Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries

21 Trendy ekonomiky a managemenTu Trends economics and managemenT issn 1802-8527 (Print) / issn 2336-6508 (online) 2017 29(2): 21–33 doi: http://dx.doi.org/10.13164/trends.2017.29.21 Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries Emenike Edwin Ekenechukwu Ezinando, Edirin Jeroh Abstract Purpose of the article: An examination of budget performance’ relationship with fiscal administration in selected countries of sub-saharan Africa is the thrust of this study. Secondary data from 12 countries in Sub-Saharan Africa (SSA) for the period (2002–2016) was sought and analysed. Methodology/methods: Analytical tools deployed for statistical analysis include the Ordinary Least Square (OLS) regression method, correlation analysis, variance inflation factor (VIF) tests for multicollinearity, and Breusch-Pagan/Cook Weisberg test for heteroskedasticity. 5% level of test of significance was employed to measure the altitude of statistical association between all variables. Scientific aim: Efforts were made to empirically ascertain by means of available statistics the relationship between budget deficits and fiscal administration in SSA. Findings: Indications from this study implied that countries SSA recorded trends of significant levels of revenue decline and fiscal balances have deteriorated despite the various acclaimed adjustments in the expenditures profiles by governments in the region. Also a significant relationship was found between budget deficit and fiscal administration among countries in SSA. Conclusions: Since budget deficit has remained a recurring decimal in SSA, countries in the region have resorted to high reliance on borrowed funds to finance the increasing amounts of budget deficits; the consequence being the exposure of countries in SSA to high cost of borrowing. Keywords: Budgeting, Fiscal Deficit, Taxation, Revenue Generation, Government Expenditure, Sub-Saharan Africa JEL Classification: H2, H24, H3, H5 Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 22 Introduction Government budgets have remained crucial to the operation of various economies, and its reciprocal linkage with economic growth has continued to generate series of debates among scholars. Aigheyisi (2013) believes that the effectiveness of governments’ actual expenditure (which presumably is a function of budgeted expenditure) in expanding and engendering rapid growth in any economy depends on whether such expenditures are productive or not. Notwitstanding however, Obinyeluaku (2013) reiterated that in all regions of Sub Saharan Africa (East Africa, Mid Africa, Southern Africa and West Africa), public expenditure is found to have exceeded revenue consistently almost throughout a studied 33 year period (1980–2012). Findings from prior studies reveal that government revenue has dropped from an average of 22 percent of GDP recorded during the 1980s; whereas, public spendings have displayed remarkable increase at unprecedented pace to reach over 28 percent of GDP on the average (Obinyeluaku, 2013; Aigheyisi, 2013; and Chude, Chude, 2013). The consequence of the perceived budget trends in Africa have resulted in widened deficits almost throughout the region. This is deemed a clear indicant of ineffective budgeting process in Sub Saharan Africa (SSA), thus resulting to failure in the promotion/ achievement of the craved culture of balanced budget, budget discipline, and qualitative budgeting for both national and economic development. This is well captured by the instances of budget variances recorded over the years which Onakoya, Somoye (2013) blamed on budget mismatch rather than inadequate provisioning. In the views of Granof, Mayper (1991), large budget variances are evidence that public funds are not spent in line with budgets. Basically, while the focus of most prior studies in Africa were majorly on budgeting and economic growth/development nexus, few studies actually related to analyzing budget deficits as it concerns fiscal administration. Scholarly as this concept may be, it is observed that very few studies have bothered on empirically determining how these fiscal administration indices drive government budgets and budgeting patterns across SSA. Also, with the problems of budgetary allocations/mismatch and variances re-occuring continuously across countries in SSA and the near absence of researches designed to comparatively analyse the subject matter across the region creates further knowledge gap which needs to be filled. It is in a bid to fill these knowledge gaps that this study empirically examined budget deficit and fiscal administration among selected SSA Countries. 