Introduction

The call for decarbonization is becoming a matter of global interest among all and sundry in the international community. The reason is not far-fetched from the intensifying needs to address how the world can avert the challenges of climate change and other environmental degradations. The consequences of the failure to take urgent and important steps in this direction have been described to be tantamount to calling for a catastrophic climate disaster (UNEP 2021; IPCC 2021). While the fight for a carbon–neutral world is a global matter, the collective efforts of individual nations would be a very crucial aspect of achieving the desired success of the decarbonization campaign.

Energy use has been on the increase in Africa in the last couple of decades even as most of the nations are rapidly urbanizing amid the fast-growing population that has seen many projecting the continent to become the future most populous continent due to high birth rates with over 1.136 billion people inhabiting Sub-Saharan Africa countries alone as of 2020 (WDI 2020; Engelman 2016). Although rapid urbanization may have its advantages in Africa, there are potential environmental challenges from this development as some studies have pointed out the risks of population growth and climate change in Africa among other issues (Ahmadalipour et al. 2019). The International Energy Agency has noted that the African continent is more vulnerable to the impact of global warming, climate variability, and other environmental challenges compared to other continents despite being among the least contributors to global carbon emission (IEA 2020). Besides, it has also been estimated that there could be a shrink in Africa’s gross domestic product (GDP) per capita growth to the tune of about fifteen percent as a result of undesirable environmental challenges that are related to climatic change in Africa (Baarsch et al. 2020).

While many economies in Africa continue to experience rapid urbanization in the era of globalization, the adverse environmental consequences are most likely going to cut across several aspects of the African economies, thereby exacerbating the numerous socioeconomic challenges that are currently confronting the continent. For instance, pronounce occurrence of droughts and longer duration of dry seasons stands to aggravate the problem of food insecurity in Africa considering that agricultural practices mostly depend on seasonal variations in the amount of rainfall. Aside from the food insecurity, instability in the agricultural sector will also connote a huge economic loss as the agricultural sector usually accounts for a significant proportion of the GDP, overall employment, and income generation in many African nations (Diao et al. 2010; Onifade et al. 2020a, b; Salahuddin et al. 2020).

Therefore, addressing environmental challenges in Africa by examining the potential contributing factors to pollution levels will help to position the continent on a sustainable path. As such, in the current study, we examine the impacts of increasing urbanization alongside its interactions with energy portfolios on environmental prospects of 15 selected African countries including the most urbanized and leading oil producers in the continent of Africa. While doing so, the study also factors in the roles of economic growth among the countries given the rising trends of globalization in our increasingly interconnected world. The study provides useful insights on relevant policy recommendations that are critical for enhancing environmental sustainability and sustainable economic growth, which are paramount focus points of most policymakers especially in developing economies by attempting to answer the following questions:

  1. i.

    Does urbanization amid economic globalization aid environmental quality in Africa?

  2. ii.

    What are the impacts of growing per capita income levels on environmental degradation in Africa?

  3. iii.

    How do energy portfolios impact environmental quality in Africa given the rapid urbanization trend in recent times?

  4. iv.

    Can African economies bypass detrimental economic consequences while implementing energy conservation policies for environmental sustainability?

To circumvent biased and spurious results, a combination of robust econometric approaches including quantile regression (QR), the dynamic ordinary least squares (DOLS), and fully modified ordinary least squares (FMOLS) were used in the current study. The first approach (QR) essentially makes it possible to observe the conditional distribution effects of the understudied environmental indicators (urbanization, globalization, income, and energy portfolios) on the pollution level among the countries and the method is robust for dealing with fundamental issues like cross-sectional dependence and outliers vis-à-vis error distribution. On the other hand, the other approaches make complementary and comparative analysis possible as they depend on the mean estimates, unlike the QR method. The combination of the approaches essentially helps to produce insightful results that informed useful policy directives.

The vast majority of empirical studies in the environmental literature often focus on advanced economies, especially the European Union (EU) and major emerging economies like China, Brazil, Turkey, and so on. Aside from providing crucial insights into environmental matters in Africa where the literature is relatively unsaturated compared to the rest of the world, this study also ensures that major oil-producing countries in the continent were accommodated in the analysis. Doing this makes the study worthwhile as many of the rapidly urbanizing African countries mainly rely on conventional energy use due to the continent’s fossil energy potential, and this aspect has often been ignored in extant studies. Fossil energy accounts for the largest proportion of the total energy consumption in the continent as oil consumption accounts for 42%, while gas consumption accounts for 28% of the total energy consumption. Also, the consumption of other fossil energy sources like coal accounts for around 22% of total energy consumption. This is not surprising given that Africa accounts for about 9.1% and 6% of global oil and natural gas production, respectively, and about 4.2% and 3.9% of the consumption, respectively (UNEP 2017).

