Disparity of Rural Income in Counties between Ecologically Functional Areas and Non-Ecologically Functional Areas from Social Capital Perspective

Abstract

In China, income disparities between regions continue to widen, especially in rural areas where environmental policies are implemented, where regional development is more underdeveloped and inequality is high. This paper provides an explanation from the perspective of social capital. Based on the panel data of 2077 counties in 2001–2015, this paper finds that the difference in social capital between ecological and non-ecological functional areas is not only from the gap in the total amount but also from the gap in the income effects. Empirical evidence shows that, although there is a positive correlation between social capital and rural income, the difference between the income effects is further caused by the lower level of social capital in ecological functional areas than in non-ecological functional areas. It is proved that there is a gap between the income effects of social capital in ecological function areas and non-ecological function areas, especially among the low-income groups of the two sectors. The results of the further decomposition of the differences show that the total difference in rural income between ecological function areas and non-ecological function areas is about 40%, of which the contribution of social capital is greater than the contribution of the two sectors. Therefore, the national key ecological functional areas need to explore new models for poverty reduction through social capital.

1. Introduction

After 40 years of rapid growth since 1978, China has become the second largest economy in the world, with far greater inequality than the United States [1]. The income gap between regions continues to widen, especially in rural areas, where it is much greater than in urban areas [2,3,4,5].

Against this context, a special phenomenon is that regional development in rural areas where environmental policies are implemented is more backward and inequality is higher [4,6]. An intuitive explanation is that most of the areas where environmental policies are implemented are located in ecologically vulnerable areas. In China, there is a high degree of geographic overlap between ecologically fragile and impoverished areas [7]. For developing or transition economies, the poor are highly dependent on ecological resources for their livelihoods [8]. Rural poverty is closely linked to environmental degradation [9,10,11,12].

Both the MEA [13] and the Economics of Ecosystems and Biodiversity (TEEB [14]) are concerned with the importance of increasing ecosystem services for poverty reduction through eco-conservation policies. In China, the construction of national key ecological functional areas with “ecological protection” as its main theme is taken seriously by the national strategy. In the Outline of the 11th Five-Year Plan for National Economic and Social Development, it was made clear that the construction of important ecological functional areas is an important task to promote the formation of the main functional areas. Subsequently, the Outline for Key Ecological Function Conservation Zone Planning (2007) and National Main Functional Area Plan (2010) classified the national key ecological functional areas as ecological protection and restricted open areas. In recent years, the coverage of key national ecological functional areas has been increasing, with the provinces enjoying ecologically transferred payments having full coverage by 2016, increasing from 436 counties in 2010 to 818 in 2018.

However, most poor areas are located in key ecological functional areas [7], where there are trade-offs between ecological conservation and economic growth. According to the list of counties with new transfer payments in the national key ecological functional areas in 2016 issued by the Ministry of Environmental Protection and the Circular of the State Council on issuing the “Ten-Three-Five-Year Plan” for the eradication of poverty, there is a high overlap between the key ecological function areas and poor areas. The vast majority of the areas covered by the key ecological function areas are old revolutionary areas, ethnic areas and frontier areas. Of these, 434 counties belonged to the National Poverty Alleviation Counties in 2016, accounting for 52.16% of the total number of key counties in poverty alleviation. While protecting the ecological environment, these areas have also lost many economic development opportunities and suffered significant economic costs. This is not special, and similar problems exist in other regions, such as nature reserves (NRs). Ecological conservation in nature reserves is often considered an excessive restriction on the social and economic development of surrounding communities [15]. The most common conflicts are related to interests such as the limited or unequal use of natural resources within NRs [15,16,17]. The development dilemma of ecological functional areas lies in the double constraints of the fragile ecological environment and the restrictions of ecological protection policy.

Although the Chinese government has introduced a series of ecological compensation policies for ecological conservation areas, it cannot prevent the gap between ecological and non-ecological functional areas from continuously widening. Since the promulgation by the Ministry of Finance of the “National Key Ecological Function Zone Transfer Payment (Pilot) Approach (2009)” [18], the efforts of the central government to transfer payments to national key ecological function zones have increased year by year, from CNY 24.9 billion in 2010 to CNY 72.1 billion in 2018. However, according to the statistical data of the China County Statistical Yearbook, in 1997–2015, the average per capita net income of rural areas in key ecological functional areas of the country was only CNY 3043.838, well below the national county average of CNY 4150.507, and the gap between the per capita income of rural areas in non-ecological functional areas increased from CNY 778.575 in 1997 to CNY 3372.587 in 2015. What causes the gap between ecological function areas and non-ecological function areas to widen? Is it the fragile ecological environment or the factor of ecological protection policy? Despite the improvement in the local ecological environment and the implementation of the ecological compensation policy, why has the gap not been narrowed?

The existing literature is limited in the study of the widening of the rural income gap between ecological protected areas and non-ecological functional areas in China. With regard to the drivers of regional differences, the role of capital and labor mobility, geography, industrial structure, agglomeration economy, nature, institutions, national policies, globalization, institutional reform, etc., have been discussed in the existing literature [19,20,21,22,23,24,25,26,27,28]. However, as far as we know, much of the existing literature on inequality in China is concentrated between regions, cities or urban and rural areas [26,27,28,29], with little attention paid to the rural gap between ecological and non-ecological reserves in China.

