1. China Western Economic Research Center, Southwestern University of Finance and Economics, Chengdu 611130, China; 2. Business Department, Luzhou Vocational and Technical College, Luzhou 646000, China

Abstract Based on survey data of nine poverty-stricken counties in Wumen mountainous area, using the ordinal Probit model, this paper analyzed the influence mechanism for the assessment of the implementation effect of the policy of relocating the poor. It obtained the following results. (i) Material security and social environment have a positive influence on the assessment of the policy of relocating the poor. (ii) In terms of material security, there is no significant relationship between the employment environment and the assessment of the policy of relocating the poor. Per capita annual income, per capita housing, income-increasing channels, and endowment insurance have significant effects on the assessment of the policy of relocating the poor. (iii) In social environment, public services, infrastructure, neighborhood relations, and environmental sanitation have significant effects on the assessment of the policy of relocating the poor, while there is no significant relationship between the awareness of poverty alleviation policies and the effect assessment of the policy of relocating the poor. (iv) The severer the poverty, the more significant the assessment of the implementation effect of the policy of relocating the poor. In accordance with the above results, it came up with recommendations for targeted poverty alleviation.

Key words Material security, Social environment, Relocating the poor, Effect assessment, Influence mechanism

1 Introduction and literature review

Relocating the poor refers to relocating the poor people living in harsh conditions to other areas with better conditions, and helping the relocated people to get rid of poverty through improving the production and living conditions in resettlement areas and expanding the income-increasing channels of relocated people. Implementing the policy of relocating the poor is an important measure for making up the short board of poverty-stricken areas and winning the poverty alleviation project. The Chinese government plans to use five years to create files and cards for poverty-stricken people, implements the policy of relocating the poor, makes effort to accomplish the mission of relocating 10 million poor people in the 13th Five-Year Plan period, to help them enter the moderately prosperous society[1]. At present and for a long term in future, the project of relocating the poor will remain at a crucial stage. It is required to solve the problem of relocation, but also required to solve the problem of resettlement. In other words, poverty alleviation and sustainable livelihood are two key issues. Then, there come problems of what about the satisfaction or assessment of relocated people with the policy implementation effect and what is the influence mechanism of the policy performance.

In foreign countries, studies about the project of relocating the poor focus on analysis on ecological or environmental migration. Since Brown LR first proposed the concept of "ecological refugees" and "environmental refugees" in 1976[2], the research direction of ecological migration (also known as environmental migration) was started. However, there are relatively few studies on the assessment of the production and living benefits of ecological migrants, and there is almost no research on China’s policy of relocating the poor. Research findings are mainly concentrated on the following points. (i) Causes of migration. For example, EI-Hhmawi[3]explained the causes for ecological migration from three aspects: migration caused by a natural disaster, migration caused by environmental collapse and migration caused by slow and degraded ecological environment. In recent years, the causes have been explained from the perspectives of climate disasters (Israel and Briones)[4]and "spatial poverty traps" (Jalan and Rvallion)[5]. (ii) Necessities of migration. For example, Scott Leckie[6]pointed out that if the migration of climatic and environmentally poor people fails to receive global attention, the consequences of government intervention and resource input will be unimaginable. (iii) Migration promotion policies or strategies. For example, Scott Leckie and Chris Huggins[7]proposed that governments at all levels should develop a more complete legal framework to support climatic migration.