1. Literature and conceptual review government budgets and budgeting A budget is a prominent financial document in all sectors of government. The word budget according to Edame, Ejue (2013) was derived from a French word “Bourgettee” which meant leather bag or wallet. Today, the meaning of the term has gone beyond just a wallet that contained papers on the financial plans for any country or organization, but has become a principal instrument of fiscal policy that exercises control over size and relationship of government receipts (revenue) and expenditures (Edame, 2010; and Edame, Ejue, 2013). Budgets are deliberate attempts specifically designed to achieve superior targets over time with available and expected resources that reflects government priorities and citizens’ preferences (Rubin, 2006). Government budgets set forth plans and broadly represent a mixture of how resources of any identified country/jurisdiction are allocated or used in various sectors. Expenditure decisions are therefore made through organised political process usually governed by sets of laws/provisions designed to Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 23 provide open deliberations about options to ensure accountability and prevent corruption, thus providing a mechanism for fiscal control (Mikesell, 2014). Just like every investor, tax payers have the right to all the necessary information that will enable them judge the level of accountability of any government in place especially as it concerns the utilisation of public/national resources. Achieving accountability in this regards has a lot to do with the appropriate basis of accounting for the budget in place at that particular point in time. This is why Omolehinwa (2012), Wynne (2007), Edame (2010) and Edame, Ejue (2013) assert that regardless of the form it takes, through budgets, governments can engage constituents in the polity’s process, and engender confidence by clear – cut representations of the masses’ interests. Hence, budgets are gauged as effective tools needed to boost economic growth among nations (Morrison, Schwartz 1996). Fiscal Administration Government as an institution, no doubt is saddled with a myriad of functions. The way and manner in which these functions are carried out however vary from one country to another. Prior to the popular Great Depression of the 1930s, the market system was generally seen as sacrosanct. The legendary law of “supply creates demand” was at the nucleus of this belief. Consequently, the market system was adjudged as being capable of allocating societal resources equitably to all manner of citizens (Medee, Nenbee, 2011). Impliedly, the general economy and stakeholders like experts, entrepreneurs, government officials, policy makers and the likes folded their hands and allowed the forces of demand and supply to dictate their economic fortune. The resulting complications under this era gave impetus for the increasing demand for a functioning fiscal administration. Fiscal administration is a complex concept with varying connotations. It is the act of managing all monetary transactions and budgets for governments, educational institutions, non-profit organizations, and other public service entities. It is clearly a concept that recognizes the divides of intent between private sector (profit oriented) and public service (non-profit) entities. It establishes a clear trail of agency responsibility for resources intended for use in the provision/rendering of public services by accommodating systems of governance in which government finances serve the interests of the citizenry (Mikesell, 2007). Fiscal administration provides fiscal discipline and responsibility, by ensuring responsible resource allocation and fostering efficiency in government operations (Campos, Pradhan, 1996). Fiscal administration basically concerns itself with how public revenues are generated and allocated, and the direction of government expenditures relative to the pool of generated revenues. It embraces policy formulation, implementation and/or evaluation of decisions on taxation and revenue administration; resource allocation, budgeting, and public expenditure; public borrowing/debt management; and accounting/auditing. Despite the above, operational fiscal policies among African countries have remained more fragile on the average than what is obtainable in other developing countries. Narrow tax bases, coupled with relatively low inflows from private capital, one-dimensionality of their industrial base, and underdeveloped financial markets leave African countries highly dependent on very volatile grants and foreign loans for financing government expenditure (Gollwitzer, 2010). Notably, a strong system of fiscal administration may promote fiscal discipline, sound allocation of public resources and technical efficiency of service delivery (Shah, 2007). This is why in Nigeria for instance, Osiyemi (2005) called for a provision in the Appropriation Act to make the implementation of approved budget compulsory in order to Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 24 insulate national objectives from undue political horse – trading. 