Energy infrastructure improvements have prompted inclusive growth in rural China

: Stable energy supply with high quality infrastructure is vital for sustainable energy consumption and inclusive growth. In this paper, we develop empirical methods to evaluate the extent to what energy infrastructure improvement towards inclusive growth, which help guide policy development to achieve Sustainable Development Goals. By investigating the national-scale energy infrastructure improvement project in rural China, we identify the general and inclusive effect of income growth attributed to improved quality of energy infrastructure. The results show that energy infrastructure improvement contributes to rural income increase up to 69 Chinese renminbi (RMB) a year and narrows the income gap by 10 RMB for residents originally with an income difference of 100 RMB. The inclusive growth effect is pronounced in the eastern region, for the poor, for the educated, and becomes remarkable over time. Energy infrastructure improvement leads to income inclusive growth through abundant working time in the eastern and western regions, and general income growth in the middle region through individual health condition. Continuous efforts to energy infrastructure improvement and investments in education and work trainings, especially in the middle and western regions, are critical to achieve inclusive growth.


Introduction
Inclusive growth firstly introduced by the Asian Development Bank in 2007 aims at fairly and reasonably balancing economic growth and narrowing income distribution gap. Prompting inclusive and sustainable social community and economic growth is one of the most critical goals in the United Nation 2030 Sustainable Development Goals (SDGs)1. One challenge for achieving inclusive growth is energy poverty, defined by the International Energy Agency (IEA) as the conditions that people mainly utilize biomass energies (including fuelwood and bagasse) and other solid fuels for cooking and heating rather than using clean energies. Although household energy consumption accounts for a small proportion of total energy use of a country, it may generate dirty emissions and consequently lead to environmental damage and health risks2,3. Due to energy poverty, it is very likely that residents have to live in a poverty trap of air pollution, illness, reduced working competence, economic poverty and continuous usage of dirty energies4,5,6.
Numerous researches have indicated the general growth effect of energy infrastructure on economic and social development7, 8,9, as the utilization of clean energies can greatly promote the growth of gross domestic product (GDP) and reduce CO2 emission. Improved energy supply system can create a living environment more beneficial for residents to reach a higher level of income. However, most of the researches on energy supply are restricted on its general effects on national development10,11, resident income12,13, and firm production14, 15,16. The role of energy infrastructure for inclusive development has received little attention. Besides, previous studies on inclusive growth focus on exploring the factors for promoting inclusive growth from the perspectives of regulatory policy, industrial structure, international trade, economic form, and technology innovation17, 18,19. The explanation of the income inclusive growth from the perspective of energy infrastructure improvement remains to be further investigated.
As the largest developing country, China also has faced a challenge to achieve inclusive growth. Per capita disposable income for urban Chinese (39250.8 RMB) is about three times larger than that of rural Chinese (14617 RMB) in 2018, implying a significant urban-rural gap in resident incomes20. China alone contributes to approximately 30% of global biomass fuels. A latest study shows that residential sector contributes 27% primary PM2.5 in mainland China and about 80% residential emissions come from rural areas21. The quality of energy supply infrastructure in rural China has risen gradually over time. With energy infrastructure improvement, the distribution of clean energy by region has converged since 2000. However, the clean energy accessibility in the middle and western regions is lower than that in the eastern region (see supplementary data 1). Thus, clean energy accessibility and income inclusive growth in rural China are critical for the sake of global sustainable development and poverty elimination in terms of energy use and income. However, there is a lack of quantitative analysis on the interactions between energy infrastructure improvement and rural inclusive growth in China.
Although increased innovation23, education24,25,26,27, financial integration28, human capital29, unemployment30, institution31 have been explored to be the reasons for income inequality and studies have revealed the relationship between energy consumption and economic growth32, we argue that income inclusive growth is derived from energy infrastructure improvement. We fill the gap by providing an empirical evidence to show that the provision of public infrastructure helps an economy to eliminate energy poverty and thus induce income growth at a micro level.
Three major questions are addressed. Firstly, whether does energy infrastructure   improvement prompt inclusive growth and thus reduce income inequality between   those with heterogeneous incomes, educations and work experiences? Secondly, how   does income inclusive growth vary in the eastern, middle and western regions? Thirdly, what are the transmission mechanisms of inclusive growth? The results are essential for accelerating inclusive growth in regions with different levels of economic developments and resource endowments so as to help to design policies for sustainable and inclusive improvements.
To answer these questions, this paper provides a quantitative estimation of the effect of energy infrastructure improvement on inclusive growth and thus reduces income inequality. We extend the traditional Mincer model22, by clearly distinguishing between general growth effect and inclusive growth effect. The general growth effect refers to the increase in absolute income, and the inclusive growth effect indicates the decrease in income gap. By using a large-scale individual-level data obtained from the China Health and Nutrition Survey (CHNS), our empirical analysis finds that energy infrastructure improvement contributes to the increase in rural income growth and helps to narrow the rural income gap. The inclusive growth effect is more pronounced in the eastern region than that in the western region. Energy infrastructure improvements prompt income inclusive growth through abundant working time in the eastern and western regions. In the middle region, energy infrastructure improvement promotes income general growth through individual health condition. The poor and the educated can benefit more than their counterparts from energy infrastructure improvements.