In China, the policy of relocating the poor is implemented as the national policy for reducing and alleviating poverty and achieving moderately prosperous society, and is now receiving wider and wider attention. For example, Kang Liang[8]studied the project of relocating the poor in Jiangxi Province and concluded that the project exerts a positive role in fundamentally eliminating the poverty, promoting poor people to increase income and get rich, and strengthening the self-development ability of poor people. For another example, through analysis on implementation methods of relocating the poor in Gansu Province since 2010, Chen Yuanetal.[9]found that the production and living conditions of the relocated people have been greatly improved, the production skills and industrial development have been developed rapidly, the employment channels have diversified and the cash income has increased steadily. (ii) There are some studies about the satisfaction or effect assessment of the implementation effect of the policy of relocating the poor, but there are few findings. For example, Jiao Yang[10]compared the satisfaction of two typical villages in the Loess Plateau and found that the satisfaction of the two villages lies in infrastructure of transportation, water and electricity, housing, children education, working opportunity, and government subsidies. (iii) Studies also involve the following aspects. (a) Policy evolution. For example, Wang Hongxinetal.[11]studied the characteristics of the evolution and support system of China’s policy of relocating the poor in the past 15 years. (b) Cause analysis. For example, through the questionnaire survey of 2019 households in 16 counties of 8 provinces in central and western areas, Zeng Xiaoxietal.[12]found that surviving relocation or developmental relocation is the main reason for relocation, but there are also concerns about the source of income after relocation. (c) Resettlement mode. For example, taking Guizhou as an example, Wang Yongpingetal.[13]divided the mode of relocation and resettlement into collective farms, small towns, tourist attractions, enterprises, enterprise promotion, and returning farmland to forests. (d) Problem analysis. Li Bo and Zuo Ting[14]pointed out that due to the deviation from the goal of precision assistance and precise management, there are differences between the institutional and administrative linkages of the practice of the policy of relocating the poor, accordingly making the poverty management fall into a fragmented predicament.

Through literature review, we found that the foreign researches are not relevant to this study topic, but they are instructive to think about the policy measures of this study. The domestic researches are extensive and comprehensive, so they have higher references than foreign researches, especially the researches about the implementation effect of the policy of relocating the poor. Domestic researches directly provide ideas for analyzing the implementation effect assessment and influence mechanism of the policy of relocating the poor. Therefore, on the basis of the above relevant research findings, we made an empirical analysis on the influence mechanism of implementation effect assessment of the policy of relocating the poor.

2 Influence mechanism and research hypothesis

The influence of material security on implementation effect assessment of the policy of relocating the poor is manifested in the following two aspects. (i) High motive force of villagers for relocation and high expectation for material security raise their satisfaction with the assessment of the relocation policy. The motive force of villagers for relocation is the joint force of four forces: driving force formed in relatively harsh living conditions of the place of departure, attraction of better conditions in the place of destination, support of government’s poverty alleviation policy, and driving force for villagers voluntarily to relocate. (ii) The relatively broad personal development prospects after the relocation influences the assessment of the effect of policy of relocating the poor. After the poor people are relocated, the original homestead is exited with compensation, land is consolidated and circulated, then agricultural industrial park and poverty alleviation apartment can be built. The housing conditions of the villagers are changed, the villagers actively are integrated into the urbanization process, their vision is expanded, the job opportunities will increase, the working environment is improved to a certain extent, and the income-increasing channels are further expanded, which will enhance the villagers’ happiness and further raise the villagers’ satisfaction with the policy of relocating the poor. In view of these, we selected the per capita annual income, per capita housing, income-increasing channels, working conditions, and endowment security to represent the material security dimension of relocated people.

Thus, we proposed Hypothesis 1: the material security is correlated with the implementation effect assessment of the policy of relocating the poor. The per capita annual income has the most significant positive influence on the implementation effect assessment of the policy of relocating the poor. Hypothesis 2: the per capita housing, income-increasing channels, working conditions, and endowment security are positively correlated with the implementation effect assessment of the policy of relocating the poor.

