Life Satisfaction and Rural-Urban Migration in China: The Role of - - PDF document

1

Life Satisfaction and Rural-Urban Migration in China: The Role

• f Income and Relative Standing

Abstract Rural-urban migrants are supposed to be less satisfied with their life, but why they still moved from rural to urban area in China for several decades? Employing national data from eight cities in China, we focus on the specific role of income and relative standing on rural-urban migrants’ life satisfaction through comparing with urban-urban migrants. The results show that the influence of income on rural-urban migrants’ life satisfaction is far stronger than relative standing. Within the relative standing, the effects of dual reference groups on their life satisfaction are asymmetric, namely, the reference group in the destination city has a larger effect than the reference group in the hometown. However, only relative standing matters for urban-urban migrants, and the effects of dual reference groups are symmetric. We take these as evidence that income increase does play an essential role in improving rural-urban migrants’ life satisfaction, which possibly offset the negative influences of poor

• bjective well-being, and thus enhances their stay in the city. Moreover, the favorable

relative standing compared with destination residents could bring rural-urban migrants with better subjective wellbeing and positive signals of future living, inspiring them to keep working in city instead of returning to their original countryside. Keywords rural-urban migrants; life satisfaction; income; relative standing; urban- urban migrants 1.Introduction One of the most widely accepted arguments in migration literature is that economic migrants move because they are not satisfied with their current situation and inclined to seek better opportunities in wealthier destination. People view the migration decision as one of utility maximization, i.e., they migrate to become better off in some subjective sense (Ziegler & Britton, 1981). Migrants’ life satisfaction is derived largely from their economic standing. To reinforce this reasoning, the literature on return migration shows that migrants will try to correct their decision if they do not realize economic gains, i.e., if they are not satisfied with their situation after migration

2

(Constant & Massey, 2002; Cassarino, 2004). Later studies extended to other potential gains, including their children’s well-being, that can affect migrants’ satisfaction. However, the rural-urban migration in China in the last few decades, one of the biggest mass migrations in recent human history, presents a different picture. Because

• f their non-native and agricultural household registration (Hukou) status, most rural-

urban migrants are excluded from local educational resources, citywide social welfare programs, and many jobs (Chan & Zhang, 1999). They are segregated from urban residents both economically and socially. The expected occupational mobility is virtually absent because of the strict Hukou system. In addition, their moving and adaptation costs are substantial. For rural migrants, working in the city suggests the huge cost of separation from their families and children. For those who brought their children with them, the difficulty of receiving education of high quality is unlikely to change in the near future. Although the government implements a series of public policies to improve rural-urban migrants’ integration, the rules preventing them from changing their rural Hukou to city registration will not be lifted in the foreseeable future. In small and medium cities, the authorities adopt more liberal policies for rural-urban migrants to register urban Hukou in the local city, whereas the big cities like Beijing, Shanghai, Guangzhou, etc., where gather the most part of rural-urban migrants still have serious restrictions of household registration. Therefore, the short-term and long- term expected gain of changing to urban Hukou status is virtually non-existent. Migrants cannot be satisfied living in such a situation. Yet, despite all these adverse factors and difficulties, migrants generally do not return to their rural villages, and new migrants continue to move to the city. In fact, official data indicate that the number of rural-urban migrants has increased over several decades, and the total number of rural- urban population reached 277.47 million in 2015, an increase of 3.52 million from

• 20141. Why does such a paradox exist?

To address the issue, we enter the debate with growing literature on the topic of life satisfaction among migrants. We highlight that most studies focus on the economic and structural situation at the destination. Yet, we argue the importance of migrants’ subjective evaluation of their economic standing through social comparisons. The study

1 The data come from Monitoring Report of Rural-urban Workers in China for 2015 which is

published online by National Bureau of Statistics of the PRC in April 28th, 2016 http://www.stats.gov.cn/tjsj/zxfb/201604/t20160428_1349713.html.

3

focuses on migrant’s perceived economic ranking relative to their counterparts in home

• rigin and in destination city, and investigate how absolute income and relative standing

relate to internal migrants’ life satisfaction in the developing countries. In addition, to identify the specific role of income and relative standing among rural-urban migrants,

• ur study considers the contextualization of urban-rural inequality inside Chinese

internal migrants, and adds the urban-urban migrants case to make a comparable analysis. 2.Rural-urban migrants’ life satisfaction Rural-urban migrants’ objective wellbeing in China has been received lot of research attention, while the studies on their subjective wellbeing, also referred to as “satisfaction with life” or “happiness” are emerging but limited. Using 2002 national data, Knight and Gunatilaka (2010, 2012) find rural-urban migrants has the lower level

