统计与数据科学学院知识库

  随着贫困治理背景的全方位转换，我国社会已经从以解决温饱为主要任务的脱贫攻坚阶段转入巩固温饱成果、缩小发展差距的乡村振兴阶段。社会救助制度作为中国基本民生保障的一项基础性制度安排，也是乡村振兴战略背景下贫困治理的主要措施，要稳步延续脱贫攻坚时期的救助成果，对弱势群体继续实施更加完善、更加全面、更加有效的社会保障。而甘肃省曾作为我国脱贫任务最重的省份之一，其偏远的地理位置和恶劣的生态环境使得经济贫困成因复杂，贫困程度深，贫困范围广，那么研究甘肃省的社会救助对象识别认定和社会救助政策实施效果便是关键之举。因此，针对乡村振兴战略背景下的甘肃省农村社会救助识别认定和政策实施效果研究具有现实意义。

  本文的研究主要集中在以下几个方面：第一，甘肃省农村社会救助对象识别效率。在与现有救助对象识别标准衔接的情况下，利用民政部门官方农户家庭微观记录数据和2020年甘肃省社会救助核查评估的真实调查数据，从基于农户基本特征信息的识别效率、基于收入视角的识别效率和基于可持续生计理论的多维救助识别体系的三个视角出发，判断农村社会救助对象的识别效率。第二，甘肃省农村社会救助对象认定准确性。将构建好的多维救助识别指标体系应用机器学习分类模型进行预测，判定目标农户的救助类型与救助级别。第三，甘肃省农村社会救助政策实施效果。主要涉及民政工作人员对政策熟知程度及政策实施效果，救助户对政策的满意程度及获得救助后的生活情况，非救助户对政策实施过程中的问题及建议等，探究不同主体对甘肃省农村社会救助政策实施效果的意见和建议。

  结果表明：甘肃省农村社会救助政策实施过程中存在“漏保误保”现象，而多维农村社会救助识别指标能显著提升社会救助对象的精准识别效率，有效降低识别过程中的漏保率和误保率。同时，生计资产框架的引入能一定程度上提高农户的认定准确度，提高救助资金和救助资源的利用率。最后，甘肃省还需进一步完善农村社会救助政策的兜底功能，加强社会救助政策的宣传和动态监管力度，推广“甘肃经验”，为乡村振兴时代完善优化农村社会救助政策提供参考依据。

With the all-round transformation of the background of poverty governance, our society has shifted from the poverty alleviation stage with food and clothing as the main task to the rural revitalization stage of consolidating the achievements of food and clothing and narrowing the development gap. As a basic institutional arrangement for basic livelihood security in China, the social assistance system is also the main measure for poverty governance under the background of the rural revitalization strategy. It is necessary to steadily continue the assistance achievements during the period of poverty alleviation, and continue to implement more perfect, more comprehensive and more comprehensive measures for vulnerable groups. Effective social security. Gansu Province was once one of the provinces with the most important poverty alleviation task in the country. Its remote geographical location and harsh ecological environment make the causes of economic poverty complex. The degree of poverty is deep, and the scope of poverty is wide. Then study the identification and identification of social assistance objects in Gansu Province and the implementation effect of social assistance policy is the key move. Therefore, under the background of rural revitalization strategy, the identification and identification of rural social assistance in Gansu Province and the research on the effect of policy implementation have practical significance.

The research of this paper mainly focuses on the following aspects: First, the identification efficiency of rural social assistance objects in Gansu Province. In the case of linking with the existing identification standards for assistance objects, using the official farmer household micro-record data of the civil affairs department and the real survey data of the social assistance verification and evaluation in Gansu Province in 2020, the identification efficiency based on the basic characteristic information of farmers and the income perspective are used. From three perspectives of identification efficiency and multi-dimensional assistance identification system based on sustainable livelihood theory, the identification efficiency of rural social assistance objects is judged. Second, the accuracy of identification of rural social assistance objects in Gansu Province. The constructed multi-dimensional rescue identification index system is applied to the machine learning classification model for prediction, and the rescue type and rescue level of the target farmers are determined. Third, the implementation effect of the rural social assistance policy in Gansu Province. It mainly involves civil affairs staff's familiarity with the policy and the policy implementation process, the assisted households' satisfaction with the policy and their living conditions after receiving assistance, and the non-assisted households' problems and suggestions in the policy implementation process. Opinions and suggestions on the implementation effect of rural social assistance policies in Gansu Province.

The results show that there is a phenomenon of "missing insurance and wrong insurance" in the implementation of rural social assistance policies in Gansu Province, and the multi-dimensional rural social assistance identification efficiency can significantly improve the precision identification efficiency of social assistance objects, and effectively reduce the missed insurance rate and error in the identification process. Guaranteed rate. At the same time, the introduction of the livelihood asset framework can improve the identification accuracy of farmers to a certain extent, and improve the utilization rate of rescue funds and resources. Finally, Gansu Province needs to further improve the basic function of rural social assistance policies, strengthen the publicity and dynamic supervision of social assistance policies, and promote the "Gansu experience", so as to provide a reference for improving and optimizing rural social assistance policies in the era of rural revitalization.