Household Energy Clean Transition Mechanisms under Market Failures: A Government Financing Perspective

Abstract

Clarifying the principles governing the transition to cleaner household energy is crucial for enhancing households’ access to clean and efficient energy, thereby enhancing households’ welfare and overall societal well-being. However, most existing energy transition theories are grounded in assumptions of perfect market functionality. This paradigm deviates from the reality of market failure and needs to comprehensively elucidate the process of transitioning to cleaner household energy. This study develops a framework for understanding the transition to cleaner household energy within the context of market failure. It investigates the effects and mechanisms of government funding and energy endowment on this transition, considering the accessibility and affordability of clean energy. The analysis is based on 20 years of data on rural energy consumption across 29 provinces in China. The findings reveal that government funding inputs significantly enhance the transition to cleaner household energy, facilitated by the clean energy supply effect, clean technology promotion effect, clean knowledge dissemination effect, and financial constraint alleviation effect. Heterogeneity analysis indicates that in regions abundant in renewable energy, government funding inputs can effectively stimulate the development and utilization of renewable energy sources, thereby enhancing the accessibility of clean energy and driving the transition to cleaner household energy. Finally, it is recommended that the market play a constructive role in the energy transition process in addition to increased government investment in energy infrastructure, extension services, and public education.

1. Introduction

One-third of the world’s population still relies on inefficient cookstoves to burn non-clean energy for household living, heating, and other energy needs [1]. This practice negatively impacts the health of household members, the labor supply, and overall economic well-being [2,3,4]. Non-clean household energy use contributes to indoor air pollution, heightening the health risks for household members [5,6]. According to the World Health Organization, indoor air pollution caused 3.2 million deaths in 2020, including more than 23,700 deaths of children under five years old. Additionally, the reliance on non-clean energy sources entails a considerable investment of time for households engaged in fuelwood collection and cooking activities. For instance, rural Indian households typically allocate around 1–2 h daily to collecting fuelwood and an additional 3 h to cooking. This expenditure of time illustrates the substantial burden imposed by reliance on non-clean energy on household routines and daily life [7]. Furthermore, the overconsumption of traditional biomass, such as fuelwood, exacerbates issues of biodiversity loss and ecosystem degradation in developing countries, thereby intensifying the effects of global warming [8]. Consequently, investigating the pathways and principles of a cleaner transition to domestic energy and promoting household access to and utilization of cleaner energy can mitigate the adverse effects of non-clean energy use and enhance both household and societal well-being [9].

Previous research has examined the impacts of various factors, including income, price, household characteristics, urbanization, and topographical features, on household energy consumption and transition [10,11,12,13]. Some studies focus on energy endowments and the effects of climate change. Resource endowment influences the type of renewable energy developed in a country or region [14], and the costs and benefits of different kinds of renewable energy elements are distinct and dynamic [15], with varying impacts on the energy mix. There is a complex relationship between climate change and household energy consumption. Additionally, these studies have developed theories such as the energy ladder theory and energy stacking theory [16,17]. Additionally, Su and Tan were the only researchers to examine the effects of regional energy transition in China through the lens of government support. They found that such support significantly facilitates energy transition, with the most significant impact observed in the central region and the smallest in the eastern region [18]. However, the study only analyzes the direct impact of government support on energy transition and lacks the test of the internal mechanism. This also indicates the necessity of an in-depth study of the mechanism of the impact of government funding on energy transition.

More critically, many of these studies operate under the assumption of perfect market development, which presupposes adequate supply and easy accessibility of clean energy to households. These studies often focus on whether households can afford clean energy at relatively high prices, as a critical factor influencing the clean energy transition in household livelihoods [19,20], namely the affordability of clean energy. This framework is more suited to elucidate energy transitions in households of developed countries or urban areas [16,21], where energy markets are well-established and clean energy, such as gas and electricity, is readily available. Nonetheless, applying this framework to comprehend energy transitions in developing countries with widespread market failures presents a significant challenge [12,22,23,24,25].

In rural areas with clean energy market failures, government intervention is a practical approach to achieve efficient rural clean energy allocation and enhance the accessibility and affordability of clean energy [18]. Government funding constitutes a primary method of such intervention. Firstly, the development cost of clean energy, such as coal and gas, is high. Moreover, the disparity between energy supply and demand necessitates laying transmission pipelines, a costly endeavor that users often find challenging to finance independently. Consequently, government funding becomes imperative in this scenario [26,27]. Secondly, technological barriers, such as clean technology and knowledge, significantly hinder the household use of clean energy. However, promoting professional technology services and public education can partially mitigate these barriers, enhancing the accessibility and affordability of clean energy [18]. While professional technical service promotion and public education can alleviate technical barriers [28,29], financial and personnel inputs rely equally on government funding support. Moreover, government funding inputs can alleviate financial constraints associated with energy infrastructure construction (e.g., base stations for hydroelectric power, wind power, etc.) and the acquisition of clean tools (e.g., energy-efficient stoves, air conditioners, etc.), thereby enhancing the accessibility and affordability of clean energy supply [30,31]. Additionally, regional disparities in energy endowments and economic conditions give rise to significant variations in rural energy funding. These variations may impact the development of clean energy and credit markets in respective regions, consequently influencing the extent of the transition to clean energy for living.

As the world’s largest developing country, China has yet to substantially transition to cleaner energy sources for rural living, leading to significant regional disparities in the extent of this transition. In rural Chinese households, firewood constitutes the primary energy source for domestic heating, accounting for 44.2%, followed by coal at 23.9% (Third Agricultural Census, 2017). The development of gas usage in rural areas lags at 28.6%, significantly lower than in urban areas (98.04%) (Bulletin on the State of Urban Construction in China, 2021) and other countries with comparable economic development levels, where gas usage exceeds 85% (World Health Organization data, 2017) (e.g., Mexico, Brazil, Argentina). Moreover, notable regional variations exist in China’s rural clean cooking energy transition, with higher clean-up rates observed in the eastern coastal provinces compared to other regions (Seventh National Census Data, 2020). In contrast, the northeastern and northwestern regions exhibit meager rates. Like many developing nations, rural areas in China experience energy and credit market failures, with some regions experiencing severe failure [32]. Population density and economic development levels significantly influence the development of inter-regional energy markets. The underdeveloped nature of energy markets, particularly in mountainous and remote rural areas, limits rural households’ access to clean energy [22]. Regarding the credit market, rural financial development varies significantly across regions, as evidenced by variations in geographic, population, and economic financial density [33]. Therefore, this study contends that empirical evidence derived from China’s rural domestic energy transition can elucidate the overarching principles of clean domestic energy transition amidst market failures. Such findings enrich the theory of energy transition and furnish a theoretical foundation for clean domestic energy transitions in developing nations.

Thus, this study develops a decision-making framework for the transition to clean living energy within market failure. It investigates the influence of government funding inputs and energy endowment on this transition, focusing on the accessibility and affordability of clean energy. The empirical analysis uses 20 years of data on energy consumption in rural living, spanning 29 provinces in China. The findings indicate that government funding inputs significantly enhance the transition to clean living energy. This conclusion holds for both per capita funding input and funding intensity. Government funding has four impacts on the shift to clean living energy: it increases the clean energy supply, promotes clean technology, spreads clean knowledge, and eases financial constraints. The impact of government funding inputs on the transition to clean living energy is influenced by renewable energy endowment. In areas abundant in renewable energy, government funding can be effectively utilized to develop and use renewable energy resources, thereby enhancing regional renewable energy supply, improving clean energy accessibility, and fostering the transition to cleaner domestic energy.

There are three potential contributions. Firstly, it pioneers examining government funding inputs’ impact on the clean energy transition for living within market failure, thereby challenging the assumption of perfect energy and credit markets prevalent in the existing literature [16,17]. This approach aligns more closely with the reality of the transition to clean energy for living in developing countries, marked by energy and credit market failures, a common phenomenon in most developing nations [22,34].

