Measuring Spatiotemporal Changes of Rural Basic Public Service in Poverty-stricken Area of China

Abstract

To address the development of rural basic public services (RBPS) among contiguous destitute areas of China, we develop a comprehensive RBPS evaluation methodology to examine RBPS development level of 728 poverty-stricken counties, using geographical information system (GIS) to describe their multiscale and multidimensional spatiotemporal change during 2010–2012; besides, we also try to reveal how RBPS interacts with county economy (CE) by integrating Tapio model and weighted Voronoi circle-layer structure. Our results show that (1) at a multiscale of area–province–county, in spite of the overall low level, RBPS is steadily growing during 2010–2012, along with a positive spatial autocorrelation and an obvious nonequilibrium that is high in east China but low in west; however, the RBPS gaps among the whole counties are gradually narrowing, shifting their development grades from a mostly relative shortage or relatively severe shortage in 2010 to a main state of relatively richness or relatively equilibrium in 2012, (2) from a multidimensional view, the RBPS gaps among most dimensions of different areas are gradually narrowing, except for the dimension of social public safety service that shows a significant regional differentiation among different areas. RBPS in Tibetan areas is the most unequalled and falls into the most obvious heterogeneity, and (3) there exist weak correlations between county-level original RBPS and original CE for each year and each circle layer, while significantly positive correlation is found only between mean RBPS and mean CE for four circle-layer subsets of counties, respectively; overall, RBPS development level lags behind that of CE as a main result of the weak decoupling between them. This study may provide a good understanding of the status, regional differences, and evolution of RBPS in poverty-stricken rural China, and serves as a scientific reference regarding decision-making in both promoting intrarural antipoverty harmonious development and constructing the new countryside of China.

Keywords

rural basic public services (RBPS)county economy (CE)spatiotemporal pattern changes coupling development Tapio decoupling model circle-layer structure

As the most populated developing country in the world, China has a wide range of rural poor people that scatter in different regions all over the country, leading to significant development gaps among different areas. In recent years, with the deepening of national antipoverty strategies of China, although rural poor population has an obvious decrease, there is still an increasing serious regional disequilibrium in rural China (Gustafsson, Li, and Sicular 2008; Liu, Lu, and Chen 2013; Ye, LeGates, and Qin 2013). It reflected not only in income difference but also in infrastructural construction, health insurance, social security, and farmers’ social status, resulting in that disequilibrium of basic public services in China becomes a well-known fact (Wang and Nie 2011; Liu, Lu, and Chen 2013), of which rural basic public services (RBPS) provision in undeveloped areas is especially such a typical example. As is known, basic public service is the most essential and critical part of public service, and its equalization is the comprehensive policy goal of political, economic, and social development ideal for a country. The government’s RBPS supply and the local economic development are the two entities that interact with and mutually influence each other in regional development. The provision of RBPS, such as clean water and sanitation, health care, schooling, public security and infrastructure, constitutes a basic human right and is an essential ingredient of economic development. Access to these basic services not only improves individual well-being but also serves as an input into aggregate economic activity and national output (Deolalikar and Jha 2013). At a national level, RBPS underpins rural human welfare and economic growth, and a reasonable equilibrium and interactive relationship between them is the fundamental prerequisite to accomplish coordinated development among rural areas for building a well-off society with Chinese style. Especially for current China, it is at the transformation period to change its economic and social model, and establishing an effective RBPS system is the interior requirement of socialism marketing economy and the basic obligation of a service-oriented government.

Chinese grand goal of building a moderately prosperous society in an all-around way requires promoting balanced development of RBPS in poverty-stricken areas. However, the insufficiency and disequilibrium of RBPS supply has represented a restricted constraint on the stability and persistence of antipoverty effects. Against this backdrop, increasing attention to inequality and differentiation from RBPS has been reflected in “China’s rural poverty alleviation and development Outline (2011–2020)” (hereinafter referred to as the “New Outline”), officially issued by the Chinese Government in 2011. The New Outline initiated the national new strategy regional development and priority poverty alleviation in fourteen concentrated contiguous destitute areas that consist of Wuling Mountains Area, Wumeng Mountains Area, Qinba Mountains Area, and so on, and are also considered as the main battleground of priority poverty alleviation in the following ten years. This strategy is different from Chinese previous regional development and poverty alleviation plan in that it takes “driving poverty alleviation by regional development; poverty alleviation facilitating regional development” as basic ideas, taking the equalization level of RBPS among poverty-stricken counties as one of the core indexes to monitor the effectiveness of poverty reduction over the next decade, so as to prompt the equalization development of RBPS. Further in 2015, Chinese government puts forward “lifting all out of poverty” strategy to stress the importance of improving RBPS allocation to help eliminate the nationwide poverty of China. In response to these policy propositions and requirements, objective and comprehensive evaluation of RBPS development has become an important issue for poverty-stricken counties, and it also assists in effectively implementing the New Outline and constructing modern countryside in which RBPS acts as a bridge to link various social and economic infrastructure constructions.

Literature Review

Public service provision has always been an important research focus in geography-related disciplines. From the perspective of qualitative policy making, previous research concentrated on examining how the economic development interact with basic public services (Li, Qin, and Li 2009; Yang and Mu 2012; Liu, Lu, and Chen 2013), as well as on their coordination degree evaluation at a province or municipality scale (Batley and Mcloughlin 2015; Gao et al. 2015). From an integrated view, some other literatures focused on economic evaluation indicators (Niu, Yang, and Bai 2010; Liao, Zhou, and Tang 2014; Zhou et al. 2014) or social public services (Higgs and White 1997; An and Ren 2008). For example, Liao, Zhou, and Tang (2014) took eighty-eight counties and county-level cities in Hunan province as a case test and employed a single economic indicator to empirically analyze the change of county economy (CE) disparity and its influencing factors by employing Theil index and Gini index. Based on the 2,352 counties’ per capita gross domestic product (GDP) indicator, Zhou et al. (2014) investigated the spatiotemporal changes of county economic development and explored the possible mechanism that responds to the changes. Batley and Mcloughlin (2015) addressed the gap among public services by proposing a framework for understanding and comparing the politics of different services, and considered that policy responses can be targeted to address service characteristics where they present opportunities or constraints to better services.

On the one hand, many rural service systems are being reorganized to provide effective, efficient, and geographically accessible services, therefore, some examined the use of GIS methods to build the model of the spatiotemporal accessibility of public service, and gave the evaluation case of a certain dimension of public service, identifying gaps in the accessibility or finding locations that maximize such accessibility to green space (Higgs, Fry, and Langford 2012), transformation infrastructure (Norden 2011; Neutens et al. 2012), sport facilities (Higgs, Langford, and Norman 2015), post office (Higgs and White 1997), education (Wang, Meng, and Ren 2014), rural health care service (Ren, Tong, and Kwan 2014; Shiikha, Kruger, and Tennant 2015; Sanders et al. 2015), and so on. For example, GIS spatial analysis technology was adopted to analyze the changes in service provision from the perspective of spatial accessibility by Higgs and White (1997), Higgs, Fry, and Langford (2012), and Higgs, Langford, and Norman (2015). Shiikha, Kruger, and Tennant (2015) adopted geospatial analysis to target rural and remote dental services shortages. Sanders et al. (2015) assessed the prevalence of rural primary care physician bypass, and used logistic regression and spatial analysis to analyze spatial bypass patterns to determine which rural communities are most affected by bypass.

