University of Ghana http://ugspace.ug.edu.gh COLLEGE OF HUMANITIES UNIVERSITY OF GHANA DISCOVERY OF OIL IN AFRICAN COUNTRIES: BLESSING OR CURSE TO THE INDIGENOUS CITIZENS? BY ANTWI EMMANUEL ASARE (10702228) THIS LONG ESSAY IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON, IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE AWARD OF MASTER OF SCIENCE (M.S.C) DEGREE IN DEVELOPMENT FINANCE AUGUST 2019 University of Ghana http://ugspace.ug.edu.gh DECLARATION I, ANTWI EMMANUEL ASARE, hereby declare that this thesis is the original research undertaken by me under the guidance of my supervisor. Neither the whole nor a part of this thesis has been presented for another degree elsewhere. All references used in this work have been accordingly acknowledged. ………………………………….. …………………………………. ANTWI EMMANUEL ASARE Date (10702228) i University of Ghana http://ugspace.ug.edu.gh CERTIFICATION I hereby certify that this thesis was supervised in accordance with procedures laid down by the University. I declare that I have supervised the above student in undertaking the study reported herein and confirm that he has my permission to submit it for assessment. ……………………….. …………………………… Dr. ELIKPLIMI K. AGBLOYOR Date (Supervisor) ii University of Ghana http://ugspace.ug.edu.gh DEDICATION To my mum (Mrs. Christiana Adjei) and dad (Mr. Samuel Antwi). GOD BLESS YOU ALL. iii University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENTS My greatest gratitude goes to the Almighty God for giving me life, good health, and strength to go through this programme. My profound gratitude goes to Dr. Elikplimi Komla Agbloyor for his good supervision, constructive criticisms, suggestions, and advice in writing this thesis. I also appreciate and thank all the lecturers at the Department of Finance for their suggestions during the proposal presentation. My next thanks go to my siblings (Mrs. Gertrude Arthur, Josephine, Christiana, and Richard) and all of my family members for their immeasurable support during my studies. To all my program colleagues and research assistants at the University of Ghana Business School who played various roles during my studies, I say a big thank you for your support. iv University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ............................................................................................................................. i CERTIFICATION .......................................................................................................................... ii DEDICATION ............................................................................................................................... iii ACKNOWLEDGEMENTS ........................................................................................................... iv TABLE OF CONTENTS ................................................................................................................ v LIST OF TABLES ........................................................................................................................ vii LIST OF ACRONYMS/ABREVIATIONS ................................................................................. viii ABSTRACT ................................................................................................................................... ix CHAPTER ONE ............................................................................................................................. 1 INTRODUCTION .......................................................................................................................... 1 1.1 Background of the Study .................................................................................................. 1 1.2 Problem Statement ........................................................................................................... 3 1.3 Objectives of the Study .................................................................................................... 4 1.4 Research Questions .......................................................................................................... 5 1.5 Significance of the Study ................................................................................................. 5 1.6 Scope of the Study............................................................................................................ 6 1.7 Limitations of the Study ................................................................................................... 6 1.8 Organization of the Study ................................................................................................ 7 CHAPTER TWO ............................................................................................................................ 8 LITERATURE REVIEW ............................................................................................................... 8 2.1 Introduction ...................................................................................................................... 8 2.2 Theoretical Review .......................................................................................................... 8 2.2.1 The General Dutch Disease Model ........................................................................................ 8 2.3 Empirical Review ............................................................................................................. 9 2.4 Natural Resource and Governance ................................................................................. 11 2.4.1 Institutional and Regulatory Design .................................................................................... 13 2.4.2 Fiscal Regimes ..................................................................................................................... 13 CHAPTER THREE ...................................................................................................................... 15 METHOD OLOGY ....................................................................................................................... 15 3.1 Introduction .................................................................................................................... 15 v University of Ghana http://ugspace.ug.edu.gh 3.2 Research Design ............................................................................................................. 15 3.3 Target Population and Sample Size ............................................................................... 15 3.4 Data Source and Description .......................................................................................... 15 3.5 Variables Description ..................................................................................................... 16 3.6 Estimation Strategy ........................................................................................................ 17 CHAPTER FOUR ......................................................................................................................... 