Secondly, unlike previous studies that primarily examine the direct effects of environmental regulations and subsidy policies on the cleaner domestic energy transition [18,35,36], this study extends beyond quantifying the impact of government intervention via funding inputs to also analyzing the underlying mechanisms that influence the cleaner domestic energy transition across four dimensions: clean energy supply, promotion of clean technology, dissemination of clean knowledge, and alleviation of financial constraints.

Thirdly, considering the variations in natural conditions and economic development among regions, this study analyzes the heterogeneous impact of government funding inputs in the context of renewable energy endowment. It offers distinct pathway options for clean energy transition in each region and provides a scientific underpinning for advancing domestic energy’s clean energy transition based on localized policies.

The remainder of this study is structured as follows: the second part presents the theoretical and analytical framework; the third part describes the econometric model setup, data sources, and descriptive statistics of the variables; the fourth part presents the estimation results; and finally, the study concludes with policy recommendations.

2. Theoretical Analysis

Under specific institutional arrangements, the market can efficiently allocate resources. However, when such arrangements fail, market failures can occur. Unlike energy and credit markets, which are relatively perfect in urban areas, clean energy and credit markets in rural areas are characterized by market failures. The degree of market failure varies considerably between regions.

Clean energy market failures manifest in four ways. The first is inadequate supply. Due to insufficient market mechanisms, the supply of clean energy in some rural areas may be limited, and private enterprises may be reluctant to enter these areas, resulting in an insufficient supply of clean energy [37]. The second is information asymmetry. Rural residents have inadequate information about clean energy technologies and products, lack detailed knowledge, and are skeptical about its benefits and feasibility [38]. The third is insufficient infrastructure. Some rural areas lack infrastructure, such as gas stations and charging stations, increasing the difficulty of clean energy adoption [39]. The fourth is a fragmented market. Rural areas are usually fragmented markets, and it is challenging to create economies of scale due to the dispersed population and geographic conditions, leading to more difficulties in promoting clean energy projects in rural areas [32].

Credit market failures manifest in insufficient rural financial supply and stringent credit constraints. Rural credit market failure primarily stems from information asymmetry and inadequate information exchange [40]. In contrast to urban credit institutions, rural credit institutions encounter scattered small-scale farmers and a multitude of rural small and medium-sized enterprises (SMEs). Low population density, market segmentation, high transaction costs due to elevated risk and seasonality, the absence of traditional collateral, high income volatility, and limited risk diversification differentiate rural credit markets from urban financial markets. These unique characteristics of rural credit markets significantly amplify information asymmetry, contributing to the failure of rural credit markets [41].

Based on the varying degrees of clean energy market failure, the clean energy transition for rural living is divided into two paths: in rural areas with minimal energy market failures, households shift from non-clean energy sources to clean commodity energy sources such as gas and electricity. At this stage, clean energy endowments such as gas and electricity and inter-area deployment influence the regional supply of clean energy, affecting its accessibility for rural households. Conversely, in rural areas experiencing severe energy market failures, households may contemplate transitioning to clean, non-commodity energy sources such as solar and biogas. However, they encounter constraints imposed by credit market failures [34]. In these contexts, rural households’ substantial financial constraints make clean energy less accessible, hindering the shift to clean domestic energy sources.

To mitigate the negative impacts of market failures on the clean energy transition for rural livelihoods, changing market behavior through government intervention is essentially the primary means to efficiently allocate rural clean energy and credit resources [37]. How does government funding mitigate rural clean energy market and credit market failures? Drawing from the existing literature [18,42], this study identifies four mechanisms: clean energy supply effects, clean technology promotion effects, clean knowledge dissemination effects, and financial constraint mitigation effects.

Clean energy supply effects. The accessibility of clean energy depends on the demand for clean energy endowments. When a region is endowed with natural resources such as gas, light, hydropower, and wind, there are potentially exploitable clean energy sources [38]. At this juncture, assuming that there is demand for clean energy in rural households, increasing the clean energy supply becomes imperative through government funding. Increased investment in clean energy infrastructure in rural areas and support from the government or relevant organizations in executing clean energy transformation projects, particularly renewable energy initiatives, is imperative. This includes the construction of micro- and small-scale power stations, upgrading grid infrastructure, biogas projects, etc., to address energy supply shortages and infrastructure deficiencies. Using clean energy endowment can be maximized by leveraging local natural resources [26]. This includes tapping into local resources or harnessing external resources such as natural gas through pipeline transportation and inter-regional deployment to enhance the regional clean energy supply and improve accessibility [27].

Clean technology promotion effects. For most rural households, clean technology represents a novel concept. Its direct adoption to replace the original energy source is challenging, as evidenced by the following factors: firstly, the complexity and diversity of geographic and climatic conditions in rural areas may necessitate an adaptive design for the new technology in different regions [43]; secondly, the instability or lack of support from government policies may hinder the diffusion of the latest technology [44]; and thirdly, the absence of maintenance and after-sales service may increase farmers’ skepticism toward new technologies [45]. Government funding is invested in promoting clean energy and technology through clean equipment and technology demonstration programs. These aim to demonstrate the actual effects of clean energy technologies on rural households, establish a technology promotion network, and provide rural residents with consultation, maintenance, and technical support. These initiatives advertise clean energy and technology and reduce the barriers to adopting clean energy technology among rural households.

Clean knowledge dissemination effects. On the technology demand side, information gaps in rural households about clean energy technology and knowledge can also affect the adoption of clean energy technology. With the rapid advancement of urbanization, the remaining rural population is mostly aging and needs more information. Their understanding of clean energy needs to be improved. On the one hand, they are constrained by traditional concepts and energy-use habits and need more initiative to actively learn about clean energy and technology [46]. On the other hand, it is difficult to distinguish valid and effective clean energy knowledge, making them vulnerable to false information or scams, such as the early spread of the “photovoltaic loan” scam in some rural areas, which hindered the promotion and installation of household photovoltaic systems in rural areas. The government has invested in disseminating and educating rural households on clean energy and associated technologies through energy training programs to enhance their understanding of clean energy technologies and raise awareness of them [47].

Financial constraint mitigation effects. In addition to barriers to the diffusion and deployment of cleaner technologies, the affordability of cleaner energy is a crucial factor affecting the cleaner energy transition for rural livelihoods. Clean energy and technologies are generally more expensive than non-clean energy sources. In the context of rural credit market failures, rural households face more significant liquidity constraints [48]. They are less likely to afford the high one-time costs associated with clean energy or technologies [34]. To address this, government funding inputs can alleviate the financial constraints of rural households in adopting clean appliances through subsidy policies, loans, and financing support [31,35,49], thereby enhancing the affordability of clean energy for households and facilitating the clean energy transition for rural livelihoods. Based on the above analysis, this study proposes the following:

H1. 

Government funding can enhance energy accessibility and affordability for rural households through clean energy supply, the promotion of clean technology and knowledge dissemination, and financial constraint mitigation, facilitating a transition to cleaner energy for rural livelihoods.

Moreover, regional heterogeneity exists in the magnitude of the impact of government funding on the clean energy transition for rural living due to variations in government funding and energy endowments across regions, particularly in areas with differing energy endowments [18]. Considering that the development and utilization of non-commodity clean energy, particularly renewable energy, relies on farmers’ consumption decisions, factors influencing the accessibility and affordability of clean energy indirectly affect rural energy consumption decisions [20]. The primary prerequisite for rural households to consume clean energy is the abundance of clean energy endowment. Due to the complex geography and dispersed population in rural areas, forming economies of scale is challenging, and rural markets often need to be more cohesive. Relying on local resource endowment to develop and utilize clean energy is a more economical transformation approach [32]. In such circumstances, the regional abundance of renewable energy will influence the allocation of government funding. In regions with abundant renewable energy, the government may prioritize promoting and disseminating clean energy and technology, aiming to achieve a clean energy transition in rural areas at minimal cost, compared to investing heavily in cross-regional gas resource deployment and energy infrastructure construction [50].

H2. 

Government funding significantly impacts the transition to clean energy for rural living in regions with abundant renewable energy sources.