Meantime, as far as China’s public services is concerned, numerous studies have analyzed China’s public service development in recent years from the multidisciplinary perspectives of geography, economics, and sociology. Almost all these research focused on the equalization on the coordinated urban–rural development (Li, Qin, and Li 2009; Yang and Mu 2012; Liu, Lu, and Chen 2013; Ye, LeGates, and Qin 2013). For example, some researchers measured regional inequality of public services such as education and health care in China (Zhang and Kanbur 2005; Chou and Wang 2009; Feng et al. 2010; Li and Wei 2014). Ma, Han, and Jiang (2011) analyzed the characteristics and spatial difference of basic public services at prefecture level in China. Wang and Nie (2011) explored the contributing deriving factors of disequilibrium from the perspectives of fiscal transfer payment system and tax system. Han, Li, and Zhang (2015) used information entropy and exploratory spatial data analysis to evaluate equalizations of basic public service in urban and rural areas of China’s thirty-one provincial level administrative units, and explored the spatial pattern of the equalization in urban and rural areas. Except that Chi and Wang (2016) only examined the disparity of RBPS in the impoverished county of Wuling Mountain Area, China, however, few of them concentrated on public service’s development quality of rural China, not to mention national poverty-stricken rural areas, resulting in that the reflection of rural development quality cannot be strengthened due to the great development difference between rural and urban areas in China. Besides, the basic public service indicators in these literatures focused on urban–rural coordination development of China at a large province or metropolis scale, rather than special rural areas’ county or township scale. They employed the whole administrative regions’ urban–rural integrated socioeconomic statistic yearbook instead of the rural areas’ special one, and targeted a given dimension of rural public service rather than a multidimensional integrated one. All these lead to the neglect of specific RBPS development for constructing the new countryside of China and eliminating multidimensional poverty, as well as the neglect of special distinctive rural indicators that both the rural development and farmer households draw attention to in reality. What’s worse, these provincial-scale or city-level studies also covered up the diversities of different county-scale RBPS development due to the fact that there are still some lacks in quantitatively evaluating RBPS and revealing their spatiotemporal differentiations.

As a result of the above studies, how about RBPS development quality in rural poverty-deprived areas of China? Does RBPS match with CE or not? Where is better and in which? It is still a mystery to answer the disequilibrium levels of RBPS development in poverty-stricken counties of China. No breakthrough has been achieved in the theory and practice of integrated RBPS in rural China to respond to the New Outline and the new countryside construction.

To address these limitations, our study aims to provide a more comprehensive and detailed RBPS evaluation for the poverty-stricken areas in rural China to meet the needs of both monitoring rural antipoverty effects of the New Outline and constructing the new countryside of China. It emphasizes the integration evaluation, spatiotemporal pattern, and dynamic process from the perspective of spatial poverty. That’s to say, compared to previous studies, however, we select 728 key poverty-stricken counties in China as our study area, designing county-level RBPS comprehensive evaluation model to systematically reveal the multiscale spatiotemporal changes of RBPS during 2010–2012. Then, we adopt weighted Voronoi circle-layer structure and Tapio decoupling model to further analyze the coupling development degree between these counties’ RBPS and their corresponding CE since the implementation of the New Outline. This study may serve as a scientific reference regarding decision-making in the optimal allocation of RBPS resources, as well as in constructing the new countryside of China.

The remainder of the article is structured as follows: we start in the third section by introducing the study area and data sets that will be used in the rest of this article. Next, in the fourth section, we present the rationale and processing issues for the development of our RBPS evaluation method. Following that, in the fifth section, we explore the measurement characteristics of RBPS in poor rural China. The last section summarizes the results of this article and presents some future research suggestions.

Study Area and Data Resource

Study Area

As shown in Figure 1, totally 728 poverty-stricken counties involve in this study and are divided into fourteen contiguous destitute areas in terms of “centralized contiguous” and “special difficult” principles of China antipoverty strategies. These counties have been considered as the main battlefield of poverty alleviation of China according to the New Outline, since the average per capita net income of farmers in these areas is RenMinBi 2,676 yuan, equivalent to only half of the national average. In the lowest-ranked 600 counties, there are 521 counties belonging to these areas, accounting for 86.8 percent.

Figure 1.

Overview of the study area.

These areas are featured by population sunk in poverty. Most of the poor are concentrated in mountainous areas, hilly areas, and restricted development zones, where poverty is persistent, intergenerational, and highly transitive. In these areas, contributing poverty factors are very complicated, coming from natural, social, ethnic, religious, historical, political constraint, and so on.

From a historical point of view, these fourteen areas mostly belong to old revolutionary areas, ethnic minority areas, and frontier areas. From the perspective of physical geography, most of them are covered by Qinghai–Tibet Plateau, desertification area, Loess Plateau and Southwest Rocky, and other harsh natural conditions. In views of antipoverty, after decades of development, the problems of survival, food, and clothing of rural residents in these areas have been basically solved, and the remarkable achievements have been gained in education, health care, public services, environmental protection, and so on; however, there are still several critical problems that can be epitomized by weak infrastructure, social undertakings lagging behind, lack of public services, and insufficiency of industrial development.

Therefore, the task of poverty alleviation in these areas is arduous, since poverty alleviation in these areas will determine the success or failure in China’s national antipoverty strategies. Some comprehensive development and evaluation means are especially needed to lift them out of poverty, according to “contiguous concentrated, focused, national coordinated, administrative zoning integrated” requirements proposed in the New Outline.

Data Resource

Instead of using regular statistic yearbook of China, the authoritative RBPS-related data source in this study are obtained from China’s top official organization of poverty reduction, that is, the State Council Leading Group Office of Poverty Alleviation and Development of China that has not initiated large-scale electronic archiving data collection of villages and counties all over the fourteen continuous destitute areas of China till 2011 to respond to the New Outline. Here, the data consist of special rural antipoverty socioeconomic data sets of 728 poverty-stricken counties, covering a wide RBPS range of culture, education, infrastructure, health, social security, production and living conditions, and so on, at a county-level statistical scale during 2010–2012. In addition, we collect 1:250,000 geographic data of China to assist in GIS-based spatiotemporal analysis for RBPS distribution pattern.

These data are preprocessed before putting into use, for example, by eliminating unreasonable socioeconomic data, spatial data’s georeferencing, clipping, as well as joining spatial and corresponding socioeconomic data.

Study Methods

To address RBPS development of China’s poverty-stricken counties and their spatiotemporal distribution, this article puts forward some effective measures to monitor intrarural RBPS diversities. To be specific, we develop a comprehensive RBPS evaluation methodology to multidimensionally reveal the spatiotemporal diversities of their RBPS at a multiple area–province–county scale during 2010–2012, consisting of the indicator selection, model specification, and multiscale spatial poverty analysis issues in the monitoring of gaps. Then, we introduce weighted Voronoi circle-layer structure and Tapio decoupling model to present coupling development degrees between these counties’ RBPS and their corresponding CE since the implementation of the New Outline.

RBPS Comprehensive Assessment

Evaluation indicators

According to the related literatures, some criteria for selecting RBPS and CE evaluation indicators are stressed here that they should be policy related, easily understood, sensitive, and should be able to be collected in a cost-effective way and on a regular basis (Higgs and White 1997; An and Ren 2008; Martinez 2009; Ma, Shen, and Song 2014; Wang and Qian 2015). To address the criteria, RBPS inequality in this research is considered to be a heterogeneous, multidimensional, and complex phenomenon. So the first step in the selection of RBPS indicators is the identification of a problem perspective, resulting in a social justice perspective (Martinez 2009). Proportional equality of RBPS in poverty-stricken counties is the policy goal inscribed in this perspective, serving the implement of the New Outline. As is stated by policy makers of the New Outline, the provision of RBPS, such as infrastructures, health care, and schooling, constitutes a basic human right and is an essential ingredient of economic development. For examples, roads and telecommunication systems lower transaction costs and hence encourage trade and economic activity; access to publicly provided (or publicly financed) health and educational services is particularly important for the poor, as they do not have the purchasing power to buy these services from private markets, and the provision of health care and schooling increases the quality of human capital, which is an important input in today’s knowledge-based economies (Deolalikar and Jha 2013).