The influence of social environment on implementation effect assessment of the policy of relocating the poor is manifested in the following three aspects. (i) The exchange and communication between the relocated villagers will affect the implementation effect assessment of the policy of relocating the poor. The harmonious relationship between family members and neighbors is often conducive to adapting to the new environment more quickly, accordingly raising the satisfaction with implementation effect assessment of the policy of relocating the poor. (ii) The optimization of environment of place of destination will raise the satisfaction with implementation effect assessment of the policy of relocating the poor. The villagers will get an intuitive feeling by comparing the changes in production, infrastructure, health conditions, children education, and medical treatment before and after the relocation, so that the optimization of the environment strengthens the recognition of the policy of relocating the poor. (iii) The understanding of the national poverty alleviation policy will increase the villagers’ attention to the policy of relocating the poor, and in turn it will strengthen the expectation of villagers for the policy implementation effect, and accordingly setting forth higher requirements for the implementation effect of the policy of relocating the poor. Based on this, according to the relevant research literature and questionnaires, we used the neighborhood relationship, infrastructure, public services, environmental sanitation, and poverty alleviation policy awareness to reflect the changes in the social environment of the relocated people.

In view of this, we proposed Hypothesis 3: social environment is correlated with the implementation effect assessment of the policy of relocating the poor, the five variables have significant influence on the implementation effect assessment of the policy of relocating the poor, and the infrastructure has a greater influence.

In addition, villagers from key national counties with many poor people, poor geographical environment, severe poverty and frequent natural disasters have a strong willingness to relocate. In view of this, we proposed Hypothesis 4: the degree of poverty is correlated with the implementation effect assessment of the policy of relocating the poor, and villagers from severe poor counties have more significant assessment on the implementation effect of the policy of relocating the poor.

3 Data,variables,and empirical model

3.1DatasourceThe data in this study were taken from the questionnaire of "relocated farmers" in concentrated poverty-stricken areas during July and August of 2017. In 2011, China took concentrated poverty-stricken areas as the main battlefield for poverty alleviation. Considering the project of relocating the poor in the 13th Five-Year Plan period involves 1 400 county-level units in 22 provinces, for the purpose of saving time and cost, we selected 9 poverty-stricken counties in Wumeng Mountainous Areas with administrative division crossing Yunnan, Guizhou, and Sichuan as the survey objects. In this survey, we adopted stratified random sampling method, the questionnaire survey method, and farmer interview method. In order to make the survey data more representative, we divided the survey fields into three groups according to the degree of poverty, selected three poverty-stricken counties in each group, and selected a total of 15 resettlement sites, distributed 50 copies for each resettlement site, 1 800 copies in total; finally, our team collected and reviewed the questionnaire, obtained 1 595 valid ones, and the valid response rate was up to 88.6%. We firstly analyzed the reliability and validity of the questionnaire. Through calculation, the reliability coefficient Cronbach α coefficient was 0.836, indicating that the questionnaire had good credibility. Then we carried out the validity analysis, KMO statistical value of the questionnaire was 0.753, showing significance at 1% level, indicating that the questionnaire had good structural validity.

3.2Selectionanddescriptionofvariables

Brownie: A brownie, brounie/Urisk (Lowland Scots) or ùruisg/brùnaidh/gruagach (Scottish Gaelic) is a legendary24 kind of elf popular in folklore around Scotland and England (especially the north). He is the Scottish counterpart of the Scandinavian tomte, the Russian domovoi or the German Heinzelm?nnchen.Customarily they are said to inhabit houses and aid in tasks around the house. However, brownies do not like to be seen and will only work at night, perhaps in exchange for small gifts or food. They usually abandon the house if their gifts are called payments, or if they are offered gifts of clothes (no matter how shabby their own clothes are). In some stories, brownies have no noses ( Brownie (elf) Wikipedia 2006).

3.2.1The explained variable is the assessment of the effect of the implementation of the policy, that is, the degree of satisfaction of the farmers with the implementation effect of the policy of relocating the poor (Assess). The explained variables are divided into three types: unsatisfied, general, satisfied, assigned with values 1, 2, and 3, respectively. Therefore, the explained variable is a discrete number of farmers’ satisfaction with the implementation effect of the policy of relocating the poor, the higher the value, the higher the satisfaction of the farmers.