Results
National inclusive estimation. Table 1 reveals the inclusive growth effect from the national perspective. Model I shows that the coefficient of energy infrastructure improvement is positive at a statistically significant level of 1%. That is, improving energy infrastructure has a positive effect on income growth. On average for the general growth, the incomes of rural residents increase up to 69 renminbi (RMB) a year. The coefficient of the interaction term is about -0.5, which implies that the inclusive growth effect of energy infrastructure improvement can narrow the income gap by 10 RMB for two residents originally with an income difference of 100 RMB.
Thus, energy infrastructure improvement does not only enhance resident income but also narrow the gap of resident income and reduce rural inequality. The schooling year has a significantly positive correlation with incomes at the level of 1%, which is consistent with the results in the literature33, 34     statistic is in parentheses. ***, ** and * are significant at the 1%, 5% and 10% levels, respectively.
There exists a huge difference from rural resident incomes, with the minimum and maximum of 1.4354 RMB and 1967213 RMB respectively, showing a dramatic income inequality. By taking the mean of rural resident incomes in the sample as a critical point, we re-estimate the models for low-income and high-income groups in columns 4-5. The general growth effect is much stronger for high-income group (145 RMB) than low-income group (36 RMB). Rural energy infrastructure improvement exhibits inclusive growth effects for both low-income and high-income groups, which helps to narrow the income gaps of the high-income group by 13%, and of the low-income group by 5%. Moreover, we find that sch and exp are significant at the level of 1% for low-income group, which implies that to improve education and work experience of the poor enhances income inclusive growth.

Regional inclusive estimation.
The results of the eastern, middle and western regions are reported in Table 2. It is found that the eastern region has higher levels of both general and inclusive growth effects than the western regions. In the eastern regions, rural income has been raised by 544 RMB a year. In terms of inclusiveness, energy infrastructure improvement helps to narrow down the income difference by 25%. The heterogeneous effect is marginal as the interactions of E with schooling year and with work experience are not significant at the level of 10%.
In the middle region, the general growth effect is significantly positive at the level of 10% and is about 20 RMB a year, while the inclusive effect is not significant at the level of 10%. Schooling year and work experience are statistically positive at the 5% and 1% levels, respectively, which indicates that improvements of energy infrastructure, education and work experience motivate income growth.
The general growth effect is about 10 RMB a year in the western region. Energy infrastructure improvement helps to narrow down the income difference by 2.2%. The heterogeneous effect is marginal as the interactions of energy infrastructure improvement with schooling year and with work experience are not significant at the level of 10%. However, schooling year and work experience are positive significant at the level of 1%.   Mechanism analysis. The mechanisms between income inclusive growth and energy infrastructure improvement are analyzed from the national and regional perspectives.
Firstly, BMI and Work are not significant at the level of 10% in with schooling year is significant at the level of 10%, which indicates that the better educated will benefit more from energy infrastructure improvement through Work than their counterparts.
The results of regional mechanism analyses are reported in Table 3 (columns 4-9).
In the eastern region, E and    In the western region, E is significant at the level of 5%. Schooling year is still significant at the level of 1% while work experience is not significant at the level of 10%, which suggests that schooling year plays a more important role than work experience when energy infrastructure improvement prompts income inclusive growth through abundant working time.

