CURING THE CURSE:
REVENUE AND REPRESENTATION
IN CONTEMPORARY NIGERIA
By
Olufunmbi M. Elemo

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Political Science
2012

ABSTRACT
CURSING THE CURSE:
REVENUE AND REPRESENTATION
IN CONTEMPORARY NIGERIA
By
Olufunmbi M. Elemo
Africans most commonly report “representing the people” as an elected official’s most
important responsibility; however, only 16% of respondents report that their legislators “listen to
what people like [them] have to say” (Afrobarometer Public Opinion, 2005). Given citizen
preferences, under what conditions are African elected officials most likely to represent their
constituents’ interests?
Scholars have continuously linked the development of representative institutions to the
taxation of citizens. Originating in studies of Western Europe, theory suggests, in order to raise
revenue for war/border protection, rulers enter into a contract with citizens: citizens agree to
provide tax revenue in exchange for an enhanced role in governance. With taxation comes the
incentive for political leaders to shift public policy and spending toward citizen interests. They
do so in order to maintain their income source. However, this previous theory ignores how, in
many instances, politicians have access to revenue from sources other than taxation. In particular,
income from natural resources serves as a large, lucrative alternative to tax revenue. This is
especially the case in the African context. How does access to natural resource wealth influence
the development of representation? The central hypothesis is that access to an external revenue
base provides an avenue that is less costly than bargaining with citizens. Those able fund their
governments without taxation can also bypass the need to defer to citizen interests, stymieing the
development of representative institutions.

In an application of this theory to Africa, analysis focuses on the Nigerian case, where
interactions between income from taxation and petroleum create varied revenue compositions
across (sub-national) states. Public opinion data as well as data collected during fieldwork in
Nigeria (October 2010 – July 2011) are utilized to this end. This includes: budgetary data (1999
– 2009) from the Central Bank of Nigeria (CBN), original interviews with state and national
legislators, and expert interviews with civil servants at the Federal Ministry of Finance, Federal
Inland Revenue Service, and various state internal tax revenue boards.
Analysis begins with an outline of the contemporary system of revenue extraction and
taxation in Nigeria. Next, utilizing Ordinary Least Squares (OLS) statistical models,
examination of CBN budgetary data confirms that as revenue from taxation increases, both local
and state government spending on public service provision increases. Following this finding,
Hierarchical Linear (Statistical) Models are used to test the taxation-representation link at the
micro-level. Evidence suggests that state legislators in more tax-reliant states spend the most
time performing constituency services. Likewise, when making decisions, these legislators are
most likely to prioritize citizens’ preferences over other considerations. Last, the investigation
considers the other side of the tax contract, investigating how ordinary Nigerians perceive
taxation and representation. Using public opinion data, analysis confirms that, in exchange for
tax payment, Nigerians expect government to shift their policy and spending to reflect public
priorities. Therefore, in circumstances where Nigerians are more satisfied with government
performance, we observe higher willingness pay taxes.
A relationship between taxation and representative governance suggests that
strengthening local and state governments’ tax administration and building citizen capacity to
monitor government budgeting can also bolster responsive and democratic governance in Africa.

Copyright by
OLUFUNMBI M. ELEMO
2012

ACKNOWLEDGEMENTS

I thank Dr. Michael Bratton for his academic and professional guidance through my
graduate school career. I also thank Dr. Eric Chang, Dr. Jeff Conroy-Krutz, and Dr. Rita Kiki
Edozie for their advice and service on my dissertation committee.
I thank the following individuals and organizations for their assistance during my
fieldwork in Nigeria: the Central Bank of Nigeria, particularly the Data Management Office and
Statistics Department, the CBN Library, Mr. B.O. Ogunlaja, Mr. Oshiba, Lady Ngozie Esther
Nwachukwu, and Mr. Adeyemi Adeboye; Mr. Rafsanjani (CISLAC); Barrister Ferdinand
Okotete (Nigerian Bar Association, Sokoto State); Dr. Bukola Adesina, Dr. Victor Isumonah,
and Professor Adigun Agbaje (University of Ibadan); Dr. Adele Jinadu (University of Lagos); Dr.
Etannibi Alemika, Mr. Innocent Chukwuma, Ms. Kemi Okenyodo, and Ms. Linda Asiquo
(CLEEN Foundation); Ms. Hahiya Saadatu (Court of Appeals, Abuja); the Honourable Members
of the Nigerian National Assembly and the House Committee on Police Affairs; the Office of the
Permanent Clerk and Honourable Members of the Lagos State House of Assembly; the Office of
the Permanent Clerk and Honourable Members of the Ekiti State House of Assembly; the Office
of the Permanent Clerk, Public Relations Office, and Honourable Members of the Oyo State
House of Assembly; the Office of the Permanent Clerk and Honourable Members of the Kano
State House of Assembly; the Office of the Permanent Clerk, Public Relations Office, and
Honourable Members of the Katsina State House of Assembly; the Office of the Permanent
Clerk, Legal Services Office, and Honourable Members of the Sokoto State House of Assembly;
the Ministry of Finance’s Director General’s Office and International Economic Relations
Department; the Federal Inland Revenue Service’s Chairman’s Office and Tax Operations
Group; the Lagos State Internal Revenue Service; the Kano State Board of Internal Revenue; the
v

Sokoto State Board of Internal Revenue; Mr. Gbenro Adesina; Dr. Carl LeVan (American
University); Improving Institutions for Pro-Poor Growth (Center for the Study of African
Economies, University of Oxford); Dr. Nic Cheesman (Jesus College, University of Oxford); Dr.
Adrienne LeBas (American University).
I thank Dr. Ric Hula, Dr. Steve Kautz, the MSU Department of Political Science, Brian
Kennedy, Jerry Lavery, my fellow graduate students, and the MSU Graduate School.
Finally, I thank my Mom, Dad, Shola, Sister Bola, and Kay for their constant love and
encouragement. I also thank my family members and close family friends in the U.S. and in
Nigeria for their support. For their remarkable friendship, encouragement, and (at times)
consolation, I thank: Dr. Patrice Ryce, Hardik Dalal, Hayley Boothe, Rev. Sarah Midzalkowski,
Julie Powers, Canterbury MSU, Danielle Carter, Dr. Niccole Pamphilis, Nick Kerr, Courtney
Heedles, Jess Caparoso, Nekol Peterson, Laura Carrero, Felicia Hardin, T.J. Barber, Mary Wells,
Beverlyn Elliott, Susi Prescott, Nicole Morrell, and the Twilight Fandom.

vi

TABLE OF CONTENTS

LIST OF TABLES

viii

LIST OF FIGURES

xv

INTRODUCTION

1

CHAPTER ONE
Theoretical Framework: Explaining Revenue and
Representation Linkages

10

CHAPTER TWO
Democracy, Governance, and Revenue Generation in
Contemporary Nigeria

36

CHAPTER THREE
Revenue Sources and Subnational Government Spending Priorities

79

CHAPTER FOUR
Leveraging Regional Variation in Tax Generation and Oil Dependence

109

CHAPTER FIVE
Popular Perceptions of Taxation and Representation:
Attitudes and Behaviors To Tax Payment

202

CHAPTER SIX
Comparing Tax Cultures in Lagos and Kano

265

CONCLUSION
Beyond Nigeria:
Explaining Africans’ Attitudes and Behavior Toward Taxation

304

APPENDIX

324

BIBLIOGRAPHY

345

vii

LIST OF TABLES

Table 2.1
Sources Of Federal Government Oil Revenue

72

Table 2.2
Vertical Allocation Of Federation Account, 1981 – Present

72

Table 2.3
Horizontal Allocation Of Federation Account, 1980 – Present

73

Table 2.4
Current Income Tax Rates

73

Table 2.5
Federal Government’s Revenue

74

Table 2.6
Sources Of Federal Government Non-Oil Revenue

75

Table 2.7
State Governments' Revenue

76

Table 2.8
Local Governments' Revenue

77

Table 2.9
Average Oil And Tax Dependence Among The States (1999 – 2009)

78

Table 3.1
Conceptual Framework

95

Table 3.2
Summary Statistics—LGA Model

96

Table 3.3
Summary Statistics—State Model

98

Table 3.4
LGA Model

99

Table 3.5
State Model

100

viii

Table 3.6
Summary Statistics—Imputed LGA Model

101

Table 3.7
Summary Statistics—Imputed State Model

102

Table 3.8
LGA Imputed Model

103

Table 3.9
State Imputed Model

104

Table 3.10
Observations with Highest Cook's D Values

105

Table 3.11
Observations with Highest Cook's D Values

106

Table 3.12
LGA Model Without Outliers

107

Table 3.13
State Model Without Outliers

108

Table 4.1
Tax Generation in the South-West Region

164

Table 4.2
Tax Generation in the North-East Region

164

Table 4.3
Tax vs. Oil Income in Nigeria’s Regions

165

Table 4.4
Tax Generation in the North-West Region

165

Table 4.5
Standard Deviation in Regional Tax and
Oil Income as a Percentage of State Budget

166

Table 4.6
Average Tax Generation in the South-West
Region, 1999 – 2009

166

ix

Table 4.7
Average Tax Generation in the North-West
Region, 1999 – 2009

167

Table 4.8
Tax Generation in the South-west Region

168

Table 4.9
Oil Dependence in the South-West Region

169

Table 4.10
Tax Generation in the North-West Region

170

Table 4.11
Oil Dependence in the North-West Region

171

Table 4.12
Legislative Productivity in the South-West Region

172

Table 4.13
Legislative Productivity in the North-West Region

172

Table 4.14
Sampling Procedures in Six State Legislatures

173

Table 4.15
Age of Legislators

173

Table 4.16
Gender of Legislators

174

Table 4.17
Ethnicity of Legislators

174

Table 4.18
Religion of Legislators

175

Table 4.19
Highest Level Legislators' Education

175

Table 4.20
Occupation Prior to Legislative Office

176

Table 4.21
Number of Years in Legislative Office

176

x

Table 4.22A
Effect of Revenue on Legislators' Execution of
Representation

177

Table 4.22B
Effect of Revenue on Legislators' Execution of
Representation

179

Table 4.23
Factor Analysis of Elite Representation Indicators
Orthogonal Varimax Principle Factors

181

Table 4.24
Factor Analysis of Elite Representation Indicators
Rotated Factor Loadings

182

Table 4.25
Effect of Revenue on Elite Representation’s
Two Dimensions

183

Table 5.1
Crosstabulation, Willingness to Pay
Taxes and Belief in Executive Justification
How Tax Income is Spent

239

Table 5.2
Chi Square Test, Differences in Rates
of Income Tax Payment

240

Table 5.3
Chi Square Test, Differences in Rates
of Property Tax Payment

240

Table 5.4
Chi Square Test, Differences in Rates
of Local Government Tax Payment

240

Table 5.5
Hypothesized Effects of Quality of
Representation and Trust on Citizens' Willingness
to Pay Taxes and Actual Tax Payment

241

Table 5.6
Summary Statistics—Effect of Quality of
Representation and Trust on Citizens' Willingness
to Pay Taxes and Actual Tax Payment

242

xi

Table 5.7
Demographic Comparisons Between
Original Sample and New Sample

243

Table 5.8
Religion of Respondents

243

Table 5.9
Ethnicity of Respondents

243

Table 5.10
Urban vs. Rural Dwellers

244

Table 5.11
Gender Distribution

244

Table 5.12
Effect of Quality Of Representation and
Trust on Citizens' Willingness to Pay Taxes
and Actual Tax Payment

245

Table 5.13
Summary Statistics—Effect of Government
Performance on Citizens' Willingness to Pay
Taxes and Actual Tax Payment

247

Table 5.14
Effect of Government Performance on Citizens'
Willingness to Pay Taxes and Actual Tax Payment

249

Table 5.15
Factor Analysis of Citizen Perceptions of
Representation Indicators
Orthogonal Varimax Principle Factors

252

Table 5.16
Factor Analysis of Citizen Perceptions of
Representation Indicators
Rotated Factor Loadings

252

Table 5.17
Factor Analysis of Citizen Satisfaction with Government Performance
Orthogonal Varimax Principle Factors

253

xii

Table 5.18
Factor Analysis of Citizen Satisfaction with Government Performance
Rotated Factor Loadings

253

Table 5.19
Effect of Quality and Representation and Satisfaction with
Government Performance Indices on Citizens’ Willingness to
Pay Taxes

254

Table 5.20
Urban Nigeria, Summary Statistics—Effect of
Government Performance on Citizens' Willingness
to Pay Taxes

255

Table 5.21
Urban Nigeria, Distribution of Household Income

256

Table 5.22
Urban Nigeria, Distribution of Respondents' Education

256

Table 5.23
Urban Nigeria, Religion of Respondents

256

Table 5.24
Urban Nigeria, Ethnicity of Respondents

257

Table 5.25
Urban Nigeria, Age Distribution

257

Table 5.26
Urban Nigeria, Gender Distribution

257

Table 5.27
Urban Nigeria, Effect of Government
Performance on Citizens' Willingness to Pay Taxes

258

Table 6.1
State Demographics (2005 – 2006)

291

Table 6.2
Distribution of Working Population by Activity

293

Table 6.3
Tax Generation in Lagos and Kano

293

xiii

Table 6.4
Correlation Between Legislator's Home State
and Belief States Should Collect

294

Table 6.5
Correlation Between Legislator’s Home State
And Belief Tax Deparment Can Compel Payment

295

Table 6.6
Urban Nigeria, Summary Statistics—Willingness
to Pay Taxes in Lagos And Kano

296

Table 6.7
Urban Nigeria, Willingness To Pay Taxes In
Lagos and Kano

298

Table 7.1
Summary Statistics—Effect of Quality of
Representation, Trust, and Government Performance
on Africans' Willingness to Pay Taxes and Actual
Tax Payment

314

Table 7.2
Effect of Quality of Representation, Trust,
and Government Performance on Africans'
Willingness to Pay Taxes and Actual Tax Payment

318

Table 7.3
Africans’ Willingness to Pay Taxes By Country

323

xiv

LIST OF FIGURES

Figure 4.1
Tax Generation in the South-West Cases (1999 – 2009)

184

Figure 4.2
Oil Dependence in the South-West Cases (1999 – 2009)

185

Figure 4.3
Tax Generation in the North-West Cases (1999 – 2009)

186

Figure 4.4
Oil Dependence in the North-West Cases (1999 – 2009)

187

Figure 4.5
Legislators’ Ratings of How Effective their State
Legislature Represents Constituents vs. Average
State Tax Generation

188

Figure 4.6
Legislators’ Ratings of How Effective their State
Legislature Represents Constituents vs. 2009
State Tax Generation

189

Figure 4.7
Legislators’ Ratings of How Effective their State
Legislature Represents Woman vs.
Average State Tax Generation

190

Figure 4.8
Legislators’ Ratings of How Effective their State
Legislature Represents Woman vs. 2009 State
Tax Generation

191

Figure 4.9
Time Legislators Report Devoting to Constituency Work
vs. Average State Tax Generation

192

Figure 4.10
Time Legislators Report Devoting to Constituency Work
vs. 2009 State Tax Generation

193

Figure 4.11
Representation is Legislator’s Most Important Job vs.
Average State Tax Generation

194

xv

Figure 4.12
Representation is Legislator’s Most Important Job vs.
2009 State Tax Generation

195

Figure 4.13
Constituents’ Views Are the Most Important
Consideration When Making Decisions vs.
Average State Tax Generation

196

Figure 4.14
Constituents’ Views Are the Most Important
Consideration When Making Decisions vs.
2009 State Tax Generation

197

Figure 4.15
In Case of Conflict, Will Oppose Party to Support
Constituents vs. Average State Tax Generation

198

Figure 4.16
In Case of Conflict, Will Oppose Party to Support
Constituents vs. 2009 State Tax Generation

199

Figure 4.17
If Legislator Ignores Constituents He/She Should
Lose Seat vs. Average State Tax Generation

200

Figure 4.18
If Legislator Ignores Constituents He/She Should
Lose Seat vs. 2009 State Tax Generation

201

Figure 5.1
“The tax department always has the right to make people pay taxes.”

259

Figure 5.2
“Which of the following statements is closest to your view?
Choose Statement 1 or Statement 2…”

260

Figure 5.3
“Who do you think actually has primary responsibility
for: Collecting income taxes?”

261

Figure 5.4
“Have you had to make any of the following payments during the past year…”

262

xvi

Figure 5.5
“Given the current government’s performance that you have seen,
do you find yourself more willing or less willing to pay taxes…”

263

Figure 5.6
(ACROSS STATES) “Given the current government’s
performance that you have seen, do you find yourself more
willing or less willing to pay taxes…”

264

Figure 6.1
“Who do you think actually has primary responsibility
for collecting income taxes?”

299

Figure 6.2
ELITES “The tax department always has the right to make people pay taxes.”

300

Figure 6.3
CITIZENS “The tax department always has the right to make people
pay taxes.”

301

Figure 6.4
“Have you had to make any of the following payments during the
past year…”

302

Figure 6.5
“Given the current government’s performance that
you have seen, do you find yourself more willing or less willing
to pay taxes…”

303

xvii

INTRODUCTION

Africans most commonly report “representing the people” as an elected official’s most
important responsibility; however, only 16% of citizens across 20 African countries report that
their legislators “listen to what people like [them] have to say” (Afrobarometer Public Opinion,
2008). Given citizen preferences, under what conditions are African elected officials most likely
to represent their constituents’ interests? How do elected officials understand their role in
government and their relationship with constituents? How do these views compare with those of
citizens? Do these perceptions of representation vary according to the nature of the fiscal
relationship between government and citizens?
Scholars link the development of representation, defined as the incorporation of citizen
interests in political decision-making, to the taxation of citizens (Huber and Powell, 1994;
Aldrich, 1995; Stokes, 1999). Theories originating in studies of Western Europe suggest that, in
order to raise revenue for war and border protection, rulers enter into a contract with citizens.
Citizens agree to provide tax revenue in exchange for an enhanced role in governance. With
taxation comes the incentive for political leaders to shift public policy and spending toward
citizen interests. They do so in order to maintain their income source. However, this theory
ignores how, in many instances, politicians have access to revenue from sources other than
taxation. In particular, income from natural resources serves as a large, lucrative alternative to
tax revenue. This is especially the case in the African context. How does access to natural
resource wealth influence the development of representation? I differentiate between domestic
sources of state revenue (income derived government taxation of citizens) and income from the
sale of natural resources as an external source of income. I hypothesize, to the extent that

1

governments rely on revenue generated from taxation in order to fund public activities, elected
officials will be representative of their constituents. In these cases, elected officials will utilize
public policy and expending public resources in a manner that satisfies citizens’ priorities. Once
again, this assertion follows the description of how representative government formed in
Western Europe (Levi 1988; Tilly 1985, 1990). “A [revenue production/tax] policy is
understood as a contract between ruler and agents or ruler and constituents,” which in exchange
for revenue, citizens have a basis for requesting reciprocal goods and services (Levi, 1988: 49).
Simply, taxation raises demands for representation. Representation, in turn, undergirds the
consolidation of democratic practice.
On the other hand, I hypothesize that in states that are more dependent on natural
resource rents, leaders are less likely to expend the resources to remain responsive to citizen
interests. Again, Levi (1988) argues that state elites have their own interests, separate from other
political actors. Being rational and seeking to maximize their own utility, these elites will seek
the most cost-effective method of acquiring resources to carry out their agenda. Access to an
external revenue base in the form of resource rents provides an avenue that is less costly than
bargaining with citizens and possibly deviating from their preferred policy outcome. Thus,
instead of bearing the transaction costs associated with taxation (e.g. enforcing compliance with
tax policies), leaders with access to resource rents are able to fund their governments while
bypassing the need to defer to citizen interests.
My research builds upon Ross’ (2001) influential work, which demonstrates reliance on
profits from natural resources is negatively associated with levels of democracy. Previous
research performs the important preliminary task of unearthing the association between resource
wealth and lack of democracy. However, my aim is to disaggregate the “oil hinders democracy”
2

hypothesis, uncovering the mechanisms of this relationship. A research direction focusing on
micro-level analysis could prove to be more theoretically and empirically fruitful. By
emphasizing individual linkages between citizens and elected officials, it is possible to
investigate the structure of incentives for elite behavior. We can also consider how oil-income
dependence influences these incentives. I hypothesize that in these resource-reliant
governments, the fiscal relationship between elected officials and citizens, usually established by
taxation, is broken. These officials are, in turn, less beholden to their constituents and the public
will.
In an application of this theory to Africa, I focus on the Nigerian case, where interactions
between petroleum and tax based income create varied revenue compositions across (subnational) states. As a result, I also expect that, within Nigeria, the extent to which elected
officials are representative of their citizens also varies.
Several scholars outline the Nigerian system of revenue generation, driven primarily by
access to income from petroleum (Okoko and Nna, 1997; Suberu, 2003; Fajingbesi et al, 2004).
For example, Ikein and Briggs-Anigboh (1998: 271 - 275) and Olaniyi (2001) provide a brief
account of the history of petroleum in Nigeria. While the exploration for oil began in 1908,
Royal Dutch Shell Incorporated made the first discovery in 1956 in Rivers State. Since 1958, oil
production has been concentrated in the following Nigerian (subnational) states: Rivers, Delta,
Edo, Imo, Abia, Akwa Ibom, and Cross Rivers. By 1990, oil rents made up almost 82% of
national government revenue (Ikein and Briggs-Anigboh, 1998: 346). Nigeria’s subnational
governments also derive revenue from the petroleum industry. As of 2002, 55% of oil revenues
accrue to the national government, 25% to the state governments, while the remaining 20% is
distributed among local government administrations (LGAs). Of the revenue allocated to the 36

3

states, distribution is based upon equity, state population, level of state social development (e.g.
education, health, water access), and state internal tax revenue effort (Usman, 2007).
Although Nigeria is heavily dependent on petroleum income, a structure for government
tax generation exists, and has origins pre-dating British colonialism. Orewa (1979) and Guyer
(1991) outline the history of taxation in Nigeria. Prior to British colonization, local chiefs
exacted tributes from their people, and individuals paid them with crops or brass rods. During
the colonial period, beginning in 1906, the British Colonial Government levied a livestock and
cattle tax. As a result of the Native Revenue Proclamation, colonial leaders extracted these taxes
as “their proper dues from their people in return for their works as rulers and judges” (Orewa,
1979: 3). Prior to independence, the Local Government Law of 1952 was enacted, allowing a
local council to tax any “income from trade, business, profession, or employment” exceeding
100 naira. This was followed by the Regional Government income tax law, which established a
personal income tax. The tax was progressive, ranging from between 3 kobo per naira (on
incomes of 400 naira or less) to 45 kobo per naira on incomes exceeding 8,000 naira (Orewa,
1979: 15 - 6).
Issues of taxation, revenue, and governance continuously inform Nigeria’s political
development. Post-independence (1960 – 1999), Nigeria’s military leadership repeatedly
curtailed subnational entities’ power to tax. This created a situation where state and local
governments were dependent on the national government for revenue. However, in the
contemporary period, after Nigeria’s 1999 return to democratic governance, state and local
governments have regained the authority to generate income, independent of federal transfers of
petroleum revenue. Nigeria’s subnational governments differ in the extent to which they have
mobilized income via taxation and/or rely on oil transfers. For example, between 1999 and

4

2009, Lagos State generated an average of 53% of their income from the taxes on citizens. This
made Lagos the strongest tax generator, and during this period, 24% of Lagos’ revenue was
derived from federal oil transfers. On the other hand, between 1999 and 2009, Ebonyi State
ranked last among all 36 states in tax generation, averaging only 3% of state income from taxes
(68% from federal oil transfers). How does the difference between these two states’ capacity to
extract taxes influence how elected officials engage in citizen representation? Based on theories
linking taxation and representative governance, I hypothesize that since the Lagos State
Government relies more on tax income, elected officials will engage in more representative
behavior and be more responsive to citizens.
Relying on subnational units of analysis, exploiting the variation in revenue generation
within Nigeria, provides empirical leverage. With this method, I am able to “increase the
number of observations, make controlled comparisons…[,] accurately code cases, and thus make
valid causal inferences.” (Snyder, 2001: 93). In addition, state and local governments play an
important role in Nigerian politics. These entities are charged with “the enhancement of
representative grassroots democracy…[and also serve as] a mechanism for participatory
integration people into the fold of democratic governance (Fajingbesi et al, 2004: 47).
Therefore, focusing this analysis on state and local governance and revenue extraction has
practical policy implications and relevance to the Nigerian context.
I utilize public opinion data as well as data collected during fieldwork in Nigeria (October
2010 – July 2011). This includes: budgetary data (1999 – 2009) from the Central Bank of
Nigeria, 109 original interviews with state legislators and civil servants at the Federal Ministry of
Finance, Federal Inland Revenue Service, and various state internal tax boards.

5

Analysis begins with an outline of the contemporary system of revenue sharing and
generation. I go on to demonstrate that when subnational government’s have greater tax capacity
and fund public activities with income derived from taxation of citizens, state legislators are
more representative of citizens. Specifically, these governments and elected officials are more
likely to engage in behaviors that are representative of their constituents’ interests, including
spending more government revenue on public services and spending more time providing
various constituency services. Furthermore, citizens in states where leaders rely on taxation are
more likely than their counterparts in resource-dependent states to view their government as
representative. Thus, when taking natural resource income into account, Western European
theories of the fiscal contract between elites and citizens and origins of representative
governance also apply in Africa.

Dissertation Outline
This dissertation is organized in the following manner:
Chapter One examines the literature on democracy, representation, and the role government
revenue plays in the development of both. In particular, I investigate previous definitions of
political representation, separating this concept from other related concepts. I also examine the
role representation plays in democratic consolidation. This chapter discusses historical
relationships between representative processes and government revenue in Western Europe.
Moreover, I identify how previous analyses overlook the influence of non-tax revenue on the
development of representation. I further examine theories of how tax-based and non-tax revenue

6

influence governance in Sub-Saharan Africa. Using this theoretical framework, I present my
analytical focus of study and the hypotheses that I will later test.
In Chapter Two, I examine the role revenue (and gaining access to income) has played in the
development of democracy and politics in Nigeria. This includes an investigation of revenue
generation in previous civilian and military regimes and in neopatrimonial relationships between
politicians and citizens. Over time, we see that gaining access to revenue influenced decisions
by civilian and military governments to consolidate or disperse political and fiscal power through
Nigeria’s federal system. Using Nigeria’s past and present constitutional frameworks, I discuss
the evolution of state and local governments’ autonomy from the national level. Thereafter, I
examine Nigerian fiscal federalism, considering revenue sharing from petroleum, the current
system of taxation (internally generated revenue, IGR), and current reforms and debates in oil
revenue sharing and the tax system.
Chapter Three investigates conditions under which Nigerian local and state governments are
most likely to engage in representative behaviors, primarily spending public revenues on public
service delivery. Local and state governments, increasingly important to the execution of
Nigerian democracy, are being granted the power to raise revenue via taxation in order to
facilitate the execution of growing duties. What determines local and state government spending
priorities? Gibson and Hoffman (2005) find that in Tanzania and Zambia, local governments
deriving a larger portion of their revenue from citizen taxation are less likely to spend money on
recurrent expenditures (e.g. government salaries) and more likely to spend revenue on public
service delivery. Using Central Bank data, I test and confirm this hypothesis in Nigeria at the
local and state government levels.

7

In Chapter Four, I rely on individual-level data to investigate the relationship between
revenue and representation. Focusing on the regional variation in tax generation (vs. oil
dependence) in the South-West and North-West regions, I select a high performer,
average/medium performer, and low performer. This results in six case-states. First, I profile
each state’s revenue effort and legislature. Next, I examine if varied government capacity to
generate tax revenue (high, average, low) influences how elites report engaging in representative
functions. Relying on interviews conducted with legislators in Lagos (high), Oyo (average),
Ekiti (low), Sokoto (high), Kano (average), and Katsina (low) States, I find that officials in high
tax-generating states spend more time performing constituency services. These legislators are
also more likely to prioritize citizen preferences over other considerations.
Chapter Five investigates the other side of the tax contract: under what conditions are
Nigerians more likely to comply with requirements for tax payment in the first place? I
hypothesize that compliance with demands for tax payment is dependent on two beliefs (Levis,
1988). First, an individual must believe that political leaders will fulfill their end of the bargain;
a citizen must perceive that, if he or she pays taxes, government will reciprocate by incorporating
citizens’ interests in policymaking. Second, a taxpayer must believe that other citizens are
complying with tax payment. If this is the case, Nigerians who identify that their elected
officials are representing their interests in government will also be more willing to pay taxes.
These individuals will also report compliance with actual tax payment at a higher rate than
Nigerians who do not feel politicians are representing them. Second, Nigerians who believe that
other citizens are complying with tax payment will be more willing to pay taxes and also follow
through with actual payment. Utilizing Ordinary Least Squares (OLS) models as tests, analysis
supports these hypotheses.

8

In Chapter Six, I go on to perform an inter-regional analysis of taxation and governance,
comparing two states in each region. Though Lagos (South-West Region) and Kano (NorthWest Region) share similar levels of socio-economic development and a comparable workforce,
there is a large gap in each state’s tax effort. I investigate these differences in Lagos and Kano’s
tax extractive capacity using archival data and interviews with members of each state’s tax
administration. Evidence suggests that elite political initiative in Lagos has empowered the state
revenue board to engage citizens in a public campaign. As a result, tax officials have been
informing Lagosians of the benefits of taxation and the process of tax payment. Conversely, in
Kano, the revenue board is largely relegated by the political arms of government and denied
resources to launch such a campaign. As a result, Lagos and Kano have two different tax
cultures, which ultimately influence their capacity to extract taxes.
In the concluding chapter, I use the Afrobarometer to expand this research to other countries
in Africa. Beyond Nigeria, Africans also recognize that taxation is a fiscal contract with elected
officials. Across Africa, when individuals are satisfied with representation from elected officials
and government performance, they are more willing to pay taxes and make actual payments. It
also holds that when an African trusts that others are complying with tax payment, she will be
more willing to pay taxes and report paying them.
This research suggests that Western European theories of revenue extraction and the fiscal
contract between elites and citizens can be expanded to Africa if these models also consider nontax income. Furthermore, once accounting for varied revenue sources, these theories are useful
in explaining attitudes, political behavior, and decision-making. In Nigeria, and across Africa,
taxation results in representative governance. This link suggests that building citizen capacity to
monitor government budgeting can bolster responsive government.

9

CHAPTER ONE

Theoretical Framework: Explaining Revenue and Representation Linkages
Background: Representation and Democracy
Under what conditions is the development of representative institutions, a key facet of the
practice of democracy, most likely to occur? In the following section, I investigate the role
representation plays in the entrenchment of democracy. Furthermore, I explore how government
access to different forms of revenue influence the expansion of representative institutions and
ultimately democracy in developing contexts.
Democracy is defined as a political regime where citizens are able to choose or remove
political leaders via competitive multi-party elections (Dahl, 1977/1989; Huntington 1991;
Schmitter and Karl, 1991; Bratton and Van de Walle, 1997). Researchers propose several
explanatory factors to account for the variation in successful democratic consolidation after
transition from an authoritarian regime. These variables include a shared sense of national
identity among citizens (Rustow, 1970; Mann, 2004), elections (Huntington, 1991; Lindberg,
2006), electoral systems (Lijphart, 1984; Reynolds, 1999), and civil society (Diamond, 1992;
Putnam, 1993). However, these theories of democratic consolidation in new democracies do not
consider how the formation of representative institutions also contributes to democratic
governance. Aldrich (1995: 3) argues that, “to be truly democratic, it is necessary for any
nation’s leadership to be harnessed to public desires and aspiration.” Representation, defined as
the process in which elected officials work to convert popular demands into public policy,
provides this harness (Aldrich, 1995; Stokes, 1999).
As Dahl (1989), Huber and Powell (1994), and Powell (2004) argue, representation is key

10

to the practice of modern democracy. According to Huber and Powell (1994: 292), “the
congruence between the preferences of citizens and the actions of policymakers constitutes a
major claim and goal of liberal democracy.” Through the practice of representation, elites in
functioning democracies channel divergent interests into government, sorting different interests,
and executing the majority will. Powell (2004) expands this, arguing that mere responsiveness is
not enough: institutions must be in place to reliably ensure that the process of representation
occurs. Political parties and elections have been cited as chief coordinators of politics, filtering
preferences, and creating a cohesive strategy for the eventual institutionalization of majority
preferences. Thus, the development of representative institutions is a vital element of
democracy.
For example, legislatures have been identified as representative institutions that contribute
to the development of democracy. In fact, some scholars argue that the success of a country’s
democracy hinges on the strength and competence of the legislature. According to Fish (2006:
5), “the presence of a powerful legislature is an unmixed blessing for democratization.”
Utilizing an index of parliamentary powers (Parliamentary Power Index, PPI), Fish (2006)
demonstrates when a country has a strong legislature it is more likely to have a strong
1

democracy. A weak legislature undermines democracy in two ways: first, it undermines
“horizontal accountability…[that is] the controls that state agencies are supposed to exercise over
other state agencies” (Fish, 2006: 12). If the legislature is unable to provide checks on executive
power, including in the bureaucracy, abusive practices are more likely to occur. Second, weak
legislatures destabilize “the growth of vertical accountability, meaning the ability of the people
to control their representatives” (Fish, 2006: 13). A strong legislature is more capable of linking
1

Legislative strength is measured with the “Parliamentary Powers Index (PPI), based on 32
items that cover the parliament’s ability to monitor the president and the bureaucracy,
11

elected officials with their constituents and acting upon citizens’ preferences. Moreover, vertical
accountability affords voters the ability to reward or punish her representatives based upon the
perception that her expectations have been met.
In addition to serving as mechanisms for accountability, legislatures are “the one institution
explicitly established to represent society’s diverse interests in government” (Barkan, 2009: 1).
Barkan (2009: 7) goes on to identify representation as a legislature’s fundamental function:
“Regardless of the type of electoral system by which the members of the legislature gain
their seats, the main purpose of individual legislators and the body to which they belong is
to represent, that is to say re-present or mimic the varied and conflicting interests extant in
society a whole. The legislature is the institutional arena where representatives of
competing interests articulate and strive to advance their respective objectives in the
policymaking process.”
Thus, legislatures serve are an example of an important representative institution that allows
elected officials to engage, manage, and act upon citizens’ interests.
The next section continues this discussion, exploring conditions under which political
representation is most likely to develop. Specifically, two questions delineate the theoretical
framework:
1. What is political representation?
2. Historically, what conditions are most conducive to the development of representative
government?

What is representation?: Linking Citizen and Elite Preferences
Political scientists (e.g. Huber and Powell, 1994; Aldrich, 1995; Stokes, 1999) define
representation as the incorporation of citizen interests into political decision-making. This
perspective takes for granted that an elected official’s primary function is to ascertain majority
will and then translate that will into political outcomes.
12

On the other hand, scholars of Western political philosophy engage in a lively debate about
the purposes of political representation, shedding light on alternative views. For example,
instead of arguing that elected officials merely reflect the citizens’ interests, Burke (1770, 1774)
asserts that an elected official cannot simply mirror their citizens’ preferences. Being better
informed, these politicians must make use of their own judgment and understanding to make
decisions that are truly in the collective’s best interest. Burke (1774) concludes that
representatives are not agents, but trustees and enforcers of one national interest. As such, they
must use their faculties to ensure that the common good prevails. However, Hamilton/Madison
(1788) argue “frequent elections are unquestionably the only policy by which [elected officials’]
dependence and sympathy [with citizens] can be effectually secured…binding the representatives
to their constituents.” This institutional structure creates a circumstance where an elected
official, at least temporarily, is concerned with her constituency’s interests. Thus, while officials
may be elected for the strength of their judgment and understanding, they must ultimately pursue
constituent interests (if only for the purpose of reelection).
Bratton (2009: 4) refers to representation as “the role of elected officials in conveying
popular demands onward to deliberative and decision-making bodies.” This is differentiated
from responsiveness, a leaders’ the willingness of to pay attention to citizen preferences and
demands. Representation is also separate from accountability, which is citizen assessment of
elected officials’ performance, for example by reelection or removal of those politicians via
periodic elections. While these three concepts are distinct, they are, in fact, related.
Accountability, occurs when “first, there is an understanding that A is obliged to act on
behalf of B. Second B is empowered by some formal institutional or perhaps informal rules to
sanction or reward A for her activities or performance” (Fearon, 1999). Following Bratton

13

(2009), when considering government accountability, elections serve as one instrument by which
citizens (B) are able to hold elected officials (A) to task. Thus, government accountability is
performance-based. Citizens’ retrospectively observe elected officials’ activity and determine
whether to re-elect (reward) or remove them from office (sanction) (e.g. Persson et al, 1997;
Cheibub and Pzeworski, 1999; Ferejohn, 1999; Manin et al, 1999).
While accountability is a citizen’s ability to demand government performance,
representation is the process by which elected officials act on behalf of constituents, gauging,
deliberating upon, and eventually incorporating citizen interests in political decision-making
(e.g. Huber and Powell, 1994; Aldrich, 1995; Stokes, 1999). However, following Pitkin’s (1967)
definition of representation, which stipulates that elected officials act in the best interest of the
people, I argue that responsiveness—paying attention to citizen interests—is nested within
representation, allowing officials to carry out the procedure of representation. In order for
leaders to confer citizen preferences into public policy, they must first be willing to take note to
their constituents and remain active listeners.
In a similar vein, principal-agent theory elaborates upon the relationship between elected
officials and citizens, and the circumstances where these elites are compelled to act in the best
interest of their constituents.
In order to coordinate collective action for the purpose of governing a society, classical
principal-agent theory suggests that citizens take advantage of expert specialization, delegating
decision-making duties to elected representatives (Pitkin, 1967; Calvert et al 1989; Pzeworski, et
al, 1999). In this case, citizens are principals, while political officials are intended to serve as
their agents. In essence, this is a contractual relationship. Principals seek to mold agents’
behavior so that they will consistently act in accordance with principals' preferences. However,

14

an information asymmetry exists, whereby agents possess expert policy knowledge that
principals do not. Not only are elected officials equipped with the authority to make laws in a
host of policy arenas, but also these political elites can compel citizens to comply with policy
decisions. Since elected officials (agents) enjoy information advantages over their constituents
(principals), elites must be policed in order to ensure adherence to the “public interest” and other
goals of the citizens-principals. Principals pay specification costs to identify acts of the agent
that would satisfy the principal's preferences and policing costs in monitoring and enforcing
compliance (Mitnick 1973, 1975, 1980: 150).
Given these circumstances, we must consider various tools citizens can use to ensure that
elected officials act in their best interests? In other words, how can citizens ensure that they are
being represented in government (Pitkin, 1967; Powell, 2004)?
Evaluating how political arms of government exert influence over non-elected
bureaucracies, another classic principal – agent relationship, can help shed light on how citizens
can, in turn, influence politicians. Primarily, these scholars explore how legislatures and
executives regulate bureaucratic behavior (in the U.S. context).
Researchers have asked: “who controls the bureaucracy?” As a result of bureaucratic
policy expertise, information asymmetry, agency mission, and constituent support, bureaucratic
agencies have some autonomy in policy making (Moe, 1984; Rourke 1984; Meier, 1985).
Moreover, bureaucrats have their own interests and goals, separate from Congress and the
president (Ogul, 1976; Wilson, 1980; Weingast, 1981). Given these circumstances, how do
elected officials exert authority over the bureaucracy?
Both Congress and the president can exert influence over bureaucratic behavior through a
variety of tools, including: nomination/approval/dismissal of bureaucrats, setting agency budgets,

15

the presidential veto, and Congressional overrides of vetoes (Calvert, McCubbins, and Weingast,
1989; Hammond and Knott, 1996). Ex-ante controls are tools that allow politicians to influence
bureaucratic behavior before agencies take a particular policy action. Congress and the president
make use of this type of control by requiring agencies to engage in public hearings, burden of
proof requirements, and Congressional or executive over-sight committees monitoring agency
activities (McCubbins, Noll, and Weingast 1987; Moe 1982, 1984, 1987; Hammond, 1986;
Epstein and O’Halloran, 1995; Schneider and Jacoby, 1996; Furlong, 1998). Ex-post controls,
allowing politicians to sanction or reward agencies after they take a policy action, incorporate
budget increases or sanctions and the reliance on constituents to voice discontent with agency
actions (Weingast, 1981, 1984; Weingast and Moran, 1983; McCubbins and Schwartz, 1984;
Calvert, Moran, and Weingast, 1989).
We are able to apply these types of controls (ex-ante, ex-post) to consider how citizens can
shape the incentives for elected officials to act in accordance with citizens’ various interests.
Free and fair elections have been cited as an ex-post tool citizens have to exercise accountability
over politicians. Citizens can vote for candidates who act on their constituents’ interests, while
throwing out officials who do not adhere to expectations (Fearon, 1999; Manin, Pzeworksi, and
Stokes, 1999). However, when asked about how well elections “ensure that the Representatives
to the National Assembly reflect the views of voters,” 71% of Nigerians report “not very well” or
“not at all” (Afrobarometer, 2008). Furthermore, international reports characterize Nigerian
presidential, legislative, and state elections (1999, 2003, 2007) as “marred by massive fraud, vote
rigging, and violence” (Freedom House, 2008). Moreover, when we specifically consider
Nigeria’s National Assembly, there is a high turnover rate of incumbent legislators. According
to Lewis (2011: 7):

16

“Each successive legislature in Nigeria since 1999 has reflected substantial turnover.
Incumbent legislators typically lose their seats by failing to be re-nominated by their
parties, rather than through defeat at the polls. [For example], Nigeria’s 2007
Assembly retained less than 20 percent of the members from the preceding
legislature…In 2011, about a third of incumbents in the Senate and a quarter of the
House of Representatives were returned to office.”
In this context, accountability, that is, rewarding/sanctioning officials via elections, is difficult to
carry out. If the threat of ex-post punishment is minimal, what other incentive does an elected
representative have to act in accordance with their constituents?
I propose that citizens have another tool at their disposal to ensure representation: the fiscal
contract: tax payment. As an ex-ante control, citizens make tax payments to government in
expectation that political outcomes will reflect citizen preferences (Huber and Powell, 1994).
This reciprocity has been extended to encompass tax payment in exchange for government
provision of public services (Levi, 1988; Brautigam, et al, 2008). The next section details the
nature of the representation-taxation link. I hypothesize that, as elected officials become fiscally
dependent on their constituents to fund government policies, they are more likely to cede to
citizens’ policy interests. When citizens pay taxes, they begin to demand an account of how their
financial contributions are spent. As a result, elected officials become increasingly willing to
yield to their citizens’ expectations, minimizing the chance that citizens will shirk or revolt
against tax payment.

17

Historically, what conditions are most conducive to the development of representative
government?
Examples from the Western Context
Scholars have continually linked the development of representative institutions in
Western Europe to the process of revenue extraction (i.e. taxation). This model begins in the
15th and 16th Centuries, when Western European rulers engaged in war, border protection, and
forging nation-states. However, a major limitation that these rulers faced was access to revenue.
According to Levi (1988: 2), access to “revenue enhances the ability of rules to elaborate the
institutions of the state [and] to bring more people within the domain of those institutions.” In
order to raise income to engage in these activities that shore up and extend their authority, rulers
turn to taxing citizens. Specifically, rulers and citizens entered into a fiscal contract: citizens
agreed to provide tax revenue in exchange for an enhanced role in government. With taxation
came the incentive for political leaders to shift policy toward citizen interests (Tilly, 1985/1990;
Bates and Lien, 1985; Levi, 1988; North and Weingast, 1989; Acemoglu and Robinson, 2005).
Tilly (1985; 1990), Bates and Lien (1985), Levi (1988), and North and Weingast (1989)
pioneer this revenue-driven perspective of representation for Western European. These scholars
regard rulers as rational actors with their own particular interests. A ruler’s ability to maximize
his or her own utility is intertwined with the ability to acquire income. War-making, statemaking, and protecting borders are the key activities for which revenue is necessary (Tilly,
1985). Thus, the state is by nature predatory: constantly “[attempting] to set the terms of trade
that maximize their personal objectives…[requiring] them to maximize state revenue” (Levi,
1988: 10).

18

A ruler’s ability to maximize and guarantee tax income depends on bargaining with
citizens for compliance (Bates and Lien, 1985; Levi, 1988). However, mass compliance relies
on a given taxpayer’s perception that other taxpayers are compliant with payment, and that the
ruler will provide benefits (e.g. providing military defense, justice; Levi, 1988). Thus, when
citizens’ have more favorable evaluations of government performance, they are more likely to
comply with demands for tax payment. Therefore, rulers have an incentive to exchange (public)
services for tax revenue (Timmons, 2005). Representation—once again defined as the
incorporation of citizen interests in political decision-making—results from this bargaining
process between rulers and citizens. Representative institutions grant legitimacy to rulers’
demand for revenue (Hoffman and Norberg, 1994).
Scholars of the fiscal contract also elaborate upon the explicit nature of the bargaining
process between rulers and citizens. Bates and Lien (1985) demonstrate how, in Western
Europe, when monarchs pursued the taxation of “moveable” property (e.g. cows, oxen, grain,
which could be transferred from one location to another), rulers directly bargained with property
owners. In exchange for a guaranteed tax base, monarchs offered deference to these citizens’
policy preferences. North and Weingast (1989: 817) provide the example of England’s 1688
Glorious Revolution, and argue “in exchange for the greater say in government, Parliament
agreed to put the government on a sound financial footing. That is, they agreed to provide
sufficient tax revenue.”
Other historical examples suggest that taxation of citizens in order to obtain revenue for
government activities that then leads to political representation of citizens’ interests. Levi (1988:
100), describing the Western Europe (e.g. France and England) during the establishment of
national tax systems, argues, “[creating] a fiscal system adequate to the new pressures

19

confronting monarchy required regular taxation [necessitated] a new conception of royal power.”
At the time “medieval monarchs lacked the right and the power to impose taxes at will” (Levi,
1988:105). As a result, rulers were forced into bargaining with citizens for access to tax revenue.
According to Tilly (1990: 101):
“Bargaining took many forms: pleading with parliaments, buying off city officials
with tax exemptions, confirming guild privileges in return for loans or fees,
regularizing the assessment and collection of taxes against the guarantee of
[citizens’] more willing payment, and so on. All this bargaining created or confirmed
individual and collective claims on the state, individual and collective rights vis-à-vis
the state, and obligations of the state to it’s citizens. It also created rights—
recognized enforceable claims—of states with respect to their citizens.”
Levi (1988) provides a concrete illustration of this process via a discussion of the
introduction of the direct income tax in Britain in 1799. “Throughout the eighteenth century,
costly foreign conflicts put increasing stress on British revenue production,” including the Seven
Years’ War (1756 – 63), the American Revolutionary War (1775 – 84), and involvement in the
French Napoleonic Wars (1793 – 1815) (Levi, 1988: 125 – 127). Levi argues that the British
government was successful in implementing the direct income tax because they provided citizens
with specific assurances: “government successfully convinced the citizenry that the income tax
was necessary to finance a popular, if costly, [Napoleonic] war” (1988: 137 - 138). Furthermore,
government assured that the funds would be used to support the military in carrying out a
successful campaign. Last, following a reorganization of the tax administration, government
could credibly guarantee that “all citizens would pay their share and that government agencies
would be honest; no one would be a sucker.” With this example, Levi (1988) demonstrates the
overt nature of the fiscal contract: in exchange for income tax payment, citizens receive
guarantees that government will act upon constituents’ policy interests (in this case, the Britain’s

20

2

successful participation in the Napoleonic wars). Thus, historically, representative government
is a result of the tax contract and explicit bargains between leaders and citizens. In other words,
representation emerged as a result of the taxation of citizens.
Bates (2008) extends the characterization of governments as “predatory” to the African
context, expanding upon the forms of revenue available to leaders. He argues that governments
have two options for acquiring income: leaders can extract revenue from citizens via taxation in
exchange for service provision, or through the predation of state institutions. Moreover, there is
a tradeoff between taxation and predation. As long as the benefits of taxation outweigh what
could be derived from the short-term gains from looting, officials will maintain their bargain
with citizens. However, Bates argues that in sub-Saharan Africa, because “predation and
corruption seemed to offer greater rewards than providing stable administration, state leaders
undermined their own economies, eventually provoking conflict and rebellions” (Bates, 2009: 5).
In these instances, predation is preferred for three reasons. First, the energy crisis of the 1980’s
created a crisis in public revenues in many African governments. Second, in the post-Cold War
era, when African governments returned to democratic practice, incumbents were confronted
with the possibility of shorter political horizons. Finally, many African countries had access to
(lootable) natural resources. These three circumstances, taken together, fostered a situation at the
end of the 20th Century where African leaders viewed predation as more attractive option that
building political stability and order.

2

Later chapters will demonstrate that the Nigerian state governments most successful in revenue
mobilization via taxation of citizens engage in a similar campaign. They explicitly bargain with
citizens in a tax contract. Having effectively overhauled the tax administration, government
educates citizens about the overall tax structure and the responsibility of tax compliance.
Moreover, these state governments assure citizens that their financial contributions will be used
for policies that benefit public interest.
21

As Bates (2008) argues, I contend that the Western European version of the “predatory
state” framework does not consider how the availability of a non-tax source of income (e.g.
natural resource revenue) would influence the development of representation. Rulers, as rational
actors seeking to maximize their own utility, will also seek the most cost-effective method of
acquiring income. In this case, access to an external revenue base, provides an avenue that is
less costly than bargaining with citizens. Rulers can avoid the costs of building tax capacity and
maintaining institutions for tax extraction and enforcing compliance. Therefore, leaders who are
able fund their governments without taxation can also bypass the fiscal contract with citizens, no
longer deferring to citizen interests. I hypothesize that dependence on natural resource revenue
ultimately stymies the development of representative institutions.
It is important to account for the impact of a lucrative, non-domestic revenue base (i.e.
natural resource rents in sub-Saharan Africa) on the interaction between political leaders and
their citizens. I hypothesize that, with the availability of resource wealth, political leaders lack
the incentive to negotiate with citizens. In other words, income from the sale of natural
resources interrupts the link between taxation and representation. Incorporating natural resource
rents into the revenue-centered perspective allows us to consider the circumstances under which
political representation will or will not emerge.

Taxation and Representation in Africa
Scholars have begun to apply the fiscal contract model in developing contexts. They
specifically consider linkages between taxation and representation, and performance-based
bargaining between governments and citizens in Africa.

22

Gibson and Hoffman (2006) investigate how sources of revenue influence government
performance and spending in Africa. Using budgetary data from Tanzanian and Zambian local
governments, they find that “revenue derived from citizens will induce politicians to expend
more funds on public services” (Gibson and Hoffman, 2006: 7). On the other hand, when local
government income is derived from sources outside of the taxation of its local constituency,
politicians are more likely to spend revenue on government salaries, allowances, and other
recurrent expenditures. Gibson and Hoffman (2006) find that non-tax based income from federal
government transfers and foreign aid are particularly linked to salary expenditures. Evidence
suggests African elected officials are more willing to utilize tax revenue to cater to public
interests. However the converse argument also holds: politicians are more likely to spend nontax income on themselves.
Guyer (1991: 14) characterizes the effects of limited taxation on the representation of local
interests in Nigeria:
“With such low contributions in rural Nigeria, financial management becomes a poor
basis for people’s demands for accountability…with no policy-making about, or
financial instruments for, local development of locally defined projects within the
government system, the extension of national plans to local area becomes an act of
fate whose financing bears no relation to the population affected by them.”
More recently, using Afrobarometer and Demographic and Health survey data, Berger
(2009) finds that Nigerian local governments areas (LGAs) that were forced to collect taxes
during the British colonial period generated stronger institutional capacity. This still manifests
today in the form of effective bureaucracies. Furthermore, citizens in LGAs with a historically
robust tax extractive capacity also perceive public service delivery as more effective and more
satisfactory when compared to their counterparts in non-tax dependent areas.

23

Lieberman (2003) and Timmons (2005) explore how relationships between upper and
lower income groups influence compliance with the tax contract. Applying this model to Brazil
and South Africa, Lieberman (2003) argues that tax policy involves the redistribution of private
wealth. When upper-income groups feel a sense of shared identity with lower-income
individuals in society, they will be more compliant with this contract. Lieberman (2003) asserts
that this is what happened in South Africa. South African leaders used race to bridge class
differences between affluent and underprivileged whites. Conversely, when there is
fragmentation in political community, and upper-income groups feel divided from other citizens,
they will be less acquiescent in tax payment. Again, Lieberman (2003) argues that in Brazil,
hyper-regionalism created polarization and fragmentation within and between income classes.
On the other hand, Timmons (2005) argues that there is an inherent divergence between
upper-income and lower-income group preferences. Acemoglu and Robinson (2005) support
Timmons’ (2005) assertion, arguing, “individuals have well-defined preferences…[which are]
inherently conflictual.” According to Timmons (2005), upper-income groups prefer policies that
provide the protection of property rights. However, lower-income groups prefer government
3

provision of basic public services, such as health care and social welfare spending. Thus, “if a
state begins taxing a group it has strong incentives to provide that group with benefits to
maintain that source of revenue” (Timmons, 2005: 531). However, “if a state is not taxing a
group for some reason, it has no incentive to cater to that group” (Timmons, 2005: 531).
Therefore, instead of political leaders acting as arbiters between upper-income and lower-income
groups (as Lieberman suggests), Timmons (2005: 562) finds “no evidence that governments
gouge the rich to benefit the poor (or vice versa).” His analysis of 90 countries determines that
3

Again, Acemoglu and Robinson (2005) corroborate Timmons (2005) assertion, claiming
poorer citizens prefer high taxes and redistribution in the form of social spending.
24

when tax revenue is primarily derived from corporate taxes, the government is more likely to
engage in the protection of property rights. Conversely, when tax income originates from
regressive taxes, government will engage in higher social spending. In these cases,
representation manifests through a translation of public needs into public policy (e.g. either
property rights protection or public service provision). Timmons (2005) shows that
representation results from government extraction of tax income from their citizenry. Moreover,
the composition of the group(s) that the government primarily taxes (either upper-income or
lower-income) will also determine the type of policy that politicians will execute. Timmons
(2005) further demonstrates that it is possible to gauge representation of a particular group
through congruence between citizen demands and public policy.
Bates and Humphreys (2005) also consider the circumstances under which government is
likely to expend resources to provide citizens with public goods: “Subgroups of citizens are
[considered to be] decisive in the sense that they can ensure the government’s tenure in office”
(Bates and Humphreys, 2005: 409). Furthermore, Bates and Humphreys (2005) discuss a game
that occurs between politicians and the "decisive" group. Depending on the nature of this group
and their demands, government is more likely to provide services to these citizens rather than
diverting resources for themselves. This work demonstrates how resource distribution is used to
ensure political tenure. Elites provide services and benefits to decisive groups of citizens as a
primary method for maintaining political loyalty.
Scholars also investigate the link between taxation, government capacity to provide
services, and democracy. Ross (2004) performs a cross-national study to examine if the need for
revenue (and taxation) results in democratization. He demonstrates that citizens have
preferences concerning rates of taxation and how government spends these funds. Furthermore,

25

citizens think of taxes in the context of the benefits and goods they receive: “When the price of
government services goes up, authoritarian regimes tend to become—or are forced to become—
more accountable to their citizens,” resulting in democratic practice (Ross, 2004: 247).
Bratton and Chang (2006: 1061) explore the relationship between state building and
democratization, finding that “[they] are best viewed neither as occurring forwards or backwards
but rather reciprocally or together.” When citizens view government as capable of solving
pertinent problems, looking after citizens’ interests, and providing fundamental services, they are
also more likely to perceive higher levels of democracy. Government provision of necessary
goods and resources remains pertinent to both representation and the development of democracy.
This previous work suggests that the mode of public revenue generation, particularly
government’s capacity (or incapacity) to extract tax-based income from citizens, is a significant
predictor of politics in Africa, and more specifically in Nigeria. Specifically, the source of
government income influences if and how politicians address citizens’ interests and public
service provision.
The next section discusses natural resource wealth as an alternative to tax income, and its
influence on political behavior in Africa.