• f happiness in comparison with both rural and urban residents. The explanation is that

rural-urban migrants raise their earning aspirations in the cities, influenced by the new surroundings, and the false expectations in relations to realistic achievement lead to their unhappiness. Nielsen et al. (2010) employ the personal well-being index (PWI) to examine the domain-level representation of subjective life satisfaction among 525 rural-urban migrants in Fujian province. The results reveal a moderate level of subjective well-being within the normative range. It is likely because for the rural-urban migrants who find it too difficult to cope in the cities, the fallback of returning to their hometown in the countryside could buffer the negative influence of those hard challenges on life satisfaction. Gao and Smyth (2011) seek to answer what keeps China’s migrant workers going despite the poor living and working conditions through examining the determinants of their happiness. Based on the data collected from 12 cities in China, they discover that the optimistic expectations as to future income has a positive effect on rural-urban migrants’ happiness. Chu and Hail (2014) investigate 1225 rural-urban migrants in Shanghai, and use Satisfaction with Life Scale (SWLS) and PWI to evaluate their life satisfaction situation. The findings show that the overall life satisfaction score among rural-urban migrants is below the normative range for Chinese populations in rural and urban areas. What’s more, there is an uncovered a U- shaped relationship between income and overall life satisfaction among rural-urban migrants. 3.Income, relative standing, and life satisfaction

4

A well-established hypothesis is that absolute income is a primary predictor of individual welfare (Diener, 1984). In economic theory, higher income is associated with higher utility function because income rise would increase individuals’ consumption level, which leads directly to greater wellbeing (Posel & Casale, 2011; Gokdemir & Dumludag, 2012). Subjective wellbeing increases with absolute income rise, not only between rich and poor members within a country, but also between richer and poorer countries (Becker et al., 2008). However, “Easterlin’s paradox” suggests that happiness levels are constant or decreasing despite income increase in the developed countries (Easterlin, 1974). As similar with this paradox, a survey of Chinese migrant workers in 2012 shows that the level of happiness and life satisfaction for migrant workers, especially young migrants, are not directly correlated with the economic growth of the destination cities, considering that four Chinese cities with the highest GDP- Shanghai, Beijing, Guangzhou and Shenzhen – are among the bottom eight in a ranking of 20 cities in the country based on how migrant workers feel about their lives2. In view of the limited effect of absolute income, researchers begin to complement

• ther explanations than income for subjective wellbeing. One of the most popular

considerations is that personal subjective well-being depends not only on absolute income but also to a large degree on relative standing (Easterlin, 1974; Diener et al.,1993; Ferrer-i-Carbonell, 2005; Luttmer, 2005). In social comparison theory, individual utility stems from comparing their material achievement with reference groups (Distante, 2013). Individuals will feel satisfied in the case that they perceive their economic standing is higher than the status of reference groups (Diener & Lucas, 2000). Chang (2013) links the mechanism between relative standing and life satisfaction, suggesting that increase in relative standing leads to a gain in positional identity of social status and thus raises the level of life satisfaction. When individuals make social comparisons, there are a series of reference groups including internal norms and external norms. Internal norms refer to using past or expected future economic standing as reference groups (McBride, 2001; Bookwalter & Dalenberg, 2010); and external norms define reference groups as other individuals with similar demographic and social characteristics (McBride, 2001; Clark et., 2008). With respect to international migration studies, on the question of whom migrants use as

2 The evidence is from the news titled “Top cities no happy haven for migrants”, which is

published online by China Daily in the April 28th, 2012. http://www.chinadaily.com.cn/business/2012-04/28/content_15168978.htm

5

external reference groups, different theories provide different views. The older theories assume that immigrants maintain reference groups in the home country to make social comparison (Piore, 1979), or that immigrants would finally employ their new country as reference group in the long run (Stark, 1991). However, in the transnational theories, it is believed that immigrants maintain a dual frame of reference (Basch et al., 1993). Gelatt (2013) analyses the association of relative standing and subjective Well-Being among Asian and Latino Immigrants in the United States. It turns out that immigrants keep simultaneous reference groups in both the United States and the country of origin, which is consist with the transnational theories. For Chinese rural-urban migration, some empirical studies have examined how migrants’ relative standing in terms of internal comparison, such as economic expectation, affects their life satisfaction and happiness (Knight & Gunatilaka, 2010, 2012; Gao & Smyth, 2011), while the researches on relationship of life satisfaction and relative standing with external reference groups are very few. Therefore, it is necessary to know whether the findings among international immigrants apply to rural-urban migration in the developing countries. 4.Contextualization of the correlates of life satisfaction within Chinese internal migrants When it comes to rural-urban migration in China, the contextualization of Hukou system must be taken into consideration. Under the Hukou system, the principle of localization management categorizes residents into natives and non-natives, and the principle of identity management categorizes residents into rural inhabitants with agricultural Hukou status and urban inhabitants with non-agricultural Hukou status. Based on the interaction of these two principles, Chinese internal migrants are mainly constituted by two groups. One is well known as rural-urban migrants, which are the temporary non-natives coming from countryside with agricultural Hukou status. The

• ther is urban-urban migrants, who come from other cities with non-agricultural Hukou

status and now live temporarily in the destination. The percentage of urban-urban population has accounted for about one third of total internal migrants in China, which become a part and parcel of internal migrants, and is believed to continue to rise in the future3.

