I. Introduction
Higher levels of electricity consumption (EC) yield numerous socio-economic benefits. Access to electricity is considered a fundamental entitlement, crucial for improving health outcomes, enhancing educational attainment (Olanrele et al., 2020), and improving overall quality of life (Bridge et al., 2016). Recognizing its importance, governments worldwide have prioritized increasing electricity accessibility, leading to steady growth in access rates and per capita consumption over recent years. Despite these advances, a considerable segment of society remains without electricity access. Conversely, accessible and affordable financial services through financial inclusion (FI) have played a crucial role in poverty alleviation (Chibba, 2009; Lal, 2018), enhancing human development, and meeting basic human needs (Barik et al., 2022; Inoue, 2019; Nanda & Kaur, 2016). Research indicates that FI positively influences EC (Iqbal & Nawaz, 2021). While previous literature has largely examined the connection between FI and the consumption of renewable energy (Ababio et al., 2023; Bakhsh et al., 2023; Cui et al., 2022; Feng et al., 2022), the connection between FI and EC remains relatively unexplored. Addressing this gap, this paper aims to investigate the interplay between FI and EC across different countries.
Diverging from earlier literature, this study adds to the existing body of research in multiple ways. First, unlike prior research, this study develops a comprehensive FI index using six different proxies and empirically assesses its impact on EC. Second, given the differing levels of FI among the sampled countries, the study investigates how these differences influence EC among higher-middle-income (HMI) and lower-middle-income (LMI) nations separately, as well as for the entire sample. Third, the study employs various panel data models in its econometric analysis to enhance the robustness of the regression results.
II. Data and Methodology
A. Data
The primary objective of this study is to examine the impact of financial inclusion (FI) on per capita electricity consumption (EC) in both high-middle-income (HMI) and low-middle-income (LMI) countries. Data was collected from 31 countries (15 LMI and 16 HMI) spanning the period from 2004 to 2021. The analysis focuses on this timeframe due to data availability constraints beyond these years. Countries are classified by income level according to the World Bank’s classification for the fiscal year 2023.
B. Methodology
The key goal of this paper is to empirically investigate the effect of FI on per capita EC in HMI and LMI countries. To achieve this objective, the following econometric model is proposed:
\[\begin{aligned} {PCELEC}_{it} &= \alpha_{0} + \beta_{1}{FIINDEX}_{it}\\ & \quad + \beta_{2}{CTRL}_{it} + µ_{it} \end{aligned}\tag{1}\]
\[\begin{aligned} {PCELEC}_{it} &= \alpha_{0} + \beta_{1\ }{FIINDEX}_{it} + {\beta_{2}Enrol}_{it}\\ & \quad + \beta_{3}{UNEMP}_{it} + \beta_{4}{PPL\ Grow}_{it}\\ & \quad + \beta_{5}{FDI}_{it} + \beta_{6}{REMIT}_{it}\\ & \quad + {\beta_{7}GDPPC}_{it}\\ & \quad + \beta_{8}{INFLATION}_{it} + \beta_{9}{TGDP}_{it}\\ & \quad + \beta_{10}{InGDP}_{it} + µ_{it} \end{aligned} \tag{2}\]
In the equation above, the dependent variable, represents per capita EC. The explanatory variable, the FI Index, represents the composite index of FI, developed through principal component analysis (PCA) utilizing six distinct FI indicators. Along with these variables, the study incorporates control variables, which are outlined in Table 1.
C. Estimating Strategy
Initially, this study employs the basic Ordinary Least Square (OLS) model, followed by fixed-effects (FE) and also random-effects (RE) models to analyze the impact of FI on per capita EC. The selection of the final model is guided by the Hausman test, which favours the random-effects model. Given the panel data’s temporal and cross-sectional dimensions, concerns such as autocorrelation and heteroscedasticity may arise. To address such problems, this paper uses panel-corrected-standard-errors (PCSEs) and the feasible generalized-least-squares-method (FGLS) to strengthen the robustness of the findings. However, these methods may not fully address potential problems such as endogeneity or omitted variables. To address these concerns, this paper utilizes the Generalized-Method-of-Moments (GMM) for re-estimation, thereby further enhancing the analysis.