2. Theoretical Framework From a normative point of view, every activity that involves budgetary decision must have a goal, explicit or implicit. Such goals are expected to provide the basis for undertaking such activities; be it the provision of new services or the expansion of existing ones. This however, is the argument of Rubin (1990) in his budget decision process theory which this study is hinged on. Rubin’s Budget Decision Process Theory (1990) According to Rubin (1990), budgeting can be construed as the means that gives viability to the activities a government undertakes to achieve a defined goal or objective. It was this ideology that gave birth to his theory, “Budget Decision Process Theory”. An important viewpoint of the Decision Process Theory (DPT) is that when any government undertakes an activity or allocates funds for it, it is often considered not in isolation but together in combination with other activities. The raison d’être of this contention is that any activity that has no appeal when considered in isolation may appear attractive when considered in combination, as a package with other activities. 3. Empirical Literature In relation to budgets and fiscal deficits, Nachega (2005) examined fiscal dominance hypothesis in the Democratic Republic (DR) of Congo. Secondary data from 1981 to 2003 were the major considerations for analytical purpose which employed multivariate co- -integration and the Vector Error Correction Model (VECM). Results indicated a strong and significant long-run relationship between budget deficits and seigniorage, and between money creation and inflation. In another study, Baharumshah, Lau, Khalid (2006) examined how fiscal and current account deficits were affected by interest and nominal exchange rates in Asia. Data from 1976 to 2000 of 4 countries (Philippines, Indonesia, Malaysia, and Thailand) were analysed by the Johansen co-integration test. Results showed that there was a symmetric long run relationship in Indonesia, Malaysia and Thailand but the reverse was the case for Philippines where the results could not find any long-run relationship. In Ethiopia, Wolde-Rufael (2008) examined the causal relationship between fiscal deficits, money growth and innovation by obtaining data from 1964–2003. Analyses were done by means of robust techniques and findings indicate the presence of a long- -run relationship between fiscal deficits and innovation. Results also show that in the short-run, fiscal deficits was not having significant effect on innovation. In Nigeria, Oladipo, Akinbobola (2011) also examined the causal link between fiscal deficits and innovation in Nigeria for the period 1970–2005. Several regression techniques/tests were conducted. Evidence from the results indicates a co-integrating long-run relationship between the macroeconomic variables of interest to the extent that gross domestic product and exchange rate had causal effect on innovation, with causation from GDP to innovation; and exchange rate to innovation. There was also evidence of unidirectional causality from fiscal deficits to innovation, thus indicating the absence of a mix of fiscal policy instrument that would achieve sustainable growth and development. Furthermore, in another econometric analysis in Nigeria Nendee (2011) examined the linkage between fiscal policy variables and Nigeria’s economic growth (1970–2009) and used GDP as the dependent variable while Federal government expenditure, Federal Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 25 government revenue, inflation rate and capital inflow were taken as the regressors. The study adopted the arcane method of Vector auto-regression (VAR) and ECM techniques. The study found a long run equilibrium relationship between these fiscal policy variables of interest and economic growth in Nigeria. Nkalu (2015) investigated budget deficits effects on macroeconomic/fiscal administration variables by focusing on Nigeria and Ghana. Time-series data were gotten from both countries during 1970 to 2013. Variables of interest included budget deficits, interest rates, inflation, and economic growth indicators. By adopting the Two-Stage Least Squares (2SLS) alongside the Seemingly Unrelated Regression (SUR) model, secondary data sourced were analysed. Findings demonstrated that budget deficit had statistical negative effect on interest rate, inflation, and economic growth, thus supporting the neoclassical argument in the literature that budget deficit slows economic growth owing to the crowding-out of resources. Finally, Nwaeke, Korgbeelo (2016) examined how budget deficits are financed in Africa with Nigeria as a study point. Several sources of deficits’ financing and their respective impact on selected variables (economic and financial) were examined also. To achieve this aim, yearly time-series data from 1981–2013 were analysed. Overall, the study found among others that deficits financed from external loans had insignificant negative influence on growth whereas; deficits financed from domestic sources had the capacity of stimulating economic growth. 