3. Data and Empirical Modeling

3.1. Data Sources

The dataset utilized in this study comprises panel data spanning 29 provinces in China over 20 years from 1995 to 2014, with a total of 580 observations. Considering that fuelwood and straw data are only available until 2014, the period of rural domestic energy consumption and cleaner transition rates accounted for in this study is 1995–2014. In addition, significant data is missing for Tibet, Hong Kong, Macao, and Taiwan, so these provinces are not considered in this study’s accounting, and data for Chongqing Municipality after 1997 are merged with Sichuan Province. Finally, a database of rural domestic energy consumption covering 29 provinces for 20 years was constructed.

Data sources from the China Energy Statistics Yearbook include provincial statistics on the production and consumption of coal, oil, natural gas, electricity, and other commodity energy sources. Information regarding fuelwood consumption, straw, solar energy, biogas, and renewable electricity is derived from the China Energy Statistics Yearbook, the China Rural Energy Statistics Yearbook, and the China Agricultural Statistics Yearbook. Additionally, data concerning investment in rural energy, rural energy management and promotion institutions, extension workers, and trainers are sourced from the China Rural Energy Statistics Yearbook. Information on the urbanization rate, rural population size, rural disposable income, rural old-age dependency ratio, rural young-adult dependency ratio, microwave ovens, and cooker hoods is obtained from the China Statistical Yearbook, China Population and Employment Statistical Yearbook, and China Rural Statistical Yearbook. The fuel sales price index is extracted from the China Rural Statistical Yearbook. Climatic data, including temperature, precipitation, and light, are sourced from the China Meteorological Data Network.

In addition, the reasons for not choosing urban areas as cases in this paper include three aspects. First, the high penetration rates and low regional variations of gas and electricity in China’s cities and towns make it challenging to use econometric modeling methods to identify the causes of small differences. Second, factors such as urban construction and environmental protection in China’s urbanization process combine to influence clean energy supply in cities and towns. Third, energy and credit markets in urban areas are more developed than in rural areas, with lower levels of market failure and smaller regional differences in the degree of market failure.

3.2. Empirical Models

This study constructs an unobserved effects synthesis data model to examine the research hypotheses. It then analyzes the impacts and mechanisms of governmental funding inputs and energy endowment on clean energy transition for rural living by estimating and testing the model parameters.

$$clea{n}_{it}={\delta }_0+{\delta }_{1}input\_go{v}_{it}+{\delta }_{2}energ{y}_{it}+{\gamma }_{1}Z+{\mu }_i+{\nu }_t+{\epsilon }_{it}$$

(1)

The subscripts of the variables in the model (1) are $i$ and $t$ for the province and year. $clea{n}_{it}$ represents clean energy transition for rural living. $input\_go{v}_{it}$ represents government funding for rural energy in each province, energy endowment in each province, and other control variables affecting the clean energy transition for household livelihoods and the mechanism variable; the model also controls for region-fixed and year-fixed effects. ${\nu }_t$ in the model also controls for province- and year-fixed effects and unobservable idiosyncratic disturbances affecting the explanatory variables that vary by time and region. In the model, ${\delta }_0$ represents the parameters to be estimated.

Explained variable. Clean energy transition for rural living is proxied by the clean energy transition rate in rural households, which is calculated as the ratio of total clean energy consumption to total domestic energy consumption, ranging from 0 to 100 [51]. Refer to the annex for detailed measurements of rural domestic energy consumption.

Key explanatory variables. The key explanatory variables include government funding and energy endowment. Government funding input is proxied by regional per capita funding input and funding input intensity [52], with detailed measurements in Appendix A. Energy endowment is proxied by the abundance of commodity energy sources in each province, encompassing both non-clean commodity energy abundance (e.g., coal and petroleum) and clean commodity energy abundance (e.g., natural gas and electricity), measured by the ratio of energy production to energy consumption [53].

Control variables. Forest stock and cultivated land area per capita are proxies for fuelwood accessibility and straw accessibility [54]. Rural disposable income per capita and demographic characteristics such as rural population size, old age dependency ratio, and child dependency ratio are included [11,21,55]. Energy price variables, measured by the retail fuel price index [22], and climate variables, including temperature, precipitation, and light, are measured by average annual temperature, rainfall, and annual light intensity for each province [10]. Given that the explanatory variable, household domestic energy clean transition, is continuous, this study employs panel fixed effects for estimation. Descriptive analysis results for each variable are presented in

Table 1. Descriptive statistics of variables.

Variables	Mean	SD	Min	Max
Explained variables
Clean energy transition rate (%)	9.57	8.753	0.20	54.55
Key explanatory variables
Non-clean commodity energy abundance	0.77	0.603	0.08	5.58
Clean commodity energy abundance	5.75	22.452	0.00	197.60
Per capita funding input (CNY/person)	28.73	62.743	0.07	586.52
Funding input intensity (CNY/Ten thousand)	6.35	6.142	0.07	44.23
Control variables
Forest stock (Ten thousand m3)	36,388	47,159	23.93	197,300
Cropland area (Thousand hm2)	4278	2882	187.65	15,865
urbanization rate (%)	45.93	16.707	12.45	92.99
Per capita disposable income of rural residents (CNY)	3495	22,181	880.34	13,583
Rural child dependency ratio	33.72	12.454	8.64	71.88
Rural old-age dependency ratio	12.29	4.285	0.99	42.51
Rural population (person)	2599	1899	206.00	7969
Retail fuel price index (100 in 1995)	106.55	7.403	85.18	156.90
Temperature (°C)	14.31	5.099	−7.80	25.40
Sunlight (h)	2009	504.237	681.60	3093
Precipitation (mm)	889.66	524.130	74.90	2679
Mechanism variables
Number of rural energy management extension agencies (Units/ten thousand persons)	1.69	1.221	0.02	5.85
Number of rural energy management extension workers (person/ten thousand person)	5.71	4.518	0.01	34.29
Number of rural energy trainers (person)	190.34	282.184	0.00	1625
Number of solar water heaters (Ten thousand m3)	116.93	179.838	0.00	1164
Installed wind power capacity (kW)	837.42	3350	0.00	26,978
Installed hydroelectricity (kW)	6054	11,054	0.00	79,469
Household biogas volume (Ten thousand m3)	24,935	37,921	0.02	203,604
Generated electrical energy (Hundred million kWh)	938.31	852.494	32.00	4655
Natural gas production (Hundred million m3)	21.34	52.491	0.00	410.11
Number of ventilators (Units/hundred households)	17.56	15.946	3.30	66.60
Number of microwave ovens (Units/hundred households)	17.02	16.645	5.70	64.80

.

3.3. Descriptive Analysis

3.3.1. Spatial and Temporal Characteristics of China’s Rural Energy Funding Inputs

Table 2. Changes in regional rural energy financing, 1995–2018.

	Per Capita Funding Inputs			Funding Input Intensity
	1995	2005	2018	1995	2005	2018
Beijing	23.1	65.3	471.9	4.6	2.3	4.5
Tianjin	0.7	15.6	257.4	0.7	1.3	5.5
Hebei	0.9	9.2	75.0	1.7	4.5	8.0
Shanxi	2.8	5.7	40.4	5.7	2.7	4.1
Inner Mongolia	0.6	2.8	98.0	1.0	1.0	5.7
Liaoning	6.5	36.8	57.7	5.4	8.8	3.8
Jilin	0.7	7.8	36.4	1.1	3.6	3.4
Heilongjiang	8.7	10.9	58.9	8.3	4.1	6.2
Shanghai	30.3	34.7	35.2	4.8	0.8	0.3
Jiangsu	1.9	4.4	70.3	1.9	0.9	2.1
Zhejiang	1.9	11.5	57.9	1.9	1.9	2.0
Anhui	1.1	7.9	34.1	2.9	5.5	3.3
Fujian	1.0	11.2	27.5	2.5	3.1	1.1
Jiangxi	2.9	5.2	34.6	7.6	3.6	3.4
Shandong	4.2	5.7	53.7	6.0	1.8	3.4
Henan	0.8	4.5	32.3	2.0	2.8	3.4
Hubei	1.5	13.5	91.2	2.8	6.8	5.8
Hunan	1.4	14.8	53.1	3.3	9.3	4.9
Guangdong	2.2	6.8	12.5	1.1	1.1	0.5
Guangxi	3.2	21.5	60.6	8.0	17.8	8.3
Hainan	3.2	25.1	109.6	5.2	12.9	9.5
Sichuan	1.8	5.1	63.1	5.2	3.9	6.8
Guizhou	0.1	20.8	95.1	0.5	29.3	13.9
Yunnan	10.3	20.0	65.8	18.3	17.9	9.0
Shaanxi	2.5	3.9	67.0	4.4	2.4	5.1
Gansu	0.7	3.4	59.5	2.4	3.2	10.6
Qinghai	5.2	12.6	59.1	9.8	8.3	6.7
Ningxia	0.3	5.0	272.7	0.5	3.0	23.5
Xinjiang	0.1	44.0	65.1	0.1	22.1	7.0
nationwide	2.4	9.7	62.9	3.4	3.9	4.2