3 The data comes from the news published online by National Health and Family Planning

Commission of PRC in October 19th, 2016. http://www.nhfpc.gov.cn/xcs/s3574/201610/a6d3a604596a4ca3acf0dad31d891c13.shtml

6

The similarity of rural-urban and urban-urban migrants lies in the same restrictions and limited access to the local public resources and social services in the destination city, such as employment, health care, children’s’ education, social insurance, etc., because both of them hold non-native Hukou under the principle of localization

• management. The disparity between rural-urban and urban-urban migrants is due to

their agricultural versus non-agricultural Hukou status under the principle of identity management, which is attached with the imbalanced development between rural and urban China. The sharp rural-urban inequality, such as education attainment, employment opportunity, occupational class, economic gain, social welfare programs, etc., is bound to generate the disadvantages of pre-migration conditions and post- migration integration for rural-urban migrants in comparison with urban-urban

• migrants. Therefore, it makes us notice such a stratification between rural-urban and

urban-urban migrants, and theoretically relevant to study whether and how the correlates of life satisfaction are contextually across systematically divided groups inside the internal migrants. To summarize, this paper seeks to gap in the literatures of rural-urban migrants’ life satisfaction as follows: first, the effects of absolute income and relative standing through external social comparison will be tested simultaneously to examine whether the conclusions for general population are applicable to rural-urban migrants in the developing countries; second, the reference groups of home origin and destination city will be included simultaneously to see the different influence types of dual reference groups on migrants’ life satisfaction; third, particular attention is devoted to identifying the specific role of income and relative standing among rural-urban and urban-urban migrants respectively to enhance our understanding of the contextualization of the correlates of life satisfaction within internal migrants in China. 5.Data, measures, and methods 5.1. Data The data used in this study comes from the National Migrant Population Dynamic Monitoring Survey in 2014 organized by National Health and Family Planning Commission of PRC. The objects of this investigation are limited to those internal migrants with an age range from 15 to 59, who as non-native Hukou holders have been migrated to the destination city for over a month. A special part investigating internal migrants’ social integration and mental health was only conducted in eight cities: Jiaxing, Shenzhen, Zhongshan, Qingdao, Xiamen, Zhengzhou, Chengdu and Chaoyang

7

District in Beijing. The PPS sampling method was used to choose respondents from each city. Internal migrant respondents in these eight cities totaled 15,999, including 13,927 rural-urban migrants and 2,072 urban-urban migrants. It’s important to point out that we use the individual standardized weight included by dataset itself when taking the analysis in order to adjust and balance the sample structure on the basis of scale and demographic characteristics of national migrant population distributing in the eight cities. 5.2. Dependent variable Life satisfaction is measured by using the Satisfaction with Life Scale (SWLS). The scale consists of five items assessing the way persons evaluate their lives. Seven- point-scale response categories are provided for the respondent to choose from: strong disagree, disagree, slightly disagree, neither disagree or agree, slight agree, agree, and strongly agree. Responses are coded from 1 to 7 so that higher values indicate more satisfied individuals feel with their life. An indicator of life satisfaction is created by taking the sum of the responses to these five statements, yielding minimum and maximum scores from 5 to 35. The Cronbach’s α for the indicator in the total sample is 0.863, indicating a good internal consistency. 5.3. Independent variables Income is estimated by calculating the common logarithm of respondent’s monthly income plus 1 CNY. Relative standing is measured by a symbolic ladder with 10 rungs representing different social status in society (Adler et al., 2000). About the economic standing relative to reference group in the home origin, the respondents are asked to, “What is the number to the right of the rung where you think you stand, relative to other relatives, friends or colleagues in your hometown?” Similarly, about the economic standing relative to reference group in the destination city, the respondents are asked to, “What is the number to the right of the rung where you think you stand, relative to other relatives, friends or colleagues in this city?” Those who have the highest income and best occupation stand at the top of the ladder, i.e., in the tenth rung; in contrast, the people who have the lowest income and worst occupations are at the bottom of the ladder, i.e., in the first rung. Higher number the respondents chosen, higher economic standing they perceived by themselves. We will treat them as the continuous variables when running the statistical models, but still give their category distributions in the description part. 5.4. Control variables

8

The analysis also uses a range of control variables including demographic characteristics, current living conditions and migration characteristics. With respect to demographic characteristics, age, gender, ethnic group and marital status are considered into the statistical models. Gender is dichotomized male versus female (reference group). Ethnic group is categorized into two groups (Han nationality and Minority nationality), where Han nationality, the larger group, is the reference group. Marital status is dichotomized married versus non-married (reference group). Married status refers to “first-married” and “remarried”. Non-married includes “single” “divorced” and “widowed”. Regarding living conditions, community environment is assessed by a dichotomized variable indicating whether the respondent lives in the “village-in-city”, “shanty town”, “suburban area”, and “rural community” or not. In those places, the house quality is usually poor, the surrounding environment is unpleasant, and the infrastructure is incomplete and worn-out. It is coded 1 if the respondent lives in the worse communities above, and coded 0 if the respondent lives in other common or better communities such as “commodity houses community”, “affordable houses community”, “enterprise community”, “public institutions community”, etc. Neighbor composition is measured by having respondents answer how their neighbors constitute, and categorized into three types including “more natives”, “half natives and half migrants”, and “more migrants”, where “more natives” is the reference group. Home

• wnership is a dummy variable which is coded into 1 if respondent has bought or built

the house by themselves in the destination city, and coded into 0 if respondent does not have any house property in the destination city. We further control for three factors relating to their migration experience. Duration

• f stay represents the number of years that migrants have worked at the destination city.