This paper aims to analyze the role of social capital in the formation of rural income difference between ecological and non-ecological functional areas in China. Statistics have shown that ecological conservation policies alone do not really solve the persistent problem of rural poverty in ecological functional areas. In China, the existence of long-term poverty is complex, and its creation is not the result of a single factor, but of a long-standing urban–rural binary structure and natural and social factors that constrain rural development [30,31]. Poverty refers not only to the scarcity of material, social and cultural resources, but also to the lack of capacity, opportunities and access to social services [31,32,33]. Social capital, as resources embedded in the individual or collective social networks [34,35,36,37,38], relates to the ability of individuals or groups to obtain benefits, and is closely linked to poverty [39,40,41,42,43]. Especially in rural areas of China, the formal system is inadequate [44], and the social capital of name lineage groups plays an important role in resource allocation [45,46,47,48].

The research contribution of this paper is of display degree. Despite the multifaceted benefits of enhanced social capital for rural development [37,49,50,51,52,53], significant regional differences in social capital [54,55] result in regional economic imbalances [56]. In this paper considering the role of social capital, the difference between rural income in ecological and non-ecological functional areas is decomposed. There is disagreement among scholars as to whether there is a wider income gap between regions or within regions in the contribution of rural income differences in China. Yao [57] considers that the inter-provincial income gap accounts for 75% and the inter-regional income gap is less than 25%, while Gustafsson and Shi [58] consider that the inter-provincial income gap is more significant. Based on the approach proposed by Melly [59], we take social capital as an important factor in the difference in rural income between the two sectors. Melly [59] proposed a method for decomposing the differences in quantiles of the unconditional distribution, which can separate the difference at each quantile of the unconditional distribution into two parts: coefficients’ difference and characteristics’ difference. Through decomposition, we divide the differences of rural income into two parts, sectoral difference and social capital difference, and then observe the role of social capital in the differences in rural income between the two sectors.

The remainder of this paper is structured as follows. Section 2 presents a literature review. Section 3 describes the materials and methods. Section 4 reports the main empirical results. Section 5 presents the conclusions and implications.

2. Literature Review

2.1. Rural Income Inequality in China

In the present literature, the research on the income gap of rural residents is mainly analyzed from the stylized factors of income structure, growth factor, system and policy [60]. Among them, Wan [61] proposed a regression-based inequality decomposition framework to quantify the contributions of various determinants to overall rural inequality in China. Recently, Zhang et al. [62] examined the impact of ecosystem service payment policies and other factors on income distribution and inequality among rural households in China. In this study, the income distribution and the causes of inequality of rural households under the Conversion of Croplands to Forest Program and Ecologic Welfare Forest Program were analyzed. The results of the study show that the income inequality of households participating in the Conversion of Croplands to Forest Program is higher than that of households not participating in the Conversion of Croplands to Forest Program, whereas the impact of the Economic Welfare Forest Program is not significant. Other factors, such as local non-agricultural work, emigration with remittances, human capital, natural capital and material capital, all play an important role in income and inequality. In another paper, Gao et al. [63] used the county per capita net income data to diagnose the factors that cause an unequal income distribution in rural China. They found that the most unequal areas tend to be geographically agglomerated, and the spatial and temporal differences in rural inequality are deeply rooted in the quadruple-transition process of globalization, decentralization, marketization and urbanization. They argued that human investment, rather than economic growth, is a key factor in reducing rural inequality in the eastern provinces.

As most rural areas in China are located in county-level areas, the differences in rural income in county areas have also gained the attention of scholars. For example, using the microdata of the household income surveys in the provinces of China, Gustafsson and Shi [58] studied income inequality within and across counties in rural areas of China. They found that, in 1995, the inequality of rural income in China mainly manifested in spatial inequality, and the uneven development of average income across counties is the main reason for the rapid increase of income inequality. Recently, He et al. [64] analyzed the spatial and temporal patterns of China’s county economy and discussed the multi-mechanism process of inequality in rural areas of China. Based on panel data from 2076 counties, 338 municipalities, 31 provincial units and 4 districts in 2000–2015, the study concluded that the multi-mechanism process of inequality in rural areas in China showed strong regional differences and temporal effects. Among them, urbanization and fiscal transfer in the central and northeast regions are negatively correlated with county economy. The positive impact of bank lending is statistically significant in all regions, while the inhibition of physical factors diminishes over time.

2.2. Social Capital

Although the definition of social capital exists in the perspective of individual and collective theory [34,37,65,66,67,68], the concept of social capital as resources embedded in social networks is clear. At the individual level, these resources can be emotional, informational, instrumental or appraisal supports [69]. At the collective level, resources are non-exclusive and targeted towards achieving a common goal [37], which can lead to instrumental returns, such as better government performance, or expressive returns [70]. These resources can be divided into different elements. As Woolcock [71] pointed out, the definition of social capital is based on different sociological traditions, with common infrastructure elements, namely, social interactions and connections [72], and cultural elements, namely, commonality of purpose, reciprocal norms, trust, civic participation and learning [73], promoting collective action and cooperation to achieve common goals. Social capital can also be divided into structural and cognitive dimensions. Structural social capital refers to the foundation, composition and participation of networks and institutions. Structural social capital manifests itself in different dimensional forms, such as bonding, bridging [74] or linking [71], which can also be divided into horizontal and vertical social capital [75]. Cognitive social capital refers to the perception of values, attitudes and social norms, such as trust and reciprocity [76,77].

2.3. Social Networks and Social Capital

In essence, human society consists of various social networks of linkages. The social network is an important dimension of social capital, essential for information exchange and confidence-building and social norms [78]. From Rostila [70], social capital is generated in networks of social linkages based on trust and reciprocity that bring social resources to individuals or societies. According to Hartmann and Herb [79], social capital has evolved from previous structures that may be dominated by strong linkages, namely, “soft infrastructure for intangible assets” [80]. In self-supporting and semi-subsistence economies, strong linkages are the basis.

