Financial inclusion is a multifaceted concept that covers various aspects, such as the extent of inclusion, service providers, target audiences, pricing, and types of financial services offered [16]. The literature on financial inclusion is shaped by different interpretations of the term, which mirror the viewpoints and goals of policymakers, regulators, financial institutions, researchers and customers [17]. It highlights the importance of covering a range of financial services that are accessible and used by both people and businesses [18]. The United Nations Capital Development Fund (UNCDF) expands upon this idea by explaining that financial inclusion is accomplished when all individuals capable of carrying out financial activities can access a broad selection of top-notch, affordable, and easy-to-manage financial products and services [19]. In the same way, Pesqué-Cela et al. [20] describe financial inclusion as offering affordable financial products and services that are easily accessible to everyone regardless of their wealth levels and organisation size.

Properly assessing financial inclusion means considering its various aspects [21]. Relying solely on the percentage of adults using banking services as a measure of financial inclusion overlooks elements such as the availability and usage of financial services beyond mere access to the financial system itself. Accessibility is critical to understanding financial inclusion; however, it also covers the access and use of services within a comprehensive financial system. Beck and De La Torre [22] claim that various other methods to assess the reach of the banking sector have been put in place, such as geographic branch penetration, per capita loan and deposit accounts, loan-income and deposit-income ratios, and comparable indicators with similar measures. These indicators offer valuable and meaningful perspectives into the inclusiveness of a financial system, encompassing a variety of financial inclusion components. However, when used alone, they might not provide an adequate grasp of how inclusive the financial system is. The scenario shows that one could misunderstand the extent of inclusion in a society if attention is solely on individual indicators instead of combined.

The financial access survey (FAS) offered by the International Monetary Fund (IMF) is widely used in research studies focusing on financial inclusion. It covers two main aspects − the usage of financial services and products and their geographical reachability. According to a report by the IMF Secretariat in 2014, the FAS is a crucial source of information, providing global data about the availability of financial services worldwide. It includes standardised measures for assessing the accessibility and usage of financial services. Since 2014, the FAS has included metrics related to mobile money. Annual data commencing in 2004 are included in this database, comprising 47 fundamental indicators and 152 time series. There are two main categories: i) The geographic scope comprises the number of commercial banks and the density of ATMs and bank branches per 1000 square metres; ii) Financial services utilisation is defined as the ratio of outstanding loans from commercial banks to the Gross Domestic Product (GDP) and outstanding deposits with commercial banks to the GDP.

The research on the link between financial inclusion (FI) and environmental quality is still in its early stages. With rising CO₂ emissions, many researchers have begun focusing on this topic and investigating its impact.

FI may impact CO₂ emissions through three channels: consumption, economic, and technological. In the case of the consumption channel, a well-developed financial system facilitates access to consumer credits, which increases energy demand by allowing individuals to purchase and use more energy-intensive home appliances and automobiles [23], which can result in CO₂ emissions from increased energy consumption. Similarly, FI facilitates access to low-cost funds, allowing businesses to purchase more machinery and equipment, construct new factories, and increase CO₂ emissions through increased energy use [24].

Technology is widely researched to reduce CO₂ emissions. Financial progress improves the environment by enabling investments in renewable energy sources and supporting eco-friendly projects that drive innovation and technology advancements to lower energy consumption and carbon emissions. The study by Li et al. [25] highlights the significance of employing green energy solutions in reducing carbon footprints after examining the effects of energy structure and the digital economy on CO₂ emissions in China, demonstrating how financial development plays a role in moderating CO₂ emissions effectively.

In their study, Grossman & Krueger [26] highlight the relationship between economic progress and environmental quality, known as the environmental Kuznets curve (EKC) concept. They suggest that although initial economic advancements may impact the environment, subsequent growth could lead to enhancements in environmental standards. In the publication by Wang et al. [27], the authors delve into the topic of the EKC, underscoring the importance of verifying its validity when formulating policies that strike a balance between preserving the environment and fostering economic development. Their research shows that the environmental impact increases as the economy grows, especially when natural resources are heavily exploited.

Since the global financial crisis, policymakers, scholars, and decision-makers have focused on how financial progress influences economic growth. Some scholars argue that financial progress is essential for development and technological innovations, but it can also substantially impact the environment’s quality, especially concerning carbon emissions [28]. However, scholars’ opinions remain divided as the data do not provide answers, making the study on the association inconclusive.

