Frontiers in Business, Economics and Management 
ISSN: 2766-824X | Vol. 7, No. 3, 2023 

 

240 

The	Impact	of	Rural	Labor	Migration	on	Relative	
Household	Poverty:	An	Empirical	Analysis	Based	on	
China	Household	Longitudinal	Survey	Data	(CFPS)	
Peiling Zong*, Wenli Zhang, Yan Wu, Jian Chu 

School of Accountancy, Anhui University of Finance and Economics, Bengbu, 233030, China 

* Corresponding author: ZONG Peiling (Email: 2711384565@qq.com) 
 

Abstract: Based on the data of China Household Tracking Survey (CFPS) in 2 016 and 2018, this paper forms a balance panel 
by using ordinary least squares (OLS). The model and the Probit model are empirical and study the impact of rural labor mobility 
on relative household poverty. The findings suggest that rural labour mobility contributes to reducing relative household poverty 
and remains reliable after robustness tests by changing the measures of the interpreted variables. Therefore, it is necessary to 
reasonably guide the mobility of rural labor, improve the quality of labor quality, and reduce the relative poverty of families. 

Keywords: Rural labor mobility, Relative poverty, Influencing factors, Empirical analysis. 
 

1. Introduction 
After China built a moderately prosperous society in 2020, 

China got rid of the dilemma of absolute poverty, has entered 
the "post-moderately prosperous era" with relative poverty 
and spiritual poverty as the main manifestations. The Fourth 
Plenum of the 19th Central Committee of the Communist 
Party of China proposed to "establish a long-term mechanism 
to solve relative poverty". Relative poverty can be changed 
and adjusted in the concept and measurement of poverty lines, 
different groups of people and dimensions, and different 
countries have different measurement standards. In any case, 
helping the people to lift themselves out of relative poverty is 
a problem that our country is currently facing. In general, the 
specific practice of China's relative poverty management will 
become a long-term undertaking with rich content and 
ambitious goals.  

With the continuous advancement of rural revitalization 
proposed in the "Thirteenth Five-Year Plan" report, more and 
more people have achieved poverty alleviation, in the context 
of the continuous development of urbanization, rural labor 
force into the city to work has become the main method of 
labor poverty alleviation, rural labor flow will further expand 
the scale of the city, improve the industrial structure, 
accelerate the urbanization process, reduce the incidence of 
poverty, thereby promoting the development process of the 
city. Further thinking, while the flow of rural labor promotes 
urban development and solves absolute poverty, what effect 
does it have on the current relative poverty of households in 
Chinese society? It is particularly important to study the 
impact of labor mobility on the relative poverty of working 
households, and answering such questions has important 
theoretical and practical significance for promoting social 
development. 

2. Research Review 
A large number of scholars' research mainly focuses on the 

impact of rural labor mobility on economic development. Tan 
Chang [1]et al. based on the analysis of China Household 
Longitudinal Survey data (CFPS) in 2018 showed that the 

mobility of rural labor significantly promoted the common 
prosperity of rural households, and it was achieved by using 
the ordered response model Ologit and propensity value 
matching ( PSM) method, the conclusions are still robust and 
reliable. Liu Chenhui [2]et al. investigated the impact of 
different types of labor mobility on regional economic 
disparities by constructing peripheral models, and found that 
skill matching is a key factor in whether the regional 
economic gap formed by labor mobility is stable, and labor 
mobility can promote the convergence of regional economic 
gaps, but when the agglomeration effect of high-skilled 
workers and technology spillover Labor mobility does not 
necessarily promote the convergence of regional economic 
disparities. Zhu Yaling [3]et al. used the China Labor Force 
Dynamic Survey Data (CLDS) for empirical testing, and the 
results showed that rural labor mobility induced income and 
cost effects, which had a positive impact on family 
development enjoyment consumption and a negative effect on 
survival consumption, and could significantly improve 
household consumption structure overall 。Li Baojun, Luo 
Jianchao[4]Based and the propensity score matching method 
(PSM), two scholars found that rural labor mobility has a 
significant multi-dimensional poverty reduction effect on 
rural households. Although there is a large number of studies 
on the impact of labor mobility on economic development and 
absolute poverty, the research on labor mobility on relative 
poverty is not complete.  