Migration scope is categorized into three groups: trans-province, trans-city, and trans-

• country4. We treat the trans-province as the reference group. Family migration status, a

dummy variable, is coded into 1 if respondent migrants with at least one family member, and coded into 0 if respondent migrants alone. The weighted descriptive statistics of all

4 In China, there are five practical levels of local government: the provincial, prefecture (mainly

including prefecture-level cities), county, township, and village. According to the administrative division, trans-province indicates that migrants come from other provinces outside the province where destination city belongs to. Trans-city indicates that migrants come from other cities inside the province where destination city belongs to. Trans-country indicates that migrants come from the countries inside the destination city.

9

variables are listed in the Table 1.

Table 1 Descriptive characteristics of internal migrants by Hukou status Total sample Rural-urban migrants Urban-urban migrants Mean life satisfaction 21.43 21.38 21.67 Mean Income (CNY) 3860.28 3626.96 4995.64 Mean relative standing: home origin 5.68 5.64 5.86 Relative standing: home origin % 1 1.57 1.57 1.57 % 2 2.15 2.30 1.42 % 3 5.61 5.74 5.00 % 4 8.50 8.51 8.43 % 5 30.85 31.63 27.03 % 6 22.94 22.85 23.37 % 7 14.39 14.09 15.88 % 8 9.06 8.55 11.53 % 9 3.19 3.16 3.36 % 10 1.74 1.61 2.41 Mean relative standing: destination city 5.39 5.36 5.54 Relative standing: destination city % 1 1.96 1.93 2.13 % 2 3.06 3.26 2.10 % 3 6.50 6.39 7.04 % 4 12.08 12.23 11.35 % 5 31.97 32.84 27.70 % 6 22.39 22.24 23.11 % 7 11.78 11.39 13.67 % 8 6.96 6.46 9.36 % 9 2.17 2.16 2.18 % 10 1.14 1.09 1.36 Mean Age 32.65 32.47 33.51 Gender % Female 50.30 50.45 49.58 % Male 49.70 49.55 50.42 Ethnic group % Han nationality 96.51 96.38 97.15 % Minority nationality 3.49 3.62 2.85 Marital status % Non-married 22.45 22.35 22.93 % Married 77.55 77.65 77.07 Community environment % Worse community 63.67 66.83 48.27 % Common or better community 36.33 33.17 51.73 Neighbor composition % More natives 11.13 10.67 13.52 % Half natives and half migrants 24.22 23.55 27.67 % More migrants 64.64 65.78 58.81 Home ownership % Having house property 9.70 7.04 22.61

10 % No house property 90.30 92.96 77.39 Mean Duration of stay (year) 4.40 4.29 4.92 Migration scope % Trans-province 66.92 66.29 69.99 % Trans-city 31.78 32.53 28.14 % Trans-country 1.30 1.18 1.87 Family migration % Migrating alone 23.46 23.25 24.47 % Migrating with families 76.54 76.75 75.53

5.5. Analytic strategy When we analysis the relationship between economic status and life satisfaction, we use the two-stage lease squares (2SLS) regressions to avoid the endogeneity produced by taking income as the independent variable. We choose the following instrumental variables: educational attainment, employment status, and inflowing

• region. Educational attainment is categorized into four levels: “primary school or

below”, “junior high school”, “senior high school or technical secondary school”, and “college or university”. In analysis, we treat “primary school or below” as the reference

• category. The Employment status is divided into four types: “the non-employed”,

“employee”, “employer”, and “the self-employed”, where “the self-employed” is the reference group. Based on the geographical positions of respondents’ destination city, the inflowing region is categized into “western area”, “central area”, and “eastern area” (reference group). The weighted descriptive statistics of instrumental variables and results of first-stage regressions are listed in the appendix respectively. We stratify the sample with Hukou status (agricultural versus non-agricultural) and perform separate analyses for the rural-urban migrants and urban-urban migrants, because we expect that income and relative standing have different effects on the life satisfaction for these two groups. Therefore, we perform this analysis in three models. Model 1 is a full model with the total sample, which tests the influence of income and relative standing on internal migrants’ life satisfaction. More importantly, we focus on the effect of Hukou status on life satisfaction after controlling other variables in order to learn whether there is a significant difference of life satisfaction level between rural- urban and urban-urban migrants. In sub-model 2 and sub-model 3, we test whether and how income and relative standing status affect life satisfaction for rural-urban and urban-urban migrants respectively.