Source: The authors’ statistical compilation was based on data from the China Rural Energy Yearbook over the years.

presents the temporal trends and regional disparities in government funding inputs for rural energy in China from 1995 to 2017. Per capita and funding input intensity exhibited an upward trend during this period. However, the per capita funding input growth rate from 1995 to 2005 was significantly slower than from 2005 to 2018. Despite the growth in funding input intensity, it was considerably less than the growth rate of per capita funding input, indicating a potential mismatch between rural energy demand and funding input growth. This disparity may hinder the transition to cleaner energy sources for rural households.

Regional disparities were also observed, with variations in per capita investment and investment intensity. Provinces such as Beijing, Tianjin, Ningxia, Hainan, and Inner Mongolia recorded the highest investments per capita, while Guangdong, Fujian, Henan, Anhui, and Jiangxi recorded the lowest. Regarding investment intensity, Ningxia, Guizhou, Gansu, Hainan, and Yunnan had the highest investment, while Shanghai, Guangdong, Fujian, Zhejiang, and Jiangsu had the lowest. Government funding inputs primarily target the development and promotion of renewable energy projects. Economically developed regions like Guangdong, Shanghai, and Fujian rely more on national commodity energy, resulting in relatively lower direct funding inputs. Conversely, regions rich in renewable resources, such as Inner Mongolia, Gansu, Yunnan, Hainan, etc., allocate more funds to renewables due to their dispersed rural populations, low density, and economic lag, rendering renewable energy development the favored option for addressing rural energy demands. As a result, local governments and households invest more in renewable energy projects.

3.3.2. Spatial and Temporal Characteristics of Clean Energy Transition for Rural Living

Figure 1 illustrates the evolutionary trend of rural domestic energy consumption and the transition to cleaner sources in China from 1995 to 2014. Both per capita rural energy consumption, per capita clean energy consumption, and per capita non-clean energy consumption exhibit a clear upward trend. Specifically, per capita rural energy consumption exhibited a downward trend followed by an upward trajectory, with a slight decline in 2006 and a sustained increase. In comparison, the per capita consumption of non-clean energy for rural areas mirrored that of total domestic energy consumption until 2007, after which its growth rate slowed considerably. Conversely, per capita clean energy consumption for rural areas has consistently risen from 1995 to 2014, with a notably accelerated growth rate post-2007. Hence, the trend in rural domestic energy consumption is primarily driven by non-clean energy consumption before 2007 and clean energy consumption after that. Rural domestic energy consumption continues to be predominantly fueled by non-clean sources, with per capita non-clean energy consumption at 468.2 kgce and per capita clean energy consumption at 107.2 kgce in 2014. The clean transition rate is 18.6%, considerably lower than the 55.8% observed in urban areas (

Table A1. Regional per capita energy consumption and clean transformation changes in rural areas by province in China from 1995 to 2014.

	Energy Consumption Per Capita (kgce)			Clean Energy Consumption Per Capita (kgce)			Clean Transition Rate (%)
	1995	2005	2014	1995	2005	2014	1995	2005	2014
Beijing	969.5	1145.1	1005.6	37.4	119.6	173.6	3.9	10.4	17.3
Tianjin	149.0	467.4	538.3	8.7	62.8	122.8	5.8	13.4	22.8
Hebei	393.4	531.8	801.1	12.2	46.3	117.4	3.1	8.7	14.7
Shanxi	470.4	481.4	696.2	7.2	18.5	96.9	1.5	3.8	13.9
Inner Mongolia	474.6	794.7	887.4	7.9	21.5	68.0	1.7	2.7	7.7
Liaoning	336.5	714.9	874.7	19.3	53.5	94.5	5.7	7.5	10.8
Jilin	473.2	676.7	890.3	12.6	27.5	57.9	2.7	4.1	6.5
Heilongjiang	1120.1	729.9	1017.6	9.9	23.1	74.2	0.9	3.2	7.3
Shanghai	449.7	667.8	638.8	23.7	67.1	155.5	5.3	10.1	24.3
Jiangsu	433.6	490.4	303.9	12.5	47.1	165.8	2.9	9.6	54.5
Zhejiang	281.3	277.8	422.2	8.6	61.7	183.9	3.1	22.2	43.6
Anhui	400.2	418.7	645.8	4.6	23.0	92.6	1.2	5.5	14.3
Fujian	157.3	286.2	562.3	3.5	44.0	161.8	2.2	15.4	28.8
Jiangxi	288.8	331.7	352.3	5.3	24.1	77.1	1.8	7.3	21.9
Shandong	279.8	423.7	499.3	11.0	46.0	142.4	3.9	10.9	28.5
Henan	294.9	338.9	325.0	3.2	22.9	86.5	1.1	6.7	26.6
Hubei	470.8	477.8	675.8	6.9	26.1	90.8	1.5	5.5	13.4
Hunan	430.7	396.0	415.1	4.8	31.9	82.0	1.1	8.0	19.8
Guangdong	800.8	422.9	274.4	13.6	42.3	118.3	1.7	10.0	43.1
Guangxi	279.9	778.5	341.8	6.0	40.0	106.6	2.1	5.1	31.2
Hainan	391.7	548.4	793.0	11.8	168.4	268.6	3.0	30.7	33.9
Sichuan	395.9	662.9	764.7	11.7	28.7	88.2	3.0	4.3	11.5
Guizhou	572.7	943.1	621.3	3.8	39.8	68.4	0.7	4.2	11.0
Yunnan	693.3	391.3	454.9	7.5	31.2	94.8	1.1	8.0	20.8
Shaanxi	484.8	293.1	519.7	10.5	19.2	82.6	2.2	6.5	15.9
Gansu	383.2	429.9	595.1	6.6	23.4	79.8	1.7	5.4	13.4
Qinghai	257.6	423.8	702.7	8.9	18.4	118.5	3.5	4.3	16.9
Ningxia	328.7	725.6	482.8	4.4	29.9	108.4	1.3	4.1	22.5
Xinjiang	666.7	626.1	398.2	5.3	11.5	43.5	0.8	1.8	10.9
nationwide	407.8	471.3	575.4	8.7	33.6	107.2	2.1	7.1	18.6

Source: Statistical compilation by the authors based on data from China Rural Energy Yearbook, China Energy Statistical Yearbook, China Rural Statistical Yearbook over the year, and data from the General Station of Agroecology and Resource Conservation of the Ministry of Agriculture and Rural Development.

).

The primary cause of the high proportion of non-clean energy consumption was significantly higher consumption of coal, fuelwood, and straw compared to other energy sources, with per capita consumption of coal, fuelwood, and straw amounting to 119.8 kg, 111.0 kg, and 196.4 kg, respectively, in 2014 (Figure 2). Although the per capita consumption of fuelwood and straw began to decline in 2004 and 2006, the significant consumption of both appears challenging to reduce substantially in the short term. Moreover, the per capita consumption of coal and petroleum in rural areas has continued to increase since 1999, with a notably higher growth rate post-2007 (Figure 2). This trend has undoubtedly heightened the challenge of transitioning to cleaner rural domestic energy sources. However, with nearly 100% rural electrification and a desire for a better lifestyle, households are replacing dirty fuels like coal, wood, and straw with electricity, purchased through the use of energy-efficient appliances such as air conditioners, water heaters, and gas stoves, which significantly reduce time spent on domestic tasks. Rural biogas and solar energy consumption increased from 1995 to 2014 (Figure 2).