18 DATA ANALYSIS AND DISCUSSION OF RESULTS ............................................................ 18 4.1 Introduction ......................................................................................................................... 18 4.2 Descriptive Analysis ...................................................................................................... 18 4.3 Empirical Results and Discussion .................................................................................. 20 4.4.1 Effect of Oil Revenue on Pro-Poor Investment .......................................................................... 22 4.4.2. The Role of Democracy ....................................................................................................... 23 CHAPTER FIVE .......................................................................................................................... 24 SUMMARY, CONCLUSION AND RECOMMENDATIONS ................................................... 24 5.1 Introduction .................................................................................................................... 24 5.2 Summary and Conclusion .............................................................................................. 24 5.3 Recommendations .......................................................................................................... 26 5.4 Delimitation of the Study ............................................................................................... 27 REFERENCES ............................................................................................................................. 28 APPENDIX ................................................................................................................................... 31 vi University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 4.1: Summary statistics of variables,………………..………………………………...... 18 Table 4.2: Correlation matrix ((Full Sample),…………………...….…..………………...…....19 Table 4.3: Correlation matrix (Democracy & High Producing) ..................................................21 Table 4.4: Correlation matrix (Democracy & Low Producing) ………………………………. 21 Table 4.5: Correlation matrix (Autocracy & High Producing)…………………………………21 Table 4.6: Correlation matrix (Autocracy & Low Producing) …………………………………21 vii University of Ghana http://ugspace.ug.edu.gh LIST OF ACRONYMS/ABBREVIATIONS DD Dutch Disease EITI Extractive Industries Transparency Initiative GDP Gross Domestic Product MDG Millennium Development Goals NR Natural Resource PPI Pro-Poor Investment RC Resource Curse SDG Sustainable Development Goals WDR World Development Report viii University of Ghana http://ugspace.ug.edu.gh ABSTRACT For a nation to grow faster, it must direct its oil revenue on pro-poor investment. It must be noted also that, reforms in the health and educational sectors are cornerstones of a nation. This means that investment in health and education is so important that a country should dedicate its natural resource revenue towards these sectors. The emphasis on investment in health and education cannot be taken for granted. As a result, this study empirically examines the influence of natural resource (oil) rent on pro- poor investment in Africa. The research used data from the World Bank’s World Development indicators and polity IV dataset for 17 African countries over the period 2000 to 2017. The study employed a simple pair-wise correlation estimation technic in the estimation. Taking the level of democracy and oil production level into consideration, the study found that oil rent influences pro-poor investment in one way or the other in Africa. The study also reveals that the democracy level plays a role in the oil rent-pro poor investment nexus, particularly investment in education. To be specific, increase in oil rent tend to increase educational investment in democratic years but detrimental to educational investment in autocratic years. We, thus, conclude that, for natural resource (oil) revenue to be effective in enhancing pro-poor investment, it requires a complementary democratic environment in Africa. The study, therefore, points out the need for domestic policymakers of oil-producing African countries to concentrate on the exploration of oils resource in their respective countries alongside ensuring democratic environment. ix University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.1 Background of the Study The government should play an active role in alleviating poverty and enhancing economic growth. ‘‘A more active role can entail enhancing budget-neutral reforms, such as improvements in legislation, regulations, and rules to attract private investments” (Harvey, 2009). This will lead to increased government expenditures, either through shifts in existing budgets or obtaining supplementary resources. However, increasing public investments raises relevant questions on the fiscal disciplines and institutional requirements to channel scarce resources into developmental projects. Additional financial resources for public investments can come from heterogeneous sources, such as domestic sources (taxes) or foreign inflows (grants, bonds, resource rents). For many African nations, the likelihood of obtaining additional revenue is confined to grants and revenues generated from natural resources given the generally restricted scope for tax increment and restricted access to international financial markets. Nonetheless, most African countries are encountering or are about to encounter a gush inflow of foreign investments. For most developing countries such as Ghana and Uganda where oil has been discovered, the governments of these nations expect the revenue generated from the oil find to augment public expenditures. However, the use of such foreign exchange inflows in financing public projects can have serious and unfavorable effects on national growth. Verifiable corroboration of whether foreign inflows are favorable or unfavorable for growth is ambiguous. Diamond and Mosbacher, 2013, found that ‘‘nations with high resource-exports-to-GDP ratios face lower growth rates” other studies 1 University of Ghana http://ugspace.ug.edu.gh shows that “resource abundance has a neutral or even positive impact on growth” (Rodríguez- Pose, 2010). The rationale for justifying the probable negative effects on growth is in two folds: Inadequate managerial capacity of the already overwhelmed public servants and the unfavorable impact of government spending on policies which do not yield any national interest. Public investment decisions are also made on the premise of their growth and poverty effect, and a salient policy question is where to invest? Public investment decisions can be geared towards rural or urban areas or sectors. The argument is perhaps best reflected in the World Development Reports (WDR) 2008 and 2009. These reports communicate two distinct views of how to speed up growth and poverty reduction in third world countries (World Bank 2008 & 2009).The WDR 2008 sees “agriculture as a vital development tool for achieving the Millennium Development Goals (MDGs)” while the World Development Reports 2009 concludes that “Growing cities, ever more mobile people, and increasingly specialized products are integral to development”. For a nation to grow faster, it must direct its oil revenue into pro-poor investment. It must be noted also that reforms in the health and educational sectors are cornerstones of a nation. This means that investment in health and in education is an important completion point trigger which a country should dedicate its natural resource revenue. The emphasis on investment in health and education cannot be taken for granted. Investment in education, for example, enhances the quality of human capital accumulation which is intertwined with economic growth (Vogl, 2014). In the view of Krueger and Lindahl (2001), education makes workers more efficient and productive, accelerates the advancement of new technologies, and enables guardians to raise skilled children, all of which boost economic growth and help reduce poverty. 