rounds of the
Afrobarometer survey, partly due to the limited availability of rounds so-far. Nevertheless, this thesis could be replicated by using a longer time-series analysis that covers longer periods and more presidents, if data comes available. This would increase the validity as the relations are studied over a longer period of time and presidents. Moreover, there seems no reason to believe that the dimensions and relations between ethnic identities cannot be studied for different dimensions of political trust.
Alongside showing the role of ethnic affinities play, this research shows that trust in both presidents is consistently determined by (economic) performance, perceptions of corruption and freeness of elections, making it a powerful additional explanation. These findings are in accordance with literature on political trust (Uslaner 2018; Lewis 2007). This finding for Nigeria has important implications, in that the solution of fostering more trust and thereby legitimacy in the political center lie in the strengthening of institutional and
democratic performance. This implicates that the current political problems of Nigeria do not necessarily lie in ‘’primordial hatreds’’ of ethnic, religious and regional identities and
animosities but are also more good-governance driven. This is positive and negative at the same time. Positive as it shows that trust and legitimacy in Nigeria’s political center can be fostered based on political interventions related to good-governance. Negative as the
challenge Nigeria faces in building a well-functioning democracy is daunting as the country is facing many socio-economic and political challenges. The fact that political trust is
important in the quality of effective governance, democratic participation (Karakoç 2013: 93; Marien & Hooghe 2011) makes studying political trust important. Last, the conclusions of this thesis go against oversimplified notions of African politics hindered by ‘tribalism’ (Kaplan 1994). It forces researchers and observes to engage in more serious way in studying the pathways that can lead to increased political legitimacy and trust in sub-Saharan African countries.
26
Bibliography
Adetiba, T. C., & Rahim, A. (2012). Between ethnicity, nationality and development in Nigeria. International Journal of Development and Sustainability, 1(3), 656-674. Afrobarometer (2015). Data codebook for a Round 5 Afrobarometer survey in 34 African
countries. [Online] http://afrobarometer.org/data/merged-round-5-codebook-34-countries-2011-2013-last-update-july-2015 [Last accessed: 05 July 2018] Afrobarometer.org (2018). Afrobarometer: let the people have a say. [Online]
http://www.afrobarometer.org/ [Last accessed: 05 July 2018] Awolowo, O. (1947). Path to Nigerian Freedom. London: Faber
Anderson, C. J., & Tverdova, Y. V. (2003). Corruption, political allegiances, and attitudes toward government in contemporary democracies. American Journal of Political
Science. 47(1), 91-109.
Ahlerup, P., & Isaksson, A. S. (2015). Ethno‐Regional Favouritism in Sub‐Saharan Africa.
Kyklos, 68(2), 143-152.
Alesina, A., Baqir, R., & Easterly, W. (1999). Public goods and ethnic divisions. The
Quarterly Journal of Economics, 114(4), 1243-1284.
Bratton, M., Bhavnani, R., & Chen, T. H. (2012). Voting intentions in Africa: ethnic, economic or partisan? Commonwealth & Comparative Politics, 50(1), 27-52.
Berg, L., & Hjerm, M. (2010). National identity and political trust. Perspectives on European
Politics and Society, 11(4), 390-407.
Berman, B.J. (2004). Ethnicity, Bureaucracy & Democracy: The Politics of Trust. In: Berman, B., Eyoh, D., & Kymlicka, W. (Eds.). (2004). Ethnicity and democracy in
Africa. Athens: Ohio University Press.
Berman, B.J. (2013). Nationalism in Post-Colonial Africa. In: Breuilly, J. (ed.). The Oxford
Handbook of the History of Nationalism. Oxford: Oxford University Press.
Berman, B.J. (2010). Ethnic diversity and economic instability in Africa: Policies for harmonious development. JICA-Rl Working Paper (22), 1-40
CIA World Factbook (2018). ‘’Nigeria: people and society’’. [Online]
https://www.cia.gov/library/publications/the-world-factbook/geos/ni.html [Last accessed: 05 July 2018]
Collier, P. (2001). Implications of ethnic diversity. Economic policy, 16(32), 128-166.
Collier, P., & Hoeffler, A. (2004). Greed and grievance in civil war. Oxford economic papers, 56(4), 563-595.
27
Daily Post (2018). Search entry: corruption president. [Online]
http://dailypost.ng/?s=corruption+president [Last accessed: 05 July 2018]
Davidson, Basil (1992). The Black Man’s Burden: Africa and the Curse of the Nation-State. New York: Random House.
Ekeh, P. (1975). Colonialism and the Two Publics in Africa: A Theoretical Statement.
Comparative Studies in Society and History. 17(1), 91-112.
Erickson, E., (1968). Identity: Youth and Crisis. New York: Norton.
Easterly, W., & Levine, R. (1997). Africa's growth tragedy: policies and ethnic divisions. The
quarterly journal of economics, 112(4), 1203-1250.
Eifert, B., Miguel, E., & Posner, D. N. (2010). Political competition and ethnic identification in Africa. American Journal of Political Science, 54(2), 494-510.
Eze, M.O. (2010). The politics of History in Contemporary Africa. New York: Palgrave Macmillan
Falola, T., & Heaton, M. M. (2008). A history of Nigeria. Cambridge: Cambridge University Press.
Green, E.D. (2017). Ethnicity, national identity and the state: evidence from sub-Saharan Africa. British Journal of Political Science.
Hale, H. E. (2004). Explaining ethnicity. Comparative political studies, 37(4), 458-485. Halperin, S. and Heath, O. (2012) Political research Methods and practical skills. New York:
Oxford University Press.
Hazen, J. & Horner J. (2007). Small Arms, Armed Violence, and Insecurity in Nigeria: The Niger Delta in Perspective. In: Small arms survey 2007. Graduate Institute of
International Studies. Geneva: Switzerland.
Hetherington, M. J. (2006). Why trust matters: Declining political trust and the demise of
American liberalism. Princeton: Princeton University Press.
Horowitz, D.L. (1985). Ethnic groups in conflict. Berkeley: University of California Press Hutchison, M. L., & Johnson, K. (2011). Capacity to trust? Institutional capacity, conflict, and political trust in Africa, 2000–2005. Journal of Peace Research, 48(6), 737-752. Johnson, K., & Hutchison, M. L. (2013). Hybridity, political order and legitimacy: Examples
from Nigeria. Journal of Peacebuilding & Development, 7(2), 37-52.
Jega, A. (2000). General Introduction Identity Transformation and the Politics of Identity under Crisis and Adjustment. In: Jega, A. (ed.) Identity transformation and identity