2.4. Social Capital and Income Inequality

Lin [81] has earlier studied the theoretical mechanism of social capital affecting income inequality, suggesting that there are two channels through which social capital can contribute to income inequality: capital deficiency and lack of returns. The first is that the poor have less social capital than the rich and are in a vulnerable position in the distribution of resources such as wealth. Because of the positive correlation between social and human capital [36,82], in particular the positive impact of education on social trust and participation [83] and the direct link between human capital and income and wealth, the poor will have less social capital. Mogues and Carter [84] also theoretically examined the important role that inequality in social capital plays in income inequality, noting that social capital (such as blood, geography or industrial ties) can be considered as an intangible asset or collateral in the case of incomplete formal markets, providing more opportunities for the owners of social capital to increase their income, and therefore, given an initial economic polarization and wealth inequality, an inequality in social capital will trigger further income or wealth inequality. Research by Chantarat and Barrett [85] revealed high costs of access to social capital, resulting in mechanisms of exclusion for the poor and preventing some poor families from using social network capital. The potential for the improvement of social networks depends fundamentally on the wider distribution of socio-economic wealth in the economy, which further leads to inequitable social capital. The second is that, due to different mobilization strategies, action efforts or institutional responses among groups give rise to a certain amount of social capital that produces different returns for different individuals. Whether the returns on social capital of the low-income groups of farmers are higher or lower than those of the rich may depend mainly on the comparison of forces in both directions: on the one hand, for social capital as an input element [39], the rate of return on social capital will decrease with the increase in stock if the rule of declining marginal output is established. If the poor have less social capital, the poor are likely to have higher returns than the rich. On the other hand, as measured by the three-dimensional criteria proposed by Lin [81], the poor lack high-quality social capital, and the poor have access to and use few social resources than the rich [86], so that social capital may also pay off less for the poor than for the rich.

Because of the influence of social capital variables through social networks [87], the role of social networks in the impact of social capital on income inequality has also gained the attention of scholars. According to the structural theory of social capital, the vulnerable position of families in social networks could affect their ability to access resources [88]. Campbell et al. [89] found that individuals at the bottom of the social fabric were also deprived of high-quality social networks, resulting in poor people being disadvantaged [90,91]. According to Pearlin [92], Willmott [93], Cattell [94], etc., high-income and middle-income farmers are more likely than low-income farmers to mobilize resources through social participation networks. Some more detailed studies explained the role of social networks in income inequality. Calvó-Armengol and Jackson [95] explained the impact of social networks on income differences through future employment rates and prospects by analyzing the differences in initial employment costs and initial employment rates. A theoretical analysis by Mckenzie and Rapoport [96] found a U-shaped relationship between regional migration rates and income inequality in the region, the formation of which depended primarily on social networks. Limited by the initial high cost of migration, only wealthy families have the opportunity to move outward, thereby exacerbating income inequality in the region. However, when social networks of places of entry are formed, subsequent migration costs will be reduced and poor households will be able to migrate and increase non-farm income, ultimately reducing income inequality in their areas. They also used MMP and ENADID data to validate the upside-down U-curve between migration rates and income inequality, and found that the curvature was mainly determined by the community’s level of development in the migration network. Foltz et al. [44] analyzed the role of Chinese lineage networks in alleviating rural income inequality and found that descent networks increase migration for all social groups by reducing costs, a pattern that is more prominent for the poor. As a result, these populations have accumulated more wealth, thereby reducing income inequality in their villages of origin.

The literature is instructive in studying the rural income gap in county areas of China and the theoretical mechanism and empirical effect of social capital on income inequality. However, there is still a gap in the research on why there is long-term poverty in the rural areas of ecological functional areas in China, and why the gap between ecological functional areas and non-ecological functional areas is widening. Due to the important role of social capital in resource allocation in rural areas of China, the research on the role of social capital in the gap between rural and non-ecological functional areas in China needs to be further enriched. This paper aims to provide theoretical and empirical explanations for the difference in rural income between ecological and non-ecological functional areas from the perspective of social capital in order to make up the shortage in the existing research.

3. Materials and Methods

3.1. Methods

3.1.1. The Standard Quantile Regression Model

In this article, we draw on Koenker and Basset [97], consider regional differences and set the quantile regression model as follows:

$$Log({Income}_i^{\tau })=S{K}_{it}{\beta }^{\tau }+D_i{\delta }^{\tau }+{\epsilon }_i$$

(1)

$$Q_{\epsilon i}^{\tau }=0$$

(2)

where $Income$ is the level of rural per capita income in area $i$ ($i=1,2$) in $t$ period. $S{K}_{it}$ is the social capital variable. $D_i$ is the virtual variable; if $D_i=1$, $i$ is the ecological function area; if $D_i=0$, it is the non-ecological function area; it reflects the sectoral difference. $\epsilon$ is the residual item under the $\tau$ quantile, $Q_{\epsilon i}^{\tau }$ is the quantile of the residual.

Given a random sample ${(Income)}_{n=1,2,\dots ,N}$, follow the Koenker and Bassett [97] solution:

$$({\widehat{\beta }}^{\tau }{\delta }^{\tau })=\arg {\min }_{\delta \in R}K{\displaystyle {\sum }_{i=1}^N{\rho }_{\tau }}(LogIncom{e}_i-scapita{l}_{it}\beta -D_i\delta )$$

(3)

where ${\rho }_{\tau }(\mu )={\tau }_{\mu }{I}_{[0,\infty )}(\mu )-{(1-\tau )}_{\mu }{I}_{(-\infty ,0]}(\mu )$ is the check function and $I(\cdot )$ is the indicator function.

