Research Article |
Corresponding author: Mduduzi Biyase ( mduduzibiyase@gmail.com ) Academic editor: Marina Sheresheva
© 2024 Mduduzi Biyase, Frederich Kirsten, Talent Zwane , Santos Bila.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Biyase M, Kirsten F, Zwane T, Bila S (2024) Tracing environmental Kuznets curves: unveiling the interplay of inequality, urbanization, GDP and emissions in BRICS nations. In: Sheresheva M, Lissovolik YD (Eds) Changing the Global Monetary and Financial Architecture: The Role of BRICS-Plus. BRICS Journal of Economics 5(1): 83-104. https://doi.org/10.3897/brics-econ.5.e117948
|
In light of environmental challenges, the BRICS countries have stepped to the forefront of economic progress versus environmental sustainability debate. Not only has energy consumption increased rapidly in these countries, but the economic progress and urbanization, mainly driven by intensive fossil fuel production, have also led to higher levels of income inequality. The dynamics of the interplay between economic growth, urbanization, and income inequality on the one hand and environmental sustainability on the other have yet to be fully understood in the BRICS context. This paper aims to contribute to the ongoing debate by assessing a combination of three Environmental Kuznets Curves (EKC) based on the GDPpc-emissions nexus, the income inequality- emissions nexus, and the urbanization-emissions nexus. Using the Autoregressive Distributed Lag (ADRL) and Panel Fully Modified Least Squares (FMOLS) models, we find an inverted U-shape EKC between GDP and carbon emissions, an inverted U-shaped EKC between income inequality and carbon emissions, and a U-shaped EKC between urbanization and carbon emissions. The inverted EKC between GDPpc and carbon emissions suggests that in the long run sustainable carbon reduction is possible alongside economic growth, but urbanization’s U-shaped impact on emissions might hinder this. Moreover, the inverted U-shaped relationship between income inequality and carbon emissions indicates a potential long-run trade-off between reducing both inequality and carbon emissions. Factors behind this relationship may vary significantly and include institutions- and country-specific factors, yet policymakers in the BRICS countries will do well attempting to better understand the dynamics behind urbanization and inequality as it will enable them to adopt more effective holistic policies aiming to improve energy efficiency, reduce fossil fuel dependence, and build economic systems contributing to faster economic growth, lower inequality and greater environmental sustainability.
По мере усугубления экологических проблем страны БРИКС активно включились в дискуссии о соотношении экономического прогресса и экологической устойчивости. Наблюдающийся в этих странах быстрый рост потребления энергии и урбанизации, а также ускорение экономического прогресса, в первую очередь ставшие результатом интенсивного производства ископаемого топлива, в свою очередь привели к более высокому уровню неравенства доходов. Динамика взаимодействия между экономическим ростом, урбанизацией и неравенством доходов, с одной стороны, и экологической устойчивостью, с другой, пока не нашла достаточно глубокого понимания в контексте БРИКС. Данная статья призвана внести свой вклад в продолжающуюся дискуссию путем оценки комбинации трех экологических кривых Кузнеца (EKC), основанных на взаимосвязи ВВП с процентными выбросами и неравенством доходов и выбросов, а также связи урбанизации и выбросов. Используя модели авторегрессии с распределенным лагом (ADRL) и панельных полностью модифицированных наименьших квадратов (FMOLS), мы находим перевернутую U-образную EKC между ВВП и выбросами углерода, перевернутую U-образную EKC между неравенством доходов и выбросами углерода и U-образную зависимость EKC между ВВП и выбросами углерода. Используя модели авторегрессии с распределенным лагом (ADRL) и панельных полностью модифицированных наименьших квадратов (FMOLS), мы находим перевернутую U-образную EKC между ВВП и выбросами углерода, перевернутую U-образную EKC между неравенством доходов и выбросами углерода и U-образную зависимость EKC между ВВП и выбросами углерода. Инвертированное соотношение EKC между ВВП на процент и выбросами углерода предполагает, что в долгосрочной перспективе устойчивое сокращение выбросов углерода возможно наряду с экономическим ростом, но U-образное воздействие урбанизации на выбросы может этому помешать. Более того, перевернутая U-образная зависимость между неравенством доходов и выбросами углекислого газа указывает на потенциальный долгосрочный компромисс между сокращением неравенства и выбросами углекислого газа. Факторы, лежащие в основе этих отношений, могут значительно различаться и включать в себя факторы, специфичные для институтов и стран, однако политики в странах БРИКС преуспеют, пытаясь лучше понять динамику урбанизации и неравенства, поскольку это позволит им принять более эффективную целостную политику, направленную на повышение эффективности энергетики, снижение зависимости от ископаемого топлива и создание экономических систем, способствующих экономическому росту, снижению неравенства и повышению экологической устойчивости.