3.2.2Main explanatory variables are material security and social environment. From the perspective of economics, material security is the key factor affecting individual utility. The problem design of material security variables includes: per capita annual income, per capita housing area, income-increasing channels, working conditions, and endowment insurance. The per capita annual income respondents selected included "less than 4 000 yuan", "4 001-6 000 yuan", and "higher than 6 000 yuan", assigned with values 1, 2, and 3, respectively. The per capita housing area respondents selected included "less than 25 m2", "25-35 m2", and "greater than 35 m2", assigned with values of 1, 2, and 3, respectively. The income-increasing channels respondents selected included "no increase", "general", and "increase", assigned with values of 1, 2, and 3, respectively. The working conditions respondents selected included "no improvement", "general", and "improvement", assigned with values of 1, 2, and 3, respectively. Another explanatory variable is the social environment. The corresponding problem design includes neighborhood relationship, infrastructure, public service, environmental sanitation, and understanding of poverty alleviation policy. Social environment selected by respondents includes "unsatisfied" (disharmony), "general" (harmony), and "satisfied" (harmony), assigned with values of 1, 2, and 3, respectively. The higher the value, the higher satisfaction the respondents with certain item.

3.2.3Other control variables. Combined with relevant literature and questionnaires, the control variables we selected in this study include gender, marital status, political identity, age, educational level, and family size. Specifically, we set the gender, marital status, and political identity as virtual variable, assigned the male with 1, female with 0, married with 1, unmarried with 0, party member with 1, and non-party member with 0. Besides, we assigned the age younger than 20, 20-60 years old, and older than 60 with values 1, 2, and 3, respectively. The family size includes small, medium, and large, assigned with values of 1, 2, and 3, respectively; the higher the value, the more family members in the respondent family. The statistical descriptive results of the variables were shown in Table 1.

Table1Assignmentanddistributionofvariables

(To be continued)

(Continued)

TypeName of variableAssignment and description MeanvalueStandarddeviationEducational levelPrimary school and below=1; junior middle school and senior middle school (or secondary school)=2; college and above=31.050.486Number of family members3 people or less=1; 4-6 people=2; more than 6 people=31.890.562Material security Per capita annual incomeLess than 4 000 yuan=1; 4 001-6 000 yuan=2; higher than 6 000 yuan=31.740.668Per capita housing areaLess than 25 m2=1; 25-35 m2=2; greater than 35 m2=31.050.529Income-increasing channels No increase=1; general=2; increase=32.170.410Employment environment conditions No improvement=1; general=2; improvement=32.080.406Endowment security Endowment insurance not bought=0; endowment insurance bought=10.860.329Social environmentNeighborhood relationshipDisharmony=1; general=2; harmony=32.330.516InfrastructureUnsatisfied=1; general=2; satisfied=32.460.537Public services Unsatisfied=1; general=2; satisfied=32.390.627Environmental sanitation Unsatisfied=1; general=2; satisfied=32.160.594Understanding of poverty alleviation policyNot known=1; general=2; well known=32.210.601

3.3ModelbuildingIn this study, the implementation effect assessment of the policy of relocating the poor belongs to subjective awareness of farmers, and the explained variables are ordinal categorical variables. We set the question "are you satisfied with the implementation effect assessment of the policy of relocating the poor" in the questionnaire, and set the options "very satisfied", "satisfied", "general", "less satisfied", and "not satisfied". For the purpose of analysis, we summarized these five conditions into three ones "unsatisfied", "general", and "satisfied", thus it is suitable to use the ordered response Probit model to conduct econometric analysis. The basic form of the model is as follows:

(1)

whereMaterialiandSocialiseparately denote changes in material security and social environment of individual villagers after relocation, andControlidenotes the control variable influencing the implementation effect assessment of individual;βis the coefficient to be estimated,εiis a random error term subject to an independent normal distribution,i.e.εi|X∽N(0,1).Assessi*is a latent variable and it corresponds to an assessment level variableAssessi, and the relationship between this level variable and the latent variable is as follows:

(2)

μi(i=1, 2, 3) is the threshold value of latent variable. Assuming that the distribution function of the random error termεiisΦ(X), the conditional probability thatAssesstakes the selected value is as follows:

Prob(Assessi=2|X)=Prob(μ1

4 Regression results and analysis

With the aid of SPSS20.0 software, we carried out ordered Probit regression processing of the sample data. In the data processing, the variables of the individual characteristics such as gender, marital status, political identity, age, family size, and educational level were taken as control variables. Firstly, we introduced the material security variables and social environment variables into the model to obtain the Model 1 and the Model 2 respectively. Secondly, we introduced the material security variables and the social environment variables jointly into the model to obtain the Model 3. Finally, we listed the estimation values and significance of the implementation effect assessment of the policy of relocating the poor in Table 2.

Table2OrdinalProbitregressionmodelandresultsofinfluenceofmaterialsecurityandsocialenvironmentontheimplementationeffectofthepolicyofrelocatingthepoor

VariableModel 1Coefficient Model 2Coefficient Model 3Coefficient Control variablesMale 0.015(0.135)0.007(0.201)0.013(0.108)Married 0.341**(0.107)0.218**(0.138)0.187**(0.132)Party member0.297**(0.063)0.273**(0.082)0.223**(0.065)Age0.153(0.087)0.158(0.093)0.173**(0.132)Educational level0.174(0.029)0.204(0.075)0.184**(0.032)Family size 0.221**(0.087)0.167**(0.076)0.186**(0.098)

(To be continued)

(Continued)

VariableModel 1Coefficient Model 2Coefficient Model 3Coefficient Material securityPer capita annual income greater than 6 000 yuan0.752***(0.217)-0.683***(0.201)Per capita annual income 4 000-6 000 yuan0.641**(0.185)-0.638**(0.192)Income-increasing channels increased0.601***(0.099)-0.664***(0.063)Income-increasing channels general 0.597**(0.062)-0.585**(0.059)Per capita housing greater than 30 m20.417**(0.207)-0.462**(0.249)Per capita housing 25-30 m20.382*(0.211)-0.381*(0.209)Endowment insurance bought 0.429**(0.206)-0.417**(0.192)Employment environment improved 0.132(0.048)-0.141(0.043)Employment environment general 0.135(0.052)-0.191(0.066)Social environmentPublic service satisfied -0.615***(0.047)0.622***(0.057)Public service general -0.548***(0.090)0.546***(0.096)Infrastructure satisfied -0.587**(0.201)0.574**(0.199)Infrastructure general -0.494**(0.089)0.497**(0.085)Neighborhood relationship harmonious -0.374**(0.251)0.372**(0.247)Neighborhood relationship general -0.328**(0.218)0.323**(0.225)Environment sanitation satisfied -0.304**(0.042)0.322**(0.068)Environment sanitation general -0.257*(0.027)0.258*(0.021)Poverty alleviation policy well known -0.073(0.017)0.081(0.029)Poverty alleviation policy generally known -0.054(0.014)0.059(0.025)Number of samples 1 5951 5951 595Chi-square test81.26**75.29**80.54**

Note: The numbers in parentheses are robust standard errors, and***,**, and*denote statistical significance at 1%, 5%, and 10%, respectively.

4.1Influenceofmaterialsecurityontheimplementationeffectofthepolicyofrelocatingthepoor

4.1.1The per capita income has the most significant influence on the effect assessment of the policy of relocating the poor. The per capita income variable passed the significance test at 1% level in Model 1 and 5% level in Model 3, and the coefficient were positive, indicating that under the premise of other conditions not changed, compared with the low per capita income group, the villagers with higher per capita income were more likely to give satisfied assessment, and the higher the per capita income, the better the policy implementation effect assessment, which is consistent with the Hypothesis 1. Many scholars have studied the positive correlation between satisfaction and income from the perspective of economics[15], and 20% of the respondents said that the per capita income after relocation was more than 10 times before the relocation, and high income is an essential foundation for poverty-stricken people to getting rid of poverty, thus it had a great influence on the effect assessment of the policy of relocating the poor.

