Sage Journals: Discover world-class research

Abstract

The study examines nexuses between carbon dioxide (CO₂) emissions, renewable energy consumption and inequality in 39 Sub-Saharan African countries for the period 2004–2014. The empirical evidence is based on Quantile regressions. First, in the 25th quantile of the inequality distributions, as long as CO₂ emissions metric tons per capita are kept below 4.700 (4.100), the Gini coefficient (Atkinson index) will not increase. These are avoidable CO₂ emissions thresholds. Second, renewable energy consumption should be complemented with other policies to: (i) reduce the Gini coefficient when renewable energy consumption is at 50.00% of total final energy consumption and (ii) mitigate the Atkinson index when renewable energy consumption is at 62.500% of total final energy consumption in the bottom quantiles of the Atkinson index distribution and at 50.00% of total final energy consumption in the 75th quantile of the Atkinson index distribution. These are renewable energy consumption thresholds for complementary policies. The novelty of this study in the light of extant literature is fundamentally premised on providing policy makers with avoidable thresholds of CO₂ emissions as well as corresponding thresholds of renewable energy consumption for complementary policies, in the nexus between the green economy and inequality.

Keywords

Renewable energy inequality Finance Sub-Saharan Africa sustainable development

Introduction

There is a growing importance of fossil fuels and renewable energy consumption surrounding policy and scholarly issues in the sustainable development agenda of most countries (Ivanovski et al., 2020; Wang et al., 2019; Yang et al., 2019). This study investigates nexuses between environmental quality and income inequality in Sub-Saharan Africa (SSA) and provides actionable policy thresholds underlying the nexuses that can be leveraged by policy makers to promote inclusive and sustainable development in the sub-region in the post-2015 development agenda. Interactions between policy variables and policy syndromes in the economic development literature can engender cut-off points or thresholds which represent actionable critical masses that policymakers can leverage upon to influence the outcome variable in one direction or the other.¹ This can be the case in the nexus between environmental quality and income inequality in a sub-region where concerns about income inequality and environmental degradation are comparatively more concerning.² The sub-region is SSA and the two underlying concerns are critically engaged in what follows in order to clearly articulate the problem statement within the context of extant literature as well as the attendant policy and scholarly concerns pertaining to the issues.

First, on the front of income distribution and by extension income inequality, it is worthwhile to emphasize that the over two decades of economic growth resurgence in SSA has not moved hand-in-glove with inclusive development in terms of equal distribution of the fruits of economic prosperity across the population. Contemporary studies supporting this perspective include Meniago and Asongu (2018), Tchamyou et al. (2019a) and Asongu and Odhiambo (2019a, 2020a). The attendant literature also maintains that close to half of countries in SSA did not reach the millennium development goal (MDG) extreme poverty target because the fruits of the over two decades of economic growth resurgence did not trickle down to the poor fractions of societies. Moreover, current projections posit that unless the glaring concern of income inequality is addressed by means of more inclusive growth, the sustainable development goal (SDG) target of mitigating extreme poverty to a threshold of below 3% would not be achieved (Bicaba et al., 2017). It follows from this narrative that inequality which is the outcome variable in this study remains a critical concern for both scholars and policy makers in the sub-region. The underlying concern can be compounded or mitigated by environmental degradation or the green economy as established by the attendant literature on the nexus between inequality and environmental quality (Demir et al., 2019; Grunewald et al., 2017; Liu et al., 2019; Uzar and Eyuboglu, 2019).

Second, concerns surrounding environmental degradation are comparatively more apparent in SSA because despite being the continent that emits the least carbon dioxide (CO₂) emissions, it is projected that the consequences of global warming would be most detrimental in the sub-region compared to other continents and regions of the world (Asongu et al., 2017, 2018). Moreover, the post-2015 policy and scholarly literature is consistent on the need to promote renewable consumption of energy in order to limit the potential unfavorable consequences of environmental degradation on economic and human developments in the sub-region (Asongu and Odhiambo, 2020b, 2020c; Bekun et al., 2019; Nathaniel and Bekun, 2020; Niranjan, 2019; O’Ryan et al., 2020; Samour et al., 2019). In the light of these underlying policy and scholarly concerns, the focus of this study is to assess how environmental degradation (in terms of CO₂ emissions) and renewable energy consumption affect inequality: a positioning that is both consistent with policy requirements and a gap in the literature.

Third, the literature on the nexus between income equality and environmental degradation has largely focused on how income inequality affects environmental outcomes. However, this study focuses on how environmental quality affects inequality outcomes. Some of the extant studies focusing on the underlying opposite nexus include, inter alia: Grunewald et al. (2017) which has examined the trade-off between CO₂ emissions and income inequality to establish that higher income inequality in lower and middle-income countries is associated with lower CO₂ emissions whereas in high-income and upper-middle income countries, per capita emissions is increased by higher levels of income inequality. Liu et al. (2019) have established that inequality in the distribution of income, especially within the remit of spatial income distribution, increases CO₂ emissions. Uzar and Eyuboglu (2019) have shown that CO₂ emissions are positively affected by income inequality. Conversely, Demir et al. (2019) conclude that income inequality mitigates environmental degradation.