income inequality, economic growth, urbanization, environmental sustainably, BRICS
неравенство доходов, экономический рост, урбанизация, экологическая устойчивость, БРИКС
In recent years, environmental challenges have escalated, with a range of pressing issues such as global warming, deforestation, loss of biodiversity, and the heightened occurrence and intensity of extreme weather events coming to the forefront. These issues, highlighted as pivotal global problems (
For emerging economies, particularly those in the BRICS group (China, India, Brazil, Russia and South Africa), economic growth and urbanization have been closely tied to fossil fuel use. The rapid economic expansion of some BRICS countries comes at the cost of higher carbon emissions. BRICS economies account for three of the world’s top five carbon emitters and have surpassed OECD countries in terms of carbon dioxide emissions
Understanding the balance and dynamics of economic growth, inequality urbanization and carbon emissions in BRICS countries is crucial and this is where the Environmental Kuznets Curve (EKC) provide valuable insight. The EKC suggests that the quality of the environment initially worsens with economic development but begins to improve once a certain income per capita level is exceeded. The EKC has been increasingly exploited to find out how economic growth impacts emissions (
Our study aims to delve deeper into the EKC phenomena within the BRICS countries, particularly focusing on the interplay between three EKC patterns: GDPpc, income inequality and urbanization. We hypothesize that each of these EKCs exhibits distinct relationships with emissions and has heterogenous patterns. This research makes several contributions to the field of environmental economics: firstly, it provides a comprehensive analysis of the EKC phenomenon by simultaneously assessing three critical variables, GDPpc, urbanization and inequality and the impact on emissions. Secondly, this study sheds new light on EKCs concerning income inequality and carbon emission in BRICS countries. Given the increasing inequality in some of these nations, the research offers valuable insights into the long-run interplay between economic disparities and environmental sustainability. This is especially significant in the BRICS context, where such dynamics have been less understood. Thirdly, the findings of the study may prove useful for formulating targeted environmental policies and economic strategies that can help these nations achieve sustainable growth without compromising their environmental commitments. Overall, the study should contribute to the existing literature by providing empirical evidence of these relationships, offering new insights into the distinct EKC patterns in a BRICS context.
The rest of the manuscript is structured as follows: Section 2 provides a detailed description of the literature for this study; Section 3 outlines the methodology used; Section 4 presents the descriptive and empirical analysis of the data. Finally, section 5 concludes the findings of the study and presents some policy recommendations.