• 6. Findings and discussion

11

6.1 Descriptions of income, relative standing and life satisfaction Table 1 provides separate descriptive characteristics for rural-urban migrants and urban-urban migrants. There is a large gap of economic standing between them. For rural-urban migrants case, their monthly income is 3626.96 CNY, which has 1368.68 CNY less than urban-urban migrants’ monthly income (4995.64 CNY). Moreover, Rural-urban migrants report lower relative standing than urban-to-urban migrants no matter with reference group in home origin or in destination city. Considering the range

• f relative standing scale is from 1 to 10, we assume that those who choose over 5 could

perceive higher economic standing than their counterparts in hometown and destination

• city. The percentage of rural-urban migrants with higher economic ranking relative to

their counterparts in home origin is smaller 6.29 percent point than that of urban-urban

• migrants. The percentage of rural-urban migrants with higher economic ranking relative

to their counterparts in destination city is still less 6.34 percent point than that of urban- urban migrants. Besides the poor situation of income and relative standing, rural-urban migrants are also in more unfavorable living conditions. Specifically, more than 60% rural-urban migrants live in the worse community environment, which is larger 18.56 percent points than urban-urban migrants. The percentage of rural-urban migrants living in the neighborhood mainly constituted by internal migrants takes up 65.78%, while that of urban-urban migrants is 58.81%. It indicates that rural-urban migrants’ habitation is more likely concentrated, closed, and homogeneous. What’s more, the percent of those who have house property in destination city among rural-urban migrants is only 7.04%, far lower than the percentage of urban-urban migrants (22.61%). All above shows that objective wellbeing of rural-urban migrants is worse in comparison with the urban-urban migrants. However, when coming to subjective wellbeing, the situation among rural-urban migrants seems not bad. Their level of life satisfaction is 21.38, falling into the category of “slightly satisfied” (21-25), which indicates that rural-urban migrants are a little bit content with their current life. In addition, the level of life satisfaction among rural-urban migrants is very close to urban- urban migrants (21.67), and the magnitude of the difference is only about 0.29. 6.2 The influence of income and relative standing on life satisfaction Table 2 illustrates the 2SLS regression results on income, relative standing and life satisfaction among internal migrants in China. The houseman tests are all significant for three models, which demonstrates that there do exist endogeneity between

12

independent and dependent variable, and 2SLS regression is necessary to replace OLS regression when we choose the statistical method. As indicated in model 1, both income and relative standing status have significant effect on internal migrants’ life satisfaction. Higher income internal migrants have, more satisfied they feel with life (coefficient=3.479, p<0.001). Higher relative standing (including both reference groups in hometown and in destination city) internal migrants perceived, higher level of life satisfaction they have (coefficient=0.401, p<0.001 and coefficient=0.762, p<0.001 respectively). More importantly, after controlling for demographic characteristics, current living conditions and many other migration characteristics, internal migrants’ Hukou status significantly influences their level of life satisfaction. Specifically, those who hold agricultural Hukou, i.e., rural-urban migrants are more satisfied their life than those who hold non-agricultural Hukou, i.e., urban-urban migrants (coefficient=1.006, p<0.001).

Table 2 Two-stage least square (2SLS) regression results Life satisfaction Model 1 Total sample Model 2 Rural-urban migrants Model 3 Urban-urban migrants Hukou status (rg. Non-agricultural) Agricultural 1.006*** Independent variables Income (log) 3.479*** 4.385*** 0.057 Relative standing: home origin 0.401*** 0.347*** 0.719*** Relative standing: destination city 0.762*** 0.753*** 0.705*** Demographic characteristics Age 0.037*** 0.044*** 0.021 Male

• 0.551***
• 0.722***

0.002 Minority nationality 0.400 0.116 1.997* Married

• 0.002
• 0.309

0.722 Living conditions Worse community

• 0.082

0.131

• 1.009***

Neighbor composition (rg. More natives) Half migrants and half natives

• 0.761***
• 0.620**
• 1.464**

More migrants

• 1.425***
• 1.363***
• 1.786***

Having house property 1.222*** 1.405*** 0.915* Migration experiences Duration of stay 0.049*** 0.058*** 0.032 Migration scope (rg. Trans-province) Trans-city migration 0.222* 0.085 0.764* Trans-country migration 0.531 0.299 1.202 Migrating with families 0.406* 0.467* 0.368 Intercept 1.270

• 0.698

12.764* N 13610 11940 1670

• adj. R2

0.125 0.107 0.205

13 Hausman test 130.94** 518.98*** 25.91* Note, instrumented: income; instruments: education attainment, employment status and inflowing region; *P<0.05, **P<0.01, ***P<0.001.