2 University of Ghana http://ugspace.ug.edu.gh In line with this school of thought, the Government of Ghana focused on spending most of the oil revenue on education, health, and infrastructure. However, there are few analytical studies that have examined the impact of the appropriation of the newfound oil wealth on pro-poor investment such as education and healthcare services by government. 1.2 Problem Statement Over the years, researchers in economic development have presupposed that discovering scarce natural resources (oil) was a blessing as posited by Norton Ginsburg in 1957. "The possession of a sizable and diversified natural resource endowment is a major advantage to any country embarking upon a period of rapid economic growth”. But experts have come to the realization that the opposite is true. Auty (2001) made an assertion that the natural resource-poor nations have out-performed their natural resource-rich counterparts by a significant margin. Natural resource (oil) dependence coupled with the volatility of prices of the resource on global markets tends to bring significant fiscal planning problems, leading to financial catastrophe and bringing down public spending quality as soon as prices fall. Aluko (2004) notes that oil nations like Ghana, Nigeria, Cameroon, and Angola tend to borrow quicker and deeper to take care of unforeseen income shortfalls compared to non-oil nations. Oil revenue as a pillar of development seems to have fallen short. Oil-reliant governments and leaders allocate an excessive amount of funds in security and military expenditure (Karl & Gary, 2004). This puts a dent on the rule of law and democracy of developing countries. Deichmann, Gill & Goh, (2010) note that ‘‘increases in oil revenue are strongly correlated with greater corruption, authoritarianism, political and economic instability, and civil war’’. 3 University of Ghana http://ugspace.ug.edu.gh In light of the natural resource curse syndrome, countries that recently discover large economic values of a natural resource have attempted to put in measures to avoid this syndrome. For instance, Ghana upon discovering petroleum resources in 2010 adopted legislative structure (Petroleum Regulatory Authority Bill) for the utilization of the expected revenue. Following the Petroleum Regulatory Authority Bill several strategic investments have been made with funds from petroleum revenue with about 1,800,000 barrels of crude sold over the past nine years. These investments have especially focused on infrastructure and public services such as education and healthcare. In 2015, Nigeria also reformed its petroleum sector aimed at reversing poor management, inappropriate management of revenue, incoherent fiscal and regulatory provisions and gross inefficiencies in managing the downstream petroleum assets. Despite these investments and new reforms in the petroleum industry across Africa, the question that remains is, however, has the natural resource-rich countries taken advantage of the proceeds of the oil revenue to increase pro-poor investment especially in Health and Education? This question requires a compelling empirical answer because there are some oil-rich countries such as Ghana, who are said to have dedicated part of their oil revenue to enhance education and health. As a result, this study tries to provide an answer to the question, whether natural resource (oil) discovery does more harm than good to African countries by exploring the effect of oil revenue on pro-poor investment (in this case, investment in education and health) and to investigate the role the level of democracy plays in African countries. 1.3 Objectives of the Study Generally, the research attempts to assess whether the discovery of a natural resource (oil) in African nations is a curse or a blessing to the citizens. The answer is sought by exploring the 4 University of Ghana http://ugspace.ug.edu.gh effect of oil revenue on pro-poor investment (in this case, investment in education and health) in African countries. Specifically, the study aims to investigate: 1. The impact of oil revenue on pro-poor investment (education and health) in African countries. 2. The role that the level of democracy plays in the link between oil revenue and pro-poor investment in African countries. 1.4 Research Questions In order to attain the objectives of this research stated above, the questions below will serve as a guide:  Does an increase in oil revenue influence pro-poor investment (education and health) in African countries?  Does the level of democracy play a role in this nexus in African countries? 1.5 Significance of the Study The outcome of this research will provide verifiable knowledge to the natural resource literature and provides an additional understanding of how African countries can improve the welfare of their citizens with the discovery of natural resources. This will be useful to academics and policymakers. This work will add to extant works in three (3) ways. Firstly, the research will add significant knowledge to the literature by assessing the relationships between oil revenue and pro-poor investment. The research will assess whether oil revenue has any notable impact on investment in education and health in developing countries and the role, 5 University of Ghana http://ugspace.ug.edu.gh the level of democracy plays. Secondly, our exclusive focus on oil-rich African nations provides new presentiments to make us understand the performance and behavior of developing countries, and this will help the development of new ways for achieving economic growth which is key in poverty lessening. Finally, the finding of this study will initiate further research, and help leaders of African countries to pay serious attention to investment in education and health which will go a long way to propel development. It will also be used as a reference by other scholars; that is, it can lessen the gap in the literature in the area of the study particularly in developing countries and will guide government policymakers on enacting appropriate policies for the effective use of the oil revenue. 