A common denominator of the above studies is that the attendant research focuses on the effect of inequality on environmental degradation. However, this study is positioned on the reverse nexus (i.e. the incidences of environmental pollution and renewable energy consumption on income inequality variables). In other words, instead of focusing on how inequality affects the green economy, the present study departs from the extant studies by focusing on how dynamics of the green economy (i.e. CO₂ emissions and renewable energy consumption) affect income inequality throughout the conditional distribution of income inequality. Moreover, beyond establishing nexuses among the macroeconomic variables, the present study also departs from the underlying strand of research by providing actionable critical masses or thresholds that policy makers can leverage upon in order to reduce income inequality, in the light of challenges of inclusive development in the post-2015 agenda. Hence, the policy-relevant thresholds this study aims to provide in view of informing policy makers on environmental quality measures that can contribute towards mitigating income inequality are: (i) avoidable CO₂ emissions thresholds and (ii) renewable energy consumption thresholds for complementary policies. The novelty of this study in the light of extant literature is fundamentally premised on providing policy makers with avoidable thresholds of CO₂ emissions as well as corresponding thresholds of renewable energy consumption for complementary policies, in the nexus between the green economy and inequality.

The theoretical underpinnings motivating this study are broadly consistent with those surrounding the “Environmental Kuznets Curve” (EKC), the Kuznets hypothesis (Grossman and Krueger, 1995; Kuznets, 1955; Selden and Song, 1994) and “Carbon Kuznets Curve” (CKC) (Xu and Song, 2011) because, for the most part, the adopted outcome variable in the study is income inequality, whereas renewable energy consumption is an independent variable of interest. Hence, a nexus being considered, inter alia, is how enhancing renewable energy consumption affects inequality. The remainder of this study is structured in the following manner. This introduction is followed by a data and methodology section before another section on empirical results. The last section concludes with implications and future research directions.

Data and methodology

Data

To examine the problem statement discussed in the previous section, the present study uses a panel of 39 countries in SSA with data from 2004 to 2014.³ The selection of countries and corresponding periodicity is motivated by constraints in the availability of the relevant information at the time of the study. The data is obtained from four principal sources, namely: (i) the World Development Indicators (WDI) of the World Bank; (ii) the World Bank’s Financial Development and Structure Database (FDSD); (iii) the Global Consumption and Income Project (GCIP) and (iv) the World Bank’s World Governance Indicators (WGI).

The outcome inequality variables are sourced from the GCIP, namely: the Gini coefficient and the Atkinson index. The Gini coefficient represents the distribution of income across the population while the Atkinson index denotes the percentage of total income that a specific society is prepared to sacrifice in order to improve the distribution of income in the attendant society. The choice of the Atkinson index to complement the Gini coefficient is motivated by the fact that the latter has been documented not to emphasize the tails or extreme points of the inequality distribution (Meniago and Asongu, 2018; Naceur and Zhang, 2016; Tchamyou et al., 2019a, 2019b). Hence, it is for the purpose of robustness that the Atkinson index is used to complement the Gini coefficient.

The independent variables of interest are obtained from WDI of the World Bank, notably: (i) CO₂emissions (metric tons per capita) and (ii) renewable energy consumption (% of total final energy consumption).⁴ The latter is a positive environmental signal while the former is a negative environmental signal. Both are used for the purpose of robustness. The choice of the environmental quality indicators is also informed by contemporary literature on the subject (Akinyemi et al., 2019; Asongu, 2018; Asongu et al., 2019a; Asongu and Odhiambo, 2020c; Nathaniel and Iheonu, 2019).

In order to control for variable omission bias, five non-dummy and two dummy variables are used in the estimation exercise. The non-dummy variables are: financial access, mobile phones, regulation quality, trade openness and urban population while the dummy variables are middle-income countries and petroleum-exporting countries. The variables which are from WDI, FDSD and WGI of the Bank and Asongu et al. (2019b), as apparent in Appendix 1, are informed by contemporary inclusive development literature (Asongu and Asongu, 2018; Asongu and Nwachukwu, 2016; Asongu and Odhiambo, 2018, 2019b; Demirgüc-Kunt and Klapper, 2012; Demirgüç-Kunt et al., 2015; Murendo et al., 2018; Tchamyou, 2019a, 2019b, 2020). With the exception of urban population and the middle-income countries dummy that are expected to increase income inequality, the other variables are anticipated to have the opposite effect. However, it is important to note that the expected signs can also be influenced by two more dynamics. On the one hand, given that quadratic terms are involved in the regression exercise, the quadratic terms can be correlated with the independent variables of interest and hence, some expected signs of variables in the conditioning information set can be affected. However, the concern of multicollinearity does not affect the main independent variables of interest because according to Brambor et al. (2006), quadratic regressions are not interpreted as linear additive models. For this purpose, net effects and thresholds underlying the quadratic specifications are computed in the empirical results section.