Fuchs Victor [5](1967) first proposed the concept of 
relative poverty, after which the understanding of poverty in 
economics gradually shifted from absolute poverty to relative 
poverty. Townsend [6](1979) rejects absolute poverty without 
the necessities of life, arguing that poverty will only be 
formed due to the lack of resources, and that poverty will not 
be formed if the population cannot reach the average living 
standard of the general public, and the relativity of poverty is 
absolute, that is, it will exist in any society and at any time. 
With the introduction of relative poverty, more scholars began 
to study how to define relative poverty and how to change the 
situation of relative poverty. Wang Sangui [7]others proposed 
to formulate a multi-dimensional relative poverty standard to 
measure relative poverty from multiple dimensions such as 



 

241 

income, education, health and living standards. Finally, it is 
concluded that in the initial stage of relative poverty, it is 
necessary to implement the "urban and rural two lines", that 
is, the relative income poverty line of urban and rural 
residents should be determined according to 40% of the 
median per capita disposable income of urban and rural 
residents. Zhou Li [8]draws on international experience and 
proposes a trial calculation of China's relative poverty line 
through C FPS data, and the calculated (adult equivalent) per 
capita household disposable income is less than 50% of the 
median disposable income of national residents through 
household size weighting, which can be regarded as relative 
poverty. Significant progress has been made in measuring 
relative poverty and related research, but little research has 
been done from the perspective of the impact of rural labour 
mobility on relative household poverty.  

In summary, there is still room for expansion in the existing 
literature on the impact of rural labor mobility on relative 
household poverty. First, from the perspective of research 
content, the current effect on rural labor mobility is based on 
absolute poverty discussions, which is sometimes led into the 
dilemma of linear thinking. It ignores that after the 
completion of absolute poverty in our country, the relative 
poverty of families is also waiting to be solved; Second, 
scholars rarely use data models to analyze the relationship 
between labor mobility and relative poverty. The main 
marginal contributions of this project are as follows: first, the 
impact of labor mobility on relative household poverty is 
studied from the perspective; Secondly, the relative poverty 
of households is analyzed in detail, and the measures of 
relative poverty are selected; Finally, the empirical analysis 
of China Household Tracking Survey Data (CFPS) was used 
to summarize the relationship between the two influences and 
make relevant suggestions. Therefore, based on the CFPS 
data of 2016 and 2018, this project intends to construct a 
relative poverty line and analyze rural labor mobility, and use 
models to prove and analyze the effect of rural labor 
migration on relative household poverty.  

3. Theoretical Analysis 
According to the Todaro model(To daro, 1969), agricultural 

laborers moved to cities mainly because of the difference in 
expected income between urban and rural areas, and when the 
difference was greater, the more people flowed into the city. 
A large number of laborers choose to work in cities in 
anticipation of higher wages and living standards, which has 
led them to shift from inefficient agriculture to more efficient 
industries and services, thereby increasing economic incomes 
[9] In this process, the theoretical mechanism of rural labor 
mobility affecting relative poverty is mainly reflected in the 
following two aspects: First, rural labor can not only obtain 
higher wage income, but also obtain better public facilities 
and education, and improve the quality of life. The "2021 
Migrant Worker Monitoring and Survey Report" shows that 
in 2021, the average monthly income of migrant workers 
abroad in China was 5,013 yuan. While the monthly income 
of migrant workers has increased, the education of their 
children has also been greatly improved, and the school 
attendance rate of children who migrate with them in the 
compulsory education age is 9 9.6 percent, an increase of 0.2 
percentage points over the previous year, and the poverty 
situation of families has been greatly improved. Second, the 
continuous influx of rural labor into the county is conducive 
to their capital accumulation, so that the next generation is no 

longer migrant workers, obtain better living conditions, and 
improve the endogenous motivation of migrant workers. 
From this a research hypothesis: 

Research hypothesis: Rural labor mobility promotes 
improvements in relative household poverty. 

4. Model Construction, Variable 
Selection and Data Sources 

4.1. Model construction 
To examine the impact of rural labour mobility on relative 

household poverty, the following econometric model is 
developed: 

 
Povertyit = α0 +γMigrationit+βXit +εit       (1) 

 

where the explanatory variable povertyit represents the 
relative poverty status of the family i in year t; The core 
explanatory variable Migrationit represents the rural labour 
mobility of households in year t; Xit represents the control 
variables at the head of household level and at the household 
level that have an impact on the relative poverty of 
households; The coefficient γ reflects the impact of rural labor 
mobility on the relative poverty of households, which is the 
coefficient that this study focuses on. εit is a random 
perturbation item.  