After aiming at the above indicator candidates that are policy relevant and policy applicable, we adopt the official rural socioeconomic antipoverty monitoring data to address the criteria of data availability. Further, we use hierarchical clustering–coefficient of variation (CV) method to address the indicators’ representativeness and significance. Hierarchical clustering is an alternative approach that builds a hierarchy from the bottom-up. It first puts each data point (each county’s certain RBPS attribute) in its own cluster and then identifies the closest two clusters and combining them into one cluster, followed by repeating the above step till all the data points are in a single cluster.

Since different indicator candidate may denote different perspective of RBPS development situation, different dimensions of indicator variables are not comparable. Although quantitative clustering method could not consider the actual meaning of one indicator, the clustering under the framework of same dimension ensures the local comparability of the indicator candidates that are instinctively associated. The interdimension clustering is more significant than randomized clustering among any intradimensions, due to the fact that it could avoid merging of unrelated indicators.

Therefore, in terms of the clustering criterion, we dealt with it from two perspectives of both subjective policy-relative requirement and objective data statistics. On the one hand, the candidate dimensions in evaluating RBMS have to be consistent with China’s RBPS comprehensive monitoring objectives that have been stated in the antipoverty New Outline of China. On the other hand, according to general indicator selection criterion, those candidate indicators in each dimension should respond to the sensibility, reliability, conciseness, and so on. Therefore, during our indicator selection, we firstly make sure that RBPS dimensions cover each important development aspect that central government of China concerns during RBPS-related antipoverty intervention. Then, using hierarchical clustering with R analysis, we classify those indicator candidates within each dimension, merging those indicator candidates with strong correlations into one subgroup. At last, for those indicator candidates within each subgroup, CV value is applied for determining whether an indicator would be removed from its subgroup, and those candidates with the largest CV will be retained.

To be specific, during the indicator selection, the first is for the identification of clusters in a given dimension, we adopt nearest neighbor method in hierarchical clustering to assure that those indicators have most similarities within same cluster, as well as most dissimilarities among different clusters. Then, we deal with those candidates within the same cluster, reserving the candidate with maximum CV value while excluding those with smaller CV. Within one cluster, a given indicator’s higher CV value means its stronger identification. On the other hand, it is possible that each indicator in different cluster may have different CV value that is relatively lower or higher. According to clustering principle that different indicators should have a lowest correlation among different clusters, as well as a largest correlation within the same cluster. At last, their internal consistency is estimated, identifying an acceptable reliability with a Cronbach’s alpha (α) coefficient of .768.

From the above, this article constructed an evaluation indexes system for RBPS, consisting of twenty-one indexes, classified into seven dimensions such as education, culture, public security, social security, health, infrastructures, and environmental protection. See Table 1(A) for further information.

Table 1.

Indicators of RBPS and CE, Respectively, as well as Their Weights.

	Dimension	Indicator	Unit	Hierarchical clustering	CV	Retained or not (Y/N)	Subject weight	Objective weight	Combined weight
(A) RBPS	Rural fundamental education	Ratio of administrative villages having one or more kindergartens or preschools	%	1	0.880	Y	0.118	0.065	0.100
		Ratio of preschool children able to get access to kindergarten	%	2	0.441	Y	0.237	0.042	0.126
		Ratio of 15- to 18-year-old children enrolling in senior high school	%	2	0.351	N
	Rural environment protection	Forest coverage ratio	%	1	0.581	Y	0.044	0.044	0.054
		Ratio of administrative villages having one or more rubbish recycling depots	%	2	1.206	Y	0.009	0.069	0.061
		Ratio of administrative villages having one or more landfill sites	%	3	2.020	Y	0.009	0.040	0.037
		Ratio of administrative villages having one or more full-time cleaners	%	3	1.978	N
	Rural public safety	Ratio of administrative villages having one or more police service rooms	%	1	1.290	Y	0.031	0.097	0.093
		Ratio of administrative villages having one or more community policemen	%	2	1.596	Y	0.031	0.091	0.088
		Ratio of administrative villages safety supervision group	%	2	1.452	N
	Rural public health	Ratio of administrative villages having clinics	%	1	0.327	Y	0.031	0.033	0.040
		Number of hospital beds per thousand people		2	1.014	Y	0.031	0.003	0.014
		Number of health physicians per thousand people		2	0.917	N
	Rural social security	Number of Social Welfare Institutes’ beds per thousand people		1	1.791	Y	0.020	0.008	0.014
		Ratio of per thousand people enrolling in the New Rural Cooperative Medical Insurance of China		2	0.182	Y	0.100	0.010	0.047
		Ratio of per thousand people enrolling in rural social endowment insurance		3	0.812	Y	0.100	0.044	0.076
		Ratio of administrative villages having one or more community service centers	%	4	1.782	Y	0.020	0.062	0.059
		Ratio of administrative villages having fund mutual cooperative	%	1	1.767	N
	Rural public culture	Ratio of administrative villages having one or more sport/culture squares	%	1	1.335	Y	0.006	0.057	0.049
		Ratio of administrative villages having fitness equipment	%	1	1.090	N
		Ratio of administrative villages having access to radio and television	%	2	0.238	Y	0.028	0.019	0.027
		Ratio of administrative villages having access to broadband network	%	3	0.705	Y	0.028	0.068	0.067
	Rural infrastructure	Ratio of administrative villages having access to cement/asphalt road	%	1	0.533	Y	0.075	0.058	0.078
		Ratio of administrative villages having access to electricity	%	2	0.190	Y	0.032	0.015	0.025
		Ratio of administrative villages having access to passenger bus	%	3	0.590	Y	0.014	0.059	0.055
		Ratio of administrative villages having access to safe drinking water	%	4	0.539	Y	0.032	0.057	0.060
		Ratio of administrative villages having agricultural greenhouse facilities	%	5	1.249	N
		Ratio of administrative villages having livestock facilities	%	5	1.623	Y	0.007	0.059	0.052
(B) CE	Economic development	GDP per capita	Ten thousand Chinese Yuan				0.375	0.334	0.371
		Fiscal revenue per capita					0.135	0.142	0.124
		Fiscal expenditure per capita					0.062	0.057	0.072
		Net income per capita					0.225	0.232	0.212
		Disposable income per capita					0.203	0.194	0.221

Note: GDP = gross domestic product; CE = county economy; CV = coefficient of variation; RBPS = rural basic public services.

Similarly, as the economic development evaluation is concerned, since a single indicator cannot systematically account for the real development condition of CE, this article refers to the above indicator selection criterion, selecting the most policy related, the most sensitive ones from the poverty-stricken core monitoring economic indicators, and resulting in a set of CE evaluation indicators that consists of five ones (see Table 1(B)). Of which, the former three ones represent the economic development from the local government’s perspective, while the latter two come from the personal view.

Combined weighting model based on game theory

To keep reasonable weight of each indicator, this article introduces a combined weighting model based on game theory, integrating subjective analytic hierarchy process (AHP) weight and objective entropy one. Essentially, the combined weight model based on the game theory is devoted to finding a consensus or compromise among different weights, and these most reliable weights can be represented in a form of a set of optimized weight value by minimizing the respective deviation between the possible actual weight and various basic weights (Wang and Qian 2015). That is, in this article, based on game theory, the combined weight can be defined as the optimal weight value between AHP weight and entropy value.