On the other hand, since the nexuses are assessed throughout the conditional distribution of the inequality variables, the attendant effects can vary depending on initial levels of income inequality. It follows from the above narrative that this study focuses on the quadratic estimated variables to compute the attendant thresholds and net effects. The study is also only concerned about the significance of estimated coefficients of variables in the conditioning information set because the study makes abstraction of the expected signs in the light of the concerns discussed above. The definitions of variables and sources of data are disclosed in Appendix 1, the corresponding summary statistics and correlation matrix are provided respectively in Appendix 2 and Appendix 3.

Methodology

In accordance with the narrative in the previous section, the quantile regression approach is adopted for this study because it enables us to assess the investigated nexuses through the conditional distribution of the inequality outcome variables. Hence, the adopted estimation approach is such that, low, intermediate and high initial levels of income inequality are articulated. Accordingly, by accounting for existing levels of income inequality, there is an assumption that a modeling approach based on mean values of the outcome variable is unlikely to produce robust and policy-relevant estimates, unless the initial levels of income inequality are taken on board in order to avoid blanket policy implications.

Following the attendant literature on the subject matter (i.e. contemporary and non-contemporary sources), the quantile estimates are obtained by solving the optimization problem presented in equation (1) below without subscripts to enhance readability and flow (Asongu and Odhiambo, 2019c; Hao and Naiman, 2007; Koenker, 2005; Koenker and Bassett, 1978; Okada and Samreth, 2012; Tchamyou and Asongu, 2017):

\min_{β \in R^{k}} [\sum_{i \in {i : y i \geq x i^{'} β}} θ | y_{i} - x_{i^{'}} β | + \sum_{i \in {i : y i < x i^{'} β}} (1 - θ) | y_{i} - x_{i^{'}} β |]

(1)

where

θ \in (0, 1)

. Contrary to Ordinary Least Squares (OLS) which is founded on minimizing the total sum of squared residuals, the Quantile regression is focused on minimizing the weighted total of absolute variations. For example, the 10th or 90th quantiles (with

θ

= 0.10 or 0.90 respectively) are estimated by weighing the residuals approximately. The conditional income inequality or

y_{i}

given

x_{i}

is:

Q_{y} (θ / x_{i}) = x_{i^{'}} β θ

(2)

where parameters that are characterized by a single slope are estimated for each the θth specific quantile. Equation (2) is analogous to

E (y / x) = x_{i^{'}} β

in the OLS slope where parameters to be estimated rely on the conditional distribution of income inequality. It is important to articulate that in the underlying equation,

y_{i}

represents income inequality (i.e. the Gini coefficient and the Atkinson index) while

x i

contains: a constant term, CO₂ emissions, renewable energy consumption, financial access, mobile phone penetration, regulation quality, trade openness, urban population, middle-income countries and petroleum-exporting nations.

Empirical results

The empirical results are provided in this section in Tables 1 and 2. While Table 1 presents linkages between the Gini coefficient, CO₂ emissions and renewable energy consumption, Table 2 discloses the corresponding findings for the Atkinson index, CO₂ emissions and renewable energy consumption. Each of the tables is divided into two main categories of quantile specifications: one corresponding to CO₂ emissions and the other to renewable energy consumption. It is important to note that from the left-hand side to the right-hand side of both tables, the inequality levels of increase. It follows that in the 10th (90th) quantile, the level of inequality is lowest (highest).

Table 1.

The Gini Coefficient, CO₂ emissions and renewable energy consumption.