4.2. Variable selection 
4.2.1. Explanatory variables 

The core explanatory variable in this paper is rural labor 
mobility, i.e. the presence or absence of migrant labor in 
households. First, by removing the urban sample in the CFPS 
database, and then screening out the data of "migrant work" 
every year, it is a proxy variable for rural labor mobility. 
Among them, since "going out to work and doing farm work" 
was a combined problem in 2016, in order to accurately 
obtain the data of "farmers going out to work", referring to 
Yin Zhichao’s [10] practice, according to the jump 
relationship of the questionnaire, the families who answered 
the question of "the amount of part-time work sent home" 
were identified as labor-migrant families, and in Consistency 
in the definition of this variable was maintained in the 2018 
data.  

4.2.2. Control variables 
Considering that this paper mainly studies the relative 

poverty of households, and refers to the selection of control 
variables by Gao Mengtao [11]et al., this study selects control 
variables from the household head and household level. 
Among them, the control variables at the head of household 
level include age, gender, years of education, whether he is 
married, whether he is employed, health status; Control 
variables at the household level included household size, net 
income per capita, proportion of elderly people in households, 
and proportion of children in households.  

4.2.3. Variables to be explained 
This paper mainly studies the impact of labor mobility on 

household relative poverty, and the interpreted variable is 
household relative poverty. Most scholars choose to use the 
average of disposable income or 40%~60% of the median to 
measure the relative poverty line, but considering that the 
average is susceptible to extreme values, this paper chooses 
to refer to the relative poverty line set by Wang Sangui [7] and 
Gao Mengtao[11] and other scholars, that is, to calculate the 



 

242 

per capita net income of rural households nationwide, and 
take 40% of the median Confirm whether the sample 
households are in static relative poverty for the poverty line. 
If the income of the sample household exceeds 40% of the 
median net income of rural households, the household is 
considered not to be in relative poverty; Conversely, the 
family is considered to be in relative poverty.  

The main variables of this study and their descriptions are 
shown in Table 1, and their statistical descriptions are shown 
in Table 2.  

4.3. Data sources 
The data in this paper are mainly derived from the China 

Household Longitudinal Survey (CFPS) funded by Peking 

University and the National Natural Science Foundation of 
China and conducted by the China Social Sciences Survey 
Center of Peking University, which reflects the changes in 
China's society, economy, population, education and health 
through tracking and collecting data at the three levels of 
individuals, families and communities. Among them, there 
are detailed data on household migrant data and income and 
expenditure in the family-level questionnaire, which can 
better provide data support for this study. The data has been 
tracked since 2010 and accessed every two years, and data are 
currently available up to 2018. Therefore, this study mainly 
uses data from 2016 and 2018 to form a balanced panel. After 
processing and combining the missing values, outliers and 
extreme values in the data, 3012 valid samples were obtained.  
 

Table 1. Main variables and descriptions of this study 

The variable name Variable definitions 
Rural labour mobility Whether there is labor mobility (yes=1, no=0). 

Head of 
household 

level 

Age/years Age of the head of household 
gender Gender of the head of household (male=1, female=0) 

Years of education/years Years of education for the head of household 
Whether you are married Whether the head of the household is married (married = 1, unmarried = 0) 

Whether or not to be 
employed 

Whether the head of the household is employed (employed = 1, unemployed = 0). 

Health status 
Health status of the head of the household (value 1~5, the larger the value, the better the 

health status) 

Family level 

Family size/person Number of household sizes 
Net income per household LN (Gross Household Income/Household Size) 
The proportion of elderly 
people in the household 

Number of people over 60 years of age in the household/size of the household 

Proportion of children in 
the household 

Number of persons under 14 years of age in the household/size of the household 

Families are relatively poor 
Whether the per capita household income exceeds 40% of the national median per 
capita household income of rural households, and does not exceed, the value is 1; 

Exceeded, the value is 0 

 

Table 2. Descriptive statistics of the main variables of this study 
variable Sample size mean standard 

deviation 
minimum maximum 

Rural labour mobility 3,012 0.500 0.500 0 1 
Age/years 3,012 50.22 11.56 18 87 

gender 3,012 0.527 0.499 0 1 
Years of education/years 2869 6.353 3.978 0 19 
Whether you are married 3,008 2.210 0.807 1 5 