Recent academic studies have highlighted the relationship between financial inclusion and economic growth and their environmental effects. The research suggests that expanding services and manufacturing operations could lead to higher carbon dioxide (CO₂) emissions, exacerbating climate change issues [4]. Based on their worldwide analysis, economic development and structure play a role in influencing carbon emissions positively or negatively. Le et al. [29] research on the impact of financial inclusion, renewable energy, and globalisation on South Asian CO₂ emissions suggests that incorporating green energy sources is a crucial strategy for reducing South Asian CO₂ emissions; their result shows that financial inclusion caused an increase in CO₂. This finding is valid even amid ongoing economic expansion and rising energy usage, leading to higher carbon emissions.

Conversely, some scholars suggest an inverse correlation between financial inclusion and CO₂ emissions. According to Yang et al. [30], financial inclusion plays a role in fighting climate change by boosting carbon sequestration capacity and enhancing industrial structure through excellent economic support.

A study by Ben Jebli & Hakimi [31] found that when financial inclusion is paired with renewable energy consumption in developed countries, it reduces carbon emissions over the long term in technologically advanced countries. Furthermore, research by Damrah et al. [32] demonstrated that implementing financial inclusion measures in oil-exporter nations has decreased CO₂ emissions. Baulch et al. [5] found that financial constraints such as capital access and lack of government support are obstacles to adopting solar home systems in Ho Chi Minh City, Vietnam.

The literature synthesis highlights the intricate and varied correlation between financial inclusion and CO₂ emissions, economic growth, human capital, consumption patterns, and technological advancements, including many factors. The relationship between financial inclusion and CO₂ emissions is vital because of its potential to exacerbate or mitigate environmental degradation. Financial inclusion is an essential but overlooked factor in environmental studies because of the challenges that are associated with the complexity of its direct and indirect effects, data limitations, and the lack of a unified framework to capture its multifaceted influence across diverse regions, emphasising the importance of implementing a well-rounded approach to attaining a balanced harmony for sustainable progress; this strategy should consider the implications for both the economy and the environment.

To this end, it is imperative to address the existing barriers so that appropriate policies for sustainable development can be formulated, especially in Sub-Saharan Africa. This situation justifies the need for more research to comprehensively address all aspects and methodologies, ensuring meaningful policy implications.

The relationship between financial inclusion and environmental quality is relatively limited and recent. The environment Kuznets curve (EKC) theory, the Investment-Saving and Liquidity Preference – Money Supply (IS-LM), and the Energy Ladder Hypothesis (ELH) explain this relationship.

The EKC framework explains the relationship between environmental degradation and economic development. Lower income levels are more susceptible to environmental issues when the economy grows because they focus more on meeting their basic needs than addressing environmental concerns. However, a debate suggests that environmental degradation decreases as the economy grows and income levels increase. According to Grossman and Krueger (1991), it is commonly observed that individuals with higher incomes tend to embrace eco-friendly technology and prioritise environmental conservation. The EKC theory is influenced by how financial inclusion affects income distribution levels [26]. This is observed in the accessibility and affordability of green finance and innovation [33]. Various studies support the EKC theory by suggesting that as developing economies strive for growth, they may relax trade and environmental regulations. This can lead to an expansion in economic size and alter its structure, influencing manufacturing practices [34]. Additionally, financial advancement can impact the EKC by stimulating and fostering growth through higher production and energy consumption levels. Financial development can fund energy-intensive consumption products, increasing environmental emissions. In contrast, it may also encourage a green environment by offering green loans and financial support for energy-efficient projects [27].

The Investment-Saving and Liquidity Preference-Money Supply (IS-LM) theory was developed by Hicks in 1937 as a visual representation of John Maynard Keynes’s significant work “The General Theory of Employment” from 1936 [35]. This model offers insights into how an economy’s money and goods markets operate and is also referred to as the Hicks- Hansen model and Money theory. The theory has three critical components: liquidity, investment, and consumption; it posits that liquidity is determined by the money supply (M) and velocity of money supply (VT). At the same time, the level of investment and consumption reflects individual actors [36]. The IS curve illustrates the causal relationship between interest rates, planned investment, national income, and output. In contrast, the LM curve represents the equilibrium level of the money exchange for various combinations of interest rates and real income.

The IS-LM theory of interaction with the environment is effective because a well-developed financial system aims to increase the money supply, lower interest rates, and promote investment and consumption by providing access to funds through credits [36]. The advancement of finance often leads to improved access to credit for individuals [37] and drives up the demand for energy-intensive goods and services such as cars, household items, and real estate [38]. As a result of this trend, in consumption patterns driven by financial inclusion, CO₂ emissions tend to rise. Businesses and industries benefit from it by enhancing their productivity and growth through marketing strategies and adopting new production lines [39]. It often involves acquiring energy-intensive machinery [40].