	Dependent variable: The Gini Coefficient
	CO ₂ emissions (CO₂)					Renewable Energy Consumption (REC)
	Q.10	Q.25	Q.50	Q.75	Q.90	Q.10	Q.25	Q.50	Q.75	Q.90
Constant	0.565***(0.000)	0.582***(0.000)	0.586***(0.000)	0.594***(0.000)	0.607***(0.000)	0.611***(0.000)	0.612***(0.000)	0.623***(0.000)	0.618***(0.000)	0.614***(0.000)
CO₂ emissions(CO₂)	–0.004(0.821)	–0.047***(0.000)	–0.013(0.168)	0.005(0.656)	0.015**(0.011)	---	---	---	---	---
REC	---	---	---	---	---	–0.002***(0.000)	–0.002***(0.000)	–0.001***(0.000)	–0.001**(0.033)	–0.0002(0.858)
CO₂ × CO₂	0.001(0.463)	0.005***(0.000)	0.002**(0.012)	0.001(0.230)	0.0005(0.339)	---	---	---	---	---
REC × REC	---	---	---	---	---	0.00002 ***(0.000)	0.00002 ***(0.000)	0.00001 ***(0.000)	0.00001 **(0.031)	4.73e-06(0.702)
Financial Access	–0.00004(0.741)	–0.00008*(0.075)	–0.0001**(0.047)	–0.0001**(0.049)	–0.0001***(0.001)	–0.0003 ***(0.004)	–0.0003 ***(0.000)	–0.0003 ***(0.000)	–0.0002*(0.060)	–0.0001(0.671)
Mobile Phone	–0.0001(0.141)	–0.0001***(0.004)	–0.0001**(0.027)	–0.00003(0.646)	0.00002(0.500)	–0.00004(0.658)	–0.00005(0.260)	–0.0001**(0.025)	–0.00002(0.729)	–0.00002(0.902)
Regulation Quality	–0.003(0.723)	0.005(0.130)	0.004(0.375)	0.004(0.515)	0.012(0.000)	0.016**(0.014)	0.018***(0.000)	0.024***(0.000)	0.033***(0.000)	0.036(0.022)
Trade Openness	–0.0001(0.112)	–0.0001***(0.000)	–0.0001**(0.032)	0.00002(0.750)	–0.00008**(0.013)	–0.00002(0.737)	2.08e-06(0.966)	–0.00004(0.446)	–0.00005(0.501)	–0.00006(0.756)
Urban Population	0.001**(0.013)	0.001***(0.000)	0.0008***(0.000)	0.0001(0.619)	0.0003**(0.010)	0.001***(0.000)	0.001***(0.000)	0.001***(0.000)	0.001***(0.000)	0.0009(0.208)
Middle Income	0.007(0.500)	0.009***(0.008)	0.002(0.560)	–0.005(0.314)	–0.014***(0.000)	0.011*(0.060)	0.015***(0.000)	0.010***(0.006)	0.005(0.285)	0.007(0.598)
Petroleum Exporting	–0.044***(0.000)	–0.010***(0.004)	–0.013***(0.007)	–0.009(0.115)	–0.005**(0.056)	–0.036***(0.000)	–0.019***(0.000)	–0.012***(0.003)	–0.005(0.328)	–0.007(0.662)
Net Effects	na	–0.037	na	na	na	0.0006	0.0006	0.0003	0.0003	na
Thresholds	na	4.700	na	na	na	50.000	50.000	50.000	50.000	na
Pseudo R²	0.304	0.312	0.295	0.385	0.623	0.331	0.265	0.202	0.213	0.360
Observations	246	246	246	246	246	245	245	245	245	245

*,**,***: significance levels of 10%, 5% and 1% respectively. Lower quantiles (e.g., Q 0.1) signify nations where inequality is least. na: not applicable because at least one estimated coefficient needed for the computation of thresholds and net effects is not significant. The mean value of CO₂ emissions (metric tons per capita) is 0.934. The mean value of Renewable Energy Consumption is 66.216.

Table 2.

The Atkinson index, CO₂ emissions and renewable energy consumption.

	Dependent variable: The Atkinson Index
	CO ₂ emissions (CO₂)					Renewable Energy Consumption (REC)
	Q.10	Q.25	Q.50	Q.75	Q.90	Q.10	Q.25	Q.50	Q.75	Q.90
Constant	0.640***(0.000)	0.672***(0.000)	0.694***(0.000)	0.693***(0.000)	0.741***(0.000)	0.738***(0.000)	0.750***(0.000)	0.671***(0.000)	0.757***(0.000)	0.743***(0.000)
CO₂ emissions(CO₂)	–0.040(0.399)	–0.041***(0.003)	0.015(0.501)	0.031**(0.040)	0.031**(0.018)	---	---	---	---	---
REC	---	---	---	---	---	–0.005***(0.001)	–0.005***(0.000)	–0.001(0.101)	–0.002***(0.000)	–0.0007(0.257)
CO₂ × CO₂	0.005(0.268)	0.005***(0.000)	–0.00003(0.989)	–0.001(0.310)	–0.002(0.104)	---	---	---	---	---
REC × REC	---	---	---	---	---	0.00004 ***(0.000)	0.00004 ***(0.000)	0.00001 **(0.018)	0.00002 ***(0.000)	9.78e-06*(0.095)
Financial Access	–0.0001(0.607)	–0.0002 ***(0.006)	–0.0003**(0.020)	–0.0003***(0.000)	–0.0004***(0.000)	–0.0008**(0.015)	–0.0008 ***(0.000)	–0.0002(0.129)	–0.0005 ***(0.000)	–0.0003**(0.025)
Mobile Phone	–0.0004(0.144)	–0.0003 ***(0.000)	–0.0002*(0.068)	0.00004(0.620)	0.0001(0.198)	–0.0001(0.462)	–0.0001(0.257)	0.00001(0.912)	0.00006(0.534)	–0.00003(0.711)
Regulation Quality	–0.003(0.881)	–0.006(0.379)	–0.006(0.587)	–0.012(0.139)	0.016**(0.017)	0.036**(0.030)	0.029***(0.000)	0.013(0.130)	0.050***(0.000)	0.051***(0.000)
Trade Openness	–0.0004(0.118)	–0.0002 ***(0.007)	–0.0001(0.370)	0.0003***(0.000)	0.0001*(0.063)	–0.00001(0.947)	0.00009(0.307)	0.00008(0.476)	0.0002 ***(0.008)	0.0001(0.156)
Urban Population	0.004***(0.000)	0.001***(0.000)	0.0006(0.183)	–0.0008***(0.008)	–0.0004(0.114)	0.004***(0.000)	0.002***(0.000)	0.001***(0.001)	0.0006*(0.057)	0.001***(0.000)
Middle Income	0.0001(0.994)	0.013*(0.051)	–0.001(0.865)	–0.010(0.181)	–0.016**(0.015)	0.009(0.545)	0.029***(0.000)	0.024***(0.004)	0.012*(0.072)	0.007(0.260)
Petroleum Exporting	–0.074***(0.002)	–0.012*(0.079)	0.004(0.716)	0.006(0.423)	0.003(0.607)	–0.060***(0.001)	–0.020***(0.005)	–0.001(0.843)	0.007(0.324)	0.016**(0.044)
Net Effects	na	–0.0316	na	na	na	0.0003	0.0003	na	0.0006	na
Thresholds	na	4.100	na	na	na	62.500	62.500	na	50.000	na
Pseudo R²	0.229	0.257	0.194	0.263	0.397	0.272	0.246	0.188	0.206	0.333
Observations	246	246	246	246	246	245	245	245	245	245