On the foundation of experts grading, we adopt AHP method in Yaahp software V7.5 to calculate the influencing factors’ derived weights that show the importance of various criteria and the evaluation indicators. Firstly, we define the problem and state the goal or objective, regarding RBPS comprehensive index as the overall object layer, and each RBPS indicator as the evaluation indicator layer, defining the criteria or factors that influence the goal and structure these factors into levels and sublevels by using 1–9 scale method. Then, we use paired comparisons of each factor and subgroup with respect to each other that forms a comparison matrix with calculated weights. The paired comparison scale between the comparison pair (a_ij) of two indicators (indicator i and indicator j) is as follows: (indicator i) 9-8-7-6-5-4-3-2-1-2-3-4-5-6-7-8-9 (indicator j). A numeric ranking of 1 means equal importance of two indicators and 9 represents extreme importance of one given indicator to another. At last, by measure of the eigenvalue that represents the relative ranking of importance attached to the criteria or objects being compared, we carry on the test of consistency to guarantee the transitivity between the estimated values. Here, consistency is provided by the largest eigenvalue, and it can be represented by a matrix algebraic property of cardinal transitivity where the equality a_ij = 1/a_ji = a_ji ⁻¹, and a_ij = a_ik a_kj for any index i, j, k. Inconsistencies arise if the transitive property is not satisfied as determined when the largest eigenvalue far exceeds the number of indicators being compared (Saaty 2008).

On the other hand, we use entropy value method, an objective weighting method, to calculate information entropy value of each indicator’s various values to judge the given indicator’s dispersion. Those indicators with bigger dispersion degrees have more impact on the evaluation object (Singh 2013). Therefore, when we calculate the combined weight, the specific process are as follows.

Firstly, the subjective AHP weight vector is as shown in equation (1), as well as of the objective entropy value weight vector in equation (2).

ω = (ω_{1}, ω_{2}, \dots, ω_{m}),

u = (u_{1}, u_{2}, \dots, u_{m}) .

The optimized weight value matrix is denoted as follows:

(\begin{matrix} ω \cdot ω^{T} & ω \cdot u^{T} \\ u \cdot ω^{T} & u \cdot u^{T} \end{matrix}) (\begin{matrix} α_{ω} \\ α_{u} \end{matrix}) = (\begin{matrix} ω \cdot ω^{T} \\ u \cdot u^{T} \end{matrix}),

where α_ω is the AHP weight and α _u is the entropy value.

So the combined weight, that is, w can be represented as follows.

w = α_{ω} \cdot ω + α_{u} \cdot u .

In addition, the weight value of each dimension and indicator should be subject to the following criterion:

ω_{1} + ω_{2} + \dots + ω_{n} = 1,

ω_{i 1} + ω_{i 2} + \dots + ω_{i m} = 1.

RBPS calculation

The comprehensive development quality of each poverty-stricken county’s RBPS can be calculated as follows.

{RBPS}_{i} = \sum_{j = 1}^{n} x_{i j} w_{j},

where x_ij is the normalized value of RBPS indicator j of poverty-stricken county i by adoption of min–max scalarization approach. w_j indicates the weight of indicator j. n represents the number of all RBPS indicators, here, n = 21. RBPS _i is the comprehensive development value of poverty-stricken county i, and RBPS value of each contiguous destitute area or province is recognized as the average value of all counties’ ones under its own jurisdiction.

In addition, to significantly measure the relative development of RBPS of poverty-stricken counties within a contiguous destitute area, as well as that of among different destitute areas, this article introduces a grading index, Z, to explore the development disparities, so that these grades across multiple years can be compared to each other.

Z = \frac{{RBPS}_{i} - \bar{RBPS}}{σ},

where $\bar{RBPS}$ is the three-year average value of RBPS during 2010–2012. σ is the standard deviation of all RBPS during 2010–2012. Z_i, county i’s RBPS grading indicator, depends on its correlation to both $\bar{RBPS}$ and σ. By which, each county could be classified into one of the following groups: relatively richness (Z ≥ 1), relatively balance (0 ≤ Z < 1), relative shortage (−1 ≤ Z < 0), relatively serious shortage (Z < −1).

Similarly, all economic indicators are integrated into a comprehensive economic index, CE, by use of the same combined weighting method with RBPS, obtained from the following equation.

{CE}_{i} = \sum_{j = 1}^{n} y_{i j} w_{j},

where y_ij is the normalized value of economic indicator j of poverty-stricken county i by adoption of min–max scalarization approach. w_j indicates the weight of indicator j. n represents the number of all the economic indicators, here, n = 5. CE _i is the comprehensive development index of poverty-stricken county i.

CE of each contiguous destitute area or province is recognized as the average value of all poverty-stricken counties under its jurisdiction, calculated by taking into consideration the population size of each poverty-stricken county within a province or area. For example, the provincial-level CE is calculated as follows.

{CE}_{p j} = \sum_{i = 1}^{n} {CE}_{i} * p_{i} / p_{t},

where CE _pj is the average CE of province j, p_i is the population of county i. n is the total number of poverty-stricken counties in province j. p_t is the total population of all poverty-stricken counties in province j. Similarly, areal-level CE is obtained by using the above principle.

It is noteworthy that, although the region’s total size or total population is also an important indicator for both reflecting a region’s comprehensive economic strength and monitoring its rural development, Chinese government has been monitoring region’s local RBPS provision in terms of per capita criterion to respond the New Outline, incapable to fully meet the rural residents’ demand for good individual accessibility and spatial accessibility to RBPS. Therefore, for CE calculation, we mainly adopt per capita indicators to respond to RBPS provision. In addition, one main purpose of this study is to monitor and compare RBPS provision level that each poverty-stricken county’s local government provides for residents living within its jurisdiction, so as to examine whether a local government has tried its best to rid all rural population out of poverty; therefore, we also take into little account the spillover effect of the adjacent counties on both RBPS and CE.

Spatiotemporal Diversity of RBPS

Gini coefficient, used to measure resource inequality (Catalano, Leize, and Pfaff 2009), is adopted here to evaluate the development differences of different indicators within a destitute area, as well as those among different destitute areas.

The G coefficient is scaled from 0 to 1, where 0 represents equal development among all the counties and a 1 means complete inequality—all the development in the hands of a single county. Usually, a G value of .2 or less means a relatively balanced development, while the value of above .6 denotes a deeply glaring disparity. G can be calculated as follows.

G = \sum_{i = 1}^{n} Y_{i} X_{i} + 2 \sum_{i = 1}^{n - 1} Y_{i} 1 - V_{i} - 1,

where Y_i is the population share of each county, X_i is the RBPS share of each dimension within one county, n is the number of all counties in the study area, V_i denotes the total number ranked by per capita indicator.

In addition, GIS spatial autocorrelation and global Moran’s I coefficient (Goodchild et al. 2000) are also introduced here to identify the spatial aggregation distribution characteristic.

Coupling Relationship between RBPS and CE

For a very long period even since the founding of People’s Republic of China, China paid more attention to urban development than to rural one, and urban people have enjoyed much more financial resource and social welfare at the expense of the rural development. Even now, almost seventy years later, the increasing gap between urban and rural areas is still obvious in China, especially for those rural poverty-stricken regions where there still exist over seventy million people under poverty that could not fully share huge China’s development dividend and the government’s rural basic public service supply. On the other hand, some local governments’ financial resources may be huge; however, it is not sure whether they had devoted adequate resource into the rural development, as a result of pure GDP achievements guidance on evaluating the achievements of the cadre. Due to the insufficiencies of central government supervision, different county may have different input share in RBPS. To address it, we examine the correlations among RBPS and CE from the views of both the global correlation analysis and local economic circle-layer analysis, in hopes that the local government could really values RBPS to meet the rural poverty elimination requirement of China’s New Outline and develop RBPS that matches its economic strength to reach a harmonious state as much as possible.

To analyze the coupling development degree between the poor counties’ RBPS and their corresponding CE since the implementation of the New Outline, weighted Voronoi diagram and Tapio decoupling model are conjunctively adopted here. Voronoi diagram, also called Voronoi tessellation, is a general and basic geometric data structure and provides a way of dividing space into a few regions, intending to find a region composed of a set of grid points, in which every point in a Euclidean space is assigned to at least one of the points (also called generators) according to a certain rule, and the resulting sets of points associated with the generators are collectively exhaustive and mutually exclusive except for boundaries (Atsuyuki, Boots, and Sugihara 2008). Each region is called a Voronoi cell, dual to the Delaunay triangulation, indicating that, for each point in a set, no other one exists inside the circumcircle of any triangle. Compared to the traditional Voronoi in which each seed has the same influence within its corresponding region, a weighted Voronoi diagram is the one in which the function of a pair of points to define a Voronoi cell is a distance function modified by multiplicative or additive weights assigned to generator points, which may reveal the more actual radiation or influence on its adjacent space (Boots and South 1997; Atsuyuki, Boots, and Sugihara 2008).