To compare energy consumption and cleaner transitions across regions, we analyze the per capita energy consumption and cleaner transition rates of each province relative to the national average, categorizing China’s 29 provinces into four groups: “high consumption, high clean (HH),” “high consumption, low clean (HL),” “low consumption, low clean (LL),” and “low consumption, high clean (LH).” According to

Table 3. Rural per capita energy consumption and clean energy transition by region, 2014.

	Energy Consumption (kgce)	Clean Energy Consumption (kgce)	Clean Energy Ratio (%)	Type
Beijing	1005.6	173.6	17.3	HL
Tianjin	538.3	122.8	22.8	LH
Hebei	801.1	117.4	14.7	HL
Shanxi	696.2	96.9	13.9	HL
Inner Mongolia	887.4	68.0	7.7	HL
Liaoning	874.7	94.5	10.8	HL
Jilin	890.3	57.9	6.5	HL
Heilongjiang	1017.6	74.2	7.3	HL
Shanghai	638.8	155.5	24.3	HH
Jiangsu	303.9	165.8	54.5	LH
Zhejiang	422.2	183.9	43.6	LH
Anhui	645.8	92.6	14.3	HL
Fujian	562.3	161.8	28.8	LH
Jiangxi	352.3	77.1	21.9	LH
Shandong	499.3	142.4	28.5	LH
Henan	325.0	86.5	26.6	LH
Hubei	675.8	90.8	13.4	HL
Hunan	415.1	82.0	19.8	LL
Guangdong	274.4	118.3	43.1	LH
Guangxi	341.8	106.6	31.2	LH
Hainan	793.0	268.6	33.9	HH
Sichuan	764.7	88.2	11.5	HL
Guizhou	621.3	68.4	11.0	HL
Yunnan	454.9	94.8	20.8	LL
Shaanxi	519.7	82.6	15.9	LL
Gansu	595.1	79.8	13.4	LL
Qinghai	702.7	118.5	16.9	HL
Ningxia	482.8	108.4	22.5	LH
Xinjiang	398.2	43.5	10.9	LL
nationwide	598.2	110.4	20.9	-

Note: Trends in regional per capita rural energy consumption and cleaner transitions, 1995–2014, are shown in Table A1. Source: Statistical compilation by the authors based on data from China Rural Energy Yearbook, China Energy Statistical Yearbook, China Rural Statistical Yearbook over the years, and data from the General Station of Agroecology and Resource Conservation of the Ministry of Agriculture and Rural Development.

, only two provinces, Shanghai and Hainan, have achieved a high consumption and clean energy transition, with Shanghai relying on gas and other commodity energy. At the same time, Hainan utilizes renewable energy for cleaner living (Figure 3). Ten provinces have achieved a low consumption, high clean energy transition. Zhejiang and Jiangsu transitioned to electricity and solar energy, while Guangxi utilized biogas (Figure 3). Additionally, Shandong, Henan, and Jiangxi adopted a combination of electricity, solar energy, and biogas for their clean transition. Provinces lacking clean energy transformation, such as Beijing, Hebei, Shanxi, Inner Mongolia, and others with high coal consumption, and Liaoning, Jilin, Heilongjiang, Anhui, Hubei, and Sichuan with high fuel wood and straw consumption, have faced challenges (Figure 3). The former group has been influenced by the national policy of shifting from coal to electricity and gas, resulting in a significant decline in coal consumption and an uptick in clean energy transition rates. Conversely, the latter group requires government assistance to transition to clean energy, mainly through biogas initiatives. The latter necessitates increased government investment in biomass energy development and utilization, leveraging the region’s biomass energy potential. This involves comprehensive strategies such as straw curing, charring, and centralized gas supply to achieve a clean energy transition for rural living.

4. Results

4.1. Benchmark Regression Results

Table 4. Estimated results of the impact of government funding and energy endowment on the clean energy transition for rural living.

	Explained Variable: Clean Energy Transition Rate in Rural Households
	(1)	(2)	(3)	(4)	(5)	(6)
Key explanatory variables
Per capita funding inputs	1.277 ***		1.004 ***		1.116 ***
	(3.770)		(2.901)		(3.259)
Funding input intensity		1.230 ***		0.978 ***		1.050 ***
		(3.516)		(2.800)		(3.001)
Non-clean commodity energy abundance	−1.937 ***	−1.842 ***			−1.827 ***	−1.740 ***
	(−3.027)	(−3.094)			(−2.964)	(−3.006)
Clean commodity energy abundance			0.029 ***	0.030 ***	0.027 ***	0.027 ***
			(3.084)	(3.102)	(2.821)	(2.869)
Control variable
Forest stock	−0.000 *	−0.000 *	−0.000 **	−0.000 **	−0.000 **	−0.000 **
	(−1.920)	(−1.935)	(−2.364)	(−2.370)	(−2.168)	(−2.191)
Cropland area	−0.002 ***	−0.002 ***	−0.001 ***	−0.001 ***	−0.002 ***	−0.002 ***
	(−5.178)	(−5.193)	(−4.310)	(−4.348)	(−4.859)	(−4.853)
Urbanization rate	0.280 ***	0.289 ***	0.306 ***	0.315 ***	0.325 ***	0.335 ***
	(3.739)	(3.835)	(3.903)	(4.010)	(4.189)	(4.298)
Rural disposable income per capita	0.002 ***	0.002 ***	0.002 ***	0.002 ***	0.002 ***	0.002 ***
	(7.467)	(7.474)	(7.283)	(7.327)	(7.504)	(7.509)
Rural child population ratio	−0.018	−0.022	0.008	0.005	−0.007	−0.010
	(−0.451)	(−0.557)	(0.185)	(0.113)	(−0.165)	(−0.238)
Rural elderly population ratio	−0.253 ***	−0.236 ***	−0.260 ***	−0.248 ***	−0.264 ***	−0.249 ***
	(−4.184)	(−4.009)	(−4.273)	(−4.184)	(−4.327)	(−4.203)
Rural population	−0.000	−0.001 *	−0.001 **	−0.001 ***	−0.001	−0.001 **
	(−1.135)	(−1.760)	(−2.329)	(−2.791)	(−1.578)	(−2.136)
Retail fuel price index	−7.163 ***	−6.968 ***	−7.194 ***	−7.050 ***	−7.278 ***	−7.103 ***
	(−5.833)	(−5.693)	(−5.752)	(−5.661)	(−5.858)	(−5.735)
Temperatures	0.208	0.223	0.214	0.226	0.210	0.224
	(0.828)	(0.871)	(0.746)	(0.777)	(0.801)	(0.837)
Sunlight	−0.000	−0.000	−0.000	−0.000	−0.000	−0.000
	(−0.285)	(−0.308)	(−0.376)	(−0.386)	(−0.206)	(−0.222)
Constant	−0.000	−0.000	−0.000	−0.000	−0.000	−0.000
	(−0.324)	(−0.318)	(−0.417)	(−0.406)	(−0.290)	(−0.283)
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Observations	580	580	580	580	580	580
Adjustment of R2	0.812	0.811	0.811	0.811	0.814	0.813
F-value	16.316 ***	16.325 ***	16.055 ***	16.096 ***	15.621 ***	15.638 ***

Note: Numbers in parentheses are Z-values; ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

provides a detailed overview of the estimation results concerning the effects of government funding inputs and energy endowment on the clean energy transition for rural living. Table 4 presents regression results regarding per capita funding inputs’ impact on the clean energy transition for rural living in columns (1), (3), and (5), while columns (2), (4), and (6) depict results for the intensity of funding inputs. The regressions are estimated using panel fixed-effects models, controlling for year and province-fixed effects across all analyses.