The critical advantage of financial inclusion advancement is that it encourages essential investment activities for a thriving stock market. Companies listed on the market can access funds, increase investment, engage in economic activities, and increase production and consumption. These factors positively and negatively affect environmental quality [41].

Another relevant theory in context is the theoretical theory known as the Energy Ladder Hypothesis (ELH), which proposes a hierarchical advancement of energy sources that communities and individuals have employed throughout history. The theory suggests a relationship between income levels and energy usage patterns among people in society. As their earnings rise steadily over time, they tend to transition from using traditional biomass resources fuel (such as wood and dung) to more modern commercial fuels, such as electricity and gas, for their energy needs. It is worth mentioning that switching to alternative energy options poses challenges due to factors that play a role in determining the preferred mix of energy sources, including factors such as availability, ease of access, pricing, dependability, convenience and personal preferences. Financial inclusion shapes the energy ladder by providing affordable and dependable energy options for individuals facing financial constraints.

Based on the literature synthesis, the study adopts the conceptual framework shown in Figure 1 to investigate the impact of financial inclusion on carbon emissions and the mechanism of household consumption. This framework will consider both dimensions of financial inclusion, relevant control variables and the mechanism variable.

Figure 1.

Conceptual framework

The study will adopt the two-way fixed effects model using the reghdfe command in Stata software (https://www.stata.com). This sophisticated estimator controls both entity and time-fixed effects to tackle unobserved heterogeneity. The model includes fixed effects for each cross-sectional unit and each period, accounting for time-invariant characteristics and common shocks that could bias the results. In addition, this method can accommodate heterogeneous panels characterised by varying periods and cross-sectional units. Furthermore, it employs statistical tests to assess the validity of the fixed effects and the consistency of the estimator. The first step of the study is the Hausman test, which will influence the choice between Fixed Effects and Random Effects models [42]; it is commonly used as a way of choosing between fixed and random effects, see Gujarati & Sangeetha, 2007 [43]. The study proceeds with the fixed effect model as detected by Hausman following the work of [44]. The next step is employing the two-way fixed estimation while accounting for both time and country’s fixed effect, followed by the instrumental variable method to account for endogeneity problems. Further analysis includes robustness tests and heterogeneity analysis. All these methods were employed to confirm the study’s outcome and make statistical inferences.

This paper constructs the following fixed-effect model to examine the impact of financial inclusion on Carbon Emission in SSA countries:

${\text{LnCO}}_{\text{2}}{}_{it}=\alpha +{\beta }_1{\text{LnFi}}_{it}+{\beta }_2{\text{Lninc}}_{it}+{\beta }_3{\text{Lnto}}_{it}+{\beta }_4{\text{Lnei}}_{it}+{\beta }_5{\text{Lnind}}_{it}+{\beta }_6{\text{LnP}}_{it}+{\beta }_7{\text{LnFDI}}_{it}+{\theta }_i+{\delta }_t+{\epsilon }_{it}$ (1)

Where subscripts i and t denote country and year, respectively, and LnCO₂ represents total CO₂ emissions. The core explanatory variable Fi is the financial inclusion index contracted using Principal Component Analysis (PCA); Lninc, Lnto, Lnei, Lnind, LnP, and LnFDI represent the control variables in this study; θ_i is country-fixed effects, δ_t is time-fixed effects, and ε_it is an unobserved error term.

The study uses panel data from 44 Sub-Saharan African countries from 2004 to 2022 and adopts nine variables described in Table 1.

Table 2 highlights the proxies used to construct the Fi composite index, as done by [29]. Financial inclusion is the most important regressor and includes a combination of various proxy measures to measure its multidimensional nature accurately. Therefore, this study adopts the PCA approach to construct a composite index for financial inclusion.

Table 4 shows that the maximum correlation coefficient value among independent variables is –0.5457, below the value of ± 0.7, generally used as the rule of thumb for high correlation. It suggests that neither collinearity nor multicollinearity is considered a concern in our data.

As shown in Table 5, the Variance Inflation Factor (VIF) further affirms the result because it is performed to test collinearity between our variables. The maximum VIF is 3.96, while the mean VIF is 2.43, much lower than the threshold value 10 generally used in academic literature, confirming that our model does not suffer from collinearity or multicollinearity problems. All variables in the econometric model for this research show a significant positive correlation with the dependent variable, where Fi has the highest correlation with LnCO₂.

The study’s primary goal is to investigate how financial inclusion impacts carbon emissions in Saharan Africa utilising the two-way fixed effect model pioneered by Mundlak in 1978 [45] based on the data sample provided.