3.1.2. Decomposition of Differences at Different Quantiles

According to the decomposition method proposed by Melly [59], the total difference in rural income between ecological and non-ecological functional areas can be obtained by three estimation steps. In the first step, the conditional distribution is obtained by quantile regression. In the second step, the conditional distribution is integrated over the social capital variables of all the samples, to obtain the estimated non-conditional distribution. In the third step, we estimate the $\tau$ quantile under the counterfactual social capital condition distribution. Through the three-step estimation steps above, the total difference at each quantile of the unconditional distribution can be decomposed into two parts: the coefficient difference and the characteristics difference.

$$\begin{array}{l}q(\tau ,S{C}^{\mathrm{Eco}},{\beta }^{\mathrm{Eco}})-q(\tau ,S{C}^{\mathrm{Neco}},{\beta }^{\mathrm{Neco}})=\\ [q(\tau ,S{C}^{\mathrm{Eco}},{\beta }^{\mathrm{Eco}})-q(\tau ,S{C}^{\mathrm{Eco}},{\beta }^{\mathrm{Neco}})]+[q(\tau ,S{C}^{\mathrm{Eco}},{\beta }^{\mathrm{Neco}})-q(\tau ,S{C}^{\mathrm{NEco}},{\beta }^{\mathrm{Neco}})]\end{array}$$

(4)

where the first part on the right in Equation (4) indicates the coefficient difference, that is, the sectoral difference, and the second part is the social capital difference. The greater proportion of social capital in the total difference means that income differences between the two sectors are mainly due to the uneven distribution of social capital in the region. Conversely, sectoral difference contributes more to the formation of total difference.

3.2. Data Collection

The key ecological function area refers to a restricted development area in the planning of the main functional areas of China. It aims to limit economic development activities in order to protect biodiversity and reduce the impact of ecological vulnerability. National key ecological function areas account for 76.52% of the total number of poverty-stricken areas in the main functional zones of poor areas [7]. Since the Main Functional Area Plan (2010) included 436 counties in the national key ecological functional areas, the list of counties in the national key ecological functional areas was not adjusted until 2015. However, this list may omit some of the counties that have fallen behind due to restrictive development policies. We note that since 2016, the transfer payment of key national ecological functional areas has expanded to 818 counties by 2018, essentially covering the ecological conservation counties of the entire country. In order, as far as possible, to avoid research deviation caused by sample selection, we take the 2018 list of key ecological function areas as the basis for the study of ecological function areas. In this way, 2077 counties can be divided into two groups: counties of ecological functional areas and counties of non-ecological functional areas. So, the paper selects the panel data of 2077 counties from 2001 to 2015 as the sample of the research. The data are from the China County Statistical Yearbook. Regarding the division of counties into national key ecological functional zones, reference is made to the policy provisions of the National Development and Reform Commission and the Ministry of Finance of the People’s Republic of China on national key ecological functional zones.

3.3. Variables

3.3.1. Rural Income

County-level per capita rural income for 2001–2015 is provided in the China County Statistical Yearbook (2002–2016) [98].

3.3.2. Social Capital

The divergence of indicators used in the literature to measure social capital has brought some complexity to the empirical study of social capital. To measure social capital, scholars adopt some aspect of social network resources, such as citizen norms [37,99], trust [100,101], information level, or synthetic index [102,103,104,105]. At the regional level, social capital (SC) is measured by substitution indicators. In this paper, we draw on Temple and Johnson [106], Ishise and Sawada [107], etc., and consider information-sharing and communication as important features of social capital. We estimated a logarithmic value of regional telephone users and used it as an alternative indicator of social capital. Specifically, this paper adopts the method of econometric estimation, using the logarithmic value of regional telephone subscribers (10,000 households) as an explanatory variable, and the variables of the relevant factors affecting social capital as explanatory variables, thus obtaining a fitted value of social capital to reflect the level of social capital in the region, as shown in Figure 1:

In Figure 1, there is a significant difference between the nuclear density estimation curve of the value of the ecological functional areas and the non-ecological functional areas in 2001–2015, and the social capital level of the ecological functional areas is significantly lower than that of the non-ecological functional areas. This gives intuitive evidence of the difference in social capital that leads to the income gap between the two regions.

4. Results and Discussion

4.1. Quantile Regression Results

In order to eliminate the endogenous problem, we adopt the Hausman [108] test method, take the number of telephone users as the explanatory variable, take the rural income as the interpretive variable for OLS regression and take the estimated social capital level variable as the tool variable for 2 SLS estimation. The tested score chi2 = 0.171 indicates that the estimated social capital level variable meets the exogenous condition. And Shea’s partial R² = 0.041 and p value 0.000 mean that the social capital level variable has a significant correlation with rural income.

Based on (1), (3), the quantile regression test is performed.

Table 1. Income effects of social capital in eco-functional areas.