Examination of columns (1), (3), and (5) reveals a significant positive effect of per capita government funding input on clean energy transition for rural living, suggesting that for each 1% increase in per capita funding input, the rate of clean energy transition in rural areas increases by 0.7% (both per capita funding and funding intensity have been taken in logarithmic terms). The results in columns (2), (4), and (6) indicate a significant positive effect of the intensity of government funding inputs on clean energy transition for rural living. Specifically, for each 1% increase in the intensity of government funding inputs, the rate of clean energy transition in rural areas increases by 0.7%, consistent with the impact observed for per capita funding inputs. This finding aligns with the conclusions drawn by Su and Tan [18]. Per capita government funding is analyzed from the perspective of population density to investigate the importance of government funding. Considering that population density not only affects the overall investment but also affects the cost and quantity of clean energy infrastructure supply and the difficulty of clean technology promotion and knowledge dissemination [42], the higher the population density, the higher the cost of clean energy infrastructure supply, the lower the quantity of supply, the more difficult it is to promote clean technology, and the fewer the chances of receiving clean knowledge education. The intensity of the funding is analyzed with regard to the importance of the government’s financial investment from the point of view of the economic level: the higher the regional economic level, the higher the corresponding financial capacity, and the more funding that can be allocated to rural energy. This can mitigate economic constraints on government funding and influence the transition to clean energy for rural living [49]. The above analysis will also be tested later in the mechanism analysis.

This study primarily examines the influence of commodity energy abundance on clean energy transition for rural living. Table 4, columns (1), (2), (5), and (6), reveal that non-clean commodity energy abundance exerts a significant negative impact on the clean energy transition for rural living. Specifically, for each 1-unit increase in non-clean commodity energy abundance, the rate of clean energy transition in rural areas decreases by 1.7%. Columns (3), (4), (5), and (6) demonstrate that clean commodity energy abundance positively influences clean energy transition for rural living [18]. Specifically, for each 1-unit increase in clean commodity energy abundance, the rate of clean energy transition in rural areas increases by 0.02%. This finding suggests that energy production significantly influences the composition of rural domestic energy, with regions abundant in coal and oil reserves supplying more non-clean energy at relatively lower prices, leading to higher consumption of non-clean energy by rural households [15,56]. Analogous to the resource curse theory, regions abundant in non-clean commodity energy sources are susceptible to an “energy curse,” negatively impacting the clean energy transition for rural living [26].

Concerning the control variables, the cultivated land area exhibits a significant negative impact on the clean energy transition for rural living, consistent with the findings of Muller and Yan [12]. The more extraordinary endowment of arable land suggests that rural households have access to straw for household cooking or heating activities at little to no cost, particularly evident in the Northeast, in Henan and Anhui, where straw consumption is notably higher than in other provinces, thereby impeding their transition to cleaner energy sources [20]. However, in recent years, this phenomenon has been mitigated significantly through the promotion of straw-for-field technology, and the advancement and promotion of comprehensive straw utilization technology are expected to reduce the reliance on straw for rural energy needs [32]. Urbanization positively influences the clean energy transition for rural living in two ways. First, urbanization facilitates the migration of rural laborers to urban areas, enhancing non-farm income opportunities and fostering clean technology knowledge [57]. Second, urbanization promotes the diffusion of clean energy technologies from urban to rural areas, for instance, by expanding the availability of LPG in townships and reducing the cost of access for rural households [58]. However, this finding is inconsistent with the fact that the resource effects of population urbanization follow a predominantly inverted “U-shaped” pattern [59]. Conversely, the regional rural old-age dependency ratio exhibits a significant negative association with the clean energy transition rate in rural households, aligning with the research findings [55,60]. This can be attributed to the income constraints and technological barriers older age groups face in adopting clean energy and their inertia in using traditional energy sources like fuelwood and straw [61]. The size of the rural population in a region significantly impedes the clean energy transition for rural living, suggesting that larger rural populations face more significant challenges in adopting cleaner lifestyles. Fuel prices exert a considerable positive impact on the clean energy transition for rural living. This effect arises because clean fuel prices typically surpass those of non-clean fuels, and the retail fuel price index, reflecting a weighted average of various fuel prices, is mainly influenced by high gas and electricity prices. Consequently, higher clean energy prices decrease the likelihood of cleaner commodity energy being selected as the domestic energy source for rural households.

4.2. Robustness Analysis

4.2.1. Clean Energy Transition

This study constructs clean and renewable energy substitution indexes [54]. The results of the panel fixed-effects model, consistent with the control variables of the baseline model, are shown in

Table 5. Robustness analysis: results of replacing clean energy transition estimates.

	Explained Variable: Clean Energy Substitution Index		Explained Variable: Renewable Energy Substitution Index
	(1)	(2)	(3)	(4)
Per capita funding inputs	2.282 ***		34.088 **
	(3.906)		(2.171)
Funding input intensity		2.743 ***		44.554 **
		(4.444)		(2.403)
Non-clean commodity energy abundance	−2.493 **	−2.641 ***	16.514	12.037
	(−2.520)	(−2.604)	(0.586)	(0.436)
Clean commodity energy abundance	0.053 ***	0.043 **	−0.128	−0.304
	(2.894)	(2.434)	(−0.325)	(−0.756)
Control variables	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes
Observations	580	580	574	574
Adjustment of R2	0.763	0.767	0.556	0.558
F-value	13.500 ***	14.220 ***	0.931	0.971

Note: Numbers in parentheses are Z-values; *** and ** represent significance at the 1% and 5% levels, respectively.

. The analysis reveals that both per capita financial input and the intensity of financial inputs significantly influence the cleaner transformation of rural household energy. Rural households tend to prefer cleaner or renewable energy over non-clean or fossil energy sources in their household energy usage. Similarly, the influence of regional energy endowments on the clean energy transition for rural living aligns with the baseline model and expectations.

4.2.2. Replacement of Energy Endowments

In addition to energy endowment abundance, this study adopts energy production as a proxy for regional energy endowment [26,53]. Specifically, regional coal production, oil production, and natural gas production are utilized as proxies for regional energy endowment. The control variables and estimation methods remain consistent with the baseline model, and the estimation results are presented in

Table 6. Robustness analysis: alternative energy endowment estimation results.

	Explained Variable: Clean Energy Transition Rate in Rural Households
	(1)	(2)	(3)	(4)
Per capita funding inputs	1.435 ***	1.257 ***	1.490 ***	1.621 ***
	(4.159)	(3.748)	(4.184)	(4.699)
Coal production	−0.000 ***			−0.000 ***
	(−5.972)			(−4.898)
Oil production		−0.002 ***		−0.001 **
		(−4.074)		(−2.008)
Natural gas production			−0.027 ***	−0.017 ***
			(−6.353)	(−3.464)
Control variables	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes
Observations	580	580	580	580
Adjustment of R2	0.815	0.815	0.817	0.821
F-value	18.437 ***	16.555 ***	17.910 ***	17.300 ***

Note: Numbers in parentheses are Z-values; *** and ** represent significance at the 1% and 5%, levels, respectively.

. The results reveal that, in line with the baseline model, financial inputs significantly influence the clean energy transition in rural households. At the same time, coal production and oil production exhibit a negative impact on this transition. However, a notable deviation is observed in the case of natural gas production, which significantly impedes the clean energy transition in rural households. This may stem from the disparity between natural gas production and marketing regions. While natural gas reserves are abundant in China’s central and western basins, accounting for 52.3% of the country’s total, the distribution is skewed towards the west and north. Despite initiatives like the “West-to-East Gas Pipeline” project, which aims to transport natural gas from the resource-rich western regions to the eastern part of the country, where reserves are scarce, natural gas consumption in the east areas remains higher [62]. This imbalance is exacerbated by factors such as the sparse population density in rural areas of the western region, resulting in the construction of natural gas pipelines with limited economies of scale and inadequate supply to meet demand. Additionally, price factors contribute to the underconsumption of natural gas, maintaining a pattern of gas abundance in the region without promoting a substantial clean energy transition.