In Fig. 1, the transport sector is expectedly the leading sectorial demand for oil consumption on the continent, followed by industrial energy demand and residential energy consumption from oil. The United Nations Environment Program (UNEP 2017) has pointed out that economic growth, population growth, and urbanization are important factors to be considered as far as energy demand is concerned in Africa. Besides, these are among the several factors that have dominated the empirical literature vis-à-vis the major driving forces for greenhouse gas (GHG) emissions as they have a high tendency to influence energy use among nations in our rapidly urbanizing world (Dogan and Turkekul 2016; Ozturk and Acaravci 2016; Leitão and Balsalobre-Lorente 2021; Alola et al. 2021; Shahbaz et al. 2020; Onifade et al. 2021a; Al-Mulali and Ozturk 2015). Other factors include the influence of interconnectedness of economies via expanding trade volumes as nations strive to maintain an upward economic growth trajectory in our vastly globalized world (Shahbaz et al. 2017, 2019; Destek 2020; Wang et al. 2020; Adebayo et al. 2021a; Saint Akadiri et al. 2020).

Fig. 1
figure 1

Source: authors’ computation using data from IEA 2020. Data are given in kilotons of oil equivalent (ktoe)

Distribution of oil consumptions across selected sectors in Africa.

Full size image

A synopsis of related studies on urbanization and its environmental impacts amid economic growth and globalization is provided in Table 1. The table summarizes methods used in the studies, the sample of studies, and the overall results alongside the conclusions. Generally, there is no unanimity in terms of results in the empirical literature as most of the studies produced varying results as seen in Table 1.

Table 1 Related studies summarized
Full size table

Hereafter, the other aspects of the study are structured into three sections in the following order: the information on data and methodology are organized in the Methodology section, while the Discussion of Results section contains the analysis and interpretations of outcomes of the simulations. The Conclusion and Policy Recommendations section concludes the research with the study’s implications and policy framework.

Methodology

An overview of the data and baseline model

To access the level of environmental sustainability in Africa amid the rapid urbanization that is being witnessed on the continent in recent times, this study essentially utilized data from the World Bank Development Indicators (WDI 2020) and the KOF Swiss Institute (Gygli et al. 2019) for a group of fifteen (15) selected African countries between 1990 and 2015, including Egypt, Nigeria, Algeria, Sudan, Tunisia, Libya, South Africa, Angola, Gabon, Congo Republic, Mozambique, Senegal, Tanzania, Kenya, and the Democratic Republic of Congo (DRC). Data sourcing is a major challenge with African countries, and the scope of the present study is limited to the extent of data available from those organizations. The country selection was carried out based on two important yardsticks vis-à-vis the aim of the study. First is the consideration for the level of urbanization in terms of the proportion of the population living in the urban settlements to the total population, and the second is the energy portfolios of countries. For the latter condition, the study prioritized the case of nations that are generally known to be rich in conventional energy sources due to their vast fossil energy resources deposits such as is the case for oil and gas-rich countries like Nigeria, Libya, Angola, Algeria, and Egypt and also the case of South Africa when considering coal resources. Notably, some of these countries apart from being rich in conventional energy resources are still among the most urbanized on the continent including others on the list. The study’s baseline model is provided in Eq. 1.

$${\mathrm{LnCO}}_{2it}={\alpha }_{0}+{\alpha }_{1}{\mathrm{LnIC}}_{it}+{\alpha }_{2}\mathrm{Ln}{{\mathrm{IC}}^{2}}_{it}+{\alpha }_{2}{\mathrm{LnUB}}_{it}+{\alpha }_{4}{\mathrm{LnUBFF}}_{it}+{\alpha }_{3}{\mathrm{LnUBRW}}_{it}+{\alpha }_{5}{\mathrm{LnEGZ}}_{it}+{\varepsilon }_{it}$$
(1)

