Theoretically, there are good reasons to expect that globalization can affect access to clean fuels and technologies for cooking either directly or indirectly. Directly, globalization can affect access to clean fuels and technologies for cooking, on the one hand, through technologies and knowledge transfer. The literature suggests that globalization increases knowledge sharing. On the other hand, globalization improves the expansion of investment in cleaner energy. Indirectly, globalization affects access to clean cooking fuels through its effects on income, economic growth, ICT and financial development. A large body of literature highlights that globalization positively affects income by improving remittances. A lack of income reduces the ability of households to be able to access the clean cooking fuels. Several studies showed that remittances have a positive impact on households income in recipient countries ( De & Ratha, 2012 ). The resulting increase income from remittances alleviates energy poverty and allows households to have access to modern energy sources ( Barkat et al., 2023 ; Djeunankan et al., 2023 ). The second channel through which globalization affects access to clean cooking fuels is through economic growth. Indeed, as argued by Acheampong (2023) , a lack of access to financial service affects access to clean fuels. The last channel is relative to ICT, ICT adoption improves cooking fuels transition from traditional dirty cooking fuels to modern clean cooking fuels. In the same vein, ICT is an important source of information concerning both the disadvantages of using dirty cooking fuel, and the advantages of using modern clean cooking fuels ( Kumar & Igdalsky, 2019 ; Acharya & Marhold, 2019 ; Murshed, 2020a ).

Draws from the existing literature, two sub-groups of the determinants of cooking fuel choices can be identified. The first category, investigated microeconomic determinants of cooking fuel choices such as economic status ( Alem et al., 2016 ), urbanization ( McLean et al., 2019b ), electrification ( Gupta & Pelli, 2021 ), price of alternative energy sources ( Waleed & Mirza, 2022 ), education ( Hou et al., 2017 ; Paudel et al., 2018 ; Liao et al., 2019 ; Gould et al., 2020 ; Zhu et al., 2022 ), information and communication technologies (ICT) ( Acharya & Marhold, 2019 ), household income, household size, fuel price, gender of the household head, and ethnic differences ( Alem et al., 2016 ; Liao et al., 2019 ; Pallegedara et al., 2021 ; Shari et al., 2022 ; Ma et al., 2022 ).

The second category focused on macroeconomic drivers. For instance, Murshed (2020b) estimates the effect of foreign direct investment (FDI) flows on access to clean cooking fuels and technologies in low- and middle-income countries of the world for the period 2000-2017. The analysis suggests that FDI improves access to clean cooking fuel and technology. More recently, Murshed (2022) highlighted the drivers of cooking fuel transition in low- and middle-income Sub-Saharan African countries for the period 2000-2016. The results suggest that energy efficiency, economic growth, environmental pollution, financial globalization, financial development, and women empowerment improve access to clean cooking fuel and technology.

Abba Yadou et al. (2023) highlight the effect of remittances on energy transition in African countries over the period 2000-2020. The empirical findings from 2SLS with instrumental variable suggest that remittances reduce energy transition in African countries. Acheampong et al. (2023b) documented that globalization contributes to improve access to clean cooking fuels and technologies for cooking in 43 Sub-Saharan African countries over the period 1990-2017. Acheampong (2023) investigates the effect of governance and credit on access and clean cooking technologies in 43 Sub-Saharan African countries over the period 2000-2017. The findings indicate that access to credit and governance variables do not facilitate clean cooking technologies usage. Onyeneke et al. (2023) have investigated the determinants of access to clean fuels and technologies for cooking in 38 African countries over the period 2000-2020. The outcomes from pooled mean group and dynamic fixed effect estimators reveal that rural population, particulate matter emission, and natural resources depletion significantly decreased access to clean cooking fuels and technologies in the long run while the gross domestic product (GDP) per capita significantly increased access to clean cooking fuels and technologies in the long run in Africa.

Acheampong et al. (2023a) investigated the effect of democracy and governance on rural electrification and rural access to clean fuels and technologies for cooking in 34 Latin America and the Caribbean countries over the period 2000-2020. The outcomes highlight that governance improves rural electrification and rural access to clean cooking fuels and technologies, while democracy of different forms limits rural electrification and rural access to clean cooking fuels and technologies. Murshed (2023) highlighted the determinants of cooking fuel accessibility divide across urban and rural areas in 14 Latin American and Caribbean countries over the period between 2000 and 2020, the outcomes demonstrate that income inequality aggravates the urban-rural inequality in clean cooking fuel accessibility.