4.2.3. Replacement of Funding Inputs

Following the methodology proposed by Zhang et al. [52], this study also incorporates the approach suggested by Shan [63] to compute the stock of funding inputs. The control variables and estimation methods remain consistent with those in the baseline model, and the estimation results are presented in

Table 7. Robustness analysis: results of alternative funding input estimates.

	Explained Variable: Clean Energy Transition Rate in Rural Households
	(1)	(2)
Per capita funding inputs—Shan Haojie method	1.431 ***
	(3.955)
Intensity of funding inputs—Shan Haojie method		1.415 ***
		(3.672)
Non-clean commodity energy abundance	−1.740 ***	−1.801 ***
	(−3.006)	(−3.097)
Clean commodity energy abundance	0.027 ***	0.021 **
	(2.869)	(2.148)
Control variables	Yes	Yes
Year fixed effect	Yes	Yes
Province fixed effect	Yes	Yes
Observations	580	580
Adjustment of R2	0.813	0.816
F-value	15.638 ***	16.056 ***

Note: Numbers in parentheses are Z-values; *** and ** represent significance at the 1% and 5% levels, respectively.

. The results demonstrate that, in line with those obtained from the baseline model, the average per capita funding inputs and the intensity of the funding inputs continue to exert a significant positive impact. Moreover, the estimated coefficients closely resemble those of the benchmark model.

4.3. Heterogeneity Analysis

Owing to the vast geographical extent of China, substantial disparities exist among regions concerning natural conditions, resource endowment, and economic development levels, leading to considerable variations in the baseline and magnitude of changes in government funding and energy endowment across the country. Given that government funding primarily influences the clean energy transition in rural households by impacting rural renewable energy consumption, this study concentrates on analyzing the disparities in renewable energy endowments. Drawing upon the methodology proposed by Xu and Sun [54], this study devises a renewable energy abundance index by considering the number of renewable energy reserves across various regions. Specifically, the index incorporates four components: theoretical reserves of hydropower resources, land-based 70 m height wind energy reserves, solar energy resources, and biomass resource distribution. Essentially, the index represents the aggregate proportion of a region’s diverse renewable energy reserves relative to the nation’s total reserves. A higher index value signifies a greater abundance of renewable energy, while a lower value indicates the opposite. Based on the descending order of the index, regions falling within the top 50 quartiles are categorized as abundant renewable energy. In contrast, those in the subsequent 50 quartiles are classified as poor renewable energy. Regression analysis is then conducted separately for pilot and non-pilot areas using the two groups of samples, and the estimation results are presented in

Table 8. Heterogeneity analysis results by renewable energy endowment.

	Abundance Renewable Energy		Poor Renewable Energy
	(1)	(2)	(3)	(4)
Per capita funding inputs	−0.653 **		2.452 ***
	(−2.553)		(4.003)
Funding input intensity		−0.425 *		3.579 ***
		(−1.666)		(4.771)
Commodity energy endowment	Yes	Yes	Yes	Yes
Control variables	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes
Observations	300	300	280	280
Adjustment of R2	0.858	0.856	0.808	0.812
F-value	13.609 ***	13.781 ***	8.152 ***	8.448 ***

Note: Numbers in parentheses are Z-values; ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

.

The findings indicate that government funding inputs in regions abundant in renewable energy significantly impede the clean energy transition for rural living. In contrast, in areas with scarce renewable energy resources, government funding inputs have a notable positive effect on the clean energy transition for rural areas. This contradicts the theoretically expected outcomes and fails to support Hypothesis 2. Given these findings, this study posits that the likely reasons are as follows: Renewable energy-abundant areas are currently characterized by relatively underdeveloped economic conditions, low income levels among rural households, and substantial financial constraints, hindering the development and utilization of renewable energy. Though government funding has been increased, it still needs to be improved compared to the investments in energy infrastructure construction in urban areas, exacerbating rural households’ financial constraints. Conversely, in renewable energy-poor regions characterized by relatively advanced economic development and higher income levels among rural families, the primary impediments to the clean energy transition are the technological thresholds and the need for more awareness regarding clean energy. Government funding can be pivotal in promoting clean technology and disseminating knowledge on clean energy in rural areas, raising awareness among rural households. These observations suggest that government funding may only sometimes yield the desired outcomes, and its effectiveness in promoting clean technology and disseminating clean energy knowledge can be optimized only when rural households have improved energy affordability.

5. Mechanism Analysis

We construct the following econometric model to discuss further the intrinsic mechanism of governmental funding inputs to the cleaner rural domestic energy transition.

$$M_{it}={\beta }_0+{\beta }_{1}input\_go{v}_{it}+{\gamma }_{2}X+{\mu }_i+{\nu }_t+{\epsilon }_{it}$$

(2)

$M_{it}$ is a mechanism variable representing clean energy supply, clean technology promotion, clean knowledge dissemination, and financial constraint alleviation in each province. Clean energy supply is proxied by clean commodity and non-commodity energy supply in each region. The former is measured by natural gas production and electricity generation [20]. At the same time, the latter is represented by the number of solar water heaters, installed wind power, hydroelectricity, and household biogas [20]. Clean technology promotion is indicated by the number of institutions and personnel engaged in energy management and promotion, assessed through the number of management and promotion institutions per capita and the number of management and promotion personnel per capita, respectively [18,43]. Clean energy knowledge dissemination is measured by the number of rural energy trainers in each region and quantified by the number of trainers per capita [18]. Financial constraint mitigation is evaluated by the number of microwave ovens and range hoods per 100 rural households in each region, serving as proxies for clean rural appliances [37]. The definitions of other variables ${\epsilon }_{it}$ are consistent with model (1). The descriptive analysis of the variables is shown in Table 1.

5.1. Clean Energy Supply Effect

This study examines the impact of government funding inputs on the supply of clean, non-commodity energy and clean commodity energy in rural areas. Clean, non-commodity energy encompasses solar energy, wind power, hydroelectric power generation, and household biogas, while clean commodity energy comprises natural gas and electricity. The estimation results are presented in

Table 9. Estimated results of the impact of government funding on rural clean non-commodity energy supply.

	Solar Water Heater		Wind Power		Hydroelectricity		Household Biogas
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Per capita funding inputs	12.620 *		468.065 ***		769.086		3874.467 ***
	(1.828)		(2.934)		(1.552)		(2.785)
Funding input intensity		25.785 ***		324.714 ***		−186.125		6740.376 ***
		(4.229)		(2.641)		(−0.391)		(5.485)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	667	667	667	667	667	667	644	644
Adjustment of R2	0.806	0.809	0.816	0.814	0.675	0.674	0.849	0.853
F-value	22.024 ***	25.992 ***	2.990 ***	2.893 ***	3.993 ***	4.060 ***	24.222 ***	25.225 ***

Note: Numbers in parentheses are Z-values; ***and * represent significance at the 1% and 10% levels, respectively.

and

Table 10. Estimated results of the impact of government funding on rural clean commodity energy supply.

	Electrical Power		Petroleum
	(1)	(2)	(3)	(4)
Per capita funding inputs	107.532 ***		9.937 ***
	(3.716)		(3.184)
Funding input intensity		83.181 ***		10.703 ***
		(3.072)		(3.954)
Control variables	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes
Observations	667	667	667	667
Adjustment of R2	0.879	0.878	0.690	0.692
F-value	36.342 ***	34.916 ***	5.884 ***	5.918 ***

Note: Numbers in parentheses are Z-values; *** represent significant at the 1% levels.

. The findings reveal that both per capita funding and funding input intensities significantly positively impact the supply of clean, non-commodity energy and clean commodity energy in rural areas. Analysis of the coefficients indicates that government funding inputs primarily stimulate the supply of rural hydropower and household biogas, aligning with the ongoing focus on advancing small hydropower and biogas in rural energy policies. Given the challenges confronting fossil energy sources due to international energy import and export tensions, it is imperative to enhance government funding to sustain the development of renewable energy sources, leveraging China’s extensive reserves of renewables like solar and wind power.