In Eq. 1, the measure of environmental quality is the level of carbon emission (LnCO2) in the countries, while the number of populations in the urban as a fraction of the total population was used to capture the level of urbanization (LnUB). Two main interaction terms were incorporated into the model to factor in the impact of energy portfolios within the urbanization context, and priority was given to both fossil and renewable energy aspects. The interaction between the level of urbanization and fossil energy resources, namely the amount of electricity production from oil, gas, and coal sources as a % of total electricity generation was represented by LnUBFF. On the other hand, the interaction between the level of urbanization and the level of renewable energy use as a proportion of total energy consumption was denoted by LnUBRW. Given the growing quest to sustain economic growth in our increasingly globalized world, the proxies for both economic growth and globalization as captured by real per capita income (LnIC) and the economic globalization (LnEGZ), respectively, were incorporated into the model. LnCO2 is given in metric tons, while the income levels are in the current US$. Lastly, the variable LnIC2 represents the square of real per capita income level. The introduction of this variable helps to simultaneously assess whether the well-known environmental Kuznets curve (EKC) hypothesis is valid for this group of countries within the context of their urbanization experience. Introducing the square value of income level aligns with existing approaches in some empirical studies (Apergis and Ozturk 2015; Bekun et al. 2021b). All of the variables were taken in the natural logarithm for ease of analysis in elasticity form and the basic statistical properties of the variables are given in the result section.

Procedures for estimation

Analysis in the empirical stage begins with conducting a combination of necessary pre-estimation tests. It is highly imperative to examine the statistical features of the panel data set as a guide for the proper choice of estimation techniques. Considering the level of interconnectedness of nations especially in our globalized world, a cross-sectional dependence (CD) test, therefore, heralds the pre-estimation tests. This action is necessary to guide the choice of proper panel unit root test as well as method selection for the cointegration test as seen in contemporary empirical studies (Chudik et al. 2016; Bekun et al. 2021a, b; Sinha and Sengupta 2019; Gyamfi et al. 2021).

$${Y}_{it}= {\delta }_{i}+{\alpha }_{i}{X}_{it}+{\mu }_{it}$$
(2)

Following a simplified panel expression between variable Y and X in Eq. 2, a cross-sectional dimension denoted by (i) is given for the panel observations ranging from 1 to N in period (t) spanning from 1 to T as shown in Eq. 3. Subsequently, a null hypothesis that supports the absence of cross-sectional dependence (correlation) in residuals, whereby Cov(\({\mu }_{it}\),\({\mu }_{jt}\)) = 0 is formulated in contrast to an alternative hypothesis that supports cross-section dependence in residuals at least in a pair of the given cross-sections whereby Cov(\({\mu }_{it}\),\({\mu }_{jt}\)) ≠ 0.

$$LM=T\sum\nolimits_{i=1}^{N-1}\sum\nolimits_{J=i+1}^N\rho\,\hat\,\begin{array}{c}2\\ij\end{array}\chi_{N(N-1)/2}^2$$
(3)

The pairwise correlation of the estimated residuals (ρˆij) obtained from the OLS results of Eq. 2 following the Lagrange multiplier (LM) approach of Breusch and Pagan (1980) is denoted by (ρˆij), while the LM test for cross-sectional dependence of Pesaran (2015) was used in line with Eq. 4 and Eq. 5. The method offers an advantage as it accounts for matters of cross-sectional dependence and slope heterogeneity, especially in not too big or relatively small sample observations.

$$\mathrm{CD}=\sqrt{\left(\frac{2T}{N(N-1)}\right)}\left(\sum\nolimits_{i=1}^{N-1}\sum\nolimits_{J=i+1}^N{\rho\,\hat\;}_{ij}\right)$$
(4)
$$ \begin{aligned}{\rho{\hat\;}}_{ji}={\rho{\hat\;}}_{ij}=\frac{\sum_{t-1}^T{\mu{\hat\;}}_{i,t}{\mu{\hat\;}}_{j,t}}{\left(\sum_{t=1}^T\mu{\hat\;}\begin{array}{c}2\\it\end{array}\right)^\frac12\left(\sum_{t=1}^T\mu{\hat\;}\begin{array}{c}2\\jt\end{array}\right)^\frac12}\end{aligned}$$
(5)