Natural Resource Wealth and Representation in Africa
The concept of a “natural resource curse,” also known as the “oil or mineral curse,”
originates in studies of democratization in the Middle East. The dilemma arises in countries that
are endowed with substantial resource deposits and derive a significant portion of revenue from
the sale of these commodities. These resources include, but are not limited to, petroleum and
26

other fuels, metal ores, diamonds and precious stones. In his most recent work, Ross (2012: 5)
outlines four characteristics of natural resource revenue that make it particularly distinctive from
other forms of government income: “their scale, source, stability, and secrecy.” These qualities
interact and result in politics and economics peculiar to these resource-rich countries.
Ross (2012: 5) argues that the scale of natural resource wealth is massive, providing
governments access to an immense amount of revenue: “On average, the governments of oilproducing countries are almost 50 percent larger (as a fraction of their country’s economy) than
the governments of non-oil countries.” Circumstances following recent discoveries of petroleum
in Asia and Africa corroborate this argument. For example, between 2001 and 2009, in
Azerbaijan, total government expenditures rose by 600% while in Equatorial Guinea, they rose
by 800% (Ross, 2012: 5). The source of this large income is also salient. These countries rely
on resource wealth and not taxes on their citizens. As a result, “they become less susceptible to
public pressure” (Ross, 2012: 5). Governments in resource-dependent countries are no longer
constrained by their citizens, which funding via taxes would engender. The stability “or rather
the instability” of oil revenues is also particular to this form of government income: “The
volatility of world oil prices, and the rise and fall of a country’s reserves, can produce large
fluctuations in a government’s finances” (Ross, 2012: 6). Reliance on revenue derived from an
unstable source often results in poor planning and management and squandering of resource
wealth. The last characteristic peculiar to this form of income is the secrecy attached to natural
resource revenues. According to Ross (2012: 6), “governments often collude with international
oil companies to conceal their transactions, and use their own national oil companies to hide both
revenues and expenditures.” Furthermore, this secrecy accounts for why natural resource income
is squandered or channeled through corrupt channels. The scale, source, stability, and secrecy of

27

natural resource wealth ultimately influences politics in these dependent countries in a way
different from those relying on other forms of income.
One mechanism used to explain how the “curse” hinders democracy is the “rentier” state.
In a “rentier” state, government is able to fund public activities through profits from the sale of
natural resources and taxes on corporations involved in resource production. Since elites can
fund themselves with rents paid by foreign companies, political leaders can bypass domestic,
public interests. This ultimately diminishes elites’ incentive to represent their constituents
(Beblawi and Luciani , 1987; Ross, 2001, 2004, 2012). Researchers explore the “natural
resource curse,” hypothesizing that a country’s dependence on resource wealth is associated with
non-democratic regimes, reduction in political competition, decline in the quality of governance,
civil war and disorder, weak political institutions, and declines in economic growth and
development. A body of work, based on large-N, statistical analyses and small-N/case studies
has emerged to investigate these hypotheses (Beblawi, 1987; Karl, 1997; Collier and Hoeffler,
1998/2004; Wantchekon, 1999; Ross, 2001/2004/2006/2012; Welden, 2001; Lam and
Wantchekon, 1999; Fearon and Laitin, 2003; Jensen and Wantchekon, 2004; Humphreys, 2005;
Lujala et al., 2005; Snyder and Bhavnani, 2005; Jones-Luong and Weinthal, 2006; Dunning,
2008; Goldberg, et al, 2008; Lujala, 2010; Bodea, 2012).
Ross (2001) investigates the validity of the “oil-impedes-democracy” claim, performing
statistical analyses of 140 countries across 25 years. He demonstrates that reliance on profits
from natural gas, coal, and precious stones is negatively associated with democratization, and
furthermore this is not concentrated to any particular geographic location. After uncovering the
validity of this effect, Ross (2001) attempts to uncover the underlying mechanisms. He finds
that resource-dependent governments tax their citizens less, consume more, and have higher

28

levels of military spending, which all work to curb popular demand for democracy. Similarly
Wantchekon (2002) uses a small-N case comparison (Norway, Botswana, Nigeria) and large-N
analysis of 120 countries to investigate the relationship between oil wealth and autocracy. He
finds that natural resource wealth sustains authoritarian regimes through the lack of transparent
budgeting processes and incumbents’ use political repression to crush their opposition.
Resource dependence is also associated with low quality of governance and weak
political institutions (Karl, 1997; Lam and Wantchekon, 1999; Jones Luong and Weinthal, 2006;
Welden, 2001; Wantchekon, 2002). For example, Karl (1997: 7) argues that “dependence on a
particular export commodity shapes not only social classes and regime types…but also the very
institutions of the state, the framework for decision-making, and the decision calculus of
policymakers.” In her case study of institutional development in Venezuela, Karl (1997) finds
that wealth from oil created vested interests, sustained the status quo, and barred opposition to
the agreements created between political elites. At the time of democratic transition in 1958, the
two dominant political parties (Accion Democratica and COPEI) bargained and compromised,
resulting in a pacted form of democracy. This elite political settlement expanded the role of the
state, enforced patterns of patronage distribution, and maintained the status quo through policy
rigidity (Karl, 1997: 93). As a result, the state remained large, rent-seeking, and interventionist,
while military spending increased without any sort of civilian check. In Venezuela, reliance on
petroleum facilitated the development of a group that profited from the system of revenue flows,
which was reinforced through the institutional framework.
Karl’s (1997: 101) analysis of petro-politics in Venezuela demonstrates how petroleum
resources influenced the protection of the status quo, allowing for the maintenance of a large and
centralized “interventionist state and oil-led development.” This essentially limited the scope of

29

elite institutional choice at the time of democratic transition. On the other hand, this focus on
elite interaction distracts attention from the relations between the state elites and citizens. We
are unsure how petroleum dependence influences the relationship between the political
leadership and citizens, the realization of citizens’ interests, and policy outcomes (Bratton and
van de Walle, 1997; Tripp, 1997).
Jones Luong and Weinthal (2006/2010) also challenge Karl’s (1997) state-centered focus.
They argue that variation in ownership over natural resource development influences the
development of fiscal and regulatory institutions. Following an examination of the Russian
Federation, Kazakhstan, Uzbekistan, and Azerbaijan, the authors differentiate between state
ownership with control, state ownership without control, private domestic ownership, and private
foreign ownership, finding that:
“strong fiscal and regulatory institutions are more likely to emerge under private
domestic ownership because it creates a set of actors who have a mutual interest
in establishing formal guarantees to increase fiscal predictability and reduce
transaction and monitoring costs” (Jones Luong and Weinthal, 2006: 242).
Therefore, Jones Luong and Weinthal (2006/2010) argue that the structure of ownership can
serve as an intervening variable between resource wealth and weak political institutions.
Although they propose the new variable of resource ownership, which can also account for the
resource curse, this explanation is not widely applicable in the developing world. For example,
by Jones Luong and Weinthal’s (2006) own admission “in most developing countries, petroleum
resources are managed through state oil companies.” Thus, it is difficult to expand Jones Luong
and Weinthal’s (2006/2010) model to other developing contexts also dealing with the resource
curse. Even though they are critical of Karl’s (1997) argument, it seems the Venezeuelan model
reflects the reality of the prevalence of state ownership of natural resource development.
30

Natural resource wealth has also been cited as a barrier to political order, thus increasing
the likelihood of civil conflict. While wealth from oil has been identified as a contributor to the
start of civil war, wealth from precious stones do not increase the chances of conflict, but rather
can sustain a war that has begun (Collier and Hoefeller, 1998/2004; Fearon and Laitin, 2003;
Ross 2004; Humphreys, 2005). Snyder and Bhavnani (2005) argue that natural resources, to the
extent that they are not taxable, decrease institutional capacity and increase chances of regime
breakdown. Creating a typology, they compare a resource’s lootability (the ease with which the
resource can be extracted) with a government’s ability to tax production (do corporations or
4

individuals engage in resource extraction?). Snyder and Bhavnani (2005) determine that in
Sierra Leone, diamonds have high lootability (low costs of extraction. Furthermore, since
individuals are more likely to engage in diamond extraction, government’s ability to tax the
individual artisans is low. In this case, natural resources create a high risk for conflict. On the
other hand, management of bauxite resources in Ghana creates a low risk of conflict. Bauxite
has low lootability because of the high cost to extract the resource. Moreover, because
corporations are primarily involved in the bauxite industry, the ability to tax them is high. We
see that generally, research conforms negative effects of resource dependence of democracy,
institutional capacity, and political order.
With that said, more recent work argues that in certain circumstances, natural resources
can encourage democratization. For example, Dunning (2008) uses comparative case studies in
Latin America (Venezuela, Chile, Bolivia, Ecuador) to criticize the dominant “oil-impedesdemocracy” argument. He finds that in a situation where transition to democracy is initiated by a

4

Snyder and Bhavnani (2005) argue that it is easier for governments to to track and tax
corporations than individuals.
31

rift between political elites (hard-liners vs. soft-liners, O’Donnell and Schmitter, 1986), resource
rents provide a source of public spending other than redistribution from rich to poor. This helps
to reduce the economic cost of democracy, abating elite incentive to block or reverse
democratization efforts: “In a phrase, resource rents can underwrite democratic stability by
reducing polarization over economic policy and particularly over redistributive tax policy”
(Dunning, 2008:55).
Morrisson (2009) also explores the stabilizing effects of non-tax revenue on regimes. He
argues that “threats to [the stability of ] democracies come from wealthy elites, whereas threats
to dictatorships come from citizens” (Morrisson, 2009: 122). Analyzing 118 countries between
1973 and 2001, Morrisson (2009) finds that leaders in both democracies and autocracies use nontax revenue to stabilize their regime. In democracies, non-tax income is used to lower the tax
rate on wealthy elites. And in dictatorships, non-tax revenue is used to increase social spending
on poorer citizens.
While Dunning (2008) and Morrisson (2009) encourage us to consider conditions under
which, counter-intuitively, oil wealth can spur democracy, their arguments may not be applicable
across contexts. For example, Bratton and van de Walle (1997) reveal that in Africa, due to the
heritage of neo-patrimonialism, the mass public drives regime change over conflicts about
accessing spoils of the state patronage network. Unlike Dunning’s (2008) assumption, elite
behavior is not always the initiating factor in democratization. Moreover, Morrisson’s (2009)
argument ignores the role of patronage systems in the distribution of resources between political
elites and citizens. Rather than using natural resource income for social spending, I argue that
leaders in these neopatrimonial regimes are more likely to divert these funds for self-enrichment
and maintenance of their patronage networks. Thus, when considering the role of natural

32

resource wealth in African politics, popular access to patronage systems and the way in which
elites sustain those networks must be considered.
Following these critical works, I argue that more work is needed to uncover the
mechanisms of how oil wealth would undermine democratic practice. Herb (2003/2005)
suggests:
“In the literature on rentier states we find a good deal of theorizing about why the
absence of taxation prevents democracy: what is needed is a convincing account
of how taxation leads to democracy in modern states that tax.”
This dissertation aims to address this gap. Previous research performs the important
preliminary task of unearthing the association between resource wealth and a non-democracy.
We know there is a link between low tax extraction (resulting from resource dependence) and
survival of non-democratic regimes. However, it is not enough to demonstrate that natural
resources have an “anti-democratic” effect. This investigation aims to uncover the mechanism
that allows for resource dependence to translate into a lack of democratic practice. In unpacking
this relationship, I hypothesize that when governments have a strong capacity to extract taxes,
political leaders will use the public agenda to address popular preferences. Taxation leads to
representation, which is inherent in democratic practice.

5

5

Scholars have also explored the endogenous nature of this relationship, investigating whether
democracies or dictatorships are more effective in tax collection (Bueno de Mesquita et al, 2003;
Cheibub, 1998; Coughlin et al, 1990; Fjeldstad, 2001; Haggard, 1990; Kasara, 2007; Levi, 1988;
Melzter and Richard, 1981; Olsen, 1993; Rakner, 2002; Ross, 2004; Weinstein, 2009). Haggard
(1990), Olsen (1993), Bueno de Mesquita et al (2003), and Kasara (2007) assert that political
competition will reduce government’s incentive and capacity to raise tax revenue. In
democracies, politicians must cater to voters who prefer to maximize their income and reduce
they’re tax burden. As a result, they will utilize low tax rates in order secure re-election. On the
other hand, since dictators are relatively autonomous from social pressures, they can maintain
high tax rates to extract the largest possible revenue from the private economy. Moreover,
dictators can ensure tax compliance via use of the police of military for coercion. However,
Meltzer and Richard (1981), Levi (1988), and Cheibub (1998) find that, in fact, democracies are
33

Analytical Focus of Study
Thus, I reiterate the central research question of this study: in sub-Saharan Africa, under
what conditions are elected officials most likely to represent their constituents? This larger
question breaks down to several component parts: What do elected officials think their job is?
How do these officials understand their relationship to constituents? Do these perceptions of
their responsibilities vary subnationally? How do elite expectations of job responsibilities
compare with citizens’ notions? How does source of government revenue influence elite
representation?
I hypothesize that in governments where a higher portion of income is derived from the
taxation of citizens, elected officials provide a greater level of representation to their
constituents’ priorities and interests. These officials will be more likely to expend public
resources on addressing citizens’ policy concerns via public service delivery. Moreover, where
leaders rely on taxation, they will be more likely to prioritize citizen policy interests. On the
other hand, when a higher portion of government income is derived from natural resource
wealth, political elites will be less likely to represent citizens. Officials in these resourcedependent contexts are less likely to spend public funds on policies benefitting their constituents
or prioritize citizens’ interests.
I expect that when tax extraction and capacity are greater, government is more likely to
prioritize constituents’ interests because it is more reliant on citizens to fund public policies. In
able to collect taxes at higher levels than dictatorships. In democracies, politicians have an
incentive to expand government size. Thus, they will increase tax rates in order to satisfy voters’
preferences for redistribution. Furthermore, representative institutions (more so than coercion)
reduce tax evasion by increasing “the monitoring of both rulers and taxpayers, promoting
cooperative arrangements among relevant actors, and permitting the establishment of realistic
and accepted sanctions for non-compliance” (Levi 1988, 179).
34

order ensure that citizens make tax payments and comply with the fiscal contract, elected official
will align their policy priorities and public expenditures with their constituents (Bates and Lien,
1985; Levi, 1988). However, the reverse relationship also holds: dependence on natural resource
revenue allows elected officials to carry out policy without fiscal reliance on their constituents.
In this case, politicians are not constrained by citizens and do not defer to citizens’ interests.
This ultimately results in a government less representative of its people.

35

CHAPTER TWO
Democracy, Governance, and Revenue Generation in Contemporary Nigeria
Background
After considering theoretical relationships between taxation, representation, and natural
resource dependence, I now investigate how these three variables have interacted in Nigeria.
Neopatrimonialism, the fusion of patronage systems and bureaucracy, has pervaded
Nigeria’s political institutions since the time of independence. This regime form ultimately
influences how the tiers of government interact with one another and how government interacts
with citizens. Nigeria’s political leadership, through cycles of civilian and military government,
has prioritized gaining access to revenue in order to distribute resources via patronage systems.
As a result, elites’ desire to access the primary mode of revenue generation motivated political
choices to centralize power in the national government or disperse authority through the federal
6

system. Thus, leaders of Nigeria’s (current) Fourth Republic find themselves attempting to
balance political and fiscal authority in a newly reestablished democratic, federal regime.
This chapter begins with a discussion of how the political leadership’s attempts to extract
revenue and control the dominant means of economic productivity influenced the breakdown and
success of democratic governance and federalism in Nigeria. Second, using Nigerian legislatures
as examples, I outline how, over time, elites have dispersed political authority to govern

6

According to Ogban-Iyam (1998): “Federalism is a system of political organization uniting
separate states of other units in such a way as to allow each to remain a political entity. A
federal system differs from other methods of organizing states in being based on a contractual
agreement by the separate governments to share power among themselves.”
36

throughout the national, state, and local levels of government. Finally, moving beyond political
federalism, I examine fiscal federalism in Nigeria. Here I consider how natural resource revenue
is accrued and distributed among the three levels of government. I also assess constitutional
powers of taxation and national, state, and local governments’ tax jurisdictions. Ultimately, this
chapter provides an overview of Nigeria’s political and fiscal contexts, allowing us to understand
how, historically, access to various forms of revenue has motivated politics in Nigeria. With this
knowledge, we can then consider the tools of revenue extraction available at different levels of
government in the contemporary era. This provides a foundation for later analyses of
subnational variation in tax capacity and the influence on political representation.

Neopatrimonialism, Federalism, and Citizen-Elite Relations in Nigeria: An Historical Overview
According to Bratton and van de Walle’s (1997) seminal work, neopatrimonialism, the
incorporation of patrimonial logic and traditional, informal authority into bureaucratic
institutions, is a “hallmark” of African politics. Three key features delineate neopatrimonal
regimes. First, there is a concentration of political power in one individual. This presidentialism
is fueled by a cult of personality, ensuring the leader’s political longevity, but weakening other
institutions. Second, neopatrimonial regimes engage in clientelism, where personal favors, jobs,
and resources are awarded in exchange for political support. Last, neopatrimonial regimes are
characterized by the leadership’s private use of state resources for political legitimation. In fact,
there is little distinction between public and private coffers. These three qualities interact to
undermine formal rules and institutions.
Since Nigeria’s independence in 1960, the country has undergone both civilian government
and military dictatorship. However, a neopatrimonial mode of administration serves as a

37

common thread linking 50 years of governance. During this time, the political regime has been
defined and altered by political elites’ attempts to gain access to revenue in order to sustain
patronage systems. Public funds have repeatedly been used to reinforce political dominance.
Therefore, particularly in Nigeria, “the political process is structured around distributional
contention and the capture of rents [e.g. government contracts, business deals, appointments to
state enterprises/ministries, illicit payments] rather than mechanisms of representation” (Lewis,
2009).

Sustaining Neopatrimonialism in Nigeria’s First Republic
In Nigeria’s First Republic (1960 – 1966), power over the dominant mode of economic
development was held subnationally. These leaders were able to translate financial power into
political gain, expanding subnational influence in such a way that the federal system and national
sovereignty were eventually challenged.
During this time, Nigeria utilized a federal parliamentary system comprised of a central
7

government and four (sub-national) regional governments. Though both levels of government
held executive, legislative, and judicial authority, fiscal power was concentrated at the regional
level through the institution of the marketing board. Marketing boards were left over from
British colonial administration. Created during World War II (in times of economic crisis), these
organizations were charged with regulation of the agricultural industry. They were intended to
use whatever funds they accumulated for the benefit of farming communities. At the time,
agriculture represented the primary economic activity in Africa, generating large amounts of
7

Four regions made up Nigeria’s sub-national system: Northern, Eastern, Western, MidWestern. The First Republic’s national and regional political structures are discussed in greater
detail later on, during the Legislatures and Government Revenue in Nigeria: Resurgence of
Political and Fiscal Federalism.
38

foreign exchange. Thus, these organizations became wealthy and influential. Marketing boards
also became a tool for regional politicians in need of revenue. Politicians would often divert
funds from these marketing boards into the public treasury (Bates, 1981: 12 – 13). Regional
leaders would go on to use revenue generated by marketing boards to fund patronage networks
and garner political support. This was achieved through “disburs[ing] loans, development funds,
and licenses in exchange for votes and kickbacks” (Welden, 2001). For example, in Western
Nigeria “persons in charge of development agencies used their powers transfer funds into banks
and corporations in which they held directorships [instead of toward agricultural subsidies],”
using the funds to award themselves large, interest-free loans (Bates, 1981: 100). Regional
politicians used financial influence to gain political leverage.
The concentration of fiscal and political influence at the regional level resulted in repeated
challenges to the Nigeria’s central authority. According to Suberu (2004: 331): “The regions
enjoyed the loyalty of their respective major ethnic communities [and] commanded relatively
8

substantial constitutional power and financial resources.” Moreover, “although the federal
government acquired more prestige and influence in relation to the regions,” it was clear that the
federation’s more talented politicians and bureaucrats remained at the regional level (Suberu,
2004: 331). Regional fiscal autonomy and political power contributed to the outbreak of
Nigeria’s Biafran Civil War (1967 – 1970), an attempt by the Igbo dominated Eastern Region to
secede from the federation. Conflict brought an end to the First Republic, as military leaders

8

In Nigeria, “demographically the Northern Region was predominantly Muslim, with an ethnic
Hausa-Fulani majority; the Western Region, mainly Yoruba, was roughly balanced among
Muslims and Christians; the Eastern Region was predominantly Igbo and overwhelmingly
Christian” (Lewis. 2011: 3).

39

from the Eastern and Northern Regions launched coups and counter-coups, either in support of
or in opposition to the secession attempt (Ihonvbere and Shaw, 1998).
The Eastern Region was unsuccessful in its challenge. Following the civil war, the military
government moved to minimize regional authority. Scrapping the regional system, Nigeria
9

expanded from 4 regions to 19 states. The military government also paid particular attention to
states’ fiscal independence. In an effort to minimize future secession attempts in the regions, the
military moved control of the marketing boards to the central government (Diamond, 1988;
Ihonvbere and Shaw, 1998). First, the power of individual sub-national entities to tax was
suspended. Furthermore, unlike in the First Republic, the national government now held control
over the sector leading economic development: the petroleum sector. As Nigeria “rode the crest
of a bounteous petroleum boom” in the 1970s, the rules of oil revenue allocation were amended
so that the national government took the bulk of these earnings (Lewis, 1996: 81). Like regional
leaders (who used revenue from marketing boards to sustain their patronage networks), national
political elites also relied on petroleum income to fund the neopatrimonal regime. During the
Gowon/Mohammed/Obasanjo military governments (1966 – 1979), “the rapid influx of cash
fostered a dramatic increase in corruption,” where 90% of the budget was used to fund inflated
military salaries and political appointments to the civil service (Lewis, 1996: 81). This
“result[ed] in an untrained, undereducated bureaucracy whose primary function was to support
the patron that granted the position and [their] secondary function was personal enrichment
(Welden, 2001: 70 – 71).

9

19 states in the Second Republic: Anambra, Bauchi, Bendel, Benue, Borno, Cross River,
Gongola, Imo, Kaduna, Kano, Kwara, Lagos, Niger, Ogun, Ondo, Oyo, Plateau, Rivers, Sokoto.
40

Neopatrimonialism and Democratic Breakdown in Nigeria’s Second Republic
After the military stepped aside, Nigeria returned to civilian rule in the Second Republic
(1979 – 1983). However, gaining access to revenue to fund patronage networks remained the
political priority. Early on in the Second Republic, “even more than in the First Republic,
government office was an opportunity for the enrichment of oneself and one’s supporter…not
only was all the wealth of the country in the hands of the [central government], but also it had
increased ten-fold” (Watts and Lubeck, 1983: 109). Scholars estimate that 60% of GDP was
used as patronage. Moreover, access to national oil revenues increased “financial dependency of
the states (and their localities) on the center [and the] consolidation of central political authority”
(Suberu, 2004: 334).
However, by 1982, revenues accruing to the national government declined as a result of a
drop in international petroleum prices. This coupled with a growing foreign debt fostered a
climate of economic decline. As a result, “patrons could no longer support broad client bases,
and concentrated instead on enriching themselves” (Lewis, 1997: 305). Since the funding of
patronage networks was now concentrated at the national level, a drop in central government
revenue damaged politicians’ ability to maintain political power via clientelist networks.
After uncovering evidence of this self-enrichment among civilian politicians, MajorGeneral Muhammadu Buhari led the military take-over of December 1983 (Welden, 2001).

10

Between 1983 and 1998, the military (under Generals Ibrahim Babangida and Sani Abacha)
embarked on a series of changes, which shifted Nigeria from “prebendalism, [or] decentralized
patrimonial rule, to predation, [that is] the consolidation of avaricious dictatorship” (Lewis,
1997: 80). For example, in an attempt to alleviate popular and elite opposition to economic
10

The Nigerian military discovered a £22 million kickback to civilian politicians for purchasing
eighteen Jaguar ground attack fighters (Williams, 1987; Graf, 1988; Welden, 2001).
41

reforms, Babangida’s government used a series of remunerative policies to distribute patronage
to vocal groups. These included the creation of new institutions, such as the Directorates for
Food, Roads, and Rural Infrastructure, and various community banks. Economic reform did
minimize the amount of government contracts, licenses, and employment, which were
traditionally used as patronage at the elite level. However, politicians devised new forms of
political patronage, including providing insider investment information and granting special
access to nascent markets during the privatization of public assets. Politicians also turned a
blind-eye while elite participation in illegal economic activity flourished (e.g. petroleum
smuggling, drug trafficking, international commercial fraud). During this time, “financial
corruption and the apportionment of privileged access…created private incentives for persons to
support the continuation” of the current regime (Reno, 1998: 184).
The “state retreat from citizens reflect[ed] the extent to which [the ruling elite] relied on
extensive personal networks, rather than effective institutions” (Reno, 1998: 153). In particular,
during General Abacha’s military government, there was no attempt to develop institutional
capacity to provide public social services. Instead Abacha opted to sustain his patronage
network though a “highly visible and extremely wealthy military-political class.” Using contract
awards within the petroleum industry, some of which the General negotiated himself, it is
estimated that $12 billion in patronage was distributed over a 6-year span (Reno, 1998: 198;
Welden, 2001: 81).
National government access to oil revenues has been the key to centralization of
clientelistic control at the cost of developing institutional capacity to govern. Furthermore:
“$5 to $10 billion a year in oil profits is controlled by a small group at the top and
distributed through patronage. To collect this income, the government need do nothing for
the people of for the domestic economy. It needn’t build roads, maintain infrastructure, or
build schools. It needn’t account for its spending nor discuss with representatives of the

42

people how funds should be distributed. It need to nothing except protect its partnerships
in oil production because this is where income is derived, not from taxing the domestic
economy” (Welden, 2001: 87).
Political Federalism and Devolution of Power in Nigeria’s Fourth Republic
Following General Abacha’s death in 1998, Nigeria returned to civilian government, and
since “federalism has been long recognized as the indispensable basis for Nigeria’s stability and
11

survival,” it was not surprising that the federal structure reemerged (Suberu, 2004: 328) . In
theory, a federal system would minimize regional and ethnic conflict by allowing subnational
units the ability to govern themselves. At the same time, a federal structure would maintain
overall stability in Nigeria’s national structure. However, “the oil-centric political economy,”
encouraged by years of “hypercentralized military rule,” has created tension in Nigerian politics
and institutions in the post-military era (Suberu, 2004: 329).
Subnational entities have repeatedly demanded the decentralization of political authority
and access to fiscal resources. For example, Nigerian governors in Southern states have called
for “true federalism,” localized control of resources, and state-led economic development. This
includes the decentralization of the national police force into state police units, increasing the
portion of oil income allocated to state and local governments, regional authority of natural
resources, and local control of public service provision (e.g. education, public housing, and
agriculture). However, this Southern demand for true federalism has prompted opposition not
only from the central government, but also from Northern political leaders. In particular, “the
landlocked and relatively more economically depressed north depends more heavily on the south

11

Elections for Third Republic “were convened under military rule in 1992, but the civilian
government was stillborn after the annulment of the presidential election in June 1993. General
Sani Abacha seized power in late 1993, and civilian rule was deferred until Abacha’s death in
1998 opened the door to a new political transition” (Lewis, 2011: 5).
43

and the present system of centralized, distributive federalism” (Suberu, 2004: 341). Thus, in
Nigeria’s (contemporary) Fourth Republic, national and state governments are continually
renegotiating the balance between national authority and decentralization of power.
It is important to note that devolution of political authority has been accompanied by a shift
in the way the political elites interact with ordinary citizens. Under neopatrimonial forms of
rule, scholars argue that citizens also participate in and expect clientelistic behaviors and seek
individualized benefits from elected officials (van de Walle, 2001; Lindberg 2006; Wantchekon,
2003). Bratton (2009: 16) analyzes using Afrobarometer public opinion data and finds that
almost one-fifth of Africans report making a “side-payment” to obtain documents. Specifically
in Nigeria, 27% of respondents admit to “paying a bribe within the last year for water or
sanitation.” With that said, recent work also finds that citizens, in fact, prefer the delivery of
public goods and services instead of individualized benefits (Young, 2009). In Nigeria, 60% of
respondents indicate that “in electing a representative to the National Assembly, [they] prefer to
vote for a candidate who can make policies that benefit everyone in [the] country (as opposed to
localized, private benefits)” (Afrobarometer, 2008).
As state and local governments emerge as more powerful players in governance and
resource allocation, we observe subnational variation in the incentives facing political elites and
the way these leaders interact with constituents. In the next section, using Nigeria’s legislatures
as an example, I outline this decentralization of political power, revenue sharing, and revenue
mobilization in the Fourth Republic.

44

Legislatures and Government Revenue in Nigeria: Resurgence of Political and Fiscal Federalism
Emergence of Legislative Autonomy in the National Assembly
As discussed in the previous chapter’s theoretical framework, legislatures and legislative
activity can serve as an indicator of the condition of affairs in a regime (Fish, 2006). Therefore,
in order to explore how federalism has developed in Nigeria, I use the relationship between
national, state, and local legislatures as an illustrative example. The following section
investigates the devolution of political and jurisdictional power within the federation vis-à-vis
the evolution of Nigeria’s legislatures.
Though Nigeria’s executive branch of government dominated political power during
previous periods of civilian government, legislatures have operated in the previous regimes.
In the First Republic (1960 – 1966), four regions (Northern Nigeria, Eastern Nigeria,
Western Nigeria, and Mid-Western Nigeria) formed “a federal parliamentary system based on
the British Westminster model.” The legislature consisted of a House of Representatives (lower
chamber, 312 members) and Senate (upper chamber, 30 members).

12

In the House, seats were

distributed through Nigeria’s three main regions on the basis of population. The majority party
or coalition within parliament selected the executive (president). According to the 1963
Constitution (Chapter 5, Part 4, Clause 69 – 83), the parliament enjoyed the power of lawmaking with respect to any areas on the Executive Legislative List, which included the
following:

12

According to the 1963 Constitution (Chapter 5, Part 1, Clause 42), twelve senators would be
selected from the regions (nominations by the Governor during a joint sitting of the regional
legislative bodies), four senators from the federal territory, and four senators selected by the
President of the Republic.
45

•
•
•
•
•
•
•

•
•
•
•
•
•
•

external affairs, diplomacy, implementation of treaties, defense
currency
customs and excise
control of exchange rate
major communications networks (railways, shipping, post, telegraphs, aviation)
deportation and extradition, immigration
higher education (specifically Universities of Ibadan and Lagos—including teaching
hospitals, the Nigerian Institution of Social and Economic Research, the Pharmacy
School at Yaba, the Forestry School at Ibadan, the Veterinary School at Vom)
confer titles of honor
income tax and estate duties
international trade and commerce
banking
electricity or gas
regulation and administration of trusts, monopolies, and estates
censorship and regulation of cinematographic films

In the First Republic, the national legislature was not only paralyzed by the politicized, ethnic
divisions within the body, but also by a lack of policy expertise. There were only three
parliamentary committees, and in each year of the First Republic, legislators served no more than
54 days, deferring the initiation and implementation of laws to the executive branch and regional
legislatures (Elaigwu, 2005).
Nigeria’s Second Republic (1979 – 1983) abandoned the Westminster model for a set of
institutions closer to the U.S. presidential system. The federal system expanded from four
regions to 19 states, thereby increasing the membership the National Assembly to 545 (from 342
in the First Republic’s Parliament). The legislature retained the House of Representatives (450
seats distributed based on population) and Senate (five from each of the states), elected from
single-member districts. Like in the First Republic, the National Assembly’s purview with
regard to the creation and passing of laws in several areas, now expanded to include:
•
•
•

national Census
fingerprints, identification, and criminal records
regulation of Local Government Councils

46

•
•
•
•
•

insurance
maritime shipping and navigation
meteorology
regulation of political parties
weights and measures

According to the 1979 Constitution, legislation could originate in either the House or
Senate. However, in order for a bill to pass into law, it had to gain a simple-majority in each
chamber. Moreover, the legislature now had the formal power to override an executive
(presidential) veto with a two-thirds majority in each chamber. Even with increased legal
authority, the Second Republic’s legislature was weak in practice. During the Second Republic’s
four-year lifespan, the legislature only introduced and passed one bill into law; the President
initiated all the other legislation (1979 Nigerian Constitution, Chapter 5, Part 1; Second
Schedule; Lewis, 2011).
With the reemergence of civilian government in 1999, Nigeria already had a framework
13

for a federal structure of governance in its Fourth Republic.

Nigeria’s 1999 Constitution

(Chapter 5, Part 1) invests legislative power in the National Assembly. In the Fourth Republic,
the number of states increased from 19 in the Second Republic to 36. Thus, the House of
Representatives (lower house) consists of 360 members elected from single-member districts (by
simple majority). The Senate (upper house) is made up of 109 members (three per state, one
14

from the Federal Capital).

The National Assembly holds the power to legislate in the same

13

“Elections for the National Assembly in the Third Republic were convened under military rule
in 1992, but never had legislative authority” (Lewis, 2011: 5).
14

36 states in the Fourth Republic: Abia, Adamawa, Akwa Ibom, Anambra, Bauchi, Bayelsa,
Benue, Borno, Cross River, Delta, Ebonyi, Edo, Ekiti, Enugu, Gombe, Imo, Jigawa, Kaduna,
Kano, Katsina, Kebbi, Kogi, Kwara, Lagos, Nasarawa, Niger, Ogun, Ondo, Osun, Oyo, Plateau,
Rivers, Sokoto, Taraba, Yobe, Zamfara.
47

areas as in the Second Republic, with additional purview in the following (1999 Nigerian
Constitution, Second Schedule, Part 1):
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

auditing of accounts within offices, courts, and authorities in the Federation
arms, ammunitions, explosives
bankruptcy
construction and maintenances of Federal truck roads
copyright, patents, trademarks
creation of states
drugs, poisons, quarantine
fishing and fisheries (not including regulation of inland waters within Nigeria)
labor (trade unions, industrial relations, safety, pensions)
legal proceedings between Governments of States or between the Federal
Government and any State, authority, or person
military (army, navy, air force)
mines and minerals (oil fields, oil mining, geological surveys, natural gas)
national parks
nuclear energy
prisons
taxation of incomes, profits, and capital gains
formation, annulment, and dissolution of marriages
wireless, broadcasting, and television

The National Assembly is constitutionally mandated to sit for at least 188 days out of the year.
Moreover, “the finances and qualifications of members [are] to be declared and vetted.” With
this brief historical overview, we can see that the “the rights and obligations of the National
Assembly [have been] enhanced over preceding governments. (Lewis, 2011: 5)”
In the Fourth Republic, the National Assembly considered over 1,100 bills and
15

resolutions between 1999 and 2010 (Lewis, 2011: 5).

Both National Assembly members and

the Executive submitted bills for the legislature’s consideration. Roughly 50% of the total bills
15

“973 bills were submitted to the floor of the House of Representatives between 1999 and
2010, of which 206 were passed. In the Senate 558 bills were introduced to the floor between the
same time period, of which roughly one-fifth were passed” (Lewis, 2011: 7;
http://www.nassnig.org/nass2/legislation.php).

48

considered by the National Assembly originated from the President; however “executive control
of the agenda has steadily declined.” For example, between 2003 and 2007, a total of 333 bills
were introduced on the floor of the House of Representatives. Of that, private Members of the
House presented 51% (169) of the bills, while 44% (146) came from the executive. And in 2008,
private Assembly members sponsored about 80% of new legislation (Lewis, 2011: 7 – 8).

16

The National Assembly is becoming more active in determining the legislative agenda
and gaining a stronger voice in Nigerian governance. With that said, unlike in previous
Republics, subnational political entities are playing an enhanced role in the development and
execution of public policy. Subnational governments have newly attained powers and
responsibilities. With this new authority, state and local governments face a path similar to that
of the national legislature, attempting to move toward a more influential position in Nigerian
politics.

Devolution of Power and Purview to State and Local Governments
Even though political power was concentrated in the national executive during Nigeria’s
military and civilian regimes, state and local legislatures did have a role in governance. The
extent to which power was delegated to these tiers varied as a result of politicians’ quests to
access public revenue. Furthermore, in the Fourth Republic, subnational governments have
become more prominent in decision-making. Particularly, state and local governments have
begun to use their new political and fiscal authority to enact public policy, rather than depending
on a federally led program (Fajingbesi, et al, 2004).
During the First Republic, subnational power was vested in the Region (Northern
Nigeria, Eastern Nigeria, Western Nigeria, and Mid-Western Nigeria). According to the 1963

16

Greater detail about legislative qualifications and production and passing of bills will be
discussed in Chapter Four.
49

Constitution (Chapter 1, Section 5), each region was governed by a regional constitution (which
held the force of law throughout that region), an Executive Governor (executive council,
executive ministers, a public service commission, and Director of Audit), a high court, and a
legislature. Regional legislatures were endowed with the power to make laws “for the peace,
order, and good government of that region with respect to matters” not specifically included in
the (national) Exclusive Legislative List. Additionally, the Concurrent Legislative List (1963
Nigerian Constitution, The Schedule, Part 2) detailed the areas in which both national and
regional legislatures could enact law. This included:
•
•
•
•
•
•
•
•
•
•
•
•

antiquities
arms and ammunition
bankruptcy and insolvency
census, scientific/industrial research, archives/public records, and statistics
commercial and industrial monopolies, combines, and trusts
drugs, poisons, and quarantine (designated by Presidential order)
fingerprints, identification, and criminal records, prisons
higher education (other than institutions referred to in the Exclusive Legislative List)
industrial development, labor, labor conditions, industrial relations, trade unions
national monuments, parks, and tourism (within a Region and by Presidential order)
public safety and order (designated by Presidential order)
execution of civil and criminal processes as dictated by the Regional High Court and
other regional courts of law

Regional legislatures were also granted “residual powers,” which were items not mentioned
expressly in the constitution. This included primary and secondary education, health, public
works, secondary roads, and marketing boards.
In general, regional legislatures “guarded their autonomy jealously,” competing with the
central government for political control. Access to a large and steady stream of revenue from the
marketing boards allowed regional bodies to increase their political scope. Bates (1981: 14 - 15)
provides an example from Nigeria’s Western Region. There, the Action Group and the National
Council of Nigerian Citizens (the two dominant political parties) formed a legislative coalition,
50

intending to consolidate financial power in the regional legislature. Seizing the Commodity
Marketing Boards, they oversaw the direct transfer of £34 million from the boards to the Western
Region’s government’s coffers. As the regions became more [financially] autonomous, fragility
and insecurity grew in Nigeria’s federal structure (Elaigwu, 2005: 60 – 61). This ultimately
coalesced in the secession attempt led by the Eastern Regional government and the interruption
of civilian government by the military.
With the disaggregation of the four regions in the Second Republic and the creation of 19
states, subnational governments also changed form. The executive structure shifted from that of
a Regional Executive to a state Governor (executive council, ministers). Each state’s legislative
authority was vested in a House of Assembly (1979 Nigerian Constitution, Chapter 1, Part 2;
Chapter 5, Part 2). Membership of a state’s House of Assembly equaled three times the total
number of seats which that state held in the Federal House of Representatives. Members of state
Houses of Assembly had authority to form committees and make laws in the same areas as their
regional predecessors. Responsibilities also expanded to include:
•
•
•
•
•
•

electoral law
electric power
exhibition of cinematograph films
industrial, commercial, or agricultural development
allocation, division, and distribution of public revenue, grants, and loan
tax collection

Additionally, the 1979 Constitution (Chapter 1, Part 2, Section 7) recognized a new local
government system made up of 301 “democratically elected local government councils”
(Fajingbesi, et al, 2004: 306). State governments were charged with “ensur[ing] local
governments’ existence under law, providing for the establishment, structure, composition,
finance, and functions of such councils.” The local government’s duties included economic

51

planning and development of the local government area (an administrative constituency
predetermined by the state government). Further duties involved:
•
•
•
•
•
•
•
•
•
•

collection of rates, radio, and television licenses
licensing bicycles, trucks, canoes, wheelbarrows, and carts
maintenance of cemeteries, burial grounds, and homes for the destitute and infirm
maintenance of markets and motor parks
construction and maintenance of roads, streets, drainage, parks as designated by the
House of Assembly
naming of roads and streets; numbering of houses
birth, death, and marriage registration
regulation of out-door advertising, movement and keeping of pets, shops and kiosks,
restaurants, and laundries
provision and maintenance of primary education and health services
development of agricultural and non-mineral natural resources
(1979 Nigerian Constitution, Fourth Schedule)

In the Second Republic, conflicts over state autonomy continued to define Nigerian
federalism, especially when it came to the creation and operation of local government councils.
“Federal sources within the [national] Executive and National Assembly had publically
contended that state governments had no right to create additional local governments,” especially
since the Federal Constitution already named and delineated the local government areas. On the
other hand, state governments believed that since they were given authority to “provide for the
establishment structure, composition, finance, and functions” of local government councils, they
inherently had the right to create, merge, and dissolve local government councils as they saw fit
(Elaigwu, 2005: 161).
However, similar to the First Republic, the major source of conflict between the federal
and state governments stemmed from finance. In the First Republic, agriculture was the major
source of Nigeria’s revenue. Regional Governments regulated this sector via control of the
marketing boards. But in the Second Republic, the national government came to dominate the
major engine for income generation: the oil sector. The 1969 Petroleum Decree established that
52

“all royalties, rents, and other revenues derived from or relating to exploration, prospecting or
searching for petroleum” would accrue directly to the federal government (Ndebbio, 2004: 113).
With the oil boom of the 1970s, states demanded a share of oil funds. In reaction to these
demands, the Revenue Allocation Act was passed in 1981. This act determined that petroleum
income would be shared among the tiers of government in the following manner: 55% to the
national government, 32.5% to be distributed among the state governments, and 10% to be
divided between local governments (Elaigwu, 2005: 287). As a result, rather than cultivating
internal sources of revenue, states grew more dependent on allocations from the national
government. “As additional states were created in Nigeria, the tendency towards greater
authority at the center became more glaring;” while the number of states increased, the newer
entities had a weaker resource base, relying on transfers from the national government for
funding. Therefore, the power of the federal center became greater (Elaigwu, 2007: 151- 155).
Thus, unlike in the First Republic where the hyper-autonomous nature of regional governments
led to secession attempts and the military’s interruption of civilian government, federalism in the
Second Republic was characterized by political and fiscal power concentrated in the center. As
international petroleum prices dropped in the 1980’s, the national government no longer had the
income to sustain the three tiers of government. This, coupled with severe personal enrichment
by national politicians, resulted in military intervention in 1982, lasting until 1998.
As previously mentioned, by the time Nigeria reinstituted civilian government (the
current Fourth Republic) in 1999, there were 36 states and 774 local government areas.
In addition to the National Assembly, the 1999 Nigerian Constitution establishes a House
of Assembly in each of the states and a Local Government Council (LGC) in each Local
Government Area (LGA). According to the constitution (Chapter 5, Part 2), “a House of
53

Assembly of a state shall consist of three or four times the number of seats which that state has in
the House of Representatives divided in a way to reflects, as far as possible, nearly equal
population…not less that twenty-four and not more than forty members.” The House of
Assembly is required to sit for at least 181 days in a year. State legislatures in the Fourth
Republic are allowed to form laws in the same areas as their predecessors in the Second
Republic, as well as in the following:
•
•
•
•

allocation/division of public revenue, grants, and loans
archives
collection of taxes (state and local governments)
trigonometrical, coastal, and topographical surveys

(1999 Nigerian Constitution, Second Schedule, Part 2).
Moreover, they hold residual powers in areas not specifically given to the National Assembly in
the constitution.
The Fourth Republic also maintains the system of local governance set up in the 1979
Constitution. LGCs are still charged with the economic planning and development in their
jurisdiction. Local councils hold purview in the same areas as their predecessors in the Second
Republic. Their duties have also been expanded to include the following:
•
•
•

control and regulation of bakeries and other places for public sale of food
licensing, regulation, and control of the sale of liquor.
provision and maintenance of adult and vocational education
(1999 Nigerian Constitution, Fourth Schedule).

State and local governance is becoming increasingly important to the execution of Nigerian
democracy. In fact, they’ve been charged with “the enhancement of representative grassroots
democracy…[and also] a mechanism for participatory integration of [ordinary Nigerians] into
the fold of democratic governance” (Fajingbesi et al, 2004: 47).

54

As the federal, state, and local tiers attempt to reorganize and redistribute authority from
the highly centralized system handed over by the previous military government, they struggle
with boundaries and overlapping jurisdictions. For example, state governments believe the
federal tier to be “too sprawling” in its in its duties and functions. State political leaders argue
that development should be driven by the state, not as a national program. On the other hand, the
federal tier warns that stripping the central government of too much authority would hearken
back to the First Republic period, where national sovereignty was undermined by hyperregionalism. Last, local governments perceive the state as overbearing, refusing to recognize
LGCs as autonomous institutions. State governors have even gone as far as removing
democratically elected LGC chairmen as if they are bureaucrats and not independent political
actors. However, from the state governments’ perspective, “local governments are the most
problematic tier in the federation; they lack executive capacity. They are
inexperienced…mistaking autonomy for independence or sovereignty” (Elaigwu, 2005: 320).
These political debates in Nigeria’s political federalism are further compounded by the
structure of fiscal federalism, that is the allocation tax authority, revenue mobilization, and
expenditure responsibilities across the three levels of government (Ogwumike and Isumonah,
2004: 259). In the next section, I outline Nigeria’s revenue sharing from petroleum and current
system taxation, highlighting the current debates and reforms.

Distributing Petroleum Wealth Across the Three Tiers
As previously mentioned, since the 1969 Petroleum Decree all revenues derived from the
production, exploration, prospecting, or searching for petroleum would accrue directly to the

55

federal government, into the federation account. From there, the funds would be distributed
between the three levels of government. Nigeria’s natural resource income is derived from two
sources: crude oil sales and oil taxes.
The Nigerian National Petroleum Corporation (NNPC), representing the government’s
business interests, has formed partnerships with private petroleum companies to find and
produce crude oil. The private oil companies and NNPC (via the federal budget) both finance
business operations and share the crude oil that is produced. NNPC then takes the government’s
share of the crude oil and sells it in domestic and international markets, which, in turn, accounts
for a major portion of Nigeria’s oil income.
In addition to taking the portion of crude oil produced in the NNPC-private partnerships,
the government also imposes taxes on oil producing companies. These oil taxes include royalties
(1), the Petroleum Profits Tax (PPT) (2), and rents (3).
First, according to the Federal Ministry of Finance (2010: 11): “in recognition of the
[Nigerian] government’s sovereign ownership of the crude oil, private companies are required to
may a fee for every barrel of crude oil they produce.” These royalty fees (1) average at a rate of
about 20% the value of the crude produced. The second type of oil tax is the PPT. According to
the Petroleum Profits Tax Act (PPTA) of 2007, all corporations deriving income from petroleum
operations (extraction, transportation) are required to pay tax on that revenue. On exports, rate
of the tax is 85%, and on domestic sales of oil and gas, the rate is 65.75%. Any profits that are
charged the petroleum tax are automatically exempt from the companies income tax. Petroleum
companies are given some allowances, including any expenditure on “equipment, pipelines,
storage facilities, buildings and drilling costs” (CITN, 2009; JHI, 2009; NIPC, 2009). Last, the

56

government also charges companies rent for the use of the land from which petroleum is
extracted. Rents also include fees for the right to lay pipelines and transport the oil produced.
Thus, Nigeria’s petroleum income is comprised of NNPC profits from crude oil sales, royalties,
the PPT, and rents paid on land ultilzed for petroleum extraction. These revenues all accrue to
the Federation Account, and are then distributed between the national, state, and local tiers.
Between 2007 and 2009, $70 billion in oil income from the previously discussed sources
accrued to the Nigerian government. Of that, 54.8% derived from the sale of crude oil (via
NNPC). 28.9% of the total petroleum revenue was derived from the PPT, and 10.5% from
17

royalties (Federal Ministry of Finance, 2010: 28).

In 2009 income from crude oil sales

continued to increase at a steady rate. On the other hand, revenue from the PPT dropped
significantly from 32.5% of oil income to 20.5%. That year saw an increase of income from the
sale of natural gas, making up for the drop in other areas.
Over the last 30 years, Nigeria has used various formulas to distribute the income from
petroleum between the three levels of government (Table 2.2). Currently (since 2002), 13% of
oil revenue goes directly to oil producing states.

18

The remaining 87% is vertically distributed

between the three tiers in the following way: 54.68% to the federal government, 24.72% to be

17

See Table 2.1 for a yearly breakdown. The remaining 5.8% is made up of profits from the sale
of natural gas (5.3%) and rents/other oil taxes (0.5%).
18

This is known as the principle of derivation, where “some funds [are] set aside to be shared by
the mineral-producing states [so that] the special contributions from these states to the resources
of the nation will be recognized” (Elaigwu, 2007: 126).

57

19

shared by state governments, and 20.60% share by LGCs

(Uche and Uche, 2004; Elaigwu,

2005).
The formula for the distribution of oil revenue among the 36 states (horizontal allocation)
has also varied, depending upon different principles through the last 30 years. Currently (as has
been the case since 1990, see Table 2.3), revenue sharing occurs based on the following:
•
•
•
•
•

40% is divided equally between all the states (equality of the states)
30% is apportioned based upon state population (more populous states obtain a
larger share)
10% is divided based on social development need (e.g. education, health, water)
10% is divided based on state size (landmass) and terrain (larger states obtain a
greater share)
10% is distributed based upon a state’s internal revenue effort (states generating
higher rates of internal revenue earn more; thus, a portion of petroleum income is
used as an incentive for state’s to increase their capacity to mobilize internally
generated tax revenue)

(Oriakhi, 2004, Usman, 2007).
Having an understanding of the sources of petroleum income for Nigerian national, state,
and local governments, I now explore that non-oil sources of public revenue.

Nigeria’s Tax System: Mobilizing Revenue At Three Levels of Government
As with the National Assembly, the 1999 Nigerian Constitution endows state and local
legislatures with the power to raise non-oil revenue via taxes on citizens (Orewa, 1979; Guyer,
1991; Suberu, 2003; Fajingbesi, et al, 2004). Several clauses, taken together, provide the
19

In the past, a portion of income has gone for the use of “special funds” for the use of
economic development (in oil and non-producing states and the federal capital), savings, and
environmental protection. However, in 2002, the Nigerian Supreme Court declared allocations to
the “special funds” from the federation account as illegal.
58

constitutional source for federal, state, and local governments’ tax authority (Sanni, 2007: 2).
For example:
•

•

•
•
•
•

•

•

•

Chapter 1, Part 2, Section 4, Clause 1/Clause 7: legislative power is vested in the
National Assembly and the State Houses of Assembly to make law for the “peace,
order, and good government” of the Federation and the States, including tax laws.
Chapter 1, Part 2, Section 7, Clause 5: each House of Assembly is to bestow its
LGCs certain functions, including “assessment of privately owned houses or
tenements for the purpose of levying rates.”
Chapter 2, Section 24, Clause F: every citizen has the duty to “declares his income
honestly to appropriate agencies and pay his tax promptly.”
Chapter 4, Section 44, Clause 2: requires a law for “the imposition or enforcement
of any tax, rate, or duty.”
Chapter 5, Part 1, Section 59, Clause 1(b): provides procedures for the passing
federal tax legislation.
Chapter 5, Part 2, Section 120/Section 163(a): revenues collected by State
Governments from personal income taxes, capital gains taxes, and stamp duties
should contribute to part of the Consolidated Revenue Fund.
Chapter 6, Part 1, Section 162, Clause 1: personal income tax paid by members of
the Nigeria armed forces, police force, Department of Foreign Affairs, and residents
of the Federal Capital Territory will not be paid into the Federation Account.
Chapter 6, Part 1, Section 163: the Federal Government must redistribute its
proceeds from personal income taxes, capital gains taxes, and stamp duties to all 36
States based on derivation.
Chapter 6, Part 1, Section 165: each State must compensate the Federal
Government for expenditures incurred for collecting taxes on behalf of the State.

According to the Federal Inland Revenue Service, the Nigerian Tax System includes
three separate prongs: the Nigerian National Tax Policy, the Nigerian Tax Law, and the Nigerian
20

Tax Administration.

1. The Nigerian National Tax Policy
The Nigerian National Tax Policy (2011) is a “statement of the government’s approach to
taxation, both from the practical [body of laws constituting Nigeria’s tax law] and normative [tax
20

Federal Inland Revenue Service (FIRS) is the agency mandated to administer and manage the
national tax regime.

59

21

administration] point of view.”

Essentially, this national policy provides a set of guidelines for

regulation of the tax system, and serves as a basis from which resultant tax legislation and
organization. According to this policy, a tax is defined as “a monetary charge imposed by the
Government on persons, entities, transactions or properties to yield revenue.” This is further
described as “the enforced proportional contributions from personal property, levied by the State
by virtue of its sovereignty for the support of Government and for all public needs” (Federal
Ministry of Finance, 2011a).
The policy goes on to outline (Federal Ministry of Finance, 2011b):
•
•
•
•

•
•

•
•
•

types of taxes that can be levied on individuals, corporations, transactions, and assets.
role federal, state, and local governments play in tax collection.
role of the executive, legislative, and judiciary branches as stakeholders in the tax
system.
responsibilities of the tax authorities—FIRS, State Boards of Internal Revenue, Joint
Tax Board—with respect to information gathering, registration of taxable persons,
filing/processing returns and refunds, payment processing and collection, record
keeping and auditing, and rewarding taxpayer compliance/sanctioning noncompliance.
role of tax consultants, practitioners, and other professional bodies.
role of taxpayers, “the bedrock of the tax system and the source of all revenue
generated by tax authorities,” providing strict and voluntary compliance, registration,
and payment (32).
special arrangements for the purpose of attracting investments.
fiscal dispute resolution and appeal mechanisms, between federal, state, and local
governments; between the Executive, Legislature, and Judiciary; with taxpayers.
institutionalizing a tax culture in Nigeria, “creating awareness about the central role
which taxation can play in National Development…creating a tax conscious
citizenry” (55).

21

The drafting of the National tax Policy was spearheaded by the Nigerian Federal Ministry of
Finance, while receiving input from various stakeholders with both government and the private
sector.
60

2. The Nigerian Tax Law
The Nigerian tax law, as a body, guides the “administration of taxes in federal, state, and
local government authorities. The following legislation comprise the bulk of the tax framework:
Federal Inland Revenue Service Establishment (FIRS) Act 2007
While the Ministry of Finance is the primary administrator of tax law at the federal level,
they operate via the Federal Inland Revenue Service (FIRS). The FIRS Act charges this
autonomous body with the administration of national tax law, investing the “powers to do all
such things as may be deemed necessary and expedient for the assessment and collection of taxes
due to the Federal Government.” FIRS may sue and be sued, and they are legally sanctioned to
“acquire, hold, and dispose of any property taken as security for or in satisfaction of any tax,
penalty, or judgment debt” (Sanni, 2007: 3).
Companies Income Tax Act (CITA) 2007
In Nigeria, “once a company is incorporated, it becomes a legal entity and is treated
under law as an artificial person, separate and distinct from its shareholders.” Consequently,
corporations pay a tax on their yearly profits at a rate of 30%. Each company performs its own
self-assessment; however, Nigerian companies are taxed on their worldwide income, while
foreign corporations are only assessed on the portion of their revenue attributed to business
operations in Nigeria (CITN, 2009: 1). Since companies in the petroleum sector already pay the
Petroleum Profits Tax (PPT), they are exempt from the companies income tax.

61

Education Tax Act 2004
All corporations are also assessed a 2% tax of their revenue, as a “social obligation places
on all companies in ensuring they contribute in developing educational facilities in the country
(NIPC, 2009: 4). This fund is disbursed in the following manner: 25% to universities; 12.5% to
polytechnic schools; 12.5% to the College of Education; 10% to secondary schools; 40% to
primary schools (JHI, 2009).
Capital Gains Tax Act 2004
This is a 10% tax on corporations and individuals on all gains from the sale, lease, or
transfer of stocks, bonds, real estate, and other investments. However, if the company or
individual is a non-Nigerian resident, the tax will only be assessed on the amount received or
brought into the country (NIPC, 2009).
Stamp Duties Act 2004
This tax is assessed on documents and transactions, the rate varying by the type of
document. When one of the parties is a corporation, FIRS levies the tax; in other circumstances,
the tax is paid to the state tax authority (CITN, 2009; NIPC, 2009).
Value Added Tax Act (VAT) 1993/2007
This consumption tax replaces the former sales tax. Those purchasing/consuming goods
and services pay 5% of the purchasing price as a tax. VAT is administered by FIRS and
collected at the federal level on behalf of the national, state, and local governments. The
following goods and services are exempted from the VAT:
•

medical and pharmaceutical products
62

•
•
•
•
•
•
•
•
•
•
•

basic food items
books and educational materials
baby products
fertilizer, locally produced agricultural and veterinary medicine, farming machinery,
and faming transportation equipment
all exports
plant, machinery and goods imported or purchased for use in Export Processing Zone,
Free Trade Zone, or downstream petroleum operations
tractors, ploughs, agricultural equipment and implements purchased for agricultural
services
medical services
services by community banks and mortgage institutions
plays and performances conducted buy educational institutions as part of learning
all exported services

As previously mentioned, all revenue from the VAT accrue directly to the federal VAT pool and
are disbursed in the following manner: 15% to national government; 50% to state governments;
35% to local governments (CITN 2009; NIPC, 2009; JHI, 2009; Federal Ministry of Finance,
2010).
Personal Income Tax Act (PITA) 1993/2007
According to this law, every Nigerian employee is supposed to pay tax on his or her
aggregate income (salaries, wages, fees, allowances, gains, benefits), derived both within Nigeria
and outside the country. In this instance, the taxpayer’s residency determines to whom their
income tax is paid.

22

(See Table 2.4 for the scale on which the personal income tax is levied).

The following are exempt from personal income tax:
•
•
•
•

medical or dental expenses incurred by employee
retirement, gratuities, and compensation for loss of office
interest on loans for developing an owner-occupied residential house
rent subsidy/allowance (maximum 150,000 naira per year)

22

Residency is defined as “a place for his domestic use in Nigeria on a relevant day, excluding
hotels and rest house. A person is deemed resident in Nigeria [of the country and of the state in
question] if he resides for a 183 days in any 12 month period” (NIPC, 2009: 8).
63

•
•
•
•

cost of passage to or from Nigeria incurred by employee (maximum 20,000 naira per
year)
meal subsidy/allowance (5,000 naira per year)
entertainment allowance (6,000 naira per year)
leave allowance (10% of annual salary, maximum of 7,500 naira per year)

(JHI, 2009; NIPC, 2009).
These eight regulations constitute the major taxation laws under which Nigeria operates (JHI,
2009).
3. The Nigerian Tax Administration
The Nigerian Tax Administration is made up of the national, state, and local bodies
charged with assessing, collecting, and accounting for all forms of taxes in accordance with the
law. The organizations involved in managing Nigeria’s tax system are the following:

Federal Inland Revenue Service (FIRS)
As previously mention, FIRS is the independent body charged with the administration of
taxation at the national level. Specifically, FIRS oversees the following levies:
•
•
•
•
•

Companies Income Tax
23
Withholding Tax (corporations, residents of Abuja, and non-Nigerian residents)
Petroleum Profits Tax
Value Added Tax
Education Tax

23

Certain activities and services are subject to Withholding Tax: during these transactions, the
individual making the payment is expected to do deduct the tax at the applicable rate and submit
it to the applicable tax authority. The following are activities subject to the Withholding Tax and
the rate: rent (10% for corporations and individuals), construction (5% for corporations and
individuals), dividends (10% for corporations and individuals), royalties (5% for corporations,
10% for individuals), commission (5% for corporations, 10% for individuals), professional fees
(5% for corporations, 10% for individuals), technical fees (5% for corporations, 10% for
individuals), consultancy fees (5% for corporations, 10% for individuals) (NIPC, 2009: 6).

64

•
•
•

Capital Gains Tax (corporations, residents of Abuja, and non-Nigerian residents)
Stamp Duties (corporations and residents of Abuja)
Personal Income Tax (member of the armed forces, members of the Nigerian Police
Force, residents of Abuja, staff of the Ministry of Foreign Affairs, and non-Nigerian
residents)

(Federal Ministry of Finance, 2011b).
State Internal Revenue Service (SIRS)
Also known as the Board of Internal Revenue, every SIRS organization is mandated with
carrying tax policy at the state level.
Specifically, SIRS has jurisdiction over the following taxes within their state:

•
•
•
•
•
•
•

•
•
•
•

24

Personal Income Tax (Pay As You Earn for individuals and Direct Taxation via
individuals’ self assessment)
Withholding Tax (individuals)
Capital Gains Tax (individuals)
Stamp Duties (documents executed by individuals)
Pools betting, Lotteries, Gaming, and Casino Taxes
Road Taxes
Business Premises Registration Fee (urban areas: 10,000 naira for registration and
5,000 naira per year for renewal; rural areas: 2,000 naira for registration and 1,000
naira per year for renewal)
Development Levy (individuals: not more that 100 naira per year on all taxable
persons)
Naming of street registration fees in State Capital
Right of Occupancy fees (land owned by State Government in urban areas)
Market Taxes and Levies (only where State finance is involved)

(Federal Ministry of Finance, 2011b).

24

The Pay As You Earn (PAYE) system was implemented with PITA 1993. Under Nigerian tax
law employers collect tax by deducting it from the salary of employees and then remit directly to
the appropriate tax authority (JHI, 2009). The PAYE system is primarily used for civil servants,
federal, and state government employees.
65

Local Government Committees on Revenue Collection
Created in each LGA within a state (under PITA 1993), these committees are given
authority to collect the following taxes on behalf of LGSs:
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

shops and Kiosks rates
tenement rates
liquor License fees
slaughter slab fees
marriage, birth, and death registration fees
nation of street registration fees (excluding streets in State Capital)
right of occupancy fees (land in rural areas, excluding those already collected by
State and Federal governments)
market taxes and levis (excluding markets where State finance is involved)
motor park levies
domestic animal license fees
bicycle, truck, canoe, wheelbarrow, and cart fees (excluding mechanically
propelled trucks)
cattle tax (payable by cattle farmers only)
merriment and road closure fees
radio and television license fees (excluding radio and television transmitters)
vehicle radio license
wrong Parking charges
public convenience, sewage, and refuse disposal fees
customary burial grounds permit fees
religious places establishment permit fees
signboard and advertisement permit fees

(Federal Ministry of Finance, 2011b).

Joint Tax Board (JTB)
Since both federal and state governments have the authority to tax income, “this makes it
necessary to have a joint forum for the discussion and resolution of issues arising in the course of
[personal income] tax administration” (Sanni, 2007: 5). Established under PITA 1993/2007, the
Joint Tax board is mandated to administer the personal income tax. They also provide advice to
federal and state governments on methods for improving the management and execution of
66

income tax assessment, collection, and remuneration. This includes the improvement of
information sharing between FIRS and the SIRSs, promoting uniformity in tax administration
across Nigeria, issues of double taxation.

25

The FIRS chairman chairs the JTB while the

directors of each state’s SIRS serve as members.
In 2007, the Nigerian Tax Administration, led by the JTB, embarked on a project aimed
at overhauling the system of tax collection (and tax payment). They identified several barriers to
the expansion of Nigeria’s tax regime. The initial problem the JTB isolated was that taxpayers
were difficult to identify. With no unified database, there was no method for identification or
acquiring information about taxpayers (Omoigui, 2007). Up until that point, FIRS/SIRSs
utilized a manual system of tax payment and remittance. This manual processing system
resulted in errors. There was no concrete way to determine who made a tax payment.
FIRS/SIRSs were also unable to identify the tax for which a collected payment was meant.
Furthermore, there was no information about where/when the tax was paid, who received the
payment, how much was paid, or the form in which the payment was made. This culminated in
the total inability to track the funds (from taxpayer to FIRS/SIRSs). As a result, there were time
lapses between tax payment and remittance to FIRS/SIRSs and difficulty in tracking defaulters.
JTB’s solution is the automation of tax collection or Project FACT (Friendly, Accurate,
Complete, and Timely). This is a joint effort between FIRS and the SIRSs via the Joint Tax
Board. Each taxable individual or corporation would register at their local FIRS/SIRS office and
receive a taxpayer identification number (TIN). TIN is “basically an electronic system of tax
25

“Nigeria has a several tax treaties referred to as ‘double taxation’ agreements with a number
of countries. This is to ensure that the tax payable to Nigeria on the profits of a Nigerian
company being remitted into the country are reduced by the amount of “foreign tax” paid abroad
and vice-versa…Some of these countries include the UK, France, The Netherlands, Belgium,
Canada, and Pakistan” (NIPC, 2009).
67

registration, which would be unique to identify taxpayers for life and would be available
nationwide” (Omoigui, 2007). They believe TIN will help to eliminate the loopholes created by
manual registration in several ways. First, the TIN system would enhance taxpayer identification
and registration. This would therefore help to bring more taxpayers into the tax net. With this
information, federal and state boards of revenue would be able to coordinate in a more effective
fashion. With TIN, tax authorities would be able to access, collate, analyze and retrieve data
about taxpayers and their payments with ease. Specifically, tax authorities could ascertain the
actual income and tax burden of all registered taxpayers. For example, “multiple taxation”
(corporate or individual payment of the same tax more than once to different tax authorities) has
been a major challenge for taxpayers and administrators. These entities would know which taxes
a given tax payer has already made, thereby reducing confusion. Due to the potential accuracy
of the taxpayer database, the system would facilitate a more efficient system of tax assessment
and collection, as well as auditing and investigation. Widespread use of the TIN system would
reduce leakages in tax collection, eventually working to minimize corruption. Long term, the
accuracy in data collection would allow FIRS/SIRSs to monitor taxpayers and authorities,
minimizing or eliminating the cost of tax compliance. Ultimately, this would help to engender
greater voluntary compliance in tax system (Omoigui, 2007).
In this new system, upon receiving their unique TIN, the taxpayer would perform a selfassessment of the tax payment owed (based on their income and allowances, tax rate determined
under PITA 1993/2007). Taxpayers would then remit the tax payment to one of the approved
collecting banks (paid directly into the FIRS/SIRS account) and obtain an electronic ticket (e-

68

26

ticket).