In the light of the empirical review, we have two mains testable hypotheses.

H1: Globalization improves access to access to clean fuels and technologies for cooking.

H2: Difference in globalization account for variations in the adoption of clean cooking technologies among urban and rural area.

By putting together the concerns of the two sub-groups, it becomes obvious that some important determinants of household’s access to clean cooking fuels and technologies have been neglected. This is particularly the case of potential drivers such as environmental degradation, and the effect globalization in adopting clean cooking fuel in rural and urban areas.

The aim of this paper is to investigate the effect of globalization on access to clean fuels and technologies for cooking in 31 African countries over the period between 2000 and 2020. The periodicity of the investigation and the sample size was constrained by the availability of data and especially of the dependent variables.

Due to missing data, especially for clean cooking fuels transition, the analysis is restricted to a balanced panel of 31 African countries with data spanning 2000-2020. This is the longest period possible and the maximum number of countries. Data are collected from KOF Swiss Economic Institute ( Dreher, 2006 ), the World Bank: World Development Indicators of World Bank (2022) . Data availability dictates the sample and periodicity. The list of countries is provided in appendix ( Table A1 ). Descriptive statistics of the variables are provided in Table 1 .

Following recent literature, we use two indicators to assess the dependent variable. The first measure is the percentage of the total population with access to clean fuels and technologies for cooking from national grids. The second indicators are the percentages of the urban and rural populations with access to clean fuels and technologies for cooking, respectively.

The independent variable of interest is globalization, which is measured by KOF globalization index. The globalization index is based on four indices of globalization, including overall globalization, economic globalization, social globalization and political globalization. Figure 1 presents a scatter plot correlation analysis between globalization and access to clean fuels and technologies for cooking in Africa.

The control variables in this study are chosen based on extant literature on the determinants of access to clean fuels and technologies for cooking. We employ GDP per capita, ICT proxied by fixed telephone, personal remittances received, environmental degradation proxied by CO₂ emissions and financial development to determine their impact on access to clean cooking fuels. GDP per capita is expected to positively affect energy transition ( Onyeneke et al., 2023 ). ICT is expected to positively influence energy transition, provided that digital technologies facilitate access to clean fuels and technologies for cooking. Remittances in energy transition literature is considered as a means through which household improve their income ( Abba Yadou et al., 2023 ). Environmental degradation is expected to positively influence the adoption of clean cooking fuel ( Murshed, 2022 ), financial development is expected to positively influence energy transition ( Nguyen et al., 2021 ). The characteristics of the variables in terms of their mean, standard deviation, minimum and maximum values are presented in Table 1 of descriptive statistics.

Source: Computation using Stata software.

Table 1 shows the descriptive statistics whereas Table 2 shows the pairwise correlations between the variables used in the paper. As shown in Table 2 , there is a significant positive association between different dimension of the globalization and clean cooking fuels transition. Figure 1 plots the relationship between globalization dimensions, and access to clean fuels and technologies for cooking. As we can see, there is a positive relationship between these two variables. This means that an increase in globalization is associated with an increase in clean cooking fuels transition. However, as correlation does not mean causality. This relationship is investigated empirically in the next section.

In order to explain the effects of globalization on access to clean cooking fuels in African countries, the paper considers a general mathematical equation. This is presented in Equation (1):

${\text{ACFT}}_{\text{it}}=f\left({\text{GLOB}}_{\text{it}},{\text{GDP}}_{\text{it}},{\text{ICT}}_{\text{it}},{\text{PRR}}_{\text{it}},{\text{CO2}}_{\text{it}},{\text{PRIVATE}}_{\text{it}}\right)$ (1)

${\text{ACFT}}_{\text{it}}$ is access to clean fuels and technologies for cooking in country i at the period t. ${\text{GLOB}}_{\text{it}}$ is globalization index in country i at period t. ${\text{GDP}}_{\text{it}}$ is the economic growth in country i at the period t. ${\text{ICT}}_{\text{it}}$ is information and communications technology proxied by fixed telephone subscriptions in country i at period t. ${\text{PRR}}_{\text{it}}$ is personal remittances received by country i at period t, ${\text{CO2}}_{\text{it}}$ is environmental degradation proxied by CO₂ is emissions, ${\text{PRIVATE}}_{\text{it}}$ is financial development proxied by domestic credit to private sector (% of GDP).