1.6 Scope of the Study The research exclusively concentrates on natural resource (oil) rent and the pro-poor investment nexus in African countries. It uses data collected on seventeen oil-producing African countries from the World Bank’s World Development Indicators and the Polity IV dataset. 1.7 Limitations of the Study The research is restricted to oil-producing African countries. It uses simple correlations to determine the link between oil rent and pro-poor investment. This study might not be able to draw a robust conclusion due to the gaps in the data set. Further, oil rent is used as a substitute for oil revenue. These problems do not mean that the results from the research will be of less use for policymaking and implementation. 6 University of Ghana http://ugspace.ug.edu.gh 1.8 Organization of the Study The research is organized into five chapters. Chapter One provides a general introduction to the research and presents the background of the research, problem statement, purpose, research objectives, research questions, the significance of the research and scope of the research. The second chapter reviews exiting literature germane to the research. Chapter Three examines the methodology of the research, which specifies the empirical model and empirical estimation methods used for the study. Chapter Four reports on the empirical estimation results and analysis. Chapter Five presents a summary, conclusions, and recommendations. 7 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.1 Introduction This chapter examines germane theoretical and empirical evidence of oil as a natural resource and economic performance. This section discusses the theoretical reviews on the general natural resource model. The review of previous studies is then presented. 2.2 Theoretical Review For a theoretical discussion of a natural resource (oil) and pro-poor investment, we need to consider the generic model of the Dutch disease. 2.3 The General Dutch Disease Model The “Dutch disease” model has to do with an open economy which considers three sectors, the first produce non-tradable goods and the last two produce goods that are tradable. The tags ‘resources’, ‘services’ as well as ‘manufacturing’ are frequently referred to. Therefore, Corden (1984) referred to these three sectors as lagging (L), booming (B) and nontradable (N). Normally, such a proposition begins in equilibrium, then one can hypothesize a shock to sector B by an expansionary policy. In the actual situation, such a shock may be temporary  or permanent, and its close source may be to find and bring  on board a new discovery  of exhaustible  oil resources (the cases  of Norwegian  or Netherlands), the growth  of a dependable services (financial) sector that is well-controlled (the cases  of UK and Switzerland),  or an upsurge in the price  of certain resources relatively (the cases  of Canadian and Australian). This can be an answer to a structure of change of demand coming from somewhere else in the global economic 8 University of Ghana http://ugspace.ug.edu.gh growth, shocks emanating from electoral uncertainty in different providing nations, or maybe asset market reactions to worries concerning policies of the macroeconomy that weaken the feasibility of the main currency in the long-run in which prices of the said resources are quoted. However, whatsoever cause the shock, generally, the model indicates that its penalties for the sector of the economy that works via a diversity of networks will have greater output and input prices in addition to a development of an action in sector B, a decrease in output and input prices and a reduction of activity in sector L, and a greater non-tradable output and input prices together with unclear penalties for the output of the sector. This fundamental framework will be able to take care of most disparities. The shock that happens to sector B might be temporary, permanent or recurring. It has the capability to generate motivations for the inflow of capital that highlight the increase of the real exchange rate; functions of production can be constructed to take care of low or high grades of fact or movements in-between sectors; as well. The labour markets may be described by the resistance to offer a real wage to make unemployment in sector L, rather than a shift of labour simply away out of it, may turn out to be one of the Dutch disease symptoms. 2.4 Empirical Review Whereas the theoretical argument of the Dutch Disease has been noticeable for over three decades, it has mixed extant literature. Auty (2001) demonstrates that the experiences of nations rich in resource are varied. Whereas some nations have undergone/experienced a ‘resource curse’, others have successfully used their resources productively. As a result, it is less astonishing that we have mixed empirical conclusions with regard to the “Dutch Disease”. 9 University of Ghana http://ugspace.ug.edu.gh Proof of the Dutch Disease exists. For example, some writers including Sachs and Warner (1995, 1997) have demonstrated that the manufacturing sector shrinks while the concurrent service sector expands. Additionally, natural resources are inversely connected to both cross country (Sachs & Warner, 1995, 2001) and within countries (Bruno & Sachs, 1982) growth. In contrast, other scholars have revealed proof opposing the theory of Dutch disease. For instance, Katarzyna Borysiak (2017) discussed the petroleum management model of Norway and its effect on establishing an equal and efficient economy. Achievements of the petroleum sector can be attributed to the ideological welfare conditions and a properly structured institutional framework. Also, continuity and predictability of the sector result in social and political trust. He concluded that the Norwegian model of oil and gas management, although very efficient, is not practicable for a worldwide framework (that is, not all countries in the world will be able to implement this design). Otaha (2012) also posit that the Norwegian manufacturing sector was really blessed by the sighting of oil resources. Nevertheless, other researchers have established that there exists no proof of any impact of oil resources on the indicators of the macroeconomy. Which is applied to both cross-country studies (Leite & Weidmann, 1999; Bean, 1988) and single-country case studies alike (Sala-i Martin & Subramanian, 2003). Bunte (2011) examined how institutional features that were in place during the time of discovery of the resource might stop the Dutch disease from happening in the first place. He refers to how capitalism works in various ways and discovered that the institutional features of market economies that are coordinated assist in preventing the Dutch Disease from happening, whereas market economies that are liberal feel its impact. The causal factors recognized and tested are 10 University of Ghana http://ugspace.ug.edu.gh three. First is the level of wage bargaining coordination, secondly, the factor specificity degree, and finally, the degree