Year	Quantiles
	0.10	0.20	0.30	0.40	0.50	0.60	0.70	0.80	0.90
2001	0.216 ***	0.184 ***	0.168 ***	0.148 ***	0.163 ***	0.158 ***	0.175 ***	0.190 *	0.088
	(0.034)	(0.036)	(0.031)	(0.034)	(0.041)	(0.042)	(0.046)	(0.106)	(0.179)
2002	0.227 ***	0.234 ***	0.238 ***	0.246 ***	0.274 ***	0.275 ***	0.258 ***	0.248 ***	0.078
	(0.036)	(0.031)	(0.027)	(0.039)	(0.045)	(0.044)	(0.040)	(0.051)	(0.068)
2003	0.241 ***	0.203 ***	0.206 ***	0.223 ***	0.230 ***	0.223 ***	0.236 ***	0.242 ***	0.096
	(0.052)	(0.029)	(0.026)	(0.030)	(0.028)	(0.043)	(0.058)	(0.031)	(0.108)
2004	0.199 ***	0.210 ***	0.214 ***	0.231 ***	0.241 ***	0.248 ***	0.287 ***	0.282 ***	0.150 ***
	(0.037)	(0.022)	(0.015)	(0.018)	(0.013)	(0.023)	(0.029)	(0.036)	(0.024)
2005	0.181 ***	0.193 ***	0.183 ***	0.178 ***	0.196 ***	0.223 ***	0.248 ***	0.243 ***	0.159 ***
	(0.037)	(0.028)	(0.019)	(0.013)	(0.015)	(0.016)	(0.026)	(0.027)	(0.027)
2006	0.169 ***	0.182 ***	0.166 ***	0.170 ***	0.170 ***	0.190 ***	0.257 ***	0.204 ***	0.123 ***
	(0.035)	(0.020)	(0.017)	(0.017)	(0.013)	(0.016)	(0.030)	(0.046)	(0.027)
2007	0.251 ***	0.235 ***	0.225 ***	0.225 ***	0.211 ***	0.234 ***	0.250 ***	0.179 ***	0.148 ***
	(0.039)	(0.039)	(0.027)	(0.022)	(0.025)	(0.038)	(0.075)	(0.038)	(0.036)
2008	0.196 ***	0.205 ***	0.190 ***	0.182 ***	0.195 ***	0.185 ***	0.216 ***	0.192 ***	0.159 ***
	(0.040)	(0.034)	(0.019)	(0.014)	(0.022)	(0.027)	(0.038)	(0.045)	(0.044)
2009	0.230 ***	0.190 ***	0.173 ***	0.178 ***	0.199 ***	0.184 ***	0.203 ***	0.218 ***	0.134 ***
	(0.026)	(0.041)	(0.020)	(0.021)	(0.035)	(0.039)	(0.036)	(0.041)	(0.046)
2010	0.241 ***	0.154 ***	0.155 ***	0.169 ***	0.170 ***	0.171 ***	0.176 ***	0.200 ***	0.151 ***
	(0.038)	(0.037)	(0.012)	(0.012)	(0.024)	(0.027)	(0.027)	(0.049)	(0.032)
2011	0.217 ***	0.188 ***	0.157 ***	0.152 ***	0.151 ***	0.113 ***	0.107 ***	0.143 ***	0.074
	(0.051)	(0.033)	(0.025)	(0.030)	(0.037)	(0.036)	(0.038)	(0.040)	(0.066)
2012	0.239 ***	0.199 ***	0.160 ***	0.172 ***	0.156 ***	0.119 **	0.124 ***	0.134 ***	0.056
	(0.025)	(0.039)	(0.022)	(0.027)	(0.030)	(0.048)	(0.041)	(0.046)	(0.065)
2013	−0.075	0.064	0.086	0.134 ***	0.137 ***	0.127 ***	0.138 ***	0.126 **	0.179 ***
	(0.127)	(0.093)	(0.073)	(0.041)	(0.041)	(0.038)	(0.046)	(0.046)	(0.048)
2014	−0.057	−0.057	−0.057	−0.107	−0.107	−0.107	0.040	0.041	0.041
	(0.057)	(0.057)	(0.086)	(0.090)	(0.089)	(0.077)	(0.083)	(0.071)	(0.064)
2015	−0.045	−0.045	−0.021	−0.021	−0.083	−0.003	0.031	0.031	0.031
	(0.061)	(0.061)	(0.063)	(0.071)	(0.084)	(0.080)	(0.068)	(0.256)	(0.257)

Note: The values in () are standard deviation. The asterisk (*), (**) and (***) indicate that the parameter is statistically significant at the 1%, 5% and 10% level, respectively.

and

Table 2. Income effects of social capital in non-ecological functional areas.