It is assumed that the obtained residuals (\(\mu\)) should be asymptomatically distributed vis-à-vis their test statistics such that CD ~ N (0, 1). Thereafter, to evaluate the long-run relationship between the panel variables of interest in the study, the study adopted a unit root analysis that can account for CD pitfalls, since the test result came out positive as fully detailed in the discussion section. As a result, the CIPS panel unit root test (Pesaran 2007) and the IPS test (Im et al. 2003) were jointly utilized in the study. Subsequently, Westerlund’s (2007) cointegration technique was applied to confirm a long-run relationship between the understudied panel variables. The Westerlund (2007) approach depends on the mechanism of error correction in Eq. 6, and this approach is also compatible for long-run tests under analysis that is marred by the CD shortfalls (Sinha and Sengupta 2019).

$$\Delta {Y}_{it}={\mathrm{\alpha }}_{i}{D}_{t}+{\mathrm{\varnothing }}_{i}{Y}_{it-1}+{\lambda }_{i}{X}_{it-1}+{\sum }_{j=1}^{pi}{\mathrm{\varnothing }}_{ij}{\Delta Y}_{\mathrm{i},t-j}+{\sum }_{j=0}^{pi}{\gamma }_{ij}{\Delta X}_{i,t-j} +{\varepsilon }_{it}$$
(6)

While \({\mathrm{\alpha }}_{t}\) stands for the vector of the individual parameters, the Dt represents the model’s deterministic arrangement in Eq. 6, and the deterministic pattern can be designed to reflect an interactive arrangement among variables without deterministic components {Dt = (0)}, or with a constant only component {Dt = (1)}, and sometimes as a model with the combination of both trend and constant {Dt = (1, t)}. The Westerlund approach produces both group statistics (Gt, Gα) and panel statistics (Pt, Pα), which assists in the assessment of the cointegration relationship based on the estimation of the error adjustment process (\({\mathrm{\varnothing }}_{i}\)).

Long-run analysis

As for the long-run analysis, a combination of robust econometric approaches including quantile regression (QR), the DOLS of Pedroni (2001a, b), and FMOLS of Pedroni (2001a, b) were used in the current study. The QR is in line with the foundational work of Koenker and Bassett (1978) as further advanced by Koenker (2004) and Powell (2016). The first approach (QR) essentially makes it possible to observe the conditional distribution effects of the understudied environmental indicators (urbanization, globalization, income, and energy portfolios) on the pollution level among the countries, and the method is robust for dealing with fundamental issues like cross-sectional dependence and outliers vis-à-vis error distribution (Nwaka et al. 2020). On the other hand, the other approaches make complementary and comparative analysis possible as they depend on the mean estimates, unlike the QR method.

$${\mathrm{QLnCO}}_{2it}(\tau /{\chi }_{it})= {\beta }_{i}^{(\tau )}+{\beta }_{1}^{(\tau )}{\mathrm{LnIC}}_{it}+{\beta }_{2}^{(\tau )} {\mathrm{LnIC}}_{it}^{2}+{\beta }_{3}^{(\tau )}{\mathrm{LnUB}}_{it}+{\beta }_{4}^{(\tau )}{\mathrm{LnUBFF}}_{it}+{\beta }_{5}^{(\tau )}{\mathrm{LnUBRW}}_{it}+{\beta }_{6}^{(\tau )}{\mathrm{LnEGZ}}_{it}+{\varphi }_{it}$$
(7)

The representations in Eq. (7) follow the interaction among variables in the baseline Eq. 1 such that \({\tau }^{\mathrm{th}}\) represent the conditional quantile of carbon emission levels as the measure of environmental pollution in the expression \({\mathrm{QLnCO}2}_{it}(\tau /{\chi }_{it})\) given that the vector of individual independent variables is represented by \({\chi }_{it}\). On the other hand, tau (\(\tau\)) represents the selected quantiles for the panel countries i in time t, while the slope parameters for the individual independent variables and the error term for the corresponding vector are denoted by \(\beta\) and \({\varphi }_{it}\) accordingly. In a nutshell, the combination of the approaches essentially helps to circumvent biased and spurious results and produce insightful results that informed useful policy directives. Finally, the analysis procedure closes with an evaluation of the direction of causality among the underlining variables of the study using the Granger causality method of Dumitrescu and Hurlin (Dumitrescu and Hurlin 2012). The findings from the empirical procedures have been structured out in the results discussion section.