of disparity in income were established as causally linked to the Dutch Disease severity. Additionally, in 2012, Dartey-Baah, Tawiah, and Aratuo, examined the Ghanaian economy in connection with its economic indicators and the achievement of the agricultural sector against reports suggesting the resource curse. They intended to furnish lawmakers with the mechanisms needed to determine the Dutch disease syndrome as the country shifts from an agricultural to a petroleum economy. Their paper questioned a World Bank (2009) report branding the economy as exhibiting the Dutch disease symptoms. It was concluded that numerous policies by the government were geared toward improving agriculture despite the emergence of oil and gas (Dartey-Baah, et.al. 2012). Tawiah, Dartey-Baah, and Osam (2015) investigated the link between the various stakeholders in the petroleum sector and advances measures aimed at turning the possibility of a curse into blessings. It was identified that expectations of stakeholders (citizens) that oil would be the remedy to solve all difficulties must be controlled. Also, Corporate Social Responsibility (CSR) policies must be planned with all stakeholders for economic growth, which will enable an amicable environment for oil companies. It was further pointed out that answerability and clarity are also necessary to improve collaboration among all stakeholders in the oil industry. 2.5 Natural Resource and Governance There has been progress regarding contract transparency in oil-producing African countries, however, the effect of the approaches made on improved accountability is largely not known. Collaborative, adaptive and flexible management practices are needed to be coupled with a good 11 University of Ghana http://ugspace.ug.edu.gh political economy understanding to adjust to varying conditions and local context. Among the key factors needed to transform resource income into sustainable development goals are sustainable and sound fiscal regimes. And leadership that prioritizes the use of revenues for the benefit of all at heart. Resource income can possibly help to achieve development goals, nevertheless, most resource- rich nations are characterized by poor governance, unsustainability, dysfunctional institutions, conflicts, corruption, and low economic growth. Resource management transparency via initiatives like the Extractive Industries Transparency Initiative (EITI) has been given recognition with the aim of increasing accountability so as to address the ‘resource curse’.  Oil resources can give a country considerable income. As it can contribute to livelihoods, improve the green economy and food security. It can also generate trade at national and international levels. But considering the huge amount of resource income at stake, corruption and poor governance are always associated with the extractives sector. The huge revenue generated by oil resources provide ways for corruption, rent-seeking, and patronage. Risks of corruption can take place at every stage of the extractive value chain, starting from the awarding of oil contracts to public spending (OECD 2016). Therefore, the governance of extractive companies is important in getting a feel of whether oil resources can be beneficiary or harmful to a nation. Different types of governance are very important to guarantee sustainable exploitation, fair distribution and to maximize the influence of natural resources to development goals. Contract transparency and income via initiatives like the EITI have been given cognizance as a way of increasing accountability in line with the management of natural resource income. 12 University of Ghana http://ugspace.ug.edu.gh Transparency alone from extractive industries is, however, not likely to achieve sustainable development goals. Transparency should be augmented by strong institutions and regulatory frameworks, collaborative and flexible management practices, local-level understanding of the political economy and committed leadership that aim to use the income to enhance sustainable development goals (DFID 2016). 2.5.1 Institutional and Regulatory Design Armstrong (2003) acknowledged that a good institutional and regulatory design aid most nations develop an economy that attracts foreign direct investments (FDI), compete in the world market, restore investor confidence, and increase and national growth. For instance, Thurber, Hults & Heller (2011) also examined the petroleum sector management model of Norway and concluded that the segregation of powers in a regulatory institution has been instrumental for the sector’s effectiveness. Rules and regulations need to be established conforming to the customs of each nation, which aid economic growth. Inyang (2009) notes that some rules and laws need to be established by various nations to help the economy more attractive to foreign direct investment. These institutional and regulatory design aid in the achievement of national objectives. 2.5.2 Fiscal Regimes Fiscal policies are important in the valuation of the resources as revenue to the State. Barma & Naazneen (2012) found that tax policy should be progressive and profit-oriented. They concluded that fiscal policies are more difficult to practice in developing nations with feeble 13 University of Ghana http://ugspace.ug.edu.gh capacity and governance than in nations operating under strong technical capacity and governance. Ushie, Oluwatosin & Akongwale (2013) note that it is relevant to have sound policies that regulate and manage the upstream sector. The fiscal policy should operate in a way that will enhance economic and fiscal discipline to the State, making the sector more lucrative to oil industry players having the technical capabilities to explore and produce. 14 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE METHOD OLOGY 3.1 Introduction This chapter begins by discussing the models and the econometric techniques employed to analyze the data for the study. It then proceeds with the discussion of the variables, sample, and source of data. 3.2 Research Design The purpose of the study is to investigate the effect of oil revenue on pro-poor investment. To achieve this purpose, the study uses a descriptive research design to obtain information about the phenomenon under study. Descriptive research was employed to examine specific behaviour as it occurs in the environment. A simple correlation technique is an econometric technique used to estimate variables of the study. 3.3 Target Population and Sample Size The population of the study consists of all oil-producing countries in Africa. The technique used to select a sample of 17 countries for the study is based on the availability of the democracy variable (P olity2). 3.4 Data Source and Description The study uses data from 17 oil-producing countries in African countries over the period 2000 to 2017 (18years). However, the data is unbalanced due to gaps in years for some of the countries. The data on all the variables under study are available on World Bank’s world development indicators website and Polity IV dataset for the polity2 data. This source of data is also used by 15 University of Ghana http://ugspace.ug.edu.gh most researchers and so it is legitimate for the purpose of this study. The list of the countries included in the study is found in appendix. 