4.1.2The income-increasing channel has a great influence on the effect assessment of the policy of relocating the poor. The income-increasing channel variable passed the 5% level significance test in both Model 1 and Model 3, and their coefficients were positive. Under the premise of other conditions not changed, compared with the group of "no increase in the income-increasing channel", villagers with more income-increasing channels were more likely to give satisfied assessment. In the questionnaire, there was a question of "content of income-increasing channels". Through statistical analysis, 41.3% respondents selected construction, catering and transportation, 21.7% selected large scale planting, breeding, village-level public welfare post, and 15% selected participation of various training, migrant work, and significant increase in economic income. The policy of relocating the poor is not just helping the poor to move out the place, but should focus on changing the poor industry and getting rid of the poor root[16]. Thus, many relocated villagers started to get on the diversified road of getting rich and have more recognition of the policy of relocating the poor.

4.1.3The per capita housing has a great influence on the effect assessment of the policy of relocating the poor. The results of Model 1 and Model 3 indicate that both passed the 5% level significance test and the coefficients were positive. Under other conditions not changed, compared with the relocated villagers with per capita housing less than 25 m2, the villagers with per capita housing of 25-30 m2were more likely to give satisfied assessment. Many relocated villagers moved from shabby and dilapidated houses to delicate houses with spacious sitting room, new-style kitchen and toilet, thus they were more likely to give higher recognition to the policy of relocating the poor.

4.1.4The endowment security has a great influence on the effect assessment of the policy of relocating the poor. The endowment insurance variable passed the 5% level significance test in both Model 1 and Model 3, and the coefficients were positive. Compared with the "no endowment insurance" control group, the villagers who had bought endowment insurance were more likely to give satisfied assessment. A perfect endowment system is favorable for forming excellent expectation for future life for respondents. Thus, the endowment security exerts a significant and indirect influence on the positive assessment of the policy of relocating the poor.

4.1.5Employment environment is not an important factor influencing the effect assessment of the policy of relocating the poor. The employment environment variable did not pass the significance test in Model 1 and Model 3, indicating that the employment environment is not an important factor influencing the effect assessment of the policy of relocating the poor. Possible reason is that implementing the policy of relocating the poor is an important measure for winning the poverty alleviation project, while the per capita income, per capita housing, and income-increasing channel have greater influence than the employment environment on the implementation effect assessment of the policy of relocating the poor.

4.2Influenceofsocialenvironmentontheimplementationeffectofthepolicyofrelocatingthepoor

4.2.1Influence of public service on the effect assessment of the policy of relocating the poor. The public service variable passed the 5% level significance test in both Model 2 and Model 3, and their coefficients were positive. Under the premise of other conditions not changed, compared with the group of "unsatisfied with public service", villagers who were satisfied with public service were more likely to give satisfied assessment. Through relying on the implementation of the policy of relocating the poor, it has built village-level organization activity places, health centers, schools, nursing homes, rural libraries, hard cultural activity squares, which further improved the public service ability including medical care, education, culture, and environmental protection, effectively solved the problems of difficulties in seeking medical care and going to schools, finally, they would have a satisfied assessment on the implementation effect of the policy of relocating the poor.

4.2.2Influence of infrastructure on the effect assessment of the policy of relocating the poor. The results of Model 2 and Model 3, under the premise of other conditions not changed, the infrastructure showed significant influence on the satisfaction of villagers with the policy of relocating the poor, and the higher the satisfaction with the infrastructure, the higher the recognition of the policy of relocating the poor. The main reason is that relocated people lived in areas with poor natural conditions in the past, while they moved to areas with better infrastructure and better living environment at present. Better infrastructure solved their difficulties in drinking, transportation, and going to the market, and effectively removed their worry for relocation.