Discussion of results

The preliminary tests

In the discussion section, an overview of the summary statistics for the panel variables is presented in Table 2 to include the mean values, the median, the maximum and minimum values, and the standard deviations of the samples. Furthermore, Table 2 also presents the correlation matrix for the sample. The correlation matrix reveals that the panel variables positively correlate with carbon emission except for the interaction between urbanization and renewable energy consumption. However, simple correlation analysis would be insufficient without accounting for the statistical properties of the panel variables through an in-depth examination. Besides, the results for the combined CD tests also validate the existence of CD among the variables for the sample countries as the null hypothesis of no cross-sectional dependency can be conveniently rejected following the significance of the probability value of the individual test statistics in Table 3. As such, the adopted unit root test approaches took into cognizance of this development in the subsequent analysis. The results in Table 4 reveal that the understudied panel variables are essentially first-order stationary variables. The IPS approach specifically rejects the evidence of stationarity at a level point for all variables, thereby corroborating the CIPS evidence of stationarity for the panel variables at first difference.

Table 2 An overview of the summary statistics and correlation matrix
Full size table
Table 3 Results for the CD test
Full size table
Table 4 Outputs for the unit root test
Full size table

The cointegration results from the Westerlund (2007) approach as shown in Table 5 reveal that the null of no cointegration among variables can be rejected following the significance of the estimates with evidence from at least each of the group and panel statistics, hence, signifying the existence of a long-run connection among the understudied panel variables. This result, therefore, precedes the need to explore the underlying long-run coefficients for the study.

Table 5 Estimates for Westerlund (2007) cointegration
Full size table

The long-run estimates and causality analysis

The results from the long-run analysis are detailed in Table 6. The findings from the QR approach in the table show the effects of the conditional distribution of the understudied environmental indicators (urbanization, globalization, income, and the interaction of urbanization with energy portfolios) on carbon emission as a measure for environmental pollution level among the African economies. To begin with the environmental impacts of urbanization and globalization, the QR coefficients for these indicators show that the rapid urbanization among the countries as well as the level of economic globalization are significantly detrimental to environmental quality as the effects of these variables are positive and very significant across all the distribution of pollution level throughout the given quantiles ranging from the lower to the mid, and upper quantiles (\(\tau\) = 0.10 to \(\tau\) = 0.30), (\(\tau\) = 0.40 to \(\tau\) = 0.60), and (\(\tau\) = 0.70 to \(\tau\) = 0.90), respectively. The results of the median observations of the QR are also in agreement with the observed findings from the mean estimate approaches of both the DOLS and FMOLS, thereby corroborating many studies where there is evidence of detrimental effects of both urbanization and globalization in the empirical literature (Anwar et al. 2021; Dogan and Turkekul 2016; Onifade et al. 2021c; Destek 2020; Wang et al. 2020).

Table 6 Long-run coefficients
Full size table

While urbanization significantly poses threat to environmental sustainability, the evidence obtained regarding its interaction with energy portfolios of the understudied countries differs. The significant detrimental environmental impacts of the interaction between urbanization and energy portfolios were only confirmed in the context of fossil energy consumption among the countries as seen in the positive coefficients across the quantiles of CO2 emissions. Contrarily, the interaction with renewables exists as a significant decarbonization channel within the framework of the increasing level of urbanization among the countries. These results were also backed up by the mean estimate of the DOLS and FMOLS as a percent rise in the interaction with renewable energy consumption corresponds with a 0.368% fall in carbon emissions, while a percent rise in the interaction with fossil energy aggravates emission levels by 0.044% accordingly. The current results uphold the position of some studies regarding the exacerbating and cushioning effects of fossil energy use and renewable energy use on environmental pollution, respectively (Shahbaz et al. 2020; Bekun et al. 2021a, b; Adebayo et al. 2021b; Kirikkaleli et al. 2021; Gyamfi et al. 2022; Leitão and Balsalobre-Lorente 2021; Adebayo and Rjoub 2021).

In addition, the countries also witnessed environmentally detrimental economic growth for the understudied period as resonated by the positive and significant coefficients of the income level across all the distribution of carbon levels among the countries for all the quantiles. The results of the median observations of the QR are also in agreement with the observed findings from the mean estimate approaches of both the DOLS and FMOLS as a percent rise in per capita income level is expected to trigger carbon emissions by about 0.168% and 0.154%, respectively. This observation upholds some empirical evidence in the literature about pollution-triggering effects of growth (Su et al. 2021). On the other hand, the significant negative coefficients of the impacts of the square income across all quantiles confirm the EKC conjecture for the study, and this is also upheld in the mean estimate of the FMOLS. The implication of the EKC conjecture is that the rising income level is aggravating pollution among the country in the meantime, but this pollution aggravation is expected to reach a peak level after which income expansion is expected to start cushioning carbon emission levels among the understudied African economies.