3.5 Variables Description The main variables used to assess the effect of oil rent on pro-poor investments are oil revenue, investment in education as a percentage share of GDP and investment in Health as a percentage share of GDP. Oil revenue is measured by oil rent as a share of GDP. (a) Pro-poor investment Variables (Education and Health) Two dependent variables are used in the empirical analysis. Investment in education as a percentage share of GDP and Investment in Health as a percentage share of GDP, both used as a proxy for pro-poor investment. The proxy for pro-poor investment, thus, health and educational investments follow the prioritization of education and health by all HIPCs in their respective definitions of pro-poor expenditure, although some include agriculture and other pro-poor initiatives based solely on government priorities (Paternostro 2007; Simson, 2012). (b)  Oil revenue (oil rent) The oil rent is a continuous variable and is used to measure revenue from oil. It is measured by oil rent as a share of GDP.  Oil rents are the difference between total production costs and crude oil production value at world prices. (c) High Oil Producing dummy This is a dummy variable that takes the value 1 if the oil rent of a country is above the sample mean of 15.6% for a particular year and 0 otherwise. 16 University of Ghana http://ugspace.ug.edu.gh (c) Democracy dummy This is a dummy variable developed from the polity2 index. The polity2 index is an autocracy- democracy index (polity2) ranging between -10 (total autocracy) and 10 (total democracy) from the Polity IV dataset. The democratic variable takes the value 1 for positive values of polity2 index and 0 for negative values. 3.6 Estimation Strategy The estimation strategy is a simple correlation estimation technique. The data are put into four other sub-groups apart from the full sample. Thus, years that recorded high  oil revenue and high democratic tendencies, years that recorded high  oil revenue and high autocratic tendencies, years that recorded low  oil revenue and low democratic tendencies, and years that recorded low  oil revenue and low autocratic tendencies. We then run a simple pairwise correlation for the various groups, pegging the significance level to 5%. 17 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR DATA ANALYSIS AND DISCUSSION OF RESULTS 4.1 Introduction In this chapter, the study presents and discusses issues concerning the analysis. The correlation analysis is done using STATA. The chapter starts with a descriptive analysis of the variables employed in the study. Then the estimations of the actual results are presented and discussed. 4.2 Descriptive Analysis In this section, a brief discussion of the fundamental statistical properties of the variables used in the model over the period 2000 to 2017 is giving. Among the summary statistics examined are the mean, standard deviation, minimum and maximum values for the pooled sample. Table 4.1 below shows the details: Table 4.1: Summary statistics  of variables Variable Obs Mean Std. Dev. Min Max  Oil rent 302 15.60135 17.3637 0 78.55187 Edu. Expenditure 94 18.0669 7.393768 7.76979 45.35416 Health Expenditure 284 4.351592 1.551648 1.282307 8.19779 Democratic 306 0.490196 0.500723 0 1 High  oil producing 306 0.385621 0.487539 0 1 Source: Author’s computation using STATA The standard deviation column from Table 4.1 above measures the dispersion of the variables from their means. The presence of outliers is indicated by large standard errors which 18 University of Ghana http://ugspace.ug.edu.gh significantly influence the data. The difference between the maximum and minimum values of the variables can also help to determine the spread. The bigger the gap of a variable, the larger the standard deviation of the said variable.   Health Expenditure, it was observed that on average the countries used in the study have health expenditure of about 4.35 percent. The maximum and minimum values of Health Expenditure are 1.28 percent and 8.20 percent respectively over the period. The average of Education Expenditure for the period is about 18.06 percent, with about 7.76 percent minimum and about 45.35 percent maximum. The mean value of Oil Rent is about 16.60 percent, with a 0 percent minimum and about 78.55 percent maximum.  On the democracy and high producing dummies, democratic years recorded about 49 percent while high producing years recorded about 39 percent. Table 4.1 shows a detailed breakdown of the rest of the variables. Table 4.2 below shows the correlation between the variables used in this study for the full sample. Table 4.2: Correlation matrix (Full Sample)  Oil Rent Educational Expenditure Health expenditure  Oil Rent 1 Educational Expenditure 0.3290* 1 Health expenditure -0.6111* -0.3837* 1 Source: Author’s computation using STATA 19 University of Ghana http://ugspace.ug.edu.gh From Table 4.2, the correlation between all the independent variables is generally low (below 0.70) but statistically significant at 5% confidence level. The general low correlations mean that there is less collinearity among the variables. The correlation coefficients indicated between educational expenditure and oil rent that an increase in oil rent leads to an increase in educational investment, however, the reverse is true for the coefficient of health expenditure and oil rent, where the results show that increase in oil rent tends to reduce health investment. This is probably explained by the inverse relationship between health and educational expenditures which suggests that increase in educational investment tends to reduce health investment and vice versa. 4.3 Empirical Results and Discussion The results shown in Tables 4.3, 4.4, 4.5 and 4.6 are obtained from estimating the four sub- categories indicated in section 3.6 above using simple correlations. As a robustness check, investments in health and education are used to measure the countries’ pro-poor investments. 