4.2.3Influence of neighborhood relationship on the effect assessment of the policy of relocating the poor. The results of Model 2 and Model 3, under the premise of other conditions not changed, the harmonious neighborhood relationship showed significant influence on the satisfaction of villagers with the policy of relocating the poor, and the higher the satisfaction with the neighborhood relationship, the higher the effect assessment of the policy of relocating the poor. Among the respondents, 94.9% selected harmonious neighborhood relationship, reflecting to a certain extent that harmonious interpersonal relationship rapidly made the villagers rapidly integrate into the resettlement sites, accordingly raising the recognition of the policy of relocating the poor.

4.2.4Influence of environmental sanitation on the effect assessment of the policy of relocating the poor. The environmental sanitation variable passed the 5% level significance test in both Model 2 and Model 3, and their coefficients were positive. Under the premise of other conditions not changed, compared with the group of "unsatisfied with environmental sanitation", villagers who were satisfied with environmental sanitation were more likely to give satisfied assessment. The relocation changed poor living environment in the past. In the resettlement sites, the garbage classification is clear, the debris stack is more reasonable, and villagers live more comfortably, accordingly the environmental sanitation greatly raised their satisfaction with the policy of relocating the poor.

4.2.5Understanding of the poverty alleviation policy is an important factor influencing the effect assessment of the policy of relocating the poor. The understanding of the poverty alleviation policy did not pass the significance test in Model 2 and Model 3, indicating that the understanding of the poverty alleviation policy is not an important factor influencing the effect assessment of the policy of relocating the poor. The possible reason is that nearly 60% of the respondents had educational level of primary school and below. They were rarely concerned about national affairs and were not good at using new media to accept new information. Therefore, they did not thoroughly understand the relevant national poverty alleviation policies and were unable to raise their satisfaction with the policy of relocating the poor.

It should be noted that in the 6 control variables, the family size, party identity, and marital status passed the significance test at 5% level, and the coefficient was positive. Under the premise of other conditions not changed, married respondents and party respondents had higher satisfaction with the policy of relocating the poor, and respondents with more family members had higher assessment of the policy of relocating the poor. The reason is that the larger the family size, the higher the benefits from the relocation, and the greater the influence on the assessment of policy of relocating the poor. Besides, party member respondents had deeper understanding of the national poverty alleviation policy, which raised their satisfaction with the policy of relocating the poor.

4.3RegressionresultsofpovertydegreeTaking respondents from counties of different poverty degrees as objects, we studied the influence of the poverty degree on implementation effect assessment of the policy of relocating the poor. According to the poverty degree differences, we divided respondents into respondents from extremely poor counties, very poor counties, and general poor counties, and we took the general poor counties as the reference group. In this study, we adopted ordinary ordinal Probit model and order Probit model to conduct the item test. Table 3 listed the threshold model regression results. According to the analysis of regression coefficient results, the two models were consistent in the direction of influence, but the overall estimate of the order Probit model was slightly higher than the Probit regression estimate. Regression data showed that samples from both extremely and very poor counties passed the significance test at 5% level and their coefficients were positive. This indicates that under the premise of other conditions not changed, compared with the group of "general poor counties", respondents from severer poor counties had higher satisfaction with the policy of relocating the poor.

Table3Regressioncoefficientresultsofpovertydegree

Variable (reference group: general poor counties)Probit modelCoefficient (standard error)Order Probit modelCoefficient (standard error)Number of samplesRespondents from extremely poor counties 0.615***(0.027)0.622***(0.027)559Respondents from very poor counties0.452**(0.021)0.455**(0.023)492

Note: The numbers in parentheses are robust standard errors, and***,**, and*denote statistical significance at 1%, 5%, and 10%, respectively.