Returning to their FIRS/SIRS office, the taxpayer would present the bank e-ticket as

their proof of payment. That e-ticket would also serve as an official FIRS/SIRS receipt,
evidence of tax payment and compliance to FIRS/SIRS agent.
The payments taxpayers have made to the collecting banks are remitted to four lead
banks (United Bank for Africa, First Bank Nigeria Plc., Union Bank of Nigeria, Zenith Bank
Plc.), which FIRS/SIRSs deal with directly. There is real-time monitoring of all tax payments.
Moreover, online tracking allows for a single view of all tax payments by a taxpayer (viewable
by the taxpayer, CBN, Ministry of Finance, FIRS/SIRS).
Through this process, FIRS/SIRSs employees do not have access to any payments or
money, curtailing corruption or fraud among tax agents. Thus, as more taxpayers (individual and
corporations) obtain their TIN number and participate in the automated tax payment system, tax
officials are able to focus on minimizing tax evasion and institutionalizing a culture of tax
payment in Nigeria.
Since 2004, FIRS/SIRSs staff have engaged in several capacity building efforts. In 2006,
1,500 unskilled staff members were released from their positions. New employees participate in
a “preliminary course for inspectors of taxes.” This 12-week course aims to educate tax
inspectors about the new TIN system, the automation of the tax collection system, recent changes
in tax policy, and new methods to examine and audit accounts. Returning employees participate
in a similar refresher course. Tax inspectors have also been sponsored to attend training
26

The following are participating collecting banks: Access Bank Nig. Plc., Afri Bank Plc.,
Diamond Bank Plc., Eco Bank Nig. Plc., Equitorial Trust Bank Plc., Fidelity Bank Plc., First
Bank Nigeria Plc., First City Monument Bank, First Inland Bank Plc., Guarantee Trust Bank,
IBTC Chartered Bank Plc., Intercontinental Bank, Nigeria Transnational Bank, Oceanic Bank
Plc., Platinum Habib Bank, Skye Bank Plc., Spring Bank Plc., Sterling Bank Plc., Union Bank of
Nigeria, United Bank for Africa, Unity Bank Plc., Wema Bank Plc., Zenith Bank Plc.
69

programs with various tax collecting agencies overseas. With an increase in capacity and
competence, FIRS/SIRSs officials are able to educate ordinary Nigerians about the tax system
and compliance. These agencies have also improved their systems for wages, benefits, and
compensation. Their “ability to pay improved compensation [will help] to enable the retention
and attraction of high caliber employees” (Omoigui, 2007). Furthermore, adequate
compensation of staff within the tax administration will reduce incentives to participate in
corrupt practices.
Having an understanding of the various methods the national, state, and local
governments generate non-oil revenue, we can investigate the nature of their extractive capacity.
According to the Central Bank of Nigeria (CBN), non-oil revenue has been growing at all levels
of Nigerian government. For example, in 2009, 34% of the federal government’s income was
derived from non-oil sources (increasing from 29% in 2008, see Table 2.5). Of the non-oil
income in 2009, 85% derived from the company income tax, custom and excise taxes, and the
VAT (See Table 2.6).

27

In 2009, an average of 15% state governments’ revenue came from

internally generated (non-oil) tax revenue (IGR, see Table 2.7). This is contrasted with 69% of
revenue derived from federal oil income transfers (federation account, stabilization and excess
crude account), 10% from federal VAT transfers, and 7% from grants (private industry and
international organizations). In 2009, the majority of LGC revenue was derived from federal and
state transfers (52%, see Table 2.8), while 3% came from taxation, 15% from federal VAT
transfers, and 32% from grants (private and international organizations).

27

The remaining 15% of non-oil income is derived from budget surpluses from federal
ministries’ and agencies’ operations.
70

Though federal, state, and local governments remain dependent on petroleum revenue,
non-oil tax generation increasing. For example, federal government, non-oil revenue grew from
13% of national income in 2006 to 34% in 2009 (Table 2.5). During the same period, internally
generated revenue from non-oil taxation increased from 8% of state governments’ revenue to
14% (Table 2.7). At the local level, tax generation is growing at a more modest rate, increasing
from 2% to 3% between 2006 and 2009 (Table 2.8).
More importantly, sub-national variation in revenue dependence exists, especially in state
governments (Table 2.9). The interaction between access to tax-based and petroleum income
results in varied revenue composition across Nigeria (Okoko and Nna, 1997; Fajingbesi, et al,
2004). Between 1999 and 2009, on average, 11% of state governments’ revenue came from
internally generated citizen taxation; 64% of their revenue was from federal oil transfers (Table
2.9). However, inter-state variation can be observed. For example, between 1999 and 2009, in
Lagos State, an average of 53% of total state income was derived from the taxation of citizens
(vs. 24% from federal oil transfers). This is compared to Ebonyi State at the other end of the
spectrum, where an average of 3% of state income between 1999 and 2009 was generated from
non-oil taxes (68% from federal oil transfers). This variation in oil income and tax reliance
within Nigeria allows for an investigation into the impact of revenue sources politics.

71

Chapter Two Tables and Figures

Table 2.1: SOURCES OF
FEDERAL GOVERNMENT
OIL REVENUE (%)
2007
ITEM
CRUDE OIL SALES
PPT
ROYALTIES
NATURAL GAS SALES
RENTS AND OTHER OIL
TAXES

2008

2009

50.8
31.7
12.3
4.5

54.4
32.5
9.4
3.5

59.2
20.5
10.3
8.1

0.7

0.2

1.9

Table 2.2:
VERTICAL
ALLOCATION OF
FEDERATION
ACCOUNT, 1981 PRESENT (%)
RECIPIENT
FEDERAL
GOVERNMENT
STATE
GOVERNMENTS
LOCAL
GOVERNMENTS
SPECIAL FUNDS

1981

1989

1990

1992

CURRENT
(FROM
May 2002)

55

55

50

48.5

54.68

30.5

32.5

30

24

24.72

10
--

10
2.5

15
5

20
7.5

20.60
--

72

Table 2.3: HORIZONTAL
ALLOCATION OF
FEDERATION ACCOUNT,
1980 - PRESENT (%)
1980
40
40
15
--

1989
40
30
10
--

5

PRINCIPLE
EQUALITY OF STATES
POPULATION
SOCIAL DEVELOPMENT
LANDMASS AND TERRAIN
INTERNAL REVENUE
EFFORTS

CURRENT
(FROM
1990)
40
30
10
10

20

10

Table 2.4: CURRENT INCOME
TAX RATES
TAXABLE INCOME (NAIRA)
First 30,000
Next 30,000
Next 50,000
Next 50,000
Over 160,000

73

RATE OF
TAX (%)
5
10
15
20
25

Table 2.5: FEDERAL
GOVERNMENT’S REVENUE
SUMMARY (% of total budget)
2006
ITEM
OIL REVENUE
NON-OIL REVENUE

2007

2008

2009

87.2
12.8

78.1
21.9

71
29

65.9
34.1

Notes: Oil Revenue = Profits from crude oil and gas exports, domestic crude oil sales,
PPT, royalties; Non-Oil Revenue = VAT, CIT, customs/excise duties, surplus from
federal ministries’ operations.
Recreated from Section 5.2.1, CBN Annual Report & Statement of Accounts for Year
Ended 31st December, 2009.

74

Table 2.6: SOURCES OF
FEDERAL GOVERNMENT
NON-OIL REVENUE
2007
ITEM
CIT
CUSTOMS/EXCISE
TAXES
VAT
FEDERAL MINISTRIES
OPERATION SURPLUSES

2008

2009

28.7

33.4

30.1

21.2
26.5

26
29.6

25.2
29.4

23.6

11

15.3

75

Table 2.7: STATE
GOVERNMENTS'REVENUE
SUMMARY (% of total budget)
2006

2007

2008

2009

ITEM
IGR
8.1
12.4
13.6
14.8
FEDERATION ACCOUNT
65.8
59.4
51.6
56.1
VAT
7.2
7
7.9
10.1
GRANTS & OTHERS
8.1
10.2
5.5
6.8
STABILIZATION & EXCESS
22.6
13.4
CRUDE
10.8
8.6
Notes: IGR = Internally Generated Revenue from citizen taxation, Federation Account and
Stabilization/Excess Crude = Petroleum Revenue from Federal Transfers.
Recreated from Table 5.1, CBN Annual Report & Statement of Accounts for Year Ended 31st
December, 2009.

76

Table 2.8: LOCAL
GOVERNMENTS' REVENUE
(%)
SUMMARY
2006

2007

2008

2009

ITEM
IGR
1.6
2.6
2.4
3.4
FEDERATION ACCOUNT
81.7
68.3
52.0
49.5
VAT
11.3
12.6
9.8
14.7
STATE GOVT TRANSFER
0.5
0.4
0.2
1.8
GRANTS & OTHERS
3.1
16.2
36.3
31.5
Notes: IGR = Internally Generated Revenue from citizen taxation, Federation Account =
Petroleum Revenue from Federal Transfers.
Recreated from Section 5.5.2, CBN Annual Report & Statement of Accounts for Year
Ended 31st December, 2009.

77

Table 2.9: AVERAGE OIL
AND TAX DEPENDENCE
AMONG THE STATES
(1999 - 2009, % OF TOTAL
INCOME)*
STATE
TAXATION
OIL TRANSFERS
LAGOS
53.4
23.5
RIVERS
22.5
63.7
OGUN
19.0
56.5
SOKOTO
17.6
63.6
OYO
16.7
59.1
OSUN
16.4
61.1
DELTA
13.8
62.1
KANO
13.2
56.6
KADUNA
12.9
59.4
FCT (ABUJA)
12.7
73.8
ANAMBRA
12.4
62.7
ABIA
12.0
60.0
EDO
11.6
63.4
ONDO
11.2
73.5
KWARA
11.0
60.3
ENUGU
10.5
64.6
CROSS RIVER
10.1
62.0
KOGI
9.6
68.0
IMO
9.3
65.2
PLATEAU
8.9
66.5
AKWA-IBOM
8.9
71.0
BENUE
8.7
66.7
BORNO
8.7
64.3
GOMBE
8.6
58.2
NIGER
7.9
62.2
ZAMFARA
7.8
66.6
JIGAWA
7.6
67.4
EKITI
7.5
71.0
KATSINA
7.1
66.2
KEBBI
6.2
65.7
BAYELSA
5.1
82.2
NASSARAWA
4.9
67.8
TARABA
4.8
64.2
ADAMAWA
4.7
69.4
YOBE
3.6
67.9
BAUCHI
3.5
64.5
EBONYI
3.2
68.3
Notes: *States are ordered from highest to lowest in terms of average tax income (% of
total revenue). The corresponding revenue from oil transfers is in next column.
78

CHAPTER THREE

Revenue Sources and Subnational Government Spending Priorities
Background
According to the Federal Inland Revenue Service’s mission statement, one of the
organization’s key goals is making taxation the pivot of Nigerian development, de-emphasizing
reliance on petroleum profits (FIRS, 2007). Tax administrators cite several reasons for pushing
Nigeria away from the dependence on oil income, centering on the instability of this revenue
source. Over the last five years, average oil prices have been declining while the costs of crude
oil production has increased. Additionally, crude oil’s average daily output has waned.
Combined with rising hostilities in the Niger Delta (location of Nigeria’s oil producing states),
Nigeria’s continued access to petroleum revenue is progressively more unpredictable. Thus,
cultivating internal revenue via taxation provides a more stable model for economic growth.
In addition to these economic reasons, I argue that there are political benefits to tax
generation (vs. oil revenue dependence). As previously mentioned, the Western European
experience demonstrates how taxation of citizens cultivates a pattern of political representation.
Political elites, financially reliant on citizens for income, shift public policy in order to satisfy
their constituents’ interests. This is done in order to curtail citizens’ shirking tax payment (Levi,
1988).
Gibson and Hoffman (2006: 7) investigate a similar hypothesis, asking: “Do sources of
revenue affect government expenditure?” Using local government budgetary data from Tanzania
and Zambia, they find that when local government revenue is derived from taxes on citizens,
politicians will tend to expend more funds on public services. On the other hand, when leaders
79

depend on funds derived from sources outside of their local constituency, (e.g. federal
government transfers, foreign aid), they are more likely to spend resources on government
salaries, allowances, and other recurrent expenditures.
Using data collected from the Central Bank of Nigeria (CBN) during fieldwork, I
investigate Gibson and Hoffman’s (2006) question in Nigerian local and state governments.
Focusing this analysis on Nigeria provides theoretical and empirical leverage to test the
revenue/representation relationship. First, Nigeria is a hard case with varying revenue profiles
across subnational entities. Local and state governments have constitutional power to tax and
they do so in a variety of ways. However, they also receive income from the sale of oil through
federal transfers. This creates subnational variation in they types of revenue state and local
governments utilize. Moreover, local and state governments are important to Nigerian politics.
As a result of decentralization of central authority, Nigerian sub-national governments have
purview over policy formulation and execution in a growing list of areas. They are also
authorized to raise revenue for those ends. What determines local and state government
spending priorities? I hypothesize that as revenue from taxation increases, local and state
government spending on public service provision increases. Conversely, as revenue from federal
transfers (natural resource income) increases, local and state government spending on
government salaries and allowances will increase.
Several scholars have investigated how the devolution of power from the central
government to sub-national units influences public service delivery and spending (Bish and
Ostrom, 1973; Rondinelli, et al., 1989; Inman and Rubinfeld, 1996; Ferejohn and Weingast
1997; Snyder, 2001). Two schools of thought have emerged: one group of scholars assert that
moving purview of public services from the national government to local and state entities will
80

decrease representation. In this argument, scholars contend that when subnational leaders have
control over public funds, it is more likely that revenue will not be spent on public service
provision, and will instead be diverted to other uses (Bardhan, 2001; Bardhan and Mookherjee,
2005a). For example, Akin, et al’s (2001) study of health care in Uganda finds that
decentralization is a detriment to service delivery. Local governments decrease expenditures for
broad-based healthcare programs targeting citizens (e.g. child health, malaria control/medicine).
Instead, leaders spend income on salaries, equipment, and vehicles for health sector employees.
On the other hand, opponents argue that devolution of power will benefit representation
in the form of public service provision. According to Gordon (1983) and Rogers and Lee (2012:
6), “devolving policymaking authority…can allow for a closer connection between local
preferences and policy.” Subnational governments are literally closer to citizens than distant
central governments. Therefore, local and state leaders are in a better position to determine
constituents’ service needs and act upon them, “tailor[ing] public services to local tastes”
(Tiebout, 1956; Oates, 1972; Weingast, 1995; Ahmed et al, 2005; Bardhan and Mookherjee,
2005b). For example, in a study of public schools in Uganda, Reinikka and Svensson (2004) find
that, while central government school grants primarily benefit well-off districts, local
governments are better able to deliver benefits to poorer areas.
Neither of these groups, however, considers how the source of revenue available to local
and state governments influences their spending priorities. Following Gibson and Hoffmann
(2006), I investigate this question in Nigeria.

81

Research Design
Dependent Variable
Ideally, the dependent variable would be measured as the portion of local and state
government expenditures utilized for the provision of public services. However, CBN reports of
Nigerian local and state governments revenue and expenditures do not explicitly account for
revenue spent on public service delivery. On the other hand, these reports do indicate
expenditures for government salaries, wages, and other personal allowances (i.e. recurrent costs).
But as Gibson and Hoffman (2006: 10) point out, “while it is tempting to infer that the share not
used for recurrent costs must be used for public services, such a conclusion is false,” especially
since non-recurrent expenditures can also include debt repayments. Thus, the portions of local
government and state government revenue used for recurrent expenditures will be the two
dependent variables utilized to test my hypothesis. I expect that as local and state governments
generate more of their income from taxes on citizens, there will be less public expenditures on
government salaries (negative relationship). However, I also expect that when local and state
governments are more dependent on external, federal transfers (e.g. profits from petroleum sales,
value added tax trances, international grants and foreign aid), they will engage in more salary
spending (Table 3.1).

28

28

I will take similar precautions as Gibson and Hoffman (2006: 10) regarding the expression of
the two dependent variables: “The most obvious way of testing our hypotheses would be to use
aggregate or per capita budget expenditures as our dependent variable. However because we
have no reason to believe that expenditures on any budget line-item will decrease as taxes and/or
transfers rise, using aggregate or per capita expenditure would be inappropriate.” Thus, for the
local and state government analyses, the dependent variables are the portion of total (1) local and
(2) state government revenue utilized for recurrent expenditures.
82

Data is obtained from the Central Bank of Nigeria Annual Report and Statement of
Accounts from 1999 – 2009.29 This report is produced yearly by the CBN; in addition to
summarizing trends, it provides local and state government budgets, including revenue and
expenditures.

Independent Variables
The key independent variables for this analysis are the four sources of revenue for local
and state governments, measured in per capita terms. The first explanatory variable is the
revenue each LGA and state government derived from non-oil taxes on citizens per capita (local
and state levies, respectively). The second explanatory variable is local and state income from
federal transfers from petroleum profits per capita. The third explanatory variable is local and
state revenue from VAT transfers per capita. The last explanatory variable is LGA and state
government income from international grants/foreign aid per capita.30, 31
Though the dependent variables are expressed as a share of total expenditures, I do not
operationalize the independent variables in the same way. First, if I measure each revenue
source as a portion of total revenue, the four shares must add up to one by definition. Therefore,
I would be unable to test the four hypotheses simultaneously. Second, creating ratios of revenue
shares would increase the correlation among the four independent variables measuring the

29

For LGAs, data is available for the years 2000 – 2003 and 2008. For states, data is available
for the years 2000-2003 and 2007 – 2009.
30
For LGAs, transfers from state governments are also included.
31

For a list of the types of taxes collected by local and state governments, see Chapter Two’s
discussion of the Local Government Committees on Revenue Collections and the State Internal
Revenue Services.

83

different sources of revenue. I am interested in testing all four sources of revenue
simultaneously, which allows me to understand which source has the greatest impact on local
and state government expenditure. Thus, I use the four sources of revenue measured in per
capita terms (Gibson and Hoffman, 2006: 12).
This data is once again obtained from the CBN’s Annual Report and Statement of
Accounts, 2000 – 2009.32 For each year, these reports provide each LGA and state government’s
total revenue, and the portion of the total that is obtained from the four key revenue sources:
taxes (local and state levies, respectively), federal transfers, transfers from the value added tax
income (VAT), and grants/foreign aid.
Like Gibson and Hoffman (2006), I include the following control and structural variables
in my models: LGA and state socio-economic capacity (measured with rates of primary
education, access to health case, mortality rates) and LGA/state size (measured with population
and land area).33 However, unlike Gibson and Hoffman’s (2006) cross-sectional analysis (100
Tanzanian and Zambian local districts in 2005), I utilize time-series data (774 LGAs across 2000
– 2004 and 2008; 36 states across 2000 – 2003 and 2007 – 2009). Thus, I also include yearly
dummy variables to control for time trends. Finally, in order to control for the size of the
economy and the possible influence the previous year’s expenditures may have on the current
year’s spending, I include the previous year’s recurrent expenditures. Including yearly dummy

32

For LGAs, data is available for the years 2000 – 2004 and 2008. For states, data is available
for the years 2000 - 2003 and 2007 – 2009.
33

Population (log of total), area (square miles), primary education (number of children
enrolled), access to health facilities (ratio between state population and total number of health
facilities in state), infant mortality (per 1,000). Obtained from the Nigerian National Bureau of
Statistics: Social Statistics in Nigeria (2005) and the Nigeria Demographic and Health Survey
(2003).
84

variables and the previous years recurrent spending allows each model to account for any serial
autocorrelation associated with utilizing time series data.

Model Specification
Therefore, the basic regression models are as followed:
LGA Recurrent Expenditures = a1 - b1(IGR/capita) + b2(Transfers/capita) + b3(VAT/capita) +
b4(Grants/capita) + b5(Log Population) + b6(Area) + b7(Primary Education) + b8(Health
Facilities/capita) + b9(Infant Mortality) + b10(Recurrent Expenditurest-1) +
b11(Year1)…b17(Year6)
State Recurrent Expenditures = a1 - b1(IGR/capita) + b2(Transfers/capita) + b3(VAT/capita) +
b4(Grants/capita) + b5(Log Population) + b6(Area) + b7(Primary Education) + b8(Health
Facilities/capita) + b9(Infant Mortality) + b10(Recurrent Expenditurest-1) +
b11(Year1)…b18(Year7)
Tables 3.2 and 3.3 present the summary statistics for all variables included in the LGA and State
models.

Analysis and Results

Tables 3.4 and 3.5 show the results from the analysis of local and state government
expenditures. Using 1999 - 2009 CBN data, the dependent variable is measured with recurrent
expenditures as a percentage of total government expenditure. The key explanatory variables are
measured in the following way: local and state income from (non-oil) tax-based revenue per
capita; local and state income from petroleum profits per capita; local and state VAT income per
capita; local and state income from grants and foreign per capita. Again, this data is obtained
from CBN reports. I discuss results from the local government analysis, followed by the analysis
of state governments.

85

As expected, local taxes per capita stand in a significant, negative relationship to the
share of local government expenditures spent of government salaries (Table 3.4). Holding all
other variables constant, a one naira increase in internally generated tax income per capita
corresponds with a 0.06% decrease in portion of local expenditures aimed at government salaries
(In 2009, one naira ! $0.01). This result provides evidence that in local governments with a
greater capacity to tax, politicians spend less public revenue on themselves and more on public
services.
On the other hand, in line with the second hypothesis, federal transfers per capita have a
strong and positive relationship to the share of local government expenditure going towards
recurrent costs. For example, holding all other variables constant, a one naira increase in federal
transfers per capita corresponds with a 0.15% increase in local government recurrent spending as
a portion of total expenditures. Grants per capita, as hypothesized, act similarly to federal
transfers, also having a strong positive relationship with local government spending on their
salaries and allowances.
Contrary to my hypothesis, local government income from VAT transfers has a
significant, but negative relationship with salary expenditures. Holding all other variables
constant, a one naira increase in VAT income per capita is associated with a 0.25% decrease in
local government spending on salaries. I initially expect a positive relationship because the VAT
is collected at the local and state levels, but remitted to a federal pool. The federal government
then distributes the funds between the different tiers of government, 774 LGAs dividing 35% of
the overall VAT income between themselves.

34

Since local governments receive this revenue as

34

According to the formula for the horizontal distribution of VAT funds among local
governments, 50% is distributed equally among the LGAs, 30% is distributed based on
population (LGAs with larger populations receive a larger share), and 20% based on derivation
86

a transfer, I expected VAT/capita, like other transfers, to have a positive relationship with salary
spending. But this significant, negative finding suggests that, even though local governments
receive VAT income as a transfer, it influences public spending in the manner of a tax. The
VAT originates as a levy on goods and services paid by citizens. Evidence suggests that the
source of this income as citizen-based tax overrides any effects from method of transfer to local
governments. These relationships between revenue and salary expenditures remain significant
even in the context of socio-economic capacity indicators, which have no effect on recurrent
expenditures in this analysis.
Turning to the analysis of state government expenditures (Table 3.5), as hypothesized,
taxed-based income has a strong, negative relationship with recurrent expenditures. Holding all
other variables constant, a one – naira increase in a state’s internally generated tax revenue (IGR)
per capita corresponds with a 0.16% decrease in that state’s contribution towards government
salaries and allowances. On the other hand, a one – naira increase in a state’s income from
federal transfers per capita is associated with a 0.14% increase in recurrent expenditures.
Like in the LGA model, VAT per capita has a negative relationship with state
governments’ salary spending; however, this is not a statistically significant result. Similarly, the
grant per capita variable bears a positive sign, but this does not indicate a significant relationship
with state recurrent expenditures. These particular non-significant results (when compared to the
significant results in the LGA model) could be attributed to the relative weight VAT and grants
income in state and local budgets. For example, in 2009, 10% of state revenue in Nigeria came
from VAT transfers and 7% was derived from grants. This is compared to LGA budgets where
15% of their revenue originated from VAT transfers and 32% was derived from grants. Taking
(that is LGAs generating more VAT income—a bigger VAT income effort—receive a larger
share).
87

this into account, we observe that income from VAT and grants transfers are a more abundant
35

revenue source for local governments than for state entities.

The LGA model and State model explain 30% and 15% of the variation in government
expenditures, respectively. While much variation remains unexplained, these are relatively
strong models. These analyses illustrate that level of development and demographic variables
(alone) are unable to account for how local and state governments spend public revenue in
Nigeria. Even though, as expected, a local or state government’s recurrent expenditures in the
previous year are positively related to spending in the current year, the source of government
revenue still matters. Non-oil, tax-based income (including VAT at the LGA level) is
consistently and negatively related to salary spending in both tiers of government. At the same
time, dependence on natural resource revenue via federal transfers increases government
spending on salaries and allowances.
As previously mentioned, both the LGA and State models have missing budgetary data.
Since LGAs are clustered by state, there should be 216 LGA/year observations (36 states over 6
years). But because of missing data, there are only 171 LGA/years accounted for. This gap is
less severe in the State model: 36 states governments over 7 years should result in 252 state/year
observations. However, due to missing data, there are 246 state/year observations. If this data is
systematically (as opposed to randomly) missing, the previous regression results could be biased.
In order to check for this bias, I use data imputation to perform a robustness check. In the case
of missing variables for LGA/year observations, I first calculate the mean values for each
variable. I then impute that mean for the year(s) in which original data is unavailable. For
instance, budgetary data (recurrent expenditures, IGR, federal transfers, etc.) is missing for local
35

Please refer back to Tables 2.7 and 2.8 in Chapter Two, which provide summaries of local and
state governments’ revenue profiles.
88

governments in Zamfara State in 2003. So, for example, in order to produce an imputed value
for recurrent expenditures, I calculate a mean using data that is available for spending in
Zamfara’s local governments in the other years. I then impute this mean value as a proxy for
recurrent expenditures in Zamfara LGAs in 2003. I repeat this for missing variables in all LGAyear and state-year observations and then conduct new regression models.
Tables 3.6 and 3.7 present the summary statistics for the imputed LGA and State models.
It is important to note that, besides the increase in observations, there are no other significant
changes in each variable’s mean, standard deviation, or minimum/maximum values.
Tables 3.8 and 3.9 show the results of the LGA and State imputed models. Local taxes
per capita still have a significant, negative relationship with the share of local government
expenditures spent of government salaries (Table 3.8). In fact, the strength of the relationship
triples in the new model. Holding all other variables constant, a one – naira increase in tax
income per capita corresponds with a 0.18% decrease in portion of local expenditures aimed at
government salaries. Similarly, federal transfers per capita still have a consistently strong and
positive influence on the share of local government expenditure going towards recurrent costs.
Again, the effect is slightly increased in the imputed model, and now exerts the most influence
on LGA recurrent spending. A one – naira increase in federal transfers per capita corresponds
with a 0.20% increase in local government recurrent spending as a portion of total expenditures.
Grants per capita maintain a positive relationship with local government spending on
salaries and allowances, though the strength decreases: a one – naira increase in grants per
capita is associated with a 0.07% increase in local government recurrent spending as a portion of
total expenditures. This is the same for VAT income per capita, which continues to have a
negative relationship with local government spending on salaries, but the magnitude decreases.

89

In the imputed model, holding all other variables constant, a one – naira increased in government
VAT income per capita corresponds with a 0.11% decrease in portion of local expenditures spent
on salaries.
In the results for imputed State model (Table 3.9), tax-based revenue continues to have a
significant, negative relationship with state government salary spending. Moreover, the strength
of this relationship increases, such that a one – naira increase in a state’s tax per capita income is
associated with a 0.17% decrease in state recurrent expenditures. The positive relationship
between state income from federal transfers and salary spending also gains strength in this new
model: if all other variables are held constant, a one – naira increase in income from federal
transfers corresponds with a 0.17% increase in expenditures contributed to government salaries.
Once again, VAT per capita has a positive relationship with state recurrent expenditures and
grants per capita a negative one, but neither are significant.
The explanatory power of the imputed LGA and State models decrease (now explaining
25% and 14% of the variation in government expenditures, respectively). With that said, these
continue to demonstrate that the sources of local and state government income remain significant
predictors of public expenditures. When revenue is derived from external sources (outside of
direct taxes on citizens), politicians are more likely to increase public spending on government
salaries and allowances. Despite missing data, these relationships still hold.
Returning to the original models, I perform additional robustness checks to explore the
possibility outlier cases that may be driving these relationships.

90

I begin with by checking for influential observations within each model.

36

I calculate

each observation’s Cook’s distance, a measure of influence that combines residuals and leverage
(Table 3.10). Since observations in Kaduna State and Lagos State exhibit a Cook’s D larger than
one, I consider them to be influential observations. I perform a similar check in the State model,
calculating each observation’s Cook’s D (Table 3.11). We see that no observation has a value
over one. With that said, Bayelsa State government in 2009 displays the highest Cook’s D.
Having determined the influential observations in each model, I re-run regressions,
removing LGAs in Kaduna State and Lagos State (LGA model) and Bayelsa State government
(State model).
Tables 3.12 and 3.13 display the results of the LGA and State models re-run without the
influential observations. Overall, the strength of the hypothesized relationships decreases in both
models; however, the expected relationships still hold, and they are significant.
As expected, local taxes per capita still have a significant, negative relationship with the
portion of government income spent on recurrent expenditures (Table 3.12). A one – naira
increase in tax income per capita corresponds with a 0.05% decrease in the local government
salary spending. Additionally, a one-naira increase in federal transfers per capita corresponds
with a 0.1% increase in local government recurrent spending as a portion of total expenditures.
Federal transfers continue have the strongest relationship with LGA recurrent spending.

36

An observation’s residual is “the difference between the predicted value (based on the
regression equation) and the actual, observed value… an outlier is an observation with large
residual. [Moreover] an observation with an extreme value on a predictor variable is a point with
high leverage. Leverage is a measure of how far an independent variable deviates from its
mean.” Therefore, an observation is considered to be influential it is an outlier (having a large
residual) and high leverage (UCLA Academic Technology Services, Statistical Consulting
Group, 2012).
91

Grants per capita maintain a positive relationship with local government spending on
their salaries and allowances, and VAT per capita continues to have a negative relationship with
local government spending on salaries.
In state governments (Table 3.13), tax-based income continues to have a significant,
negative relationship with salary expenditures. Again, the strength of this effect has decreased: a
one – naira increase in a state’s tax per capita income is associated with a 0.1% decrease in state
recurrent expenditures. The positive relationship between state income from federal transfers
and salary spending is also still significant: if all other variables are held constant, a one – naira
increase in income from federal transfers corresponds with a 0.1% increase in spending on
government salaries. VAT per capita has a negative relationship with state recurrent
expenditures and grants per capita a positive relationship, but like in the original model, neither
is significant.
Despite conducting a series of robustness checks, the hypothesized relationships continue
to maintain significance. Differentiating between tax based and non-tax based income, the
source of local and state government revenue matters when predicting expenditures. When
government revenue is derived from external, federal government transfers and grants, local and
state leaders are more likely to spend income on themselves via salaries and allowances.
However, when revenue comes from the internally generated income from the taxation of
citizens, politicians spend less on their salaries. While accounting for missing values and
influential observations, this relationship is still significant in LGAs and states.

92

Conclusion
In this chapter, I have investigated the influence of government revenue sources on
budgeting and expenditures in Nigerian local and state governments. Using budgetary data
between 1999 and 2009, I find that when local and state governments derive more of their
revenue from internally generated income through taxes on citizens, the are less likely to spend
public resources on government salaries and allowances. Instead, these governments contribute
more resources to non-recurrent expenditures, including the provision of public services. By
contrast, a higher level of dependence on natural resource revenue via federal transfers is
associated with higher levels of recurrent spending. Going against the initial hypothesis, I find
that VAT income act as an internally generated revenue. Even though these revenues are
collected at the local and state level, pooled at the federal level, and transferred to LGAs and
state governments, VAT income has a negative relationship with LGA salary spending. Thus, as
VAT income increases, local governments are less likely to use public funds for recurrent
expenditures. Last, just as revenue from federal petroleum transfers, as the grants and foreign
aid contribution to local budgets increase, LGAs spend more on government salaries.
This analysis suggests that theories of revenue extraction must consider not only taxation
as a source of government income, but also petroleum and other external sources (e.g. grants,
foreign aid). Once these models taken other sources into account, they can be expanded to the
developing world, especially Africa. Furthermore, once accounting for varied revenue sources,
these theories are useful in explaining political behavior and decision-making, especially in
budget priorities. In Nigeria, government officials are more likely to incorporate citizen
preferences when making choices about how to spend public funds. A relationship between
taxation and representative government suggests that building citizen capacity to monitor

93

government budgeting can bolster responsive government and influence the consolidation of
democratic governance.

94

Chapter Three Tables and Figures

Table 3.1:
CONCEPTUAL
FRAMEWORK
DEPENDENT
VARIABLES
LOCAL
STATE
GOVERNMENT GOVERNMENT
RECURRENT
RECURRENT
EXPENDITURES EXPENDITURES
INDEPENDENT TAX-BASED
VARIABLES
REVENUE
FEDERAL
TRANSFERS
(PETROLEUM
PROFITS)
VAT TRANSFERS
INTERNATIONAL
GRANTS &
FOREIGN AID

+

+

+
+

95

-

+
+

Table 3.2: SUMMARY
STATISTICS—LGA
MODEL
VARIABLE

37

OBSERVATIONS

MEAN

STANDARD
MIN
DEVIATION

MAX

RECURRENT
EXPENDITURES (%
TOTAL
EXPENDITURES)

171

68

14.1

4

97.1

TAXES/CAPITA
(NAIRA)

171

4775.1

25141

4.5

287439.4

TRANSFERS/CAPITA
(NAIRA)

171

123022.7

531835.6

97

4273885

VAT/CAPITA
(NAIRA)

171

17893.3

82236.4

7.7

751986.4

GRANTS/CAPITA
(NAIRA)

171

4337

31611.1

0

382737.8

37

Due to the nature of the data, LGAs are clustered by state. This should result in LGA observations in 216 state/year (36 states over
6 years); however there is missing data in the following: Plateau State 2002/2003/2004/2008; Abia 2003/2004, Adamawa 2003/2004,
Anambra 2003/2004, Edo 2003/2004/2008, Ekiti 2003/2004, Enugu 2003/2004, Jigawa 2003/2004, Kwara 2003/2004, Niger
2003/2004, Ogun 2003/2004/2008, Oyo 2003/2004, Rivers 2003/2004, Taraba 2003/2004, Yobe 2003/2004, Zamfara 2003, AkwaIbom 2004/2008, Borno 2004/2008,Ebonyi 2004/2008, Ondo 2004/2008, Osun 2004/2008. This results in LGA clusters in 171
state/years.

96

Table 3.2
(cont’d):
SUMMARY
STATISTICS—
LGA MODEL
VARIABLE
LOG
POPULATION
AREA (SQ.
MILE)
PRIMARY
EDUCATION
HEALTH
FACILITIES/
CAPITA
INFANT
MORTALITY
PREVIOUS
YEAR'S
RECURRENT
EXPENDITURES
(%)

MEAN

STANDARD
DEVIATION

MIN

MAX

171

6.6

0.2

6.2

7

171

9828.9

6970.6

2440.2

29483.9

171

546077.9

234241.5

731611

1443848

171

6129.9

3628.1

2466

25577

171

100.8

23.1

66

125

171

69.4

12.1

4

97.1

38

OBSERVATIONS

38

Due to the nature of the data, LGAs are clustered by state. This should result in LGA observations in 216 state/year (36 states over
6 years); however there is missing data in the following: Plateau State 2002/2003/2004/2008; Abia 2003/2004, Adamawa 2003/2004,
Anambra 2003/2004, Edo 2003/2004/2008, Ekiti 2003/2004, Enugu 2003/2004, Jigawa 2003/2004, Kwara 2003/2004, Niger
2003/2004, Ogun 2003/2004/2008, Oyo 2003/2004, Rivers 2003/2004, Taraba 2003/2004, Yobe 2003/2004, Zamfara 2003, AkwaIbom 2004/2008, Borno 2004/2008,Ebonyi 2004/2008, Ondo 2004/2008, Osun 2004/2008. This results in LGA clusters in 171
state/years.

97

Table 3.3: SUMMARY
STATISTICS—
STATE MODEL
MEAN

STANDARD
DEVIATION

MIN

MAX

56.0

14.4

14.6

88.8

1262.0

2269.1

31.0

15719.4

7635.0

10040.6

900.4

87409.7

792.0

620.9

116.1

4433.4

1014.0

2019.6

0.0

19651.2

6.6
9828.9

0.2
6970.6

6.2
2440.2

7.0
29483.9

546077.9

234241.5

6129.9

3628.1

2466.0

25577.0

100.8

23.1

66.0

125.0

58.0

13.3

17.3

88.8

39

OBSERVATIONS
VARIABLE
RECURRENT
EXPENDITURES
246
(STATES, %)
TAXES/CAPITA
246
(NAIRA)
TRANSFERS/CAPITA
246
(NAIRA)
VAT/CAPITA
246
(NAIRA)
GRANTS/CAPITA
246
(NAIRA)
LOG POPULATION
246
AREA
246
PRIMARY
246
EDUCATION
HEALTH
246
FACILITIES/CAPITA
INFANT
246
MORTALITY
PREVIOUS YEAR'S
RECURRENT
EXPENDITURES (%
246
TOTAL
EXPENDITURES)
39

731611.0 1443848.0

In this model, I’m including 36 states over 7 years, which should result in 252 state/year observations. However, due to missing
data in Edo 2002/2003, Niger 2002/2003, Jigawa 2007/2008, there are 246 state/year observations.
98

Table 3.4: LGA
40
MODEL
RECURRENT
EXPENDITURES
(% TOTAL
EXPENDITURES)
TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

-0.059*
(0.001)
0.154*
(0.005)
-0.247*
(0.0004)
0.109*
(0.009)

AREA

-0.667
(7.201)
-0.019
(0.0001)

PRIMARY
EDUCATION

0.146
(0.0006)

HEALTH
FACILITIES/CAPITA

-0.109
(0.0003)

LOG POPULATION

INFANT
MORTALITY

-0.139
(0.053)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.072*
(0.107)

CONSTANT
100.411*
N
171
R2
0.303
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

40

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
99

Table 3.5: STATE
41
MODEL
RECURRENT
EXPENDITURES
(% TOTAL
EXPENDITURES)
TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

-0.160*
(0.0006)
0.136*
(0.0001)
-0.017
(0.003)
0.052
(0.0005)

AREA

-0.069
(7.795)
.0011
(0.0001)

PRIMARY
EDUCATION

-0.035
(0.0004)

HEALTH
FACILITIES/CAPITA

-0.034
(0.0003)

LOG POPULATION

INFANT
MORTALITY

-0.102
(0.048)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.019*
(0.075)

CONSTANT
93.097*
N
246
R2
0.135
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

41

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
100

Table 3.6 : SUMMARY
STATISTICS-IMPUTED LGA
MODEL
VARIABLE
RECURRENT
EXPENDITURES (%
TOTAL
EXPENDITURES)
TAXES/CAPITA
(NAIRA)
TRANSFERS/CAPITA
(NAIRA)
VAT/CAPITA
(NAIRA)
GRANTS/CAPITA
(NAIRA)
LOG POPULATION
AREA (SQ. MILE)
PRIMARY
EDUCATION
HEALTH
FACILITIES/CAPITA
INFANT
MORTALITY
PREVIOUS YEAR'S
RECURRENT
EXPENDITURES (%)

OBSERVATIONS

MEAN

MIN

MAX

68.4

STANDARD
DEVIATION
13.6

216

4.0

97.1

216

4616.4

23784.7

4.5

287439.3

216

121196.7

506820.6

97.0

4273885.0

216

17376.7

78027.1

7.7

751986.4

216

4001.5

29834.0

0.0

382737.8

216
216
216

6.6
9828.9
546077.9

0.2
6970.6
234241.5

216

6129.9

3628.1

2466.0

25577.0

216

100.8

23.1

66.0

125.0

216

69.8

11.6

4.0

97.1

101

6.2
7.0
1341.7
29483.9
177830.0 1443848.0

Table 3.7: SUMMARY
STATISTICS-IMPUTED STATE
MODEL
VARIABLE
RECURRENT
EXPENDITURES
(STATES, %)
TAXES/CAPITA
(NAIRA)
TRANSFERS/CAPITA
(NAIRA)
VAT/CAPITA
(NAIRA)
GRANTS/CAPITA
(NAIRA)
LOG POPULATION
AREA
PRIMARY
EDUCATION
HEALTH
FACILITIES/CAPITA
INFANT
MORTALITY
PREVIOUS YEAR'S
RECURRENT
EXPENDITURES (%
TOTAL
EXPENDITURES)

OBSERVATIONS

MEAN

STANDARD
DEVIATION

252

56.4

14.3

14.6

88.8

252

1257.9

2255.9

31.0

15719.4

252

7608.2

9983.9

900.4

87409.7

252

791.1

617.4

116.1

4433.4

252
252
252

1012.9
6.6
9828.9

2008.9
0.2
6968.3

0.0
6.2
1341.7

19651.2
7.0
29483.9

252

556954.7

258852.5

252

6515.2

4214.2

2466.0

25577.0

252

100.8

23.1

66.0

125.0

252

57.5

13.2

17.3

88.8

102

MIN

MAX

177830.0 1443848.0

Table 3.8: LGA
42
IMPUTED MODEL
RECURRENT
EXPENDITURES (%
TOTAL EXPENDITURES)
-0.181*
(0.0009)
0.198*
(0.0002)
-0.110*
(0.003)
0.075*
(0.0009)

TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

AREA

-0.048
(6.379)
.0011
(0.0001)

PRIMARY
EDUCATION

-0.131
(0.0004)

HEALTH
FACILITIES/CAPITA

-0.117
(0.0002)

LOG POPULATION

INFANT
MORTALITY

-0.139
(0.049)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.065*
(0.103)

CONSTANT
89.578*
N
216
2
R
0.249
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

42

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
103

Table 3.9: STATE
43
IMPUTED MODEL
RECURRENT
EXPENDITURES (%
TOTAL EXPENDITURES)
-0.174*
(0.0006)
0.167*
(0.0001)
-0.055
(0.003)
0.032
(0.0005)

TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

AREA

-0.068
(7.721)
.006
(0.0001)

PRIMARY
EDUCATION

-0.033
(0.0004)

HEALTH
FACILITIES/CAPITA

-0.049
(0.0003)

LOG POPULATION

INFANT
MORTALITY

-0.098
(0.046)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.014*
(0.075)

CONSTANT
90.169*
N
252
2
R
0.139
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

43

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
104

Table 3.10:
OBSERVATIONS WITH
HIGHEST COOK'S D
VALUES
LGA-STATE YEAR
COOK'S D
BAYELSA 2004
0.102
CROSS RIVER 2004
0.038
DELTA 2004
0.225
GOMBER 2004
0.045
KADUNA 2004
96.836
KANO 2004
0.564
LAGOS 2004
9.776
DELTA 2008
0.026
ENUGU 2008
0.029
LAGOS 2008
0.154
NASSARAWA 2008
0.030
YOBE 2008
0.033
Notes: For interpretation of the references to color in this and all other
tables and figures, the reader is referred to the electronic version of this
dissertation.

105

Table 3.11:
OBSERVATIONS WITH
HIGHEST COOK'S D
VALUES
STATE YEAR
NIGER 2000
YOBE 2002
BAUCHI 2003
KANO 2003
BAYELSA 2007
KANO 2007
RIVERS 2007
BAYELSA 2008
EBONYI 2008
EBONYI 2008
KANO 2008
RIVERS 2008
TARABA 2008
ABIA 2009
AKWA-IBOM 2009
BAYELSA 2009
EBONYI 2009

106

COOK'S D
0.023
0.029
0.020
0.027
0.031
0.068
0.038
0.019
0.029
0.024
0.024
0.035
0.024
0.010
0.067
0.133
0.022

Table 3.12: LGA
MODEL WITHOUT
44
OUTLIERS
RECURRENT
EXPENDITURES (%
TOTAL EXPENDITURES)
-0.051*
(0.0002)
0.097*
(0.0004)
-0.221*
(0.009)
0.083*
(0.0009)

TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

AREA

-0.585
(7.001)
.0009
(0.0064)

PRIMARY
EDUCATION

-0.023
(0.0003)

HEALTH
FACILITIES/CAPITA

-0.112
(0.0002)

LOG POPULATION

INFANT
MORTALITY

-0.115
(0.056)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.054*
(0.112)

CONSTANT
97.474*
N
169
R2
0.266
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

44

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
107

Table 3.13: STATE
MODEL WITHOUT
45
OUTLIERS
RECURRENT
EXPENDITURES (%
TOTAL EXPENDITURES)
-0.099*
(0.0006)
0.093*
(0.0001)
-0.011
(0.003)
0.036
(0.0005)

TAXES/CAPITA
TRANSFERS/CAPITA
VAT/CAPITA
GRANTS/CAPITA

AREA

-0.041
(7.721)
.0010
(0.0001)

PRIMARY
EDUCATION

-0.023
(0.0004)

HEALTH
FACILITIES/CAPITA

-0.032
(0.0003)

LOG POPULATION

INFANT
MORTALITY

-0.092
(0.046)

PREVIOUS YEAR'S
RECURRENT
EXPENDITURES

0.018*
(0.075)

CONSTANT
90.169*
N
245
R2
0.131
Notes: * significant at the 0.05 level.
Standard errors are in parentheses below the coefficients.

45

Ordinary Least Squares regressions run with Stata 12.0; panel corrected for standard errors.
The regression is run with dummy variables for every year (but one) covered by the data.
108

CHAPTER FOUR

Leveraging Regional Variation in Tax Generation and Oil Dependence
Background
In the previous chapter, I established how taxation and representation interact at a macrolevel in Nigeria. The pattern of revenue reliance shapes budgetary priorities. In states and local
governments that derive more internally generated income from non-oil related taxes on citizens,
politicians spend less on government salaries and allowances. On the other hand, states and
LGAs with a higher level of dependence on external sources of revenue (e.g. federal transfers of
oil revenue and foreign aid) engage in more salary spending. As I continue to investigate
revenue and representation linkages in Nigeria, it becomes important to consider how tax
extraction and dependence on oil income vary within and between regions. In doing so, I can
examine how sources of government income may influence individual elected officials’ behavior
at the micro-level.
In this chapter, I will investigate how reliance on different forms of public revenue
influences individual elected officials. Scholars argue that in Western Europe, when
governments generate income from taxes on citizens, elected leaders are more likely to shift
public policy decisions to reflect the populace’s priorities. Political elites do so in order to fulfill
their part of the fiscal contract with citizens (Tilly, 1985/1990; Bates and Lien, 1985; Levi,
1988). I rely on original elite surveys with legislators to test this theory in Nigeria. Due to
limited research resources, I narrow down my cases, focusing on elites in regions with the most
variation in state tax generation. Therefore, this chapter begins with an analysis of regional
variation of tax generation and oil dependence in Nigeria. After selecting two regions (and six

109

states) as cases, I profile revenue extraction and elite representation in each state. From here, I
outline my sampling procedures and survey methodologies. Using data from these elite surveys,
I test the hypothesis that politicians in states with higher levels of tax generation are more
representative of their constituents.

Regional Variation in Tax and Oil Dependence: Case Selection
Although Nigeria abandoned use of the “region” as a subnational unit in the Second
Republic, regional characteristics continue to play a role in Nigerian politics. According to
Suberu (2001: 219), “in order to promote the equitable and stable distribution of powers and
resources in the country,” political leaders aggregate the Nigerian states into geopolitical zones.
The North-West, North-East, Middle Belt (i.e. Lower North), South-West, South-South, and
South-East constitute these regions.

46

The six units reflect broad regional, cultural, and ethno-

political cleavages, which historically characterize the federation (Paden, 1990). For example,
each of Nigeria’s three major ethno-linguistic groups dominates specific regions.
Demographically, the Northern Region (e.g. North-West, Middle Belt, and North-East zones) is
predominantly Muslim, with an ethnic Hausa-Fulani majority. The South-Western Region is
largely Yoruba and roughly balanced among Muslims and Christians. Last, the Eastern Region,
made up of the South-South and South-East zones, is predominantly Igbo and overwhelmingly
46

The contemporary Nigerian states included in each zone:
North-West: Jigawa, Kaduna, Kano, Katsina, Kebbi, Sokoto, and Zamfara.
Middle Belt: Federal Capital Territory (Abuja), Benue, Kogi, Kwara, Nasarawa, Niger, and
Plateau.
North-East: Adamawa, Bauchi, Borno, Gombe, Taraba, and Yobe.
South-West: Ekiti, Lagos, Ogun, Ondo, Osun, and Oyo.
South-South: Akwa-Ibom, Bayelsa, Cross River, Delta, Edo, and Rivers.
South-East: Abia, Anambra, Ebonyi, Enugu, and Imo.

110

Christian (Lewis, 2011: 3). Given the ethno-political salience of zonal distinctions, how do tax
income generation and oil dependence vary within and between regions?
In Chapter Two, Table 2.9 presents the variation in average tax and oil dependence across
Nigeria’s 36 states between 1999 and 2009. However, since scholars (e.g. Paden, 1997; Suberu,
2001) argue politicians take regional groupings into account when distributing resources, it is
important to consider variation in state governments’ reliance on tax income (vs. petroleum
income) within these subnational categories.
Overall, states in the Southern portion of Nigeria generate more income from taxes on
citizens than their counterparts in the Northern portion of the country. On average, between
1999 and 2009, Southern States obtained 15% of their income from taxation of citizens.
However, during this same period, Northern States derived 8% of their revenue from taxes.
Between 1999 and 2009, states in Nigeria’s South-West Region were the most capable of
extracting tax revenue. Table 4.1 presents tax revenue as a percentage of total income in the
South-Western States in recent years. By 2009, four out of the six states in this region were
generating over 20% of their income from taxation of citizens. In fact, Lagos (located in the
South-West) was the only state in the federation where income from direct taxation of citizens
superseded petroleum revenue (between 1999 and 2009, 54% of their revenue from taxes, 24%
from oil transfers).
On the other hand, The North-East Region generated the least amount of their income from
taxation of its citizens: from 1999 to 2009, in this region, only 5% of state revenue came from
non-oil, tax sources. As seen in Table 4.2, between 2005 and 2009, all of the North-East states
managed a steady increase in their tax generation. In fact, Gombe and Taraba States were able to
at least double taxes as percentage of their total revenue. But, with the exception of Gombe,
111

North-Eastern states still derived less than 10% of income from taxes on citizens.
This finding that Southern states out performing Northern states in tax generation is
partially contradictory to previous research. Scholars typically argue that states in Northern
Nigeria have a historical experience as strong tax generators.
Berger (2009) asserts that, from the time of British colonialism, government units in
Northern Nigeria have been characterized with a higher capacity to raise income from the direct
taxation of citizens:
“The government of the Southern Protectorate [of Nigeria] had an obvious tax handle in
import tariffs levied at the ports…[But] Since the Protectorate of Northern Nigeria had no
seacoast and thus no ports the northern government was unable to raise money from tariffs.
Therefore, Lugard [colonial governor in the North] raised the money he needed by a series
of direct taxes. In the Muslim areas farther to the north there were property taxes and taxes
on livestock. However, in what they referred to as the “pagan areas” they used a simple
poll-tax” (Berger, 2009: 10).
He goes on to argue that this earlier experience with developing and maintaining a tax
administration manifests in contemporary Nigeria. Berger (2009) finds that, today, local
governments north of the 7! 10" N line have more efficient bureaucracies that are better able to
47

deliver public services.

However, the budgetary data presented and discussed here do not corroborate Berger’s
(2009) assertions. Southern states generate tax revenue at a higher rate. Moreover, besides
states engaged in oil production, Northern governments are the most dependent on petroleum
income from federal transfers (Table 4.3).

48

It is especially interesting to note that oil-producing

47

In 1899, the British Colonial Authority divided Nigeria into the Northern and Southern
Protectorate using the 7! 10" N line (Berger, 2009).
48

The following Nigerian states are oil producing: Abia and Imo in the South-East; Akwa Ibom,
Bayelsa, Cross River, Delta, Edo, and Rivers in the South-South; Ondo in the South-West
112

states still out perform Northern governments in terms of tax generation. In oil producing states,
between 1999 and 2009, on average, 12% of government income was derived from non-oil taxes
on citizens. However, in Northern states, about 8% of revenue was generated from taxes.
During this same period, the Southern, non-oil producing states were the most capable of
extracting tax revenue: 18% of government income was derived from internally generated tax
income. As analysis in Chapter Three demonstrates, states and local governments that extract
more income from tax revenue are less likely to spend income on politicians’ salaries. Inferring
from the pooled analyses in the previous chapter, it is probable that because Southern, non-oil
producing states have a greater tax-capacity, they also spend less resources on government
salaries and allowances. On the other hand, contrary to Berger (2009), Northern states and
Southern oil producing states would be more likely to contribute a greater share of revenue to
salary spending.
In the majority of Nigeria’s states, the tax administration inherited from their colonial
legacy has stagnated or declined. In the immediate post-colonial era, Nigeria underwent a
centralization of political and financial authority (Suberu, 2004). Berger (2009) discusses how
during the colonial period, subnational entities had authority to generate their own income (e.g.
through direct taxation of citizens, tariffs, and income from agricultural marketing boards).
However, as Nigeria gained independence, regional hyper-autonomy repeatedly threatened
federal authority, coalescing in the Biafran Civil War.

49

As a result, Nigeria’s subsequent

military and civil governments held power closely in the center, ensuring the states’ dependence
(Oviasuyi and Uwadiae, 2010). In addition to the standard allocation of petroleum revenue,
these states receive an additional 13% of national petroleum income, beyond the standard federal
allocation.
49

See Chapter Two: Neopatrimonialism, Federalism, and Citizen-Elite Relations in Nigeria: An
Historical Overview.
113

on the national government. As Suberu (2001: 27) argues: “the central authorities became
responsible for the legal administration and collection of the most lucrative revenue sources.”
This included exclusive rights to all revenue produced from the petroleum industry (enforced by
the 1969 Petroleum Decree) and import, export, excise, business, and mining taxes. Suberu
(2001: 47) also asserts “the allocation of the most important tax jurisdictions to the center
[created] heavy reliance on the redistribution of centrally collected revenues [to the states].”
Thus, as state authority over taxation decreased, these politicians substituted oil revenue for their
income. This combination of factors resulted in the deterioration of state tax administrations.
Since Nigeria’s return to democratic governance and the creation of the (contemporary)
Fourth Republic in 1999, constitutional authority over taxation has been dispersed, the federal
50

government sharing power with states.

States vary in the way they have generated tax revenue

under this new system. For example, states in Southern Nigeria (particularly in the South-West
Region) have been able to steadily increase the capacity of their tax administration. However, as
the data suggests, this has not been the case in much of Northern Nigeria. Suberu (2001: 69,
2004: 341) provides justification for why this may be the case:
“the landlocked and relatively more economically depressed North depends more heavily
than the South on the present system of centralized distributive federalism. Thus, while
they [Northern political leaders and citizens] have supported the transfer of more federally
collected revenue to the subfederal authorities, the Northern states are averse to any
proposals for regional resource control [or autonomous revenue generation within the
states].”
Berger (2009) previously notes that during the colonial period, the landlocked, Northern states
turned to direct taxation in order to raise the money needed to govern. However, in
contemporary Nigeria, like Suberu (2001, 2004) argues, access to petroleum revenue from
50

Chapter Two Nigeria’s Tax System: Mobilizing Revenue At Three Levels of Government
outlines the division of federal, state, and local tax authority in contemporary Nigeria.
114

federal transfers allows Northern politicians to bypass developing their internal revenue
generation capacity via taxes on citizens.
With that said, some Northern states, particularly in the North-West Region, have been
able to make advances in increasing their tax administration’s capacity. For example, between
2005 and 2009, Jigawa, Kaduna, and Kano States were able to double the percentage of total
revenue made up from taxes (Table 4.4). It is important to note that these three states began
from a low base. In 2005, tax revenue made up 7% of Jigawa and Kaduna States total income.
However, by 2009, in these two states, about 15% of revenue was derived from taxes on citizens.
The same could be said for Kano State, where between 2005 and 2009, tax income as a portion
of total revenue doubled, growing from 11% to 22%. Furthermore, taxes as a portion of total
income increased by almost five times in Kebbi, Sokoto, and Zamfara States. In Kebbi and
Zamfara States, in 2005, roughly 3% of total revenue was obtained via taxes on citizens. But by
2009, taxes as a portion of state income grew to 14% in Kebbi State and 15% in Zamfara State.
Similarly, in 2005, 9% of Sokoto State’s income came from taxes. By 2009, tax revenue was
48% of Sokoto’s total income. In fact, in 2008 and 2009, Sokoto followed Lagos as the most tax
dependent state in the Nigerian federation. Unlike in the South-West Region, where by 2009,
four out of six states were generating over 20% of their income from taxes on citizens, this is
only true of two North-Western states (Kano and Sokoto). Even though this is the case, unlike
other regions in Northern Nigeria, where tax revenue generation has remained relatively
stagnant, the North-West Zone has been able to make some progress.
Given this intra and inter-regional variation in tax generation and oil dependence, what
are the consequences for elite attitudes and behavior toward representation of citizens? In
Chapter Three, I have demonstrated how, overall in Nigeria, when a state obtains more of its

115

income from taxes, it spends less on government salaries and allowances. Instead, these states
spend more resources on non-recurrent expenditure, including investing in public services.
Looking within regions, does this relationship also hold? Does varied reliance on tax vs. oil
income change individual elite policy behavior?
By leveraging intra-regional variation, it is possible to investigate the effect of revenue on
the incentives and preferences of individual elected officials. I hypothesize that elected officials
in states with a greater capacity to generate tax income will be more representative of their
constituents. I expect this based on theories of revenue extraction and representation in Western
Europe (Tilly, 1985/1990; Bates and Lien, 1985; Levi, 1988; North and Weingast, 1989;
Acemoglu and Robinson, 2005) and Africa (Fjeldstad and Semboja, 2001; Fjeldstad, 2001;
Gibson and Hoffman, 2006; McGuirk, 2012). Elite opinions about revenue and representation
are not readily available for the purpose of analysis. Thus, I have conducted interviews with
Nigerian state legislators. These interviews gauging legislators’ policy perspectives and
51

priorities are an original contribution of this project.

Ideally, I would have completed

interviews with legislators in all 36 Nigerian states; however, due to limited resources, I select
states from two regions with contrasting patterns of tax capacity and oil dependence to test my
hypotheses. By focusing on the regions with the greatest variability in revenue generation, I can
more effectively isolate any influence tax capacity or oil dependence has on elite perceptions of
51

The African Legislatures Project (ALP) is an academic and policy project, located in the
Centre for Social Science Research at University of Cape Town (South Africa). This project
utilizes both qualitative and quantitative approaches to uncover “how and why do African
legislatures function as they do” and “what constitutes ‘best practice’ for the purpose of
strengthening African institutions which are an essential component of democratization on the
continent.” So far, analysis has been conducted in Kenya, Malawi, Mozambique, Namibia,
Tanzania, Uganda, and Zambia. While research in Nigeria is currently underway, it is important
to note that this project focuses exclusively on National Legislatures. Therefore, my focus on
state legislators is an original contribution.
116

representation. The next section outlines the case selection strategy.
Looking at average tax and oil income distribution (as percentage of total government
revenue) between 1999 and 2009, Table 4.5 presents the standard deviation each of Nigeria’s six
regions. A larger standard deviation indicates more within-region variation. As we can see, the
South-West Region has the highest level of variability in tax and oil revenue reliance between
1999 and 2009. During this period, South-Western states’ average tax generation (tax income as
a percentage of total revenue) was 21%. On average, states in the South-West varied from this
mean by 17%. In terms of oil dependence, states’ in the South-West derived 57% of their
income from petroleum transfers. On average, these states varied from the regional mean by
18%. The South-South Zone followed with the next highest level of variation. In this region,
between 1999 and 2009, on average, 12% of states’ income was derived from taxes on citizens.
On average individual South-South states varied from the regional mean by 6%. During this
same time frame, 67% of the South-South states’ revenue came from oil income. On average,
states varied from the regional mean by 8%. However, since states in this region produce
petroleum, they receive an additional 13% of oil income from federal transfers beyond the
standard federal allocation.