Discussions
The results obtained at the national level reveal that energy infrastructure improvement promotes income inclusive growth in rural China. We also find that individual schooling year and work experience improve income more for low-income group than for high-income group, which indicates that promoting education development and skill training are crucial to income growth for low-income group. The results on reverse causality analysis are consistent with those reported in Table   1. The interaction of schooling year with energy infrastructure improvement still has a positive effect on inclusive growth after a combination of macro-micro data.
The instrumental variable method is further used to address the endogeneity issue.
The results confirm the existence of inclusive growth effect of energy infrastructure improvement. The inclusive effect between energy infrastructure improvement and resident income is robust.
The empirical findings show that improving energy infrastructure, with access to modern energy sources, contributes to promoting income growth and narrowing Thirdly, as we have found the existence of regional heterogeneity, each region shall design its own appropriate policies tailored with the local characteristics. There is a huge difference on energy structure across regions. Therefore, it is necessary to take into consideration the regional resource endowment for the development planning.
For the area with easy accessibility to clean fuels like the eastern China, it is essential to consider how to improve the utilization efficiency of energies further and reduce the cost of energies. Under exploited area like the western China shall focus more on the first access to electricity, liquefied natural gas, and natural gas. Inclusive growth in area with rich resource endowment of solid fuels like the middle China should enhance the access to electricity and natural gas while pay attention on resident health.
We acknowledge that the indicator measured energy infrastructure improvement includes only household cooking fuels. Besides, our mechanism analyses provide limited information about the transmission mechanism between inclusive growth effect and energy infrastructure improvement.
Further research could measure energy infrastructure improvement from multi indicators in case of data availability. As the inclusive growth effect is only discussed between rural residents, it could further investigate the effect of energy infrastructure improvement on income inclusive growth between residents in urban and rural areas.
We focus on the mediating effect in mechanism analyses. The further discussion on moderating effect would be valuable. Another area of future study is whether public infrastructures such as power station, public transportation and motorway prompt income inclusive growth.
Our quantitative results provide references and inspirations for inclusive development, which can be applied by national and regional policy-makers to identify the essential development areas. The proposed approach lays a foundation for estimating the inclusive growth effect of energy public provision. Moreover, the holistic analysis presented provides insights into the national and regional transmission mechanisms of energy infrastructure improvement aiming at improving income inclusive growth. This helps to obtain potential drivers of sustainable and inclusive growth, and consequently contributes to achieving SDGs.

Methods
Inclusive growth effect model. This paper extends the traditional Mincer model22 to quantitatively explore the effects of energy infrastructure improvement on income inclusive growth. Inclusiveness is defined as a narrower gap in resident income, compared to the previous year. Therefore, the proposed quantitative estimation is to investigate whether energy infrastructure can promote income growth, and at the same time, whether income gap can be narrowed between the rich and the poor. If the answers to both questions are positive, we could conclude that energy infrastructure improvement indeed promotes inclusive growth.
The overall growth effect of it E on it Y is expressed as: (2) is expanded as: to what extent energy infrastructure improvement prompts income growth through BMI and Work . Eq. (4) is extended to Eq. (7) to examine the transmission mechanisms from national and regional perspectives.  (1989, 1991, 1993, 1997, 2000, income at the constant price of 2015 is used as the explained variable. The Catalog of Polluting Fuels issued by the Ministry of Environmental Protection of China in 2017 stipulates that polluting energy includes fuels that could burn directly for energy consumption such as raw coal, coal cinder and biomass fuels (including fuelwood, straw, and bagasse). If a respondent of a family uses efficient and clean fuels for cooking, including electricity, liquefied natural gas, and natural gas, the family is regarded as improved in terms of energy infrastructure. Our key variable energy infrastructure improvement indicator is recorded as 1. If the family member uses dirty energies (including fuelwood, charcoal, and briquette) for cooking, the family suffers from energy poverty, recorded as 0. Fig. 1   Control variables consist of individual schooling year, individual work experience and its quadratic term, accessibility to rural infrastructure, whether having a child or not and the location of the family38,39. Work experience cannot be measured directly.
We follow Andrews and Buchinsky (2001) to calculate it as the years after normal schoolings. If we use " sch " to indicate the number years in schooling, and seven is the normal age to go to school, the individual work experience is thus (0, 7) max Age sch −− 40.
Rural infrastructure includes electric power, hydraulic engineering, and transportation. According to follow-up survey, the penetration rate of electric lighting in the sample reached more than 90% during 1989~2015, with no significant variations. Therefore, we use the accessibility to water-supply facilities as a proxy for the accessibility to rural infrastructure3.
Data availability. The input data of Fig. 1, the descriptive statistics of all the 3 According to CHNS data, the proportions that rural residents utilize power lamps, kerosene, petroleum (fuel oil), candle and other energies for illumination during 1989~2015 reached 97.23%, 1.15%, 0.11%, 0.11% and 0.09%, respectively.