The causal association between income inequality and environmental quality through greenhouse gas emissions has been contested in empirical studies (see for example,
Although recent empirical work has shown that it is not only income that plays a role here, its distribution is also a fundamental element that determines the level of aggregate emissions and so the quality of the environment (
There are various approaches through which income distribution might explicitly impact per capita carbon dioxide emissions. In his study,
It is interesting to note that
The alternative purported theory to describe the association between income inequality and carbon emissions is the marginal propensity to emit (MPE) (see for example,
Several empirical works have explored the impact of national income and its distribution on per capita carbon emissions; other studies have incorporated the effect of urbanization into the equation (
The previous research recognizes the critical role of urbanization, inequality and GDP growth in increasing emissions. For example, a study by
Among recent contributions,
As concerns inequality, in their seminal paper
In their paper, Hailemariam et al (2019) explored the economic growth- emissions using data on top income inequality measured by the share of pretax income earned by the richest 10% of the inhabitants of the OECD nations and Gini coefficients, as these two measures account for diverse structures of the income distribution (Hailemariam et al, 2019). The authors used panel data models and found a positive association between top-income inequality and emissions (Hailemariam et al., 2019). Likewise,
Recent empirical studies suggest that there is a negative causal association between urbanization, inequality, GDP growth and carbon emission. Several empirical works, such as
Subsequent empirical work, such as
There are studies, however, that have found no significant relationship between these variables. For instance,
While the relationship and non-linearity effect of per capita GDP has been explored quite extensively under the EKC perspective in the environmental field (
To fulfil the objective, the study builds from the EKC, as previously used by
CO 2 = f (IE, GDP, EU, URBN, AGRVA, MANVA, SERVA), (1)
Where
the acronyms denote the following: CO2 (carbon dioxide) IE (income inequality), GDP (gross domestic product per capita), EU (energy use), URBN (urbanization), AGRVA (agricultural value added), MANVA (manufacturing value added) and SERVA (services value added). The CO2 is measured in kilotons, IE measured by the gini coefficient, GDP measured in 2015 constant terms to ensure that it is introduced into the equation in real terms, EU measured in kilograms of oil equivalent (kgoe) per capita, URBN measured as a percentage of the total population, AGRVA, MANVA and SERVA are measured as a percentage of GDP.
Employing the natural logarithm transformation to improve the interpretability of the results, we assemble the linear model as follows:
(2)
Moreover, , and are introduced into the equation to account for the non-linearity in the income inequality, GDPpc and urbanization nexus.
Where β1, β2 , β3 , β4 , β5 , β6 , β7 , β8 , β9 , β10 are the coefficients of the independent variables that shall be estimated, it is the intercept, and αi, eit are country specific effect and error terms. In order to explore the long-run and short-run relationship between these variables, equation (1) is restated as the ARDL equation below:
(3)
Where α0 signifies the intercept, ∆ is the first difference operator, ∅1 , ∅2 , ∅3 , ∅4 , ∅5 , ∅i , ∅6 , ∅7 , ∅8 , ∅i , ∅9 , ∅10 and ∅11 , shows the short-run estimated coefficients of our variables. The long-run estimated coefficient are then deduced from β1, β2 , β3 , β4 , β5 , β6 , β7 , β8 , β9 , β10 , β11 . Once we ascertain the presence of the long-run association between variables, then we estimate the error correction model (ECM) as shown below:
(4)
Equation (4) is similar to equation (3) except for the Δ which represent the coefficient for the speed of adjustment to equilibrium. Due to the fact that different specification estimator can somehow be sensitive, we further use the Full Modified Ordinary Least Square for robustness check.
Table
LCO2 | LIE | LGDPPC | LEU | LURBN | LAGRVAR | LMANVA | LSERVAR | |
Mean | 1.167961 | 3.842233 | 8.313581 | 2.994025 | 3.992577 | 25.06955 | 2.815272 | 27.16802 |
Median | 0.870993 | 3.829722 | 8.704638 | 3.492551 | 4.078765 | 24.86812 | 2.746723 | 27.26665 |
Maximum | 2.475273 | 4.149464 | 9.245531 | 3.969348 | 4.466747 | 27.72248 | 3.479772 | 29.67941 |
Minimum | -0.434712 | 3.190476 | 6.270796 | 1.156881 | 3.240520 | 22.07652 | 2.335294 | 25.33718 |
Std. Dev. | 0.930702 | 0.208024 | 0.908828 | 0.945841 | 0.404001 | 1.613991 | 0.314657 | 1.078579 |
Skewness | -0.052304 | -0.416868 | -1.041112 | -0.782898 | -0.605522 | -0.258362 | 0.673869 | 0.280718 |
Kurtosis | 1.528344 | 3.096453 | 2.626329 | 2.306465 | 1.923973 | 2.185258 | 2.613835 | 2.789717 |
Jarque-Bera | 10.88358 | 3.522101 | 22.37643 | 14.66355 | 13.12230 | 4.654049 | 9.827610 | 1.797143 |
Probability | 0.004332 | 0.171864 | 0.000014 | 0.000654 | 0.001414 | 0.097586 | 0.007344 | 0.407151 |
Observations | 120 | 120 | 120 | 120 | 120 | 120 | 120 | 120 |
The variables of interest in this study focus on uncovering the relationship between carbon emissions, GDPpc, income inequality and urbanization in BRICS countries. Figures
The next focus is on the interplay between income inequality and emissions in Figures
The next step involves checking the stationarity levels of the variables. Table
Variables | Levin, Lin & Chu t* | Im, Pesaran and Shin W-stat | ADF - Fisher Chi-square | PP - Fisher Chi-square |
At level | ||||
LCO2 | -1.67758 | -0.90828 | 15.4086 | 11.6805 |
0.0467 | 0.1819 | 0.1179 | 0.307 | |
LIE | -0.98971 | 1.10143 | 7.2595 | 11.201 |
0.1612 | 0.8646 | 0.7007 | 0.3421 | |
LGDPPC | -2.76163 | 0.74998 | 7.33142 | 6.12064 |
0.0029 | 0.7734 | 0.6938 | 0.805 | |
LEU | -1.62047 | -0.29859 | 8.22229 | 8.23377 |
0.0526 | 0.3826 | 0.6071 | 0.606 | |
LURBN | -1.14935 | 2.45741 | 5.10244 | 57.3877 |
0.1252 | 0.993 | 0.8842 | 0.0000 | |
LAGRVAR | 0.25994 | 3.24924 | 1.65914 | 1.83709 |
0.6025 | 0.9994 | 0.9983 | 0.9974 | |
LMANVA | -0.96007 | 0.69085 | 5.52638 | 5.53209 |
0.1685 | 0.7552 | 0.8534 | 0.8529 | |
LSERVAR | -3.26707 | 0.55771 | 6.90884 | 6.60551 |
0.0005 | 0.7115 | 0.734 | 0.7621 | |
First difference | ||||
LCO2 | -0.5281 | -2.10034 | 20.7494 | |
0.2987 | 0.0178 | 0.0229 | ||
LIE | -0.22653 | -1.25229 | 23.4824 | 44.7113 |
0.4104 | 0.1052 | 0.0091 | 0.0000 | |
LGDPPC | -3.35888 | -3.75591 | 34.7426 | 53.3553 |
0.0004 | 0.0001 | 0.0001 | 0.0000 | |
LEU | 0.92865 | -1.46923 | 23.3907 | 49.181 |
0.8235 | 0.0709 | 0.0094 | 0.0000 | |
LURBN | -1.239 | 14.4334 | 20.118 | |
0.1077 | 0.1541 | 0.0282 | ||
LAGRVAR | -10.9098 | -10.8644 | 80.3604 | 136.242 |
0.0000 | 0.0000 | 0.0000 | 0.0000 | |
LMANVA | -4.31095 | -4.67288 | 42.04 | 59.6569 |
0.0000 | 0.0000 | 0.0000 | 0.0000 | |
LSERVAR | -2.61077 | -2.57699 | 25.6582 | 45.3358 |
0.0045 | 0.005 | 0.0042 | 0.0000 |
Utilizing the unit root findings, in table 3 we now present the test results of Pedroni cointegration for the GDPpc-; inequality- emissions and urbanization- emissions nexus. Predictably the majority of the statistics (such as the Panel PP-Statistic, Panel ADF-Statistic, Group PP-Statistic and Group ADF-Statistic) fail to accept the null hypothesis of no cointegration. In other words, the Pedroni cointegration results confirm that the variables have a long-run relationship.