52

As another measure of control, I will not to focus on petroleum-

producing states in my analysis. Therefore, the zone with the next highest level of within-region
variability in revenue generation is the North-West. In this region, from 1999 to 2009, on
average, tax income made up 10% states’ income. Individual states varied from this regional
mean by 4%. In this same period, North-Western states, obtained 64% of their income from
petroleum wealth. On average, these states varied from the regional mean by 4%.
Thus, in order to maximize variation, I select the South-West and North-West Zones for
52

Referred to as the derivation principle, please see Chapter Two: Distributing Petroleum
Wealth Across the Three Tiers for further discussion.
117

my continued analysis of the revenue and representation relationship at the micro-level.
Focusing on these two regions, I investigate the hypothesis that, elected officials in tax-reliant
states will be more representative of their constituents.
Having selected the South-West and North-West Regions for analysis, this narrows down
my possible cases from 36 to 13 states (six states in the South-West, seven states in the NorthWest). Once again, due to limited resources, I limit my cases further, choosing a high performer,
53

average performer, and low performer with respect to tax generation in these two.

As a result,

I am able to control for regional effects by looking at high, average, and low tax generators
within both regions. Tables 4.6 and 4.7 present the South-West and North-West’s average tax
generation and oil income dependence between 1999 and 2009. Based on this data, I select high,
average, and low performers in each region. Therefore, the cases in the South-West are Lagos
(high), Oyo (average), and Ekiti (low).

54

The cases in the North-West are Sokoto (high), Kano

(average), and Katsina (low).
In the next section, I describe these six cases, providing background information on the
structure of each state’s revenue sources since Nigeria’s return to democratic rule. I will also
discuss the organization of each state’s legislature. Within this context, we can delve deeper into
the revenue and representation linkages, considering elite perspectives.

53

High, average, and low performance is with respect to the national average of tax and oil
income as percentage of total state revenue. Between 1999 and 2009, on average taxes make up
13% of state income, and average oil dependence is 64%.
54

Since Ondo State is a petroleum producing state and a recipient of derivation funds, I do not
include it in my analysis.
118

Revenue in the South-West and North-West Cases
As previously mentioned, since Nigeria’s return to democratic governance in 1999, the
South-West and North-West Regions have led the nation in tax generation. However, referring
to Tables 4.8 and 4.10, we can observe variation in tax generation within these zones.

Tax Generation and Oil Dependence in the South-West
Between 1999 and 2009, Lagos State (high-performer) generated an average of 53% of
state income from taxes on citizens. This made Lagos the most capable with tax extraction
during this period, not only in the South-West, but also in the Nigerian Federation. Turning to a
graphic representation of the data (Figure 4.1), we see, for the most part, the portion of Lagos’
state income derived from taxes steadily rose, beginning with taxes comprising 41% of revenue
in 1999 and reaching 64% in 2009.
Comparing Lagos to Oyo State, the average-performer, we observe a noticeable gap. From
1999 to 2009, Oyo State earned a mean of 17% of revenue from taxes. Though tax generation in
Oyo also steadily increased during this period, it did not reach the same rates as Lagos State
(Figure 4.1). Starting with taxes making up 10% of state income in 1999, Oyo reached a peak of
22% in 2007, then slightly reducing to 21% in 2009 (Table 4.8).
Finally, in Ekiti State, the low-performer, taxes constituted an average of 8% of revenue
between 1999 and 2009 (Table 4.8). However, as evident in Figure 4.1, Ekiti (like Lagos and
Oyo) experienced an increase in tax dependence during this time frame. With that said, taxes as
a percentage of total revenue increased at a notably slower rate than the other two case-states.
Beginning with taxes making up 3% of Ekiti’s income in 1999, tax generation crested at 13% in
2009.

119

The picture of petroleum income dependence in the South-Western cases is markedly
different. While tax generation in Lagos, Oyo, and Ekiti States increased at relatively steady,
albeit different rates, oil income as a portion of state revenue was more volatile. This confirms
Ross’s (2012) assertion that the volatility of world oil prices can produce fluctuations in
government finances. Figure 4.2 provides a graphic representation of the South-Western oil
dependence data in Table 4.9.
On average, Lagos State was the least oil dependent government in this zone, as well as
among all Nigerian states. From 1999 to 2009, oil revenue from federal transfers made up 24%
of Lagos’ government income (Table 4.9). Oil dependence reached a peak of 39% in 2001,
decreasing to petroleum income comprising 20% of Lagos State’s revenue. In 2005, oil revenue
once again increases to 26%, bottoming out in 2007 at 11%, before rising to 15% of state income
in 2009. Though petroleum revenue as portion of Lagos’ income declined overall during this
period, this source of funding was also unstable from year to year.
During the same time period, oil income made up a larger portion of state revenue in Oyo
and Ekiti States. However, like in Lagos, petroleum wealth remained a volatile component of
Oyo and Ekiti’s income, declining over time. For example, in 1999 54% of Oyo State’s revenue
was obtained via federal transfers of petroleum income. Oyo’s oil dependence peaked in 2001
(78% of state income derived from oil revenue). By 2009, the portion of income from petroleum
decreased to 49%, the lowest point in Oyo during this 10-year frame. Ekiti State, the most oil
dependent state of my cases, followed a similar pattern. In 1999, 74% of Ekiti’s income was
derived from oil wealth. Peaking in 2001 (at 86%), Ekiti’s oil dependence continued to
fluctuate, before then declining to 58% of state revenue coming from petroleum transfers in
2009.

120

Overall, in the South-West cases, we see that as tax generation and reliance has increased,
the trend in oil dependence has decreased. Moreover, evidence confirms previous scholars’
assertions about the instability in petroleum income and the volatile influence on state budgets.

Tax Generation and Oil Dependence in the North-West
Turning to the North-West cases, we observe similar patterns in the growth of states’ tax
dependence and volatility, but gradual decrease in oil income reliance. Unlike in the South-West
Zone, where the high performing state is far ahead of the other cases in tax generation, Sokoto,
Kano, and Katsina States started off at similar levels of tax dependence. Referring to Table 4.10
and Figure 4.3, we see that, in 1999, tax income made up between 3% and 7% of revenue in
these cases. In Sokoto (high-performer), between 1999 and 2005, tax dependence ranged from
5% to 9% of state income. However, starting in 2007, tax income as a portion of total revenue
jumped to 19%, and reached a new height of 48% in 2009. During this same period, Kano State
(average performer) experienced a relatively more stable growth in tax reliance. Beginning in
1999, when taxes comprised 7% of Kano’s state income, over this ten-year time frame, tax
dependence peaked at 22% by 2009. Last, in Katsina State (low performer), we also observed a
stable, but comparatively modest rate of growth in tax dependence. In 1999, 3% of Katsina’s
income was derived from taxes on citizens. This increased, reaching its highest point at the end
of the decade, where 12% of Katsina’s state revenue was tax-based. Thus, starting in 2007,
Sokoto State began to surpass other North-Western states in tax generation. And by 2009,
Sokoto was the second most tax-dependent state in Nigeria (following Lagos States). Likewise,
by the end of this time period, the internal variation in tax reliance within the North-West Zone is
evident.

121

Oil reliance in the North-West cases follows the same trend of instability as in SouthWestern states. In 1999, Kano State was the least tax dependent, with 52% of revenue coming
from federal petroleum income transfers (Table 4.11, Figure 4.4). At its lowest point, 39% of
Kano’s revenue cam from oil income. Oil dependence in Kano reached its peak two years later,
when federal transfers made up 77% of state revenue. Between 1999 and 2009, oil dependence
in Kano State fluctuated, and by the end of that time period, 53% of revenue was derived from
petroleum wealth. Similarly, over ten years, Katsina’s access to petroleum income via federal
transfers was volatile. In 1999, 76% of Katsina’s revenue came from oil-based income. This
remained relatively steady before dropping to 47% in 2002. Oil dependence peaked in 2005
when 78% of Katsina’s income was obtained from oil wealth. By 2009, federal transfers as a
potion of state income dropped to 61%. On the other hand, in Sokoto, after oil dependence
reached its height in 2000 (84% of revenue from oil income), reliance on federal transfers
decreased. By 2009, 34% of Sokoto’s revenue was derived from petroleum funds, once again
making this state one of the least oil dependent in the Federation.
As previously mentioned, tax generation has been on the rise in Nigeria since 1999.
Though occurring at different rates, this trend applies to both South-Western and North-Western
states (Figures 5.7 and 5.9). I argue the expansion of the state revenue boards’ authority to
collect taxes and punish evaders explains this development. Moreover, the leap in Sokoto’s
capacity to extract taxes provides a concrete illustration.
Efforts to improve tax capacity and remove barriers to tax management across Nigeria
have been spearheaded by a Joint Tax Board (JTB) campaign. Though Project FACT (Friendly,
Accurate, Complete, and Timely), state tax administrations (SIRS) have begun automating tax

122

collection.

55

As individuals and corporations register and receive taxpayer identification

numbers (TIN), state tax administrations can identify and access information about taxpayers.
This brings more taxpayers under the tax net, allowing tax authorities and citizens to determine
actual income and tax burden. Taxpayers then remit the tax payment to one of the approved
collecting banks and receive a ticket as proof of payment. Through this process, tax agents are
removed from accepting direct payments or money. This has helped to curtail fraud. Thus, as
more taxpayers (individual and corporations) have obtained their TIN numbers and participated
in the automated tax payment system, state tax administrations have been able to increase states’
income from taxes.
Sokoto State provides a good example of Project FACT’s positive influence. In an
interview with a high-level administrator in the Sokoto SIRS, he attributes the spike in Sokoto’s
tax generation, in part, to the success of Project FACT: “With automation, now the taxpayer
feels more comfortable. We [at SIRS] do not take cash. The taxpayer pays taxes to the bank and
they get a receipt. Now that there is computerization, they can check their own records”
(Director, Sokoto State Board of International Revenue, Personal Communication, June 21,
2011).
In addition to federal support and coordination, the rise in state tax boards’ authority and
effectiveness is also attributed to political initiative from state legislators. Returning to the
Sokoto State tax administrator: “sustaining the success of internally generated tax revenue is a
problem without politicians’ support.” Again, Sokoto provides a good example of the role
political support can play in cultivating a strong tax administration. In 2007, members of

55

Chapter Two’s discussion of Joint Tax Board (under Nigeria’s Tax System: Mobilizing
Revenue At Three Levels of Government) provides a comprehensive outline of Project FACT.
123

Sokoto’s tax administration (supported by Governor Aliyu Wamakko) began exploring best
practices in tax extraction and methods of increasing state tax capacity. Petitioning the Sokoto
State House of Assembly, tax administrators joined members of the House Committees on Public
Accounts and Finance and Appropriation to craft legislation to empower the Board of Internal
Revenue. In 2008, the House of Assembly passed the Sokoto State Board of Internal Revenue
Law. This law established the Board’s financial and legal autonomy. It also provided the tax
administration with the legal framework to collect taxes from citizens and pursue and punish tax
evaders. According to the afore-mentioned Sokoto State tax administrator: “With this autonomy
law, the board can now act,” and this action has taken several forms. With their legislative
authority, SIRS has used “carrot and stick” methods to encourage tax payment. In terms of
carrots, Sokoto SIRS engages in campaigns to educate citizens about the tax burden and the new
process for payment. The Sokoto State tax administrator explains, “[Sokoto] SIRS has tax
collection committees. They are made up of field officers who go door to door to educate
[citizens] about taxes. They also have public meetings with unions and other local groups.” He
also highlights the SIRS use of media services via TV and radio jingles: “Tax payers should
know what they should be paying. Education first, then enforcement.” The “sticks” SIRS now
has the authority to use include the power to audit individuals and corporations, fine those who
are non-compliant, and use the state legal system prosecute tax evaders in court.
But more so than force, this tax administrator believes: “people will pay taxes if they feel
government is working. There must be projects and infrastructure on-ground. When government
performs, the citizen feels compelled to pay [taxes]. It is the responsibility of the [government]
administration to be wise with the new tax revenue.” Thus, even though the Joint Tax Board
aims to increase tax capacity across all Nigerian states, political initiative via legislation factors

124

into a given state’s level of success in increasing tax income. In Chapter Six, I continue this
investigation of the role of political initiative in broadening state tax capacity. An inter-regional
comparison of Lagos and Kano states corroborates these findings in Sokoto.

Representation in the South-West and North-West Cases
As examined in Chapter One, legislatures are the chief political institution engaged in
representation. According to Barkan (2009: 7), “legislatures are the institutional mechanism
through which societies realize representative governance on a day-to-day basis.” Different from
the executive, who “as heads of state, are expected to synthesize, balance, and aggregate interests
into ‘the national interest’,” a legislator is beholden to the interests of his/her constituency. Thus,
the legislative body, as a whole, serves as an arena where various constituent preferences are
conveyed, debated, and converted into public policy. In my analysis of representation in Nigeria,
and the influence of revenue on representative practice, I concentrate on state legislatures.
Barkan (2009: 2) argues, “most African legislatures, like legislatures worldwide, remain
weak in relation to the executive.” In particular, the post-independence era of neopatrimonial
rule curtailed the development of legislatures in Africa. For example, during this period,
Nigeria’s legislature was completely disbanded in lieu of military-authoritarian governance
(although there were brief moments of civilian rule, i.e. First Republic (1960 – 1966), Second
Republic (1979 – 1983); see Chapter Two). In other cases (e.g. Zambia), the legislature
functioned as an executive tool for delivering patronage to citizens in an effort to legitimize the
authoritarian regime (Barkan, 2009: 13 – 15). Rather than serving as an independent branch of
government—providing oversight or generating policy—legislatures were wholly dependent on
the executive. In both circumstances, as African countries returned to multi-party democracy in

125

the 1990s, legislatures emerged from the “downward spiral of conditions that had undermined
them for a quarter-century” (Barkan, 2009: 15).
While this previous scholarship has focused exclusively on national legislatures, I argue
that the same assertions of legislative development apply subnationally. Especially in Nigeria,
the return to democracy in 1999 has been accompanied by a two-pronged federalism. First, there
has been a distribution of authority between branches of government, equipping the national
legislature with the ability to form policy independently while serving as a check on the
executive. Second, political power has been dispersed through levels of government. State
legislatures are empowered, by law, to govern, independently of the federal government and the
state executive (governor).
According to Nigeria’s 1999 Federal Constitution, state legislatures (Houses of Assembly)
are unicameral bodies endowed with the power to “make laws for the good governance of the
state” (Omoyele, 2011: 8). The membership of each state’s House of Assembly varies,
comprised of between 24 and 40 members (three of four times the number of seats the state has
in the House of Representatives; 1999 Nigerian Constitution, Chapter 5, Part 2, Section 91). The
members of each House of Assembly also nominate and elect from among themselves several
leading officers. The Federal Constitution specifically names the Speaker (presiding over the
activities of the state legislature, keeping order and decorum, and ensuring House rules are
obeyed) and the Deputy Speaker (assisting the Speaker in his/her duties, presiding over the
Assembly in the Speaker’s absence) (1999 Nigerian Constitution, Chapter 5, Part 2, Section 92).
However, there are provisions allowing state legislators to elect other officers to assist the
Assembly in conducting business. Generally, this includes the Majority Leader and Deputy
Majority Leader (serving as the chief strategists of the majority party, in charge of government

126

business and the legislative schedule); the Minority Leader(s) and Deputy Minority Leader (s)
(representing the interests of the minority party (or parties)); and the Party Whips (responsible
for maintaining decorum during sessions, and ensuring legislators’ support for party programs
and strategies) (Omoyele, 2011: 22 – 24).
As with the National Assembly, each state legislature “performs three basic functions—
law-making, oversight functions, and representation” (Omoyele, 2011: 14; Barkan, 2009).
Generally, Houses of Assembly are expected to “make laws with respect to any matter within
[their] legislative competence, correct any defects existing law, and expose corruption,
inefficiency, or waste in the administration of laws” (Omoyele, 2011: 15).

56

Second, state legislatures are given the authority to check the other branches of
government, and they provide this oversight in several ways. Houses of Assembly control and
monitor public funds, such that “No money shall be withdrawn from any public fund of the State,
unless the issue of those moneys has been authorized by a Law of the House of Assembly of a
State” (1999 Nigerian Constitution, Chapter 5, Part 2, Section 120). Moreover, the Executive is
charged with preparing an annual budget, “estimate[ing] revenues and expenditures of the State
of the following financial year” (1999 Nigerian Constitution, Chapter 5, Part 2, Section 121).
This budget is presented to the House of Assembly to investigate and approve. State legislatures
also provide oversight through their power to confirm and approve Executive appointments. This
includes “screening and approval of people nominated by the Governor of the State for public
offices, such as Commissioners, Chief Judge of the State, and Chairmen and members of
statutory bodies or Commissions” (Omoyele, 2011: 16). Conversely, Houses of Assembly also
56

Chapter Two: Devolution of Power and Purview to State and Local Governments outlines the
evolution of sub-national authority, and describes the current areas under state government
jurisdiction to make laws.
127

have the power to conduct investigations of public officials and public complaints. And with
their power of impeachment, the Governor, Deputy Governor, and Chief Judge of each State are
also subject to investigation and punishment of any misconduct.
Last, in the execution of representation, members of House of Assembly use deliberations
and the passage of bills into laws to give expression to the interests of their constituents and the
people of the State at large (Omoyele, 2011: 15). Barkan (2009: 7) expands representation to
include constituency service. This can include regular visits by legislators to their districts to
assist constituents with individual needs. Legislators can also sponsor development projects to
provide public services to their district, “including roads, water supply systems, schools, health
clinics, and meeting halls.”
Returning to the discussion of representation in my six case-states, I profile each of the
legislatures in the six cases. Beginning from Nigeria’s return to democratic governance in 1999
(Fourth Republic), I provide background for each House of Assembly. In particular, I discuss
the nature of legislative membership, the committee system, and provide examples areas in
which legislators have effectively used their formal powers to make laws. By providing this
background information, I explore how state legislatures in the Fourth Republic operate in a reestablished federal system and use the constitutional authority they have been granted.

Legislatures in South-West States
Lagos State’s 20 Local Government Areas (LGAs) are divided into 40 constituencies and
represented by 40 legislators in the House of Assembly. Lagos’ first legislative assembly in the
Fourth Republic convened from 1999 to 2003, in which 37 members belonged to the Alliance for
Democracy Political Party (93%). The remaining three members (7%) were part of the All

128

People’s Party. This changed slightly in the second legislative assembly (2003 – 2007), where
39 out of 40 members (98%) belonged to Alliance for Democracy, while one member was a part
of the People’s Democratic Party (PDP). In 2006, after Alliance for Democracy merged with the
Justice Party and the Advance Congress of Democrats to form the Action Congress of Nigeria
(ACN), Lagos’ third legislative assembly was dominated by the ACN. While 37 members (93%)
belonged to ACN, PDP was able to increase their presence to three members (7%) (Omoyele,
2011).
For the purpose of facilitating legislative business, the Lagos State House of Assembly is
made up of 34 committees. The Special House Committees “are established by the Rules of the
[Lagos State] House to assist in the smooth running of legislative operations.” These seven
committees include:
•
•
•
•
•
•
•

Selection
Business, Rules, and Standing Orders
Ethics, Protocols, and Privileges
House Services
State Public Accounts
Local Government Public Accounts
Human Rights and Public Petitions

The House Standing Committees are the “means by which the legislature effectively and
efficiently carries out its statutory oversight functions of supervising and scrutinizing the
activities of the Executive” (Omoyele, 2011: 74). Thus, there is a Standing Committee
corresponding with each executive ministry/agency (27). These committees monitor the
activities of the agencies and conduct investigations as necessary. Standing committees also
analyze bills dealing with issues under the purview of that particular committee. These include:
• Agriculture and Co-Operatives
• Central Business Districts
• Chieftaincy and Boundary Adjustment
Matters

•
•
•
•
129

Commerce and Industry
Economic Planning and Budget
Education
The Environment

• Establishment, Training, Pensions,
And Public Service
• Finance
• Health
• Home Affairs and Culture
• Housing
• Information, Strategy, and Security
• Infrastructure
• Judiciary
• Lands
• Local Government Administration
• Physical Planning and Urban
Development

•
•
•
•
•
•
•
•
•

Rural Development
Science and Technology
Special Duties
Tourism and Inter-Governmental
Relations
Transportation
Waterfront Infrastructure
Development
Women Affairs and Poverty
Alleviation
Works
Youth, Sports, and Social
Development

In terms of legislation, the Lagos State House of Assembly produced 149 different pieces.
Breaking this number down, the first legislative assembly (after Nigeria’s return to democracy,
1999 – 2003) passed 30 laws. This legislation included measures to modify the local
government system, reforms to the state electoral commission, and various allocations for
education and judiciary spending. The second legislative assembly (2003 – 2007) more than
doubled its productivity, passing 76 laws. These laws focused on establishing commissions to
ensure workers’ safety, the welfare of disabled persons, and HIV/AIDS management. The
second assembly also legislated on issues of road construction and safety, public transportation,
reforming the penal code, and public education. The third legislative assembly (2007 – 2011)
produced 43 laws by December 2010. Various pieces of legislation created state microfinance
institutions, reformed the state judiciary, established a casino/gaming regulatory body, and
established/mended traffic laws. Of particular interest is Lagos State’s 2009 law “imposing a tax
on goods and services consumed in hotels, restaurants and event centers within the territory of

130

57

Lagos State.”

Turning to the Oyo State House of Assembly, 33 LGAs are represented by 32 elected
officials in the state legislature. In Oyo’s first legislative assembly, post-return to democratic
governance (1999 – 2003), 30 out of 32 members (94%) belonged to the Alliance for Democracy
Party. The remaining two members (6%) were a part of the PDP. This drastically changed in the
second legislative assembly (2003 – 2007): 31 legislators (97%) were members of the People’s
Democratic Party and only one (3%) now belonged to Alliance for Democracy. Finally, in the
third legislative assembly (2007 – 2011), PDP maintained the majority, though not as big as the
previous assembly. 23 out of 32 legislators (72%) were members of PDP, eight were members
of the Action Congress of Nigeria, and one was a member of the All Nigeria Peoples Party
(Okoosi-Simbine, 2007).
The Oyo State Legislature is made up of 27 Standing Committees:

57

Information about the Lagos State House of Assembly’s activities was obtained via archival
research and data collection at the Assembly’s library and Clerk’s Office in Ikeja, Lagos (April –
June 2011). Due to the nature of the data, information is only available for legislation produced
between June 1999 and December 2010.
131

•
•
•
•
•
•
•
•
•
•
•
•

Accounts and Appropriation
Agriculture
Commerce and Cooperatives
Culture and Tourism
Education
Employment and Labour
Environment
Establishment
Foreign Relations
Fund Allocation
Health
Human Rights

•
•
•
•
•
•
•
•
•
•
•
•

Information and Orientation
Investment Monitoring Strategy
Land and Survey
Local Government
Science and Technology
Security and Strategy
Solid Minerals
Special Duty
Water Resources
Women Affairs
Works and Transport
Youth and Sports

Between 1999 and 2005, the Oyo State House of Assembly passed 70 bills into law. The
first legislative assembly (1999 – 2003) passed 40 pieces of legislation. These laws included
reforms to primary education, amending pensions for Oyo State employees, and expanding
electricity to rural areas of the state. Other measures established a State Housing Corporation,
the Ministry of Environment and Water Resources, the State Board for Technical and Vocational
Education,, and Urban and Regional Planning Board. From 2003 to 2005, Oyo State legislators
passed 30 laws. These bills involved the creation of a State Agency for Youth Development, the
state Road Maintenance Agency, the state’s Independent Electoral Commission (SIEC), and the
Ministry of Information and Culture. Legislators also made reforms to the State College of
58

Nursing and Midwifery and the Oyo State College of Education (Okoosi-Simbine, 2007).

The Ekiti State House of Assembly is made up of 26 members, representing the state’s 16
LGAs divided into 26 constituencies. In Ekiti’s first legislative assembly, post return to
democracy, 22 out of 26 members 85% belonged to the Alliance for Democracy Party (AD).

58

In addition to secondary sources (cited in text), information about the Oyo State House of
Assembly’s activities was obtained via archival research and data collection at the Assembly’s
library and Clerk’s Office in Ibadan, Oyo (June 2011). Due to the nature of the data, information
is only available for legislation produced between June 1999 and December 2005.
132

Three legislators were a part of the People’s Democratic Party (PDP, 11%), while one member
belonged to the All People’s Party (4%). The second legislative assembly experienced a
dramatic shift in membership, as PDP became the dominant party. In fact, 24 out of 26 members
of this assembly belonged to PDP (92%). Now, one legislator was elected from the AD Party
(4%), while one member was a part of the National Conscience Party (4%). In the beginning of
Ekiti’s third legislative assembly, 14 members belonged to PDP (54%), while 12 legislators were
a part of the Action Congress of Nigeria (ACN, 46%). However, during the legislative session,
Honourable Adebayo Morakinyo switched from PDP to ACN, creating a 50-50 split between the
two parties in the House of Assembly.
The Ekiti State legislators also divide themselves among 27 committees in order to conduct
business. They are as followed:
• Selection
• Fund Management and SelfAccounting Law
• HIV/AIDS
• State Independent Electoral
Commission and Other Agencies
• Environment, Lands, and Housing
• Works and Transport
• Public Petitions
• Due Process, Millennium
Development Goals (MDGs), and
Multilaterl Relations
• Governor’s Office and Government
House
• Finance and Appropriation
• Education, Science, and Technology
• Rule and Business
• Public Accounts
• Local Government and Chieftaincy
Affiairs

• Health and Human Services
• Ethics, Privileges, and Constitutional
Review
• Public Utilities and Infrastructures
• Women Affairs and Social
Development and Gender Equality
• House Service
• Agriculture and Rural Development
• Information and Civic Orientation
• Physical, Urban, and Regional
Planning
• Public Service Matters, Employment,
and Human Capital Development
• Judiciary and Legal Matters
• Commerce, Industry, and
Cooperatives
• Anti-Corruption
• Culture, Arts, and Tourism

Between 1999 and 2010, the Ekiti State House of Assembly produced and passed 90 pieces
133

of legislation. The first legislative assembly (1999 – 2003) passed 29 bills, including reforms to
the state electoral commission and local government administration. In addition to appropriation
bills, laws were passed creating a poverty alleviation agency, a waste management authority,
Council of Traditional Rulers, and a road works agency. During this period, legislators also
passed bills to protect the rights of widows and prohibit female circumcision and genital
mutilation. The second legislative assembly (2003 – 2007) passed 23 laws, including those to
build, manage, and coordinate roads (especially in rural areas), amend state electoral law, and
reform local government administration law. Legislation was also passed to establish a state
education board, abolish secret religious cults, and provide funding for environmental health and
sanitation matters. By the end of 2010, the third legislative assembly (2007 – 2011) had already
produced 38 laws. Different bills regulated community watch groups, the University of AdoEkiti, as well as state micro-credit. Legislators also used laws to establish the Job Creation and
Employment Agency, the Community and Social Development Agency, and Emergency
Management Agency. Moreover, bills passed during this time provided for the House of
Assembly’s fiscal autonomy, allocated funds to protect the rights of persons with disabilities, and
59

also manage the creation and distribution of birth and death certificates.

Overall, laws passed in these South-West legislatures were for the purpose of regulating
and funding government activities. However, the bills produced by these Houses of Assembly
also aimed to build and manage various public services, while protecting the rights of typically
vulnerable groups (e.g. children, widows, disabled persons). Table 4.12 compares the number of
laws passed in Lagos, Oyo, and Ekiti States between 1999 and 2010. Data for Oyo State are
59

Information about the Ekiti State House of Assembly’s activities was obtained via archival
research and data collection at the Assembly’s library and Clerk’s Office in Ado, Ekiti (May
2011). Due to the nature of the data, information is only available for legislation produced
between June 1999 and December 2010.
134

only available between 1999 and 2005, so we cannot fully compare across all three legislative
assemblies. However, if we consider the first two assemblies (June 1999 – May 2003), Oyo
State Legislature passed the most bills (40). During the same time, Lagos passed 30 bills into
law, while Ekiti State passed 29. In the second assembly, legislators in Lagos passed 76 laws;
Ekiti State legislature was only able to pass 23 bills (June 2003 – May 2007). Between June
2003 and December 2005, Oyo State had already passed 30 laws. If we focus on Lagos and Ekiti
States (where there is data until December 2010), Lagos had a slightly more productive third
assembly, passing 43 bills compared to 38 in Ekiti State. But due to increased productivity in
earlier assemblies, Lagos is overall the more productive: in this time (1999 – 2010), the Lagos
State House passed 149 bills into law, while Ekiti State legislators passed 90.

60

Legislatures in North-West States

Turning to the case-states in the North-West Region, I begin with the Sokoto State House
of Assembly. Sokoto’s 23 LGAs form the 30 constituencies represented in the state legislature.
In the third legislative assembly after Nigeria’s return to democracy (2007 – 2011), the PDP
dominated the House’s membership: 22 out of 30 members (73%) belonged to this party. The
other 8 legislators (27%) were part of the Democratic People’s Party.
Sokoto’s House of Assembly is comprised of 30 Standing Committees, which execute
legislative duties. They include the following:

60

Unfortunately, in the course of conducting archival research in the North-West cases, I was
only able to obtain information about legislative membership and productivity for the third
legislative assembly (2007 – 2011) in Sokoto, Kano, and Katsina Houses of Assembly.
135

Selection
Rules and Business
Public Accounts
House Services/Ethics and Privileges
Public Petitions
Finance and Appropriation
Education
Health
Works and Transport
Agriculture
Information
Commerce, Industry, and Tourism
Judiciary
Local Government and Community
Development
• Lands and Housing
• Science and Technology
•
•
•
•
•
•
•
•
•
•
•
•
•
•

•
•
•
•
•
•
•
•
•
•
•
•
•
•

Women Development
Public Service Matters
Religious and Chieftaincy Affairs
Rural Development
Urban Development
Emergency Relief and Disasters
Multilateral Agencies
Social Welfare, Sports, and Culture
Special Services
Youth Empowerment and Career
Development
Solid Minerals and Petroleum
Resources
Environment
Poverty Alleviation and Economic
Development
Water Resources

During its tenure, the third assembly of the Sokoto State Legislature passed 36 bills into law. In
their first session (June 2007 – May 2008), legislators passed six laws, including those to provide
pensions for state employees and restructuring the administration of local governments. The
Sokoto State House was able to pass 12 laws during its second session (June 2008 – May 2009),
making it the most productive. In addition to establishing a College of Agriculture and Animal
Science and reorganizing the state judiciary, the legislature also passed the Sokoto State Board of
Internal Revenue Law. As previously discussed, this law formalized the state revenue service’s
autonomy and authority to collect taxes, pursuing and punishing evaders to that end. During the
third session (June 2009 – May 2010), the state legislature produced eight bills, establishing
agencies such as the Sokoto State Agency for the Control of AIDS, the Sokoto State Emergency
Management Agency, and the Primary Health Care Development Agency. Finally, in the third
assembly’s last session (June 2010 – May 2011), the Sokoto House passed 10 laws. These
included a bill to establish a State College of Nursing Sciences and various bills suggesting

136

amendments to the 1999 Federal Constitution.

61

Considering the Kano State House of Assembly, the state’s 44 LGAs are divided and
represented by 40 legislators. The membership of the third legislative assembly, convened
between 2007 and 2011, was made up of members representing the All Nigeria People’s Party
(ANPP) and PDP. 25 out of 40 members (63%) belonged to ANPP, while the remaining 15
legislators (37%) were part of PDP.
In order to perform their legislative and oversight functions, Kano State legislators divide
themselves into 32 standing committees:

61

Information about the Sokoto State House of Assembly’s activities was obtained via archival
research and data collection at the Clerk’s Office in Sokoto (City), Sokoto (June 2011). Due to
the nature of the data, information is only available for legislation produced between June 2007
and May 2011.
137

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Pension Funds
Agriculture and Natural Resources
Judiciary
Higher Education
Monitoring
Environment
Guidance and Counseling
Religious Affairs
Information
Ethics and Privileges
Public Accounts
Public Service
Emergency Management
Appropriation
Rural and Community Development
Primary Education

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

Works and Transport
Land and Physical Planning
Housing
Science and Technology
House Services
Health
Finance and Economic Development
Hajj
Security
Public Petitions
Commerce and Industry
Youth and Sport
Women Affairs
Water Resources
Local Government and Chieftaincy
Education

During its tenure, Kano’s third assembly passed 36 pieces of legislation. In the first
session (June 2007 – May 2008), legislators passed 13 bills, including various appropriation laws
and reforms to the local government administration. Laws that were produced also enacted a
pension fund for state employees and created an Emergency Relief and Rehabilitation Agency.
During the second session (June 2008 – May 2009), nine bills were passed. These bills focused
on funding various local government efforts, including those to improve the power sector.
Moreover, in this session legislators passed a measure to create the Kano State Drugs and
Medical Consumables Supply Agency. In the third session (June 2009 – May 2010), legislators
passed eight laws, including the Kano State Revenue Administration Law. This law aimed to
strengthen and professionalize the Kano State Internal Revenue Board, while optimizing the
collection of taxes. Finally, in the fourth session (June 2010 – May 2011), the third assembly
passed six laws, including proposals for amendments to the federal constitution, executive

138

appropriation bills, and a bill to establish the Sa’adatu Timi College of Education and Law.

62

Last, in the Katsina State House of Assembly, 34 legislators represent the 34 Local
Government Areas. During the third legislative assembly, the People’s Democratic Party (PDP)
dominated the Katsina State Legislature, with 28 out of 34 members (82%). Three out of 34
legislators (9%) were members of the Congress for Progressive Change Party, two members
(6%) were a part of the All Nigeria People’s Party, while one legislator (3%) was a member of
the Action Congress of Nigeria.
Katsina legislators divide themselves between 30 Standing Committees in order to perform
their legislative and operational duties:

62

Information about the Kano State House of Assembly’s activities was obtained via archival
research and data collection at the Clerk’s Office in Kano (City), Kano (March 2011, July 2011).
Due to the nature of the data, information is only available for legislation produced between June
2007 and May 2011.
139

Religious Affairs
Agriculture
Local Governments
Commerce and Industries
Water Resources
Education
Information
Justice and Judiciary
Lands and Environment
Administration
Millennium Development Goals,
Community, and Social
Development
• Security and Intelligence
• Inter-Parliamentary
• Multilateral and Donor Agencies

Roads Maintenance
Habitat and Emergency Relief
Public Petitions
Public Accounts
Anti-Corruption
Finance
Appropriation
House Services
Business, Rules, Ethics, and Privileges
Science and Technical Education
Higher Education
Primary Health
Agricultural and Rural Development
Authority
• Farmer’s Supply
• Rural Electricity

•
•
•
•
•
•
•
•
•
•
•

•
•
•
•
•
•
•
•
•
•
•
•
•

Between 2007 and 2010, Katsina’s third legislative assembly passed 30 bills into law. In
their first session (June 2007 – June 2008), legislators passed eight laws. These included
measures to reform public employee pension plan and public education (e.g. creation of the
Secondary School Education Board). During the second session (July 2008 – June 2009),
legislators in the Katsina State House passed nine laws, including the creation of the Roads
Maintenance Agency, a Sanitation Monitoring Committee, and a Community and Social
Development Agency. In the third session (July 2009 – June 2010), legislators passed an
additional nine bills. These laws focused on reforms to the state penal code, environmental
protection standards, and the administration of local governments. During this period, the
Katsina State House also established a state micro-finance institution, a Religious Education and
Preaching Board, and a state Agency for the Control of AIDS. Between July 2010 and
December 2010 (the beginning of the fourth session), legislators passed six bills into law. These
concentrated on the creation of a Child Rights Law, amendments to the state electoral laws, and

140

63

establishing an Urban and Regional Planning Board.

Like in the South-West, the North-West legislatures passed bills focused on government
appropriations (e.g. executive branch, judiciary, local government). However, several laws
created new institutions for the purpose of improving public service provision (emergency relief,
education, roads, urban planning, drug administration/health care). In terms of productivity,
Table 4.13 compares the number of bills passed by the Sokoto, Kano, and Katsina State
Legislatures during the tenure of the third legislative assembly. Since data are only available for
the first six months of Katsina’s fourth session, it is difficult to compare the three states’ overall
productivity. However, if we look at the first three sessions (June 2007 – May 2010), we see the
Kano State House passed the most bills (30). In this same period, Sokoto and Katsina State
Legislatures each passed 26 laws. If we focus on Sokoto and Kano (where there is data for all
four sessions), we observe Sokoto had a more productive fourth session (passing 10 bills,
compared to Kano’s six). As a result, Sokoto and Kano tied in productivity: in this time (2007 –
2011), both state legislatures passed 36 bills into law.
Having profiled the legislatures and the legislative productivity in the six cases, I return to
my initial hypothesis:
Hypothesis: At higher levels of tax generation capacity, Nigerian legislators will be more likely
to represent their constituents.
In order to text this hypothesis, I rely on data from interviews conducted with state
legislators in the six case-states during fieldwork in Nigeria (October 2010 – July 2011). In the
next section, I describe sampling procedures, provide summary statistics about the sample of
63

Information about the Katsina State House of Assembly’s activities was obtained via archival
research and data collection at the Clerk’s Office in Katsina (City), Kano (June 2011). Due to the
nature of the data, information is only available for legislation produced between June 2007 and
October 2010.
141

legislators, and outline the research design.

Research Design: Revenue and Elite Perceptions of Representation
Table 4.14 presents a summary of the sampling procedures. The total population of
legislators in the six selected cases is 202. Ideally, I would have included all 202 legislators in
the sample; however, given limited research resources, the maximum feasible sample size was n
64

= 101.

These 101 legislators resulted from including 50% of each state’s legislature in the

sample. This broke down to: 20 legislators in Lagos, 16 legislators in Oyo, 13 legislators in
Ekiti, 15 legislators in Sokoto, 20 legislators in Kano, and 17 legislators in Katsina. By
maintaining proportionality, sampling legislators in proportion to the total size of each
legislature, findings from they analysis can be generalized to the larger population of legislators
in the six cases. Moreover, with around 100 legislators, I am able to use percentages to
summarize and characterize the sample.
In addition, I made the selection of legislators as random as possible, whereby each
legislator had an equal chance of being included in the sample. This way, I could avoid
systematic bias between elites I interviewed and those I did not. For example, since I conducted
interviews during an election season, it was possible that legislators involved in an intense or
competitive campaign were absent from the House of Assembly, instead spending time in their
constituency. Thus, when I arrived at each state legislature, I first obtained a list of legislative
members from the Office of the Permanent Secretary/Clerk of the House. Using that list, I
assigned each legislator a unique number. With a table of random numbers generated that day, I
64

As the principal investigator, I conducted all interviews myself. Nigeria held state elections
on April 23, 2011. As a result, my interviews were conducted in the midst of an election season
where legislators were spending less time at the State House of Assembly, instead, campaigning
in their local constituency.
142

serially selected two-digit, consecutive random numbers from the list.

65

If the two-digit number

matched one of the legislators’ numbers, I added that legislator to my sample. I continued until I
selected enough legislators to fulfill the sample size in the particular state in question. For
example, in Lagos State, there were 40 legislators; therefore, the sample size was 20. When I
obtained the list of legislators from the Lagos State House Clerk, I assigned each legislator a
unique number between one and 40. Using a table of random numbers, I matched the two-digit
random numbers from the list to any corresponding legislators’ numbers. I continued until 20
legislators were selected for my sample. I utilized these same procedures in the other five
legislatures.
Tansey (2007) identifies elite interviews as a useful tool for establishing “attitudes, values,
and beliefs...[and for making] inferences about a larger population’s characteristics and
decisions.” With that said, scholars also acknowledge that legislators can be “reluctant
respondents.” This literature provides several tools for ensuring validity and reliability of elite
responses (Dean and Whyte, 1970; Aberbach, Chesney, and Rockman, 1975; Berry, 2002; Leech
and Goldstein, 2002; George and Bennett, 2005). When drafting questions for legislative
interviews, I used open-ended/grand tour questions, which have been identified as “potentially
the most valuable type of elite interviewing in more politically unstable environments” (Berry,
2002: 679). Open-ended questions afforded the best opportunity to let respondents identify the
relevant and important issues. In addition to original, open-ended questions, I replicated
questions from the African Legislatures Project and the Afrobarometer survey instruments.
Doing this allowed me to make use of questions that have already been tested in the field and
found to be successful in gathering information from potentially unreliable respondents.
65

Tables of random numbers were generated using the following website:
http://stattrek.com/statistics/random-number-generator.aspx .
143

Furthermore, by using more structured and closed-response questions, I was able to somewhat
dictate the direction of the interview. This also ensured that legislative responses could be
66

systematically compared to each other for analysis.

With the survey, I sought data about legislators’ demography (e.g. age, ethnicity,
education, gender), their activities while they were members of their House of Assembly,
legislators’ opinions about their relationship to constituents, and their law-making/decisionmaking process.
The full survey/interview instrument was submitted to and approved by Michigan State
University’s Institutional Review Board. Before conducting each legislative interview, I
provided respondents with a consent form, which briefly outlined the research project and the
purpose of the interview. I assured each respondent that his or her responses would be kept
strictly confidential. I did not write the legislators’ names on the questionnaires, and I assured
them that their names would never be linked with their responses. I also assured respondents
that they were free to refuse participation, and once we started, he or she was free to answer each
question openly. Finally, I provided legislators with my contact information, as well as that for
the chair of my dissertation committee (Dr. Michael Bratton) and the Michigan State University
IRB. They would be free to express any questions or concerns, and inquire a more complete list
of their rights as a research participant.

67

After reading and discussing the consent form and any

questions or concerns, respondents verbally indicated their agreement to conduct the interview.

66
67

See Appendix for the list of complete survey questions.
See Appendix for the full Consent Form.
144

Each legislative interview took between 30 and 35 minutes to complete.

68

Repeated calls

ensured random selection of legislators.

Summarizing the Legislators: Describing Demographic Factors
Beginning with a description of the sample’s various demographic attributes provides
useful descriptive information. Legislators in the six-state sample range from 33 from 63 years
of age (Table 4.15). While the mean legislator is 47 years old, legislators in Lagos State are, on
average, the oldest at 49 years old. On the other hand, Sokoto State legislators are, on average,
the youngest at 44 years old.
In terms of gender, 96% of the legislators are male, while 4% are female (Table 4.16). In
Ekiti, Kano, Katsina, and Sokoto Houses of Assembly, 100% of legislators are male. However,
in Lagos State, three out of 20 legislators are female; in Oyo State, one out of 16 legislators are
female. Even though women, on average make up between 48% and 50% of the population in
these six states, we can see that females are severly under-represented in these state
69

legislatures.

Considering ethnicity, 48% of legislators are Yoruba, while 51% are from the HausaFulani, ethnic group (Table 4.17). All legislators in Kano, Katsina, and Sokoto (North-West) are
Hausa-Fulani. Similarly, all of the legislators included from the Ekiti and Oyo State Houses are
Yoruba. In the Lagos State Legislature, 19 out of 20 legislators are Yoruba, while one legislator
is from the Ogu (Egun) ethnic group. We can see that these state legislatures in both regions are

68

Interviews were conducted with members of the third legislative assembly in each of the State
Houses (2007 – 2011 term).
69
This statistic is from a 2006 report from the Nigerian National Bureau of Statistics, entitled
Women and Men in Nigeria: Facts and Figures.
145

relatively ethnically homogenous.
Referring to Table 4.18, we see 66% of legislators affiliate with the Muslim religion, while
33% characterize themselves as Christians. Officials from Kano, Katsina, and Sokoto States
(North-West) are all Muslim. 13 out of 13 Ekiti State legislators are Christian. However, in the
Lagos State Assembly, 11 out of 20 legislators identify as Christians, while eight out of 20 are
Muslim. Likewise, nine out of 16 Oyo State legislators are of the Christian faith, and the
remaining seven identify Islam as their religious affiliation.
37% of legislators have completed a post-graduate degree (e.g. Master of Business
Administration, Master of Public Administration; Table 4.19). For 31% of legislators, the
highest level of education completed is a bachelor’s degree (or equivalent). 17% of legislators
have completed post-secondary school training and/or an advanced diploma. The remaining
15% have completed secondary school. In the South-West, 13 out of 14 Ekiti State legislators
(93%) have completed at least a bachelor’s degree. In Lagos State, 18 out of 20 (90%) have
attained a university degree (or more), while in Oyo State, this number is 14 out of 16 (88%). In
the North-West Region, 11 out of 15 state legislators (73%) have completed a bachelor’s degree
or more. In Kano State, 9 out of the 20 officials (45%) at least completed university, while in
Katsina State, 5 out of 17 (29%) attained a bachelor’s degree or higher.
Moving to legislators’ prior occupational experiences, we see the majority of officials were
business people (26%), civil servants (23%), or local councilors (21%) before joining the State
Houses of Assembly (Table 4.20). Other occupations include: commercial farmers (7%); other
professionals (e.g. doctor, engineer, 6%); teachers/head maters (5%); supervisors/mid-level
managers (4%); bankers (3%); university lecturers (2%); lawyers, accountants, general
managers, and journalists (1% each).

146

Finally, turning to the number of years in legislative office, officials have been members of
their respective State Houses for a minimum of two years and a maximum of 12 (Table 4.21).
The average legislator has held his or her position for five years. On average, Lagos State
legislators have held office for the longest, at a mean of 6 years. On the other hand, legislators in
Oyo, Ekiti, and Katsina States have, on average, spent the least time in the legislature, holding
only one four-year term.

Dependent Variable: Elite Perceptions of Representation
As previously stated, I hypothesize legislators in states that generate more income from
non-oil taxes will be more representative of their constituents. To also reiterate, representation is
the process whereby elected officials gauge, deliberate upon, and incorporate citizen interests in
the execution of decisions and policy (Pitkin 1967, 1969; Huber and Powell, 1994; Aldrich,
1995; Stokes, 1999). In the performance of their representative duties, politicians must prioritize
their constituents’ preferences. Thus, when operationalizing elite perceptions of representation, I
utilize seven indicators. These variables separately measure the extent to which an individual
legislator believes his or her representative function and constituents’ opinions and interests take
precedence over other considerations.
The first indicator measures a legislator’s evaluation of how effective their state House of
Assembly does with representing constituents. Specifically, the item asks, “how well or badly
would you say the State Assembly is doing its job of representing constituents?” Respondents
can answer with the following: “very bad, bad, good, or very good.” A second indicator asks

147

70

this question, but specifically about “representing women’s interests.”

The third variable I use to measure representation is legislators’ self-report of the
percentage of time he or she engages in constituency work (out of 100% of their time, and in
comparison to time spent doing plenary work, committee work, party work, and—if they have
one—their other job).

71

A fourth indicator is a dummy variable measuring whether a legislator believes
“representing their constituents’ views in parliament” and performing constituency services (e.g.
bringing development to his/her constituency, assisting constituents with their personal
problems, and soliciting funds for his/her constituency) are “the most important part of being a
72

member of the State Assembly.”

A fifth indicator is also a dummy variable that captures “in general, when [he/she] takes a
position about an issue in the State Assembly” how much of a priority is a legislator’s

70

“I want you to rate the effectiveness of the State Assembly like a score card. For each of the
following areas, how well or badly would you say the State Assembly is doing its job?”
Representing constituents? Representing women’s interests? 1=Very bad, 2=Bad, 3=Good,
4=Very good.
71

“What percentage of your time is devoted to each of the following:” Plenary work?
Committee work? Constituency work? Party work (outside your constituency)? Your other job
(including ministerial work)? PERCENTAGES SHOULD TOTAL TO 100%.
72

“In your opinion, which of these following jobs is the most important part of being member
of the State Assembly?” 0=Debating bills and passing laws, 0=Making public policy by writing
laws, 0=Overseeing the executive, 1=Representing constituents’ views in parliament, 1=Bringing
development to your constituency, 1=Assisting constituents with their personal problems,
1=Soliciting funds for your constituency.
25 “

In general, when you take a position about an issue in State Assembly, which of the following
is most important? 0=The views of your party leader, 0=The views of your party, 0=The national
interest, 0=Your knowledge about the issue, 0=Your personal convictions, 1= The views of your
constituents.
148

73

constituents’ views? This is in comparison to other considerations.

The sixth variable asks

specifically about legislators’ opinions of “what MPs should (generally) do if there is a conflict
74

between their party’s position and the views of their constituents.”

The last measure of elite representation asks legislators what should happen if “a member
of the State Assembly ignores what his/her constituents have to say.” Should he/she remain in
office or lose their seat in the next election?

75

These seven indicators each gauge the importance a given legislator places on executing
representative functions and prioritizing his/her constituents’ preferences.

Independent Variable: State Tax Dependence
The key explanatory variable in this analysis is state tax generation: that is, the percentage
of total state revenue derived from taxes on citizens. To capture this, I include a variable for
taxes as a percentage of state income in 2009 (this is the most recent data available). I also use
average tax dependence variable, which represents the mean value of taxes as a portion of state
government revenue between 1999 and 2009. I utilize both of these measures in an attempt to
capture both proximate and longer-term effects of tax reliance on shaping legislators’ incentives
and choices. I expect a positive relationship between elite perceptions of representation and state

74

“What should MPs do if there is a conflict between their political party’s position and the
views of their constituents?” 0=Support the party position 0=Abstain from voting, 1=Oppose the
party position.
75

“If an Member of the State Assembly ignores what his/her constituents have to say, do you
think that he/she would still remain in office or lose the next election?” 0=Remain in office,
definitely, 0=Remain in office, probably, 1=Lose their seat, probably, 1=Lose their seat,
definitely.
149

reliance on tax revenue. It follows that legislators in states that generate more income from nonoil taxes on citizens (e.g. Lagos and Sokoto States) will be more likely to identify representation
and constituency service as their job priorities. As Timmons (2005: 531) argues, in these taxreliant states, officials face “strong incentives to provide benefits to maintain the source of
revenue [that taxes on citizens provide].” These benefits manifest in how legislators establish
their representative duties as precedence. This also includes placing constituents’ interests ahead
of other considerations when making policy decisions (Bates and Lien, 1985; Levi, 1988).
Based on this reasoning (returning to the dependent variables), I expect legislators in tax-reliant
states will be more likely to:
Ha: rate their State House of Assembly as being highly effective in representing constituents.
Hb: rate their State House of Assembly as being highly effective in representing women’s
interests.
Hc: indicate a greater percentage of their time is spent on constituency work.
Hd: identify representation and constituency services are the most important part of being a
member of the State Assembly.
He: report that the views of their constituents are the most important consideration when taking a
position about an issue.
Hf: oppose their party position if there is a conflict with the views of their constituents.
Hg: believe that if a Member of the House of Assembly ignores their constituents’ views, he/she
should lose their seat in the next election.

150

Analysis and Results: Tax Reliance and Elite Execution of Representation
Preliminary Analysis of Correlation
Given the small sample size, I rely on a preliminary analysis of scatterplots and correlation
to test the aforementioned hypotheses. First, I separate the 101 legislators by state. In each of
the six clusters of legislators, I calculate the average value for each indicator of representation
(seven). I then plot the average value of the dependent variables against each state’s average tax
reliance between 1999 and 2009 (explanatory variable 1) and the percentage of state income
derived from taxes in 2009 (explanatory variable 2). According to my hypotheses, legislators
from states that generate higher rates of income from taxes should express attitudes that prioritize
representation and constituency services. In other words, I expect a positive relationship
between elite perceptions of representation and state tax reliance. Figures 4.5 through 4.18
present the results.
Generally, there is a weak relationship between state tax capacity and how legislators rate
the effectiveness of their state legislature in representing constituents (Figures 4.5 and 4.6).
Though the line of best fit moves in a positive direction (as hypothesized), the six data points are
scattered widely. Moreover, a Pearson’s coefficient reveals a weak correlation between elite
perceptions of how well their House of Assembly represents constituents and average tax
generation (r = 0.43). The correlation between ratings of the state legislature and state tax
reliance in 2009 is even weaker (r = 0.26). Average tax generation accounts for 18% of variation
2

in elite perceptions of how well their state legislature represents constituents (r = 0.18). State
2

tax capacity in 2009 only accounts for 7% of the variation (r = 0.07). Thus, there is a weak
positive relationship between reliance of tax income and elite views of how effective their state
legislature represents constituents.
151

On the other hand, elite views of how effective the state legislature represents women has a
strong and positive relationship with states’ capacity to extract taxes (Figures 4.7 and 4.8). A
Pearson’s coefficient confirms the strength of this correlation: legislators’ views of how well the
House of Assembly represents women is positively related to average tax generation (r = 0.87).
These perceptions positively related to state tax reliance in 2009 (r = 0.94). Moreover, a state’s
average tax generation (1999 – 2009) explains for 76% of variation in elite perceptions of how
2

effective the state legislature represents women (r = 0.76). The portion of state income
2

generated from taxes in 2009 also accounts for 88% of the variation (r = 0.88) in legislators’
ratings. This strong, positive relationship supports the hypothesis that legislators in tax-reliant
states are more likely to rate their State House of Assembly as being effective in representing
women’s interests.
The remaining five indicators of elite representation reveal a similar robust and positive
relationship with state tax capacity. According to Figures 4.9 and 4.10, there is a positive
relationship between states that generate more income from taxes between 1999 and 2009 and
legislators who report spending a greater percentage of their time engaged in constituency work
2

(r = 0.75, r = 0.56). There is also a positive relationship whereby elites from states that report a
greater capacity to extract taxes in 2009 are also more likely to claim that they spend more time
2

doing directing serving their constituents (r = 0.97, r = 0.94).
There is a strong, positive link between states that generate more of their average income
(1999 – 2009) from taxes and state legislators that are more likely to identify representation and
2

constituency work as their most important job (r = 0.78, r = 0.61) (Figure 4.11). Similarly,
legislators from a state that was more tax reliant in 2009 tend to report representation as a
152

2

legislator’s primary concern (Figure 4.12; r = 0.96, r = 0.92). Legislators from states with a
high capacity to generate tax income (1999 – 2009) also report that their constituents’ views are
2

the most important consideration when making decisions (Figure 4.13; r = 0.73, r = 0.53). This
positive relationship is even stronger when considering state tax generation is 2009 (Figure 4.14;
2

r = 0.93, r = 0.86). State tax reliance in 2009 accounts for 86% of the variation in whether elites
identify constituents’ views as the most significant concern in their decision-making.
A legislator from a state that is (on average) a strong generator of tax income is more likely
to report that, given a conflict between the views of her political party and her constituents, she
2

would oppose her party (r = 0.64, r = 0.41) (Figure 4.15). There is also a positive relationship
between state tax reliance in 2009 and the likelihood legislators will choose their constituents’
2

views over their party’s position when making policy (Figure 4.16; r = 0.89, r = 0.79). Finally,
a state’s tax capacity (evidenced by the percentage of income between 1999 and 2009 derived
from taxes on citizens) is positively related to the likelihood that legislators will articulate that
2

officials who ignore their constituents should lose their seats (Figure 4.17; r = 0.80, r = 0.64).
Similarly, legislators from states that were strong tax generators in 2009 are also more likely to
2

report this opinion (Figure 4.18; r = 0.86, r = 0.74).
Taken together these scatterplots and analyses of correlation support my initial hypotheses.
There is a positive relationship between a state’s reliance on income from taxes on citizens and
various indicators of state legislators’ perceptions of representation. This preliminary analysis
lends support to the assertion that elected officials in states with a strong tax capacity are more
likely to report prioritizing representation and constituency services.
153

Given this initial finding of a positive relationship, I now employ a more rigorous test of
the hypotheses.
While the various elite representation variables are measured at the level of the individual
legislator, the two indicators of state tax reliance (explanatory variables) are measured at the
state (group) level. In order to perform a more robust analysis, I rely on hierarchical linear
modeling (HLM). Since the legislators included in my sample are clustered by state, and tax
dependence is measured at the state level, legislative attitudes and behaviors (within a state) are
likely correlated. This violates the OLS assumptions of the independence of observations and
homogenous distribution of variance. With this type of data, “OLS regression would not
produce correct standard errors; therefore, HLM needs to be used. It takes the issue of correlated
errors into consideration and provides more realistic and conservative statistical testing” (Kaz,
2005). Moreover, though HLM parameter estimates rely on maximum likelihood instead of least
squares, parameters are not “drastically different from OLS…However, standard errors would be
larger for HLM, as HLM considers sources of errors more rigorously than OLS” (Kaz, 2005;
Shamosh and Farach, 2007).
In order to isolate the effect of state tax dependence on legislators’ attitudes and behavior
toward representation, I include two other control variables. I use the number of years a
76

legislator has been a member of the State House of Assembly as a control variable.

I expect a

positive relationship between legislators’ total years in office and their willingness to prioritize
representation duties. As an official spends more time as a member of the legislature, he/she
gain more experience in interacting with constituents, gauging preferences, problems, or interests
76

“How many total years have you been a member of the State Assembly?” The variable is
given in years.
154

and acting upon them.
I also include a measure of party affiliation, specifically, whether the legislator in question
is a member of the majority party in his/her state legislature.

77

Particularly, with rating his/her

Assembly’s effectiveness in representing constituents, a legislator in the majority party could
perceive a more favorable evaluation than legislators in the opposition.
Taken together, this results in the following model specification:
Elite Representation = a + b1 (Average % Taxes) + b2 (2009 % Taxes) + b3 (Years in
Office) + b4 (Member of Majority Party).
“Elite Representation” is operationalized using the seven afore-mentioned dependent variables.
78

This results in seven separate analyses.

The average tax reliance variable is captured by the

mean value of taxes as a portion of state government revenue between 1999 and 2009. I also use
taxes as a percentage of state income in 2009 as a key independent variable.
For the six analyses, where the dependent variables are binary, dummy variables (Ha, Hb,
Hd, He, Hf, Hg), I utilize multi-level mixed effects logistic regression. For the remaining
analysis where the dependent variable is continuous (percentage of time legislator’s report doing
constituency work, Hc), I use multi-level mixed effects linear regression. I execute these
analyses with STATA 12.0 with clustered robust standard errors (clustered by the six casestates).

77

This is measured as a dummy variable, where a respondent will receive a value of 1 if he/she
is a member of the majority party in their state legislature and a 0 otherwise. In the case of Ekiti
State, where membership is evenly divided between two parties, all respondents receive a value
of 0 because neither party is a majority.
78
I restructure legislators’ ratings of the House of Assembly’s effectiveness in representing
constituents and representing women (originally ordinal variables) as a binary dummy variable
(very bad/bad = 0; very good/good = 1).
155

Multi-Level Analysis of Elite Representation and Revenue
Tables 4.22A and 4.22B present the results of the seven analyses investigating the
relationship between Nigerian legislators’ perceptions of representation and tax revenue
dependence. Since my sample size is small (n = 101), in order to avoid false precision, I only
present the coefficients to demonstrate the directionality of the relationships and whether the
results support my hypotheses. In terms of substantive interpretation of the HLM coefficients, I
will discuss the broad trends.
Evidence confirms the expectations outlined in this chapter: there is a positive relationship
between increased state tax reliance and legislators from those states prioritizing representation
and constituent preferences. Elected officials in states with a greater capacity to extract tax
income also express attitudes that prioritize representation of citizens and constituency service.
Furthermore, across multiple measures, as hypothesized, the positive relationship between
representation and higher rates of government tax generation endures.
Turning to the first column in Table 4.22A, we see both measures of state tax dependence
exert a positive influence on legislators’ evaluation of their State Assembly’s ability to represent
constituents. With that said, neither coefficient is statistically significant at the 90% level.
However, this lack of significance could be a result of minimal variation in this particular
variable. Most legislators believe their House of Assembly is doing a good/very good job of
representing constituents. In fact, in this sample, 93% of legislators (irrespective of state and
region) believe their State Assembly is doing a good or very good job of representing
constituents. However, with a larger sample, I would anticipate the positive relationship between
higher levels of state tax generation and elite execution of representation (measured with
legislative ratings of the legislature’s effectiveness) would meet the significance threshold.