5.2. Clean Technology Promotion Effects

This study separately estimated the impact of government funding inputs on rural energy management extension institutions and personnel, with the estimation results presented in

Table 11. Estimated results of the impact of government funding on rural energy management promotion and training.

	Managing Extension Agencies		Management Promoter		Trainers
	(1)	(2)	(3)	(4)	(5)	(6)
Per capita funding inputs	0.608 ***		1.618 ***		36.363 *
	(8.777)		(5.810)		(1.711)
Funding input intensity		0.502 ***		1.304 ***		49.332 **
		(8.049)		(6.171)		(2.467)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Observations	667	667	667	667	667	667
Adjustment of R2	0.663	0.654	0.632	0.626	0.696	0.698
F-value	33.428 ***	27.361 ***	21.649 ***	19.731 ***	22.575 ***	23.719 ***

Note: Numbers in parentheses are Z-values; ***, **, and * represent significance at the 1%, 5%, and 10% levels, respectively.

. The findings indicate that both per capita funding and funding input intensities significantly increase the number of rural energy management extension institutions and personnel. The rise or fall of government funding input impacts the number of extension institutions and personnel; however, whether it influences their quality requires further investigation. As the central government’s institutional reform progresses, the consolidation of institutions and personnel has emerged as a prevailing trend [43]. Additionally, the preceding input description indicates a downward trajectory in recent years in the number of rural energy management extension institutions and personnel. Nonetheless, continual increments in government funding for the diffusion of rural clean energy technologies are crucial to overcoming the final-mile challenge of technology diffusion, especially in rural areas. This could entail enhancing the caliber of diffusion personnel and updating diffusion technologies.

5.3. Clean Knowledge Dissemination Effects

The impact of government funding inputs on rural energy trainers was estimated in this study, with the results presented in Table 11. The findings indicate that both per capita funding inputs and funding input intensities significantly positively impact rural energy trainers. Rural energy training predominantly encompasses roles such as biogas production workers, property management workers, rural energy conservationists, solar energy utilization workers, and other rural energy utilizers, focusing on advancing and utilizing renewable energy. Government investment in clean energy and technology knowledge dissemination and education, facilitated by expanded rural energy training and vocational skills certification, can foster a cohort effect to simplify the adoption of clean technologies for rural households. Nevertheless, akin to the promotion of other agricultural technologies, clean energy technologies encounter challenges such as financial constraints and low participation motivation. Hence, augmenting government funding to subsidize training and education is imperative to facilitate the transition to cleaner energy in rural areas.

5.4. Financial Constraint Mitigation Effects

The impact of government funding inputs on rural households’ adoption of cooker hoods and microwave ovens was estimated in this study, with the results presented in

Table 12. Estimates of the impact of government funding on rural appliance use.

	Ventilators		Microwaves
	(1)	(2)	(3)	(4)
Per capita funding inputs	2.823 **		2.165 **
	(2.192)		(2.176)
Funding input intensity		2.622 **		2.170 **
		(2.145)		(2.302)
Control variables	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
Province fixed effect	Yes	Yes	Yes	Yes
Observations	232	232	232	232
Adjustment of R2	0.974	0.974	0.983	0.983
F-value	4.122 ***	4.101 ***	4.287 ***	4.346 ***

Note: Numbers in parentheses are Z-values; *** and ** represent significance at the 1% and 5% levels, respectively.

. The findings indicate that per capita funding and funding input intensity significantly impact rural households’ adoption of cooker hoods and microwave ovens. This implies that government funding inputs can partially mitigate the financial constraints associated with household clean technology upgrades for cooking, heating, and water boiling by offering acquisition subsidies. The positive impact of household biogas, as shown in Table 9, can also corroborate this mechanism. Currently, subsidies for rural households to adopt clean technologies still need to be improved. Despite the nationwide “Home Appliances to the Countryside” campaign in 2007, which offered subsidies for the replacement of clean and energy-saving home appliances in select provinces, there remains a shortage of dedicated funds to subsidize the cleaner replacement of rural energy sources [48], thereby hindering the promotion of a cleaner energy transition for rural life. It is imperative that government funding be specifically allocated for the adoption and replacement of clean energy sources.

6. Conclusions and Policy Implications

Distinguishing itself from previous studies that solely analyze the effects of income, price, and individual characteristics on the clean energy transition under the assumption of perfect markets [11,12,13,16,17,60], this study constructs a decision-making framework for the clean energy transition in rural households within the context of market failure. It examines regional disparities in clean energy transition in rural households in China, considering governmental funding inputs and energy endowment, focusing on the accessibility and affordability of clean energy. The study empirically tests these factors using up to 20 years of provincial panel data. The conclusions are as follows.

The results show that government funding has a significant positive effect on the clean energy transition of rural livelihoods, with each 1% increase in per capita funding increasing the rate of clean energy transition of rural livelihoods by 0.7%. Each 1% increase in funding intensity increases the rate of clean energy transition of rural livelihoods by 0.7%. As mentioned by Su and Tan, government support can reduce the consumption of traditional fossil energy sources and increase the consumption of renewable energy sources, promoting energy transition. This conclusion has also been confirmed by several studies [48]. However, we need to move beyond merely proving the importance of government support, as studies have long pointed out that in the context of market failures, the role of the visible hand, i.e., appropriate government intervention, is necessary [37]. Furthermore, where government intervention takes place is a crucial and essential part of the discussion. Therefore, this paper further analyzes the intrinsic mechanism of the impact of government funding on the clean energy transition of rural livelihoods and obtains four effects: the clean energy supply effect, the clean technology promotion effect, the clean knowledge dissemination effect, and the financial constraint alleviation effect. The first three effects mainly increase the accessibility of rural clean energy, and the financial constraint alleviation effect increases the affordability of clean energy to play the role of government funding inputs on the cleaner transformation of rural domestic energy.

In addition, the heterogeneity results show that in regions with abundant renewable energy, government funding can be fully utilized for the development and utilization of renewable energy, increasing the supply and accessibility of rural renewable energy and promoting a cleaner energy transition for rural livelihoods. Distinguishing from previous studies that only analyze the heterogeneity of government support by region [18,43], we argue that energy endowment is an essential factor influencing regional government funding inputs without a deeper understanding of where the inter-regional sources originate. Therefore, attention should be paid to inter-regional variability and intrinsic sources at the theoretical and practical levels [20,33]. Moreover, consideration should be given to regional renewable energy resources when advancing clean energy supply, technology promotion, and knowledge dissemination, with tailored fiscal and energy policies to enhance the effectiveness of government investments.

The findings of this study are of practical significance for policy instruments to promote cleaner domestic energy transition. The recommendations include two aspects: first, respecting the objective law of energy transition and playing a positive role in the market for living energy transition. This includes improving the clean energy market, promoting the grid and new energy service system, enhancing household clean energy accessibility [20], improving the credit market, innovating financing methods and service modes, and giving differentiated preferences to high-quality energy projects to enhance the affordability of clean energy for rural households [43]. Second, increasing the government’s financial investment in energy construction, extension services, publicity, and education. This includes formulating differentiated fiscal and energy policies to alleviate the economic constraints of energy construction [35], improving and perfecting the energy service system, actively exploring a new model based on market-oriented operation and supplemented by government support, reducing the technical threshold of clean energy transition [42], deepening energy transition publicity and education, and using publicity brochures and door-to-door publicity, clean energy equipment experience, and other ways to increase residents’ clean energy and technology cognition [32].

Moreover, several limitations in this study warrant attention in future research endeavors. Firstly, data insufficiencies persist as a critical drawback, necessitating the collection of more comprehensive and updated energy consumption data to explore the nexus between governmental funding inputs and clean energy transition in rural households. Secondly, governmental funding inputs exhibit hierarchical structures, with involvement from central, local, or federal levels, potentially resulting in varying roles across different governance tiers; analyzing how funding allocations are distributed among different levels of government in rural energy initiatives warrants further investigation. Lastly, further analysis of the mechanisms of household decision making on the consumption of various types of energy sources and the role of the government therein at the micro level, such as household demographic factors such as population aging, household miniaturization, residential separation, and the solitary living of young people, is a direction for future discussion.