Disaggregating ${\text{GLOB}}_{\text{it}}$ into overall globalization index ( ${\text{KOFGI}}_{\text{it}}$), economic globalization index ( ${\text{KOFEcGI}}_{\text{it}}$), social globalization index ( ${\text{KOFSoGI}}_{\text{it}}$), political globalization index ( ${\text{KOFPoGI}}_{\text{it}}$). To avoid heteroskedasticity and produce consistent and efficient results all variables are transformed into natural logarithms ( Wang & Dong, 2019 ). Thus, we have the following equations:

$\begin{array}{c}{\text{LnACFT}}_{\text{it}}={\text{δ}}_0+{\text{δ}}_1{\text{LnKOFGI}}_{\text{it}}+{\text{δ}}_2{\text{LnGDP}}_{\text{it}}+{\text{δ}}_3{\text{LnICT}}_{\text{it}}+{\text{δ}}_4{\text{LnPRR}}_{\text{it}}\\ +{\text{δ}}_5{\text{LnCO2}}_{\text{it}}+{\text{δ}}_6{\text{LnPRIVATE}}_{\text{it}}+{\text{ε}}_{\text{1it}}\end{array}$ (2)

$\begin{array}{c}{\text{LnACFT}}_{\text{it}}={\text{β}}_{\text{0}}+{\text{β}}_{\text{1}}{\text{LnKOFEcGI}}_{\text{it}}+{\text{β}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{β}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{β}}_{\text{4}}{\text{LnPRR}}_{\text{it}}\\ +{\text{β}}_{\text{5}}{\text{LnCO2}}_{\text{it}}+{\text{β}}_{\text{6}}{\text{LnPRIVATE}}_{\text{it}}+{\text{ε}}_{\text{2it}}\end{array}$ (3)

$\begin{array}{c}{\text{LnACFT}}_{\text{it}}={\text{φ}}_{\text{0}}+{\text{φ}}_{\text{1}}{\text{LnKOFSoGI}}_{\text{it}}+{\text{φ}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{φ}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{φ}}_{\text{4}}{\text{LnPRR}}_{\text{it}}\\ +{\text{φ}}_{\text{5}}{\text{LnCO2}}_{\text{it}}+{\text{φ}}_{\text{6}}{\text{LnPRIVATE}}_{\text{it}}+{\text{ε}}_{\text{3it}}\end{array}$ (4)

$\begin{array}{c}{\text{LnACFT}}_{\text{it}}={\text{η}}_{\text{0}}+{\text{η}}_{\text{1}}{\text{LnKOFPoGI}}_{\text{it}}+{\text{η}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{η}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{η}}_{\text{4}}{\text{LnPRR}}_{\text{it}}\\ +{\text{η}}_{\text{5}}{\text{LnCO2}}_{\text{it}}+{\text{η}}_{\text{6}}{\text{LnPRIVATE}}_{\text{it}}+{\text{ε}}_{\text{4it}}\end{array}$ (5)

To address the endogeneity, Equations (1), (2), (3), (4) and (5) are estimated using instrumental-variable (IV) two-stage least squares (2SLS) approach (IV-2SLS).

Source: computation using Stata software.

The baseline results in Table 3 present outcomes corresponding to the effect of overall globalization, economic globalization, social globalization and political globalization on access to clean fuels and technologies for cooking.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

We find that globalization has a positive effect on access to clean fuels and technologies for cooking in African countries. These results mean that overall globalization, economic globalization, social globalization and political globalization improve access to clean cooking fuels. The results suggest that an increase in overall globalization, economic globalization, social globalization and political globalization by 1 unit leads to an increase in the indicator of access to clean cooking fuels by 3.733, 3.991, 3.153 and 1.398 units respectively. These results can be explained by the fact that globalization is associated with knowledge and technologies transfer ( Doytch & Uctum, 2016 ). The outcome matches those of Acheampong et al. (2023b) for 43 Sub-Saharan African countries.