Then we created the sub-models for rural-urban migrants and urban-urban migrants respectively to study how the income and relative standing affect their life satisfaction, and to summarize the specific role of these determinants for rural-urban

• migrants. In sub-model 2, both income and relative standing is significantly predictive
• f rural-urban migrants’ life satisfaction. With rural-urban migrants’ income or relative

standing increasing, their level of life satisfaction correspondently get enhanced. There are two important points deserve our attention. Frist, the coefficient of income is over 12 times larger than the coefficient of relative standing with reference group in home

• rigin, and is nearly 6 times larger than the coefficient of relative standing with

reference group in destination city. It indicates that income has a deeper influence on rural-urban migrants’ life satisfaction than relative standing, considering that the variation of rural-urban migrants’ life satisfaction with absolute income increase is much larger than with relative standing increase. This result is contrast to the general findings that suggest the effect of absolute income on subjective wellbeing is small and marginal, while relative standing tends to have a larger and more important influence (Posel and Casale, 2010). The possible reason is that rural-urban migrants are the representative of economic-driven migrants in the context of clear rural-urban segregation in China, and they always think of making money as their primary and essential goal of working in the city. Thus, the material economic gains in income, rather than the enhancement of perceived social ranking, is much more significant to make them satisfied. Second, each unit increase in relative standing with the reference group in hometown is significantly associated with 0.347 units increase in rural-urban migrants’ life satisfaction, while each unit increase in relative standing with the reference group in destination is significantly associated with 0.753 units increase in rural-urban migrants’ life satisfaction. It indicates the reference group in destination city, i.e., urban residents, has more larger effect on life satisfaction than reference group in hometown, i.e., rural residents. This infers that although rural-urban migrants also keep simultaneously dual reference groups, which is as similar as the international immigrants do, they prefer to compare themselves with the new reference group in destination city. The possible explanation is also dependent on the obvious hierarchy

14

between rural and urban residents. If rural-urban migrants perceive that their economic standing is higher than the urban residents, this sense of inspiration and accomplishment is deeper, which easily make them more satisfied. For urban-urban migrants, the effects of income and relative standing on life satisfaction are very different. Income generates positive influence on life satisfaction, but it is not significant in the statistic (coefficient=0.057, p>0.05). The result for the urban-urban case is in accordance with the “Easterlin Paradox” -- income is increasing while subjective levels are constant. This “threshold effect” of income can be explained by three psychological approaches summarized by Bartram (2011). First is about

• adaptation. The positive influence of income on urban-urban migrants’ life satisfaction

will become limited when they adjust to or get accustomed to the higher income. Second is about aspiration. An increase in income possibly lead to the equal increase in urban-urban migrants’ income aspiration. Therefore, the gap between expectation and achievements still make their life satisfaction remain constant. Third is about social

• comparison. Income rise may have urban-urban migrants employ the richer people as

reference group, which undermines the potential improvement of life satisfaction. Contrast to the role of income, relative standing plays a more important role on urban-urban migrants’ life satisfaction. Both reference groups in hometown and destination city are positively associated with the level of life satisfaction among urban- urban migrants (coefficient=0.719, p<0.001 and coefficient=0.705, p<0.001 respectively). What’s more importantly, the coefficient of relative standing with reference group in hometown is as approximate as the coefficient of relative standing with reference group in destination city, suggesting that the effects of dual reference groups on urban-urban migrants’ assessment of life satisfaction are symmetric. It is reasonable because urban-urban migrants come from urban area, and their reference groups in both hometown and destination city are urban residents where there is no such a striking structural cleavage between them. Therefore, it will not cause such a disparity

• f psychological wellbeing when making social comparisons with them.

In the aspect of living environments, worse community has a negative effect on urban-urban migrants’ life satisfaction (coefficient=-1.009, p<0.001), while it does not have the significant influence on rural-urban migrants’ life satisfaction. Neighbor composition significantly affects both rural-urban and urban-urban migrants’ life

• satisfaction. In comparison with natives-based neighborhood, the neighborhood

composed by half migrants or most migrants are negatively related to migrants’ life

15

• satisfaction. Nevertheless, such influences are still smaller for rural-urban migrants than

that for urban-urban migrants. Having house ownership is positively associated with internal migrants’ life satisfaction, namely, the migrants who have bought or built their

• wn house in the destination city are more satisfied with their life than those who do
• not. The differences in coefficients and significance levels of having house property

between model 2 (coefficient=1.405, p<0.001) and model 3 (coefficient=0.915, p<0.05) indicate that the influence of whether owning house property on life satisfaction is deeper among rural-urban migrants than urban-urban migrants. All these results show that for rural-urban migrants, the environment of community and neighborhood do not affect their life satisfaction as strongly as the way for urban-urban migrants. By contrast, the house property has larger effect on rural-urban migrants’ life satisfaction, which strengthen to some degree the fact that material needs, such as income and house property etc., are more likely to promote rural-urban migrants’ life satisfaction. 7.Conclusion This paper seeks for answering the question: rural-urban migrants are supposed to be less satisfied with their life, but why they still moved from rural to urban area in China for several decades. Analyzing national data from eight cities in China, we focus