Here, this article uses the circle-layer structures that are generated by weighted Voronoi diagrams in ArcGIS 10.3, reflecting the impact differences of different cities with different strength grade and size, to reveal the radiating and driving influence of both the provincial capitals and the planned key counties on the adjacent counties. Specifically, setting each provincial capital or the planned key county that are located within the contiguous destitute areas as the seed, introducing its corresponding CE as the weights of the distances, weighted Voronoi diagrams are constructed to divide their spatial influence ranges. In this article, there are seventy-six seeds, consisting of twenty-one provincial capitals and fifty-five planned key counties, participating in constructing weighed Voronoi diagram. Then we use GIS adjacent analysis in ArcGIS 10.3, to divide each Voronoi diagram into four circle zones according to the distance from the seed, from inside out, respectively, followed by the center circle layer, inner circle layer, middle circle layer, and outer circle layer, center circle layer being the closest one to the seed. Then similarly by overlay analysis, 728 stricken-poverty counties in this study area are classified into these four circle zones.

Further, Tapio decoupling elastic model, representing dilute or even completely deviated response relationship among interrelated variables, is introduced here to reveal the associations between RBPS and CE of those counties within each circle zone. Referring to the definition given by Tapio (2005), the decoupling index between RBPS and CE during a period of t, EC _t , can be measured as the ratio of the growth change of RBPS to the growth change of CE in a given time period, as shown in equation (11):

{EC}_{t} = \frac{Δ {RBPS}_{t}}{Δ {CE}_{t}} = \frac{{RBPS}_{t s} - {RBPS}_{t e}}{{CE}_{t s} - {CE}_{t e}},

where EC _t denotes the decoupling elastic coefficient. ΔRBPS _t and ΔCE _t are, respectively, the growth change of RBPS and CE between a base year 0 to a target year t. RBPS _ts and RBPS _te denote RBPS of the beginning year, and that of a target year t, respectively. Similarly, CE _ts and CE _te mean the CE index of a base year 0, and that of a target year t, respectively.

Further, the decoupling state can be classified in reference to the studies of Tapio (2005) and Vehmas, Luukkanen, and Kaivo-oja (2007), as shown in Table 2. Decoupling type means the situation when RBPS growth is lower than CE growth, indicating that the dependence of RBPS growth on CE is weakening. Furthermore, this decoupling type can be divided into three subtypes as recessive decoupling, strong decoupling, and weak decoupling, in terms of the differences among RBPS _t , CE _t _, and EC _t . Coupling type indicates that the linkage between RBPS growth and CE growth is reinforced. Similarly, it can be divided into two subtypes as growing coupling and recessive coupling. Negative decoupling type is the situation when RBPS growth is faster than CE growth, and also can be divided into three subtypes. For all six subtypes in negative decoupling and decoupling types, recessive coupling and recessive decoupling ones mean that RPBS growth is lower than CE growth and other four are unsatisfactory because of the negative CE growth.

Table 2.

Coupling Type Based on Tapio Decoupling Model.

Coupling type	Classification reference	Difference subtype	Definition
Negative decoupling	ΔRBPS _t > 0; ΔCE _t > 0; EC _t > 1.2	Expanded negative decoupling	RBPS growth
	ΔRBPS _t > 0; ΔCE _t < 0; EC _t < 0	Strong negative decoupling	CE damage
	ΔRBPSt < 0; ΔCE _t < 0; 0 < EC _t < 0.8	Weak negative decoupling	CE lag
Decoupling	ΔRBPS _t > 0; ΔCE _t > 0; 0 < EC _t < 0.8	Weak decoupling	CE growth
	ΔRBPS _t < 0; ΔCE _t > 0; EC _t < 0	Strong decoupling	RBPS damage
	ΔRBPS _t < 0; ΔCE _t < 0; EC _t > 1.2	Recessive decoupling	RBPS lag
Coupling	ΔRBPS _t > 0; ΔCE _t > 0; 0.8 < EC _t < 1.2	Growing coupling	Cooperative growth
Coupling	ΔRBPS _t < 0; ΔCE _t < 0; 0.8 < EC _t < 1.2	Recessive coupling	Cooperative loss

Note: CE = county economy; RBPS = rural basic public services; EC = elastic coefficient.

Results and Analysis

Spatiotemporal Distribution of RBPS

The spatiotemporal distribution of RBPS during 2010–2012 in fourteen poverty-stricken areas of China can be acquired by using the above study methods. As shown in Figure 2, in 2010, RBPS in most poverty-stricken counties of China is mainly featured by relative shortage and relatively serious shortage, roughly distributed in the west side of Hu Line (also called Hu Hwan Yong Line, or Aihui–Tengchong Line) that marks a striking difference in the distributions of both China’s population and poor population. In 2011, the number of the counties with RBPS of relative shortage and relatively serious shortage is similar to that of those counties with RBPS of relative equilibrium and relative enrichment, respectively, distributed roughly in both sides of Hu Line. Further, in 2012, most poverty-stricken counties of China have the characteristics of relative equilibrium or relative enrichment, and RBPS of the east of Hu Line is significantly higher than that of the west. All these indicate that, with the further promotion of the New Outline, RBPS in most poverty-stricken counties have developed rapidly, transiting their RBPS from relatively serious shortage or relative shortage to relative equilibrium or relative enrichment.

Figure 2.

Rural basic public services grades of poverty-stricken counties during 2010–2012, respectively.

At an area scale, among different contiguous destitute area, there has been a significant RBPS development difference (Figure 3). For examples, thanks to the synchronous deepening of “central rise” national strategy, RBPS of Luoxiao Area is overall the highest in fourteen destitute areas over these years, 62.5 percent of counties within this area is relatively rich by 2012, obviously higher than those of the surrounding area. However, due to the poor foundation, although with a biggest increase of RBPS in Tibetan areas of four provinces (i.e., Qinghai, Gansu, Sichuan, Yunnan, and Tibet Autonomous Region), there still remains a large and contiguous area of relatively serious shortage; while other regions have basically eliminated the relatively serious shortage area. Overall, RBPS distribution that shows lower in the west versus higher in the east has little change in fourteen contiguous destitute areas, roughly increasing along northwest–southeast, also increasing along Hu Line from southwest to northeast.

Figure 3.

Rural basic public services of fourteen contiguous destitute areas and their year-on-year growths.

From a provincial scale, there is a progressive increase in RBPS for the most of twenty-one provinces within the fourteen areas, as shown in Table 3, still along with an obvious inequality. Counties in Jiangxi Province have a highest RBPS level over these years, in contrast to those in Qinghai and Tibet with a lowest level. By 2012, fourteen of twenty-one provinces have the higher development level more than the national mean. On the other hand, it is particularly noteworthy that the counties in Jilin Province show an unexpected decrease in RBPS level from 2011 to 2012, dropping from the second rank to the last one, in sharp contrast to Anhui Province.

Table 3.

RBPS Ranking of Each Province during 2010–2012.