5 Conclusions and recommendations

5.1SignificantinfluenceofmaterialsecurityontheimplementationeffectofthepolicyofrelocatingthepoorThe per capita annual income has the most significant relationship with the implementation effect assessment of the policy of relocating the poor, followed by the income-increasing channels, per capita housing, and endowment security, while there is no significant relationship between the employment environment and the implementation effect assessment of the policy of relocating the poor. Therefore, the government should play a leading role, adhere to both the relocation and development, take the employment and income increase as the core, and make effort to raise the sustainable development capability of relocated people. (i) For the disabled and half-disabled relocated poor people, it is recommended to take social protection policies to provide basic living guarantees, increase government transfer payments, establish dynamic poverty standard adjustment mechanisms, and raise the purchasing power of poverty-stricken people. (ii) For relocated people who have the working ability, it is recommended to make full use of the characteristic advantages of the resettlement area to foster the development of industries with large employment capacity. Local governments should encourage more people to work in the resettlement areas through financial funds and other supports. (iii) It is recommended to promote and help relocated people to develop production and increase income through strengthening the support for farmers’ specialized cooperatives and leading enterprises. (iv) It is recommended to increase the non-agricultural income of relocated people through providing employment services and skill training, especially trainings with high pertinence and income, and necessary skills for migrant work.

5.2SignificantinfluenceofsocialenvironmentontheimplementationeffectofthepolicyofrelocatingthepoorThe public service has the most significant relationship with the implementation effect assessment of the policy of relocating the poor, followed by the infrastructure, neighborhood relationship, and environmental sanitation, while there is no significant relationship between the understanding of the poverty alleviation policy and the implementation effect assessment of the policy of relocating the poor. Therefore, it is recommended to focus on the precision poverty alleviation, and take the public service and infrastructure as the new breakthrough in the relocation project, to mobilize the poor people to fully participate in the poverty alleviation process and stimulate the endogenous force of poverty alleviation. (i) It is recommended to gradually increase the financial input of governments at all levels, improve the public services and infrastructure of water, electricity, roads, education, medical care,etc., which have the most direct relationship with the living and development of the people at the resettlement areas, and focus on resolving the housing, medical care, social security, and children education of the relocated people. (ii) It is recommended to fully mobilize the enthusiasm of the relocated people to participate in poverty alleviation and mobilize the poverty-stricken people to actively participate in the design, decision-making, implementation and supervision of poverty alleviation projects. (iii) It is recommended to strengthen the poverty alleviation from concept, actively guide the poverty-stricken people to set up the concept of relying on their own efforts, to actually realize the transformation from "helping us to get rid of poverty" to "wanting to get rid of poverty".

5.3RelocatedpeoplefromextremelypoorcountieshavinghigherassessmentfortheimplementationeffectofthepolicyofrelocatingthepoorThrough survey, we found that the reasons for poverty mainly include harsh natural and ecological environment, backward infrastructure and social development, low human capital stock, and low social civilization. On June 23, 2017, General Secretary Xi Jinping delivered a speech, at the Session on the Fight against Poverty in Deep Poverty-Stricken Areas, that deep poverty-stricken area is the top priority of poverty alleviation project, and it is required to strengthen the support security system, increase the policy inclination effort, and ensure the poverty-stricken people to enter the moderately prosperous society. Therefore, in addition to the investment in the resettlement sites, it is recommended to make effort to develop industries, increase income, and improve public services and infrastructure, so as to improve the living and production conditions of relocated people. The fundamental point is to improve human quality and strengthen the human resource development. (i) It is recommended to ensure children of poverty-stricken families to receive regular education, to cut off the poor roots through education-oriented poverty alleviation. (ii) It is recommended to provide healthy and nutritious life for the relocated people, and strengthen and improve the new rural cooperative medical care system. Only in this way, may it be able to promote the improvement of self-development ability of relocated people through the accumulation of human capital.