The following findings can be established from Table 1. First, CO₂ emissions unconditionally (conditionally) decrease (increase) inequality in the 25th quantile of the conditional distribution of the Gini coefficient. The corresponding overall net effect of growing CO₂ emissions on the Gini coefficient is −0.037 while the threshold at which the negative unconditional effect changes to positive is 4.700.

It is worthwhile to clarify that net effects and attendant thresholds are computed because, as documented by Brambor et al. (2006), in order to avoid the pitfalls of quadratic specifications, corresponding estimates from the independent variables of interest should not be interpreted as in linear additive models. The procedure for understanding and interpreting quadratic specifications is consistent with contemporary literature based on quadratic regressions (Asongu and Odhiambo, 2020d, 2020e). Hence, this study clarifies the computation of net effects and corresponding thresholds used to assess the overall effects and tendencies of enhancing environmental quality dynamics on income inequality. From the third column of Table 1, the net effect of enhancing CO₂ emissions on the Gini coefficient is −0.037 (2×[0.934 × 0.005] + [−0.047]). In the computation of the underlying net effect, it is important to note that −0.047 is the unconditional effect of CO₂ emissions on the Gini coefficient, 0.934 is mean value of CO₂ emissions per capita, 0.005 is the marginal impact of CO₂ emissions per capita on the Gini coefficient whereas the leading 2 term is from the quadratic derivation. The CO₂ emission threshold at which the overall effect of CO₂ emissions on inequality changes from negative to positive is 4.700 =[−0.047/(2 × 0.005)]. It follows that at a critical mass of 4.700 CO₂ emissions per capita, the overall net effect is 0 (2×[4.700 × 0.005] + [−0.047]). Hence, above the 4.700 threshold of CO₂ emission per capita, the net effect on inequality becomes positive. It follows that in the 25th quantile of the inequality distribution, as long as CO₂ emissions are kept below the 4.7 per capita critical limits, inequality would decrease.The same analytical scope is employed to understand estimation techniques on the right-hand side of Table 1 as well as corresponding estimations on both the right-hand and left-hand sides of Table 2.

The following findings are apparent on the right-hand side of Table 1. With the exception of the highest quantile (i.e. 90th quantile), the estimations are significant throughout the conditional distribution of the Gini coefficient. Second, renewable energy consumption unconditionally (conditionally) decreases (increases) the Gini coefficient and in order to maintain the dampening role of renewable energy consumption on inequality, renewable energy consumption should be complemented with other policies to reduce inequality when renewable energy consumption is at 50.00% of total final energy consumption. This threshold of renewable energy consumption for complementary policies is the same throughout the conditional distribution of the Gini coefficient.

The following findings can be established from Table 2. First, CO₂ emissions unconditionally (conditionally) decrease (increase) the inequality in the 25th quantile of the conditional distribution of the Atkinson index. The corresponding overall net incidence of growing CO₂ emissions on the Atkinson index is −0.0316 while the threshold at which the negative unconditional effect changes to positive is 4.100. Hence, above the 4.100 threshold of CO₂ emission per capita, the net effect on inequality becomes positive. It follows that in the 25th quantile of the inequality distribution, as long as CO₂ emissions are kept below the 4.100 per capita critical limits, inequality would decrease.