Year	Quantiles
	0.10	0.20	0.30	0.40	0.50	0.60	0.70	0.80	0.90
2001	0.422 ***	0.419 ***	0.418 ***	0.403 ***	0.378 ***	0.363 ***	0.356 ***	0.326 ***	0.287 ***
	(0.023)	(0.016)	(0.011)	(0.012)	(0.013)	(0.014)	(0.014)	(0.028)	(0.013)
2002	0.419 ***	0.414 ***	0.415 ***	0.412 ***	0.384 ***	0.365 ***	0.359 ***	0.326 ***	0.291 ***
	(0.017)	(0.012)	(0.014)	(0.010)	(0.013)	(0.014)	(0.011)	(0.021)	(0.017)
2003	0.433 ***	0.411 ***	0.420 ***	0.414 ***	0.386 ***	0.362 ***	0.348 ***	0.312 ***	0.290 ***
	(0.032)	(0.018)	(0.013)	(0.012)	(0.016)	(0.015)	(0.011)	(0.013)	(0.014)
2004	0.380 ***	0.401 ***	0.391 ***	0.394 ***	0.372 ***	0.346 ***	0.317 ***	0.280 ***	0.273 ***
	(0.015)	(0.018)	(0.011)	(0.012)	(0.011)	(0.011)	(0.013)	(0.014)	(0.016)
2005	0.414 ***	0.404 ***	0.392 ***	0.377 ***	0.353 ***	0.334 ***	0.303 ***	0.281 ***	0.269 ***
	(0.013)	(0.010)	(0.008)	(0.009)	(0.009)	(0.011)	(0.021)	(0.013)	(0.020)
2006	0.406 ***	0.414 ***	0.410 ***	0.372 ***	0.350 ***	0.315 ***	0.283 ***	0.263 ***	0.250 ***
	(0.016)	(0.009)	(0.012)	(0.012)	(0.016)	(0.017)	(0.016)	(0.012)	(0.016)
2007	0.460 ***	0.443 ***	0.435 ***	0.400 ***	0.373 ***	0.333 ***	0.298 ***	0.271 ***	0.242 ***
	(0.017)	(0.012)	(0.009)	(0.011)	(0.011)	(0.016)	(0.013)	(0.013)	(0.016)
2008	0.428 ***	0.406 ***	0.404 ***	0.387 ***	0.368 ***	0.320 ***	0.287 ***	0.273 ***	0.256 ***
	(0.020)	(0.017)	(0.018)	(0.017)	(0.019)	(0.017)	(0.019)	(0.016)	(0.021)
2009	0.421 ***	0.397 ***	0.398 ***	0.379 ***	0.360 ***	0.321 ***	0.291 ***	0.271 ***	0.250 ***
	(0.018)	(0.014)	(0.014)	(0.014)	(0.016)	(0.016)	(0.016)	(0.021)	(0.019)
2010	0.413 ***	0.392 ***	0.371 ***	0.345 ***	0.319 ***	0.297 ***	0.273 ***	0.250 ***	0.222 ***
	(0.018)	(0.018)	(0.013)	(0.013)	(0.014)	(0.012)	(0.012)	(0.016)	(0.014)
2011	0.390 ***	0.375 ***	0.368 ***	0.351 ***	0.335 ***	0.307 ***	0.267 ***	0.252 ***	0.226 ***
	(0.020)	(0.014)	(0.012)	(0.008)	(0.009)	(0.014)	(0.015)	(0.010)	(0.011)
2012	0.385 ***	0.378 ***	0.366 ***	0.354 ***	0.329 ***	0.308 ***	0.276 ***	0.239 ***	0.222 ***
	(0.022)	(0.014)	(0.011)	(0.010)	(0.014)	(0.012)	(0.014)	(0.014)	(0.016)
2013	0.453 ***	0.419 ***	0.374 ***	0.285 ***	0.212 ***	0.210 ***	0.188 ***	0.169 ***	0.153 ***
	(0.098)	(0.048)	(0.052)	(0.056)	(0.050)	(0.045)	(0.044)	(0.038)	(0.042)
2014	0.408 ***	0.413 ***	0.391 ***	0.314 ***	0.236 ***	0.217 ***	0.184 ***	0.162 ***	0.149 ***
	(0.101)	(0.065)	(0.071)	(0.068)	(0.060)	(0.045)	(0.039)	(0.036)	(0.044)
2015	0.598 ***	0.469 ***	0.457 ***	0.406 ***	0.296 ***	0.231 ***	0.183 ***	0.188 ***	0.139 ***
	(0.127)	(0.047)	(0.079)	(0.093)	(0.055)	(0.062)	(0.055)	(0.042)	(0.046)

Note: The values in () are standard deviation. The asterisk (***) indicate that the parameter is statistically significant at the 1% level.

report the regression results of two groups of samples from ecological and non-ecological functional areas. The test results show that social capital has a significant positive effect on rural income, which indicates that the accumulation of social capital has positive significance for the promotion of rural income levels in different regions. However, the income effects of social capital differ between the two regions. The influence effect of social capital on rural income in non-ecological functional areas is higher than that of social capital in ecological functional areas.

In Table 1, the level of income effects in eco-functional areas has decreased over time. In particular, the income effects of social capital in 2002–2006 showed a downward trend under quantiles = 0.2, 0.3…0.8. We note that 2007 is a key node in the changing stage of the income effects of social capital in eco-functional areas. Although the income effects of social capital reached a brief “peak” in 2007, after that, the income effects of social capital showed a more obvious overall downward trend under quantiles = 0.2, 0.3…0.7. This suggests that, even when the Chinese government has implemented ecological poverty alleviation policies, such as transfer payments in key ecological functional areas since 2008, it has not changed the low-income effect of rural social capital but may widen the gap with non-ecological functional areas. In 2001, 2005–2009 and 2012, the income effects of social capital in eco-functional areas decreased under quantiles = 0.2, 0.3, 0.4, and increased under quantiles = 0.5, 0.6, 0.7. This means that in most years, the increase in social capital of the relatively higher-income groups in the eco-functional areas has a greater income effect than that of the middle- and lower-income groups. The income effects of social capital of middle- and low-income groups are “hovering” at the low level. Accordingly, we judge that low-income groups in eco-functional areas face the “low-income trap” caused by inadequate social capital. The judgment is also based on the fact that after 2012, the level of income effects of social capital in the key ecological functional areas has decreased markedly, even having a negative effect, which shows that it is difficult for the key ecological functional areas to get rid of the “dilemma” of the decline in income effects caused by insufficient social capital.

In Table 2, the income effects of social capital in non-ecological functional areas had a brief “rebound” after 2007. However, the income effects of social capital for low-income groups (quantiles = 0.1, 0.2, 0.3, 0.4) have been increasing since 2012, which is significantly different from eco-functional areas. And the income effects of social capital of the high-income group (quantiles = 0.6, 0.7, 0.8, 0.9) has an obvious decreasing trend. The income effects of social capital in non-ecological functional areas decrease with the increase in quantiles, that is, the income improvement effect of social capital of the low-income group is higher than that of the high-income group. It is clear that there is no “bottleneck” of income enhancement due to insufficient social capital for low-income groups in non-functional areas, and that increasing social capital is beneficial to get rid of the low-income dilemma. Therefore, it can be judged that social capital in non-ecological functional areas plays an important role in reducing the gap in income effects between the low-income group and high-income group, which is significantly different from ecological functional areas.