20 University of Ghana http://ugspace.ug.edu.gh Table 4.3: Correlation matrix (Democracy & High Producing)  Oil Rent Educational Expenditure Health expenditure  Oil Rent 1 Educational Expenditure -0.2711 1 Health expenditure -0.3129* -0.5837* 1 Source: Author’s computation using STATA Table 4.4: Correlation matrix (Democracy & Low Producing)  Oil Rent Educational Expenditure Health expenditure  Oil Rent 1 Educational Expenditure 0.3602* 1 Health expenditure -0.4419* -0.6573* 1 Source: Author’s computation using STATA Table 4.5: Correlation matrix (Autocracy & High Producing)  Oil Rent Educational Expenditure Health expenditure  Oil Rent 1 Educational Expenditure -0.5358* 1 Health expenditure -0.5523* 0.0244 1 Source: Author’s computation using STATA Table 4.6: Correlation matrix (Autocracy & Low Producing)  Oil Rent Educational Expenditure Health expenditure  Oil Rent 1 Educational Expenditure -0.2128* 1 Health expenditure -0.2988* -0.5712* 1 Source: Author’s computation using STATA 21 University of Ghana http://ugspace.ug.edu.gh 4.3.1 Effect of Oil Revenue on Pro-Poor Investment From Table 4.3, where we consider democratic and high producing years, the correlation coefficient between oil rent and educational investment is negative but not statistically significant. Therefore, there is not enough evidence to draw any conclusion from it. In Table 4.4, where we consider democratic and low producing years, the correlation between oil rent and educational investment is positive and statistically significant, indicating that, all things being equal, an increase in oil rent leads to an increase in educational investment. However, the correlation coefficients between oil rent and educational investment are negative and statistically significant in the Tables 4.5 and 4.6 where we consider autocratic and high producing years as well as autocratic and low producing years respectively. This indicates that all else equal, an increase in oil rent tends to decrease educational investment. With regard to oil rent and health investment, the coefficient is negative and significant across all tables. This indicates that an increase in oil rent reduces health investment, all else equal. Table 4.3, 4.4 and 4.6 also indicate a trade-off between health and educational investments with the coefficient being insignificant in Table 4.5. This partly explained the positive coefficient in Table 4.4 for democratic and low producing years, in the sense that increase in educational investment as a result  of increase in  oil rent will affect investment negatively. To conclude, our results suggest that oil rent influences pro-poor investment (investments in health and education) in one way or the other, depending on the democracy and oil revenue levels in African countries. 22 University of Ghana http://ugspace.ug.edu.gh 4.3.2 The Role of Democracy A careful look at the results the Tables 4.3, 4.4, 4.5 and 4.6 indicates that democracy plays a role in the oil revenue and pro-poor investment nexus, particularly for oil rent and educational investment. Whereas the correlation between  oil rent and health investment is negative throughout the tables, regardless  of the democracy and oil production levels; that  of  oil rent and educational investment varies in direction, depending largely  on the level  of democracy. In Table 4.4, where the coefficient is significant for democratic years, the results suggest that, in the presence of democracy, increase in oil rent tends to increase educational investment. There is not enough evidence to reach any conclusion on the insignificant coefficient in Table 4.3. On the contrary, in Table 4.5 and 4.6, for autocratic years, and regardless of the amount of oil rent, the results indicate that increase in oil rent reduces educational investment. In conclusion, our results suggest that the level of democracy does play a role in the oil rent pro- poor investment (particularly investment in education). Whereas oil revenue is beneficiary to African countries in the presence of democracy, it is detrimental in autocratic regimes. 23 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE SUMMARY, CONCLUSION AND RECOMMENDATIONS 5.1 Introduction This final chapter presents a summary and conclusion of the study. Presenting the major findings on the impact of oil rent on pro-poor investment in African countries. It also suggests appropriate recommendations for policy implementation, for practitioners and identifies areas for future research. Limitations of the study are also discussed here. 5.2 Summary and Conclusion The discovery of oil can be a curse rather than a blessing to some citizens, especially in the developing world. Most countries in the developing world have similar political misbehavior which directs the oil revenue away from pro-poor expenditure. There are always obvious accountability issues that pave way for corrupt politicians to try to stay in their positions forever. Poor governance, feeble institutions, and mismanagement of public funds are some of the reasons the oil wealth is unevenly distributed among citizens of these nations. Contrary to the above argument, some oil-dependent developing nations have tried to demonstrate fairness towards the utilization of the oil revenue. The building of road infrastructures is springing up, health facilities and services seem to have improved as well as investment in education. Nations that were already democratic and strongly adhere to the rule of law before the discovery of oil have avoided the natural resource curse syndrome. These nations comprise the UK, the 24 University of Ghana http://ugspace.ug.edu.gh USA, Canada, Norway and so forth. A different group of countries has successfully moved to oil blessed from oil doom. They are countries such as Dubai, Mexico, Indonesia and so forth. It must be noted that reforms in the health and educational sectors are cornerstones of a nation. Hence, the study was carried out to examine the impact of oil revenue on pro-poor investment (investment in health and education) in oil-producing African countries, and to investigate the role the level of democracy plays. To achieve these objectives, the study used a quantitative approach and a simple correlation approach. The study used health and educational investments (health expenditure as a percentage of GDP and education expenditure, tertiary (percentage of total) due to availability of data points, as well as oil revenue (oil rent) as main variables to analyze the impact of oil revenue on pro-poor investment in African countries. The study also used democratic and high producing oil rent dummies to assess the role of democracy and production level. Using a simple correlation, the study estimated four sub-groups of the data for 17 oil-producing African countries over the period 2000-2017. Thus, years that recorded high oil revenue and democratic, years that recorded high oil revenue and autocratic, years that recorded low oil revenue and democratic, and years that recorded low oil revenue and autocratic. The study presented a pair-wise correlation between the variables pegging the significance level to 5%. There is not enough evidence to draw a conclusion for democratic and high producing years because the correlation coefficient between oil revenue and educational investment is negative but not statistically significant. In democratic and low producing years, the correlation between oil rent and educational investment is positive and statistically significant, indicating that, all things being equal, an increase in oil rent leads to an increase in educational investment. The 25 University of Ghana http://ugspace.ug.edu.gh correlation coefficients between oil rent and educational investment are negative and statistically significant in autocratic and high producing years as well as autocratic and low producing years. This indicates that all else equal, an increase in oil rent tends to decrease educational