Lastly, following the causality evidence among urbanization, economic growth, carbon emission, and globalization in Table 7, a one-way causality flows from per capita income level to carbon emission levels and urbanization among the countries and not the reverse direction. This further buttresses the long-run results that economic growth is the driver of emission among the countries. On the other hand, there is a two-way causality flow between urbanization, emission, and globalization. The causality channel, therefore, implies that carefully orchestrated emission reduction schemes will have little or no detrimental effects on sustainable economic growth. This is a welcome result; however, the matters of the detrimental pollution effects of urbanization must be carefully addressed to ensure environmental sustainability, since urbanization granger causes carbon emission.

Table 7 Granger causality
Full size table

In a nutshell, from the study, urbanization plays a significant role in the environmental pollution dynamics of Africa as evidenced in the understudied African countries, and this calls for authorities and policymakers to be proactive in addressing urbanization issues as a crucial matter not just for economic gains but also for environmental sustainability.

Conclusion and policy recommendations

This study focuses on assessing the environmental implications of the rapid urbanization being witnessed among African states amid the dynamics of energy use in our increasingly globalized world. To this end, the study applies a combination of approaches for empirical data analysis for a sample scope ranging from 1990 to 2015 for a total of fifteen African countries. The evidence obtained from the empirical analysis shows that the duo of urbanization and economic globalization reduces the quality of the environment by inducing CO2 emissions. The countries also witnessed environmentally detrimental economic growth for the understudied period. The evidence obtained regarding the effects of the interaction between urbanization and energy portfolios differs as it supports a favorable environmental effect when considering interaction with renewable energy use but a detrimental effect concerning fossil energy production. Furthermore, a one-way causality flows from per capita income level to carbon emission levels and urbanization among the countries and not the reverse direction, while a two-way causality flow between urbanization, emission, and globalization.

The causality channel, therefore, implies that carefully orchestrated emission reduction schemes will have little or no detrimental effects on sustainable economic growth. While this is a welcome result, however, the matters of the detrimental pollution effects of urbanization must be carefully addressed to ensure environmental sustainability, since urbanization granger causes carbon emission. As such, to mitigate environmental pollution and attendant climate crisis that can be exacerbated by urbanization amid economic globalization among the African countries, it is recommended that authorities in these nations should promote investments in green energy resources. Renewables significantly mitigate carbon emissions in these nations, and supporting advancements of energy production and consumption from renewable resources would therefore foster decarbonization and ultimately facilitate environmental sustainability on the African continent.

Africa has a huge natural advantage of benefitting from renewable resources ranging from hydro resources potential to solar energy, wind energy, and even geothermal. For example, the continent’s vast renewable potential in solar and wind energy as shown in Fig. 2 in the Appendix can be tapped into and fully maximized especially for energy production rather than focusing on energy generation from fossil resources, which have been confirmed to be enhancing carbon emission growth and ultimately detrimental to environmental sustainability.

Fig. 2
figure 2

Source: African Development Bank (AfDB 2014)

A map of Africa showing solar and wind energy potential.

Full size image

Furthermore, the selected African countries also need to pursue green infrastructural development plans for crucial sectors of the economy where energy demands are pronounced such as the transportation sector and the industrial sector among others as seen in Fig. 1. Such plans would bring two advantages for the understudied countries if implemented in both urban and rural settlements. Firstly, green infrastructural investments would trigger environmental benefits by ensuring a cleaner and quality environment from reduced greenhouse gas emissions, and secondly, it would further assist the African countries in combatting the growing pressure on urban infrastructure due to rural–urban migration, thereby enhancing the overall quality of life in the urban settlements toward fostering the realization of SDG 11 that emphasizes the need for sustainable cities and communities.

Limitation of the study and future recommendation

The current study is mainly constrained in terms of the scope since not all African countries were accommodated in the empirical analysis. Future studies can therefore expand on the current framework to cover more African economies, and attention can also be paid to the exploration of the roles of the understudied variables on a country-specific basis. Doing this may yield more advantages in addressing environmental pollution challenges beyond a group analysis considering some possible influence of country-specific differences.