Province	2010	2011		2012
Province	RBPS	Ranking	RBPS	Ranking	RBPS	Ranking
Jiangxi	0.6604	1	0.6931	1	0.7205	1
Jilin	0.6322	2	0.6386	2	0.5386	14
Shanxi	0.5787	3	0.6206	5	0.6625	4
Hebei	0.5786	4	0.5991	8	0.6395	7
Ningxia	0.5700	5	0.5997	7	0.6445	6
Anhui	0.5686	6	0.6363	3	0.6880	2
Shannxi	0.5643	7	0.6032	6	0.6696	3
Henan	0.5633	8	0.6208	4	0.6581	5
Heilongjiang	0.5602	9	0.5761	9	0.6240	9
Chongqing	0.5218	10	0.5493	11	0.6152	10
Hubei	0.5078	11	0.5453	12	0.6149	11
Xinjiang	0.4972	12	0.5537	10	0.6337	8
Yunnan	0.4865	13	0.5431	13	0.5814	13
Guangxi	0.4543	14	0.5021	14	0.5989	12
Hunan	0.4378	15	0.4622	17	0.5204	15
Gansu	0.4365	16	0.4834	15	0.5191	16
Guizhou	0.3975	17	0.4715	16	0.5184	17
Inner Mongolia	0.3810	18	0.4053	19	0.4554	19
Sichuan	0.3767	19	0.4206	18	0.4829	18
Qinghai	0.3683	20	0.3817	21	0.4267	21
Tibet	0.3132	21	0.3879	20	0.4300	20
Mean	0.481		0.527		0.574

Note: RBPS = rural basic public services.

At a county level, it can be inferred that Nierong County grasps twice awful opportunities of the last rank both in 2010 and 2011, while Geer County shows another lowest RBPS in 2012, and both of the two counties are located in Tibet Area. The highest RBPS from 2010 to 2012 is shown in Qinba Area, respectively, corresponding to Luanchuan County (in 2010) and twice Ruyang County (in 2011 and 2012, respectively). From 2010 to 2012, as shown in Figure 4, RBPS disparity among the highest and the lowest, respectively, reaches fifty-five times, fifty times, and twenty-two times, showing a gradually narrowing gap. The number of those counties with RBPS value that is more than the average value is 366,375,377, respectively, accounting to 50.3 percent, 51.5 percent, and 51.8 percent of the whole stricken-poverty counties. All these denote that, overall, RBPS is positively developing though it is still not so high. On the hand, from Figure 4, it can be seen that, despite the fact that average RBPS of the national poverty-stricken counties is obviously lower than that of area-level counties, all of them are trying hard to improve their RBPS. As a matter of fact, these counties had shifted their RBPS grades from a mostly relative shortage or relatively severe shortage in 2010 to a state of mostly relatively richness or relatively equilibrium in 2012.

Figure 4.

Rural basic public services comparison between nation-level poverty-stricken counties and contiguous destitute-level ones.

Spatiotemporal Diversity of RBPS

By adopting global Moran’s I, spatial aggregation distribution characteristic of fourteen areas can be identified. We calculated Moran’s I by using of first-order polygon contiguity in ArcGIS 10.3, assigning 1 to those polygons that have common edge so as to be spatially adjacent, otherwise 0. Those polygons that have no common edges with the object ones (poverty-stricken counties) will not join in the calculation. As a result, the global Moran’s I indexes of RBPS yearly from 2010 to 2012 are, respectively, 0.39, 0.41, and 0.40, showing a positive and steady spatial autocorrelation.

As far as the spatial diversity of RBPS is concerned, Figure 5 represents a shrinking trend of Gini coefficient gap for the RBPS of the fourteen areas from 2010 to 2012, denoting that the RBPS gaps among different areas are narrowing. From the special perspective of different areas, Tibet Area has a largest Gini value, as well as a significant difference in Gini values among its different dimensions. That is because of the bad natural environment characterized by high altitude, inclement climate, the lack of production and living conditions, and frequent natural disasters, resulting in the extremely unbalance of its internal development. Internal RBPS difference of Lvliang Mountain Area is small, comprehensive development level tending to be relatively balanced. Gini coefficients of each dimensions are relatively small, respectively, within South Area of Great Khingan Mountains, Dabie Mountain Area, Border Area in Western Yunnan, Rocky Desertification Area in Yunnan, Guizhou and Guangxi Provinces, and Wumeng Mountain Area, indicating a relatively balanced development. In terms of rural environment protection, there exists a big difference within South Xinjiang Area, showing no obvious improvement. In addition, most RBPS dimensions of other areas have been getting better except for rural public safety service.

Figure 5.

Gini coefficients of each dimension in rural basic public services for fourteen contiguous destitute areas during 2010–2012, respectively.

From the views of the dimensions that bring the RBPS internal diversities, the dimension of rural public safety service has a biggest Gini coefficient value of over 0.6, indicating that this dimension has a great development gap with the characteristic of prominent polarization, as well as an obvious regional disparity among different areas; therefore, it acts as the most contributing factor that cause RBPS disparity. In addition, two dimensions of rural basic education and rural environment protection haven’t been provided with a relatively reasonable and fair resource allocation, compared to those dimensions of rural public health, rural public culture, rural social security, and rural infrastructure with their Gini values ranging from 0.3 to 0.4.

The result indicates that since 2010, special rural antipoverty policy has played a beneficial role in effectively resolving problems facing Chinese poverty-stricken rural areas and fostered coordinated and harmonious intrarural development. That’s to say, Chinese government has paid more attention to RBPS in recent years, resulting in the promotion of rural health services, the improvement of rural social security system, the development of rural infrastructure, as well as the initial success of some strategies for precise poverty alleviation. However, the basic public safety services cannot be guaranteed among all the areas. Therefore, the latter measures should be adjusted to local conditions to provide the corresponding differential antipoverty support policy, according to the vulnerable situation of different areas.

Correlation between RBPS and CE

As a whole, when we use Spearman’s rank correlation to examine the relation between RBPS and CE throughout all the counties, we find that there exists a Spearman’s coefficient of .185 in 2011, .201 in 2012, and .234 in 2013, respectively, indicating a weak linear correlation, as well as an overall disorder and instability.

On the other hand, in order to further reflect the influence of the economic development on RBPS, circle-layer structures of seventy-six provincial capitals and planed key counties are calculated according to the above weighted Voronoi diagram method. As shown in Figure 6, similarly, the correlation tests among each county’s original RBPS and its corresponding original CE among each of four circular regions during 2010–2012 also denote a weak linear correlation with an R value of no more than .4, respectively. Further, we compute the correlation coefficient between mean RBPS and mean CE at a circle-layer scale for four subsets of counties, respectively. It was noteworthy that these two variables have a significant positive correlation coefficient (R) as high as.9 during 2010–2012. As shown in Figure 7, R changes from 2010 to 2012 are universal, gradually increasing from 2010 to 2012. From Table 4, it can be seen that the farther from the center circle, the weaker the mean CE, the lower the mean RBPS. This may result from the fact that, for a long time, RBPS provision resources have been gathered to relatively developed areas, the local government excessively concerned about city rather than about rural area, taking into less account rural livelihood projects. Which, however, is also a response to China’s New Outline in order to monitor the work performance and administrative program of government officials, helping officials understand their relative situation.

Figure 6.

The weighted Voronoi circle-layer structure.

Figure 7.

Correlation between mean rural basic public services and mean county economy for circle-layer structures during 2010–2012, respectively.

Table 4.

Mean RBPS and Mean CE within Different Weighted Voronoi Circle Layers.

Weighted Voronoi circle layer	2010		2011		2012
Weighted Voronoi circle layer	CE	RBPS	CE	RBPS	CE	RBPS
Center circle layer	0.5160	0.3844	0.6068	0.5601	0.6985	0.7204
Inner circle layer	0.4333	0.3487	0.5297	0.5153	0.6382	0.6996
Middle circle layer	0.4177	0.3319	0.5197	0.4862	0.6086	0.6767
Outer circle layer	0.3255	0.3091	0.3998	0.4681	0.4878	0.6143

Note: CE = county economy; RBPS = rural basic public services.