Within-dimension | ||||
Weighted | ||||
Statistic | Prob. | Statistic | Prob. | |
Panel v-Statistic | -0.880466 | 0.8107 | -3.122421 | 0.9991 |
Panel rho-Statistic | 3.690677 | 0.9999 | 3.995235 | 1.0000 |
Panel PP-Statistic | -5.312261 | 0.0000*** | -5.306100 | 0.0000*** |
Panel ADF-Statistic | -1.785919 | 0.0371*** | -1.958545 | 0.0251*** |
Between-dimension | ||||
Statistic | Prob. | |||
Group rho-Statistic | 4.617716 | 1.0000 | ||
Group PP-Statistic | -8.756492 | 0.0000*** | ||
Group ADF-Statistic | -2.202279 | 0.0138*** |
Since we confirm the long-run cointegration between emissions and the variables of interest, we now perform an PADRL model to assess the long-run results of the study. The results of the PADRL model in Table
Variable | Coefficient | Std. Error | t-Statistic | Prob.* |
Long Run Equation | ||||
LIE | 2.927244 | 1.037676 | 2.820961 | 0.0068 |
-0.383767 | 0.136866 | -2.803958 | 0.0072 | |
LGDPPC | 1.183061 | 0.591942 | 1.998611 | 0.0511 |
-0.005808 | 0.032126 | -0.180771 | 0.8573 | |
LEU | -0.624737 | 0.019281 | -32.40173 | 0.0000 |
LURBN | -5.392398 | 1.422524 | -3.790725 | 0.0004 |
0.518098 | 0.167952 | 3.084797 | 0.0033 | |
LSERVAR | -0.141797 | 0.095791 | -1.480283 | 0.1451 |
LAGRVAR | 0.201318 | 0.050426 | 3.992313 | 0.0002 |
LMANVAR | -0.052382 | 0.023306 | -2.247572 | 0.0290 |
Short Run Equation | ||||
COINTEQ01 | -0.905449 | 0.249165 | -3.633936 | 0.0007 |
D(LCO2(-1)) | -0.051051 | 0.327283 | -0.155985 | 0.8767 |
D(LIE) | -196.9917 | 239.6183 | -0.822106 | 0.4149 |
D() | 23.68570 | 29.29755 | 0.808453 | 0.4227 |
D(LGDPPC) | 56.34348 | 38.60980 | 1.459305 | 0.1507 |
D() | -3.163045 | 2.138215 | -1.479293 | 0.1453 |
D(LEU) | 0.078170 | 0.258001 | 0.302983 | 0.7632 |
D(LURBN) | -2531.321 | 2499.045 | -1.012915 | 0.3160 |
D() | 292.6175 | 288.0523 | 1.015848 | 0.3146 |
D(LSERVAR) | 0.236631 | 0.555148 | 0.426249 | 0.6718 |
D(LAGRVAR) | -0.390277 | 0.481624 | -0.810335 | 0.4216 |
D(LMANVAR) | 0.113816 | 0.092361 | 1.232293 | 0.2236 |
Root MSE | 0.015314 | Mean dependent var | 0.019657 | |
S.D. dependent var | 0.043268 | S.E. of regression | 0.023724 | |
Akaike info criterion | -4.591847 | Sum squared resid | 0.028142 | |
Schwarz criterion | -2.965810 | Log likelihood | 345.5108 | |
Hannan-Quinn critter. | -3.931506 |
This study’s main contribution is testing the dynamics of three different Environmental Kuznets Curves (EKC): GDP and CO2 emissions, inequality and CO2 emissions and urbanization and CO2 emissions. The results support the inverted U-shape EKC for inequality and GDP, where both LIE and LGDPpc are positively and significantly related to CO2 emission, while their squared terms ((LIE2 LGDPpc2) are negatively related. Initially, an increase in GDP and income inequality leads to higher levels of CO2 emissions; in the long run, however, this relationship reverses, and higher levels of GDP and income inequality lead to lower CO2 emissions. For GDP and CO2 emissions, BRICS countries would initially experience higher levels of CO2 emissions, but economic development will eventually lead to lower CO2 emissions once GDP surpasses a certain threshold. The results of an inverted U-shape EKC within BRICS countries are consistent with other studies like
The relationship with income inequality is less clear. Given the high level of inequality in these countries, reducing both inequality and their carbon footprint is crucial. However, the inequality- relationship shows an inverted U-shape EKC, where initially income inequality and CO2 emissions have a positive and significant relationship, suggesting the potential for simultaneous reduction. However, over time the relationship becomes negative and higher inequality is associated with a lower carbon footprint. This suggests a potential trade-off between inequality and carbon reductions in the current development path of BRICS nations. The findings of an inverse relationship between income inequality and environmental degradation are not uncommon and some have found it to be linked to middle-income countries (