156

Additionally, legislators from the majority party are more likely than minority members to report
that their state legislature is effective in its representation function (b = 0.1).
When considering legislators’ evaluations of the legislature’s success in representing
women, we observe a significant and positive relationship (column 2). State average tax
generation is positively related to officials’ evaluation of the state legislature (b = 0.045).
Furthermore, a state’s tax reliance in 2009 is also positively linked to legislators’ rating of the
State Assembly’s ability to represent women (b = 0.026). Elected officials in states that generate
higher portions of income from taxes on citizens are more likely to rate the state legislature as
effective in representing women.
As hypothesized, legislators in more tax-reliant states report spending a greater percentage
of their time conducting constituency service (column 3). As a state’s average tax income
increases, the likelihood legislators report spending more time doing constituency work also
increases (b = 0.105). Similarly, as in states that generated more of the 2009 revenue from taxes
on citizens, legislators are also indicate spending more time engaged in constituency work (b =
0.269).
Looking now at Table 4.22B, we see state tax generation (both measures) also exerts a
strong, positive influence on the likelihood a legislator will indicate “representing constituents’
views in parliament” or some form of constituency service as a legislator’s most important job
(column 4). Both the average tax income variable (b = 0.047) and a state’s 2009 income from
taxes (b = 0.043) exert a positive influence. Legislators in tax-reliant states are more likely than
their counterparts in other states to choose representation as their primary job.
When making decisions, legislators in tax-reliant states are more likely to designate the
views of their constituents are their most important consideration (column 5). As a state’s

157

average tax reliance (b = 0.043) and 2009 tax income generation (b = 0.095), the probability
legislators will prioritize constituents’ views in decision-making also increases.
In the event of a conflict between a legislator’s political party position and his/her
constituents’ views, officials in more tax dependent states are more likely to oppose their
political parties (column 6). Average tax generation (b = 0.076) and tax as a portion of state
revenue in 2009 (b = 0.086) share positive and significant relationships with the likelihood
legislators will choose to represent their citizens’ preferences, even if those preferences differ
from the party position.
Finally, legislators in states that generate more income from taxes are report that a member
of the State Assembly should lose their seat if he/she ignores what his/her constituents have to
say (column 7). Evidence suggests a positive relationship between a state’s average percentage
of tax generation (b = 0.092) and state tax reliance in 2009 (b = 0.093) and the liklihood
legislators in the state will express that officials who ignore their citizens should not be reelected.
Contrary to my initial hypothesis, results indicate that length of time in office has a
negative influence on legislators’ perceptions of representation. As an official spends more years
as member of the state legislature, he/she spends less time performing constituency services (b =
-0.387). This legislator is also less likely to identify representation and constituency work as a
legislator’s most important job (b = -0.054) or identify constituents’ views as his/her most
important consideration in decision-making (b = -0.043). As the amount of time a legislator
spends in office increases, he/she is less likely to oppose their party position even if it conflicts
with constituent views (b = -0.093). Last, more years in office correspond with the likelihood a
legislator will be less likely to believe a member of the Assembly should lose his/her seat if
he/she ignores constituents (b = -0.074). These findings could be the result of senior legislators

158

choosing to take on leadership positions in the House of Assembly and/or their political party. In
this instance, a legislator would spend more time in administration, rather than focusing on
representation and constituency service. As legislators gain leadership positions, it is also
possible this role competes with the representative function. Conversely, it is possible junior
legislators, in an effort to secure reelection, may be more committed to constituency work and
representation. In this scenario, senior legislators may believe they have already demonstrated
their representative function, and now choose prioritize other factors.
These analyses demonstrate source of revenue is a significant determinant of individual
legislators’ attitudes and behaviors toward the execution of the representation function. Chapter
Three establishes the positive relationship between taxation and representation at a macro-level
in Nigeria: tax revenue shapes budgetary priorities, and states generating more revenue from
taxes are more likely to spend income on non-recurrent expenditures, like public service
delivery. However, the previous findings confirm that sources of government trickle down to the
micro-level. State reliance on taxes (or other sources of revenue) shape legislators’ incentives to
prioritize (or not prioritize) representation and constituent interests. As scholars have argued in
other contexts, when government is generates income from taxes on citizens, there is a greater
inducement for public officials to shift policy decisions to their constituents’ interests. Elected
leaders do so in order to maintain income from their tax base (Tilly, 1985/1990; Bates and Lien,
1985; Levi, 1988; North and Weingast, 1989; Hoffman and Norberg, 1994; Timmons, 2005).
My analysis demonstrates this theory can also be used to explain how Nigerian elites perceive
and carry out representation in state legislatures. Nigerian legislators in settings with a greater
capacity to generate tax income also prioritize representation of constituents’ interests and spend
more time performing constituency services.

159

Dimensions of Elite Representation
In the previous analysis, legislators’ perceptions of representation are captured with seven
separate indicators. This section investigates the possibility that these variables scale together,
such that there is a latent construct of “representation” in the minds of Nigerian legislators. In
other words, what are the dimensions of representation? To answer this question, I perform a
factor analysis using STATA 12.0. This analytical method will allow me to group variables into
homogenous categories and identify the fundamental elements of representation (Garrett-Mayer
and Onicescu, 2009).
“Representation” is operationalized with the following variables: legislators’ ratings of
how effective their State House of Assembly is in representing constituents; legislators’ ratings
of how effective their State House of Assembly is in representing women’s interests; the
percentage of time legislators report performing constituency work; whether legislators believe
representation and performing constituency services are “the most important part of being a
member of the State Assembly; whether legislators should prioritize their constituency’s views
over other considerations when making decisions; whether legislators should support or oppose
their party’s position if it conflicts with constituents’ views; whether legislators should remain in
office if they ignore what their constituents have to say. I include these seven indicators in the
factor analysis and perform an orthogonal varimax rotation of the principal factors (so that the
factors are not correlated with one another). Tables 4.23 and 4.24 present the results.
Factor analysis identifies two dimensions of representation. We can see that the first factor
is defined the following five variables: the amount of time legislators report performing
constituency services (loading = 0.581); the likelihood legislators identify representation and
performing constituency services as the most important part of their job (loading = 0.463); the

160

likelihood legislators report prioritizing constituents’ views over other considerations (loading =
0.546); the likelihood legislators will oppose their party’s position if it conflicts with
constituents’ views (loading = 0.317); legislators’ belief that officials should lose their seat if
they ignore what constituents have to say (loading = 0.409). Based on the variables included, it
seems that this dimension of representation focuses on the extent to which legislators prioritize
their constituents’ views and needs. This factor explains about 82% of the observed variance.
Moreover, a test of internal consistency results in a Cronbach’s alpha of 0.76. Taken together,
this suggests that the five variables do scale together and can be combined into reliable index.
The second factor is defined by the remaining two variables: legislators’ ratings of how
well their state legislature represents constituents, in general, (loading = 0.500) and women,
specifically, (loading = 0.551). Factor Two explains 47% of variance and yields a Cronbach’s
alpha of 0.721. This dimension of legislators’ perception of representation embodies overall
ratings of how well their state legislature is carrying out the representative function.
Based on these results, I generate two indices to capture the two dimensions of
representation. The first index assembles the five variables in Factor One, and focuses on
legislators’ reports of how they prioritize constituents’ interests and public services. The second
index combines the two variables in Factor Two, and concentrates on legislators’ perceptions of
how well their House of Assembly represents constituents and women. Using these two indices
as indicators for representation (dependent variable), I conduct two more multi-level mixed
effects linear regressions (clustered robust standard errors). In these analyses, I hypothesize that
in states that derive higher levels of government income from taxes, state legislators will be more
representative of constituents. In states with higher levels of tax generation, officials will also
report their state House of Assembly is more effective in carrying out the representative function.

161

Table 4.25 presents the results of analysis and supports the two hypotheses. First we
observe a significant relationship whereby the constituency priorities index is positively linked to
average tax generation (b = 0.149) and state tax reliance in 2009 (b = 0.148). Evidence suggests
that Nigerian legislators in states with higher levels of tax generation are more likely to prioritize
their constituents (as captured by the index of Factor One variables).
In a similar vein, in states deriving a larger portion of their income from taxes on citizens,
legislators are more likely to give a favorable rating of how the House of Assembly represents
their people. This positive and significant relationship holds across the two measures of tax
generation—average tax generation between 1999 and 2009 (b = 0.058) and state tax reliance in
2009 (b = 0.054).
These analyses demonstrate that there are two dimensions of representation: the first
focuses on legislators’ prioritizing their constituents’ views and needs, while the second factor
centers on how effective these officials think their House of Assembly represents their people.
Furthermore, in states that generate a higher level of income from taxes on citizens, legislators
are more likely to identify constituents’ interests and needs as important. Legislators in these
states are also more likely to view their state legislature as effective in carrying out the
representative function.

Conclusion
This chapter investigates the relationship between revenue at representation at the microlevel. In Chapter Three I establish how source of revenue shapes government budgeting and
expenditures. However, the chapter (Four) demonstrates the incentives created by reliance on
different forms of public revenue shape individual elected official’s priorities and decision-

162

making. As previous scholars have theorized in Western Europe, when government is more
reliant on taxes, elected leaders are more likely to say that they would shift public policy
decisions to reflect citizens’ interests. They make this shift as a part of the contract with
constituents: tax payment in exchange for a greater influence in governance (Tilly, 1985/1990;
Bates and Lien, 1985; Levi, 1988). Utilizing original elite surveys and Nigeria’s intra-regional
and inter-regional variation, we have seen this theory holds in an African context: Nigerian elite
attitudes change in the context of the different restraints produced by tax generation. Nigerian
legislators in more tax-reliant states present themselves as most likely to represent citizens: they
report spending more time performing constituency services and prioritizing constituents’
preferences and interests above other considerations (e.g. party position, personal views, etc).
Moreover, factor analysis confirms establishes two dimensions of how legislators think about
representation: the first is constituency oriented, while the second focuses on how effective their
state legislature is in representing various citizens. Multi-level analysis once again confirms my
hypotheses that legislators in states with higher levels of tax generation are more likely to
identify constituents’ interests and needs as important. These officials are also more likely to
view their House of Assembly as effective in carrying out the representative function.
Having analyzed the relationship between revenue and elite representation at the individual
level, the next chapter considers Nigerian citizens’ perceptions of the tax contract. Using public
opinion data, I will consider ordinary Nigerians’ attitudes and behaviors toward taxation and
representation in government.

163

Chapter Four Tables and Figures

Table 4.1: TAX GENERATION
IN THE SOUTH-WEST
REGION (TAX REVENUE AS
A PERCENTAGE OF TOTAL,
%)
STATE
EKITI
LAGOS
OGUN
ONDO
OSUN
OYO

Table 4.2: TAX
GENERATION IN THE
NORTH-EAST REGION
(TAX REVENUE AS A
PERCENTAGE OF TOTAL,
%)
STATE
ADAMAWA
BAUCHI
BORNO
GOMBE
TARABA
YOBE

2005
9.3
51.2
N/A
10.1
14.9
18.3

2005
3.3
2.9
3.8
4.6
2.3
2.5

164

2007
10.1
62.2
11.4
15.2
15.2
22.3

2007
4.7
2.9
4.8
5.2
6.7
2.5

2008
11.9
63.5
27.5
16.4
16.2
20.0

2008
4.979
3.6
5.3
7.1
8.2
2.4

2009
13.1
64.3
29.3
N/A
22.0
20.6

2009
5.1
5.8
7.5
11.1
9.0
3.0

Table 4.3: TAX VS. OIL
INCOME IN NIGERIA’S
REGIONS (PRECENTAGE OF
TOTAL REVENUE, 1999 –
2009, %)
REGION
NORTHERN STATES
SOUTHERN OIL STATES
SOUTHERN, NON-OIL STATES

Table 4.4: TAX
GENERATION IN THE
NORTH-WEST REGION
(TAX REVENUE AS A
PERCENTAGE OF TOTAL,
%)
STATE
JIGAWA
KADUNA
KANO
KATSINA
KEBBI
SOKOTO
ZAMFARA

2005
7.4
7.7
11.2
9.2
3.1
8.9
2.8

165

TAX
8.2
11.6
18.4

2007
N/A
9.1
14.7
8.4
5.9
19.3
7.9

OIL
64.5
67.6
53.3

2008
12.0
9.8
17.2
10.9
9.4
46.7
8.3

2009
15.6
15.4
21.6
11.8
13.9
47.7
15.4

Table 4.5: STANDARD
DEVIATION IN REGIONAL
TAX AND OIL INCOME AS A
PERCENTAGE OF STATE
BUDGET (1999 – 2009, %)

REGION
NORTH-WEST
MIDDLE BELT
NORTH-EAST
SOUTH-WEST
SOUTH-SOUTH
SOUTH-EAST

STANDARD
DEVIATION IN
TAX INCOME
4.252
2.062
2.380
16.568
5.931
3.726

Table 4.6: AVERAGE TAX
GENERATION IN THE SOUTHWEST REGION, 1999 - 2009 (% OF
TOTAL REVENUE)
STATE
EKITI
LAGOS
OGUN
ONDO
OSUN
OYO

TAX
7.5
53.4
18.9
11.2
16.4
16.7

166

STANDARD
DEVIATION IN
OIL INCOME
4.106
3.192
3.885
17.977
7.989
3.054

OIL
71.0
23.5
56.5
73.5
61.1
59.1

Table 4.7: AVERAGE TAX
GENERATION IN THE NORTHWEST REGION, 1999 - 2009 (% OF
TOTAL REVENUE)
STATE
JIGAWA
KADUNA
KANO
KATSINA
KEBBI
SOKOTO
ZAMFARA

167

TAX
7.6
12.9
13.2
7.1
8.2
17.6
7.9

OIL
67.4
59.4
56.6
66.2
65.7
63.6
66.1

Table 4.8: TAX
GENERATION IN THE
SOUTH-WEST
REGION (TAX
REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE,
1999 – 2009, %)
STATE
LAGOS (HIGH
PERFORMER)
OYO (AVERAGE
PERFORMER)
EKITI (LOW
PERFORMER)
Table 4.8 (cont’d): TAX
GENERATION IN THE
SOUTH-WEST
REGION (TAX
REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE,
1999 – 2009, %)
STATE
LAGOS (HIGH
PERFORMER)
OYO (AVERAGE
PERFORMER)
EKITI (LOW
PERFORMER)

1999

2000

2001

2002

2003

40.8

41.8

49.7

54.3

53.4

10.0

11.5

13.9

15.9

17.7

2.8

3.2

5.3

6.0

6.0

2005

2007

2008

2009

MEAN

51.2

62.2

63.5

64.3

53.4

18.3

22.3

20.0

20.6

16.7

9.3

10.1

11.9

13.1

7.5

168

Table 4.9: OIL DEPENDENCE
IN THE SOUTH-WEST
REGION (PETROLEUM
REVENUE AS A
PERCENTAGE OF TOTAL
REVENUE, 1999 – 2009, %)
STATE
LAGOS (HIGH PERFORMER)
OYO (AVERAGE
PERFORMER)
EKITI (LOW PERFORMER)
Table 4.9 (cont’d ): OIL
DEPENDENCE IN THE
SOUTH-WEST REGION
(PETROLEUM REVENUE AS
A PERCENTAGE OF TOTAL
REVENUE, 1999 – 2009, %)
STATE
LAGOS (HIGH PERFORMER)
OYO (AVERAGE
PERFORMER)
EKITI (LOW PERFORMER)

1999

2000

2001

2002

2003

22.95

39.09

39.26

20.95

20.37

54.33
73.85

66.12
82.03

77.65
85.73

51.58
62.47

65.37
73.69

2005

2007

2008

2009

MEAN

26.09

10.62

16.43

15.34

23.5

62
80.37

55.38
65.4

49.95
58.04

49.2
57.57

59.1
71.0

169

Table 4.10: TAX
GENERATION IN THE
NORTH-WEST REGION
(TAX REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE, 1999
– 2009, %)
STATE
SOKOTO (HIGH
PERFORMER)
KANO (AVERAGE
PERFORMER)
KATSINA (LOW
PERFORMER)

1999

2000

2001

2002

2003

5.4

6.8

6.7

8.0

8.9

6.8

10.4

11.4

12.8

12.7

2.9

3.7

4.4

6.0

6.08

Table 4.10 (cont’d): TAX
GENERATION IN THE
NORTH-WEST REGION
(TAX REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE, 1999
– 2009, %)
STATE
SOKOTO (HIGH
PERFORMER)
KANO (AVERAGE
PERFORMER)
KATSINA (LOW
PERFORMER)

2005

2007

2008

2009

MEAN

8.9

19.3

46.7

47.7

17.6

11.2

14.7

17.2

21.6

10.8

9.2

8.4

10.9

11.8

7.1

170

Table 4.11: OIL
DEPENDENCE IN THE
NORTH-WEST
REGION (PETROLEUM
REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE,
1999 – 2009, %)
STATE
SOKOTO (HIGH
PERFORMER)
KANO (AVERAGE
PERFORMER)
KATSINA (LOW
PERFORMER)

Table 4.11 (cont’d): OIL
DEPENDENCE IN THE
NORTH-WEST
REGION (PETROLEUM
REVENUE AS A
PERCENTAGE OF
TOTAL REVENUE,
1999 – 2009, %)
STATE
SOKOTO (HIGH
PERFORMER)
KANO (AVERAGE
PERFORMER)
KATSINA (LOW
PERFORMER)

1999

2000

2001

2002

2003

77.08

84.01

76.32

72.37

73.48

51.53

70.87

50.58

54.59

39.28

75.95

75.35

77.06

46.71

69.8

2005

2007

2008

2009

MEAN

66.25

53.42

35.28

34.41

76.7

76.79

63

49.55

53.22

53.4

78.42

48.62

63.2

60.69

69.0

171

Table 4.12:
LEGISLATIVE
PRODUCTIVITY IN
THE SOUTH-WEST
REGION (NUMBER OF
BILLS PASSED INTO
LAW, 1999 - 2010)
FIRST
SECOND
THIRD
ASSEMBLY ASSEMBLY ASSEMBLY
30
76
43
40
30
29
23
38

STATE
LAGOS
OYO
EKITI

TOTAL
149
70
90

Table 4.13:
LEGISLATIVE
PRODUCTIVITY IN
THE NORTH-WEST
REGION (NUMBER
OF BILLS PASSED
INTO LAW, 2007 2011)

STATE
SOKOTO
KANO

79

KATSINA

FIRST
SESSION
6
13

SECOND
SESSION
12
9

8

9

79

THIRD FOURTH
SESSION SESSION TOTAL
8
10
36
8
6
36
9

6

32

Due to the nature of the data, information is only available for the first six months (June 2010
– December 2011) of the Katsina State House of Assembly’s fourth session.
172

Table 4.14:
SAMPLING
PROCEDURES IN SIX
STATE
LEGISLATURES

STATE
LAGOS
OYO
EKITI

TOTAL NUMBER
OF LEGISLATORS
IN STATE
LEGISLATURE
40
32
26
Subtotal
98

SOKOTO
KANO
KATSINA

30
40
34
104

TOTAL

15
20
17
52

202

Subtotal

Table 4.15:
AGE OF
LEGISLATORS
(YEARS), n = 101
STATES
LAGOS
OYO
EKITI
SOKOTO
KANO
KATSINA
TOTAL SAMPLE

NUMBER OF
STATE
LEGISLAORS
IN SAMPLE
20
16
13
49

101

MINIMUM
38
37
35
33
38
40
33

173

MAXIMUM
63
55
55
58
57
55
63

MEAN
49
47
46
44
48
48
47

Table 4.16: GENDER OF
LEGISLATORS, n = 101
STATES
LAGOS
OYO
EKITI
SOKOTO
KANO
KATSINA
TOTAL SAMPLE

MALE
17/20
15/16
13/13
15/15
20/20
17/17
96%

FEMALE
3/20
1/16
0/13
0/15
0/20
0/17
4%

Table 4.17: ETHNICITY
OF LEGISLATORS, n =
101
STATES
LAGOS
OYO
EKITI
SOKOTO
KANO
KATSINA
TOTAL SAMPLE

YORUBA
19/20
16/16
13/13
0/15
0/20
0/17
48%

174

HAUSAFULANI
0/20
0/16
0/13
15/15
20/20
17/17
51%

OTHER
1/20
0/16
0/13
0/15
0/20
0/17
1%

Table 4.18: RELIGION
OF LEGISLATORS, n =
101
STATES
CHRISTIANITY
LAGOS
11/20
OYO
9/16
EKITI
13/13
SOKOTO
0/15
KANO
0/20
KATSINA
0/17
TOTAL SAMPLE*
33%
* 1% of respondents declined answering.

ISLAM
8/20
7/16
0/13
15/15
20/20
17/17
66%

Table 4.19:
HIGHEST
LEVEL
LEGISLATORS'
EDUCATION, n
= 101

STATES
LAGOS
OYO
EKITI
SOKOTO
KANO
KATSINA
TOTAL
SAMPLE

POSTPOSTSECONDARY SECONDARY BACHELOR'S GRADUATE
SCHOOL
DIPLOMA
DEGREE
DEGREE
0/20
2/20
7/20
11/20
0/16
2/16
7/16
7/16
0/13
1/13
1/13
11/13
4/15
0/15
9/15
2/15
3/20
8/20
5/20
4/20
6/17
6/17
2/17
3/17
15%

17%

175

31%

37%

Table 4.20:
OCCUPATION PRIOR
TO LEGISLATIVE
OFFICE (%), n = 101
OCCUPATION
BUSINESS PERSON
CIVIL SERVANT
LOCAL COUNCILLOR
COMMERCIAL
FARMER
OTHER
PROFESSIONAL
TEACHER
SUPERVISOR/MIDLEVEL MANAGER
BANKER
UNIVERSITY
LECTURER
LAWYER
ACCOUNTANT
GENERAL MANAGER
JOURNALIST

Table 4.21: NUMBER
OF YEARS IN
LEGISLATIVE OFFICE
(YEARS), n = 101
STATES
LAGOS
OYO
EKITI
SOKOTO
KANO
KATSINA
TOTAL SAMPLE

MINIMUM
4
3
4
4
4
2
2

176

%
26
23
21
7
6
5
4
3
2
1
1
1
1

MAXIMUM
12
8
4
12
12
4
12

MEAN
6
4
4
5
5
4
5

Table 4.22A: EFFECT
OF REVENUE ON
LEGISLATORS'
EXECUTION OF
REPRESENTATION

1

2

REPRESENTING
CONSTITUENTS

3

TIME SPENT
REPRESENTING
DOING
WOMEN
CONSTITUENCY
SERVICE

STATE LEVEL
AVERAGE STATE TAX
DEPENDENCE (1999 2009)
2009 STATE TAX
DEPENDENCE
INDIVIDUAL LEVEL
YEARS IN OFFICE
MEMBER OF
MAJORITY PARTY
Notes: *p < 0.10, two-tailed.

0.035
(0.002)

0.045*
(0.046)

0.198*
(0.117)

0.008
(0.002)

0.046*
(0.037)

0.269*
(0.094)

-0.015
(0.009)
0.100*
(0.038)

-0.019
(0.019)
0.049*
(0.083)

-0.387*
(0.098)
0.546
(2.110)

177

Table 4.22A (cont’d):
EFFECT OF REVENUE
ON LEGISLATORS'
EXECUTION OF
REPRESENTATION

1

2

REPRESENTING
CONSTITUENTS
CONSTANT
STANDARD ERROR
RANDOM EFFECTS
RESIDUAL STANDARD
DEVIATION
INTRACLASS
CORRELATION
LOG LIKLIHOOD
OBSERVATIONS
STATES
Notes: *p < 0.10, two-tailed.

0.942*
0.048

3

TIME SPENT
REPRESENTING
DOING
WOMEN
CONSTITUENCY
SERVICE
0.762*
24.092*
0.104
2.653

0.163

0.353

9.035

0.011

0.025

0.636

-138.164

-76.512

-52.81

101
6

101
6

101
6

178

Table 4.22B:
EFFECT OF
REVENUE ON
LEGISLATORS'
EXECUTION OF
REPRESENTATION

4

REPRESENTATION
IS MOST
IMPORTANT JOB
STATE LEVEL
AVERAGE STATE
TAX DEPENDENCE
(1999 - 2009)
2009 STATE TAX
DEPENDENCE
INDIVIDUAL
LEVEL
YEARS IN OFFICE
MEMBER OF
MAJORITY PARTY

5

6

VIEWS OF
OPPOSE PARTY
CONSTITUENTS
TO SUPPORT
ARE MOST
CONSTITUENTS'
IMPORTANT
VIEWS
CONSIDERATION

7

MPs WHO
IGNORE
CONSTITUENTS
SHOULD LOSE

0.047*
(0.005)

0.095*
(0.059)

0.076*
(0.012)

0.092*
(0.006)

0.043*
(0.004)

0.043*
(0.005)

0.086*
(0.009)

0.093*
(0.005)

-0.054*
(0.023)
0.107
(0.122)

-0.043*
(0.055)
0.089
(0.107)

-0.093*
(0.053)
0.115
(0.223)

-0.074*
(0.038)
0.174*
(0.102)

Notes: *p < 0.10, two-tailed.

179

Table 4.22B (cont’d):
EFFECT OF
REVENUE ON
LEGISLATORS'
EXECUTION OF
REPRESENTATION

4

5

6

7

REPRESENTATION
IS MOST
IMPORTANT JOB

VIEWS OF
CONSTITUENTS
ARE MOST
IMPORTANT
CONSIDERATION

OPPOSE PARTY
TO SUPPORT
CONSTITUENTS'
VIEWS

MPs WHO
IGNORE
CONSTITUENTS
SHOULD LOSE

-0.313*
0.122

0.340*
0.135

0.475*
0.280

0.505*
0.129

0.415

0.459

0.954

0.438

0.029

0.032

0.067

0.031

-54.548

-64.718

-80.646

-59.985

101
6

101
6

101
6

101
6

CONSTANT
STANDARD ERROR
RANDOM EFFECTS
RESIDUAL
STANDARD
DEVIATION
INTRACLASS
CORRELATION
-2 x LOG
LIKLIHOOD
(DEVIANCE)
N
STATES
Notes: *p < 0.10, two-tailed.

180

Table 4.23: FACTOR
ANALYSIS OF ELITE
REPRESENTATION
INDICATORS
(Orthogonal Varimax
!!
Principle Factors)
FACTOR
VARIANCE
FACTOR 1
1.045
FACTOR 2
0.595
OBSERVATIONS
101
RETAINED
FACTORS
2
NUMBER OF
PARAMETERS
18
CHI2 (21)
73.65
PROB > CHI2
0

181

!!

PROPORTION
0.818
0.466
101
2
18
73.65
0

Table 4.24: FACTOR
ANALYSIS OF ELITE
REPRESENTATION
INDICATORS
(Rotated Factor
Loadings)
VARIABLE
REPRESENTING
CONSTITUENTS
REPRESENTING
WOMEN
TIME SPENT DOING
CONSTITUENCY
SERVICE
REPRESENTATION
IS MOST
IMPORTANT JOB
VIEWS OF
CONSTITUENTS
ARE MOST
IMPORTANT
CONSIDERATION
OPPOSE PARTY TO
SUPPORT
CONSTITUENTS'
VIEWS
MPs WHO IGNORE
CONSTITUENTS
SHOULD LOSE

FACTOR 1

FACTOR 2

0.055

0.500

0.050

0.551

0.581

0.057

0.463

0.157

0.546

-0.029

0.317

-0.036

0.409

0.111

182

Table 4.25: EFFECT
OF REVENUE ON
ELITE
REPRESENTATION'S
TWO DIMMENSIONS

1

2

RATING STATE
PRIORITIZING
LEGISLATURE'S
CONSTITUENTS REPRESENTATIVE
FUNCTION
STATE LEVEL
AVERAGE STATE
TAX DEPENDENCE
(1999 - 2009)
2009 STATE TAX
DEPENDENCE
INDIVIDUAL LEVEL
YEARS IN OFFICE
MEMBER OF
MAJORITY PARTY
CONSTANT
STANDARD ERROR
RANDOM EFFECTS
RESIDUAL
STANDARD
DEVIATION
INTRACLASS
CORRELATION
LOG LIKLIHOOD
OBSERVATIONS
STATES
Notes: *p < 0.10, two-tailed.

0.149*
(0.064)

0.058*
(0.016)

0.148*
(0.047)

0.054*
(0.013)

-0.072*
(0.029)
0.261
(0.149)
5.049
0.287

-0.048*
(0.011)
0.176
(0.071)
3.380
0.070

2.341

0.565

0.253

0.127

-53.233
101
6

-85.635
101
6

183

Figure 4.1: Tax Generation in the South-West Cases (1999 – 2009)

70

% OF STATE REVENUE

60
50
LAGOS (HIGH PERFORMER)

40

OYO (AVERAGE
PERFORMER)

30

EKITI (LOW PERFORMER)
20
10
0

1999 - 2009

184

Figure 4.2: Oil Dependence in the South-West Cases (1999 – 2009)

100
90

% OF STATE REVENUE

80
70
60

LAGOS (HIGH PERFORMER)

50

OYO (AVERAGE
PERFORMER)

40

EKITI (LOW PERFORMER)

30
20
10
0

1999 - 2009

185

Figure 4.3: Tax Generation in the North-West Cases (1999 – 2009)
60

% OF STATE REVENUE

50

40
SOKOTO (HIGH
PERFORMER)
30

KANO (AVERAGE
PERFORMER)
KATSINA (LOW
PERFORMER)

20

10

0

1999 - 2009

186

Figure 4.4: Oil Dependence in the North-West Cases (1999 – 2009)
90
80

% OF STATE REVENUE

70
60
50

SOKOTO (HIGH
PERFORMER)

40

KANO (AVERAGE
PERFORMER)

30

KATSINA (LOW
PERFORMER)

20
10
0

1999 - 2009

187

Figure 4.5: Legislators’ Ratings of How Effective their State Legislature Represents Constituents vs. Average State Tax
Generation

4
LAGOS

AVERAGE RATING OF STATE
LEGISLATURE

3.9

KATSINA

3.8
3.7
3.6
3.5

EKITI

3.4
SOKOTO

3.3
OYO

3.2

KANO

3.1
3
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.43
2
r = 0.18

188

60

Figure 4.6: Legislators’ Ratings of How Effective their State Legislature Represents Constituents vs. 2009 State Tax
Generation

AVERAGE RATING OF STATE
LEGISLATURE

4

LAGOS

3.9

KATSINA

3.8
3.7
3.6
3.5
3.4

EKITI

3.3
3.2

KANO

3.1

SOKOTO

OYO

3
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.26
2
r = 0.07

189

60

70

Figure 4.7: Legislators’ Ratings of How Effective their State Legislature Represents Woman vs. Average State Tax Generation

AVERAGE RATING OF STATE
LEGILATURE

4
3.8
3.6

LAGOS

SOKOTO

3.4
3.2

EKITI

3

OYO

2.8

KANO

2.6
2.4

KATSINA

2.2
2
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.87
2
r = 0.76

190

60

Figure 4.8: Legislators’ Ratings of How Effective their State Legislature Represents Woman vs. 2009 State Tax Generation

AVERAGE RATING OF STATE
LEGISLATURE

4
LAGOS

3.8
3.6
3.4

SOKOTO

OYO

3.2

EKITI

3
2.8

KANO

2.6

KATSINA

2.4
2.2
2
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.94
2
r = 0.88

191

60

70

Figure 4.9: Time Legislators Report Devoting to Constituency Work vs. Average State Tax Generation

AVERAGE % TIME DEVOTED TO
CONSTITUENCY WORK

55
50

LAGOS

SOKOTO

45
40
35
KANO
30

OYO

EKITI
25

KATSINA

20
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.75
2
r = 0.56

192

60

Figure 4.10: Time Legislators Report Devoting to Constituency Work vs. 2009 State Tax Generation

AVERAGE % TIME DEVOTED TO
CONSTITUENCY WORK

60
55

SOKOTO

50
LAGOS

45
40
35

KANO

30

OYO
EKITI

25

KATSINA

20
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.97
2
r = 0.94

193

60

70

Figure 4.11: Representation is Legislator’s Most Important Job vs. Average State Tax Generation

REPRESENTATION IS MOST
IMPORTANT JOB (AVERAGE)

0.8
0.7

SOKOTO

LAGOS

0.6
0.5
0.4
KANO
0.3

OYO

KATSINA

0.2

EKITI
0.1
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.78
2
r = 0.61

194

60

Figure 4.12: Representation is Legislator’s Most Important Job vs. 2009 State Tax Generation

REPRESENTATION IS MOST
IMPORTANT JOB (AVERAGE)

0.8
0.7

SOKOTO

LAGOS

0.6
0.5
0.4
KANO
0.3

KATSINA

OYO

0.2
EKITI
0.1
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.96
2
r = 0.92

195

60

70

Figure 4.13: Constituents’ Views Are the Most Important Consideration When Making Decisions vs. Average State Tax
Generation

CONSTITUENTS ARE MOST
IMPORTANT CONSIDERATION
(AVERAGE)

0.9
0.8
SOKOTO

0.7

LAGOS

0.6
0.5
OYO

0.4

KANO

0.3
EKITI

0.2
0.1

KATSINA

0
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.73
2
r = 0.53

196

60

Figure 4.14: Constituents’ Views Are the Most Important Consideration When Making Decisions vs. 2009 State Tax
Generation

CONSTITUENTS ARE MOST
IMPORTANT CONSIDERATION
(AVERAGE)

0.9
0.8

SOKOTO

LAGOS

0.7
0.6
0.5

OYO

0.4

KANO

0.3

EKITI

0.2
0.1

KATSINA

0
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.93
2
r = 0.86

197

60

70

Figure 4.15: In Case of Conflict, Will Oppose Party to Support Constituents vs. Average State Tax Generation

OPPOSE PARTY TO SUPPORT
CONSTITUENTS (AVERAGE)

1
0.9
SOKOTO

0.8

LAGOS

0.7
OYO

0.6

EKITI

0.5
0.4

KANO

0.3
0.2

KATSINA

0.1
0
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.64
2
r = 0.41

198

60

Figure 4.16: In Case of Conflict, Will Oppose Party to Support Constituents vs. 2009 State Tax Generation

OPPOSE PARTY TO SUPPORT
CONSTITUENTS (AVERAGE)

1

SOKOTO

0.9
0.8

LAGOS

0.7
0.6

OYO

0.5

EKITI

0.4

KANO

0.3
0.2

KATSINA

0.1
0
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.89
2
r = 0.79

199

60

70

Figure 4.17: If Legislator Ignores Constituents He/She Should Lose Seat vs. Average State Tax Generation

1

LAGOS

LOSE SEAT IF IGNORES
CONSTITUENTS (AVERAGE)

0.9
0.8
SOKOTO

0.7
KANO

0.6

OYO

0.5

EKITI

0.4
0.3
0.2

KATSINA

0.1
0
0

10
20
30
40
50
TAXES AS A % OF STATE INCOME (AVERAGE, 1999 - 2009)

Correlation = 0.80
2
r = 0.64

200

60

Figure 4.18: If Legislator Ignores Constituents He/She Should Lose Seat vs. 2009 State Tax Generation

1

LAGOS

LOSE SEAT IF IGNORES
CONSTITUENTS (AVERAGE)

0.9
0.8

SOKOTO

0.7
OYO

KANO

0.6
0.5

EKITI

0.4
0.3
0.2

KATSINA

0.1
0
0

10

20
30
40
50
TAXES AS A % OF STATE INCOME (2009)

Correlation = 0.86
2
r = 0.74

201

60

70

CHAPTER FIVE

Popular Perceptions of Taxation and Representation: Attitudes and Behaviors To Tax
Payment
Background
In the last two chapters, I have presented evidence supporting the thesis of this dissertation
at the macro and micro levels. I have found that when Nigerian subnational governments
generate more of their revenue from non-oil taxes on citizens, they seem to be more
representative. This relationship is evident in the following ways:
1. As local and state governments rely more on income generated from taxes, they are less
likely to spend income on politicians’ salaries and allowances. Instead, they are more
likely to expend resources on non-recurrent expenditures, including public service
delivery.
2. Individual legislators in states that generate more income from non-oil taxes on citizens
are more likely to prioritize and execute their representative function. Officials in these
states say that they spend more time performing constituency services. They also claim
to rank citizens’ views above other considerations when making policy decisions.
Given these findings about how tax revenue shapes elite attitudes and behavior, under what
circumstances are ordinary Nigerian citizens willing to pay taxes in the first place?
Following Levi (1988), I argue that the relationship between government and citizens
undergirds politicians’ ability to extract tax revenue from citizens. In order to maximize
government’s capacity to collect taxes, it is necessary to establish compliance among the
citizenry. The use of force or coercion is one, albeit expensive, method of garnering this mass

202

compliance. However, Levi (1988: 52 - 60) identifies quasi-voluntary compliance as the modal
technique of obtaining citizens’ obedience toward paying taxes. This form of compliance is
“voluntary because taxpayers choose to pay.” But it is also considered quasi-voluntary “because
the noncompliant [who do not pay their taxes] are subject to coercion—if they are caught.”
According to Levi (1988), a citizen’s quasi-voluntary compliance is dependent on two
beliefs: first, an individual must believe that political leaders will fulfill their end of the bargain.
A citizen must perceive that, if he or she pays taxes, government will reciprocate by incorporate
citizens’ interests in policymaking. Second, a taxpayer must believe that other citizens are
complying with tax payment. Levi (1988: 53) argues that “taxpayers are strategic actors who
will cooperate only when they can expect others to cooperate as well. The compliance of each
depends on the compliance of others. No one prefers to be a ‘sucker.’” I hypothesize that these
kinds of calculations occur among the Nigerian populace. If so, and citizens do perceive tax
payment as a contract with government in exchange for representation, we should be able to
observe at least two empirical relationships. First, citizens who perceive that elected officials are
representing their interests in government will be willing to pay taxes. These individuals will
also report compliance with actual tax payment at a higher rate than Nigerians who do not feel
politicians are representing them. Second, Nigerians who believe that other citizens are
complying with tax payment will be more willing to pay taxes. They will also follow through
with actual payment.
Before testing these hypotheses, it is important to consider Nigerians’ general perspectives
on taxation and compliance. Using Nigerian public opinion data from Afrobarometer Round
Four (2008), we can paint an overall picture of ordinary citizens’ perceptions.
The Afrobarometer (2008) asks Nigerian respondents the extent to which they believe “the

203

tax department always has the right to make people pay taxes,” thus gauging their latent
willingness to pay taxes (Figure 5.1). A majority of Nigerians agree or strongly agree with this
statement (66%). This is contrasted with 23% of Nigerians who disagree or strongly disagree
with the tax department’s right to compel tax payment from citizens. Moreover, this belief in the
government’s right to compel tax payment holds despite individual beliefs about whether the
executive should account for how tax income is spent (Figure 5.2). For example, 53% of
Nigerians believe “the National Assembly should ensure that the President explains to it on a
regular basis how [the] government spends taxpayers’ money.” On the other hand, 43% of
respondents believe “the President should be able to devote his full attention to developing the
country rather than wasting time justifying his actions” (Afrobarometer, 2008). A crosstabulation (Table 5.1) shows that even though Nigerians are closely split in their opinions about
whether the president must justify public spending, the majority still agrees that it’s the tax
department’s responsibility to make citizens pay taxes. For example, of the Nigerians who
strongly agree that the executive must justify to the National Assembly how tax income is being
spent, 70% also agree or strongly agree that the tax department has the right to compel tax
payment. Similarly, of Nigerians who strongly believe the executive does not have to account
for tax spending (free to act on their own), 76% also agree or strongly agree that people must pay
taxes. A Cramer’s V test result of 0.113 demonstrates that there is a weak relationship between
Nigerians’ willingness to pay taxes and their perceptions of executive accountability. Thus,
Nigerians’ latent willingness to comply with government demands for taxes is relatively
independent and endures despite differing views in executive justification of public spending.
Now knowing that a majority of Nigerians agree that government is justified in compelling
tax payment, whom do citizens believe has the primary responsibility for collecting (specifically)

204

income taxes: federal government, state governments, local governments, members of the
community, or traditional leaders (Figure 5.3)? First, it is important to point out Nigerians
overwhelmingly believe income tax collection should primarily be a government responsibility.
Only 6% of respondents believe a non-government entity (i.e. community members, traditional
leaders), should be primarily responsible for collecting income taxes. Generally, Nigerians
perceive income tax collection as a state and local responsibility, 36% of respondents choosing
the state government and another 36% choosing the local government. On the other hand, 21%
of respondents believe income tax collection is the federal government’s responsibility.
According to Nigeria’s tax law, income tax collection is a state government responsibility.
However, there are exceptions: federal government employees, members of the military, and
residents of the Federal Capital Territory pay income taxes to the federal government.
Figure 5.4 presents results from respondent reports of actual payment of various taxes in
the last year. In 2008, about 30% of Nigerians report making income tax payments during this
period. 27% of respondents indicate making property tax payments, while 31% of Nigerians say
that in the last year they paid local government fees taxes/fees. These rates of payment may, at
first seem, low, but in comparison to other African nations, Nigeria is not statistically different.
Comparing Nigeria to other Afrobarometer countries, we see that the rates of tax payment are
slightly higher. With respect to income tax payment, even though only 30% of Nigerians indicate
paying in the last year, this is higher than the mean (22%) in the 20 countries surveyed by the
Afrobarometer. Similarly, the 27% who report paying property taxes in Nigeria is higher than
the mean (24%) in Afrobarometer countries overall. With that said, this difference is close to the
margin of sampling error in Afrobarometer surveys (+/- 3% at a 95% confidence level). Last, in
Nigeria, 31% who report paying local government taxes is higher than the rate of 25% in all

205

countries included in the Afrobarometer.
Since the gaps between Nigeria and the Afrobarometer countries (on average) are so small,
I perform three chi-square tests to examine the significance of these differences. Tables 5.2, 5.3,
and 5.4 show results of the comparisons in tax payment (income, property, and local taxes)
between Nigeria and the averages in all Afrobarometer countries. We see that there is no
statistically significant difference between rates of tax payment in Nigeria and other
Afrobarometer countries (in each test, p = 1). Thus, Nigeria remains similar to other African
nations.
Returning to the initial, question, I hypothesize the following:
Hypothesis 1: Nigerians are more likely to express willingness to pay taxes as they perceive
higher quality of representation from government.
Hypothesis 2: Nigerians are more likely to express willingness to pay taxes as they express
higher quality of trust in other citizens.
Hypothesis 3: Nigerians are more likely to report actual tax payment as they perceive higher
quality of representation from government.
Hypothesis 4: Nigerians are more likely to report actual tax payment as they express higher
quality of trust in other citizens.
These hypotheses are derived from Levi’s (1988) theory of citizen quasi-voluntary
compliance with tax payment in Nigeria. If an individual Nigerian believes government is
fulfilling its end of the bargain, she will be more accepting of government’s authority to tax.
This will also manifest in her actual tax payment. Similarly, if individuals trust other Nigerians
are likely to comply with demands for tax payment, they will also be more willing to acquiesce
to the government’s right to tax. Once again, they will also be more likely to report actually

206

paying taxes.
In order to test these hypotheses, I rely on Nigerian public opinion data from
Afrobarometer Round Four. These data allow me to establish citizen expectations regarding
taxation and representation, especially in the context of government corruption.

Research Design
Dependent Variables
The following selected variables gauge Nigerians’ attitudes and behaviors toward tax
payment. Willingness to pay taxes, the dependent variable in Hypotheses 1 and 2, is
operationalized with an indicator of the extent to which the respondent believes, in principle,
80

people must pay taxes.

The dependent variable in Hypotheses 3 and 4 is captured with three

indicators that measure an individual’s actual tax payment within the last year. This includes
payment of income taxes, property taxes, and local government taxes.

81

These four indicators

result in four separate analyses.
Independent Variables
In Hypotheses 1 and 3, quality of representation is the explanatory variable. To reiterate,
representation is the process through which elected officials gauge, deliberate upon, and
incorporate citizen interests in political decision-making (Pitkin 1967, 1969; Huber and Powell,
1994; Aldrich, 1995; Stokes, 1999). However, in order to practice representation, politicians
80

“The tax department always has the right to make people pay taxes.” 1=Strongly disagree,
2=Disagree, 3=Neither agree nor disagree, 4=Agree, 5=Strongly agree.
81

“Have you had to make any of the following payments during the past year: License fees to
local government e.g., for a bicycle, cart, or market stall? Property rates or taxes? Income taxes?”
0=No, 1=Yes

207

must be first willing to pay attention to citizen interests and preferences. This includes listening
to constituents and remaining in contact with them. When operationalizing citizen perceptions of
the quality of representation they receive from elected officials, I use six indicators. These
measures individually capture respondents’ evaluations of local and national representatives’
ability to assess and act upon citizens’ needs.
First, I include an indicator measuring a Nigerian’s belief in the likelihood that she “could
get together with others and make [their] elected local councilor listen to [their] concerns about a
matter of importance to the community.” I also include a second indicator measuring this
perception, but with respect to the their representative to the National Assembly.

82

I hypothesize

that when Nigerians believe it is somewhat/very likely they can make their local and national
representatives listen to them, they will also indicate being more willing to pay taxes (Hypothesis
1). I also expect a positive relationship between a Nigerian’s perceived ability to make her local
and national representative listen and her actual tax payment (Hypothesis 3). As Levi (1988)
argues, a citizen’s compliance with government demands for taxes is, in part, dependent on her
belief political leaders will reciprocate by incorporating ordinary people’s interests in policy
making. Therefore, if a Nigerian feels they can get together with others and make her local and
national officials listen to her concerns, she will be more likely to perceive government as
fulfilling their end of the tax contract. As a result, they will be more willing to pay taxes.
The second set of explanatory variables also gauges Nigerians’ perceptions of how well
their local and national representatives listen to them. In this case, the question asks the
82

“In your opinion, how likely is it that you could get together with others and make: Your
elected local councilor listen to your concerns about a matter of importance to the community?
Your representative to the National Assembly listen to your concerns about a matter of
importance to the community?” 0=Not at all likely, 1=Not very likely, 2=Somewhat likely,
3=Very likely
208

respondent, in her opinion, “how much of the time [does she] think elected local government
councilors try their best to listen to what people like [her] have to say.” A second question asks
the same about representatives of the National Assembly.

83

While the previous set of questions

ask about a respondent’s general efficacy in making elected officials listen to her, these questions
are focused on the regularity with which the respondent believes her representatives, of their own
volition, spend listening to ordinary people. Again, I hypothesize that when Nigerians believe
their local and national representatives often/always listen to what ordinary people have to say,
they will be more willing to pay taxes (Hypothesis 1). I also expect Nigerians will be more
likely to pay income, property, and local taxes when they believe local and national leaders are
often/always listening to people like them (Hypothesis 3). Following the theory of quasivoluntary compliance, I expect these positive relationships. Nigerians will be more willing to
pay taxes and more compliant with tax payment when they believe political leaders will
reciprocate by including citizens’ preferences in government deliberations. As Nigerians believe
government to be in fulfillment of the tax contract, they will be more willing to yield to the tax
authority. They will also be more likely to report actual tax payment.
The last set of independent variables measures how often a respondent has contacted her
local government councilor in the last year to discuss a problem or a view on an issue. A second
question asks the same about contact with her representative to the National Assembly in the last

83

“How much of the time do you think the following try their best to listen to what people like
you have to say: Representatives to the National Assembly? Elected Local Government
Councilors?” 0=Never 1=Only sometimes, 2=Often, 3=Always

209

year.

84

I expect a positive relationship between the frequency of Nigerians’ contact with their

local and national representatives and their willingness to pay taxes (Hypothesis 1). I also expect
a positive relationship between the frequency of a respondent’s contact with her local and
national leaders and her actual payment of taxes (Hypothesis 3). When a Nigerian believes she
can contact her local and/or national representative to share a problem or view, and then actually
follows through with the action, she is more likely to view these officials as interested in her
problems and concerns. In this case, the respondent will view her leaders as attentive, willing to
help her solve a problem or willing to represent her views on an issue within the local
government or National Assembly. If Nigerians believe this, they will be more likely to view
government as complying with its end of the bargain. Therefore, they will be more willing to
yield to demands for taxes and actually pay those taxes.
In Hypotheses 2 and 4, the explanatory variable is Nigerians’ belief that other individuals
are complying with tax payment. As Levi (1988) argues, this is the second tenet on which an
individual’s quasi-voluntary compliance with tax payment relies. I hypothesize Nigerian
taxpayers operate perform a similar calculation. When they trust other taxpayers to comply, they
will also be willing to acquiesce to government demands for taxes. These Nigerians will also be
likely to report actual tax payment.
Round 4 of the Afrobarometer does not ask a question about respondents’ trust in the
context of tax payment. However, Afrobarometer has two indicators measuring generalized trust

84

“During the past year, how often have you contacted any of the following persons about some
important problem or to give them your views: A Local Government Councillor? A
Representative to the National Assembly?” 0=Never, 1=Only once, 2=A few times, 3=Often
210

of people they know and, more widely, other Nigerians.

85

These two indicators of a

respondent’s trust in people they know and Nigerians, at large, can serve as a proxy for
measuring if respondents trust other taxpayers to comply with payment. It stands to reason that
if a respondent trusts people they know or other Nigerians in a general sense, they will also trust
them to pay their taxes. On the other hand, if a respondent does not generally trust people they
know or other Nigerians, they will not trust them to, specifically, pay taxes.
I hypothesize that when Nigerians express trust of people they know and other Nigerians,
they will also be willing to pay taxes (Hypothesis 2). Additionally, I expect a positive
relationship between respondents’ trust of others (people they know and other Nigerians) and
their actual payment of taxes (Hypothesis 4). Following the theory of quasi-voluntary
compliance, when Nigerians trust other taxpayers to comply with payment, they will be more
willing to cooperate. This will manifest in respondents’ willingness to pay taxes and their
reports of making actual tax payment.

Control Variables
I also include control variables identified by previous literature as predictors of citizens’
willingness to pay taxes and actual tax payment.
Fjeldstad and Semboja (2001), in a survey of tax compliance in Tanzania, find an African’s
ability to pay is a significant predictor of tax compliance. When respondents are relatively better
off, they tend to be more tax compliant. In their analysis, they use the number of wage earners in
the respondent’s household as an indicator. Fjeldstad and Semboja (2001) find respondents in
85

“How much do you trust each of the following types of people: Other people you know (nonfamily members? Other Nigerians?” 0=Not at all, 1=Just a little, 2=I trust them somewhat, 3=I
trust them a lot
211

households with more than one income earner are more likely to pay taxes. On the other hand,
respondents in households with only one wage earner are less likely to be tax compliant. The
Afrobarometer does not specifically ask about employment within a respondent’s household.
86

However, there is a question about the respondent’s employment status.

Therefore, I

hypothesize employed respondents will be willing to pay taxes and report actual tax payment.
Furthermore, I hypothesize Nigerians with full-time employment will be more willing to pay
taxes and report tax payment when compared to respondents with part-time employment.
Previous work on tax compliance in Western countries find younger individuals are less
likely to pay taxes, while the most tax compliant are respondents are over 60 years old (Spicer
and Lundstedt, 1976). However, in Fjeldstand and Semboja’s (2001) analysis of Tanzania, they
find younger respondents are more tax compliant. In Africa, elderly people tend to be
unemployed and live in less affluent households. As a result, older Africans have a lower ability
to pay when compared with younger Africans. I include age as a control variable in my
analysis.

87

Following Fjeldstand and Semboja (2001), I hypothesize a negative relationship

between age and willingness to pay taxes. I also expect a negative relationship between age and
reports of actual tax payment.
Last, I control for respondents’ perception of corruption within the local government,

86

“Do you have a job that pays a cash income? Is it full-time or part-time? And are you
presently looking for a job (even if you are presently working)?” 0=No (not looking), 1=No
(looking), 2=Yes, part time (not looking), 3=Yes, part time (looking), 4=Yes, full time (not
looking), 5=Yes, full time (looking)
87

“How old are you?”

212

National Assembly, and among tax officials.

88

I hypothesize all three types of corruption will

have a negative influence on citizens’ willingness to pay taxes. I also expect a negative
relationship between Nigerians’ perceptions of corruption and their tax compliance. The
justification of this negative expectation is tied to the two tenets on which quasi-voluntary
compliance is based. First, if a respondent believes members of their local government and the
National Assembly are corrupt, she will be less likely to believe government will fulfill its part
of the tax contract. If the respondent does not believe leaders will incorporate citizens’ priorities
in decision-making, she will be less willing to pay taxes. She will also be less likely to report
actually paying her taxes. Second, if Nigerians perceive members of the tax administration as
corrupt, they are also less likely to believe the tax department is pursuing and punishing tax
evaders. If a respondent feels other Nigerians are getting away with noncompliance, she will be
less willing to pay taxes and less likely to report tax payment.
Once again, data for all variables are obtained from Round Four of the Afrobarometer.
Table 5.5 summarizes each of the hypothesized relationships between the dependent,
explanatory, and control variables. Table 5.6 presents each variable’s summary statistics.

Describing the Sample
There are originally 2400 respondents in the Afrobarometer survey in Nigeria. However,
after excluding respondents who answered “don’t know” or refused to answer any questionindicator (dependent, independent, control variables), the sample size drops to 1,641. In order to
ensure the removal of these cases has not introduced bias to the sample, I compare the
88

“How many of the following people do you think are involved in corruption, or haven’t you
heard enough about them to say: Members of Parliament (National Assembly)? Elected Local
Government? Tax officials? 0=None, 1=Some of them, 2=Most of them, 3=All of them”
213

demographic makeup of the original and new samples (Tables 5.7, 5.8, 5.9, 5.10, 5.11). Table
5.7 presents a comparison of means across various demographic variables between the original
and new sample. Respondents in the new sample are, on average, only slightly older (31.5 years
old) than those in the original sample (31.3 years old). Respondents in the new sample, like in
the original, have, on average, completed some/all of secondary school. Both respondents in the
new and original sample are, on average, employed part-time.
Table 5.8 compares the distribution of religious affiliation in the two samples. The newer
sample has about a 1% increase in those who identify as “Christian only” and 1% increase in
those identifying as “Muslim only.” There is a slight 3% decrease in the newer sample of
respondents identifying themselves as “Roman Catholic.” With that said, like in the original
sample, Christians remain the dominant religious group, followed by Muslims, and Roman
Catholics in the new distribution.
When considering the distribution of ethnicities in the two samples, Hausa respondents
increase by about 1%, while Yoruba respondents increase by 2% between the original and new
sample (Table 5.9). Those identifying as Igbo decrease by 1%. However, like in the original
sample, these ethnic groups remain the three most dominant in the new sample. Moreover, like
in the original distribution, Hausas are the most numerous in the sample, followed by Yorubas,
and then Igbos.
In the original sample, most respondents live in large, urban areas (51%), while about
36% live in small, urban areas (Table 5.10). Large, urban dwellers still remain the majority,
though decreasing to 47%. Those living in small, urban areas increase to about 39% of the
sample. Last, gender distribution in the original sample is about a 50-50 split between males and

214

females (Table 5.11). However, the new sample has more male respondents (53% male
compared to 47% female).
Based on these demographic distributions, I argue that the new sample remains similar to
the original. However, the 6% gap between male and female respondents could introduce bias to
the analysis. In order to mitigate this, I add gender as a control variable.

Analysis and Results
For analysis, I rely ordered logistic regression and logistic regression, both with robust
standard errors. Since “willingness to pay taxes” (the first of the dependent variables) is
measured at the ordinal level with five meaningful and sequential categories for response,
ordered logit would provide the best test. The remaining three dependent variables (whether the
respondent paid local, property, or income taxes in the last year) have yes/no responses.
Therefore, these variables are binary, dummies, and I utilize logit for analysis.
Table 5.12 presents results from the four analyses of the effect of citizens’ perceptions of
representation and trust on willingness to pay taxes and on actual tax payment. Since each
question-indicator (dependent and independent variables) is measured on a different scale, I
report the proportional odds ratios (POR). These ratios will allow us to compare the relative
strength of the various predictors in their influence on popular compliance with tax payment.
There is strong and positive evidence that Nigerians are willing to pay taxes when they
perceive their elected officials as representing their interests. A higher quality of perceived
representation among Nigerians also corresponds with higher rate of tax payment. Additionally,
Nigerians with higher levels of trust are more likely to report greater willingness to pay taxes and
actual tax payment. These findings directly support the hypotheses outlined earlier in this

215

chapter. Perceived quality of citizen representation and trust in government are significant
predictors across multiple measures of Nigerians’ willingness to pay taxes and actual tax
payment. Moreover, as hypothesized, the theory of quasi-voluntary compliance provides
leverage in analyzing Nigerian attitudes and behaviors toward taxation.
When considering the conditions under which Nigerians are most likely to yield to the
authority of the tax department, as hypothesized, the ability to make local and national
representatives listen are among the strongest predictors (Table 5.12). As a Nigerian believes
she (and other people like them) can get together and share concerns with their local councilor,
she is more likely to concede to the government’s power to tax (POR = 1.278). In fact, of all
variables included, the ability to make one’s local representative listen exerts the strongest
influence. This influence is positive, as hypothesized. Similarly, when a Nigerian believes she
can get together with others and make a national representative listen, she is more willing to pay
taxes (POR = 0.998). Once again, this relationship is strong and positive, as hypothesized.
The time a Nigerian perceives local and national representatives spend listening to ordinary
people also exerts a positive influence on willingness to pay taxes. As Nigerians believe their
local councillor (POR = 1.191) and representative to the National Assembly (POR = 0.862)
spend more time listening to what ordinary citizens have to say, they are also more likely to
agree to the tax department’s right to compel tax payment. Moreover, when a Nigerian indicates
making contact with a local councilor to discuss a problem or share a view, she is more willing
to pay taxes (POR = 1.064). When a Nigerian indicates contact with a national representative in
the last year, she is also more willing to concede to the tax department’s authority (POR =
0.877). These findings support Hypothesis 1, demonstrating Nigerians are more willing to yield
to the government’s authority to tax as they perceive higher quality of representation.

216

Evidence also supports Hypothesis 3, and as Nigerians indicate they trust others, they are
also more willing to pay taxes. Quasi-voluntary compliance specifically identifies trust in
others’ compliance with tax payment as a tenet. However, the use of generalized trust as a proxy
bears out empirically. Evidence suggests Nigerians’ trust in other Nigerians (not other people
they know) significantly, and positively, influences their willingness to pay taxes (POR = 1.101).
When contemplating issues of taxation and tax compliance, it seems Nigerians use a larger
reference group. Rather than focusing on a smaller group on individuals they know, a Nigerian
thinks about the larger population and if they trust other Nigerians, at large. In accordance with
quasi-voluntary compliance, those indicating more trust in Nigerians are also more likely to yield
to government’s power to tax. Even when taking ability to pay (employment status), age, and
corruption into account, quality of representation and trust are more powerful and significant
explanatory variables. For example, perceptions of corruption in the local government (POR = 0.856), National Assembly (POR = -1.115), and the tax administration (POR = -1.135) have a
significant, negative influence on Nigerians’ willingness to comply with tax payment. In fact,
opinions about tax officials’ corruption are the third most powerful predictor of popular beliefs
on the government’s authority to tax. As hypothesized, when Nigerians believe local, national,
and tax officials are engaged in corrupt practices, they are less willing to comply with demands
for tax payment. In this instance, ordinary citizens are less likely to believe government will
fulfill its end of the fiscal contract or punish evaders.
Turning to predicting actual tax payment (Hypotheses 2 and 4), quality of representation
and trust also exert a strong, positive influence. As hypothesized, a Nigerian is more likely to
report paying local taxes when she also believes she is able to make her local councillor listen to
community concerns (POR = 1.191). Similarly, when Nigerians believe ordinary people can

217

make local representatives listen, they are more likely to pay property taxes (POR = 1.264) and
income taxes (POR = 1.087). In fact, the ability to make their local councillor listen is among
the most powerful predictors in explaining each of these forms of tax payment. For example,
Holding all other variables constant, a one unit increase in a Nigerian’s belief that she can make
her local representative listen (e.g. moving from “not very likely” to “somewhat likely”),
increases the likelihood that she will have paid local taxes by 1.191 times. This one-unit
movement also increases the likelihood of property tax payment by 1.264 times and income tax
payment by 1.087 times.
When a Nigerian believes her local councillor spends more time listening to what ordinary
people have to say, she is more likely to have paid local taxes (POR = 0.986), property taxes
(POR = 1.031), and income taxes (POR = 1.133) in the last year. Increased contact with a local
representative is a positive and significant explanatory variable. If Nigerians have contacted
their councillor with a problem or to discuss a view on an issue, they are more likely to report
payment of local taxes (POR = 1.299), property taxes (POR = 1.263), and income taxes (POR =
1.283). Making contact with one’s local councillor is the most powerful predictor of Nigerians’
local tax and income tax payment.
In the case of actual tax payment, it seems Nigerians evaluate representation from their
local officials more heavily than the quality of representation received from their national
leaders. Scholars argue citizens feel closer to their local governments (i.e. Tiebout, 1956 and
Oates, 1972). So, theoretically, Nigerians might consider both local and national representation
when determining their willingness to pay taxes. However, when it comes to actual tax payment,
local government representation is the key explanatory variable. In addition, most Nigerians pay
income, property, and local taxes directly to their state and local government. It would make

218

sense for them to reference quality of representation from their local government when
considering tax compliance.
Nigerians’ level of trust also positively predicts their likelihood of actual tax payment
(Hypothesis 4). When a respondent indicates a high level of trust in other Nigerians, she is also
more likely to report paying local taxes (POR = 1.097), property taxes (POR = 1.071), and
income taxes (POR = 1.049) in the last year. Once again, this indicator of general trust serves as
a proxy, but still provides support for the second tenet in quasi-voluntary compliance. When
Nigerians trust that other taxpayers are complying with payment, they will also be more likely to
comply.
As hypothesized, corruption in government exerts a negative influence on actual tax
payment in the last year. Specifically, perceptions of corruption in local government have a
significant, negative relationship with payment of local government taxes (POR = -1.150),
property taxes (POR = -1.072), and income taxes (POR = -1.075). Beliefs about corruption in
the tax administration also exert a negative influence on reported payment of taxes. Perceptions
on corruption among tax officials are the third most powerful predictor of income tax payment
(POR = -1.117). Holding other variables constant, a one-unit increase in a Nigerian’s perception
that tax officials are engage in corruption (e.g. “some of them” to “most of them”) decreases the
likelihood that she will have paid income taxes by 1.117 times. As hypothesized, when
Nigerians believe local and tax officials are participating in corrupt practices, they do not think
government will satisfy its side of fiscal contract. However, even in this context, quasi-voluntary
compliance provides explanatory leverage. These four models, taken together, offer strong
evidence that, like in Western Europe, quasi-voluntary compliance and its implications, apply in
Africa. Nigerians expect representation of their interests in exchange for their tax payment.