However, it also seems that these test results, respectively, from original and mean values are conflicting. As a matter of fact, they also reflect China’s actual situation on urban–rural dual development and public service provision. For quite a long time, due to inadequate incentives mechanism inhabited in fiscal transfer payment and tax system reforms, governments at various levels in China kept their eyes glued on economic growth in day-to-day work, providing RBPS according to their willingness, rather than to the rural residents’ general demand. What’s worse, they devoted less attention to RBPS than to urban public service, resulting in that public service provision differences in urban–rural China are obvious. Rural population shared such fewer fruits of China’s economic reform and development that RBPS development level could not denote one county’s intrinsic economic development quality.

From Figure 8, it can be found that 63.05 percent of the whole counties within fourteen contiguous destitute areas are in the weak decoupling (relative decoupling) state from 2010 to 2012, belonging to the economic growth subtype in which the growth speed of RBPS is lower than that of CE, despite both are in improvement; 13.19 percent of the whole counties are featured with growing coupling degree, in which both CE and RBPS are in cooperative growth; 15.66 percent of the whole counties are featured with expanded negative decoupling, in which the growth speed of RBPS is faster than that of CE. In addition, 6.04 percent and 1.79 percent of the counties are, respectively, in the strong negative decoupling-CE damage and strong decoupling-RBPS damage subtype, these counties only focus on a single development of RBPS or CE, gravely deviating from the coordinated state. It’s worth mentioning that, only Dangchang County in Qinba Mountain Area is in recessive coupling, where both RE and RBPS reduced roughly and RBPS development lags behinds that of CE. In addition, Yushu County in Qinghai Province is in weak negative decoupling, where CE development has a larger decrease than RBPS.

Figure 8.

The decoupling relation between rural basic public services and county economy during 2010–2012.

On the other hand, what most counties within fourteen contiguous destitute areas have in common is that both their RBPS and CE have improved over these years, showing a weak decoupling in which RBPS lagged behind CE, far from ideal cooperative development-oriented type. Similarly, there is no decoupling recession or weak negative decoupling between them, indicating that, by the implementation of New Outline, China has achieved initial success on poverty alleviation; however, due to the long-term existence of the urban–rural dual structure of China’s economic and social development, some local governments only pursued one-sided inertia economic growth rather than RBPS growth, resulting in that RBPS development level still could not be compared to the rapid process of CE development, though Chinese government’s related policy supports had created a huge rift that can be seen today between China’s selected urban areas and vast swathes of impoverished rural hinterlands. Therefore, in the follow-up process, Chinese government should take a more active and effective targeted measures to coordinate the relationship between them. For examples, for the counties with a strong negative decoupling, their economic development should be particularly strengthened; for those counties with strong decoupling, they should strengthen the investment inclination of economy to RBPS.

Conclusion and Discussion

In this article, to respond to RBPS equalized development requirement in rural China’s precise poverty alleviation policy, we contributed to building a comprehensive assessment model to measure RBPS level from the view of spatial poverty and to systematically exploring multiscale and multidimensional spatiotemporal diversities of RBPS, which help provide an important decision basis for the central and local governments of China to design the corresponding differential antipoverty support policy, according to the vulnerable situation of different areas.

To be specific, our main work in this article is to construct a county-level RBPS evaluation indicator system, consisting of seven dimensions and twenty-one indexes that effectively reflect different RBPS aspects of economy, society, culture, and so on, and then we measured comprehensive RBPS index to identify the multiscale RBPS development levels of poverty-stricken rural China, constructing poverty maps to describe their multiscale and multidimensional spatiotemporal change. Besides, we also revealed how RBPS interact with CE by introducing Tapio model under an analysis framework of weighted Voronoi circle-layer structure. Our results show that (1) during 2010–2012, RBPS was gradually increasing, also denoting a positive spatial aggregation and an obvious nonequilibrium that was high in east but low in west at any scale of area, province or county, as well as a steady growth over time; however, the development gap was gradually narrowing, shifting their RBPS grades from a mostly relative shortage or relatively severe shortage in 2010 to a main state of relatively richness or relatively equilibrium in 2012. (2) RBPS gaps within different dimensions among different areas were mostly gradually narrowing; however, the dimension of social public safety service is an exception, which showed an obvious development gap and a significant regional differentiation, compared to those of other dimensions. RBPS of Tibetan areas felled into the most obvious heterogeneity. (3) There exist weak correlations between county-level original RBPS and original CE for each year and each circle layer, and significantly positive correlation is found only between average RBPS and average CE for four subsets of counties, respectively. Overall, RBPS development lagged behind that of CE as a main result of the weak decoupling between them.

By measuring RBPS spatiotemporal changes for 728 Chinese poverty-stricken counties during 2010–2012 from the multiscale and multidimensional perspectives, this study provides a good understanding of the status, regional differences, and evolution of RBPS in poverty-stricken rural China, and serves as a scientific reference regarding decision-making in promoting intrarural harmonious development and in constructing the new countryside of China. The promotion and implementation of RBPS will also enable the overall rural development at the aspects of economy, society, culture, and so on, which may finally help shape a more equitable rural coordination development pattern in rural China.

It is noteworthy that China is currently at the transformation period to change its economic and social model to concentrate on building a well-off society in an all-around way. Establishing an effective and impartial RBPS provision system is the interior requirement of socialism marketing economy and the basic obligation of a service-oriented government, as well as an important impetus of China’s current structural reform of supply side. From the above, it can be concluded that, as a whole, China has achieved gradual growth on the development of RBPS to respond to the implementation of New Outline; however, one of the challenges is that the inequality among multiscale regions still remains. RBPS development still cannot be compared to CE development due to many subjective and objective reasons. Therefore, this study emphasized that China government should devote itself to implementing suitable region-target and intrarural harmonious RBPS development policies, adjusting to local conditions to promote the corresponding differential antipoverty support, according to vulnerable situation of different areas. For example, by improving public safety service facilities, stimulating the driving radiation of key large and medium-sized cities with comprehensive strength, increasing the investment in RBPS, China may improve RBPS provision system and accelerate the comprehensive integration of RBPS with CE, so as to achieve the goal of comprehensive poverty elimination in 2020.

On the other hand, according to China’s current policy, Chinese government is taking new-type urbanization that puts emphasis on urban–rural coordinate development and antipoverty rural development, as one of the strategic focuses for restructuring rural economic and removing severe urban–rural dual development structure. Accompanying with new-type urbanization progradation of current China, increasing the provision of RBPS is being an inevitable choice to accelerate rural reform and development. However, compared to other no-poverty counties, it is still a slow progress for poverty-stricken counties that are located in mountainous area, where backward environment–economy ecosystem, barren land, harsh natural environment, and uncultured rural adults have been hindering the poverty-stricken regions’ urbanization process, indirectly leading to the low rural income, insufficient financial capitals, and weak public service provision. In most poverty-stricken counties of China, as the most key pillar industry, agriculture has still been the main driving force of poverty-stricken counties’ economic growth, from which, however, farmers only can make a limited income due to harsh eco-environment and scarce natural resource, and most adult farmers have to feed their families mostly by working far away from home as migrant workers, which is nearly helpless to increase the local government’s RBPS provision. Therefore, actively addressing the climate challenge, energetically developing green economy, and fostering new sources of economic growth so as to accelerate RBPS development, are being a longtime challenge to rid all out of poverty.

In this article, spatiotemporal change of RBPS are examined quantitatively. Nevertheless, due to the data obtainment, the three continuous periods of panel data could not yet fully show the change degree and the trend of RBPS. In addition, the calculations of both CE and RBPS have not taken into full consideration the population size and spillover between adjacency counties, having not made a clear statement at a finer spatial or administrative scale. Therefore, our following work is to collect and dispose the specific socioeconomic data, such as county size or population control on the RBPS metrics, analyzing more detailed grid-level heterogeneity that involves a detailed socioeconomic spatialization issue. On the other hand, spatial poverty trap, spillover effect, driving factors, and the optimal allocation mechanism of RBPS antipoverty resources for different scale of regions remain to be explored in depth.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Ren

. 2008. “The Construction of the Index System of the Equalization of Public Services in China: A Quantitative Analysis Based on the Perspective of Regional Differences.” Finance and Trade Economics 6:79–82.