In contrast to GDPpc and income inequality, we find that urbanization has a U-shaped EKC, where the urbanization-CO2 emission nexus starts with an initial negative relationship that becomes positive after a certain threshold. This indicates that increased levels of urbanization in BRICS countries initially lead to lower levels of CO2 emissions, but in the long run there will be a trade-off between further urbanization and CO2 emissions. These results are in line with Pianoing and Kaneko (2010) and
The negative effect of LSERVAR and LMANVA on emissions could be explained by technological advancements in manufacturing that reduce carbon footprints or by the shift of economic structures toward service-orientated sectors that produce markedly less carbon emissions (
Table
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
LIE | 2.177887 | 0.052296 | 41.64546 | 0.0000 |
-0.337711 | 0.053209 | -6.346830 | 0.0000 | |
LGDPPC | 5.398281 | 0.077035 | 70.07602 | 0.0000 |
-0.265393 | 0.043559 | -6.092692 | 0.0000 | |
LEU | -0.289742 | 0.068172 | -4.250157 | 0.0000 |
LURBN | -14.86930 | 0.016053 | -926.2708 | 0.0000 |
1.916970 | 0.004907 | 390.6931 | 0.0000 | |
LSERVAR | -0.436541 | 0.044440 | -9.823173 | 0.0000 |
LAGRVAR | 0.108272 | 0.041084 | 2.635346 | 0.0097 |
LMANVA | -0.082952 | 0.007529 | -11.01834 | 0.0000 |
R-squared | 0.996464 | Mean dependent var | 1.171670 | |
Adjusted R-squared | 0.995969 | S.D. dependent var | 0.926083 | |
S.E. of regression | 0.058800 | Sum squared resid | 0.345744 | |
Long-run variance | 0.000593 |
In this paper, our main purpose is to investigate the impact of GDPpc, income inequality, and urbanization through separate Environmental Kuznets Curves. The PARDL and FMOLS regression results both suggest that GDPpc follows an inverted U-shaped EKC, urbanization a U-shaped EKC, and income inequality an inverted U-shape. These findings highlight the importance of understanding long-run environmental sustainability among BRICS nations. For example, the inverted U-shape between GDPpc and carbon emissions suggest a future threshold turning point for BRICS nations. This turning point could see greater public demand for environmental quality, more resources available for cleaner technologies and a transition from manufacturing and agriculture-centered industries towards service-based industries. These components support the notion that the long run GDPpc-carbon emission relationship will eventually reach a threshold, leading to lower carbon emissions with greater economic growth. Ultimately this aligns with the goal of BRICS nations to harmonize economic development and environmental sustainability.
However, the results of the other two EKCs carry vital implications for the growth-emissions nexus. We find that there is a U-shaped relationship between urbanization and carbon emissions, indicating that initial urbanization does not pose a threat to carbon emissions. Only after a certain urbanization threshold is breached will long-run urbanization lead to higher levels of carbon emissions, potentially disrupting the environmental sustainability targets of BRICS nations. Since urbanization is a natural process of economic development, policymakers should focus on sustainable urban planning, renewable energy technology and innovations to reduce the harmful impact of the current urbanization trajectory. Without such measures, the ideal scenario of high growth and low emissions might remain unattainable.
Moreover, the inverted U-shape relationship between income inequality and carbon emissions may point to an alarming scenario where future carbon reductions might come at the expense of higher inequality. Reducing inequality remains a crucial objective for BRICS countries; therefore, policymakers should consider growth strategies that will lead to lower inequality while ensuring environmental sustainability. Overall, the success of BRICS countries in reaching climate change targets highly depends on policymakers’ ability to maneuver between economic growth, urbanization, rising inequality and increasing carbon emissions.