• n the role of income and relative standing on rural-urban migrants’ life satisfaction,

and complement the previous studies in the following aspects. First of all, it integrates the examination of absolute income and relative standing through external social comparisons in the background of internal migration in developing countries. In addition, it incorporates both hometown and destination city into migrants’ reference groups when considering the measurements of relative standing, and further test their different effects on migrant’s life satisfaction. Last but not the least, this study emphasizes the economic and social division within Chinese internal migrants, and contextualize the correlates of life satisfaction between rural-urban and urban-urban migrants. It finds that the situation of rural-urban migrants’ life satisfaction is better than that

• f urban-urban migrants after controlling for other social-economic and non-economic
• factors. Both income and relative standing are positively associated with rural-urban

migrants’ life satisfaction, while the influence of income is far stronger. Furthermore, within the relative standing, the effects of reference groups in home origin and in destination city on rural-urban migrants’ life satisfaction are asymmetric. In other words, rural-urban migrants are more likely to be inspired by comparing themselves

16

with urban residents. Contrary to rural-urban migrants, urban-urban migrants’ life satisfaction does not relate to their income, but largely stem from perceiving higher economic standing with reference groups in hometown and destination city. Moreover, the effects of dual reference groups on urban-urban migrants’ life satisfaction are symmetric, suggesting the comparisons with people in the home origin and in destination city have nearly equal effect on urban-urban migrants’ life satisfaction. To sum up, we come back to the question we mentioned at the beginning. It is common to have a misunderstanding that rural-urban migrants’ life satisfaction should be worse to prevent them from working in the city. In the previous studies, we only notice that rural-urban migrants’ subjective wellbeing is worse than the natives, the rural residents and the normal level. However, in fact, their life satisfaction level is not completely at the bottom in the destination city, considering that they at least feel more satisfied than urban-urban migrants. Such a life satisfaction status may provide an argument for supporting the continuous growth of rural-urban migration in the past

• decades. Rural-urban migrants’ life satisfaction mainly derives from the absolute

income instead of relative standing. Unlike urban-urban migrants and general suggestions, the increase in income does play a more essential role in enhancing rural- urban migrants’ life satisfaction, which possibly offset the negative influences of poor

• bjective well-being, strengthen the positive influences of fine subjective wellbeing,

and thus enhances their stay in the city. In addition, contrary to the balanced effect of dual reference groups for urban-urban migrants, the reference group in the destination city is more stimulating for rural-urban migrants’ life satisfaction compared with the reference group in the hometown. That means the favorable relative standing compared with urban residents could bring rural-urban migrants with much more sense of achievement and positive signals of future living, which could encourage them to work in the city instead of returning to their original countryside.

References Adler, N. E., Epel, E. S., Castellazzo, G., & Ickovics, J. R. (2000). Relationship of subjective and

• bjective social status with psychological and physiological functioning: Preliminary data in

healthy, White women. Health psychology, 19(6), 586. Bartram, D. (2011). Economic migration and happiness: Comparing immigrants’ and natives’ happiness gains from income. Social Indicators Research, 103(1), 57-76.

17 Basch, L., Schiller, N. G., & Blanc, C. S. (1993). Nations Unbound: Transnational Projects, Postcolonial Predicaments and Deterritorialized Nation-States. London: Routledge. Becker, G. S., Rayo, L., & Krueger, A. B. (2008). Economic growth and subjective well-being: Reassessing the Easterlin paradox. Comments and discussion. Brookings Papers on Economic Activity, 2008, 88-102. Bookwalter, J. T., & Dalenberg, D. R. (2010). Relative to what or whom? The importance of norms and relative standing to well-being in South Africa. World Development, 38(3), 345-355. Cassarino, J. P. (2004). Theorizing return migration: The conceptual approach to return migration

• revisited. International Journal on Multicultural Societies 6(2), 253–79.

Chan, K. W., & Zhang, L. (1999). The Hukou system and rural-urban migration in China: Processes and changes. The China Quarterly, 160, 818-855. Chang, W. C. (2013). Climbing up the social ladders: Identity, relative income, and subjective well-

• being. Social indicators research, 113(1), 513-535.

Chu, R., & Hail, H. C. (2014). Winding road toward the Chinese dream: The U-shaped relationship between income and life satisfaction among Chinese migrant workers. Social indicators research, 118(1), 235-246. Clark, A. E., Frijters, P., & Shields, M. A. (2008). Relative income, happiness, and utility: An explanation for the Easterlin paradox and other puzzles. Journal of Economic literature, 46(1), 95-144. Constant, A., & Massey, D. S. (2002). Return migration by German guestworkers: Neoclassical versus new economic theories. International migration, 40(4), 5-38. Diener,E. (1984). Subjective well-being. Psychological Bulletin, 95, 542-575. Diener, E., & Lucas, R. E. (2000). Explaining differences in societal levels of happiness: Relative standards, need fulfillment, culture, and evaluation theory. Journal of Happiness Studies, 1(1), 41-78. Diener, E., Sandvik, E., Seidlitz, L., & Diener, M. (1993). The relationship between income and subjective well-being: Relative or absolute? Social Indicators Research, 28(3), 195-223. Distante, R. (2013). Subjective well-being, income and relative concerns in the UK. Social indicators research, 113(1), 81-105. Easterlin, R. A. (1974). Does economic growth improve the human lot? Some empirical

• evidence. Nations and households in economic growth, 89, 89-125.