Regarding the control variables, we find that they all have the expected signs. Specifically, we find that GDP per capita, ICT, remittances, environmental degradation and financial development have a positive and statistically significant effect on access to clean fuels and technologies for cooking. In other words, high levels of economic growth, ICT, remittances, environmental degradation and a well-developed financial system are associated with high access to clean technologies for cooking in Africa. These results are similar to those of Murshed (2022) , showing that economic growth has a positive impact on access to clean fuels and technologies for cooking in Sub-Saharan African countries. In addition, the outcomes reveal that improving technologies remains imperative for clean cooking fuels technology in Africa. These results support the findings of Acharya and Marhold (2019) concerning cooking fuels transition in Nepal. The coefficient associated with the remittances is positive and statistically significant. This implies that an increase in remittances leads to an improvement in clean cooking technology, which is true for African countries, according to the World Bank, Africa is one of the world’s largest recipient of remittances. That contributes to increase income in households ( Pan et al., 2020 ), improve human capital ( Gyimah-Brempong & Asiedu, 2015 ), reduce poverty ( Masron & Subramaniam, 2018 ; Aloui & Maktouf, 2021 ), and increase financial inclusion ( Abba et al., 2021 ; Eggoh & Bangake, 2021 ). Our results corroborate with the finding of Hosan et al. (2022) who showed that remittance inflows reduce energy poverty in Bangladesh, Barkat et al. (2023) who disclosed that remittances contribute to alleviate energy poverty in low and middle-income countries, Djeunankan et al. (2023) , who confirm the same for developing countries, and Abba Yadou et al. (2023) for African countries.

Environmental degradation improves access to clean cooking fuel transition. These outcomes mean that environmental degradation improves access to clean cooking fuels technologies. It is established that unclean cooking fuels increase CO₂ emissions. As a result, higher CO₂ emissions influence the production of clean cooking fuels. In addition, environmental degradation may also be correlated to a high environmental regulation, which at least to a certain extent contributes to ameliorating the access to clean cooking fuels. The result corroborates the findings of Murshed (2022) , who documented that environmental degradation improves access to clean cooking fuel in Sub-Saharan Africa.

Rural residents in developing countries, still rely heavily on solid fuels ( Aemro et al., 2021 ). To model that whether or not the place of residence matter in the access of household to clean technologies for cooking. We disaggregate ${\text{ACFT}}_{\text{it}}$ into rural ( ${\text{ACFTR}}_{\text{it}}$) and urban ( ${\text{ACFTU}}_{\text{it}}$) access to clean fuels and technologies for cooking.

${\text{LnACFTR}}_{\text{it}}={\text{η}}_{\text{0}}+{\text{η}}_{\text{1}}{\text{LnOGI}}_{\text{it}}+{\text{η}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{η}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{η}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{1it}}$ (6)

${\text{LnACFTR}}_{\text{it}}={\text{α}}_{\text{0}}+{\text{α}}_{\text{1}}{\text{LnEGI}}_{\text{it}}+{\text{α}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{α}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{α}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{2it}}$ (7)

${\text{LnACFTR}}_{\text{it}}={\text{λ}}_{\text{0}}+{\text{λ}}_{\text{1}}{\text{LnSGI}}_{\text{it}}+{\text{λ}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{λ}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{λ}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{3it}}$ (8)

${\text{LnACFTR}}_{\text{it}}={\text{β}}_{\text{0}}+{\text{β}}_{\text{1}}{\text{LnPGI}}_{\text{it}}+{\text{β}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{β}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{β}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{4it}}$ (9)

${\text{LnACFTU}}_{\text{it}}={\text{δ}}_{\text{0}}+{\text{δ}}_{\text{1}}{\text{LnOGI}}_{\text{it}}+{\text{δ}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{δ}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{δ}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{1it}}$ (10)

${\text{LnACFTU}}_{\text{it}}={\text{ω}}_{\text{0}}+{\text{ω}}_{\text{1}}{\text{LnEGI}}_{\text{it}}+{\text{ω}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{ω}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{ω}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{2it}}$ (11)

${\text{LnACFTU}}_{\text{it}}={\text{ξ}}_{\text{0}}+{\text{ξ}}_{\text{1}}{\text{LnSGI}}_{\text{it}}+{\text{ξ}}_{\text{2}}{\text{LnGDP}}_{\text{it}}+{\text{ξ}}_{\text{3}}{\text{LnICT}}_{\text{it}}+{\text{ξ}}_{\text{4}}{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{3it}}$ (12)

${\text{LnACFTU}}_{\text{it}}={\vartheta }_0+{\vartheta }_1{\text{LnPGI}}_{\text{it}}+{\vartheta }_2{\text{LnGDP}}_{\text{it}}+{\vartheta }_3{\text{LnICT}}_{\text{it}}+{\vartheta }_4{\text{LnPRR}}_{\text{it}}+{\text{ε}}_{\text{4it}}$ (13)

The results from the panel data regression analyses for Equations (6), (7), (8), (9), (10), (11), (12), and (13) are provided in Table 4 . The effect of overall globalization, economic and social globalizations on clean cooking fuels technologies in rural area is non-significant. The result is also non-significant for economic globalization in urban area.