Second, with the exception of the highest quantile (i.e. 90th quantile) and the median (i.e. 50th quantile), the estimations are significant throughout the conditional distribution of the Atkinson index. Renewable energy consumption unconditionally (conditionally) decreases (increases) the Atkinson and in order to maintain the dampening role of renewable energy consumption on inequality renewable energy consumption should be complemented with other policies to reduce inequality when renewable energy consumption is at 62.500% of total final energy consumption in the bottom quantiles of the Atkinson index distribution and at 50.00% of total final energy consumption in the 75th quantile of the Atkinson index distribution.

The presented findings are novel in the light of extant literature because policy makers have been provided with avoidable thresholds of CO₂ emissions as well as corresponding thresholds for complementary policies in renewable energy consumption, in the nexus between the green economy and inequality. It follows that the non-linear tendency of the findings support the perspective that contrary to the attendant literature discussed in the introduction, the present study provides avenues along which policy shifts can be formulated and implemented in view of mitigating income inequality. This corresponding literature include, inter alia: Li et al. (2019) on the perilous role of gender inequality in the consumption of green households; Liu et al. (2019) and Uzar and Eyuboglu (2019) on the relationship between CO₂ emissions and income inequality; the conclusions of Demir, Cergibozan and Gok (2019) on the linkage between environmental degradation and income inequality as well as the plethora of studies on the nexus between the distribution of income and renewable energy consumption (Bai et al., 2020; McGee and Greiner, 2019; Uzar, 2020).

In the light of the above, it follows that, as opposed to the underlying studies on the inequality-“green economy” nexus which for the most part provide blanket policy implications based on mean values of the outcome variable, the dynamic nature of the present findings can be articulated on two fronts. On the one hand, the fact that the inequality variables are assessed throughout their conditional distributions in order to articulate initial levels of income inequality portrays aspects of non-linearity. On the other hand, thresholds of the independent variables of interest also reflect some non-monotonic features given that quadratic specifications have been adopted for the study.

Concluding implications and future research directions

The study has examined nexuses between CO₂ emissions, renewable energy consumption and inequality dynamics in terms of the Gini coefficient and the Atkinson index. It has focused on 39 countries in Sub-Saharan Africa using data from 2004–2014. Quantile regressions are adopted as estimation strategy in order to articulate conditional distributions of the inequality dynamics. The following findings are established. First, CO₂ emissions unconditionally (conditionally) decrease (increase) inequality at the 25th quantile of the conditional distributions of the Gini coefficient and the Atkinson index. The corresponding overall net effect of growing CO₂ emissions on the Gini coefficient (Atkinson index) is −0.037 (−0.0316) while the threshold at which the negative unconditional effect changes to positive is 4.700 (4.100). Hence, in the 25th quantile of the inequality distributions, as long as CO₂ emissions metric tons per capita are kept below 4.700 (4.100), the Gini coefficient (Atkinson index) will not increase.

Second, with the exception (s) of the highest quantile (median and 90th quantile) of the Gini coefficient (Atkinson index), renewable energy consumption unconditionally (conditionally) decreases (increases) income inequality and in order to maintain the dampening role of renewable energy consumption on inequality, renewable energy consumption should be complemented with other policies to: (i) reduce Gini coefficient when renewable energy consumption is at 50.00% of total final energy consumption and (ii) mitigate the Atkinson index when renewable energy consumption is at 62.500% of total final energy consumption in the bottom quantiles of the Atkinson index distribution and at 50.00% of total final energy consumption in the 75th quantile of the Atkinson index distribution.

It follows from the above findings that avoidable CO₂ emission thresholds have been provided as well as renewable energy consumption thresholds for complementary policies. The thresholds make economic sense and have policy relevance because they are within their respective statistical ranges disclosed in the summary statistics. Hence, the policy implications of this study are straight forward and directly apparent. The novelty of this study in the light of extant literature is fundamentally premised on providing policy makers with avoidable thresholds of CO₂ emissions as well as corresponding thresholds of renewable energy consumption for complementary policies, in the nexus between the green economy and inequality.

Future studies can examine how the established findings are relevant to country-specific frameworks in order to provide findings with country-specific policy options. It would also be worthwhile to assess complementarities between renewable energy consumption and other policy variables that can mitigate income inequality, especially at established thresholds for complementary policies.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Simplice A Asongu

Nicholas M Odhiambo

Notes

Appendix 1. Definitions of variables

Appendix 2. Summary statistics (2004–2014)

Appendix 3. Correlation matrix (uniform sample: 245)

References

Akinyemi

Uchenna

Asongu

S, et al.

(2019) Renewable energy, trade performance and the conditional role of finance and institutional capacity in Sub-Sahara African countries. Energy Policy 132(September): 490–498.