Ferrer-i-Carbonell, A. (2005). Income and well-being: an empirical analysis of the comparison income effect. Journal of Public Economics, 89(5), 997-1019. Gao, W., & Smyth, R. (2011). What keeps China's migrant workers going? Expectations and happiness among China's floating population. Journal of the Asia Pacific Economy, 16(2), 163-182. Gelatt, J. (2013). Looking down or looking up: status and subjective well‐being among Asian and Latino immigrants in the United States. International Migration Review, 47(1), 39-75.

18 Gokdemir, O., & Dumludag, D. (2012). Life satisfaction among Turkish and Moroccan immigrants in the Netherlands: The role of absolute and relative income. Social indicators research, 106(3), 407-417. Knight, J., & Gunatilaka, R. (2010). Great expectations? The subjective well-being of rural–urban migrants in China. World Development, 38(1), 113-124. Knight, J., & Gunatilaka, R. (2012). Aspirations, adaptation and subjective well-being of rural-urban migrants in China. Adaptation, poverty and development: The dynamics of subjective well- being, 91-110. Luttmer, E. F. (2005). Neighbors as negatives: Relative earnings and well-being. The Quarterly journal of economics, 120(3), 963-1002. McBride, M. (2001). Relative-income effects on subjective well-being in the cross-section. Journal

• f Economic Behavior & Organization, 45(3), 251-278.

Nielsen, I., Smyth, R., & Zhai, Q. (2010). Subjective well-being of China’s off-farm

• migrants. Journal of Happiness Studies, 11(3), 315-333.

Piore, M. J. (1979). Birds of Passage: Migrant Labor and Industrial Society. Cambridge: Cambridge University Press. Posel, D. R., & Casale, D. M. (2011). Relative standing and subjective well-being in South Africa: The role of perceptions, expectations and income mobility. Social Indicators Research, 104(2), 195-223. Stark, O. (1991). The Migration of Labor. Cambridge, MA: Basil Blackwell. Ziegler, J. A., & Britton, C. R. (1981). A Comparative Analysis of Variations in Measuring the Quality of Life. Social Science Quarterly, 62(2), 303.

Appendix

Table 1 Descriptions of instrumental variables by specific Hukou status Total sample Rural-urban migrants Urban-urban migrants Educational attainment % Primary or below 8.28 9.47 2.52 % Junior high 46.29 52.12 17.90 % Senior high 28.69 28.15 31.31 % College or university 16.74 10.26 48.27 Employment status % Employee 60.22 59.27 64.87 % Employer 7.74 7.44 9.24 % Self-employment 19.37 20.42 14.25 % Non-employed 12.66 12.87 11.64 Inflowing region % Eastern area 90.49 89.74 94.15 % Central area 4.38 4.98 1.43 % Western area 5.13 5.28 4.42

19 Table 2 First-stage regressions results Income(log) Model 1 Total sample Model 2 Rural-urban migrants Model 3 Urban-urban migrants Hukou status (rg. Non-agricultural) Agricultural

• 0.057 ***

Independent variables Relative standing: home origin 0.009*** 0.006** 0.024*** Relative standing: destination city 0.009*** 0.012***

• 0.004

Demographic characteristics Age

• 0.001***
• 0.002***

0.001 Male 0.089*** 0.090*** 0.086*** Minority nationality

• 0.030**
• 0.027*
• 0.027

Married 0.067*** 0.078*** 0.013 Living conditions Worse community

• 0.022***

0.014

• 0.025

Neighbor composition (rg. More natives) Half migrants and half natives 0.007 0.038***

• 0.016

More migrants 0.028***

• 0.024 ***
• 0.009

Having house property 0.080*** *** 0.058 *** 0.113*** Migration experiences Duration of stay 0.001* 0.001* 0.000 Migration scope (rg. Trans-province) Trans-city migration

• 0.027***
• 0.016**
• 0.067***

Trans-country migration

• 0.079***
• 0.074***
• 0.085*

Migrating with families

• 0.027**
• 0.036 ***

0.005 Instrumental variables Education attainment (rg. Primary or below) Junior high 0.030*** 0.029***

• 0.017

Senior high 0.078*** 0.081*** 0.047 College or university 0.158*** 0.127*** 0.174*** Employment status (rg. The self-employed) Employee

• 0.058***
• 0.065***
• 0.014

Employer 0.128*** 0.130*** 0.139*** The non-employed

• 0.564***
• 0.604***
• 0.192*

Inflowing area (rg. Eastern area) Middle area

• 0.046***
• 0.045***
• 0.126**

Western area

• 0.098***
• 0.093***
• 0.139***

Intercept 3.434*** 3.387 *** 3.353 *** N 13610 11940 1670 Adj R-squared 0.219 0.187 0.293 Note, instrumented: income; *P<0.05, **P<0.01, ***P<0.001