4. Methodology This study is based on the ex-post facto design. Our justification for this choice is because the design enabled us to observe and measure existing data that are both quantitative and qualitative in nature over a long period of time without any form of manipulation of data. It also helped in establishing the cause-and-effect relationship between the variables of interest in this study. The population of the study consists of the fifty (50) countries of the Sub Saharan African (SSA) region (Library of Congress, 2010). These countries are similarly referred to in this work as economies. They are all rated and classified as developing economies within the four regions of SSA (East Africa, Mid Africa, Southern Africa and West Africa). The sample size consists of twelve (12) countries of SSA. The economies sampled are Kenya, Ethiopia, Tanzania, Angola, D.R. Congo, Cameroon, South Africa, Botswana, Namibia, Nigeria, Ghana and Cote D’Ivoire. We selected the samples using Purposive/ Judgmental Sampling Technique to cover the four regions of SSA. Using the size of GDP at current $USD as at 2014, the largest 3 economies in each region (East Africa, Mid Africa, Southern Africa and West Africa) as at December, 2015, were selected and included as the sample for this study. The size of GDP which ranged between $12.9 billion (Namibia) and $568.5 billion (Nigeria) was based on available data of the African Development Bank (2015, 2017). Basically, for dependable and unbiased analysis, secondary data were sourced for 15 years (2002–2016). The data were however sourced from publications of respective ministries of finance and monetary authorities including apex banks of the selected countries. Data were also sourced from the African Statistical Year Book (a publication of the African Development Bank) for the relevant years. Several analytical tools have been deployed in this study. They include the Ordinary Least Square (OLS) regression method, correlation analysis, variance inflation factor (VIF) tests for multicollinearity, and Breusch-Pagan/Cook Weisberg test for heteroskedasticity etc. In measuring the altitude of statistical association between all variables, a 5% level of test of significance was employed. Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 26 5. Hypotheses and Specification of Models In furtherance of this study, the following hypothesis was made HO: There is no significant relationship between budget deficit and fiscal administration among countries in Sub Saharan Africa. Model Specification This study’s composite model established the link between budget performance and fiscal administration indices like revenue generated, expenditure incurred, budget variance and fiscal deficit financing of countries in SSA. This relationship is specified in its implicit form as follows: BuDeft – ƒ(GovRevt , GovExpt , Bvart , FisDefFt ) (1) where: BuDeft Budget Deficit in year t, GovRevt Total Government Revenue in year t, GovExpt Total Government Expenditure in year t, Bvart Budget Variance in year t (measured by 1 for surplus and 0 for deficit), FisDefFt Total Amount of Borrowed Fund used in financing budget deficit in year t. The above model is however restated in an explicit form below: BuDefit – β0 + β1GovRevit + β2GovExpit + + β3Bvarit + β4FisDefFit + Uit (2) where: BuDefit Budget Deficit for country i in year t, GovRevit Total Government Revenue for country i in year t, GovExpit Total Government Expenditure for country i in year t, Bvarit Budget Variance for country i in year t (measured by 1 for surplus and 0 for deficit), FisDefFit Total Amount of Borrowed Fund used in financing budget deficit for country i in year t, β0, β1– β4 Regression Coefficients, Uit Error term. Since total revenue basically comprise of tax and non-tax revenue and expenditure is the summation of capital and recurrent expenditure, Eqns (1) and (2) were further modified to produce Eqn (3) which in its explicit form is stated thus: BuDefit – β0+β1TaxRevt+β2NTaxRevt+ +β3CapExpt+β4RecExpt+β5Bvart+ +β6FisDefFt + Uit (3) where: Budefit Budget deficit of country i in year t, TaxRevit Tax revenue for country i in year t, NTaxRevit Non Tax