Asongu

(2018) ICT, openness and CO₂ emissions in Africa. Environmental Science and Pollution Research International 25(10): 9351–9359.

Asongu

(2018) The comparative exploration of mobile money services in inclusive development. International Journal of Social Economics 45(1): 124–139.

Asongu

Iheonu

Odo

(2019a) The conditional relationship between renewable energy and environmental quality in Sub-Saharan Africa. Environmental Science and Pollution Research International 26(36): 36993–37000.

Asongu

Le Roux

Biekpe

(2017) Environmental degradation, ICT and inclusive development in Sub-Saharan Africa. Energy Policy 111(December): 353–361.

Asongu

Le Roux

Biekpe

(2018) Enhancing ICT for environmental sustainability in Sub-Saharan Africa. Technological Forecasting and Social Change 127(February): 209–216.

Asongu

Nwachukwu

(2016) The role of governance in mobile phones for inclusive human development in Sub-Saharan Africa. Technovation 55–56(September–October): 1–13.

Asongu

Nwachukwu

Pyke

(2019b) The comparative economics of ICT, environmental degradation and inclusive human development in Sub-Saharan Africa. Social Indicators Research 143(3): 1271–1297.

Asongu

Odhiambo

(2018) Human development thresholds for inclusive mobile banking in developing countries. African Journal of Science, Technology, Innovation and Development 10(6): 735–744.

10.

Asongu

Odhiambo

(2019a) Inclusive development in environmental sustainability in Sub‐Saharan Africa: Insights from governance mechanisms. Sustainable Development 27(4): 713–724.

11.

Asongu

Odhiambo

(2019b) How enhancing information and communication technology has affected inequality in Africa for sustainable development: An empirical investigation. Sustainable Development 27(4): 647–656.

12.

Asongu

Odhiambo

(2019c) Governance and social media in African countries: An empirical investigation. Telecommunications Policy 43(5): 411–425.

13.

Asongu

Odhiambo

(2020a) Economic development thresholds for a green economy in Sub-Saharan Africa. Energy Exploration & Exploitation 38(1): 3–17.

14.

Asongu

Odhiambo

(2020b) Governance, CO₂ emissions and inclusive human development in Sub-Saharan Africa. Energy Exploration & Exploitation 38(1): 18–36.

15.

Asongu

Odhiambo

(2020c) Enhancing governance for environmental sustainability in Sub-Saharan Africa. Energy Exploration & Exploitation.

16.

Asongu

Odhiambo

(2020d) Insurance policy thresholds for economic growth in Africa. The European Journal of Development Research 32(3): 672–689.

17.

Asongu

Odhiambo

(2020e) How enhancing gender inclusion affects inequality: Thresholds of complementary policies for sustainable development. Sustainable Development 28(1): 132–142.

18.

Bai

Feng

Yan

, et al. (2020) Will income inequality influence the abatement effect of renewable energy technological innovation on carbon dioxide emissions? Journal of Environmental Management 264(June): 110482.

19.

Bekun

Emir

Sarkodie

(2019) Another look at the relationship between energy consumption, carbon dioxide emissions, and economic growth in South Africa. The Science of the Total Environment 655: 759–765.

20.

Bicaba

Brixiova

Ncube

(2017) Can extreme poverty in Sub-Saharan Africa be eliminated by 2030? Journal of African Development 19(2): 93–110.

21.

Brambor

Clark

Golder

(2006) Understanding interaction models: Improving empirical analyses. Political Analysis 14(1): 63–82.

22.

Demir

Cergibozan

Gok

(2019) Income inequality and CO2 emissions: Empirical evidence from Turkey. Energy & Environment 30(3): 444–461.

23.

Demirgüc-Kunt

Klapper

(2012) Measuring financial inclusion: The Global Findex database. World Bank Policy Research Working Papers No. WPS 6025, Washington.

24.

Demirgüç-Kunt

Klapper

Van Oudheusden

(2015) The Global Findex database2014. World Bank Policy Research Working Paper No. 7255, Washington, DC.

25.

Grossman

Krueger

(1995) Economic growth and the environment. The Quarterly Journal of Economics 110(2): 353–377.

26.

Grunewald

Klasen

Martínez-Zarzoso

, et al. (2017) The trade-off between income inequality and carbon dioxide emissions. Ecological Economics 142(December): 249–256.

27.

Hao

Naiman

(2007) Quantile Regression. London, UK: Sage.

28.

Ivanovski

Hailemariam

Smyth

(2020) The effect of renewable and non-renewable energy consumption on economic growth: Non-parametric evidence. Journal of Cleaner Production ( November): 124956.

29.

Koenker

(2005) Quantile Regression. Cambridge, UK: Cambridge University Press.

30.

Koenker

Bassett

Jr (1978) Regression quantiles. Econometrica 46(1): 33–50.

31.