219

Moreover, Nigerians’ willingness to pay taxes and actual tax payment is directly related to their
perceived level of trust in other taxpayers to comply. Evidence suggests perceptions of
corruption within the local government, National Assembly, and among tax officials have a
significant, negative influence on citizens’ attitudes and behavior toward taxation. With that
said, the hypothesized relationship between tax payment, quality of citizen representation, and
trust remains positive and significant.

Expanding the Concept of Representation: Citizen Satisfaction With Government
Performance
This analysis using Afrobarometer public opinion data supports Levi’s (1988) assertion
about the linkages between representation, trust, and citizen compliance with government
demands for taxes. In exchange for tax payment, citizens expect government to pay attention to
their concerns and incorporate citizen interest in public policy. However, scholars have
expanded substantive representation to include public expenditures on service delivery.
Following Levi (1988), Fjelstad and Semboja (2001: 2061) argue: “Individuals pay taxes
because they value the goods provided by the government, recognizing that their payments are
necessary both to help finance the goods and services and to get others to contribute.” Therefore,
satisfaction with these services is likely to increase the probability that taxpayers will voluntarily
comply with requests for tax payment. Fjeldstad (2001) and Fjeldstad and Semboja (2001)
investigate this hypothesis, using Tanzania and South Africa as their cases. They find Africans’
compliance with actual tax payment is positively linked to their satisfaction with government
performance in the arena of public service delivery. As Africans perceive higher levels of
service provision, they also indicate a higher propensity to pay taxes. When public service

220

delivery improves in quality and provision becomes more equitable, Africans perceive the terms
of trade with government as fair exchange, and will consequently be more likely to comply with
tax payment.
I now test this hypothesis in Nigeria. While Fjeldstad and Semboja (2001) and Fjeldstad
(2004) focus on actual tax payment as their dependent variable, I follow the previous analysis. I
investigate the influence of satisfaction with public service delivery on Nigerians’ willingness to
pay taxes (attitude) and actual tax payment (behavior). In circumstances where citizens are more
satisfied with government performance, we should observe a greater willingness to contribute to
government via tax payment (Hypothesis 1). Conversely, when citizens are less satisfied with
government service provision, they are also less willing to pay taxes. We should also discern a
higher rate of actual tax payment among satisfied individuals (Hypothesis 2). In addition,
unsatisfied Nigerians will be less likely to comply with tax payment. Like Levi (1988), Fjeldstad
(2001), and Fjeldstad and Semboja (2001), I assert that Nigerians perceive taxation and tax
payment as a contract with government. Individuals pay taxes in expectation that government
will provide representation via expending resources to deliver public services. When Nigerian
citizens believe their elected officials are fulfilling their end of the bargain, in this case by
providing services, they will be more willing to comply with government demands for taxes.

Research Design: Operationalizing Satisfaction with Service Provision
For this analysis, I operationalize Nigerians’ satisfaction with government provision of
public services with several indicators of respondents’ opinions about how well the current
government is managing certain matters. The Afrobarometer asks this question regarding

221

several areas in a general sense, for example, managing the economy, improving conditions for
poor citizens, and job creation. With these larger policy areas, respondents may consider them
more abstractly, providing a general opinion. Rather, I incorporate Nigerians’ opinions about
specific services, allowing me hone in on those with which respondents have daily experience. It
is more likely Nigerians use their everyday experiences with these more specified services to
gauge satisfaction with government performance. As a result, I include the following measures
of citizens’ perceptions of how well the current government is managing basic health services,
89

water and sanitation, roads and bridges, and supplying electricity.

I also incorporate indicators

for satisfaction with (specifically) local government management of local roads and local
markets.

90

I hypothesize Nigerians will be willing to comply with tax payment as they are more

satisfied with how the current/local government handles the aforementioned services (Hypothesis
1). I also hypothesize respondents with higher levels of satisfaction with government
performance in these areas will be more likely to report actual payment of income, property, and
local taxes (Hypothesis 2).
This analysis builds on the initial four models, using the same indicators for willingness to
pay taxes and reported payment of income, property, and local taxes in the last year. In the new
model, the indicators for quality of representation and trust (the dependent variables in the first
89

“How well or badly would you say the current government is handling the following matters,
or haven’t you heard enough to say: Improving basic health services? Providing water and
sanitation services? Maintaining roads and bridges? Providing a reliable supply of electricity?”
1=Very badly, 2=Fairly badly, 3=Fairly well, 4=Very well
90

“What about local government? I do not mean the Federal government or State government. I
mean your local government council. How well or badly would you say your local government is
handling the following matters, or haven’t you heard enough about them to say: Maintaining
local roads? Maintaining local market places?” 1=Very badly, 2=Fairly badly, 3=Fairly well,
4=Very well

222

model) will serve as controls. I also still include controls for employment status; respondent age;
perceptions of corruption in the local government, National Assembly, and tax administration;
and gender.
Table 5.13 presents summary statistics for the variables included in the second analysis.
Once again, I exclude “don’t know” responses and respondents who refused to answer. This
results in a sample size of n = 1596 (reduced by 45 from the first analysis). Like in the first
analysis, I utilize ordered logit and logistic regressions with robust standard errors.

Analysis and Results
Turning to the results of analysis, evidence confirms that Nigerians’ attitudes and
behaviors toward taxation are positively influenced by their perceptions of government
performance (Table 5.14). Once again, the proportional odds ratios (POR) are reported. A
Nigerian is more likely to concede to the authority of the tax department to compel payment
when she is more satisfied with how the government is managing service provision. In addition,
Nigerians with higher levels of satisfaction with government service management are more likely
to report actual payment of local, property, and income taxes. These findings suggest quasivoluntary compliance has a strong performance component: part of taxpayers’ assessment
whether government is fulfilling its end of the bargain is based upon individual evaluation of
public service delivery. Furthermore, while the government performance variables are strong
and positive, the quality of representation and trust variables also remain strong and positive.
Ultimately, quasi-voluntary compliance remains an applicable theory in Nigeria.
As hypothesized (Hypothesis 1), a Nigerian is more willing to yield to tax department’s
authority to tax when she is more satisfied with how the government is handling health service

223

provision (POR = 0.972). Similarly, at higher levels of satisfaction with government
management of water and sanitation (POR = 1.122) and local government management of (POR
= 1.063), Nigerians are also more willing to pay taxes. However, the performance-based
variables with the strongest explanatory power are government management of electricity (POR
= 1.256), maintenance of roads and bridges (POR = 1.142), and management of local roads
(POR = 1.168). For example, a one-unit shift in a Nigerian’s satisfaction with government
management of electricity services (e.g. moving from “fairly badly” to “fairly well” increases the
likelihood that she will be willing to comply with tax payment by 1.256 times. Overall, even
controlling for corruption, a Nigerian is more willing to pay taxes when she is more satisfied
with government performance, more satisfied with the quality of representation she receives
from her local and national leaders, and when she trusts other Nigerians are complying with tax
payment.
Analysis also confirms Hypothesis 2: when Nigerians are satisfied with how the
government is handing service provision, they are also more likely to report tax payment in the
last year. Looking at determinants of local tax payment, satisfaction with government
management of health services (POR = 1.118), water and sanitation services (POR = 1.052), and
local government maintenance of markets (POR = 1.009) are all positive predictors. The most
powerful explanations of why a Nigerian will pay her local taxes (among the government
performance variables) are her satisfaction with electricity supply (POR = 1.247), government
maintenance of roads and bridges (POR = 1.168), and her satisfaction with local government
maintenance of local roads (POR = 1.183). Once again, respondents’ satisfaction with the
quality of representation they receive from their local councillor and their trust of other Nigerians
also positively influence their payment of local taxes. In accordance with quasi-voluntary

224

compliance, Nigerians are more likely to pay their local taxes when they believe government is
fulfilling its side of the fiscal contract, providing quality representation and services. In addition,
a Nigerian who trusts other Nigerians are paying their taxes is also more likely to pay her local
taxes.
Property tax payment is also significantly and positively explained by Nigerians’
satisfaction with government performance. Respondents are more likely to have paid property
taxes in the last year when they are satisfied with government management of health services
(POR = 0.994), water and sanitation services (POR = 1.197), road and bridge maintenance (POR
= 1.247), and electricity supply (POR = 1.292). Nigerians are also more likely to pay property
taxes when they are satisfied with local government maintenance and local roads (POR = 1.058)
and markets (POR = 0.858). Even taking into account a respondent’s satisfaction with the
quality of representation from her local leaders and her trust in other Nigerians (both positive and
significant explanatory variables), her satisfaction with government provision of electricity is the
strongest predictor of her property tax payment in the last year. This model confirms
government performance matters in individuals’ calculations of whether or not to follow through
with tax payment.
Finally, income tax payment follows a similar pattern: Nigerians are more likely to report
payment in the last year when they are satisfied with public service provision. When it comes to
payment of income taxes, Nigerians’ satisfaction with how government maintains the electricity
supply (POR = 1.349), roads and bridges (POR = 1.358), and local roads (POR = 1.349) are
more significant than their perceptions of corruption, quality of representation, and trust.

225

Results of these four analyses confirm my hypotheses and reinforce the initial findings
about quasi-voluntary compliance and tax payment. This theory, first utilized to explain
attitudes and behaviors toward taxation in Western Europe, also applies in the African context.
Nigerians are, clearly, like tax payers in other nations, viewing taxation as a bargain with
government. Individuals require certain expectations to be met in exchange for their tax
payment. Like taxpayers in other contexts, Nigerians must believe government will fulfill their
terms of the trade: they expect substantive representation of their interests and public services.
Thus, when a Nigerian is more satisfied with the quality of representation she obtains from her
local and national leaders she will be more willing to pay taxes and comply with actual tax
payment. It also follows that when this taxpayer is more satisfied with how government is
managing public service provision, her attitude and behavior toward tax payment will be more
compliant. Evaluations of government performance are a part of quasi-voluntary compliance’s
first tenet. And last, when a Nigerian taxpayer trusts other individuals are paying taxes, she will
be more willing to pay taxes and more likely to report tax payment. These three major findings
demonstrate the applicability of theories of revenue and representation in Africa.

Dimensions of Citizen Representation and Satisfaction
In the previous chapter, factor analysis uncovered a latent “elite representation” variable
with two dimensions. When Nigerian legislators think about the representative function, they
think of prioritizing constituency interests/needs and also how effective their state legislature is
in representing various constituencies. In this section, I explore the possibility that there is a
single dimension of quality of representation among citizens. Is there a single construct of

226

representation when ordinary Nigerians evaluate their elected officials? Similarly, when
thinking about government performance with public service provision, is there a single
dimension along which citizens rate officials? Once again, I rely on factor analysis to group
variables and identify essential components of representation and satisfaction among citizens.
Factor analysis identifies one dimension of citizens’ perceptions of representation (Tables
5.15 and 5.16). All six variables help to define this factor: ability to make local councilors listen
to citizen concerns (loading = 0.601); ability to make national representatives listen to citizen
concerns (loading = 0.621); perceptions of the amount of time local officials spend listening to
ordinary people (loading = 0.657); perceptions that national officials spend time listening to
ordinary people (loading = 0.644); the amount of contact respondents had with local
representatives (loading = 0.549); the amount of contact respondents had national representatives
in the last year (loading = 0.496). This factor explains about 47% of the observed variance.
Moreover, a test of internal consistency results in a Cronbach’s alpha of 0.74. This suggests that
these six variables move together and can be combined into reliable index.
I perform a factor analysis test with the six variables previously used to measure Nigerians’
satisfaction with government service provision (Tables 5.17 and 5.18). Similarly, analysis
confirms one dimension in this group of variables. Essentially, citizens’ satisfaction is defined
by all six variables: satisfaction with government health services (loading = 0.559); satisfaction
with government water and sanitation services (loading = 0.607); satisfaction with government
maintenance of roads and bridges (loading = 0.544); satisfaction with government electricity
services (loading = 0.552); satisfaction with local government maintenance of local roads
(loading = 0.707); satisfaction with local government maintenance of local markets (loading =
0.699). This factor explains 61% of observed variance, and a Cronbach’s alpha of 0.80 suggests

227

that the six variables can be assembled into an index.
Based on these results, I produce two indices. The first captures the six variables
characterizing one dimension of citizen perceptions of representation. The second index
encompasses the other six variables that signify Nigerians’ satisfaction with government
performance. Using these two indices as independent variables, I conduct an OLS regression
(with robust standard errors). I will again investigate the influence of citizen beliefs about
representation and service provision on their willingness to pay taxes and actual tax payment. In
this analysis, I hypothesize that Nigerians who perceive a higher level of representation from
their elected officials (captured by the citizen representation index) will be more willing to pay
taxes. These individuals will also report higher rates of actual tax payment. I also hypothesize
that Nigerians who are more satisfied with government provision of public services (captured by
the government performance index) will be more willing to comply with tax payment and follow
through with these payments.
Table 5.19 reports the results of this analysis and supports these two hypotheses. Nigerians
who perceive higher levels of representation from their elected officials are more willing to pay
taxes (beta = 0.107). These citizens are also more likely to report making local tax payments
(beta = 0.092), property tax payments (beta = 0.143), and income tax payments (beta = 0.114) in
the last year. Furthermore, the index of citizen representation is the strongest predictor of
Nigerians’ willingness to comply with tax payment and actual payment of property and income
taxes. In terms of payment of local taxes, citizen perceptions of representation are the second
most powerful predictor (next to individuals’ employment status, beta = 0.146).
Similarly, the index of citizen satisfaction with government performance also shares a
significant, positive relationship with Nigerians’ compliance with tax payment. Respondents

228

who are satisfied with how government is managing the provision of services are more willing to
pay taxes (beta = 0.085). Satisfied Nigerians also report paying local taxes (beta = 0.077),
property taxes (beta = 0.107), and income taxes (beta = 0.098) in the last year.
This analysis first establishes that citizen perceptions of the quality of representation from
elected officials center around one dimension. Nigerian perspectives of how government
manages various public services also characterize a single factor. Furthermore, after creating
two indices, evidence suggests that “citizen representation” and “citizen satisfaction” help to
explain the conditions under which Nigerians are most willing to pay taxes and follow through
with actual tax payment. Individuals who observe higher levels of representation from their
elected officials and who believe government is effective in handling service provision are also
the most compliant with tax payment.

Robustness Check: Cross-Examining Government Performance and Willingness to Pay Taxes
In order to crosscheck the role of government performance in Nigerians’ quasi-voluntary
compliance, I retest these hypotheses utilizing another public opinion dataset. In December 2011,
an international consortium of researchers conducted a public opinion survey of 2,750 Nigerian
91

citizens in 11 cities across the country.

Topics of interest included perceptions of identity,

security, personal and public finance, and government performance.

92

Unlike the Afrobarometer

91

Collaboration between Drs. Adrienne LeBas (American University), Etannibi Alemika
(University of Jos), Nic Cheeseman (University of Oxford), Olufunmbi Elemo (Graduate Student,
Michigan State University; funded by the Improving Institutions for Growth (iiG) DFID
Research Programme Consortium.
92

Cities included: Aba (in Abia State), Bauchi (in Bauchi State), Enugu (in Enugu State), Ibadan
(in Oyo State), Jos (in Plateau State), Kaduna (in Kaduna State), Kano (in Kano State), Lagos (in
229

Public Opinion Survey, which focuses on both urban and rural areas, this new survey focuses
exclusively on cities. Urban centers are considered particularly interesting and worth singular
focus. This is the especially the case when discussing individual perceptions of government
performance and finance. Urban dwellers are more likely to pay taxes, have direct contact with
government, and have expectations of public service delivery (Bates, 1981/1983; Herbst, 2000).
Using this new dataset, I once again investigate the conditions under which Nigerians are
most willing to pay taxes.

Research Design: Dependent, Independent, and Control Variables
Individual attitudes and behavior remain the key dependent variables. This new survey
does not include questions asking the respondents about specific payment of taxes. For that
reason, “willingness to pay taxes” will serve as the key dependent variable.

93

Specifically, the

survey asks respondents to compare their current State government with the last administration:
“Given the current government’s performance that you have seen, do you find yourself more
willing or less willing to pay taxes than you were during the last civilian administration?”
Turning to Figure 5.5, we see 45% of urban Nigerians describe themselves as less willing to pay
taxes now than in the past. On the other hand, 55% of respondents say they feel the same about
paying taxes or are more willing to pay now than in the past. With that said, across states, there
Lagos State), Lafia (in Nasarawa State), Onitsha (in Anambra State), and Sokoto (in Sokoto
State).
93

“Given the current government’s performance that you have seen, do you find yourself more
willing or less willing to pay taxes than you were during the last civilian administration in _____
State? Or would you say your attitude toward taxes is about the same?” 1=Less Willing,
2=About the Same, 3=More Willing
230

is variation in urban Nigerians’ willingness to pay taxes (Figure 5.6). For example, Lagos and
Plateau States have the highest percentage of city-dwellers who are more willing to pay taxes
now than in the past. In Lagos, 45% of urban Nigerians indicate greater willingness, while in
Plateau, 40% of respondents are more willing to pay now. Among other states, there is a
significant drop in the percentage of individuals more willing to pay taxes now than under the
past administration. The state with the third highest percentage of urban Nigerians who are more
willing to pay taxes is Kaduna (22%). Of the states included in the survey, Aba has the smallest
percentage of city-dwellers who are more willing to pay taxes (5%).
Following previous analyses, I hypothesize that variation in urban Nigerians’ willingness
to pay taxes is determined, in part, by their belief government will fulfill its end of the fiscal
contract. Additionally, trust that other Nigerians will comply with tax payment also influences
willingness to pay taxes. This new survey does not ask about respondents’ trust (neither
generally, nor specifically in terms of taxation). However, there are a series of indicators
gauging respondents’ satisfaction with government performance and service delivery. So, while
this data set will not allow me to test both tenets of quasi-voluntary compliance, I will be able to
cross-examine how citizen perceptions of government performance influence attitudes toward
taxation. Moreover, since the Afrobarometer focuses on local and national government
performance, this chapter’s previous analyses provide evidence that local and national
performance factor into citizens’ calculations about their tax compliance. However, this new
survey specifically asks urban Nigerians’ about their evaluation of state government. Now, I
will be able to investigate whether Nigerians also consider their state government’s performance
when determining their attitudes toward taxation.

231

Satisfaction with government performance (main explanatory variable) is operationalized
using five individual indicators. First, respondents are asked how well (or badly) they think the
current state administration is spending tax revenue.

94

I hypothesize that when Nigerians are

more satisfied with how their state government is spending tax revenue, they will also be more
willing to pay taxes. Next, respondents are asked to compare their current state government
administration with the past one, and assess the conditions of the roads, public schools,
95

government hospitals and clinics, and the water supply.

I expect that when a Nigerian believes

the state government is managing road maintenance better than the previous administration, she
will also be more willing to pay taxes. I also hypothesize that Nigerians who favorably evaluate
state government management of public schooling, government hospitals/clinics, and supplying
water will also be more willing to pay taxes. A Nigerian who is more satisfied with how her
state government handles provision of these services will view her government as following
through with the terms of the tax contract. As a result, she will be more willing comply with tax
payment.
As for control variables, I use the respondent’s (monthly) household income as a measure
96

of her ability to pay taxes.

I expect urban Nigerians with higher household incomes will have

94

“Here in _____ State, how satisfied are you with the way that the current administration has
spent tax revenue?” 1=Not at all satisfied, 2=Not very satisfied, 3=Somewhat satisfied, 4=Very
satisfied
95

“Compared with the past administration in ______ State, have the following things in this area
gotten better, worse, or have they stayed about the same? Condition of roads? Condition of
public schools? Condition of government hospitals and clinics? Condition of water supply?”
1=Worse, 2=About the same, 3=Better
96

“Please, can you estimate the monthly income for your whole household?” 1=I don’t earn an
income myself—other people support me, 2=Less than N5,000, 3=N5,000-N10,000, 4=N11,000232

a greater ability to pay taxes and will be more willing to pay taxes. I include respondent age as a
control variable again, hypothesizing that older urban Nigerians will be less wiling to pay taxes.
Last, I control for corruption, reasoning that individuals who believe their government to be
corrupt are less likely to believe government will fulfill their terms of the tax contract. Nigerians
who view the government as corrupt are also less likely to believe tax evaders are being caught
and punished. As a result, they will be less willing to pay taxes. In the new survey, respondents’
experiences with government corruption will serve as the indicator. The survey asks
respondents: “during the past 12 months, did any government official ask you or expect you to
pay a bribe for his/her service?”

97

I expect that respondents who report having this experience

with corruption will be less likely to pay taxes.
Table 5.20 presents each variable’s summary statistics. Like in the previous analyses, I
have excluded respondents who respond “don’t know” or refuse to answer the any of the
questions I use as an indicator (dependent, independent, control variables). Therefore, sample
size drops from 2,750 to 1,638. Once again, to ensure the exclusion of respondents has not
introduced bias into the analysis, I compare various demographic distributions in the original and
new sample (Tables 5.21, 5.22, 5.23, 5.24, 5.25, 5.26).
Referring to Table 5.21, we see in the original sample, most respondents’ report their
household income as either between N16,000 and N25,000 (15%), between N26,000 and
N15,000, 5=N16,000-N25,000, 6=N26,000-N35,000, 7=N36,000-N45,000, 8=N46,000N55,000, 9=N56,000-N75,000, 10=N75,000-N95,000, 11=N96,000-N120,000, 12=More than
N120,000
97

“In some areas, there is a problem of corruption among government or public officials. During
the past 12 months, did any government official, for instance a customs officer, police officer,
traffic officer, court official, pensions officer or building inspector ask you or expect you to pay a
bribe for his/her service?” 1=No, 2=Yes
233

N35,000 (17%), or between N36,000 and N45,000 (14%). In the new sample, these three
categories remain the most frequently reported. Turning to respondents’ educational attainment
(Table 5.22), 66% of urban Nigerians have completed secondary school. In the new sample, this
percentage is slightly higher (67%). However, this difference is within the margin of sampling
error. In the original sample, respondents most frequently identify as Catholic (17%), Christian
(12%), or Muslim (10%) (Table 5.23). These three remain religious identities remain the three
most reported, even maintaining the same distribution (with the exception of “Christian,” which
drops by one percentage point, within the margin of sampling error). Hausa (29%), Igbo (32%),
and Yoruba (17%) ethnic groups make up the majority of the original sample (Table 5.24). This
is still this case in the new sample; however, the percentage of Hausa respondents slightly
decreases to 28%. The Igbo distribution increases by one percentage point and those identifying
as Yoruba increase by two percentage points (Once again, both of these changes are within the
sampling error). Table 5.25 describes the age distribution in the original and new sample. In
both, most respondents report being between 25 and 34 years old. And last, while the gender
distribution is evenly split between male and female respondents (50-50) in the original sample,
the new sample is made up of 54% males and 46% females (Table 5.26). Given this 8% gap, I
add gender as a control variable to mitigate any possible bias. Following with the previous
analyses, I use ordered logistic regression analysis with robust standard errors.

Analysis and Results
Table 5.27 presents results of the analysis investigating the relationship between
willingness to pay taxes and perceptions of government performance. The proportional odds

234

ratios are reported. As hypothesized, Nigerians’ satisfaction with their state government’s
performance and service provision has significant and positive relationships to their willingness
to pay taxes. Evidence confirms that when an urban Nigerian is satisfied with how the state
government is managing tax revenue, she is more willing to pay taxes. In addition, across
various measures of satisfaction with government service provision, when this respondent more
favorably evaluates how her state government is handling the delivery, she will also be more
compliant with tax payment.
Urban Nigerians who are more satisfied with how their state government is managing tax
revenue are also more willing to pay taxes (POR = 1.000). Similarly, Nigerians who give more
favorable reports of their state government’s road maintenance are also more willing to comply
98

with tax payment (POR = 1.461).

A one unit shift in a Nigerian’s perception of the current

government’s management of roads (e.g. moving from “about the same as the last
administration” to “better”) increases the likelihood that she will be more willing to pay taxes by
1.461 times. In fact, of all the variables included, satisfaction with how government is managing
roads is the strongest predictor of urban Nigerians’ willingness to pay taxes.
Urban Nigerians who believe their current state government is managing public schools
better than the previous administration are more willing to pay taxes (POR = 1.246).
98

In this analysis of urban Nigerians’ attitudes, satisfaction with road maintenance is the second
most powerful predictor of willingness to pay taxes. Similarly, in the Afrobarometer analysis in
the previous section, satisfaction with maintenance of roads, bridges, and local roads have strong
positive relationships with Nigerian’s willingness to pay taxes and actual tax payment (Table
5.14). The strength of this relationship is only surpassed by individuals’ satisfaction with public
provision of electricity. Evidence suggests that access to good roads (perhaps for the purpose of
travelling to market/work or transporting goods/wares to market in rural and urban areas) and
electricity serve as a sort of litmus test for tax payment. Moreover, evidence suggests that when
Nigerians across urban and rural settings are satisfied with government maintenance of roads,
they are also willing to pay taxes and more likely to follow through with actual payment. This
relationship holds across multiple indicators and two datasets.
235

Respondents’ satisfaction with their state government’s management of public hospitals and
clinics is also positively linked to their willingness to pay taxes (POR = 1.413). Finally, when
urban Nigerians are more satisfied with the current government’s management of water supple,
they are also more willing to comply with tax payment (POR = 1.270).
As hypothesized in the literature, a respondent with more ability to pay (higher household
income) are most more willing to pay taxes. However, contrary to Fjeldstad and Semboja
(2001), age is significant and positively linked to Nigerians’ willingness to pay taxes ((POR =
1.114). Older Nigerians seem to be more willing to comply with tax payment, which is more in
line with previous literature on Western attitudes toward taxation (Spicer and Lundstedt, 1976).
While corruption’s negative influence on tax compliance is as expected, the relationship is not
significant. This may be a result of using a measure gauging experience with corruption rather
than perception (as in the previous analyses using Afrobarometer data). For example, only 30%
of urban Nigerians indicate that, in the last year, they have been asked by various government
officials to pay a bribe for services. It seems that many more urban Nigerians have not had a
direct experience with government corruption. Thus, this variable may not influence their
willingness to pay taxes. Yet, ultimately, the government performance variables exert the
strongest (and positive) influence on taxpayers’ attitudes toward tax payment.
This analysis further corroborates previous findings: Nigerians clearly have expectations
of their local, state, and national governments before making tax payment. Taxpayers assess
their leaders and the quality of their performance, be it substantive representation of their
interests or provision of public services. When Nigerians determine that their government (at all
levels) is in compliance with its end of the bargain, they will also be more willing to pay taxes.
Although the new dataset does not allow me to test quasi-voluntary compliance’s trust

236

component, the previous two analyses using Afrobarometer data confirm Nigerians who trust
that other Nigerians are paying taxes are also more willing to pay taxes. The Afrobarometer
analyses provide evidence that quasi-voluntary compliance explains attitudes toward taxation
and actual tax payment. The analysis using the urban Nigeria dataset checks previous findings
and corroborates the results. As hypothesized, when considering tax payment (both theoretically
and actual payment), Nigerians use their assessments of government performance and trust to
determine their attitudes and beliefs. At higher levels of satisfaction with local and national
representation, Nigerians are more willing to pay taxes and more likely to follow through with
tax payment. In addition, when Nigerians are more satisfied with local, state, and national public
service delivery, they are also more willing to comply with tax payment. Last, when Nigerians
trust other Nigerians are complying with tax payment, they are more willing to pay taxes and
actually pay.

Conclusion
In sum, I have demonstrated the conditions under which citizens are most willing to pay
taxes. Nigerians clearly understand their relationship to government as a contract: they
undertake tax payment in expectation of representation of their interests in public policy and
expenditures. Now understanding this, I characterize the exchange between leaders and citizens
as a fiscal contract. Nigerian governments that generate more income from non-oil taxes are
more likely to supply representation, incorporating their citizens’ preferences into policy and via
service provision. And Nigerian citizens are more willing to pay taxes to government, as they
perceive elected officials are meeting demands for representation of public interests and service
delivery. Given these two processes inform one another, it important to rule out the possible

237

problem of endogeneity. In other words, are governments more representative as a result of
taxing citizens, or, are representative governments better able to extract revenue from citizens via
taxation?
I argue that political leaders in Nigeria need access to revenue; however, in this context,
income can be obtained via taxes on citizens and petroleum wealth. With that said, access to
revenue from natural resources is not guaranteed. As international oil prices fluctuate and
decline and violence in the Niger Delta (location of Nigeria’s oil producing states) increased,
Nigeria’s continued access petroleum revenue is unpredictable. Therefore, (subnational) states
have begun cultivating internal revenue via taxation.
In the next chapter, I explore the reasons why elected officials and members of Nigeria’s
tax administration identify non-oil taxes as the optimal revenue source for development. I also
outline how the movement from oil dependence to increased state tax capacity is an elite-driven
process, propelled by elected officials’ political initiative. Political elites make the first move.
Because they want access to taxes from citizens, elected officials engage in representative
behaviors and engage citizens in a public dialogue. In this bargain, elected officials make certain
guarantees to citizens, assuring them that government will use tax revenue to meet public
interests and needs. Through a comparison of Lagos and Kano States, I investigate why some
states have been better at generating tax revenue than others. Ultimately, political leadership and
initiative makes a difference and influences the nature of tax compliance among citizens.

238

Chapter Five Tables and Figures

Table 5.1:
CROSSTABULATION,
WILLINGNESS TO PAY TAXES
AND BELIEF IN EXECUTIVE
JUSTIFICATION HOW TAX
INCOME IS SPENT (Row %)
TAX DEPARTMENT
HAS A RIGHT TO
MAKE PEOPLE PAY
TAXES
PRESIDENT MUST JUSTIFY
SPENDING (1) VS. PRESIDENT
IS FREE TO ACT ON OWN (2)
AGREE VERY STRONGLY
WITH 1
AGREE WITH 1
AGREE WITH NEITHER
AGREE WITH 2
AGREE VERY STRONGLY
WITH 2
CRAMER’S V

STRONGLY
DISAGREE

DISAGREE

NEITHER AGREE NOR
DISAGREE

AGREE

STRONGLY
AGREE

6

12

11

42

28

5
6
3

23
16
17

13
18
19

44
36
45

15
25
17

4

8

12

47

29

0.113

239

Table 5.2: CHI SQUARE
TEST, DIFFERENCES IN
RATES OF INCOME TAX
PAYMENT (%)
PAID IN LAST YEAR
DID NOT PAY IN LAST
YEAR

Table 5.3: CHI SQUARE
TEST, DIFFERENCES IN
RATES OF PROPERTY TAX
PAYMENT (%)
PAID IN LAST YEAR
DID NOT PAY IN LAST
YEAR

Table 5.4: CHI SQUARE
TEST, DIFFERENCES IN
RATES OF LOCAL
GOVERNMENT TAX
PAYMENT (%)
PAID IN LAST YEAR
DID NOT PAY IN LAST
YEAR

NIGERIA
30

ALL
AFROBAROMETER
COUNTRIES
(MEAN)
22

CHI SQUARE TEST
RESULT
chisq(1) is 0, p = 1

70

78

chisq(1) is 0, p = 1

NIGERIA
27

ALL
AFROBAROMETER
COUNTRIES
(MEAN)
24

CHI SQUARE TEST
RESULT
chisq(1) is 0, p = 1

73

76

chisq(1) is 0, p = 1

NIGERIA
31

ALL
AFROBAROMETER
COUNTRIES
(MEAN)
25

CHI SQUARE TEST
RESULT
chisq(1) is 0, p = 1

69

75

chisq(1) is 0, p = 1

240

Table 5.5: HYPOTHESIZED
EFFECTS OF QUALITY OF
REPRESENTATION AND TRUST
ON CITIZENS' WILLINGNESS TO
PAY TAXES AND ACTUAL TAX
PAYMENT

TIME LOCAL COUNCILLORS
SPEND LISTENING
TIME MPs SPEND LISTENING
CONTACT LOCAL COUNCILLOR
CONTACT MP
TRUST OTHER PEOPLE YOU
KNOW
TRUST OTHER NIGERIANS
EMPLOYMENT STATUS
AGE
LOCAL GOVERNMENT
CORRUPTION
PARLIAMENT CORRUPTION
TAX OFFICIALS CORRUPTION

2

3

PEOPLE MUST
PAY TAXES
MAKE LOCAL COUNCILLOR
LISTEN
MAKE MPs LISTEN

1

PAID
LOCAL
GOVERNME
NT TAXES

PAID
PROPERTY
TAX

+
+
+
+
+
+
+
+
+
-

+
+
+
+
+
+
+
+
+
-

+
+
+
+
+
+
+
+
+
-

241

4

PAID
INCOME
TAX

+
+
+
+
+
+
+
+
+
-

Table 5.6: SUMMARY STATISTICS-EFFECT OF QUALITY OF
REPRESENTATION AND TRUST
ON CITIZENS' WILLINGNESS TO
PAY TAXES AND ACTUAL TAX
PAYMENT
VARIABLE
PEOPLE MUST PAY TAXES
PAID INCOME TAX
PAID PROPERTY TAX
PAID LOCAL TAXES
MAKE LOCAL COUNCILLOR
LISTEN
MAKE MPs LISTEN
TIME LOCAL COUNCILLORS
SPEND LISTENING
TIME MPs SPEND LISTENING
CONTACT LOCAL COUNCILLOR
CONTACT MP
TRUST OTHER PEOPLE YOU
KNOW
TRUST OTHER NIGERIANS
EMPLOYMENT STATUS
AGE
LOCAL GOVERNMENT
CORRUPTION
PARLIAMENT CORRUPTION
TAX OFFICIALS CORRUPTION

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

1641
1641
1641
1641

3.613
0.294
0.264
0.306

1.151
0.456
0.441
0.461

1
0
0
0

5
1
1
1

1641

1.410

0.995

0

3

1641

1.237

1.005

0

3

1641

0.751

0.846

0

3

1641
1641
1641

0.631
0.421
0.182

0.788
0.810
0.563

0
0
0

3
3
3

1641

1.349

0.920

0

3

1641
1641
1641

1.004
2.113
31.53

0.932
1.725
11.355

0
0
18

3
5
86

1641

1.756

0.806

0

3

1641
1641

1.682
1.729

0.793
0.778

0
0

3
3

242

Table 5.7: DEMOGRAPHIC
COMPARISONS BETWEEN
ORIGINAL SAMPLE AND
NEW SAMPLE
AGE
EDUCATION
EMPLOYMENT STATUS

ORIGINAL
SAMPLE
(n = 2400)
31.303
4.402
2.013

NEW
SAMPLE
(n = 1641)
31.530
4.535
2.113

Table 5.8: RELIGION OF
RESPONDENTS (%)
CHRISTIAN ONLY
MUSLIM ONLY
ROMAN CATHOLIC

ORIGINAL
SAMPLE
(n = 2400)
35
31
12

NEW
SAMPLE
(n = 1641)
36
32
9

Table 5.9: ETHNICITY OF
RESPONDENTS (%)
HAUSA
IGBO
YORUBA

ORIGINAL
SAMPLE
(n = 2400)
23
16
21

NEW
SAMPLE
(n = 1641)
24
15
23

243

Table 5.10: URBAN VS.
RURAL DWELLERS (%)
RURAL
SMALL URBAN
LARGE URBAN

ORIGINAL
SAMPLE
(n = 2400)
13
36
51

NEW
SAMPLE
(n = 1641)
14
39
47

Table 5.11: GENDER
DISTRIBUTION (%)
MALE
FEMALE

ORIGINAL
SAMPLE
(n = 2400)
50
50

NEW
SAMPLE
(n = 1641)
53
47

244

Table 5.12: EFFECT OF QUALITY
OF REPRESENTATION AND TRUST
ON CITIZENS' WILLINGNESS TO
PAY TAXES

1

2

PEOPLE MUST
PAY TAXES
MAKE LOCAL COUNCILLOR
LISTEN

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX

1.278*
1.191*
(0.089)
(0.086)
0.988*
0.917
MAKE MPs LISTEN
(0.067)
(0.072)
TIME LOCAL COUNCILLORS
1.191*
0.986*
SPEND LISTENING
(0.101)
(0.013)
0.862*
0.961
TIME MPs SPEND LISTENING
(0.077)
(0.108)
1.064*
1.299*
CONTACT LOCAL COUNCILLOR
(0.079)
(0.111)
0.877*
1.123
CONTACT MP
(0.079)
(0.127)
TRUST OTHER PEOPLE YOU
0.929
1.087
KNOW
(0.068)
(0.099)
1.101*
1.097*
TRUST OTHER NIGERIANS
(0.076)
(0.095)
1.043*
1.198*
EMPLOYMENT STATUS
(0.031)
(0.039)
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.

245

3

1.264*
(0.103)
0.819
(0.066)
1.031*
(0.104)
1.088
(0.114)
1.263*
(0.111)
1.139
(0.133)
0.889
(0.081)
1.071*
(0.098)
1.171*
(0.039)

4

PAID
INCOME
TAX
1.087*
(0.089)
0.899
(0.073)
1.133*
(0.123)
1.038
(0.106)
1.283*
(0.109)
1.133
(0.124)
0.997
(0.092)
1.049*
(0.094)
1.201*
(0.041)

Table 5.12 (cont’d): EFFECT OF
QUALITY OF REPRESENTATION
AND TRUST ON CITIZENS'
WILLINGNESS TO PAY TAXES

1

2

PEOPLE MUST
PAY TAXES
1.101*
(0.004)
(-)0.856*
(0.069)
(-)1.115*
(0.091)
(-)1.135*
(0.064)
1.023 (
0.097)

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX

1.101*
(0.004)
LOCAL GOVERNMENT
(-)1.150*
CORRUPTION
(0.096)
(-)0.970
PARLIAMENT CORRUPTION
0.113)
(-)1.017*
TAX OFFICIALS CORRUPTION
(0.088)
0.651*
GENDER
(0.076)
0.211*
CONSTANT
-(0.071)
LOG PSEUDOLIKELIHOOD
-2268.455
-952.058
N
1641
1641
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.
AGE

246

3

1.101*
(0.004)
(-)1.072*
(0.166)
(-)0.907
(0.113)
(-)1.095*
(0.088)
0.082*
(0.099)
0.151*
(0.053)
-896.705
1641

4

PAID
INCOME
TAX
1.102*
(0.004)
(-)1.075*
(0.103)
(-)0.975
(0.098)
(-)1.117*
(0.097)
0.903*
(0.105)
0.108*
(0.037)
-938.464
1641

Table 5.13: SUMMARY STATISTICS-EFFECT OF GOVERNMENT
PERFORMANCE ON CITIZENS'
WILLINGNESS TO PAY TAXES AND
ACTUAL TAX PAYMENT
VARIABLE
PEOPLE MUST PAY TAXES
PAID INCOME TAX
PAID PROPERTY TAX
PAID LOCAL TAXES
SATISFACTION WITH
GOVERNMENT HEALTH SERVICES
SATISFACTION WITH
GOVERNMENT WATER AND
SANITATION SERVICES
SATISFACTION WITH
GOVERNMENT MAINTENANCE OF
ROADS AND BRIDGES
SATISFACTION WITH
GOVERNMENT ELECTRICICTY
SERVICES
SATISFACTION WITH LOCAL
GOVERNMENT MAINTENANCE OF
LOCAL ROADS
SATISFACTION WITH LOCAL
GOVERNMENT MAINTENANCE OF
LOCAL MARKETS

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

1596
1596
1596
1596

3.610
0.294
0.264
0.308

1.147
0.456
0.441
0.462

1
0
0
0

5
1
1
1

1596

2.329

0.909

1

4

1596

2.048

0.896

1

4

1596

2.182

0.928

1

4

1596

1.781

0.893

1

4

1596

1.987

0.906

1

4

1596

2.118

0.899

1

4

247

Table 5.13 (cont’d): SUMMARY
STATISTICS--EFFECT OF
GOVERNMENT PERFORMANCE ON
CITIZENS' WILLINGNESS TO PAY
TAXES AND ACTUAL TAX PAYMENT
VARIABLE
MAKE LOCAL COUNCILLOR LISTEN
MAKE MPs LISTEN
TIME LOCAL COUNCILLORS SPEND
LISTENING
TIME MPs SPEND LISTENING
CONTACT LOCAL COUNCILLOR
CONTACT MP
TRUST OTHER PEOPLE YOU KNOW
TRUST OTHER NIGERIANS
EMPLOYMENT STATUS
AGE
LOCAL GOVERNMENT
CORRUPTION
PARLIAMENT CORRUPTION
TAX OFFICIALS CORRUPTION
GENDER

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

1596
1596

1.407
1.223

0.992
1.005

0
0

3
3

1596

0.754

0.846

0

3

1596
1596
1596
1596
1596
1596
1596

0.632
0.411
0.175
1.354
1.008
2.112
31.510

0.787
0.799
0.546
0.919
0.935
1.723
11.362

0
0
0
0
0
0
18

3
3
3
3
3
5
86

1596

1.756

0.804

0

3

1596
1596
1596

1.684
1.727
1.463

0.796
0.782
0.499

0
0
1

3
3
2

248

Table 5.14: EFFECT OF
GOVERNMENT PERFORMANCE ON
CITIZENS' WILLINGNESS TO PAY
TAXES AND ACTUAL TAX
PAYMENT

1

PEOPLE MUST
PAY TAXES

2

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX
1.118*
0.994*
(0.076)
(0.072)

SATISFACTION WITH
0.972*
GOVERNMENT HEALTH SERVICES
(0.055)
SATISFACTION WITH
1.122*
1.052*
GOVERNMENT WATER AND
(0.064)
(0.073)
SANITATION SERVICES
SATISFACTION WITH
1.142*
1.168*
GOVERNMENT MAINTENANCE OF
(0.065)
(0.077)
ROADS AND BRIDGES
SATISFACTION WITH
1.256*
1.247*
GOVERNMENT ELECTRICICTY
(0.075)
(0.085)
SERVICES
SATISFACTION WITH LOCAL
1.168*
1.183*
GOVERNMENT MAINTENANCE OF
(0.072)
(0.089)
LOCAL ROADS
SATISFACTION WITH LOCAL
1.063*
1.009*
GOVERNMENT MAINTENANCE OF
(0.065)
(0.076)
LOCAL MARKETS
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.

249

3

4

PAID
INCOME
TAX
1.139*
(0.075)

0.858*
(0.088)

1.019*
(0.070)

1.247*
(0.083)

1.358*
(0.081)

1.292*
(0.089)

1.376*
(0.095)

1.197*
(0.072)

1.349*
(0.105)

1.058*
(0.105)

0.966*
(0.076)

Table 5.14 (cont’d): EFFECT OF
GOVERNMENT PERFORMANCE ON
CITIZENS' WILLINGNESS TO PAY
TAXES AND ACTUAL TAX PAYMENT

1

2

3

PEOPLE
PAID LOCAL
PAID
MUST PAY GOVERNMENT PROPERTY
TAXES
TAXES
TAX
1.250*
1.129*
1.276*
MAKE LOCAL COUNCILLOR LISTEN
(0.089)
(0.092)
(0.109)
1.158*
0.922
0.842
MAKE MPs LISTEN
(0.069)
(0.074)
(0.071)
TIME LOCAL COUNCILLORS SPEND
1.173*
1.021*
1.278*
LISTENING
(0.102)
(0.112)
(0.110)
1.073*
0.974
1.131
TIME MPs SPEND LISTENING
(0.082)
(0.113)
(0.123)
1.067*
1.321*
1.278*
CONTACT LOCAL COUNCILLOR
(0.079)
(0.115)
(0.110)
1.068*
0.881
0.912
CONTACT MP
(0.082)
(0.133)
(0.123)
0.951
0.918
0.914
TRUST OTHER PEOPLE YOU KNOW
(0.072)
(0.103)
(0.087)
1.241*
1.292*
1.285*
TRUST OTHER NIGERIANS
(0.076)
(0.099)
(0.116)
1.033*
1.201*
1.172*
EMPLOYMENT STATUS
(0.031)
(0.039)
(0.041)
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.

250

4
PAID
INCOME
TAX
1.113*
(0.095)
0.906
(0.077)
1.193*
(0.103)
0.906
(0.077)
1.204*
(0.137)
0.862
(0.132)
0.827
(0.096)
1.267*
(0.099)
1.219*
(0.042)

Table 5.14 (cont’d): EFFECT OF
GOVERNMENT PERFORMANCE
ON CITIZENS' WILLINGNESS TO
PAY TAXES AND ACTUAL TAX
PAYMENT

1

2

PAID LOCAL
GOVERNMENT
TAXES
1.006*
1.011*
AGE
(0.004)
(0.005)
LOCAL GOVERNMENT
(-)0.974*
(-)1.182*
CORRUPTION
(0.067)
(0.098)
(-)1.116*
(-)0.904
PARLIAMENT CORRUPTION
(0.094)
(0.114)
(-)1.091*
(-)1.218
TAX OFFICIALS CORRUPTION
(0.067)
(0.091)
1.051
0.864*
GENDER
(0.102)
(0.081)
CONSTANT
-0.317
LOG PSEUDOLIKELIHOOD
-2202.473
-921.025
N
1596
1596
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.
PEOPLE MUST
PAY TAXES

251

3

4

PAID
PROPERTY
TAX
1.019*
(0.005)
(-)1.151*
(0.101)
(-)0.842
(0.117)
(-)1.130
(0.110)
0.885*
(0.110)
0.246
-852.962
1596

PAID
INCOME
TAX
1.021*
(0.005)
(-)1.230*
(0.104)
(-)0.823
(0.105)
(-)1.089
(0.097)
0.936*
(0.112)
0.110
-903.239
1596

Table 5.15: FACTOR ANALYSIS OF
CITIZEN PERCEPTIONS OF
REPRESENTATION INDICATORS
(Orthogonal Varimax Principle Factors)
FACTOR
FACTOR 1
OBSERVATIONS
RETAINED FACTORS
NUMBER OF PARAMETERS
CHI2 (15)
PROB > CHI2

VARIANCE
1.332
1596
1
15
3371.3
0

Table 5.16: FACTOR ANALYSIS OF
CITIZEN PERCEPTIONS OF
REPRESENTATION INDICATORS (Rotated
Factor Loadings)
VARIABLE
MAKE LOCAL COUNCILLOR LISTEN
MAKE MPs LISTEN
TIME LOCAL COUNCILLORS SPEND
LISTENING
TIME MPs SPEND LISTENING
CONTACT LOCAL COUNCILLOR
CONTACT MP

252

PROPORTION
0.467
1596
1
15
3371.3
0

FACTOR 1
0.601
0.621
0.657
0.644
0.549
0.496

Table 5.17: FACTOR ANALYSIS OF
CITIZEN SATISFACTION WITH
GOVERNMENT PERFORMANCE
INDICATORS (Orthogonal Varimax
Principle Factors)
FACTOR
FACTOR 1
OBSERVATIONS
RETAINED FACTORS
NUMBER OF PARAMETERS
CHI2 (15)
PROB > CHI2

VARIANCE
1.395
1596
1
15
2819.4
0

Table 5.18: FACTOR ANALYSIS OF
CITIZEN SATISFACTION WITH
GOVERNMENT PERFORMANCE
INDICATORS (Rotated Factor
Loadings)
VARIABLE
SATISFACTION WITH
GOVERNMENT HEALTH SERVICES
SATISFACTION WITH
GOVERNMENT WATER AND
SANITATION SERVICES
SATISFACTION WITH
GOVERNMENT MAINTENANCE OF
ROADS AND BRIDGES
SATISFACTION WITH
GOVERNMENT ELECTRICICTY
SERVICES
SATISFACTION WITH LOCAL
GOVERNMENT MAINTENANCE OF
LOCAL ROADS
SATISFACTION WITH LOCAL
GOVERNMENT MAINTENANCE OF
LOCAL MARKETS

253

PROPORTION
0.613
1596
1
15
2891.4
0

FACTOR 1
0.559
0.607
0.544
0.552
0.707
0.699

Table 5.19: EFFECT OF QUALITY OF
REPRESENTATION AND SATISFACTION
WITH GOVERNMENT PERFORMANCE
INDICES ON CITIZENS' WILLINGNESS
TO PAY TAXES

1

PEOPLE MUST
PAY TAXES
CITIZEN REPRESENTATION
0.107*
CITIZEN SATISFACTION
0.085*
TRUST OTHER PEOPLE YOU KNOW
0.013
TRUST OTHER NIGERIANS
0.056*
EMPLOYMENT STATUS
0.019
AGE
0.032
LOCAL GOVERNMENT CORRUPTION
-0.092*
PARLIAMENT CORRUPTION
-0.050*
TAX OFFICIALS CORRUPTION
-0.057*
GENDER
0.008
CONSTANT
3.591
STANDARD ERROR
0.195
R2
0.210
N
1596
Notes: Values are standardized OLS regression coefficients (betas).
*p < 0.05

254

2

3

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX
0.092*
0.143*
0.077*
0.107*
0.04
0.029
0.054*
0.057*
0.146*
0.118*
0.065*
0.098*
-0.063*
-0.053*
-0.043
-0.019
-0.052*
-0.053*
-0.087*
-0.029
0.228
0.193
0.075
0.069
0.268
0.264
1596
1596

4

PAID
INCOME
TAX
0.114*
0.098*
0.007
0.058*
0.154*
0.109*
-0.058*
-0.003
-0.066*
-0.017
0.208
0.072
0.264
1596

Table 5.20: URBAN NIGERIA,
SUMMARY STATISTICS-EFFECT OF GOVERNMENT
PERFORMANCE ON
CITIZENS' WILLINGNESS
TO PAY TAXES

!!

VARIABLE

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

1638

1.744

0.752

1

3

1638

2.260

0.967

1

4

1638

2.073

0.837

1

3

1638

2.045

0.757

1

3

1638

2.062

0.777

1

3

1638

1.822

0.766

1

3

1638
1638

6.676
3.184

2.345
1.336

1
1

12
7

1638

1.305

0.46

1

2

1638

1.544

0.498

1

2

WILLINGNESS TO PAY
TAXES
SATISFACTION WITH GOVT.
TAX SPENDING
SATISFACTION WITH ROAD
MAINTENANCE
SATISFACTION WITH
PUBLIC SCHOOL
MAINTENANCE
SATISFACTION WITH
HOSPITAL MAINTENANCE
SATISFACTION WITH
WATER SUPPLY
HOUSEHOLD INCOME
AGE
EXPERIENCE WITH
GOVERNMENT
CORRUPTION
GENDER

!!

255

!!

!!

!!

Table 5.21: URBAN
NIGERIA, DISTRIBUTION
OF HOUSEHOLD INCOME
N16,000 - N25,000
N26,000 - N35,000
N26,000 - N45,000

Table 5.23: URBAN
NIGERIA, RELIGION OF
RESPONDENTS (%)
CATHOLIC
CHRISTIAN (GENERAL)
MUSLIM (AHMADDIYA)

256

NEW
SAMPLE
(n = 1638)
14
16
14

ORIGINAL
SAMPLE
(n = 2750)

NEW
SAMPLE
(n = 1638)

10

6

10

10

66

67

ORIGINAL
SAMPLE
(n = 2750)
17
12
10

Table 5.22: URBAN
NIGERIA, DISTRIBUTION
OF RESPONDENTS'
EDUCATION (%)
PRIMARY SCHOOL
COMPLETED
SOME SECONDARY
SCHOOL
SECONDARY SCHOOL
COMPLETED

ORIGINAL
SAMPLE
(n = 2750)
15
17
14

NEW
SAMPLE
(n = 1638)
17
11
10

Table 5.24: URBAN
NIGERIA, ETHNICITY OF
RESPONDENTS (%)
HAUSA
IGBO
YORUBA

ORIGINAL
SAMPLE
(n = 2750)
29
32
17

NEW
SAMPLE
(n = 1638)
28
33
19

Table 5.25: URBAN
NIGERIA, AGE
DISTRIBUTION (%)
20 - 24 YEARS OLD
25 - 34 YEARS OLD
35 - 44 YEARS OLD

ORIGINAL
SAMPLE
(n = 2750)
23
36
16

NEW
SAMPLE
(n = 1638)
20
37
17

Table 5.26: URBAN
NIGERIA, GENDER
DISTRIBUTION (%)
MALE
FEMALE

ORIGINAL
SAMPLE
(n = 2750)
50
50

NEW
SAMPLE
(n = 1638)
54
46

257

Table 5.27: URBAN NIGERIA, EFFECT OF
GOVERNMENT PERFORMANCE ON
CITIZENS' WILLINGNESS TO PAY TAXES
SATISFACTION WITH GOVT. TAX
SPENDING
SATISFACTION WITH ROAD
MAINTENANCE
SATISFACTION WITH PUBLIC SCHOOL
MAINTENANCE
SATISFACTION WITH HOSPITAL
MAINTENANCE

WILLINGNESS
TO PAY TAXES

1.000*
(0.002)
1.461*
(0.101)
1.246*
(0.068)
1.413*
(0.116)
1.270*
SATISFACTION WITH WATER SUPPLY
(0.085)
1.033*
HOUSEHOLD INCOME
(0.019)
1.114*
AGE
(0.039)
EXPERIENCE WITH GOVERNMENT
-0.991
CORRUPTION
(0.099)
LOG PSEUDOLIKLIHOOD
-1765.149
N
1638
Notes: Values are proportional odds ratios, robust standard errors in parentheses.
*p < 0.05, two-tailed test.

258

Figure 5.1: “The tax department always has the right to make people pay taxes.”

50
45
40
35
%

30
25
20
15
10
5
0
!"#$%&'()
*+!,&#--)

*+!,&#--)

%-+".-#)
,&#--)%$#)
*+!,&#--)

259

,&#--)

!"#$%&'()
,&#--))

Figure 5.2: “Which of the following statements is closest to your view? Choose Statement 1 or Statement 2.
Statement 1: The National Assembly should ensure that the President explains to it on a regular basis how his government
spends taxpayers’ money.
Statement 2: The President should be able to devote his full attention to developing the country rather than wasting time
justifying his actions.
30
25

%

20
15
10
5
0
AGREE VERY AGREE WITH AGREE WITH AGREE VERY AGREE WITH
STRONGLY
1
2
STRONGLY
NEITHER
WITH 1
WITH 2

260

Figure 5.3: “Who do you think actually has primary responsibility for: Collecting income taxes?”
40
35
30

%

25
20
15
10
5
0
FEDERAL
STATE
LOCAL
MEMBERS OF TRADITIONAL
GOVERNMENT GOVERNMENT GOVERNMENT
THE
LEADERS
COMMUNITY

261

NONE OF
THEM

Figure 5.4: “Have you had to make any of the following payments during the past year…”

80
70
60

%

50

NO: NIGERIA ONLY

40
YES: NIGERIA ONLY
30
YES: ALL AFROBAROMETER
COUNTRIES

20
10
0
INCOME TAX

PROPERTY TAX

LOCAL
GOVERNMENT
FEES

262

Figure 5.5: “Given the current government’s performance that you have seen, do you find yourself more willing or less willing
to pay taxes than you were during the last civilian administration in _____ State? Or would you say your attitude toward taxes
is about the same?”
50
45
40
35

%

30
25
20
15
10
5
0
LESS WILLING

ABOUT THE SAME

263

MORE WILLING

Figure 5.6: (ACROSS STATES) “Given the current government’s performance that you have seen, do you find yourself more
willing or less willing to pay taxes than you were during the last civilian administration in _____ State? Or would you say your
attitude toward taxes is about the same?”
50
45
40
35

%

30
25
20
15
10
5
0

STATES

264

CHAPTER SIX

Comparing Tax Cultures in Lagos and Kano
Background
Analyses in previous chapters demonstrate that, in Nigeria, revenue and representation
are linked. At macro and micro levels, state governments that generate more income from nonoil taxes on citizens are also more representative of citizens. These states spend less on
government salaries and allowances, instead devoting public resources to non-recurrent
expenditures, including service provision. Likewise, legislators in states with greater capacity to
extract taxes report spending more time performing constituency services. They are also more
likely to report prioritizing the views of their constituents in decision-making. Chapter Five
moves beyond the how revenue influences elite representation, demonstrating how ordinary
citizens perceive the fiscal contract. Nigerians are more willing to pay taxes and report actual
tax payment when they observe a higher quality of representation from elected officials.
Similarly, a Nigerian who is more satisfied with government performance (national, state, local)
is also more willing to pay taxes and follow through with actual tax payment. Looking within
regions and using intra-regional variation, these revenue-representation relationships manifest
across Nigeria.
Now I consider these linkages across regions in Nigeria by comparing Lagos and Kano
States. Though Lagos (South-West) and Kano (North-West) share similar levels of development,
there is a large gap between the two states in terms of tax effort. Over the last 10 years, Lagos,
on average, generated 53% of state income from taxes on citizens. On the other hand, Kano
State averaged only 13% of state income from taxation. This chapter begins with an assessment
265

of the levels of socio-economic development in Lagos and Kano, establishing that the two states
are comparable. Next, I consider why there is a difference between each state’s capacity to
generate tax income. I rely on original interviews with tax administrators and archival
documents to perform an analysis of the Lagos and Kano State Revenue Boards. Ultimately, I
argue the higher rate of tax generation in Lagos (in comparison to Kano) can be attributed to elite
political initiative. Through legislation, Lagos’ Governor Fashola and members of the Lagos
House of Assembly have empowered the State Internal Revenue Board (SIRB) to educate the
public about the fiscal contract while pursuing/punishing evaders. On the other hand, in Kano,
rather than strengthening the SIRB’s capacity, Governor Shekarau and political elites have
instead pursued efforts to increase the state’s share of petroleum income. Finally, I investigate
how the differences between Lagos Kano and result in two different tax cultures, manifested in
elite and citizen perceptions of the tax contract.

Comparing Lagos and Kano: Socio-Economic Indicators
In the period since Nigeria’s 1999 return to democracy, I find that Lagos and Kano share
99

similar demographic, social, and economic attributes.

Looking at Table 6.1, we see both states

have around nine million citizens. Taking into account Nigeria as a whole, Lagos State citizens
make up 6.4% of the national population. Similarly, Kano’s state population is also about 6.7%
of the whole. With that said, while Kano State has an area of roughly 7,700 square miles, Lagos
State is one-sixth that size (1300 square miles). In terms of population density, there are about

99

The statistics presented are from two 2006 and 2007 (respectively) reports from the Nigerian
National Bureau of Statistics: Social Statistics in Nigeria, 2005 and 2006 Population Census.
266

6,718 people per square mile, compared to 1,207 people per square mile in Kano. Moreover, in
Lagos, 93% of individuals reside in urban areas, while in Kano, this percentage is smaller (35%).
Lagos and Kano are also similar in regards to education, health, and employment.
Students enrolled in public primary school make up about 14% of Lagos and Kano’s state
populations. Those enrolled in secondary school are roughly 16% of Lagos State’s populace; in
Kano, this figure is closer to 12%. Looking at the ratio of state population to number of primary
health care facilities, in Lagos there are about 8,000 people to 1 facility. This ratio is higher in
Kano state where there are 10,000 people to each primary health care facility.
Finally, looking at employment, about 37% of the populations that are of working age in
Lagos and Kano State are employed. Table 6.2 describes the distribution of activities in which
the working population is engaged. Private business via wholesale and retail trade is the
dominant mode of employment in both states. 40% of Lagos’ employed citizens work in this
sector, compared with 33% of Kano’s labor force. In Lagos, occupations in the service industry
(e.g. hotels and restaurants, financial services, social and personal services) are the second-most
common. 20% of employed Lagosians work in services, while 14% of Kano’s labor populace
also works in this industry. In Lagos State 8% of the workforce are engaged in manufacturing
and 7% in public administration. In Kano, between 2% and 3% of working people are in these
two industries.
With that said, there are key differences between Lagos and Kano, both in terms of their
working populations and ethno-religious characteristics. For example, 25% of Kano’s working
population is engaged in agriculture, making it the second-most popular in the state. On the
other hand, in Lagos, only 2% of employed individuals work in this area. According to Nigerian

267

tax administrators, agriculture (particularly, smaller-scale farming) tends to be a difficult sector
to tax (for example, because of the informal nature of record-keeping). On the other hand,
individuals employed in the service sector generally work in more formal business settings.
These employers are easier to pursue for tax information and payments.
In addition, there are large differences in the ethnicities of Lagos’ and Kano’s populations.
In Lagos, 69% of the population is Yoruba, while 16% identifies as Igbo. On the other hand, in
Kano, 93% of the population belongs to the Hausa-Fulani ethnic group. These disparities also
apply to religion: in Lagos, the majority of the population (61%) identify themselves as
Christians. On the other hand, in Kano, most people (40%) are Muslim, 16% are members of the
Izala movement (an Islamic society), and 12% identify as Sunni Muslims.
Despite these differences, I argue that a comparison between Lagos and Kano would be
fruitful. These two states are the most populous in the federation with the same percentage of
employed persons. Lagos and Kano have access to tax-bases of similar size. Taken together
with their socio-economic similarities, one would also expect similar levels of tax effort.
However, this is not the case. For example, between 2005 and 2009, the yearly portion of Lagos’
100

income derived from taxes was between three and five times that of Kano State (Table 6.3).