revenue for country i in year t, CapExpit Capital Expenditure for country i in year t, RecExpit Recurrent Expenditure for country i in year t, Bvarit Budget Variance for country i in year t (measured by 1 for surplus and 0 for deficit), FisDefFit Total Amount of Borrowed Fund used in financing budget deficit for country i in year t, β0, β1– β6 Regression Coefficients, Uit Error term. 6. Analyses, Results and Discussions The results from all analyses made are highlighted in different phases and presented in sections as follows: 6.1 Results and Descriptive Statistics Panel data analysis was the basis of the regression analysis employed. Both descriptive statistics, correlation matrix and varian- Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 27 ce inflation test were conducted. Panel data regression corrected for heteroskedasticity was also conducted. Variables analysed include Budget Deficit (BuDef) measure (dependent variable) alongside components of the independent variables which included Tax revenue (TaxRev), Non Tax revenue (NTaxRev), Capital Expenditure (CapExp), Recurrent Expenditure (RecExp), Budget Variance (Bvar) and Borrowed Fund used in financing budget deficit (FisDefF). 6.2 Descriptive Statistics for Entire Panel Data This section presents results’ summary of the descriptive statistics, for the entire panel data. Table1 above reveals the mean (average) for all the variables and their degrees of dispersion. The result presented in Table 1 provides insight into the nature of the selected countries that were examined by this study. As can be observed, tax revenue recorded the highest average with a mean of 24.5678, and was followed by recurrent expenditure which recorded a mean of 22.7314, and debt/ borrowed funds (11.9659), capital expenditure (8.2922), non tax revenue (3.3765) and budget variance (0.35). The level of debt incurred by Governments to finance budget deficits recorded the highest dispersion with a standard deviation of 36.1664, while the least was budget variance (0.4783) and tax revenue (4.1469). The value of dispersion of 36.1664 for debt/borrowed funds shows that the level of reliance by governments of countries in SSA is dispersed from each other. 6.3 Correlation Analysis for Entire Panel Data The correlation analysis was conducted to show the direction and relationship among the variables of concern. Table 2 presents the results for the correlation analysis. Table 1. Descriptive statistics. Variables Mean Degree of Dispersion Min. Value Max. Value Budget Deficit (BuDef)–1.1948 12.5582 –78.81 33.08 Tax Revenue (TaxRev)24.5681 29.9445 0.95 213 Non Tax Revenue (NTaxRev)3.3765 4.1469 0 20.69 Rec. Expenditure (RecExp)8.2922 16.2803 0.08 128 Cap. Expenditure (CapExp)22.7314 30.6836 1.53 209.01 Debt/Borrowed Funds (FDefF)11.9659 36.1664 0 275.33 Budget Variance (BVar)0.3500 0.4783 0 1 Source: Authors’ computation using Stata 13.0 Software, 2017. Table 2. Correlation analysis (Entire Panel Data). Variables BuDef TaxRev NTaxRev CapExp RecExp FDefF Bvar BuDef 1.0000 TaxRev –0.6857 1.0000 NTaxRev –0.0294 0.3374 1.0000 CapExp –0.7277 0.8730 0.3442 1.0000 RecExp –0.7098 0.8046 0.2863 0.8292 1.0000 FDefF –0.7131 0.7069 0.5414 0.7784 0.7977 1.0000 Bvar 0.5005 –0.0368 0.3082 –0.0959 –0.1662 –0.0082 1.0000 Source: Authors’ computation using Stata 13.0 Software, 2017. Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 28 As revealed from the table, the dependent variable (BuDef) is positively related to budget variance (Bvar). The result also indicates that GovPerf is negatively related to government revenue (TaxRev and NTaxRev), government expenditure (CapExp and RecExp) and government debt/borrowed funds (FDefF). The correlation matrix also revealed that no two explanatory variables were perfectly correlated. 6.4 Multicollinearity Test for Entire Panel Data In studies involving panel data, where multicollinearity exist between variables (independent) arriving at wrong signs or implausible magnitudes in the estimated model coefficients is inevitable. The long run effect is making wrong conclusion from bias standard errors and coefficients. On this note, multicollinearity test for the independent variables was conducted using the Variance Inflation Factors (VIF) test (see Table 3 for the result). From the result in Table 3 above, the value of the Mean VIF of 5.91 is less than the accepted value of VIF (10), which shows that the specified models (Eq. 1a, 1b and 1c) are fit and devoid of multicollinearity problem. 