海南省农村多维贫困及影响因素的空间分异

贫困具有多维属性,根据不同社会群体和背景从多维视角定义贫困已成为贫困问题研究的共识。依据Alkire-Foster多维贫困框架,拓展精准扶贫的“两不愁,三保障”识别标准,建立了涵盖教育、健康、居住、生活和收入指标的海南省农户多维贫困评估指标体系,基于海南省70个乡镇、134个贫困村3924户入户调查数据,采用双重临界值法评估了农户及村域多维贫困状况,进而运用地理加权回归(Geographically Weighted Regression,GWR)模型,分析了村域多维贫困影响因素的空间分异。结果显示,调查农户多维贫困率达18.22%,多维贫困程度严重的村多维贫困发生率不一定高,“两不愁、三保障”及收入指标对多维贫困指数的贡献率低。中、西部连片贫困地区多维贫困主要表现为较差的资产状况、不清洁的炊事燃料、较高的家庭成员患病率和较低的家庭成员最高学历。GWR模型分析表明,作为多维贫困最重要的影响因素,户主性别、户主受教育水平、女性劳动力占比和抚养比4个变量估计系数的空间分异明显。总体上,女性户主和低学历户主为主的地区倾向于更易发生多维贫困,二者的影响分别表现为从东到西、从北到南有所增强。女性劳动力占比为负向影响,抚养比为正向影响,呈现出自北向南增强的趋势,体现了海南贫困地区劳动力弱、女性相对更勤劳等典型地域特征。

Spatial difference of multidimensional poverty and its influencing factors in the rural areas of Hainan Province

Poverty has multidimensional attributes, and it has become a consensus to study poverty from a multidimensional perspective according to different social groups and backgrounds. In order to measure the multidimensional poverty situation in the rural areas where the poor population is concentrated in Hainan Province, we expanded the index system based on the exit criteria for targeted poverty alleviation fulfilling the basic needs of food and clothing and guaranteeing compulsory education, basic medical care, and housing, and established a multidimensional poverty assessment conceptual model for rural households in Hainan Province that covers education, health, housing, livelihood, and income indicators. Then, based on household survey data from 3924 households in 70 towns and 134 poor villages of Hainan Province in 2018, we used the double threshold Alkire-Foster (A-F) method to evaluate the multidimensional poverty status of rural households and villages, and then used the geographically weighted regression (GWR) model to analyze the spatial heterogeneity of the influencing factors of multidimensional poverty in villages. The study results show that: 1) The incidence of multidimensional poverty of the surveyed households was 18.22%. But the incidence of multidimensional poverty in villages with severe multidimensional poverty is not necessarily high. 2) The four indicators of farming households' asset status, cooking fuels, family members' diseases, and family members’ highest academic qualifications contribute the most to multidimensional poverty, while the contribution ratio of indicators belonging to the standard of fulfilling basic needs of food and clothing and guaranteeing compulsory education, basic medical care, and housing, as well as income are generally not high. The multidimensional poverty in the contiguous poverty areas in the central and western regions of the province is mainly manifested by poor asset conditions, unclean cooking fuels, high prevalence of disease of family members, and lower education levels. 3) The GWR model analysis showed that as the most important influencing factors of multidimensional poverty, spatial heterogeneity of the estimated coefficients of the four variables, gender of the household head, education level of the household head, ratio of female labor force, and dependency ratio, have very obvious impacts. In general, areas with more female-headed and low-education attainment individual headed households tend to be more prone to multidimensional poverty, and their impacts increased from east to west and from north to south, separately. With an increasing trend from north to south, the effect of the proportion of female labor force is negative and that of the dependency ratio is positive, which reflects the typical regional characteristics of weak labor force and relatively more industrious women in Hainan poverty-stricken areas.