Why has Lagos State been able to augment its capacity to extract taxes, while taxes (as a portion
of total revenue) in Kano have increased at a more modest pace? Using evidence from
interviews with members of Lagos’ and Kano’s tax administration and, I investigate the origins
of these differences.
100

For a more in-depth discussion of Lagos and Kano’s tax and oil dependence, see Chapter
Four: Regional Variation in Tax and Oil Dependence: Case Selection and Revenue in the SouthWest and North-West Cases.

268

Explaining Lagos’ Tax Capacity: Political Leadership and Civic Education
Over the last ten years in Nigeria’s post-authoritarian period, the Lagos State Internal
Revenue Service (LIRS), acting on behalf of the state government, launched several projects to
increase tax generation. In interviews with two high-level LIRS administrators, internally
generated revenue via taxation and tax-led development are identified as superior to petroleum
income. As one official asserts, “Oil is subject to international fluctuations and economic
downturns. We can’t always rely on it. Lagos is among the first states in Nigeria to understand
this” (Director #1, Lagos Internal Revenue Service, June 1, 2011). A second tax administrator
agrees:
“For sustainable and meaningful development, taxation is necessary. Nigeria cannot only
have an oil economy. Natural resources are not there forever. It’s not sustainable for
economic development. Taxes come yearly. As long as we invest appropriately in
development, taxes will regenerate. Taxes are the only viable means” (Director #2,
Lagos Internal Revenue Service, Personal Communication, June 8, 2011).
LIRS officials have specified how political initiative from the executive and legislative arms of
government is paramount to the success of taxation in Lagos:
“Recently, there’s even more realization that states can’t survive without internally
generated revenue. Politicians have put pressure on the civil service to improve the tax
administration. They have also been encouraging citizens to make their payments.”
(Director #2, Lagos Internal Revenue Service, Personal Communication, June 8, 2011).
In fact, the Lagos State Government has been active in engaging citizens in an open
dialogue about the tax contract and current reforms. For example, in an address to the Fourth
Lagos State Taxation Stakeholder’s Conference, Governor Babatunde Fashola was very clear
about the government’s stance:
“In the great social contract which binds all organized societies, the ability of government
and public officers to play their part depends largely on resources made available to
them; and the most certain, most durable, and most sustainable of these resources are the

269

public funds which the people themselves contribute by way of taxation…In Lagos, in
Nigeria, and throughout the world, it is becoming increasingly clear that a government
cannot do more than it is financially empowered to do, and that taxation remains the
golden key to economic development…beware of politicians who are coming to promise
you that they can achieve development without enforcing the collection of taxes. Sweet
as it may sound, the promise of lower and lower taxes must translate to lower and lower
resources for government and eventually undermine government’s ability to tackle even
the basic developmental challenges that daily confront us in the great megacity of ours
(December 8, 2010).
Under Governor Fashola’s leadership, The House of Assembly has joined Governor
Fashola in affirming the importance of taxation to the state through legislation. For example,
after LIRS conducted an investigation into citizen and civil society complaints to identify the
most prevalent problems in tax administration. The results of this inquiry revealed that tax
collection at the local level was the most problematic. The lack of information between local
authorities and the taxpayers led to the collection of illegal taxes by fake tax collectors. And in
the case of legal taxes, there were even reports of “unscrupulous revenue officers in the habit of
varying the amounts payable to facilitate unlawful negotiations and extorting bribes from
taxpayers” (Ipaye, 2010). As a response, LIRS officials collaborated with the Lagos State
Legislature to draft the Local Government Levies Law (passed into law in April 2010). This
law lists the levies state and local governments can collect. It also stipulates that local and state
tax officials are required to present official identification when interacting with citizens. Tax
collecting authorities must explicitly publish the legal taxes and the how they will be
administered and collected. Moreover, “the use of roadblocks and other obstructions for
purposes of tax collection” is strictly prohibited.
Members of Governor Fashola’s political party in the state legislature (Action Congress
of Nigeria, also the majority party in the House of Assembly) supported the executive and
bureaucratic effort to increase tax capacity at the local level. As one legislator remarks:

270

“We cannot give in to the attempts to slow down our tax drive. We must ensure that we
administer taxes correctly, and this new law will allow us to do so. We must clear any
areas of doubt or confusion in taxation. We can make sure that tax payment reaches
government coffers and can be used for our common benefit” (Lagos State Honourable
#2, Lagos State House of Assembly, Personal Communication, May 10, 2011).
With that said, there was some opposition from members of the minority People’s Democratic
Party. According to another legislator,
“it is unconstitutional for local governments and even state governments to be collecting
their own revenue. All revenue should be collected in one pool. All taxes should go to
the Federation Account. From there, it can be disbursed to the federal state, and local
level” (Lagos State Honourable #19, Lagos State House of Assembly, Personal
Communication, May 16, 2011).
Although the majority party prevailed and the Local Government Levies Law passed,
politicians and tax administrators recognized that reforms to tax collection could not be achieved
without popular support and understanding of taxation. Armed with legal authority, LIRS has
taken this message directly to Lagosians. By focusing on income tax payment and collection,
officials have engaged in a campaign to educate citizens about the fiscal contract, conveying how
paying one’s taxes is a civic responsibility with clear benefits in terms of representation and
service provision. In an interview, a tax administrator comments that:
“Before government can charge taxes, we must show how the revenue is used to provide
services. Government must demonstrate to citizens that the money from taxes is used to
develop infrastructure and facilitate business through access to shops, land, and credit.
We use the money from taxes to create an enabling environment for citizens. Once
people understand this, they will pay” (Director #2, Lagos Internal Revenue Service,
Personal Communication, June 8, 2011)
For example, in an address to the Lagos State Branch of the Medical Women’s Association of
Nigeria on International Women’s Day, Mr. Tunde Fowler, Executive Chairman and CEO of the
Lagos State Internal Revenue Service, outlines how taxation benefits women:
“Women cannot be separated from the general society; when we talk about the society,
we are talking about women and when we speak of the women, we are tabling about the

271

society. This is so because of the critical roles women play in the economic, political,
and social development of the society… ‘Health with the spirit of a mother’ is only
attainable because government provides the necessary enabling environment, which is
financed from taxation. Taxation is the bedrock of financing government activities and
the benefits to society derives from the provision of these services” (March 8, 2009).
Likewise, in a presentation to the National Association of Banking and Finance Students,
Mr. Fowler emphasizes the contractual nature of tax payment, and what non-compliance would
mean for the ordinary citizen:
“The payment of taxes on income is a first principle of the social contract between the
government and the governed. As that great American Statesman, Franklin D. Roosevelt,
puts it, ‘taxes, after all, are dues that we pay for the privileges of membership in an
organized society.’ It is the price we pay and must continue to pay for a civilized society.
The people are at liberty to question the administration of the tax laws and demand
accounting through their elected representatives. But no viable state can exist without tax.
If government did not impose taxes, there would be nothing public. No highways, no
public hospitals and schools, no law enforcement, no courts, no fire service, no waste
collection, and so on” (October 28, 2010).
LIRS stresses how every able citizen must make their contribution through tax payment,
“all income earners should pay tax. We all pay, no matter how small our incomes” (Ipaye, 2009).
As a result, members of the tax administration have worked to expand the tax net while making
compliance easier. This is particularly the case when dealing with the informal sector. Mr. Ade
Ipaye, Special Adviser to the Governor of Lagos State on Taxation and Revenue, outlines these
efforts in a 2009 report. The informal economy refers to economic activities that occur outside
of government records and regulations. Because it involves “small-scale, largely selfemployment activities, the informal sector is difficult to measure.” With that said, Lagos State
tax administrators assert value in the informal economy:
“No doubt [the informal sector is] highly dynamic and pervasive, contributing
substantially to employment and personal or household income. As a result, the informal
sector is a major contributor to economic growth…it also provides competition in the
economy and enhances innovation and adaptation, mobilizing capital and human
resources which would otherwise been laid waster and idle” (Fowler, 2010).

272

Since LIRS promotes payment from all income-earning individuals, no matter how small their
income, the tax administration has made reforms in order to enhance compliance from those
working in the informal sector. For example, LIRS has increased the number of offices and
locations. Individuals can file and pay taxes at over 120 bank branches, 36 tax stations, and
mini-stations located in various markets. In addition, “forms can be obtained in all these
locations and there are officers ready, able, and willing to explain the process absolutely free of
charge” (Ipaye, 2010). Now, instead of travelling long distances and missing work in order to
pay taxes, the process can be done at the taxpayer’s convenience. Second,
“The informal sector, as the name connotes, [is made up of] very informal people. They
don’t keep accurate records. We have had to have discussions with them and agree on
what is called a minimum tax expectation based on their income…a minimum of N2,500,
then we’ve moved up gradually from there” (Fowler, 2010).
Furthermore, LIRS provides a table on the back of the tax forms where individuals can assess
their income and expenditures with reference to others in their line of business. Then a taxpayer
can make the corresponding income tax payment. The new Revenue Complaints and
Information Unit investigates any problems, including any problems a taxpayer may experience
with LIRS staff members. Finally, LIRS has urged members of the informal sector to join
trade/business associations:
“This makes communication easier as we can most easily pass information or receive
information from [those employed in the informal sector] through those
associations…[For example,] staff members of LIRS have always attended the market
men and women association meetings, and this offers the opportunity to answer members’
questions, take messages to government, and clear up outstanding issues. We have also
arranged meetings with several other associations and trade groups in order to achieve the
same end” (Ipaye, 2009).

273

Through community awareness campaigns, commercials, billboards/placards, films,
documentaries, radio jingles, and other various means, LIRS (on behalf of the Lagos State
Government) has entered into an explicit bargain with citizens. Pay taxes and government will
use that income to address citizens’ needs, improve infrastructure, and develop the state. A highlevel director at the Nigerian Ministry of Finance cites Lagos State as an example for the
federation to follow:
“How I wish there were Fasholas in all states of Nigeria. Taxation is a mutual
relationship: government must provide social amenities and people must pay tax. Every
citizen has an obligation and civic responsibility to pay. It’s like voting. But if leaders
can demonstrate what tax revenue can do, the community willingly pays. People in
Lagos have seen the changes, the good roads and infrastructure. In Lagos, government
functions for the people. Now people are eager to pay taxes because they see the benefits.
They are diversifying the tax base by bringing the informal economy under the tax net
and stopping the leakages. Lagos shows that in Nigeria, the will is there” (Director,
Nigerian Ministry of Finance, Personal Communication, March 29, 2011).
Lagos officials identify tax income as the most optimum form of revenue, engaging in an
explicit bargain with citizens. These officials aim to educate citizens about the role of taxation in
society and the benefits ordinary people derive with tax income in regularized. Ultimately, these
efforts have started to shift the tax culture in Lagos State. Through political leadership (e.g.
putting taxation on the public agenda, legislation) and civic education, Lagos State engages
citizens and has entered into a tax contract. By outlining the benefits of taxation to society as a
whole and specific groups and simplifying the tax payment process, LIRS works to change
Lagos’ increase tax compliance.

274

A Contrast with Kano: The Lack of Political Will and an Adversarial Tax Culture
Turning to Kano State, we observe a markedly different relationship between the state
revenue service and political arms of government. As in Lagos, tax administrators in the Kano
Board of Internal Revenue (KBIR) identify tax revenue as superior to petroleum income;
however, State Governors have not prioritized taxation in the public agenda or reforms to the
state tax administration. Without political leadership and resources, KBIR has not been able to
engage or educate citizens about the tax contract. Kano’s tax administration, in conjunction with
the World Bank, has outlined various strategies to place tax reform and compliance on the
political agenda, working to shift elite and public perceptions.
In an interview with a top KBIR director, internally generated income through taxation is
the best way to fund development:
“Certainly yes. It is through tax that we can build infrastructure. But in Kano, if the
statutory allocation [federal transfers of petroleum income)] was not there, how would
Kano survive? This is the big question. We can only survive through tax revenue. We
must tap taxation to survive in future” (Director, Kano Board of Internal Revenue,
Personal Communication, July 5, 2011).
Like his counterparts in Lagos State, this KBIR administrator believes tax-based development is
the most reliable and sustainable. However, he emphasizes that Kano is not meeting its
potential:
“We should collect five billion naira per month in taxes, but we collect far less, between
five and six hundred million. We are not collecting enough. In fact, it is grossly
101
inadequate.”

101

This is compared to Lagos State, which reports generating an average of 14.5 billion naira
per month in tax revenue (Fowler, 2010).
275

When asked to describe the nature of KBIR’s relationship with elected officials,
particularly the Governor and Kano’s House of Assembly, this director describes an adversarial
one, in which the state tax administration has been relegated:
“For now, [politicians] have not grasped the benefits to accrue from taxation. There is a
total lack of support for the tax administration. The facilities don’t meet our needs. The
board [KBIR] is treated as if we are not a ministry. We are not given resources. We
don’t have vehicles or computers. There is not enough for the adequate payment for our
salary. There is no capacity building. There is no political pressure on citizens to pay
tax.”
Rather than develop state tax capacity, Kano’s State Governors (as a part of the Northern
Governors Forum) have called for reforms to the formula for allocating federal oil income. Over
the last four years, these state executives have sought to increase the percentage of oil income
coming to Northern States. In particular, Governor Kwankwaso of Kano (1999 – 2003) opposed
the derivation principle, whereby states that produce petroleum collect an additional 13% of oil
income (on top of revenue from the standard allocation). According to Governor Kwankwaso:
“The South-South [location of petroleum producing states] get more than their northern
counterparts from [oil] incomes emanating from the Federation Account. It implies that
some states in the country are getting richer while others get poorer by the day. This has
been unfair to majority of Nigerians especially the northern states. It is unfair for oil
states to be receiving huge sums of money. The North today is in a very grave situation
where illiteracy, poverty and general backwardness are on the rise. This is a result of the
unfavorable federation allocation structure in which the Northern states are at a great
disadvantage” (February 24, 2012, Address to Northern Governors Forum during the
Inauguration of the Sir Ahmadu Bello Memorial Foundation Advisory Council).
However, members of Kano’s tax administration have criticized Governor Kwankwaso, instead
arguing that:
“It is irresponsible to ask for more revenue based on what is given to oil bearing states.
Rather than ask for more funds from the federal government, Kano should explore ways
of more money for itself. Leaders should concentrate on diversifying of our revenue base
and use sustainable alternatives like taxes” (Director, Kano Board of Internal Revenue,
Personal Communication, July 5, 2011).

276

In 2009, KBIR collaborated with the World Bank and the UK Department for
International Development (DFID) to assess the current tax administration and develop a
modernization plan to reform KBIR and tax generation in the state. As the KBIR director argues,
this report confirms the lack of political leadership from state Governors and the lack of
autonomy and resources in Kano’s tax administration:
“[The KBIR needs] considerable autonomy in control and decision making in the areas of
fiscal as well as human and physical resource management…As regards physical
infrastructure, the Kano State BIR offices are housed in rented premises (with the
exception of one) making it very difficult for long term planning. The buildings, furniture
and general office equipment are largely in a dilapidated state and are unsuitable for
modern ways of working, service delivery, and the installation of modern computer
systems” (KBIR Modernization Plan, 2009).
Both the KBIR director and the Modernization Plan explain the tax administration’s
relationship with citizens in the same adversarial fashion:
“Citizens have no desire to pay [taxes]. This is probably because of the years of no tax
payment during the military regime. People are now skeptical. They won’t pay tax
unless they understand why” (Director, Kano Board of Internal Revenue, Personal
Communication, July 5, 2011).
The Modernization Plan corroborates this characterization, emphasizing:
“Above all, a fight against public ignorance of the taxation laws and their obligations as
taxpayers is a central pillar in increasing effectiveness in revenue generation as well as
propelling a positive corporate image of the BIR” (KBIR, 2009).
In line with the Modernization Plan, the KBIR director argues Kano residents must be
educated about taxation as a civic duty with benefits for society as a whole:
“The old method of forced compulsion no longer applies. No one pays taxes with a smile,
but we must educate people to voluntarily pay. This civic tax education can enlighten
taxpayers to make them aware of how to make payments and what they gain from it. If
you want to reform and boost taxes, taxpayers’ education through advertising and public
hearings is necessary” (Director, Kano Board of Internal Revenue, Personal
Communication, July 5, 2011).

277

Like administrators in Lagos, KBIR officials recognize that before citizens will pay taxes,
government must make the first move, demonstrating how regular tax income allows
government to engage in public development:
“Taxpayers doubt the government. They do not believe the money they will pay will be
used judiciously. Government must first show what they have done, improving
electricity, water supply, roads, and security. Government must create an enabling
environment where people can carry out their business. It is a give and take. If people
believe when they pay tax, they will get these things in return, citizens willingly pay.”
To address the limited political resources available to KBIR, the board received
assistance from the World Bank/DFID tax team to draft the Kano State Board of Internal
Revenue Autonomy Act. This legislation was submitted to the Kano State House of Assembly
and passed on February 1, 2012. As the title suggests, this act would afford the KBIR political
and fiscal autonomy to act as independent revenue board:
“[With] freedom from the traditional Civil Service institutional arrangements, [KBIR
can] adopt new ways of working and funding so as to embrace better and modern
techniques of tax administration. Foremost, management plans will reorganize the BIR
governance structure so as to become more productive, efficient and taxpayer focused”
(KBIR, 2009).
With this autonomy, the KBIR would be able improve its office infrastructure and computerize
and automate the tax collection process. Administrators believe “[reducing] human interventions
in tax transactions,” will curtail corrupt practices. In this scenario (similar to that in Lagos), tax
officials can focus on civic education and improving interactions with taxpayers (KBIR, 2009).
Thus, the bulk of KBIR’s plan is aimed directly at citizens, with the goal of improving
voluntary taxpayer compliance, primarily thorough public awareness campaigns. A Public
Relations & Protocol Unit will be charged with developing reference materials for citizens and
organizing taxpayer education, which includes the following methods: conferences, consultative
meetings, trade show stands, print media, advertisements, bill boards, radio jingles, drama
278

productions, leaflets, seminars, workshops, clinics, and talk show appearances (KBIR, 2009).
KBIR would further public promotions by introducing an annual Revenue Week and a taxpayer
appreciation event. A new Public Affairs and Taxpayer Services Department (PATS) will
conduct an annual taxpayer perceptions surveys. In addition, this department will allow KBIR
to:
“Capture data and provide feedback on taxpayer inquiries and complaints; Engage with
taxpayers and the general public in order to secure their recommendations and feedback on
service delivery in the BIR operations; Conduct regular surveys to determine level of
satisfaction and public confidence in the service delivery by the BIR; Sensitize the BIR Staff
on the new procedures for handling taxpayer inquiries and complaints while executing their
work; Institute a customer care monitoring and evaluation system” (KBIR, 2009).
Kano State Board of Internal Revenue’s tax generation capacity has been hindered
through the deficient political support and public awareness and compliance with the tax
payment process. However, as in Lagos, the KBIR identifies legislation (granting the board
political and fiscal leadership) and public engagement as prescriptions. Through instilling public
confidence in KBIR and the tax system, the tax administration aims to shift Kano’s tax culture
from indifferent and adversarial to compliant.

A Tale of Two Tax Cultures: Elite and Citizen Perceptions in Lagos and Kano
The previous analysis of civil servant interviews and reports compares Lagos and Kano’s
tax administration from the perspective of those actively engaged in the two revenue boards. In
light of these characterizations, I consider elite and citizen perceptions of taxation in these two
states. Using original legislative interviews, I compare legislators in Lagos and Kano: how do
they perceive state (and Nigeria’s) reliance on oil revenue? Do legislators believe the tax
department has the right to compel tax payment? Which level of government should engage in
279

income tax collection? Based on the analysis in the previous section, I hypothesize legislators in
Lagos will be more likely than their counterparts in Kano to identify tax revenue as superior to
oil income (1). Legislators in Lagos will be more likely to indicate the state government has the
primary responsibility for collecting income taxes (2). Last, legislators in Lagos will be more
likely to agree that the tax department has the right to compel tax payment (3). I expect these
relationships as a result of LIRS officials’ characterization of politicians as cooperative. LIRS
administrators identify political elites as supportive, placing tax policy on the public agenda and
collaborating with the tax administration tax generation. On the other hand, KBIR officials (for
now) describe politicians as indifferent and unsupportive of the tax administration.
Similarly, I expect citizens in Lagos State to be more willing to pay taxes when compared
to those in Kano State (4). I also expect that Lagosians will be more likely to report tax payment.
Since Governor Fashola has led the effort to empower LIRS, the state tax administration has
successfully engaged in public awareness campaigns, educating citizens about the benefits of
tax-based development and the process of tax payment. LIRS has also worked to bring the
informal sector under the tax net. On the other hand, in Kano, Governor Kwankwaso has
focused state efforts on augmenting oil income dependence, rather than generating tax income.
As a result, KBIR have yet to launch such a campaign, instead describing citizens as unaware
and unknowledgeable about the tax contract. I hypothesize that Lagosians, having a firmer
understanding of the tax contract, will be more compliant with tax payment.

280

Comparing Legislative Perspectives of Tax Dependence and Payment
Relying on the data collected in interviews with members of the Lagos and Kano Houses
of Assembly, evidence corroborates civil servants’ characterizations of their respective state
political elites. We observe that, generally, legislators in both states prioritize tax dependence
over reliance on federal petroleum income transfers. However, in Kano, this belief does not
transfer into support of the state tax administration’s right to compel and collect taxes.
All 20 of the legislators interviewed at the Lagos House of Assembly indicate that
Nigeria is too reliant on oil revenue and that diversifying via tax revenue would be a more
optimal path.
According to one Lagos State legislator: “for now, Nigeria has a monotonous economy.
Taxes and a strong tax regime are necessary” (Lagos State Honourable #2, Lagos State House of
Assembly, Personal Communication, May 10, 2011). Another official agrees, arguing: “Nigeria
is stupidly reliant on oil, thinking it’s something in perpetuity. Petrol decreases as it’s utilized,
until its dried up. It’s myopic. Nigeria must exploit other areas that allow us to build the tax
base” (Lagos State Honourable #3, Lagos State House of Assembly, Personal Communication,
May 10, 2011). In an interview, one legislator separates Lagos from the remainder of the
federation, reporting: “Yes, Nigeria is too reliant on oil, but not in Lagos. There is too much
laziness in other states. People must learn that it’s best for them to contribute to their
government and pay taxes” (Lagos State Honourable #6, Lagos State House of Assembly,
Personal Communication, May 11, 2011). Finally, another legislator from the Lagos House of
Assembly laments Nigeria’s prospects for development: “It is only through taxation that we can
have the resources to govern. If we are unable to raise internally generated revenue with tax,

281

Nigerian cannot stay above water. It is sad for development” (Lagos State Honourable #7, Lagos
State House of Assembly, Personal Communication, May 12, 2011).
Likewise, 18 out of 20 legislators from Kano assert Nigeria is over-reliant on oil income
102

to the country’s detriment.

As one legislator states: “If oil goes, there will be no money to

sustain the country. In fact, it is the military government that bred this culture. People must
contribute to government in order for government to perform its functions (Kano State
Honourable #1, Kano State House of Assembly, Personal Communication, March 9, 2011).
Another member of the Kano State Assembly agrees saying: “States need to tax so government
can provide infrastructure and maintain it. But internally generated revenue is meager. Without
federal accounts, states would collapse” (Kano State Honourable #13, Kano State House of
Assembly, Personal Communication, July 6, 2011). In another interview, a Kano legislator
remarks: “Truly, there are no earnings from taxes. Taxes are the backbone of the country, but no
one wants to pay. We have forgotten other means of revenue besides oil” (Kano State
Honourable #4, Kano State House of Assembly, Personal Communication, March 9, 2011). Last,
an official specifies that Northern Nigeria is particularly worse off as a result of dependence on
oil income: “There is too much dependence on oil, especially in Northern Nigeria. In the North,
we are behind other states in taxation. I don’t even think internally generated revenue is a
priority. This has made Nigerians lazy and to abandon innovation. Without federal allocation,
most states and local governments would fail (Kano State Honourable #6, Kano State House of
Assembly, Personal Communication, March 8, 2011).

102

Two legislators from the Kano State House of Assembly indicate that they “do not know.”
282

Thus, contrary to Hypothesis 1, evidence from legislative interviews suggest that officials
from both Lagos and Kano say that they prioritize tax-based development as a superior option.
The majority of legislators from the two states agree that generating income from taxes should be
a priority.
With that said, turning to Hypothesis 2, we find a divergence in legislators’ beliefs about
which level of government should collect (income) taxes. In Lagos, officials’ opinions about tax
generation being a priority also align with their belief in the state’s primacy as collectors of
income taxes from citizens. In fact, this responsibility constitutionally belongs to states (as
opposed to the federal government, local government, traditional leaders, or members of the
community).

103

Evidence suggests this is not the case in Kano (Figure 6.1).

For example, in Lagos 13 out of the 20 legislators (65%) believe the state government has
the primary responsibility for collecting income taxes. This is in line with Nigeria’s Federal
Constitution and the 1993/2007 Personal Income Tax Act, in which collection of income taxes is
under the jurisdiction of state governments. There are exceptions for federal employees,
members of the military, and residents of the Federal Capital Territory (Abuja), who pay income
taxes to the federal government. On the other hand, six out of 20 legislators (30%) in the Lagos
State Legislature believe the federal government should collect income taxes, while the
remaining member (5%) believes local governments should be the primary collector.
Conversely, in Kano, state legislators are more likely to place the federal government in
the role of primary collector of income taxes. 16 out of 20 Kano legislators (80%) indicate that
the federal government should have the primary responsibility of collecting income taxes. Two
103

Chapter Two, Legislatures and Government Revenue in Nigeria: Resurgence of Political and
Fiscal Federalism, Nigeria’s Tax System: Mobilizing Revenue At Three Levels of Government for
a longer discussion of federal, state, and local tax jurisdictions.
283

legislators (10%) believe the state government should have this task, while the remaining two
members (10%) place this responsibility at the local level.
A Pearson’s corelation (r = 0.57) indicates a strong, positive association: legislators from
Lagos are more like than those from Kano to believe state government has the primary
responsibility for collecting income taxes (Table 6.4).

104

The observation that legislators from

Kano believe the federal government should collect income taxes corresponds with Suberu’s
(2001, 2004) argument: Northern governments prefer a centralized distributive system.
According to Suberu (2004: 341), Northern political leaders support the collection of taxes and
revenue at the federal level, with transfers to subnational authorities. In fact, one legislator from
Kano comments: “By definition, tax revenue is any money derived from the execution of the law.
The federal government should collect everything, then states can collect from the Federation
Account.” (Kano State Honourable #15, Kano State House of Assembly, Personal
Communication, July 6, 2011).
Thus, we find find support for Hypothesis 2: Lagosian legislators report being more
supportive of the state’s power to collect income taxes.
When testing Hypothesis 3, we observe a similar relationship when considering elite
perceptions of the tax department’s authority to compel tax payment (Figure 6.2). 20 out of 20
(100%) legislators from Lagos agree or strongly agree that “the tax department always has the
right to make people pay taxes.” On the other hand, in Kano, there’s less support for the tax
department’s right to compel tax payment. 12 out of 20 members (60%) of legislators from
Kano indicate they disagree or strongly disagree that “the tax department always has the right to
104

STATE is a dummy variable where Kano = 0 and Lagos = 1. Similarly, PRIMARY
RESPONSIBILITY FOR COLLECTING INCOME TAXES is a dummy variable where Federal
Government, Local Government = 0 and State Government = 1.
284

make people pay taxes.” The remaining eight members (40%) agree or strongly agree with this
statement. But even then, only three out of 20 (15%) legislators interviewed strongly agree with
the tax department’s authority to compel citizens to pay. Another analysis of correlation (Table
6.5) confirms legislators from Lagos are more likely than their counterparts in Kano to agree (or
105

strongly agree) that the tax department has the authority to compel citizens to pay taxes.

With

a Pearson’s r of 0.69, this is a strong, positive association.
Empirical analysis confirms the assertions made by tax administrations in Lagos and
Kano States. Contrary to Hypothesis 1, elites from both Houses of Assembly express the belief
that tax-led development is superior to dependence on petroleum income. However, supporting
Hypothesis 2, I find that in Lagos, legislators also support the state government’s authority to
collect income taxes and the tax department’s right to compel tax payment. This is not the case
in Kano, where elites are more likely to indicate the federal government has the responsibility for
collective income taxes. Likewise, legislators in Kano are less likely to support the tax
department’s authority to make citizens pay taxes (Hypothesis 3). This political support (or lack
thereof, as in the case of Kano) has real consequences for the tax administration’s ability to
educate citizens about taxation and execute tax policy. In Lagos, as a result of political
leadership and initiative, taxation and reform has been placed on the public agenda, providing
legal and financial resources to LIRS. This has not yet occurred in Kano, where KBIR officials
describe politicians as indifferent and at times adversarial toward the state tax administration.

105

STATE is a dummy variable where Kano = 0 and Lagos = 1. TAX DEPARTMENT CAN
COMPEL TAX PAYMENT is an interval variable where, Strongly Disagree = 1, Disagree = 2,
Neither Agree, Nor Disagree = 3, Agree = 4, Strongly Agree = 5.
285

Comparing Nigerians’ Willingness to Pay Taxes and Actual Tax Payment
Returning to the two hypotheses regarding citizens’ attitudes and behaviors toward
taxation, I hypothesize citizens in Lagos State to be more willing to pay taxes than citizens in
Kano State (4). I form these expectations following the interviews with members of the Lagos
and Kano state tax administrations. While LIRS has been able to engage in public awareness
campaigns with citizens, KBIR has not yet done so. In fact, according to KBIR officials, citizens
in Kano remain generally unknowledgeable about the tax contract. As in Chapter Five, I rely on
Nigerian public opinion data from Afrobarometer Round 4 (2008) and Project Oxford/DFID
(2011) survey to test these hypotheses.
Results from the Afrobarometer (2008) provide evidence that citizens of Lagos are more
willing to pay taxes. Lagosians also report actual tax payment at a higher rate. Looking at
Figure 6.3, 59% of respondents from Lagos agree or strongly agree that the tax department has
the right to make people pay taxes. On the other hand, 36% of respondents from Kano report
this opinion. Rather, 45% of people from Kano disagree or strongly disagree with the tax
department’s authority to compel tax payment (compared with of respondents from 36% in
Lagos).
Comparing respondents’ reported tax payment (Figure 6.4), we see Lagosians are more
likely to report actual tax payment in all three types of levies. 44% of respondents from Lagos
say that they have paid income taxes in the last year, compared to 18% of respondents from
Kano. 45% of Lagosians indicate paying property taxes in the last year, while only 5% of
respondents from Kano say they made property tax payments. Last, 42% of respondents from
Lagos report paying local government taxes, whereas 24% from Kano also paid local levies.
Considering Nigeria as a whole, Lagos rates of tax payment exceed those across the federation.
286

On the other hand, reports of actual tax payment in Kano are below the national average.

Robustness Check with Alternative Data: Willingness to Pay Taxes in Urban Lagos and
Kano
I also explore the influence of a resident’s state on her willingness to pay taxes using
results of the Oxford/DFID survey in urban Nigeria. Once again, comparing Lagos and Kano, I
expect Lagosians to be more willing to pay taxes.
“Willingness to pay taxes” remains the key dependent variable (Figure 6.5).

106

This is

measured with the following question: (in regard to their State government) “Given the current
government’s performance that you have seen, do you find yourself more willing or less willing
to pay taxes than you were during the last civilian administration…or would you say your
107

attitude toward taxes is about the same?”

In Lagos, 44% of respondents report being more

willing to pay taxes, while 36% say they feel about the same now as under the last administration.
On the other hand, only 9% of respondents from Kano are more willing to pay taxes now, while
52% feel about the same.
The main independent variable is a dummy variable that captures a respondent’s state,
either Lagos or Kano.

108

I hypothesize there will be a positive relationship, whereby

106

The Oxford/DFID survey does not include questions asking the respondents about specific
payment of taxes.
107

Given the current government’s performance that you have seen, do you find yourself more
willing or less willing to pay taxes than you were during the last civilian administration in _____
State? Or would you say your attitude toward taxes is about the same?” 1=Less Willing,
2=About the Same, 3=More Willing
108

In this variable, Kano = 0 and Lagos = 1.
287

respondents from Lagos will be more willing to pay taxes than citizens of Kano. I make this
expectation again, based on interviews with civil servants in the Lagos and Kano State tax
administrations. Since LIRS officials have been more successful with educating citizens about
the process of taxation, this will translate into a greater willingness to comply with payment at
the individual level.
Since results of Chapter Five confirm that satisfaction with government performance
influences Nigerians’ willingness to pay taxes, I include this set of variables as controls. In this
case, satisfaction with government performance is operationalized using five individual
indicators: an indicator asking respondents how well (or badly) they think the current state
administration is spending tax revenue;

109

an indicator asking respondents to assess the current

conditions of the roads, public schools, government hospitals and clinics, and the water
supply.

110

I expect when a Nigerian has more favorable evaluations of how the state

government manages these areas, she will also be more willing to pay taxes.
I also include several control variables: I expect urban Nigerians with higher household
111

incomes will have a greater ability to pay taxes and will be more willing to pay taxes.

I also

109

“Here in _____ State, how satisfied are you with the way that the current administration has
spent tax revenue?” 1=Not at all satisfied, 2=Not very satisfied, 3=Somewhat satisfied, 4=Very
satisfied
110

“Compared with the past administration in ______ State, have the following things in this
area gotten better, worse, or have they stayed about the same? Condition of roads? Condition of
public schools? Condition of government hospitals and clinics? Condition of water supply?”
1=Worse, 2=About the same, 3=Better
111

“Please, can you estimate the monthly income for your whole household?” 1=I don’t earn an
income myself—other people support me, 2=Less than N5,000, 3=N5,000-N10,000, 4=N11,000288

hypothesize older urban Nigerians will be more wiling to pay taxes. Finally, I control for
corruption: a respondent who reports experiences with government corruptions with be less
willing to pay taxes. Table 6.6 presents this model’s summary statistics.
I utilize ordered logistic regression with robust standard errors to perform analysis, and
results are reported in Table 6.7 (proportional odds ratios). As hypothesized, respondents from
urban Lagos are more willing than those in urban Kano to pay taxes (POR = 3.157). In fact, a
respondent’s home state is the most powerful predictor in her willingness to pay taxes. Holding
all other variables constant, the likelihood a respondent will report being more willing to pay
taxes is 3.157 times greater if she respondent is from Lagos (as opposed to Kano). This finding
corroborates accounts from LIRS and KBIR administrators: as a result of LIRS efforts, citizens
in Lagos are more engaged in discourse about taxation. Understanding the role taxes play in
development and procedures for payment, respondents from Lagos are more willing to comply
with payment. On the other hand, in Kano, tax administrators have been unable to educate
residents about the tax contract. As a result, citizens in Kano are less willing to pay taxes. This
analysis confirms that these relationships still exist when we consider urban settings. Taking
satisfaction with government performance, income, age, and corruption into account does not
diminish the positive and significant relationship between respondents’ state and their
willingness to yield to the tax authority. Analysis of Nigerian public opinion data affirms the
perceptions expressed by members of Lagos’ and Kano’s State tax administrations.

N15,000, 5=N16,000-N25,000, 6=N26,000-N35,000, 7=N36,000-N45,000, 8=N46,000-N55,000,
9=N56,000-N75,000, 10=N75,000-N95,000, 11=N96,000-N120,000, 12=More than N120,000.
289

Conclusion
This chapter suggests that differences in political leadership can help explain why Lagos
and Kano have two different tax cultures, which ultimately results in two different rates of tax
generation. Evidence from interviews with tax administrators suggests that in Lagos, Governor
Fashola has lead efforts of create a relationship of cooperation and compliance between political
elites, tax administrators, and citizens in Lagos. On the other hand, Governor Kwankwasu in
Kano has prioritized increasing the state’s access to natural resource income. As a result,
relationships between legislators, tax officials, and citizens in Kano are based on indifference
and a lack of engagement. The nature of these relationships has real consequences for tax
generation in Lagos and Kano: in comparison to citizens of Kano, Lagosians are more willing to
pay taxes and report making payments at higher rates. For example, by opening up dialogue
between political elites, tax administrators, and citizens, Governor Fashola has placed the
taxation and notions of the fiscal contract on the public agenda. Through this engagement, LIRS
officials have bargained with citizens, publicizing the individual and societal benefits of tax
payment. Analysis of public opinion data provides evidence that citizens of Lagos have
responded to this campaign with increased tax compliance. This suggests that political initiative,
particularly from governors, can give momentum to developing state tax capacity.

290

Chapter Six Tables and Figures

Table 6.1: STATE
DEMOGRAPHICS
(2005 – 2006)

STATE
LAGOS
KANO

POPULATION
AS % OF
AREA
POPULATION
TOTAL
(SQUARE
(2006)
NIGERIA
MILES)
9, 013, 534
6.44
1,341.70
9, 383, 682
6.70
7,772.60

291

Table 6.1 (cont’d):
STATE
DEMOGRAPHICS
(2005 – 2006)

STATE
LAGOS
KANO

%
POPULATION
ENROLLED
IN PUBLIC
PRIMARY
SCHOOL
13.95
13.76

%
POPULATION
ENROLLED IN
PUBLIC
SECONDARY
SCHOOL
15.98
12.40

292

POPULATION
PER
PRIMARY
HEALTH
CARE
FACILITY
8,160
10,070

%
POPULATION
EMPLOYED
37.19
36.82

Table: 6.2
DISTRIBUTION OF
WORKING
POPULATION BY
ACTIVITY (%)
AGRICULTURE
FISHING

LAGOS
2
1

KANO
25
1

MANUFACTURING
CONSTRUCTION
TRADE
TRANSPORT

8
5
40
7

12
2
33
2

7
5

2
3

3
20
3

1
14
2

PUBLIC
ADMINISTRATION
EDUCATION
HEALTH/SOCIAL
WORK
SERVICES
OTHER

Table 6.3: TAX
GENERATION
IN LAGOS AND
KANO (Tax
Revenue As A
Percentage of
Total Revenue, %)
STATE
LAGOS
KANO

2005
51.2
11.2

2007
62.2
14.7

293

2008
63.5
17.2

2009
64.3
21.6

MEAN
(1999 – 2009)
53.4
13.2

112

Table 6.4 :
CORRELATION
BETWEEN
LEGISLATOR'S
HOME STATE AND
BELIEF STATES
SHOULD COLLECT
INCOME TAXES (%,
n = 40)

STATE
KANO
LAGOS

112

!!

FEDERAL
80
30

2

WHICH LEVEL OF
GOVERNMENT
SHOULD COLLECT
INCOME TAXES
LOCAL
10
5

Pearson Chi (2) = 13.5884, Pr = 0.001; Pearson’s r = 0.568.
294

!!

STATE
10
65

Table 6.5:
CORRELATION
BETWEEN
LEGISLATOR'S
HOME STATE AND
BELIEF TAX
DEPARMENT CAN
COMPEL PAYMENT
113
(%, n = 40)

!!

!!

!!

!!

TAX DEPARMENT
CAN MAKE PEOPLE
PAY TAXES
!!

!!

!!

STATE
KANO
LAGOS

113

STRONGLY
DISAGREE

DISAGREE

30
0

30
0

2

Pearson Chi (3) = 20.0000, Pr = 0.000; Pearson’s r = 0.685.
295

NEITHER AGREE
NOR DISAGREE
0
0

AGREE
25
25

STRONGLY
AGREE
15
75

Table 6.6: URBAN
NIGERIA, SUMMARY
STATISTICS-WILLINGNESS TO PAY
TAXES IN LAGOS AND
KANO

!!

VARIABLE

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

364

2.017

0.747

1

3

364

0.548

0.498

0

1

SATISFACTION WITH
GOVT. TAX SPENDING

364

2.696

0.980

1

4

SATISFACTION WITH
ROAD MAINTENANCE

364

2.179

0.819

1

3

SATISFACTION WITH
PUBLIC SCHOOL
MAINTENANCE

364

2.223

0.723

1

3

SATISFACTION WITH
HOSPITAL
MAINTENANCE

364

2.267

0.719

1

3

WILLINGNESS TO PAY
TAXES
STATE

!!

!!

296

!!

!!

Table 6.6 (cont’d):
URBAN NIGERIA,
SUMMARY
STATISTICS-WILLINGNESS TO
PAY TAXES IN
LAGOS AND KANO

!!

VARIABLE

OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

364

1.960

0.757

1

3

364

6.697

2.336

1

12

364

3.194

1.172

1

7

364

1.233

0.423

1

2

SATISFACTION
WITH WATER
SUPPLY
HOUSEHOLD
INCOME
AGE
EXPERIENCE WITH
GOVERNMENT
CORRUPTION

!!

297

!!

!!

!!

Table 6.7: URBAN
NIGERIA,
WILLINGNESS TO
PAY TAXES IN
LAGOS AND KANO

WILLINGNESS
TO PAY
TAXES
3.157*
(0.717)

STATE
SATISFACTION WITH
GOVT. TAX
SPENDING
SATISFACTION WITH
ROAD
MAINTENANCE
SATISFACTION WITH
PUBLIC SCHOOL
MAINTENANCE
SATISFACTION WITH
HOSPITAL
MAINTENANCE
SATISFACTION WITH
WATER SUPPLY
HOUSEHOLD
INCOME
AGE

2.309*
(0.281)
1.338*
(0.143)
0.940*
(0.172)
1.247*
(0.236)
1.075*
(0.169)
0.974*
(0.145)
1.125*
(0.099)

EXPERIENCE WITH
(-)0.945
GOVERNMENT
(0.392)
CORRUPTION
LOG LIKELIHOOD
-336.113
N
364
Notes: Values are proportional odds ratios, robust standard errors in
parentheses.
*p < 0.05, two-tailed test.

298

Figure 6.1: “Who do you think actually has primary responsibility for collecting income taxes?”

20

# OF LEGISLATORS

18
16
14
12
10
8
6
4

LAGOS

2

KANO

0

299

Figure 6.2: ELITES “The tax department always has the right to make people pay taxes.”
20
18
16

# OF LEGISLAORS

14
12
10

LAGOS

8

KANO

6
4
2
0
STRONGLY
DISAGREE

DISAGREE

NEITHER
AGREE NOR
DISAGREE

300

AGREE

STRONGLY
AGREE

Figure 6.3: CITIZENS “The tax department always has the right to make people pay taxes.”

114

45
40
35
30
%

25
LAGOS

20

KANO

15
10
5
0
STRONGLY
DISAGREE

114

DISAGREE

NEITHER
AGREE NOR
DISAGREE

2

Pearson Chi (4) = 39.8884, Pr = 0.000.
301

AGREE

STRONGLY
AGREE

Figure 6.4: “Have you had to make any of the following payments during the past year…”

115

50
45
40
35

%

30
LAGOS: YES

25

KANO: YES

20

NIGERIA: YES

15
10
5
0
INCOME TAXES

115

PROPERTY
TAXES

LOCAL TAXES

2

2

INCOME TAXES: Pearson Chi (1) = 24.1717, Pr = 0.000; PROPERTY TAXES: Pearson Chi (1) = 62.2832, Pr = 0.000;
2

LOCAL TAXES: Pearson Chi (1) = 20.4561, Pr = 0.000.

302

Figure 6.5: “Given the current government’s performance that you have seen, do you find yourself more willing or less willing
to pay taxes than you were during the last civilian administration in _____ State? Or would you say your attitude toward taxes
is about the same?”

60

50

%

40
LAGOS

30

KANO
20

10

0
LESS WILLING

ABOUT THE SAME

303

MORE WILLING

CONCLUSION

Beyond Nigeria: Explaining Africans’ Attitudes and Behavior Toward Taxation
Background
In the previous chapters, analyses of government budgeting priorities, elite and popular
attitudes, and interviews with civil servants demonstrate that standard theory of revenue and
representation is applicable in the African context. Bargaining between government and citizens
seems to occur and an implicit fiscal constant apparently results. First, governments in Nigeria
with a greater capacity to generate income from non-oil taxes tend to be more representative of
their constituents. In this instance, elites shift public policy and expenditures to meet the
interests of their citizens. Elected leaders do so in order to maintain the source of tax income.
Second, Levi’s (1988) theory of quasi-voluntary compliance can also explain popular attitudes
toward taxation in Nigeria. Ordinary Nigerians are more willing to pay taxes (and follow
through with actual payment) when satisfied with the level of representation received from
government. Moreover, when a Nigerian trusts that fellow citizens are complying with tax
payment, she is more likely to yield to the tax department’s authority and report making tax
payments in the last year. Finally, I have defined the quality of representation to include
evaluations of government performance. Nigerians who are more satisfied with government
provision of public services are also more compliant with tax payment. But do results based on
the experience of Nigeria travel elsewhere?
Moving beyond Nigeria, Afrobarometer Public Opinion data allows us to perform an
exploratory investigation about citizen perceptions of the tax contract in other countries across
304

116

the continent.

Across Africa, under what conditions are ordinary people willing to pay taxes

and make actual tax payments? I hypothesize that the findings from Nigeria also apply to other
African countries. Generally, Africans will be more willing to pay taxes when: (1) they are
satisfied that government is fulfilling its end of the bargain, providing representation; (2) they
trust that other countrymen are complying with tax payment; (3) they are satisfied with
government provision of public services. I also expect positive relationships between
perceptions of representation, trust, satisfaction with service provision and the likelihood
Africans will follow through with tax payments. I form these hypotheses based on Levi’s (1988)
theory of citizen quasi-voluntary.
In order to test these hypotheses, I rely on public opinion results from Round Four of the
Afrobarometer (2008). In addition to Nigeria, the Afrobarometer conducted surveys in 19 other
African countries in 2008 – 2009.

117

Because the Afrobarometer asks respondents across all

these countries about perceptions of taxation, representation, and government performance, I am
able to investigate these hypotheses cross-nationally.
I replicate the analysis I conducted in Chapter Five, whereby Africans’ theoretical
willingness to pay taxes and their reported payment of local, property, and income taxes are the
dependent variables. Once again, the explanatory variables I include capture various facets of
Africans’ (1) perceptions of representation, (2) trust, and (3) government performance. I expect
a positive relationship between Africans’ perceptions of the quality of representation from local

116

The analysis in this chapter is an expansion of Chapter Five: Nigerian Perceptions of
Taxation and Representation: Attitudes and Behaviors To Tax Payment.
117

Countries included in Afrobarometer Round Four: Benin, Botswana, Burkin Faso, Cape
Verde, Ghana, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali , Mozambique, Namibia,
Nigeria, Senegal, South Africa, Tanzania, Uganda, Zambia, Zimbabwe (20).
305

and national officials and their willingness to pay taxes. I hypothesize these indicators will also
yield a positive relationship with Africans’ actual payment of taxes. I expect that when Africans
express trust of others, they will also be more willing to pay taxes. I also expect a positive
relationship between respondents’ trust of others and actual payment of taxes. Last, I
hypothesize that Africans who have more positive evaluations of how their government is
managing the provision of services will be more likely to believe government is fulfilling its end
of the bargain. As a result, they will be more willing to pay taxes and report making such
payments.
As in Chapter Five, I include the control variables for employment status (positive
influence on willingness to pay taxes and actual tax payment); age (positive relationship between
age and willingness to pay taxes); and perceptions of corruption within the local government,
National Assembly, and among tax officials (negative relationship between Nigerians’
perceptions of corruption and wiliness to pay taxes and actual tax payment). In order to control
for cross-national differences, I include dummy variables for each country. For example, in the
Benin dummy variable, all respondents from Benin receive a value of “1”, and all other
118

respondents are given “0.”

118

Originally, Round Four of the Afrobarometer includes 27,713 respondents. However, after
excluding respondents who answer “don’t know” or refused to answer any of the questions
utilized for dependent, independent, or control variables, the sample size drops to 15,485. After
comparing demographic descriptors in both samples, I find that the new sample has 4% more
urban respondents than in the old sample (Table 7.4). In addition, while the old sample is split
evenly between male and female respondents, there is an 8% gap in the new sample (Table 7.5).
In order to control for these two shifts, I add controls for urban vs. rural dwellers and gender to
my model. Table 7.1 presents each model’s summary statistics.
306

Analysis and Results
For analysis, I again rely on ordered logistic and logistic regressions with robust standard
errors. Table 7.2 presents results.
The four analyses presented gauge the effect of Africans’ perceptions of quality of
representation, trust, and government performance on willingness to pay taxes and actual tax
payment. As hypothesized, evidence confirms the theory of quasi-voluntary compliance
explains Africans’ attitudes and behaviors toward taxation across the continent. There is a strong
and positive relationship whereby Africans are more willing to pay taxes when they judge
elected officials to be representing their interests. Second, Africans who perceive a higher level
of representation are also more likely to comply with actual tax payment. As the second tenet of
quasi-voluntary compliance theorizes, in Africa, respondents with higher levels of trust are also
more likely to report being more tax compliant, in theory and practice. Finally, when an African
is more satisfied with government provision of services, she is also more willing to pay taxes and
report making those payments in the last year. These findings directly support my hypotheses.
Across multiple measures, we see Africans make decisions about taxation directly in relation
how they evaluate representation, trust, and government performance.
Comparing Nigeria with the pooled analysis, we can see differences in the factors that
ordinary people rely on to determine their tax compliance. For example, in Chapter Five’s
analysis of Nigerian attitudes toward tax payment, satisfaction with government management of
roads and electricity are the most powerful predictors. In addition, when a Nigerian’s perceives
that she can make her local councillor and/or national representative listen to her, she is more
willing to pay taxes and report doing so. These three elements are all significant, positive
predictors of tax compliance in the larger African sample. However, in the pooled analysis,

307

contact with local representatives and members of Parliament are more salient. Holding all other
variables constant, a one-unit shift in an African’s reported contact with her local councillor (e.g.
moving from making “only once” to “a few times”) increases the likelihood that she will report
paying local taxes by 1.161 times. This shift in contact with a local representative also increases
the likelihood of property tax payment (1.187 times greater) and income tax payment (1.148
times greater). Contrary to Nigerian respondents, representation from national officials plays a
bigger role in influencing tax payment in the pooled analysis. For example, in Nigeria, contact
with a national representative is not a significant predictor of actual tax payment. However, in
the cross-national sample, contact with a member of Parliament is among the most influential
factors. An African who has more frequent contact with her national representative is 1.177
times more likely to report paying local taxes. She is also more likely to pay property tax (POR
= 1.112) and income taxes (POR = 1.154). This difference could be a result of different
governing systems. While Nigeria has a federal structure of governance, other countries in
Africa are unitary states. In Nigeria, where state (and to an extent local) governments have
authority to make laws, govern, and tax, citizens seem to rely more on evaluations of subnational
representatives when gauging their attitudes toward taxation. On the other hand, in unitary
systems, where the central government holds complete law-making authority (e.g. South Africa,
Malawi, Zambia), ordinary citizens may reference their national representatives as their focal
point. In this case, a citizen would use quality of representation from a member of parliament to
determine tax compliance.
When we consider the variation in willingness cross-nationally, dummy variables reveal
more divergence in attitudes toward tax payment. In Table 7.3, we see that the coefficients for
country dummy variables representing Benin, Burkina Faso, Cape Verde, Ghana, Liberia, Mali,

308

Senegal, and Uganda are significant and positive. Generally, citizens in these countries seem to
be more willing to pay taxes when compared to Africans from other countries. On the other
hand, the coefficients for Kenya, Lesotho, Madagascar, Mozambique, South Africa, Zambia, and
Zimbabwe are significant and negative. In these countries, ordinary people appear to be less
willing to agree that the tax department has the right to compel payment. Why do citizens of
West African countries seem to be more willing to pay taxes than their counter-parts in Southern
Africa? How have political elites approached tax administration and the fiscal contract in these
two regions? Is the authority to collect taxes shared between levels of government (as in
Nigeria) or concentrated at the national level? How have these differences shaped representation
and expectations between elected officials and citizens? Future research is needed to investigate
these questions and tease out explanations for these cross-national differences.
Quasi-voluntary compliance not only explains attitudes and behaviors in Nigeria, but also
provides leverage in analysis across Africa. These findings, taken together, provide evidence in
support of my hypotheses. Western European theories of a tax contract also apply in Africa. In
Nigeria and across Africa, citizens have expectations of their local and national governments.
Evaluations of these expectations correspond with attitudes and behavior towards taxation. As
with in other contexts, African taxpayers appraise their leaders’ performance with both
substantive representation of citizens’ interests and public service delivery. Once determining
that government has fulfilled it’s part of the bargain, Africans across the continent are also more
willing to pay taxes and actually make such payments. Moreover, analysis of public opinion
data affirms that Africans who trust that others are also paying taxes will be more compliant with
payment of taxes. Quasi-voluntary compliance provides analytical leverage in explaining how
Africans perceive taxation and their relationship to government in light of the tax contract.

309

Curing the Curse?: Moving Forward in Nigeria
This research demonstrates the role revenue plays in the development of elite-citizen
relationships, both historically and in contemporary Nigeria. In fact, politicians’ quest to access
revenue and control of the primary economic engine have motivated decisions to either
concentrate power in the federal government or diffuse it through state and local levels. As
Nigeria’s democratic Fourth Republic emerges from military governance, balancing fiscal
authority and the distribution of revenue remain key issues. Furthermore, relationships between
elected leaders and citizens vary across the federation in response to government access to
various income streams.
In Western Europe, representative institutions emerged as a result of bargains between
rulers and citizens: tax income in exchange for a greater role in policy-making. However, in
order to consider these theories in non-Western contexts, one must account for the fact that elites
elsewhere also have access to income other than taxes, specifically revenue from natural
resources. This is the case in Nigeria, where reliance on taxes and petroleum income creates
varied subnational revenue profiles. Relying on theories of representation and using this
subnational variation, I have tested how tax vs. oil dependence influences government
representation in Nigeria.
In Chapter Three, using budgetary data between 1999 and 2009, evidence confirms that
in Nigeria, as local and state government dependence on tax income increases, spending on
government salaries and allowances decreases. Instead, these governments expend revenue on
public service. On the other hand, reliance on natural resource income via federal transfers
corresponds with increased recurrent expenditures on salaries and allowances.

310

After establishing how taxation and representation interact at a macro-level in Nigeria
(tax reliance shaping budgetary priorities), Chapter Four evaluates these linkages regionally and
at the micro-level. Analysis of original interviews with individual elites in six states (across two
regions) demonstrates that Nigerian legislators in tax-reliant states are more likely to represent
citizens. These officials spend a greater percentage of their time performing constituency
services and prioritize constituents’ preferences above other considerations (e.g. party position,
personal views, etc).
With the understanding of how revenue shapes elite behavior, in Chapter Five I evaluate
ordinary Nigerians’ attitudes and behavior towards the tax contract. Relying of public opinion
data from the Afrobarometer and a new survey of urban Nigerians, I investigate conditions under
which citizens are the most tax compliant. Ordinary Nigerians are more willing to pay taxes
when: (1) they perceive elected officials are representing public interests; (2) they trust that other
Nigerians are complying with tax payment; (3) they are satisfied with government provision of
services.
Chapter Six presents an inter-regional analysis of tax cultures, comparing Lagos and
Kano States. Though Lagos and Kano have similar levels of socio-economic development and a
comparable workforce, there is a large gap between in their tax generation. According to
interviews with members of the tax administrations in Lagos and Kano, political initiative
explains these differences. In Lagos, politicians have provided the state revenue board with the
resources necessary to place taxation on the public agenda. With this support, elites and LIRS
officials launched a political campaign, engaging Lagosians and educating them about the
benefits of tax-led development and the payment process. On the other hand, in Kano, as a result
of politicians’ indifferent (and at times adversarial) attitude, the state tax service has not received

311

the support or resources needed to engage in such a campaign. In this instance, citizens remain
largely uneducated about the tax contract. Using evidence from Lagos and Kano’s experience, I
untangle the temporal nature of the taxation-representation linkages between elites and citizens.
I argue that government makes the first move. In an attempt access to income taxes on citizens,
elites shift policy and expenditures to reflect popular preferences. Citizens, viewing credible
signals of government’s willingness to comply with their end of the tax contract, will also fulfill
their part, and make tax payments. This is also the case on Nigeria, where officials engage in
explicit bargains with citizens for tax revenue. Returning to the original elite interviews,
analyses of correlation confirm that elites from Lagos are more supportive of the state’s tax
administration when compared to their counterparts in Kano. Furthermore, across multiple
measures and data from two different citizen surveys, we see Lagosians are more tax compliant
than residents of Kano.
In this concluding chapter, I use Afrobaromter data to expand analysis of citizens’
attitudes and behaviors toward taxation. Evidence suggests that notions of the tax contract and
quasi-voluntary compliance have predictive power across Africa. Generally, Africans are more
willing to pay taxes and follow through with payment when they believe elected officials are
representing their interests. In a similar vein, satisfaction with government maintenance of
public services also positively influences tax compliance in Africa. However, explaining crossnational differences in citizens’ perceptions of the tax department’s authority to tax open up
other avenues for future study.
Taken together, this research provides strong evidence that revenue extraction (in all
variations) influences the development of representation in Africa. A relationship between
taxation and representative governance suggests that strengthening local and state governments’
312

tax administration and building citizen capacity to monitor government budgeting can also
bolster responsive and democratic governance in Africa. As Lagos State already demonstrates,
citizen civic education must continue as a tool for educating citizens about the taxation as a part
of citizenship. Representation includes both citizen/civil society pressure and government
responsiveness to those demands. Therefore, building capacity across these three different
groups is important to sustained collaboration—especially Africans’ capacity to interact with
government (including tax payment) (Wolpe and McDonald, 2006).

