Energy poverty, characterized by limited access to and affordability of clean energy, is a significant issue, especially in rural areas of developing countries. While the energy ladder hypothesis suggests that economic development leads to cleaner energy choices, empirical evidence shows a more complex relationship, highlighting the role of socio-cultural factors. Religious beliefs, as a fundamental aspect of culture, shape social norms and individual behavior, potentially influencing energy choices. This study addresses the gap in understanding the impact of religious beliefs on energy poverty in rural China, where religious adherence is prevalent. Existing research on the topic yields conflicting conclusions, with some studies suggesting religious beliefs exacerbate energy poverty, while others suggest alleviation. This paper aims to clarify this discrepancy by rigorously examining the causal relationship between religious beliefs and energy poverty in rural China, focusing on both accessibility and affordability of clean energy. The study uses a large national sample and employs instrumental variables to address potential endogeneity issues. The findings contribute to the literature by providing micro-level evidence from rural areas, exploring potential mediating mechanisms (low income and education), examining the heterogeneous impacts of different religions, and utilizing robust econometric techniques.

The literature on energy poverty in China highlights the significant disparity between urban and rural areas, with rural regions facing challenges in both accessibility and affordability of clean energy. Studies show that even with increasing income, households may not fully transition to modern fuels, suggesting the influence of non-economic factors. Research on non-economic factors influencing energy poverty points to the roles of education, gender, ethnicity, and social capital. While extensive research explores the relationship between religiosity and environmental behavior, the impact of religious beliefs on energy poverty remains understudied. Existing studies focusing on this relationship offer conflicting conclusions: one suggesting exacerbation and another suggesting alleviation of energy poverty. The existing literature on religious beliefs and environmental behavior reveals varied and sometimes opposing effects depending on the region and context. This highlights the need for specific research focused on the unique socio-religious context of rural China.

This study utilizes data from three waves (2012, 2014, and 2016) of the China Labor-force Dynamics Survey (CLDS), a nationally representative household survey covering 29 provinces. The final sample comprises 13,773 rural adults after data cleaning. Two dependent variables are used to measure energy poverty: (1) a binary variable indicating whether a household uses solid fuels (firewood and coal) as the primary energy source (accessibility), and (2) a continuous variable representing energy expenditure divided by total income (affordability). The independent variable is a binary variable indicating whether an individual has religious beliefs. Low income and education are included as mediating variables. Control variables include demographic factors (gender, age, marital status, ethnicity), household characteristics (family size), Communist Party of China membership, and regional fixed effects. To address the potential endogeneity of religious beliefs, the study employs instrumental variables. The primary instrumental variable used is the density of places of worship (churches, temples, mosques) at the village/community level. This instrumental variable is justified based on its correlation with religious beliefs and its lack of direct impact on household energy consumption. A historical instrumental variable (number of temples at the provincial level in 1820) is also utilized as a robustness check. The study uses Probit and Tobit models for the accessibility and affordability measures respectively, along with their instrumental variable counterparts (IV-Probit and IV-Tobit). A mediation effect model (Baron and Kenny, 1986) is used to test the mediating effects of low income and education. Sensitivity analysis is performed using the method of Altonji et al. (2005) and Oster (2019) to assess potential bias from unobservable factors. Heterogeneity analysis is conducted to explore variations in the effects across different subgroups.

The baseline results (Table 1) show a statistically significant positive association between religious beliefs and energy poverty, affecting both accessibility and affordability. The instrumental variable estimations confirm this relationship, addressing the potential endogeneity concerns. The robustness checks (Table 2) using selection on observable factors suggest that the estimated effects are unlikely to be substantially affected by unobservable factors. The historical instrumental variable (Table 3) also supports the main findings. Further analysis shows that the effect of religious beliefs on energy poverty is more pronounced among males, ethnic minorities, and low-income individuals (Table 6). The mechanism analysis (Table 7) confirms that both low income and education mediate the relationship between religious beliefs and energy poverty. Low income is found to be a more significant mediator than education. Additional analysis (Table 8) demonstrates a positive relationship between the frequency of religious activities and energy poverty. Analysis of the three major religions (Buddhism, Protestantism, and Islam) reveals that Islam has the strongest positive association with accessibility energy poverty (Table 9).

The findings of this study confirm the hypothesis that religious beliefs exacerbate energy poverty in rural China. The significant positive association persists even after controlling for various factors and addressing endogeneity issues. The mediating roles of low income and education offer valuable insights into the underlying mechanisms. The greater effect of low income suggests that religious practices may divert resources from secular consumption, including energy-related expenses. The mediating effect of education highlights the importance of human capital in mitigating energy poverty. The heterogeneous effects across different religious groups suggest the need for tailored interventions. The findings challenge the simplistic view of economic development as the sole determinant of energy choices, highlighting the complex interplay of cultural and religious factors.

This study provides strong evidence that religious beliefs contribute to energy poverty in rural China, influencing both accessibility and affordability. Low income and education mediate this relationship, suggesting that interventions should focus on improving economic opportunities and educational attainment among religious communities. The findings highlight the need for culturally sensitive energy poverty alleviation strategies that account for the role of religion. Future research could explore the specific religious practices and beliefs associated with energy consumption patterns and investigate the effectiveness of interventions targeting the identified mediating mechanisms.

The study relies on self-reported data on religious beliefs and energy consumption, which might be subject to reporting biases. The instrumental variable approach, while mitigating endogeneity concerns, does not entirely eliminate the possibility of unobserved confounding factors. The study's focus on rural China limits the generalizability of the findings to other contexts. Future research should consider using more objective measures of energy consumption and exploring alternative instrumental variables.