III. Empirical Findings
The dataset includes data from both HMI and LMI countries. Our analysis of this combined sample shows a statistically significant positive effect of FI on EC across all countries (see Table 2). This relationship suggests that increased FI improves financial accessibility, enhances economic conditions, and boosts buying capacity, which ultimately leads to a higher per capita EC and improved standards of living.
Upon examining the control variables, it is evident that Foreign Direct Investment (FDI), school enrollment, and unemployment are positively associated with EC across the countries studied. Higher FDI stimulates economic activity, leading to greater electricity demand. Similarly, increased school enrollment indicates more students engaged in learning activities, increasing electricity demand in educational institutions and households.
Though the connection between unemployment and EC seems counterintuitive, this apparent contradiction can happen due to several factors, such as unemployed individuals spending more time on household activities like lighting, heating, cooling, gaming, watching television, cooking, etc., which may increase the usage of household appliances.
Inflation and population growth show negative associations with per capita EC across the sample countries. The impact of inflation on EC is straightforward: higher inflation typically leads to increased electricity prices, reducing purchasing power and thus decreasing electricity demand. In contrast, the opposite relationship between population growth and EC can be attributed to factors such as energy efficiency measures, urbanization trends, policy interventions, technological advancements, changes in industrial practices, cultural shifts, and demographic changes.
For the remaining control variables, the observed effects vary depending on the specific econometric model used and the level of estimation precision achieved. Generally, variables such as Per Capita GDP (PCGDP), Trade GDP (TGDP), Industry GDP (InGDP), and Remittance Inflow are expected to positively influence EC. This expectation arises from their correlation with increased economic activity and enhanced purchasing power.
Table 3 highlights the effect of FI on EC in HMI countries. Similar to the findings for the aggregated sample, our analysis of HMI countries reveals a strong positive impact of FI on EC, statistically significant at the 1 percent level.
Regarding the control variables specific to HMI countries, school enrollment shows a positive but non-significant impact. This observation is primarily due to the higher median/average school enrollment rates in HMI countries, albeit with limited variation. Additionally, our analysis indicates that coefficients for unemployment and Foreign Direct Investment (FDI) are positively and significantly correlated, consistent with our findings in the aggregate sample.
Furthermore, population growth consistently presents negative and significant coefficients across the majority of econometric models used. This reaffirms the established relationship between population growth and EC.
Moreover, per capita GDP, as an indicator of economic development, exhibits a robust and significant positive impact on EC. It is noteworthy that while coefficients associated with inflation show positive trends, they often lack statistical significance in our analysis.
This section explores the linkage between FI and EC within the context of LMI countries. Our findings closely align with those observed in the aggregate and HMI country data. Specifically, we observe positive coefficients for the FI index, statistically significant at the 1 percent level.
Additionally, the outcomes related to control variables within LMI countries closely resemble those observed in HMI countries. Similar to HMI countries, we observe a positive association between unemployment and EC, while population growth exhibits a negative correlation with EC in LMI countries (see Table 4).
Furthermore, the positive and significant coefficient associated with remittances is particularly noteworthy. Remittances play a crucial role in lower-income economies by providing essential economic support, alleviating poverty, stimulating consumption and investment, and ultimately contributing to increased EC (see Table 4).
IV. Conclusion
This study explores the intricate relationship between FI and EC across both HMI and LMI countries. Using aggregate panel data from 31 countries spanning 2004 to 2021, this work empirically demonstrates that FI positively impacts per capita EC across all three groups of countries (Overall sample, HMI Countries, and LMI Countries). This paper recommends the development of targeted FI policies to enhance per capita EC and promote sustainable livelihoods in both HMI and LMI countries.