Presented in Table 6 are the results from eq. (1); columns (1) and (2) display the outcomes of the fixed and random effects, respectively, employed for the Hausman test. With a p-value of 0.000, the null hypothesis is rejected in favour of the fixed effect model under the Hausman test. It is evident that, observing Columns (3)–(5), all coefficients of the financial inclusion index used in the study consistently showed positive and statistically significant effects on carbon emissions at a significance level of 1%. The correlation between financial inclusion and Carbon Emission suggests that more significant financial activity is linked to higher levels of atmospheric carbon emissions.

The regression analysis in columns (4) and (5) demonstrates statistically significant correlations between several economic variables and CO₂ emissions. In column (4), a statistically significant positive relationship exists between Economic Growth (Lninc) and emissions at a 1% significance level. This suggests that wealthier countries produce CO₂ emissions on a larger scale than others with lower incomes. The coefficient of trade openness (Lnto) has a statistically negligible negative impact. However, the coefficients of energy intensity (Lnei) and industrial output (Lnind) positively increase emissions at a significance level of 1%. This highlights the impacts associated with energy consumption and industrial activities. The population size (LnP) has a negative but statistically insignificant correlation. In contrast, foreign direct investment (LnFDI) has a modest but statistically significant negative impact at a significance level of 5%, suggesting that FDI might help reduce emissions. Including fixed effects for both country and year in column (5) enhances the precision of existing correlations. The association between the financial inclusion index and carbon emissions remains positive but with a reduced magnitude of impact. Results show a notable positive correlation between trade openness and emissions, suggesting that countries with higher levels of openness might experience increased emissions rates. While the impact of energy intensity (Lnei) and industrial production (Lnind) has decreased somewhat in magnitude by incorporating the county and year-fixed effect in column (5), they still play roles in driving CO₂ emissions; the correlation between population size (LnP) and emissions levels is positive, suggesting that larger populations are linked to higher emissions levels (all results at a significance level of 1%). The influence of Foreign Direct Investment (FDI) becomes insignificant. At the same time, the high R squared value in column (5) indicates that the model adequately accounts for a considerable amount of the variability in CO₂ emissions. This emphasises the need to consider country-specific and time-specific elements when analysing the influence of financial inclusion on the region’s carbon emissions while maintaining significant control variables.

While the two-way fixed effects model effectively accounts for unobserved heterogeneity across nations over time, endogeneity issues can still be affected due to biases arising from omitted variable bias and measurement error. Even though the empirical approach has managed to consider factors affecting a country’s carbon emissions accurately, a chance of bias remains due to missing variables that stem from unobservable elements. The study addressed these problems using the instrumental variable (IV) method. In particular, the financial inclusion index (Fi) lag was utilised as an instrument. By using the exogenous variance in the lagged financial inclusion index, which is associated with the present financial inclusion index but uncorrelated with the error term, this approach helps to reduce endogeneity. The results obtained from the instrumental variable (IV) estimation in column (2) of Table 7 are consistent with those obtained from the two-way fixed effects model. This adds weight to our conclusions and confirms the significant impact of financial inclusion on carbon emissions.

This section discusses the approaches to testing the reliability of the results of the presented research.

This part of the study delves into the connection between inclusion and its influence on carbon emissions via the role of household consumption.

Financial inclusion enhances the availability of services and facilitates access to credit and other financial offerings for households and businesses. Due to improved access to finances, households and firms may boost their consumption levels, ultimately driving up the demand for products and services, increasing production activities and leading to higher carbon emissions. Consumption is measured by the natural logarithm of the Country’s Households and NPISHs’ final consumption expenditure per capita, denoted as LnCon.

Table 10 presents the relationship between Financial Inclusion and CO₂ emissions, evaluated with consumption as a mediating variable. The analysis was conducted in three stages: First column (1), the direct impact of financial inclusion on CO₂ emissions was evaluated; second column (2), the effect of financial inclusion on consumption; and third column (3), estimating the relationship of both financial inclusion and consumption on CO₂ emissions. The research findings revealed financial inclusion positively and significantly impacts consumption and CO₂ emissions. In addition, consumption was also discovered to mediate the relationship between financial inclusions, with approximately 12.5% of the effect being explained through consumption. It indicates that financial inclusion directly impacts CO₂ emissions, but some of the effects are through increased consumption.

The pathway through which financial inclusion impacts carbon emissions is better understood by linking these two analyses. The advancement of inclusion provides real income to households and businesses by easing financial restrictions and enhancing spending. Nevertheless, this increased spending results in heightened carbon emissions, underlining the importance of implementing policies harmonising progress with environmental conservation.

This part of the study discusses the heterogeneity of financial ecosystems of SSA countries and its impact on the level of countries’ carbon emissions.