6.5 Comparative Results of Sampled Countries The comparative results for the sampled SSA countries are also presented. Table 3. Result of multicollinearity test (Entire Panel Data). VIF Variables Coefficient Variance Centered VIF TaxRev 6.44 0.155315 NTaxRev 1.80 0.556345 CapExp 8.02 0.124723 RecExp 13.87 0.072107 FDefF 4.09 0.244494 Bvar 1.26 0.791172 Mean VIF 5.91 Source: Authors’ computation using Stata 13.0 Software, 2017. Table 4. Government Revenue (TaxRev and NTaxRev) and Budget Performance (BuDef). Country Tax Revenue (TaxRev) Non-Tax Revenue (NTaxRev) Coefficient Std. Error T-stat. Prob Coefficient Std. Error T-stat. Prob Kenya 1.005701 0.1808706 5.56 0.001 1.0481250 0.2350708 4.46 0.002 Ethiopia 0.099720 0.0041232 242.46 0.000 1.0009430 0.0026378 379.47 0.000 Tanzania –0.0062791 0.1693154 0.971 0.494 –2.4038980 3.369888 –0.71 0.494 Angola 1.0118640 0.0137762 73.45 0.000 0.8991780 0.0976766 9.21 0.000 DR. Congo 0.1096335 0.2516469 0.44 0.675 –1.0393890 1.052176 –0.99 0.352 Cameroon –0.0736513 0.0776337 –0.95 0.371 0.9278371 0.9069982 1.02 0.336 South Africa 1.59996700 0.2933873 5.45 0.002 0.7976875 1.223511 0.65 0.539 Botswana 0.76450490 0.2993235 2.55 0.034 0.9545712 0.4945752 1.93 0.090 Namibia 0.88256636 0.3310790 2.67 0.029 0.5393500 1.013558 0.53 0.609 Nigeria 0.95176860 0.6724450 14.15 0.000 5.8767160 1.625357 3.62 0.009 Ghana 0.87201770 0.0052402 166.41 0.000 1.7359630 0.0080392 215.94 0.000 Cote D’Ivoire 0.16073960 0.8587370 1.87 0.098 0.8082574 0.7443051 1.09 0.309 Source: Authors’ computation using Stata 13.0 Software, 2017. Emenike Edwin Ekenechukwu Ezinando et al.: Budget Deficit and Fiscal Administration in Selected Sub-Saharan African Countries 29 Table 4 reports the comparative analysis of the results of sampled countries. It is clear from the table above that the coefficients of all the sampled countries except Tanzania (–0.0062791) and Cameroun (–0.0736513) are carrying positive signs for tax revenue. The coefficients of non-tax revenue for Tanzania (–2.403898) and DR. Congo (–1.039389) are also carrying negative signs. The negative sign in the coefficient for Tanzania and Cameroon (Tax Revenue) is an indication that government revenue negatively influenced budget performance for the period. However, government revenue has some insightful revelations; first, tax revenue (TaxRev) is stastistically significant for countries like Kenya, Ethiopia, Angola, South Africa, Botswana, Namibia, Nigeria, Ghana and Cote D’Ivoire except that for countries like Tanzania, DR. Congo and Cameroon, the results were statistically flawed (see p- -values for tax revenue). Secondly, non-tax revenue (NTaxRev) was found to be statistically significant for some of the countries (Kenya, Ethiopia, Angola, Botswana, Nigeria and Ghana). Although, for countries like Tanzania, DR. Congo, Cameroon, South Africa, Namibia and Cote D’Ivoire the results were statistically flawed (see p-values for non-tax revenue). On the whole, we found that budget performance was positively influenced by government revenue for most of the countries in SSA. Table 5 reports the comparative analysis of the results of sampled countries with respect to budget variance, fiscal deficit financing and their relationship with budget performance. It is clear from the table above that apart from Kenya, Angola, South Africa and Nigeria, the coefficients of most of the sampled countries carried positive signs for budget variance. The coefficients of fiscal deficit financing for Tanzania (5.088799), DR. Congo (2.29913), South Africa (0.1048067) and Cote D’ Ivoire (2.851573) are also carrying positive signs. These positive signs indicate that government budget performance had positive relationship budget variance and fiscal deficit financing where applicable. However, for countries like Kenya, Angola, South Africa and Nigeria, we observe that budget variance had a negative relationship with budget performance. Similarly, fiscal Table 5. Relationship between Budget Variance (Bvar), Fiscal Deficit Financing (FisDefF) and Budget Performance (BuDef). Country Budget Variance Fiscal Deficit Financing Coefficient Std. Error T-stat. Prob Coefficient Std. Error T-stat. Prob Kenya –0.1503788 0.5004901 –0.30 0.771 –1.222106 2.439577 –0.50 0.630 Ethiopia 0.000 0.0000 0.000 N/A –0.0018934 0.0035775 –0.53 0.611 Tanzania 0.000 0.0000 0.000 N/A 5.088799 7.919222 0.64 0.537 Angola –0.897006 0.198481 –0.45 0.665 –0.0821954 0.0545451 –1.51 0.176 DR. Congo 6.374883 0.8587954 7.42 0.000 2.29913 2.178947 1.06 0.322 Cameroon 8.740541 7.052753 1.24 0.250 –0.1302394 0.0751636 –1.73 0.121 South Africa –0.4208329 0.7350332 –0.57 0.588 0.1048067 0.5709052 0.18 0.860 Botswana 3.38071 2.996075 1.13 0.292 –1.183426 0.7109324 –1.66 0.135 Namibia 1.299217 1.748617 0.74 0.479 –0.0617102 0.6490467 –0.10 0.927 Nigeria –0.8726716 0.8122976 –1.07 0.318 –14.87401 2.083834 –7.14 0.000 Ghana 0.000 0.0000 0.000 N/A –0.1107976 0.0028459 –38.93 0.000 Cote D’Ivoire 2.362154 0.701214 –1.57 0.155 2.851573 7.041296 0.40 0.696 Source: Authors’ computation using Stata 13.0 Software, 2017.