Kuznets

(1955) Economic growth and income inequality. American Economic Review 45(1): 1–28.

32.

Zhang

, et al. (2019) Does gender inequality affect household green consumption behaviour in China? Energy Policy 135(December): 111071.

33.

Liu

Wang

Zhang

, et al. (2019) Examining the effects of income inequality on CO2 emissions: Evidence from non-spatial and spatial perspectives. Applied Energy 236(February): 163–171.

34.

McGee

Greiner

(2019) Renewable energy injustice: The socio-environmental implications of renewable energy consumption. Energy Research & Social Science 56(October): 101214.

35.

Meniago

Asongu

(2018) Revisiting the finance-inequality nexus in a panel of African countries. Research in International Business and Finance 46(December): 399–419.

36.

Murendo

Wollni

De Brauw

A, et al.

(2018) Social network effects on mobile money adoption in Uganda. The Journal of Development Studies 54(2): 327–342.

37.

Naceur

Zhang

(2016) Financial development, inequality and poverty: Some international evidence. IMF Working Paper, Washington, DC.

38.

Nathaniel

Bekun

(2020) Electricity consumption, urbanization and economic growth in Nigeria: New insights from combined cointegration amidst structural breaks. Journal of Public Affairs.

39.

Nathaniel

Iheonu

(2019) CO₂ abatement in Africa: The role of renewable and non-renewable energy consumption. The Science of the Total Environment 679: 337–345.

40.

Niranjan

(2019) Green growth: Africa chooses between renewables and fossil fuels. DW Made for Minds. Available at: www.dw.com/en/green-growth-africa-chooses-between-renewables-and-fossil-fuels/a-51510277 (accessed 30 April 2020).

41.

Okada

Samreth

(2012) The effect of foreign aid on corruption: A quantile regression approach. Economics Letters 115(2): 240–243.

42.

O’Ryan

Nasirov

Álvarez-Espinosa

(2020) Renewable energy expansion in the Chilean power market: A dynamic general equilibrium modeling approach to determine CO₂ emission baselines. Journal of Cleaner Production 247(February): 119645.

43.

Samour

Isiksal

Resatoglu

(2019) Testing the impact of banking sector development on Turkey’s CO2 emissions. Banking and Finance Department, Near East University, Nicosia, Northern Cyprus. Available at: http://aloki.hu/pdf/1703_64976513.pdf (accessed 21 November 2020).

44.

Selden

Song

(1994) Environmental quality and development: Is there a Kuznets curve for air pollution emissions? Journal of Environmental Economics and Management 27(2): 147–162.

45.

Tchamyou

(2020) Education, lifelong learning, inequality and financial access: Evidence from African countries. Contemporary Social Science 15(1): 7–25.

46.

Tchamyou

(2019a) The role of information sharing in modulating the effect of financial access on inequality. Journal of African Business 20(3): 317–338.

47.

Tchamyou

(2019b) Essays on inequality, knowledge economy and financial development in Africa . Unpublished PhD Thesis, University of Antwerp, Antwerp.

48.

Tchamyou

Asongu

(2017) Conditional market timing in the mutual fund industry. Research in International Business and Finance 42(December): 1355–1366.

49.

Tchamyou

Asongu

Odhiambo

(2019b) The role of ICT in modulating the effect of education and lifelong learning on income inequality and economic growth in Africa. African Development Review 31(3): 261–274.

50.

Tchamyou

Erreygers

Cassimon

(2019a) Inequality, ICT and financial access in Africa. Technological Forecasting and Social Change 139(February): 169–184.

51.

Uzar

Eyuboglu

(2019) The nexus between income inequality and CO₂ emissions in Turkey. Journal of Cleaner Production 227(August): 149–157.

52.

Uzar U (2020) Is income inequality a driver for renewable energy consumption? Journal of Cleaner Production 255: 120287.

53.

Wang

Lin

, et al. (2019) Formation and rupture mechanisms of visco-elastic interfacial films in polymer stabilized emulsions. Journal of Dispersion Science and Technology 40(4): 612–626.

54.

Song

(2011) An empirical study on the environmental Kuznets curve for China’s carbon emissions: Based on provincial panel data. Chinese Journal of Population Resources and Environment 9(3): 66–76.

55.

Yang

Shuangqing

Bing

, et al. (2019) Layout optimization of large-scale oil–gas gathering system based on combined optimization strategy. Neurocomputing 332: 159–183.

The green economy and inequality in Sub-Saharan Africa: Avoidable thresholds and thresholds for complementary policies

Abstract

Keywords

Introduction

Data and methodology

Data

Methodology

Empirical results

Concluding implications and future research directions

Footnotes

Declaration of conflicting interests

Funding

ORCID iDs

Notes

Appendix 1. Definitions of variables

Appendix 2. Summary statistics (2004–2014)

Appendix 3. Correlation matrix (uniform sample: 245)

References