investment in autocratic regime. Concerning oil rent and health investment, the coefficient is negative and significant across all four sub-groups. This indicates that an increase in oil rent reduces health investment, all else equal. There is a trade-off between health and educational investments with regards to high oil revenue and democratic year, low oil revenue and democratic year, and low oil revenue and autocratic years. Whereas the coefficient in high oil revenue and autocratic years being insignificant. This partly explained the positive coefficient for democratic and low producing years, in the sense that increase in educational investment as a result  of increase in  oil rent will affect investment negatively. The study found that oil revenue influences pro-poor investment in one way or the other and that the level of democracy plays a role in the oil revenue-pro poor investment nexus. In particular, in the presence of democracy, oil revenue propels pro-poor investment (particularly investment in education) and detrimental to pro-poor investment in autocratic regimes. The study, thus, concludes that for natural resource (oil) revenue to be effective in leading to pro- poor investment, it requires a complementary democratic environment. 5.3 Recommendations The results of this study have provided further insights into the exploration decisions of African countries producing oil. Oil revenue is crucial to sustainability of pro-poor investment and needs 26 University of Ghana http://ugspace.ug.edu.gh to be given much prominence in the quest to change macroeconomic hardships in oil-producing developing countries. The study, therefore, makes the below recommendations: For policymakers, the results of the study reveal that oil rent can influence pro-poor investment positively in the presence of democracy. Therefore, domestic policymakers of oil-producing African countries should concentrate on the exploration of oil resources in their respective countries alongside ensuring a democratic environment. Policymakers should also try as much as possible to reduce, if not completely eradicate, the corrupt practices on the oil sector of their respective countries to achieve the Sustainable Development Goals (SDGs) in Africa. 5.4 Delimitation of the Study The key limitation of this study is that regression analyses were not employed which provide effective use of the panel data and the controlling for other cofactors. Therefore, the study recommends that future works in this area should consider the use of regression analyses and such a study can add more impetus to this study. 27 University of Ghana http://ugspace.ug.edu.gh REFERENCES Aluko, S. (2004). Making natural resources into a blessing rather than a curse in covering oil. A reporter’s guide to energy and development. New York: Open society Institute. Armstrong, M. (2003). A handbook of human resource management: Kogan Page. Auty, R. (2001). The political economy of growth collapses in mineral economies: Mineral and Energy Press. Auty, R. (2004). Political Man. The social basis of politics. Maryland: John Hopkins University Press. Barma, H. N. (2012). Rents to riches? The political economy of natural resource-led development, (New York: World Bank, 2012). Bruno, M. & Sachs, D. (1982). Energy and resource allocation: A dynamic model of the Dutch Disease: Review of economic studies. Bunte, J. (2011). Why does the severity of the Dutch disease vary across countries? University of Minnesota, Department of Political Science. Corden, M. (1984). Booming sector and the Dutch disease: Survey and consolidation:  Oxford Economic papers, 36 pp. 359-380. Dartey-Baah, K., Tawiah, K. & Aratuo, K. (2012). Emerging Dutch disease in emerging oil economy: Ghana's perspective. Society and business review, Vol. 7(2), pp.185-199. Deichmann, U., Gill, I. & Goh, C. (2010). World Development Report 2009: a practical economic geography. Economic geography, 86, 371-380. Diamond, L. & Mosbacher J. (2013). Petroleum to the People: Africa’s coming resource curse, and how to avoid it: Foreign Affairs, 92(5), 86–98. 28 University of Ghana http://ugspace.ug.edu.gh Ginsburg N. (1957). Annals of the association of American geographers. Atlas of economic development, Vol. 47(3), pp. 197-212. Harvey, D. (2009). Reshaping economic geography. World development report 2009. Development and Change, 40, 1269 - 1277. Inyang, B. (2009). Nurturing corporate governance system: the emerging trends in Nigeria journal of business systems, Governance and Ethics, 4(2), 1-13. Karl & Gary (2004). Oil and blood. The way to take over the World. New York: Norton and Co. Katarzyna, B. (2017). Model of state management of petroleum sector: a case of Norway International studies interdisciplinary political and cultural Journal, Vol. 20(1), 97-112. Krueger, A. & Lindhal, M. (2001). Education for growth: Why and for whom? Journal of economic Literature, 6(2): 289-339. Leite, C. & Weidmann, J. (1999). Does mother nature corrupt? Natural resources, corruption, and economic growth: IMF Working Paper 99/85.   Otaha, J. (2012). Dutch disease and Nigeria’s oil economy. Africa Research Review, Vol. 6 (1), Serial No. 24. Paternostro, S., Rajaram, A. & Tiongson, E. (2007). How does the composition of public spending Matter? Oxford Development Studies, Vol. 35(1), March 2007. Rodriguez-Pose, A. (2010). Local and regional development. Local quality of government and voting with one’s feet: The annals of regional science, 55, 2-3. Sachs, D. & Warner, A. (1995). Natural resource abundance and economic growth: NBER Working paper 5398. Sachs, D. & Warner, A. (1997). Fundamental sources of long-run growth. American economic review, 87(2):184-188. 29 University of Ghana http://ugspace.ug.edu.gh Sala-i-Martin, X. & Subramanian, A. (2003). Addressing the natural resource curse: An illustration from Nigeria. Tawiah, K., Dartey-Baah K., & Osam, K. (2015). Turning potential collision into cooperation in Ghana’s oil industry. Society and Business Review, Vol. 10(2), 118-131. The Economist (2006). The Paradox of Plenty. December 24. Thurber, M., Hults, & Heller P. (2010). NNPC and Nigeria oil patronage ecosystem: Oil and governance. Program on energy and sustainable development: Working paper no. 95. United Nations Development Program (UNDP), Human Development Reports, [Accessed October 5, 2016], http://hdr.undp.org/en/countries/ Ushie, V., Oluwatosin, A. & Akongwale, S. (2013). Oil revenue, institutions and macroeconomic indicators in Nigeria. OPEC energy review, Vol. 37(1), 30-52. Vogl, T. (2014). Education and health in developing economies. In A. J. Culyer (Ed.), Encyclopedia of health economics (1st ed., pp. 246–249). Boston, MA: Newnes. World Bank Governance Indicators (2006). 30 University of Ghana http://ugspace.ug.edu.gh APPENDIX Table 1: List of countries included in the study 1. Algeria 10. Gabon 2. Angola 11. Ghana 3. Cameroon 12. Libya 4. Chad 13. Mauritania 5. Congo, Dem. Rep. 14. Niger 6. Congo, Rep. 15. Nigeria 7. Cote d'Ivoire 16. South Africa 8. Egypt, Arab Rep. 17. Sudan 9. Equatorial Guinea 31