Whether or not to be employed 3,003 1.222 0.735 -8 3 
Health status 3,012 3.169 1.346 -8 5 

Family size/person 3,012 4.101 1.890 1 14 
Net income per household 3,011 15055 18256 0 400000 

The proportion of elderly people 
in the household 

2,942 0.166 0.330 0 3 

Proportion of children in the 
household 

2,942 0.303 0.382 0 3 

Families are relatively poor 3012 0.114 0.317 0 1 
 

5. Empirical Analysis 

5.1. Regression analysis 
Based on the 2016 and 2018 CFPS databases, this study 

mainly uses the ordinary least squares (OLS) model for 
empirical analysis, and based on the relative poverty line 
divided under the 40% standard. The impact of rural labour 
mobility on relative household poverty was studied and data 
from Table 3 were obtained.  

Table 3 shows the regression results based on the OLS 
model. Specifically, after controlling for the relevant 

variables at the head of household level and the household 
level, the coefficient of influence of the core explanatory 
variable rural labor mobility on the relative poverty of the 
households of the interpreted variable was negative, that is, as 
rural labor mobility intensified, the relative poverty of 
households weakened and was 1%. This indicates that labour 
mobility has a significant impact on relative poverty. This 
verifies the poverty alleviation effect of labor outflow, which 
is consistent with the hypothesis proposed in the previous 
research paper, indicating that as the number of rural laborers 
going out to work increases, the relative poverty of 



 

243 

households will gradually weaken and the quality of life will also improve.  
 

Table 3. OLS model regression results 

 Families are relatively poor 

Rural labour mobility 
-0.03060*** 

(0.01137) 

age 
-0.00001 
(0.00059) 

gender 
0.02966** 
(0.01208) 

Years of education 
-0.00690*** 

(0.00163) 

Whether you are married 
-0.00363 
(0.00806) 

Whether or not to be employed 
0.00559 

(0.00731) 

Health status 
0.01428*** 
(0.00491) 

Family size 
0.00161 

(0.00358) 

Net income per household 
-0.00000*** 

(0.00000) 

The proportion of elderly people in the household 
0.05713*** 
(0.02099) 

Proportion of children in the household 
0.02329 

(0.01969) 

Constant terms 
0.14244*** 
(0.04124) 

Sample size 2791 
R2 0.07595 

Note: *, **, **** indicate that they are significant at the level of 1 0%, 5% and 1%, respectively; Standard error in 

parentheses. 

 

Table 4. Robustness test results 

 Probit model 

variable Families are relatively poor 

Rural labour mobility 
-0.166** 
(-2.54) 

age 
-0.008*** 

(-3.37) 

gender 
0.053 
(1.03) 

Years of education 
-0.004 
(-0.62) 

Whether you are married 
0.008 
(0.24) 

Whether or not to be employed 
-0.002 
(-0.05) 

Health status 
0.006 
(0.30) 

Family size 
0.027* 
(1.79) 

Net income per household 
-0.000*** 

(-6.41) 

The proportion of elderly people in the household 
-0.187** 
(-2.42) 

Proportion of children in the household 
-0.159** 
(-2.05) 

Constant terms 
0.590*** 

(3.49) 
Sample size 2,791 

Note: *, **, **** indicate significant at the level of 10%, 5%, 1%, respectively, with the t statistic in parentheses. 
 

5.2. Robustness test 
To further improve the reliability of the results, robustness 

tests are carried out in this paper. 
After drawing on the practices of Cheng Witt [12] and 



 

244 

others, this paper raises the measure of relative poverty from 
40% to 60% of the per capita net income of rural households, 
that is, the per capita net income of households is less than 60% 
of the median  For relative poverty, vice versa, it is not 
counted. In addition, in the robustness test, the variables were 
reduced by 1% and 9 9%. The following regression is 
performed for each variable after changing the relative 
poverty line, and Table 4 shows the regression results based 
on the Probit model.  

As can be seen from Table 4, after replacing the measure of 
the relative poverty line, the coefficient of rural labor mobility 
for the core explanatory variable is -0.166, which is still 
negative and significant at the level of 5%, that is, as the 
number of rural labor mobility increases, the relative poverty 
of households is decreasing, and rural labor mobility 
promotes the improvement of relative poverty of households. 
The regression results of this study were robust and reliable. 