The above results are carried out on Africa as a sample. However, according to Table A2 there is a difference in the proportion of the population with access to clean cooking fuels across geographical regions of the continent. For instance, in 2020, only 17% of the population had access to clean cooking fuels in Sub-Saharan Africa, while in North Africa we have about 95%.

The outcomes are discussed and presented in Table 5 . The results reveal some disparities at the sub-regional level. Globalization positively and significantly determines access to clean cooking fuels in Sub-Saharan Africa and North Africa. This result corroborates the findings of Acheampong et al. (2023b) .

The KOF index consists of three dimensions (economic, social, and political) and specifies amid between de jure and de facto aspects; together with its subindexes, which allows for the shedding of more light on the correlation between globalization and economic sophistication. Furthermore, the use of each dimension and/or aspect of the spread of globalization provides different results ( Aluko et al., 2021 ; Fotio & Nguea, 2022 ). Table 5 reports the results, showing the positive effect of globalization on access to clean cooking technologies.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

Table 6 above offers a comprehensive analysis with instrumental variables for estimating the impact of globalization on access to clean cooking fuel (lnacft) using eight varying model specifications. The findings indicate considerable disparities between de facto and de jure measures of globalization that have profound implications for energy policy in the African continent. The economic aspect of de facto globalization (lnkofecgidf) has the largest effect (2.974), suggesting trade and foreign direct investment as primary drivers for improved clean cooking access. The social (2.313) and political (0.966) aspects have smaller yet significant effects. Most striking is that all de jure globalization measures have larger coefficients than de facto measures, and the overall de jure index (lnkofidj) is even 4.387.

Table 7 and Table 8 present the results for de facto versus de jure globalisation in rural Africa and urban Africa, respectively.

Table 7 presents the findings of eight instrumental variable regression specifications that examine the differential impacts of dimensions of globalization on access to clean cooking fuels (lnacftr) in African rural areas. The findings present several significant results that require close interpretation.

The economic dimension of de facto globalization (lnkofecgidf) shows a statistically significant positive impact (2.129), though much smaller than the overall de facto index (lnkofgidf at 3.440). Surprisingly, the social dimension of de facto globalization (lnkofsogidf) yields an insignificant statistic (0.854), whereas its political dimension (lnkofpogidf) exerts a significant, albeit small, impact (1.043). This is a pattern that suggests that in rural areas, economic integration through foreign trade and investment flows can have an even more dominant role in determining access to clean cooking than social or political integration. The de jure indicators exhibit a more nuanced scenario. The economic aspect (lnkofecgidj) shows an extraordinarily strong impact (5.608), nearly triple its de facto counterpart. The political aspect (lnkofpogidj) still remains significant (2.050), while the social aspect (lnkofsogidj) has only borderline significance (1.657). The overall de jure index (lnkofidj at 4.866) supports the general trend that policy-driven globalization measures have larger coefficients than their de facto versions.

Table 8 presents instrumental variable regression results examining the effects of de facto versus de jure globalization on clean cooking fuel access (lnacftu) across urban Africa. The analysis reveals several key insights that differ meaningfully from rural patterns, highlighting the importance of context-specific policy approaches.

The economic dimension of de facto globalization (lnkofecgidf) shows a strong positive effect (3.093), significantly larger than the overall de facto index (lnkofgidf at 1.003). This suggests that in urban areas, economic integration through trade and investment flows drives clean cooking access more than composite globalization measures. The social dimension (lnkofsogidf at 2.797) also demonstrates substantial impact, while the political dimension (lnkofpogidf at 0.311) appears relatively weak. This pattern implies urban populations benefit more from economic and social globalization than from political integration alone.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

Source: Computation using Stata software. Robust standard errors in parentheses ***p < 0.01, **p < 0.05, *p < 0.1.

De jure measures reveal striking results. The social dimension (lnkofsogidj at 5.239) emerges as the strongest predictor overall, exceeding even economic de jure effects (4.600). This suggests formal policies promoting social integration—such as cultural exchanges, educational partnerships, or technology transfers—may yield exceptional returns in urban contexts. The political dimension remains modest (0.601), while the composite de jure index (1.348) underperforms its components, potentially indicating offsetting effects when dimensions combine.