Comparing the income effects of social capital between ecological and non-ecological functional areas, it can be found that there are obvious differences between the two. For the non-ecological functional areas, the income effect of social capital decreases with the increase in income level, which is in accordance with the law of declining marginal compensation of capital in the economic sense. Because of the openness of the formation of social capital in non-ecological functional areas, the flow and free allocation of social network resources is more beneficial to the income of low-income groups. However, the situation of ecological functional areas is special. Although, similar to the non-ecological functional areas, the income effect of the national key ecological functional areas also has the characteristics of decreasing with time, the income effect of social capital of the low-income group is “hovering” at the low level for a long time, caused by the lack of social capital. The social network of the key ecological function areas of the country is semi-closed and cannot freely carry out the flow and allocation of social network resources. The semi-closed nature of social networks will inevitably make it difficult for low-income groups to escape from the lack of social capital, and thus they may fall into the “trap” of low-income levels.

Figure 2 shows the difference in the regression coefficients of social capital on rural income between ecological and non-ecological functional areas. The difference in coefficients highlights negative characteristics; that is, the income effect of social capital in ecological functional areas is lower than that in non-ecological functional areas. Since 2004, especially in the 2005–2010 and 2013–2015 periods, the coefficients difference has increased overall under quantiles = 0.1, 0.2, 0.3, 0.4. It is worth mentioning that the largest coefficient differences are found between 2013 and 2015. The coefficient differences averaged over the 2013–2015 period amounted to −0.55, −0.45, −0.40 and −0.33 for quantiles = 0.1, 0.2, 0.3 and 0.4, respectively. This result predicts the largest differences between the two sectors between 2013 and 2015 among the low-income groups. It is clear that the policies implemented by the Chinese Government on fiscal transfers and payments in the key ecological functional areas have not narrowed the gap in the income effects of social capital between the low-income groups in the two sectors, but rather the difference is more obvious. In relative terms, the regression coefficient of social capital under quantiles = 0.5, 0.6, 0.7, 0.8 is small and decreases with the increase in quantiles, which shows that the income effect of social capital between high-income groups in the two sectors is small. However, there is no significant sign that the difference decreases. The years 2013–2015 are used as an example for illustration. When the quartile = 0.5, the value of the coefficient difference increases from 0.075 in 2013 to 0.379 in 2015; when the quartile = 0.6, the value of the coefficient difference increases from 0.083 in 2013 to 0.234 in 2015; when the quartile = 0.7, the value of the coefficient difference increases from 0.050 in 2013 to 0.152 in 2015; and when the quartile = 0.8, the value of the coefficient difference changes from 0.043 in 2013 to 0.157 in 2015.

The conclusion shows that there is not only a gap in the amount of social capital between ecological and non-ecological functional areas, but also a widening gap in the income effects of social capital, especially between low-income groups. The social network of eco-functional areas is semi-closed, and it is at a disadvantage in the competition with non-eco-functional areas to seize resources, which further exacerbates the problem of insufficient income effects caused by the lack of social capital in eco-functional areas.

4.2. Decomposition of Differences at Different Quantiles

Table 3. Decomposition of rural income difference.

Year	Decomposition of Differences	Quantiles
		0.10	0.20	0.30	0.40	0.50	0.60	0.70	0.80	0.90
2001	Social capital difference	0.479	0.481	0.453	0.392	0.323	0.266	0.242	0.194	0.206
	Sectoral difference	0.079	0.114	0.146	0.174	0.149	0.156	0.129	0.131	0.086
	Total difference	0.558	0.595	0.599	0.566	0.472	0.422	0.371	0.325	0.292
2002	Social capital difference	0.344	0.346	0.35	0.351	0.348	0.344	0.339	0.331	0.317
	Sectoral difference	0.034	0.062	0.079	0.093	0.103	0.116	0.126	0.138	0.147
	Total difference	0.378	0.408	0.429	0.444	0.451	0.46	0.465	0.469	0.464
2003	Social capital difference	0.358	0.363	0.364	0.363	0.358	0.356	0.347	0.34	0.326
	Sectoral difference	0.032	0.066	0.088	0.102	0.116	0.128	0.144	0.162	0.173
	Total difference	0.39	0.429	0.452	0.465	0.474	0.484	0.491	0.502	0.499
2004	Social capital difference	0.437	0.414	0.401	0.39	0.38	0.36	0.345	0.328	0.313
	Sectoral difference	0.015	0.055	0.076	0.093	0.103	0.113	0.125	0.14	0.158
	Total difference	0.452	0.469	0.477	0.482	0.48	0.475	0.47	0.468	0.471
2005	Social capital difference	0.45	0.419	0.397	0.38	0.37	0.35	0.334	0.321	0.31
	Sectoral difference	0.041	0.09	0.119	0.138	0.151	0.165	0.182	0.201	0.221
	Total difference	0.491	0.509	0.516	0.519	0.516	0.515	0.516	0.522	0.531
2006	Social capital difference	0.428	0.402	0.386	0.37	0.35	0.33	0.316	0.302	0.292
	Sectoral difference	0.045	0.103	0.13	0.145	0.154	0.165	0.182	0.205	0.228
	Total difference	0.473	0.505	0.516	0.514	0.506	0.499	0.498	0.507	0.52
2007	Social capital difference	0.375	0.362	0.353	0.34	0.326	0.311	0.295	0.281	0.267
	Sectoral difference	0.076	0.124	0.149	0.166	0.175	0.182	0.194	0.215	0.223
	Total difference	0.451	0.486	0.502	0.506	0.501	0.493	0.489	0.496	0.49
2008	Social capital difference	0.4	0.373	0.356	0.341	0.326	0.309	0.293	0.279	0.27
	Sectoral difference	0.034	0.081	0.106	0.121	0.13	0.139	0.152	0.171	0.188
	Total difference	0.434	0.454	0.462	0.462	0.456	0.448	0.445	0.45	0.458
2009	Social capital difference	0.365	0.35	0.337	0.322	0.308	0.294	0.28	0.268	0.26
	Sectoral difference	0.056	0.092	0.115	0.131	0.141	0.15	0.164	0.181	0.199
	Total difference	0.421	0.442	0.452	0.453	0.449	0.444	0.444	0.449	0.459
2010	Social capital difference	0.32	0.32	0.312	0.3	0.29	0.27	0.258	0.247	0.235
	Sectoral difference	0.106	0.138	0.156	0.17	0.18	0.19	0.204	0.216	0.225
	Total difference	0.426	0.458	0.468	0.469	0.465	0.462	0.462	0.463	0.46
2011	Social capital difference	0.256	0.267	0.271	0.27	0.26	0.25	0.239	0.23	0.224
	Sectoral difference	0.086	0.12	0.142	0.157	0.168	0.176	0.184	0.201	0.222
	Total difference	0.342	0.387	0.413	0.424	0.427	0.426	0.423	0.431	0.446
2012	Social capital difference	0.252	0.272	0.275	0.27	0.26	0.25	0.242	0.231	0.224
	Sectoral difference	0.101	0.123	0.14	0.153	0.162	0.17	0.175	0.189	0.207
	Total difference	0.353	0.395	0.415	0.424	0.424	0.422	0.417	0.42	0.431
2013	Social capital difference	0.263	0.277	0.273	0.28	0.288	0.282	0.292	0.306	0.321
	Sectoral difference	0.182	0.136	0.136	0.127	0.102	0.096	0.086	0.071	0.076
	Total difference	0.445	0.413	0.409	0.407	0.39	0.378	0.378	0.377	0.397
2014	Social capital difference	0.384	0.348	0.345	0.31	0.25	0.23	0.242	0.202	0.186
	Sectoral difference	0.097	0.204	0.202	0.217	0.225	0.222	0.193	0.193	0.175
	Total difference	0.481	0.552	0.547	0.523	0.471	0.454	0.435	0.395	0.361
2015	Social capital difference	0.479	0.481	0.453	0.392	0.323	0.266	0.242	0.194	0.206
	Sectoral difference	0.079	0.114	0.146	0.174	0.149	0.156	0.129	0.131	0.086
	Total difference	0.558	0.595	0.599	0.566	0.472	0.422	0.371	0.325	0.292