Atsuyuki

Boots

Sugihara

. 2008. Spatial Tessellations: Concepts and Applications of Voronoi Diagrams. Chichester, UK: John Wiley & Sons.

Batley

Mcloughlin

. 2015. “The Politics of Public Services: A Service Characteristics Approach.” World Development 74:275–85.

Boots

South

. 1997. “Modeling Retail Trade Areas Using Higher-order, Multiplicatively Weighted Voronoi Diagrams.” Journal of Retailing 73:519–36.

Catalano

M. T.

Leize

T. L.

Pfaff

T. J.

. 2009. “Measuring Resource Inequality: The Gini Coefficient.” Numeracy 2:1–22.

Chi

Wang

. 2016. “Disparity Research on the Equalization of Rural Basic Public Services in the Impoverished County of Wuling Mountain Area.” Journal of Geo-information Science 18:298–306.

Chou

W. L.

Wang

Z. J.

. 2009. “Regional Inequality in China’s Health Care Expenditures.” Health Economics 18:S137–46.

Deolalikar

A. B.

Jha

. 2013. “Empowerment and the Delivery of Public Services: A Note.” Empowerment and Public Service Delivery in Developing Asia and the Pacific, Asian Development Bank policy report, Manila. Accessed November 14, 2015. http://www.un.org/esa/socdev/egms/docs//2013/EmpowermentPolicies/Empowerment%20and%20the%20Delivery%20of%20Public%20Services.pdf.

Feng

X. L.

Zhu

Zhang

Song

Hipgrave

Guo

Ronsmans

Guo

Yang

. 2010. “Socio-economic Disparities in Maternal Mortality in China between 1996 and 2006.” BJOG: An International Journal of Obstetrics & Gynaecology 117:1527–36.

10.

Gao

Yan

Zhang

Fan

. 2015. “Dynamic Quantitative Studies on the Coordination of Public Service Provision and Economy Development in Henan Province.” Journal of Xinyang Normal University National Science Edition 28 (2): 186–90.

11.

Goodchild

M. F.

Anselin

Appelbaum

R. P.

Harthorn

. 2000. “Toward Spatially Integrated Social Science.” International Regional Science Review 23:139–59.

12.

Gustafsson

B. A.

Sicular

. 2008. Inequality and Public Policy in China. Cambridge, MA: Cambridge University Press.

13.

Han

Zhang

. 2015. “Evaluation and Spatial Analysis of the Equalization of Basic Public Service in Urban and Rural Areas in China.” Geographical Research 34:2035–48.

14.

Higgs

Fry

Langford

. 2012. “Investigating the Implications of Using Alternative GIS-based Techniques to Measure Accessibility to Green Space.” Environment and Planning B: Planning and Design 39:326–43.

15.

Higgs

Langford

Norman

. 2015. “Accessibility to Sport Facilities in Wales: A GIS-based Analysis of Socioeconomic Variations in Provision.” Geoforum 62:105–20.

16.

Higgs

White

S. D.

. 1997. “Changes in Service Provision in Rural Areas. Part 1: The Use of GIS in Analysing Accessibility to Services in Rural Deprivation Research.” Journal of Rural Studies 13:441–51.

17.

Qin

. 2009. “An Analysis of Regional Social Public Services Development Difference in China.” Economic Geography 29:887–93.

18.

Wei

Y. D.

. 2014. “Multidimensional Inequalities in Health Care Distribution in Provincial China: A Case Study of Henan Province.” Tijdschrift Voor Economische En Sociale Geografie 105:91–106.

19.

Liao

Zhou

Tang

. 2014. “The Empirical Study on the Change of County Economy Disparity in Hunan.” Economic Geography 34:35–41.

20.

Liu

Chen

. 2013. “Spatio-temporal Change of Urban–rural Equalized Development Patterns in China and Its Driving Factors.” Journal of Rural Studies 32:320–30.

21.

H. Q.

Han

Z. L.

Jiang

H. X.

. 2011. “The Characteristics and Spatial Difference of Basic Public Services of Cities at Prefecture Level and above in China.” Economic Geography 31:212–17.

22.

Shen

Song

. 2014. “The Gaps between Urban and Rural Basic Public Service and Its Obstacle Factors in Jiangsu Province.” Human Geography 29:89–93.

23.

Martinez

2009. “The Use of GIS and Indicators to Monitor Intra-urban Inequalities. A Case Study in Rosario, Argentina.” Habitat International 33:387–96.

24.

Neutens

Delafontaine

Scott

D. M.

De Maeyer

. 2012. “A GIS-based Method to Identify Spatiotemporal Gaps in Public Service Delivery.” Applied Geography 32:253–64.

25.

Niu

Yang

Bai

. 2010. “Analysis of Regional Disparities of Rural Economic Development Level in China.” Economic Geography 30:479–83.

26.

Ren

Tong

Kwan

M. P.

. 2014. “Space–time Measures of Demand for Service: Bridging Location Modelling and Accessibility Studies through a Time-geographic Framework.” Geografiska Annaler: Series B, Human Geography 96:329–44.

27.

Saaty

T. L.

2008. “Decision Making with the Analytic Hierarchy Process.” International Journal of Services Sciences 1:83–98.

28.

Sanders

S. R.

Erickson

L. D.

Call

V. R. A.

McKnight

M. L.

Hedges

D. W.

. 2015. “Rural Health Care Bypass Behavior: How Community and Spatial Characteristics Affect Primary Health Care Selection.” The Journal of Rural Health 31:146–56.

29.

Shiikha

Kruger

Tennant

. 2015. “Rural and Remote Dental Services Shortages: Filling the Gaps through Geo-spatial Analysis Evidence-based Targeting.” Health Information Management Journal 44:39–44.

30.

Singh

V. P.

2013. Entropy Theory and Its Application in Environmental and Water Engineering. Hoboken, NJ: John Wiley & Sons.

31.

Tapio

2005. “Towards a Theory of Decoupling: Degrees of Decoupling in the EU and the Case of Road Traffic in Finland between 1970 and 2001.” Transport Policy 12:137–51.

32.

Vehmas

Luukkanen

Kaivo-oja

. 2007. “Linking Analyses and Environmental Kuznets Curves for Material Flows in the European Union 1980–2000.” Journal of Cleaner Production 15:1662–73.

33.

Wang

Meng

Ren

. 2014. “MTFCAM_based Spatial Accessibility to the Educational Resources from the Primary Schools in the Impoverished Mountain Areas.” Energy Education Science and Technology Part A: Energy Science and Research 32:5859–70.

34.

Wang

Nie

. 2011. “Study on the Equalization of Basic Public Services in China.” In 2011 International Conference on Management and Service Science (MASS), August 12–14, 2011, 1–5. Wuhan, China: IEEE. doi:10.1109/ICMSS.2011.5998749.

35.

Wang

Qian

. 2015. “A PPI-MVM Model for Identifying Poverty-stricken Villages: A Case Study from Qianjiang District in Chongqing, China.” Social Indicators Research 124:1–26. doi:10.1007/s11205-015-1190-4.

36.

Yang

. 2012. “Theoretical and Empirical Research on Relationship between Basic Public Services and Economic Growth.” Science of Science and Management of Science Technology 33:96–101.

37.

LeGates

Qin

. 2013. “Coordinated Urban-rural Development Planning in China—the Chengdu Model.” Journal of the American Planning Association 79:125–37.

38.

Zhang

Kanbur

. 2005. “Spatial Inequality in Education and Health Care in China.” China Economic Review 16:189–204.

39.

Zhou

. 2014. “Evolution of Spatial-temporal Pattern of County Economic Development in China during 1982–2010.” Progress in Geography 33:102–13.