313

Conclusion Tables and Figures

Table 7.1: SUMMARY
STATISTICS--EFFECT OF
QUALITY OF
REPRESENTATION, TRUST,
AND GOVERNMENT
PERFORMANCE ON AFRICANS'
WILLINGNESS TO PAY TAXES
AND ACTUAL TAX PAYMENT
OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

PEOPLE MUST PAY TAXES

15,485

3.671

1.211

1

5

PAID INCOME TAX
PAID PROPERTY TAX
PAID LOCAL TAXES

15,485
15,485
15,485

0.258
0.273
0.289

0.438
0.445
0.453

0
0
0

1
1
1

MAKE LOCAL COUNCILLOR
LISTEN

15,485

1.693

1.074

0

3

MAKE MPs LISTEN

15,485

1.434

1.102

0

3

TIME LOCAL COUNCILLORS
SPEND LISTENING

15,485

1.138

0.96

0

3

TIME MPs SPEND LISTENING

15,485

0.887

0.889

0

3

15,485

0.589

0.959

0

3

15,485

0.271

0.694

0

3

VARIABLE

CONTACT LOCAL
COUNCILLOR
CONTACT MP

314

Table 7.1 (cont’d): SUMMARY
STATISTICS--EFFECT OF
QUALITY OF
REPRESENTATION, TRUST,
AND GOVERNMENT
PERFORMANCE ON
AFRICANS' WILLINGNESS
TO PAY TAXES AND ACTUAL
TAX PAYMENT
OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

TRUST OTHER NIGERIANS

15,485

1.306

0.999

0

3

SATISFACTION WITH
GOVERNMENT HEALTH
SERVICES

15,485

2.574

0.994

1

4

SATISFACTION WITH
GOVERNMENT WATER AND
SANITATION SERVICES

15,485

2.329

0.969

1

4

SATISFACTION WITH
GOVERNMENT
MAINTENANCE OF ROADS
AND BRIDGES

15,485

2.325

1.008

1

4

VARIABLE

315

Table 7.1 (cont’d):
SUMMARY STATISTICS-EFFECT OF QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE ON
AFRICANS' WILLINGNESS
TO PAY TAXES AND
ACTUAL TAX PAYMENT
OBSERVATIONS

MEAN

STANDARD
DEVIATION

MIN

MAX

SATISFACTION WITH
GOVERNMENT
ELECTRICICTY SERVICES

15,485

2.143

0.985

1

4

SATISFACTION WITH
LOCAL GOVERNMENT
MAINTENANCE OF LOCAL
ROADS

15,485

2.114

0.975

1

4

SATISFACTION WITH
LOCAL GOVERNMENT
MAINTENANCE OF LOCAL
MARKETS

15,485

2.222

0.943

1

4

EMPLOYMENT STATUS

15,485

1.671

1.627

0

5

VARIABLE

316

Table 7.1 (cont’d): SUMMARY
STATISTICS--EFFECT OF
QUALITY OF
REPRESENTATION, TRUST,
AND GOVERNMENT
PERFORMANCE ON
AFRICANS' WILLINGNESS
TO PAY TAXES AND
ACTUAL TAX PAYMENT
OBSERVATIONS

MEAN

AGE

15,485

35.098

STANDARD
DEVIATION
13.499

LOCAL GOVERNMENT
CORRUPTION

15,485

1.322

PARLIAMENT CORRUPTION

15,485

VARIABLE

TAX OFFICIALS
CORRUPTION
URBAN/RURAL
GENDER

MIN

MAX

18

99

0.828

0

3

1.308

0.814

0

3

15,485

1.492

0.875

0

3

15,485
15,485

1.588
1.464

0.492
0.497

1
1

2
2

317

Table 7.2: EFFECT
OF QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE
ON AFRICANS'
WILLINGNESS TO
PAY TAXES AND
ACTUAL TAX
PAYMENT

1

PEOPLE
MUST PAY
TAXES

2

3

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX

4

PAID
INCOME
TAX

MAKE LOCAL
COUNCILLOR
LISTEN

1.076*
(0.021)

1.025* (0.025)

1.043*
(0.027)

1.059*
(0.027)

MAKE MPs LISTEN

1.047*
(0.020)

1.041* (0.024)

1.005*
(0.024)

1.045*
(0.025)

TIME LOCAL
COUNCILLORS
SPEND LISTENING

1.085*
(0.024)

1.082* (0.027)

1.049*
(0.029)

1.079*
(0.029)

TIME MPs SPEND
LISTENING
CONTACT LOCAL
COUNCILLOR

1.039*
(0.022)
1.025*
(0.019)

1.039*
(0.029)
1.187*
(0.029)

1.105*
(0.031)
1.148*
(0.028)

1.107* (0.036)
1.161* (0.027)

CONTACT MP

1.001*(0.024)

1.177* (0.036)

1.112*
(0.035)

1.154*
(0.037)

TRUST OTHER
COUNTRYMEN

1.033*
(0.017)

1.088* (0.022)

1.040*
(0.021)

1.019*
(0.021)

318

Table 7.2 (cont’d):
EFFECT OF
QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE
ON AFRICANS'
WILLINGNESS TO
PAY TAXES AND
ACTUAL TAX
PAYMENT

1

2

3

PEOPLE
PAID LOCAL
PAID
MUST PAY GOVERNMENT PROPERTY
TAXES
TAXES
TAX
SATISFACTION
WITH
GOVERNMENT
HEALTH SERVICES
SATISFACTION
WITH
GOVERNMENT
WATER AND
SANITATION
SERVICES
SATISFACTION
WITH
GOVERNMENT
MAINTENANCE OF
ROADS AND
BRIDGES

4

PAID
INCOME
TAX

1.129*
(0.022)

1.030*
(0.024)

0.934*
(0.024)

0.959*
(0.025)

1.062*
(0.020)

1.065*
(0.024)

1.089*
(0.027)

1.053*
(0.027)

1.125*
(0.022)

1.036*
(0.025)

1.033*
(0.027)

1.078*
(0.029)

319

Table 7.2 (cont’d):
EFFECT OF
QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE
ON AFRICANS'
WILLINGNESS TO
PAY TAXES AND
ACTUAL TAX
PAYMENT

1

2

3

PEOPLE
PAID LOCAL
PAID
MUST PAY GOVERNMENT PROPERTY
TAXES
TAXES
TAX
SATISFACTION
WITH
GOVERNMENT
HEALTH SERVICES
SATISFACTION
WITH
GOVERNMENT
WATER AND
SANITATION
SERVICES
SATISFACTION
WITH
GOVERNMENT
MAINTENANCE OF
ROADS AND
BRIDGES

4

PAID
INCOME
TAX

1.129*
(0.022)

1.030*
(0.024)

0.934*
(0.024)

0.959*
(0.025)

1.062*
(0.020)

1.065*
(0.024)

1.089*
(0.027)

1.053*
(0.027)

1.125*
(0.022)

1.036*
(0.025)

1.033*
(0.027)

1.078*
(0.029)

320

Table 7.2 (cont’d):
EFFECT OF
QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE
ON AFRICANS'
WILLINGNESS TO
PAY TAXES AND
ACTUAL TAX
PAYMENT

1

2

3

PEOPLE
PAID LOCAL
PAID
MUST PAY GOVERNMENT PROPERTY
TAXES
TAXES
TAX

4

PAID
INCOME
TAX

LOCAL
GOVERNMENT
CORRUPTION

(-)1.062*
(0.026)

(-)1.006*
(0.030)

(-)1.012*
(0.033)

(-)1.091*
(0.034)

PARLIAMENT
CORRUPTION

(-)0.978*
(0.024)

(-)1.116*
(0.036)

(-)1.043*
(0.034)

(-)1.034*
(0.035)

TAX OFFICIALS
CORRUPTION

(-)0.973
(0.022)

(-)1.044
(0.028)

(-)1.026
(0.028)

(-)1.056
(0.030)

URBAN/RURAL

0.913*
(0.029)

0.733* (0.030)

0.429*
(0.018)

0.502*
(0.021)

GENDER

0.898*
(0.028)

0.753*
(0.029)

0.805*
(0.032)

0.813*
(0.033)

321

Table 7.2 (cont’d):
EFFECT OF
QUALITY OF
REPRESENTATION,
TRUST, AND
GOVERNMENT
PERFORMANCE ON
AFRICANS'
WILLINGNESS TO
PAY TAXES AND
ACTUAL TAX
PAYMENT

1

PEOPLE
MUST
PAY
TAXES
CONSTANT
LOG LIKLIHOOD
N

-20801.331
15,485

2

3

PAID LOCAL
PAID
GOVERNMENT PROPERTY
TAXES
TAX
0.470*
(0.071)
-8501.995
15,485

322

0.548*
(0.085)
-8042.136
15,485

4

PAID
INCOME
TAX
0.294*
(0.046)
-7694.818
15,485

Table 7.3: AFRICANS'
WILLINGNESS TO PAY
TAXES BY COUNTRY
(Country Dummy Variables)
WILLINGNESS
TO PAY TAXES
0.667*
0.088
0.786*
0.602*
0.951*
-0.211*
-1.071*
0.608*
-0.529*
-0.199
0.338*
-0.447*
-0.210
0.382*
-0.216*
-0.091
0.655*
-0.529*
-0.411*

COUNTRY
BENIN
BOTSWANA
BURKINA FASO
CAPE VERDE
GHANA
KENYA
LESOTHO
LIBERIA
MADAGASCAR
MALAWI
MALI
MOZAMBIQUE
NAMIBIA
SENEGAL
SOUTH AFRICA
TANZANIA
UGANDA
ZAMBIA
ZIMBABWE

323

APPENDIX

324

119

Interview Instrument for Nigerian State Legislators
Local Councillor/State Legislator/National Legislator (Circle One)
Local Government Area (Name, if applicable):
State (Name):
TODAY’S DATE:
TIME INTERVIEW STARTED (Hour, Minute):

1. When were you first elected to the [ Local Government Council/State Assembly/National
Assembly ] (Year Elected)?

2. How many total years have you been a [member of the Local Government Council/State
Assembly/National Assembly ]?

3. What was your main occupation before entering [Local Government Council/State
Assembly/National Assembly ]? DO NOT READ OUT OPTIONS
A.
B.
C.
D.
E.
F.
G.
H.
I.
J.
K.
L.
M.

Never Had A Job
Farmer (produces only for home consumption)
Commercial Farmer (produces mainly for sale)
Businessperson (owner of a business)
Lawyer
Accountant / Auditor
Other professional (for example, doctor, nurse, engineer, etc.)
Supervisor / Mid-level manager / Foreman
Clerical Worker / Secretary / Clerk
Miner
Teacher or Headmaster / Headmistress
Government Worker / Civil Servant
Armed Services/ Police / Security Personnel

119

Adapted from instruments developed by the African Legislatures Project (University of
Capetown, Center for Social Science Research-Democracy in Africa Research Unit) and the
Afrobarometer Public Opinion Survey (Round 4, Afrobarometer Network, East Lansing,
MI).

325

N.
O.
P.
Q.
R.
S.
T.
U.
V.
W.
X.

Local councillor/Local government employee
Traditional Authority/ Chief / Village Headman
Student
Housewife / Works In the Household
General manager / managing director of a company or a parastatal or NGO
Banker
University Lecturer or Professor
Pastor/Evangelist
Guerilla / Liberation fighter
Musician
Journalist

4. I want you to rate the effectiveness of the [Local Government Council/State
Assembly/National Assembly] like a score card. For each of the following areas, how
well or badly would you say the [Local Government Council/State Assembly/National
Assembly] is doing its job? (Very Bad, Bad, Good, Very Good, Do Not Know)
A. Making laws
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

B. Representing constituents
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

C. Reviewing the national budget (the budget allocations, not how it’s spent).
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

D. Monitoring how the Executive spends money
A.
B.
C.

Very Bad
Bad
Good

326

D.
E.

Very Good
Do Not Know

E. Fighting corruption in government
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

F. Assuring development reaches the poorest people
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

G. Fighting HIV/AIDS
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

H. Representing women’s interests
A.
B.
C.
D.
E.

Very Bad
Bad
Good
Very Good
Do Not Know

5. Regardless of what others say the job of an member of the [ Local Government
Council/State Assembly/National Assembly ] is, how would you describe the job of being
an member of the Local Government Council/State Assembly/National Assembly: what
are the most important responsibilities of an member of the Local Government
Council/State Assembly/National Assembly? WRITE DOWN ANSWERS VERBATIM,
EVERYTHING, NOT JUST PHRASES
6. What percentage of your time is devoted to each of the following? GIVE CARD!
Percent
A. Plenary Work
_____
B. Committee Work
_____

327

C. Constituency Work
D. Party Work (outside your constituency)
E. Your other job (including ministerial work)
PERCENTAGES SHOULD TOTAL TO

_____
_____
_____
100%

7. In your opinion, which of these following jobs is the most important part of being member
of the [ Local Government Council/State Assembly/National Assembly ]? READ OUT
OPTIONS - PICK ONLY ONE GIVE CARD!
A.
B.
C.
D.
E.
F.
G.
H.

Debating bills and passing laws
Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ

Second Most Important?
A.
B.
C.
D.
E.
F.
G.
H.

Debating bills and passing laws
Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ

Third Most Impotant?
A.
B.
C.
D.
E.
F.
G.
H.

Debating bills and passing laws
Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ

8. For you personally, which one brings you the most satisfaction? READ OUT OPTIONS –
PICK ONLY ONE GIVE CARD!
A. I don’t find my work rewarding
B. Debating bills and passing laws

328

C.
D.
E.
F.
G.
H.
I.

Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ
Second Most?

A.
B.
C.
D.
E.
F.
G.
H.
I.

I don’t find my work rewarding
Debating bills and passing laws
Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ

Third Most?
A.
B.
C.
D.
E.
F.
G.
H.
I.

I don’t find my work rewarding
Debating bills and passing laws
Making public policy by writing laws
Overseeing the executive
Bringing development to your constituency
Representing constituents’ views in parliament
Assisting constituents with their personal problems
Soliciting funds for your constituency
Do Not Know DO NOT READ

9. In general, when you take a position about an issue in the [ Local Government
Council/State Assembly/National Assembly ], which of the following is most
important?READ OUT OPTIONS – PICK ONLY ONE GIVE CARD!
A.
B.
C.
D.
E.
F.
G.
H.
I.

I am an independent DO NOT READ: MARK ONLY IF MEMBER IS INDEPENDENT
The views of your party leader
The views of your party
The views of your constituents
The national interest
Your knowledge about the issue
Your personal convictions
The views of BOTH the party and party leader DO NOT READ
Do Not Know DO NOT READ

329

.
.
.
.
.
.
.
.

.
.
M.
.
.
.

10. What should MPs do if there is a conflict between their political party’s position and:
The national interest (what’s best for the nation)
A. Support the Party Position
B. Oppose the Party Position
C. Abstain From Voting/Not Attend
D.
Don’t Know DO NOT READ
Their personal convictions
A. Support the Party Position
B. Oppose the Party Position
C. Abstain From Voting/Not Attend
D.
Don’t Know DO NOT READ
The views of their constituents
A. Support the Party Position
B. Oppose the Party Position
C. Abstain From Voting/Not Attend
D.
Don’t Know DO NOT READ
11. Please indicate whether you agree more with Statement A or Statement B, or neither.
PICK ONLY ONE OPTION (Agree Very Strongly With A, Agree With A, Agree Very
Strongly With B, Agree With B)
1.
A. In our country these days, citizens should show more respect for authority.
B. Citizens should be more active in questioning the actions of their leaders.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

2.
A. People are like children; the government should take care of them like a father.
B. Government is an employee; the people should be the bosses who control the
government.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

3.
A. People should look after themselves and be responsible for their own success in life.
B. The government should bear the main responsibility for the well-being of people.

330

A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

4.
A. Since leaders represent everyone, leaders should not favor their own family or group.
B. Once in office, leaders are obliged to help their home community.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

5.
A. In our country, women should have equal rights and receive the same treatment as men do.
B. Women have always been subject to traditional laws and customs, and should remain so.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

6.
A. Government should be able to ban any organization that goes against its policies.
B. Citizens should be able to join any organization, whether or not the government approves of
it.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

7.
A. Government should close newspapers that print false stories or misinformation.
B. The news media should be free to publish any story that they see fit without fear of being shut
down.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

8.
A. Government should not allow the expression of political views that are fundamentally
different from the views of the majority.

331

B. People should be able to speak their minds about politics free of government influence, no
matter how unpopular their views may be.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

9.
A.. The costs of reforming the economy are too high; the government should therefore abandon
its current economic policies.
B. In order for the economy to get better in the future, it is necessary for us to accept some
hardships now.
A.
B.
C.
D.

Agree Very Strongly With A
Agree With A
Agree Very Strongly With B
Agree With B

12. Do you live in your constituency, or area that you represent?
A. No
B. Yes
13. Would you say that you “come from” your constituency? Is it where you lived before you
became an MP?
A. No
B. Yes
14. Do you have a constituency office?
A. No -- Don’t have a constituency office
B. Yes
If yes: Who pays for it?
A. Free (in your home, in your business premises)
B. Yes -- Rent paid by you / Owned by you
C. Yes -- Rent paid / space provided by Local/State/National Assembly
D. Yes -- Rent paid / space provided by your political party
15. Do you have someone who represents you when you are not in your constituency?
A. No
B. Yes -- I pay salary
If yes: How do you compensate them?

332

A.
B.
C.
D.

I pay salary.
Party pays salary.
Local/State/National Assembly pays salary.
No salary/Volunteer.
16. Regardless of whether you have an office or not, does having an office in one's
constituency improve the relationship with constituents, make no difference, or only
make things worse?
A. Helps a great deal
B. Helps somewhat
C. Makes no difference
D. Makes things worse
E. Makes things much worse
F.Do Not Know DO NOT READ
17. How would you describe the burden on your personal time due to traveling back to your
constituency?
A.
B.
C.
D.
E.

Major burden
Minor burden
Not a burden
I live in my constituency because it is in or near [ site of Assembly ]
Do Not Know DO NOT READ

18.
How would you describe the burden of financial costs of traveling back to your
constituency?
A.
B.
C.
D.
E.

Major burden
Minor burden
Not a burden
I live in my constituency because it is in or near [ site of Assembly ]
Do Not Know DO NOT READ
19. How much time do you spend in your constituency?

A.
B.
C.
D.
E.
F.

Never
At least once a year
At least once a month
At least weekly
Almost all of your time
Don’t know
20. When [ Local Government Council/State Assembly/National Assembly ] is in session:

333

A. How many trips did you make to your constituency during a typical month?
____________________________________________________ TRIPS PER MONTH
B. On average, how long did you stay (in days)?
____________________________________________________ DAYS PER TRIP
21. When [ Local Government Council/State Assembly/National Assembly ] is not in
session?
A. How many trips did you make to your constituency during a typical month?
____________________________________________________ TRIPS PER MONTH
B. On average, how long did you stay (in days)?
____________________________________________________ DAYS PER TRIP
22. Would you regard your constituency as mostly:
A. Urban / Suburban
B. Peri urban / semi urban
C. Rural
D. Do Not Know DO NOT READ
23. Thinking of the distance between your constituency and [site Local Government
Council/State Assembly/National Assembly ]:
A. How many kilometers is it?
_________________________________________ KMS FROM [ SITE of Local
Government Council/State Assembly/National Assembly ] TO CONSTITUENCY
B. How long does it take you to travel this distance? (In Hours)
____________________________________________________HOURS PER TRIP
24. Thinking of the distance between your headquarters in your constituency to the furthest
boundary of your constituency:
A. How many kilometers is it?
_________________________________________ KMS FROM [ SITE of Local Government
Council/State Assembly/National Assembly ] TO CONSTITUENCY
B. How long does it take you to travel this distance? (In Hours)
____________________________________________________HOURS PER TRIP
25. How do you spend most of your time when you are in your constituency? GIVE CARD!
A.
B.
C.
D.
E.

Giving speeches to groups at meetings and rallies
Listening to constituents
Attending local political party meetings
Attending functions, like weddings and funerals
Assisting, visiting and inspecting development projects

334

F.
Meeting with government officials
G. Other: __________________________________________________________
H. Do Not Know DO NOT READ
Second most time consuming?
A.
B.
C.
D.
E.
F.
G.
H.

Giving speeches to groups at meetings and rallies
Listening to constituents
Attending local political party meetings
Attending functions, like weddings and funerals
Assisting, visiting and inspecting development projects
Meeting with government officials
Other: __________________________________________________________
Do Not Know DO NOT READ

Third most time-consuming?
A.
B.
C.
D.
E.
F.
G.
H.

Giving speeches to groups at meetings and rallies
Listening to constituents
Attending local political party meetings
Attending functions, like weddings and funerals
Assisting, visiting and inspecting development projects
Meeting with government officials
Other: __________________________________________________________
Do Not Know DO NOT READ

26. What is the primary method you use when you consult your constituents? GIVE CARD!
A.
B.
C.
D.
E.
F.
G.
H.

Attend general public meetings (open to all)
Attend meetings with religious leaders
Hold meetings with traditional local authorities
Hold meetings with constituency party officials
Hold meetings with provincial/regional party officials
Live in the constituency
Other: __________________________________________________________
Do Not Know DO NOT READ

27. Overall do you think the amount of time you spend with constituents is too much, too
little, or about right?
A.
B.
C.
D.
E.

Far too little
Too little
About right
Too much
Far too much

335

F.

Do Not Know DO NOT READ

28. In general, how interested do you think the Members of the [ Local Government
Council/State Assembly/National Assembly ] are in hearing what their constituents have
to say?
A.
B.
C.
D.
E.

Very interested
Interested
Not very interested
Not interested at all
Do Not Know DO NOT READ

29. On an average week, how many of your constituents contact you here?
A. None
B. One or two per week
C. One or two per day
D. Many people every day
E.
Do Not Know DO NOT READ
30. In general, how do your constituents most often contact you when you are in [the city
where Local Government Council/State Assembly/National Assembly meets]?
A.
B.
C.
D.
E.
F.
G.

Cell phone
Landline
Through constituency office
In person
Letters
E-mail
Do Not Know DO NOT READ

31. How easy or difficult is it for ordinary people to obtain the following services? (Very
Difficult, Difficult, Easy, Very Easy, Do Not Know)
A. An identity document (such as a birth certificate or passport)
A. Very Difficult
B. Difficult
C. Easy
D. Very Easy
E. Do Not Know
B. A place in primary school for a child.
A. Very Difficult
B. Difficult
C. Easy
D. Very Easy
E. Do Not Know
C. Household services (like piped water, electricity or telephone)

336

A. Very Difficult
B. Difficult
C. Easy
D. Very Easy
E. Do Not Know
32. If an Member of [ Local Government Council/State Assembly/National Assembly ]
ignores what his/her constituents have to say, do you think that he/she would still remain
in office or lose the next election?
A.
B.
C.
D.
E.

Remain in office, definitely
Remain in office, probably
Lose their seat, probably
Lose their seat, definitely
Do Not Know DO NOT READ

33. Think about how elections work in practice in this country. How well do elections ensure
that the Representatives to the National Assembly reflect the views of voters?
A.
B.
C.
D.
E.

Not at all well
Not very well
Well
Very well
D Not Know. DO NOT READ.
34. Think about how elections work in practice in this country. How well do elections enable
voters to remove from office leaders who do not do what the people want?

A.
B.
C.
D.
E.

Not at all well
Not very well
Well
Very well
D Not Know. DO NOT READ.
35. In your opinion, what are the most important problems facing this country that
government should address? ACCEPT TOP 3

36. In your opinion how much of a democracy is Nigeria today?
A.
B.
C.

Not a democracy
A democracy, with major problems
A democracy, but with minor problems

337

D.
E.

Full democracy
Do Not Know DO NOT READ
37. Overall, how satisfied are you with the way democracy works in Nigeria? Are you:

A.
B.
C.
D.
E.
F.

Nigeria is not a democracy
Not at All Satisfied
Not Very Satisfied
Fairly Satisfied
Very Satisfied
Do Not Know DO NOT READ
38. In your opinion, how likely is it that Nigeria will remain a democratic country?

A.
B.
C.
D.
E.
F.

Nigeria is not a democracy
Not at All Likely
Not Very Likely
Likely
Very Likely
Do Not Know DO NOT READ

39.

Do you think that public revenues are adequate to development needs?

40.

Is Nigeria is too reliant on oil revenues? Why or why not?

41. Who do you think actually has primary responsibility for collecting income taxes?
A.
B.
C.
D.
E.
F.

The Federal government
The State government
The Local government
Traditional leaders
Members of the community
None

42.
A. Do you think low-income citizens are under or over-taxed with respect to their income?
Can you provide examples?
B. Do you think high-income individuals are under or over-taxed with respect to their
income? Can you provide examples?

338

C. Do you think corporations in Nigeria are under or over-taxed? Can you give examples?
43.

In your opinion, how big of a problem is tax avoidance? Can you give examples?

44. How many Tax Officials do you think are involved in corruption, or haven’t you heard
enough about them to say?
A.
B.
C.
D.
E.

None
Some of Them
Most of them
All of them
Do Not Know DO NOT READ

45. Have you had to make any income tax payments during the past year?
A.
B.
C.

No
Yes
Do Not Know DO NOT READ

46. Please tell me whether you disagree or agree: The tax department always has the right to
make people pay taxes.
A.
B.
C.
D.
E.
F.

Strongly Disagree
Disagree
Neither Agree, Nor Disagree
Agree
Strongly Agree
Do Not Know DO NOT READ

47.

What is your current political party affiliation?

48.

How old are you?

49.

What is your home language?

50.

What is your highest level of education (Primary school, secondary school, university)?

339

IF WENT TO UNVERSITY OR COLLEGE SPECIFY EACH DEGREE.
51.

What is your ethnic group or tribe?

52.

What is your religious denomination, if any?

Thank you very much for your time.
TIME INTERVIEW ENDED (Hour, Minute):
Gender:

340

120

Interview Instrument for Civil Servants
Ministry:
Position:
TODAY’S DATE:
TIME INTERVIEW STARTED (Hour, Minute):

1. What is your position at this ministry? What are your daily responsibilities?

2. How long have you been at this post?

3. Do you think public revenues are adequate to meet development needs?

4.
a. Why have officials identified “internal revenue generation through taxation” as a superior
alternative to petroleum income?
b. What strategies have been adopted to achieve this goal?
c. How have elected officials (e.g. legislators) reacted to reform efforts?
d. How have citizens reacted to reform efforts?

5. Do you think low-income citizens are under or over taxed with respect to their income?

6.
a. Do you think high-income citizens are under or over taxed with respect to their
income?
120

Adapted from instruments developed by the African Legislatures Project (University of
Capetown, Center for Social Science Research-Democracy in Africa Research Unit) and the
Afrobarometer Public Opinion Survey (Round 4, Afrobarometer Network, East Lansing, MI).

341

b. Do you think corporations are under or over taxed with respect to their income?

c. In your opinion, how big of a problem is tax-avoidance? Examples?

7. Who do you think actually has primary responsibility for collecting income taxes?
A.
B.
C.
D.
E.
F.

The Federal government
The State government
The Local government
Traditional leaders
Members of the community
None
8. How many Tax Officials do you think are involved in corruption, or haven’t you heard
enough about them to say?

A.
B.

None
Some of Them
C. Most of them
D. All of them
E. Do Not Know DO NOT READ
9. Please tell me whether you disagree or agree: The tax department always has the right to
make people pay taxes.

A.
B.
C.
D.
E.
F.

Strongly Disagree
Disagree
Neither Agree, Nor Disagree
Agree
Strongly Agree
Do Not Know DO NOT READ

Thank you very much for your time.
TIME INTERVIEW ENDED (Hour, Minute):

342

Research Participant Information and Consent Form: Legislators

You are being asked to participate in a research study of elite (public and elected officials)
opinions by answering questions about your views of current political and economic issues
facing Nigeria today. This is part of my larger doctoral dissertation project, based at Michigan
State University (East Lansing, MI, USA). The results of this study will be used to that end.
You have been selected to participate in this survey because of your position as a (Councilor,
Representative, Senator) in Nigeria. You are not obligated to take part in this study. You may
refuse to answer any question. All of your answers will be kept strictly confidential. Your name
is not on this questionnaire, and your name will NEVER be used with the answers you provide.
Your answers will be put together with the answers of your colleagues who have also been
selected for this survey to get an overall picture of the views and opinions of the elected officials
across Nigeria.
Please feel free to answer openly and honestly. If something is unclear to you, or you want
further explanation, please ask me. Most importantly, there are no right or wrong answers.
Instead, I am interested in what you think and your opinions.
If you have concerns or questions about this study, such as scientific issues or how to do any part
of it, please contact Olufunmbi M. Elemo via email at elemoolu@msu.edu, or regular mail at
Olufunmbi M. Elemo Department of Political Science, Michigan State University, 323 South
Kedzie Hall, E. Lansing, MI 48824. You may also contact Dr. Michael Bratton via email at
mbratton@msu.edu or regular mail at Dr. Michael Bratton Department of Political Science,
Michigan State University, 323 South Kedzie Hall, East Lansing, MI 48824.
If you have questions or concerns about your role and rights as a research participant, would like
to obtain information or offer input, or would like to register a complaint about this study, you
may contact, anonymously if you wish, the Michigan State University’s Human Research
Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail
at 207 Olds Hall, MSU, East Lansing, MI 48824.
According to the research ethics policy of Michigan State University, I am obligated to tell
you that there is no penalty for refusing to participate, and to ask you whether you wish to
continue with this interview.
You indicate your voluntary agreement to participate by beginning this interview.

343

Research Participant Information and Consent Form: Civil Servants

You are being asked to participate in a research study of elite (public and elected officials)
opinions by answering questions about your views of current political and economic issues
facing Nigeria today. This is part of my larger doctoral dissertation project, based at Michigan
State University (East Lansing, MI, USA). The results of this study will be used to that end.
You have been selected to participate in this survey because of your position as a civil servant in
Nigeria. You are not obligated to take part in this study. You may refuse to answer any question.
All of your answers will be kept strictly confidential. Your name is not on this questionnaire,
and your name will NEVER be used with the answers you provide. Your answers will be put
together with the answers of your colleagues who have also been selected for this survey to get
an overall picture of the views and opinions of the elected officials across Nigeria.
Please feel free to answer openly and honestly. If something is unclear to you, or you want
further explanation, please ask me. Most importantly, there are no right or wrong answers.
Instead, I am interested in what you think and your opinions.
If you have concerns or questions about this study, such as scientific issues or how to do any part
of it, please contact Olufunmbi M. Elemo via email at elemoolu@msu.edu, or regular mail at
Olufunmbi M. Elemo Department of Political Science, Michigan State University, 323 South
Kedzie Hall, E. Lansing, MI 48824. You may also contact Dr. Michael Bratton via email at
mbratton@msu.edu or regular mail at Dr. Michael Bratton Department of Political Science,
Michigan State University, 323 South Kedzie Hall, East Lansing, MI 48824.
If you have questions or concerns about your role and rights as a research participant, would like
to obtain information or offer input, or would like to register a complaint about this study, you
may contact, anonymously if you wish, the Michigan State University’s Human Research
Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail
at 207 Olds Hall, MSU, East Lansing, MI 48824.
According to the research ethics policy of Michigan State University, I am obligated to tell
you that there is no penalty for refusing to participate, and to ask you whether you wish to
continue with this interview.
You indicate your voluntary agreement to participate by beginning this interview.

344

BIBLIOGRAPHY

345

BIBLIOGRAPHY

Aberbach, J., J. Chesney, and B. Rockman. 1975. “Exploring Elite Political Attitudes: Some
Methodological Lessons.” Political Methodology, 2: 1 – 27.
Acemoglu, D. and J. Robinson. 2005. Economic Origins of Dictatorship and Democracy. New
York: Cambridge University Press.
Adams, J., M. Clark, L. Ezrow, and G. Glasgow. 2004. “Understanding Change and Party
Ideologies: Do Parties Respond to Public Opinion of to Past Election Results?” British
Journal of Political Science 34(4): 589 – 610.
Afrobarometer Network. 2008. Round 4 Afrobarometer Surveys in 20 African Countries [Data
file and codebook]. Retrieved from http://www.afrobarometer.org/data/data-roundsmerged.
Afrobarometer Network. 2008. Round 4 Afrobarometer Survey in Nigeria [Data file and
codebook]. Retrieved from http://www.afrobarometer.org/data/data-by-country-nz/nigeria?start=10.
Ahmad, J., S. Devarajan, S. Khemani, S. Stuti, and S. Shah. 2005. “Decentralization and Service
Delivery.”Washington, DC: World Bank.
Akin, J., P. Hutchinson, and K. Strumpf. 2001. “Decentralization and Government Provision of
Public Goods: The Public Health Sector in Uganda.” MEASURE Evaluation Working
Paper 01-35. Carolina Population Center, University of North Carolina at Chapel Hill.
Aldrich, J. 1995. Why Parties?: The Origins and Transformation of Political Parties in America.
Chicago: University of Chicago Press.
Bardhan, P. 2001. "Understanding Underdevelopment: Challenges for Institutional Economics
From the Poor Country Perspective." In G. Kochendörfer-Lucius and B. Pleskovic (eds.)
Institutional Foundations of a Market Economy, Berlin: German Foundation for
International Development (DSE).
Bardhan, P. and D. Mookherjee. (2005a), “Decentralizing Delivery of Anti-Poverty Programs in
Developing Countries.” Journal of Public Economics, forthcoming.
Bardhan, P. and D. Mookherjee. 2005b. Decentralization and Local Governments in Developing
Countries: A Comparative Perspective. Cambridge: MIT Press.
Barkan, J. 2009. “African Legislatures and the ‘Third Wave’ of Democratization.” In J. Barkan
(ed.) Legislative Power in Emerging African Democracies, Boulder: Lynne Rienner.

346

Barro, R. 1999. "Determinants of Democracy." Journal of Political Economy, 107 (S6): 158 –
183.
Bates, R. 1981. Markets and States in Tropical Africa. Berkley: University of California Press.
Bates, R., and D. Lien. 1985. "A Note on Taxation, Development, and Representative
Government." Politics and Society, 14 (1): 53 - 270.
Bates, R. and M. Humphreys. 2005. "Political Institutions and Economic Policies: Lessons from
Africa." British Journal of Political Science, 35: 403 - 428.
Bates, R. 2009. The Logic of State Failure: Learning from Late-Century Africa (Dealing With
Failed States). New York: Routledge.
Beblawi, H. and G. Luciani. 1987. The Rentier State. New York: Croom Helm.
Berger, D. 2009. “Taxes, Institutions and Local Governance: Evidence from a Natural
Experiment in Colonial Nigeria.” Department of Political Science: New York University.
Berry, J. 2002. “Validity and Reliability Issues in Elite Interviewing.” PS: Political Science, 35
(4): 679 – 682.
Bish, R. and V. Ostrom. 1973. Understanding Urban Government: Metropolitan Reform
Reconsidered. Washington. D.C.: American Enterprise Institute for Public Policy
Research.
Bodea, C. 2012. “Natural Resources, Weak States, and Civil War: Can Rents Stabilize Coup
Prone Regimes?” World Bank, Washington, DC.
Boix, C. and S. Stokes. 2003. "Endogenous Democratization." World Politics, 44: 517 - 49.
Bollen, K. and R. Jackman. 1985. “Political Democracy and the Size Distribution of Income.”
American Sociological Review, 50: 438 – 457.
Bratton, M., and N. van de Walle. 1997. Democratic Experiments in Africa. Cambridge:
Cambridge University Press.
Bratton, M., R. Mattes, and E. Gyimah-Boadi. 2005. Public Opinion, Democracy, and Market
Reform in Africa. Cambridge: Cambridge University Press.
Bratton, M. and E. Chang. 2006. "State Building and Democratization in Sub-Saharan Africa."
Comparative Political Studies 39 (9): 1059 - 1083.
Bratton, M. 2009. “Citizen Perceptions of Local Government Responsiveness in Sub-Saharan
Africa,” Paper Prepared under World Bank Contract No. 715131.

347

Brautigam, D., O. Fjeldstad, and M. Moore. 2008. Taxation and State-Building in Developing
Countries: New York: Cambridge University Press.
Bueno de Mesquita, B., A. Smith, R. Siverson, and J. Morrow. 2003. The Logic of Political
Survival. Cambridge, MA: MIT Press.
Burke, E. 1770. Thoughts on the Cause of Present Discontents. London: Printed for J. Dodsley in
Pall-Mall.
Burke, E.1774. Speech to the Electors of Bristol. (November 3, 1774). Retrieved from
http://press-pubs.uchicago.edu/founders/documents/v1ch13s7.html.
Calvert, R., M. McCubbins, and B. Weingast. 1989. “A Theory of Political Control and Agency
Discretion,” American Journal of Political Science, 33: 588 - 611.
Central Bank of Nigeria. 2000 – 2010. Annual Report and Statement of Accounts [Summaries of
State Government Finances for 1999, 2000, 2001, 2002, 2003, 2005, 2007, 2008, 2009
fiscal years; Summaries of Local Government Finances for 1999, 2000, 2001, 2002, 2003,
2008 fiscal years]. Retrieved from the Central Bank of Nigeria Headquarters, Abuja,
Nigeria.
Chartered Institute of Taxation of Nigeria (CITN). 2009. “CITN: Nigerian Tax Guide and
Statutes, Tax Laws and Regulations.”
Cheibub, J. 1998. “Political Regimes and the Extractive Capacity of Governments: Taxation in
Democracies and Dictatorships.” World Politics 50(3): 349.
Cheibub, J. and A. Przeworski. 1999. “Democracy, Elections, and Accountability for Economic
Outcomes,” in A. Przeworski, S. Stokes, and B. Manin (eds) Democracy, Accountability,
and Representation. New York: Cambridge University Press.
Collier, P. and A. Hoeffler. 1998. “On the economic causes of civil war.” Oxford Economic
Papers, 50 (4): 563 - 573.
Collier, P. and A. Hoeffler. 2004. "Greed and Grievance in Civil War." Oxford Economic Papers.
56 (4): 563 – 595.
Coughlin, P., D. Mueller, and P. Murrell. 1990. “Electoral Politics, Interest Groups, and the Size
of Government.” Economic Inquiry, 28: 682 - 705.
Dahl, R. 1977. “On Removing Certain Impediments to Democracy in the United States,”
Political Science Quarterly, 92: 1-20.
Dahl, R. 1989. Democracy and Its Critics. New Haven: Yale University Press.

348

Dean, J. and W. Whyte. 1970. “What Kind of Truth Fo You Get?” in L. Dexter (ed.) Elite and
Specialized Interviewing Chicago: Northwestern University Press, 119 – 138.
Diamond, L. 1987. “Nigeria: Pluralism, Statism, and the Struggle for Democracy.” In L.
Diamond, J. Linz, S. Lipset, Democracy in Developing Countries: Africa, Boulder:
Lynne Rienner.
Diamond, L. 1988. “The 1983 General Elections.” In V. Ayeni and K.
Soremekun (eds), Nigeria's Second Republic: Presdientialism, Politics and
Administration in a Developing State, Lagos: Daily Times.
Diamond, L. 1992. "Economic Development and Democracy Reconsidered." American
Behavioral Scientist, 35 (4/5): 450 – 499.
Dunning, T. 2008. Crude Democracy: Natural Resource Wealth and Political Regimes.
Cambridge: Cambridge University Press.
Elaigwu, J. 2005. The Politics of Federalism in Nigeria. Jos: Aha Publishing Ltd.
Elaigwu, J. 2007. Fiscal Federalism in Nigeria: facing the Challenges of the Future. Jos: Aha
Publishing Ltd.
Epstein, D., R.Bates, J. Goldstone, I. Kristensen and S. O’Halloran. 2006. “Democratic
Transition,” American Journal of Political Science, 50 (3): 551 - 569.
Epstein, D. and S. O’Halloran. 1995. “A Theory of Strategic Oversight: Congress, Lobbyists,
and the Bureaucracy.” Journal of Law, Economics, and Organization 11: 227-255.
Fajingbesi, A., A. Odusola, and E. Uga. 2004. “Revenue Mobilization at the State and Local
Government Levels.” in Fiscal Federalism and Democratic Governance in Nigeria, I.
Taiwo and A. Fajingbesi (eds). Ibadan: National Centre for Economic Management and
Administration.
Fearon, J. 1999. “Electoral Accountability and the Control of Politicians: Selecting Good Types
versus Sanctioning Poor Performance,” in A. Przeworski, S. Stokes, and B. Manin (eds)
Democracy, Accountability, and Representation. New York: Cambridge University Press.
Fearon, J. and D. Laitin. 2003. “Ethnicity, Insurgency, and Civil War.” American Political
Science Review, 97 (1): 75 – 90.
Federal Inland Revenue Service (FIRS). 2007. Federal Inland Revenue Service Establishment
Act. Abuja: Federal Inland Revenue Service.
Federal Ministry of Finance. 2010. A Citizens’ Guide to Understanding the 2009 Federal Budget.
Abuja: Budget Office of the Federation.

349

Federal Ministry of Finance. 2011a. “The National Tax policy.” Abuja: Federal Inland Revenue
Service.
Federal Ministry of Finance. 2011b. “The Nigerian Tax System.” Abuja: Federal Inland Revenue
Service.
Federal Republic of Nigeria. 2000. The Nigerian Constitutions: 1963, 1979, 1999. Lagos:
Law Lords Publications.
Ferejohn, J. 1999. “Accountability and Authority: Toward a Theory of Political Accountability,”
in A. Przeworski, S. Stokes, and B. Manin (eds) Democracy, Accountability, and
Representation. New York: Cambridge University Press.
Ferejohn, J, and B. Weingast. 1997. The New Federalism: Can the States be Trusted?
Stanford, California: Hoover Institution Press.
Fish, M. 2006. “Stronger Legislatures, Stronger Democracies.” Journal of Democracy, 17(1): 5 –
20.
Fjeldstad, O. 2001. “Taxation, Coercion and Donors: Local Government Tax Enforcement in
Tanzania.” The Journal of Modern African Studies, 39 (2): 289 - 306.
Fjeldstad, O. and J. Semboja. 2001. “Why People Pay Taxes: The Case of the Development Levy
in Tanzania.” World Development, 29 (12): 2059 - 2074.
Fjeldstad, Odd-Helge and L. Ranker. 2003. “Taxation and Tax Reforms in developing countries:
Illustrations from sub-Saharan Africa.” Working Paper. Chr. Michelsen Institute.
Fjeldstad, O. and B. Tungodden. 2003. “Fiscal Corruption: A Vice or Virtue?” World
Development, 31 (8): 1459-67.
Freedom House. 2008. Freedom in the World 2008: Nigeria. Retrieved from
http://www.freedomhouse.org/report/freedom-world/2008/nigeria.
Frey, B. and F. Schneider. 1978. “An Empirical Study of Politico-Economic Interaction in the
United States.” The Review of Economics and Statistics 60 (2): 174 - 183.
Furlong, S. 1998. “Political Influence on the Bureaucracy: The Bureaucracy Speaks.” Journal of
Public Administration Research and Theory, 8 (1): 39 - 65.
Geddes, B. "What Do We Know About Democratization After 20 Years?" Annual Review of
Political Science 2: 115 - 144.
George, A. and A. Bennett. 2005. Case Studies and Theory Development in the Social Sciences.
Cambridge: MIT Press.

350

Gibson, C. and B. Hoffman. 2006. "Political Accountability and Fiscal Governance: Evidence
from Tanzania and Zambia," under review, Journal of Politics.
Goldberg, E., E. Wibbels, and E. Mvukiyehe. 2008. “Lessons from Strange Cases: Development
and the Resource Curse in the U.S. States.” Comparative Political Studies, 41 (4/5): 477
– 514.
Gordon, R. 1983. “An Optimal Taxation Approach to Fiscal Federalism.” Quarterly Journal of
Economics, 98: 567 – 586.
Graf, W. 1988. The Nigerian State: Political Economy, State, Class, and Political System in the
Post-Colonial Era. London: James Currey.
Guyer, J. 1991. "Representation without Taxation: An Essay on Democracy in Rural Nigeria,
1952-1990," In African Studies Center: Working Papers. Boston: Boston University.
Haggard, S. 1990. Pathways from the Periphery: The Politics of Growth in Newly Industrialized
Countries. Princeton: Princeton University Press.
Hamilton, A., J. Madison, and J. Jay. 1788. The Federalist Papers. Retrieved from
http://www.foundingfathers.info/federalistpapers/.
Hammond, T. 1986. “Agenda Control, Organizational Structure, and Bureaucratic Politics.”
American Journal of Political Science, 30 (2): 379 - 420.
Hammond, T. and J. Knott. 1996. “Who Controls the Bureaucracy?: Presidential Power,
Congressional Dominance, Legal Constraints, and Bureaucratic Autonomy in a Model of
Multi-Institutional Policymaking.” Journal of Law, Economics, and Organization, 12 (1):
121 - 168.
Herb, M. 2003. “Taxation and Representation.” Studies in Comparative International.
Development, 38 (3): 3 - 31.
Herb, M. 2005. “No Representation without Taxation? Rents, Development, and Democracy.”
Comparative Politics, 37 (3): 297 – 316.
Hoffman, P. and K. Norberg. 1994. Fiscal Crises, Liberty, and Representative Government,
1450-1789. Stanford: Stanford University Press.
Huber, J. and G. Bingham Powell, Jr. 1994. “Congruence between Citizens and Policymakers in
Two Visions of Liberal Democracy,” World Politics 46: 291 – 326.
Humphreys, M. 2005. “Natural Resources, Conflict, and Conflict Resolution.” Journal of
Conflict Resolution, 49 (4): 508 – 537.

351

Huntington, S. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman:
University of Oklahoma Press.
Ihonvbere, J. and T. Shaw. 1998. Illusions of Power: Nigeria in Transition. Lawrenceville, New
Jersey: Africa World Press.
Ikein, A. and C. Briggs-Anigboh. 1998. Oil and Fiscal Federalism in Nigeria: The Political
Economy of Resource Allocation in a Developing Country. Aldershot: Ashgate
Publishing Company.
Inman, R. and D. Rubinfeld. 1996. “Designing Tax Policy in Federalist Economies: An
Overview.” Journal of Public Economics, 60: 307 - 334.
Jensen, N. and L. Wantchekon. 2004. “Resource Wealth and Political Regimes in Africa.”
Comparative Political Studies, 37 (7): 816 – 841.
JHI. 2009. “Doing Business in Nigeria: Taxation.” Retrieved from
http://jhi.com/news/publications/JHI_DoingBusinessNigeria.pdf.
Jones Luong, P. and E. Weinthal. 2006. "Rethinking the Resource Curse: Ownership Structure,
Institutional Capacity, and Domestic Constraint." Annual Review of Political Science, 9:
241 - 263.
Jones Luong, P. and E. Weinthal. 2010. Oil Is Not a Curse: Ownership Structure and Institutions
in Soviet Successor States. Cambridge: Cambridge University Press.
Kano State Board of Internal Revenue (KBIR) Modernization Plan. 2009. Kano: World Bank.
Karl, T. 1997. The Paradox of Plenty: Oil Booms and Petro-States. Los Angeles: University of
California Press.
Kasara, K. 2007. "Tax Me If You Can: Ethnic Geography, Democracy, and the Taxation of
Agriculture in Africa.” American Political Science Review, 101 (1): 19 - 172.
Lam, R. and L. Wantchekon. 1999. “Political Dutch Disease.” Working Paper. Yale University.
LeBas, A. 2012. “Taxation and Trust in Urban Nigeria.” Working Paper. Department of
Government, American University.
Leech, B. and K. Goldstein. 2002. “Symposium: Interview Methods in Political Science.” PS:
Political Science and Politics, 35 (4): 663-672.
Levi, M. 1988. Of Rule and Revenue. Berkley: University of California Press.

352

Lewis, P. 1996. “From Prebendelism to Predation: The Political Economy of Decline in Nigeria.”
Journal of Modern African Studies, 34 (1): 79 – 103.
Lewis, P. 1997. “Politics and the Economy: A Downward Spiral.” In P. Beckett and C. Young
(eds.) Dilemmas of Democracy in Nigeria, Rochester: Univeristy of Rochester Press.
Lewis, P. 2009. “Rules and Rents in Nigeria’s National Assembly.” In J. Barkan
(ed.) Legislative Power in Emerging African Democracies, Boulder: Lynne Rienner.
Lewis, P. 2011. “Africa Legislatures Project: Nigeria Country Report.” Center for Social Science
Research, University of Cape Town.
Lieberman, E. 2003. Race and Regionalism in the Politics of Taxation in Brazil and South
Africa. Cambridge: Cambridge University Press.
Lijphart, A. 1984. Democracies: Patterns of Majoritarian and Consensus Government in
Twenty-One Countries. New Haven: Yale University Press.
Lindberg, S. 2006. Democracy and Elections in Africa. Baltimore: Johns Hopkins
University Press.
Linz, J. and A. Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern
Europe, South America, and post-Communist Europe. Baltimore, MD: Johns Hopkins
University Press.
Lipset, S. 1959. "Some Social Requisites for Democracy: Economic Development and Political
Legitimacy," American Political Science Review 53 (1): 69 - 105.
Lujala, P. 2010. “The Spoils of Nature: Armed Civil Conflict and Rebel Access to Natural
Resources.” Journal of Peace Research, 47 (1): 15 - 28.
Lujala, P., N. Gleditsch, and E. Gilmore. 2005. “A Diamond Curse?: Civil War and a Lootable
Resource.” Journal of Conflict Resolution, 49: 538 – 562.
Manin, B., A. Przeworski, and S. Stokes. 1999. “Elections and Representation,” in A. Przeworski,
S. Stokes, and B. Manin (eds) Democracy, Accountability, and Representation. New
York: Cambridge University Press.
Mann, M. 2004. The Dark Side of Democracy. Cambridge: Cambridge University Press.
McCubbins, M. 1985. “The Legislative Design of Regulatory Structure.” American Journal of
Political Science, 29: 721- 748.
McCubbins, M. and T. Schwartz. 1984. “Congressional Oversight Overlooked: Police Patrols
Versus Fire Alarms.” American Journal of Political Science, 28 (1), 165 - 179.

353

McCubbins, M., R. Noll, and B. Weingast. 1987. “Administrative Procedures as Instruments of
Political Control.” Journal of Law, Economics, and Organization, 3 (1), 243 - 277.
McGuirk, E. 2012. “The Illusory Leader: Natural Resources, Taxation and Accountability.”
Public Choice, forthcoming.
Meltzer, A. and S. Richard. 1981. “A Rational Theory of the Size of Government.” The Journal
of Political Economy, 89 (5): 914 - 927.
Meier, K. 1985. Regulation: Politics, Bureaucracy, and Economics. New York: St. Martin's
Press.
Mitnick, B. 1973. “Fiduciary Rationality and Public Policy: The Theory of Agency and Some
Consequences,” Paper presented at the 1973 Annual Meeting of the American Political
Science Association: New Orleans, LA.
Mitnick, B. 1975. "The Theory of Agency: The Policing 'Paradox' and Regulatory Behavior."
Public Choice, 24: 27 - 42.
Mitnick, B. 1980. The Political Economy of Regulation. New York: Columbia University Press.
Moe, T. M. 1982. “Regulatory Performance and Presidential Administration.” American Journal
of Political Science, 26: 197 - 224.
Moe, T. M. 1984. “The New Economics of Organization.” American Journal of Political Science,
28: 739 - 777.
Moe, T. M. 1987. “An Assessment of the Positive Theory of Congressional Dominance.”
Legislative Studies Quarterly, 1: 475 - 520.
Moore, B. 1966. The Social Origins of Dictatorship and Democracy. Boston: Beacon Press.
Morrison, K. 2009. “Oil, Nontax Revenue, and the Redistributional Foundations of Regime
Stability.” International Organization, 63: 107 - 138.
Morrison, K. 2011. “Nontax Revenue, Social Cleavages, and Authoritarian Stability in Mexico
and Kenya: ''Internationalization, Institutions, and Political Change'' Revisited”
Comparative Political Studies, 44: 719 - 747.
Ndebbio, J. 2004. "Financial Deepening, Economic Growth and Development: Evidence from
Selected Sub-Saharan African Countries." Research Papers RP_142, African Economic
Research Consortium.
North, D. and B. Weingast. 1989. "Constitutions and Commitment: The Evolution of Institutions
Governing Public Choice in Seventeenth-Century England." The Journal of Economic
History XLIX (4): 803 - 832.

354

Nigeria Investment Promotion Commission (NIPC). 2009. “Taxation and Fiscal Regulations in
Nigeria.”
Nigerian National Bureau of Statistics. 2006. “Social Statistics in Nigeria, 2005.”
Nigerian National Bureau of Statistics. 2007. “Population Census, 2006.”
Oates, W. 1972. Fiscal Federalism. New York: Harcourt Brace Jovanovich.
Ogban-Iyam, O. 1998. “Federalism in Nigeria: Past, Present and Future.” In T. Babawale, K.
Olufemi, and F. Adewumi (eds) Re-inventing Federalism in Nigeria: Issues and
Perspectives. Lagos: Friederich Ebert Inventing Nigeria Foundation.
Ogul, M. 1976. Congress Oversees the Bureaucracy: Studies in Legislative Supervision.
Pittsburgh: University of Pittsburgh Press.
Ogwumike, F. and V. Isumonah. 2004. “Fiscal Federalism and the Alleviation of Poverty: The
Experience So Far.” In I. Taiwo and A. Fajingbesi (eds.) Fiscal Federalism and
Democratic Governance in Nigeria, Ibadan: NCEMA.
Okoko, A.B., and N.J. Nna. 1997. "Federalism and Resource Allocation: The Nigerian
Experience," Nigerian Journal of Oil and Politics, 1 (1): 16 - 36.
Olaniyi, O. 2001. "The Oil Sector and Growth of Non-Oil Sectors in Nigeria." Journal of
Economics and Allied Fields, 1 (2): 1 - 11.
Okoko, A. and N. Nna. 1997. "Federalism and Resource Allocation: The Nigerian Experience."
Nigerian Journal of Oil and Politics, 1 (1): 16 - 36.
Okoosi-Simbine, A. 2007. Assessing the Role of the Legislature in A Democracy: A Case Study
of Oyo State House of Assembly.” Ibadan: NISER.
Olsen, M. 1993. “Dictatorship, Democracy, and Development.” The American Political Science
Review, 87 (3): 567 - 576.
Omoigui, I. 2007. “Federal Inland Revenue Service.” Presented at FIRS Training Program on
Reform, Abuja: March 16, 2007.
Omoyele, O. 2011. Lagos Legislative Fact Finder. Lagos: Straight Path Communications
Company.
Orewa, G. 1979. Taxation in Western and Mid-Western Nigeria. Ibadan: Nigeria Institute of
Social and Economic Research.
Oriakhi, D. 2004. Introduction to Public Finance. Benin City: Mindex Publishing Co.

355

Oviasuyi, P. and J. Uwadiae. 2010. “The Dilemma of Niger-Delta Region as Oil Producing
States of Nigeria.” Journal of Peace, Conflict, and Development, 16: 110 – 126.
Paden, N. 1997. “Nigerian Unity and the Tensions of Democracy: Geo-Cultural Zones and
North-South Legacies.” In P. Beckett and C. Young (eds.), Dilemmas of Democracy in
Nigeria, Rochester: University of Rochester Press.
Persson, T., G. Roland, and G. Tabellini. 1997. “Separation of Powers and Political
Accountability.” Quarterly Journal of Economics, 112 (4): 1163 - 1202.
Pitkin, H. 1967. The Concept of Representation. Berkley: University of California Press.
Powell, G. Bingham, Jr. 2004. “Political Representation in Comparative Politics,” Annual
Review of Political Science, 7: 273 – 296.
Powell, G. Bingham, Jr. 2004. “The Chain of Responsiveness.” Journal of Democracy, 15 (4):
91–105.
Przeworski, A., S. Stokes, B. Manin. 1999. Democracy, Accountability, and Representation.
New York: Cambridge University Press.
Przeworski, A., M. Alvarez, J. Cheibub and F. Limongi. 2000. Democracy and Development:
Political Institutions and Well-Being in the World, 1950-1990. New York: Cambridge
University Press.
Putnam, R. 1993. Making Democracy Work. Princeton: Princeton University Press.
Rakner, L. 2002. “The Politics of Revenue Mobilization: Explaining Continuity in Namibian Tax
Policies.” Forum for Development Studies, 28 (1): 125 - 145.
Reinikka, R. and J. Svensson. 2004. "Local Capture: Evidence From a Central Government
Transfer Program in Uganda." Quarterly Journal of Economics, 119 (2): 678 - 704.
Reno, W. 1999. Warlord Politics and African States. Boulder: Lynne Rienner.
Reynolds, A. 1999. Electoral Systems and Democratization in Southern Africa. Oxford: Oxford
University Press.
Riker, W. 1980. “Implications from the Disequilibrium of Majority Rule for the Study of
Institutions.” American Political Science Review 74: 432 - 446.
Rogers, M. and D. Lee. 2012. “Uneven Development, Unequal Education: The Divergent Effects
of Federalism on Education Spending.” Working Paper. Department of Politics and
Policy, Claremont Graduate University.

356

Rondinelli, D., J. McCullough, and R. Johnson. 1989. “Analyzing Decentralization
Policies in Developing Countries: A Political Economy Framework.” Development and
Change, 20 (1): 57 - 87.
Ross, M. 2001. "Does Oil Hinder Democracy?" World Politics, 53: 325 - 361.
Ross, M. 2004. "Does Taxation lead to Representation?" British Journal of Political Science, 34:
229 - 49.
Ross, M. 2006. "Is Democracy Good for the Poor?" American Journal of Political Science, 50
(4): 860 – 874.
Ross. E. 2012. The Oil Curse: How Petroleum Wealth Shapes the Development of Nations.
Princeton: Princeton University Press.
Rourke, F. 1984. Bureaucracy, Politics, and Public Policy. (3rd ed.) Boston: Little, Brown.
Rustow, D.A. 1970. "Transitions to Democracy: Toward a Dynamic Model," Comparative
Politics 2 (3): 337 - 63.
Sanni, A. 2007. “Tax Adminisration and the Constitution: Issues on Multiplicity of Taxes and
Federation Account.” Presented at FIRS Training Program on Tax Enforcement and
Appeal Procedures, Abuja: June 13 – 15, 2007.
Schmitter, P. and T. Karl. 1991. “What Democracy is -- and Is Not.” Journal of Democracy, 2:
75 - 88.
Schneider, S. and W. Jacoby. 1996. “Influences on Bureaucratic Policy Initatives in the
American States.” Journal of Public Administration Research and Theory, 6 (4): 495-522.
Snyder, R. 2001. “Scaling Down: The Subnational Comparative Method.” Studies in
Comparative International Development, 36 (1): 93 – 110.
Snyder, R. and R. Bhavnani. 2005. “Blood, Diamonds, and Taxes: Lootable Wealth and Political
Order in Sub-Saharan Africa.” Journal of Conflict Resolution 49 (4): 563 – 597.
Spicer, M. and S. Lundstedt. 1976. “Understanding Tax Evasion.” Public Finance, 21 (2): 295 –
305.
Stokes, S. 1999. “Political Parties and Democracy.” Annual Review of Political Science 2 (1):
243 - 267.
Suberu, R. 2001. Federalism and Ethnic Conflict in Nigeria. Washington D.C.: United States
Institute of Peace Press.

357

Suberu, R. 2003. “Nigeria: Dilemmas of Federalism ” In W. Reno (ed) Nigeria After
Military Rule: Federalism, Resource Allocation, and Ethnic Politics, Evanston: Program
of African Studies, Northwestern University.
Suberu, R. 2004. “The Politics of Fiscal Federalism in Nigeria.” In U. Amoretti and N. Bermeo
(eds) Federalism and Territorial Cleavages, Baltimore: Johns Hopkins University Press.
Tansey, O. 2007. “Process Tracing and Elite Interviewing: A Case for Non-Probablilty
Sampling,” PS: Political Science and Politics, 40 (4).
Tiebout, C. 1956. “A Pure Theory of Local Expenditures.” Journal of Political Economy, 64:
416 – 424.
Tilly, C. 1985. "War Making and State Making as Organized Crime." In P. Evans, D.
Rueschemeyer and T. Skocpol (eds) Bringing the State Back In, London: Cambridge
University Press.
Tilly, C. 1990. Coercion, Capital, and European States, AD 990-1992. Cambridge: Blackwell.
Timmons, J. 2005. "The Fiscal Contract: States, Taxes, and Public Services." World Politics, 57:
530 - 567.
Uche, C. and C. Uche. 2004. “Oil and the Politics of Revenue Allocation in Nigeria.” African
Studies Centre, Lieden Netherlands, ASC Working Paper 54/2004.
Usman, P. 2007. “Resource or Revenue Sharing in Nigeria.” In J. Elaigwu (ed) Fiscal
Federalism in Nigeria: Facing the Challenges of the Future, Jos: Aha Publishing House.
van de Walle, N. 2001. African Economies and the Politics of Permanent Crisis, 1979-1999.
New York: Cambridge University Press.
Wantchekon, L. 1999. "Why Do Resource Dependent Countries Have Authoritarian
Governments?" Department of Political Science: Yale University.
Wantchekon, L. 2002. "Why Do Resource Abundant Countries Have Authoritarian
Governments?" Department of Political Science: Yale University.
Wantchekon, L. 2003. “Clientelism and Voting Behavior: Evidence from a Field Experiment in
Benin.” World Politics, 55: 399 - 422.
Watts, M. and P. Lubeck. 1983. “The Popular Classes and the Oil Boom: A Political Economy of
Rural and Urban Poverty.” In I. Zartman (ed.) The Political Economy of Nigeria. New
York: Praeger.
Weingast, B. 1981. “Regulation, Reregulation, and Deregulation: The Political Foundations of
Agency Clientele Relationships.” Law and Contemporary Problems, 44: 147-177.

358

Weingast, B. 1984. “The Congressional-Bureaucratic System: A Principal-Agent
Perspective (with Applications to the SEC).” Public Choice, 44: 147-177.
Weingast, B. 1995. “The Economic Role of Political Institutions: Market-Preserving Federalism
and Economic Growth.” Journal of Law, Economics and Organization, 11: 1 – 31.
Weingast, B. and M. Moran. 1983. Bureaucratic Discretion or Congressional Control?
Regulatory Policymaking by the Federal Trade Commission. Journal of Political
Economy, 91: 765-800.
Weinstein, L. 2008. “The Politics of Tax Mobilization: Evidence from Sub-Saharan Africa.
Qualifying Field Paper. Department of Political Science, University of California, Los
Angeles.
Welden, S. 2001. Independent Sources of State Revenue in Developing Countries: The Impact on
Democracy, Development, and Human Rights (Doctoral Dissertation). Ann Arbor:
ProQuest/UMI (Publication No. 3000226).
Williams, R. 1987. Political Corruption in Africa. Aldershot, U.K.: Gower.
Wilson, J. 1989. Bureaucracy New York: Basic Books.
Wolpe, H. and S. McDonald. 2006. “Training Leaders for Peace.” Journal of Democracy. 17 (1):
132 - 138

359