reports the results of the decomposition of total rural income difference between ecological and non-ecological functional areas. In different years, the difference in total rural income between ecological and non-ecological functional areas is about 40%. From 2001 to 2012, the total difference between ecological and non-ecological areas was above 40%. The “peak” occurred in 2005–2007, and the total difference reached about 50%. There is no significant sign of a decrease in the total difference between ecological and non-ecological areas. However, the total rural income difference between low-income groups in the two sectors increased after 2013. In the composition of total rural income difference between ecological and non-ecological functional areas, the contribution of the social capital difference is higher than that of the sector, which is the main reason for the formation of total difference. Because of the lack of social capital, the endogenous development ability of the national key ecological function areas is lower than that of the non-ecological function areas for a long time. This provides empirical evidence for the explanation of the persistent disparity in rural income between the ecological function areas and the non-ecological function areas.

5. Conclusions and Implications

The core aim of this paper is to provide a social capital perspective on the difference between rural income in China’s ecological function areas and non-ecological function areas. Specifically, this paper first provides a theoretical model to explain the formation of social capital differences between ecological and non-ecological functional areas and its effect on rural long-term income. Then, through quantile regression and decomposing the differences, it provides an empirical basis for the contribution of social capital to rural income differences between ecological and non-ecological functional areas.

Through an empirical estimation, we find that the social capital gap between ecological and non-ecological functional areas results not only from the gap in the total amount but also from the gap in the income effect. The empirical evidence shows that, although there is a positive correlation between social capital and rural income, the difference between the income effects is further caused by the lower level of social capital in ecological functional areas than in non-ecological functional areas. It is proved that there is a gap between the income effects of social capital in ecological function areas and non-ecological function areas, especially among the low-income groups of the two sectors. The income effect of the social capital of the low-income group in the national key ecological function areas has long been “hovering” at the low level, caused by a lack of social capital. The results of the further decomposition of differences show that the total difference in rural income between ecological function areas and non-ecological function areas is about 40%, of which the contribution of social capital is greater than the contribution of the two sectors.

The results indicated that the ecological compensation policies implemented in China do not reduce the rural income gap between ecological and non-ecological functional areas. From the perspective of social capital, we explain that there exists a “double” difference in the total amount and income effects of social capital between ecological and non-ecological functional areas.

The policy implications derived from this research are clear. The national key ecological functional areas need to explore the new model of poverty alleviation of social capital. Through the integration of social network resources of enterprises, universities, governments and social organizations, China should make full use of the ecological resources of the ecological functional areas and develop their capacity for sustainable development with maximum potential. At the same time, it is necessary to embed open Internet resources and develop electric industry, in order to change the inadequacy of the semi-closed network of traditional agricultural society in ecological functional areas.

This paper gives an explanation of the differences in rural income between ecological and non-ecological functional areas in China, which is of significance to the understanding of long-term poverty and regional differences in deeply backward regions of China. However, the research sample is limited to county level, and the lack of specific micro-research data will have a certain effect on the depth of the research, which needs to be further enriched and improved. We believe that, building upon this study, future research will provide more in-depth and objective findings on long-term poverty and regional differences in the deeply backward regions of China, and that the ideas of this paper will also apply to research in other aspects of regional differences, such as wage differences, differences in economic growth, etc.