6. Conclusions and Policy 
Recommendations 

6.1. Conclusion 
Based on the China Household Tracking Survey (CFPS) 

data of 2 016 and 2018 rural households in 2018 form a 
balanced panel, by using the ordinary least squares (OLS) 
model and sum The Probit model is empirical and robustness 
tested to study the impact of rural labor mobility on relative 
household poverty. The results show that rural labour mobility 
contributes to the reduction of relative poverty among 
households, based on the relative poverty lines under the 40 
per cent and 6 per cent criteria. In the robustness test, it was 
found that in terms of significance level, the 40% standard 
was more effective than the 60% standard.  

6.2. Policy recommendations 
On the one hand, it is necessary to speed up urbanization, 

realize "urban-rural integration" at an early date, improve 
urban infrastructure construction and public services, strive 
for more welfare policies for farmers, make them have better 
income, better treatment, and a greater sense of belonging to 
the city, so that the lives of migrant workers will be more 
stable. On the other hand, in the context of the implementation 
of land transfer, it is more necessary to increase investment in 
agriculture and scientific and technological construction, 
improve relevant policies, provide more non-agricultural jobs 
for farmers, and reasonably guide the flow of rural labor, so 
as to promote economic development and improve the 
relative poverty of families. In short, it is necessary to jointly 
promote the stability of rural labor mobility and employment 
through multifaceted changes, so as to improve the relative 
poverty of households. 

Acknowledgment 
Fund: Anhui University of Finance and Economics College 

Students Innovation and Entrepreneurship Training Program 
"The Impact of Rural Labor Mobility on Relative Household 
Poverty: An Empirical Analysis Based on China Household 
Longitudinal Survey Data (CFPS)" (No. 202110378059). 

References 
[1] TAN Chang,WU Haitao,PENG Yan. The impact of rural labor 

mobility on the common prosperity of rural households[J/OL]. 
Agricultural Resources and Regional Planning in China. 

[2] LIU Chenhui, CHEN Changshi Labor mobility, skill matching 
and regional economic gap[J] Economic Research Journal, 
2022, 57(07): 45-63. 

[3] ZHU Yaling, ZHANG Bin, MA Yifei The impact of rural labor 
migration on household consumption structure[J]. Consumer 
Economics, 2022, 38(03): 52-63 

[4] Li Baojun, Luo Jianchao An empirical study on the impact of 
rural labor mobility on multidimensional poverty of rural 
households[J]. Statistics and Decision,2022,38(08):22-
26.DOI:10.13546/j.cnki.tjyjc.2022.08.004 

[5] Fuchs V R.Redefiningpoverty and redistributing income[J]. 
The Public Interest,1967,14(08):88. 

[6] Townsend P.Poverty in the United Kingdom:Asurvey of 
household resources and standards of living [M]. 
California:University of California Press, 1979. 

[7] Wang Sangui, Sun Junna China's Relative Poverty Standards, 
Measurement and Targeting after the Comprehensive 
Construction of a Moderately Prosperous Society: An Analysis 
Based on the Data of 2018 China Household Survey [J]. China 
Rural Economy, 2021(03): 2-23 

[8] Zhou Li International experience and enlightenment of relative 
poverty criterion delineation[J] People's Forum ꞏ Academic 
Frontiers,2020(14):70-
79.DOI:10.16619/j.cnki.rmltxsqy.2020.30.019 

[9] YIN Zhichao,LIU Taixing,ZHANG Cheng The influence of 
rural labor migration on household savings rate[J] China 
Industrial Economics, 2020 (01): 24-42. 
DOI:10.19581/j.cnki.ciejournal.2020.01.003 

[10] YIN Zhichao, LIU Taixing, YAN Yu Whether Labor Mobility 
Can Alleviate the Mobility Constraints of Farmers: An 
Empirical Analysis Based on the Perspective of Social 
Network [J]. China Rural Economy, 2021, No.439(07): 65-83 

[11] Gao Meng Tao The alleviation effect of digital economy 
development on relative poverty of households[J]. China 
Circulation Economy, 2023, 37(01): 24-
37.DOI:10.14089/j.cnki.cn11-3664/f.2023.01.003 

[12] CHENG Weit, WU Haitao, JIANG Fan Measurement and 
decomposition of multidimensional relative poverty of urban 
and rural households[J] Statistics and Decision, 2021, 37 (08): 
68-72.DOI:10.13546/j.cnki.tjyjc.2021.08.014