Hukou, social-spatial inequality and migration intention Biqing Li 1 - - PDF document

Hukou, social-spatial inequality and migration intention

Biqing Li1 Yan Tan2 Dianne Rudd3 Susan Oakley4

• 1. Introduction

Rural-to-urban migrants in China work or live in cities without urban household registration status (hukou) (Lin et al. 2011). Rural migrants amounted to 250 million in 2012 due to rapid urbanisation in China. One of the major driving force of this massive movement is inequality (Black et al. 2005; Lipton 1980). In the literature on migration intention (or settlement intention), many factors that measure social inequality has been studied (Cai & Wang 2008; Cao et al. 2015; Yue et al. 2010), so did factors that measure one type of spatial inequality—regional inequality (East-Middle-West) (Zhu 2003). Most studies addressed inequality issues from three dimensional differences— between rural and urban areas, between farmers and urbanites, and between eastern and western regions. There is little research into spatial inequality between tiered cities within the Chinese urban clusters. Spatial inequality in this paper refers only to the hierarchical disparities between tiered cities. Three crucial changes have been overlooked in current research on migration intention. First, the tiered city system calls for more detailed classification of migration intention. Migration intention of migrants was measured primarily through “stay or leave” or “stay in this city, go back to hometown

• r go to other cities” (Cai & Wang 2008; Fan 2011; Hao et al. 2015; Tang & Feng 2015; Zang et al.

2015; Zi-cheng & Wei-guo 2013). These measurements neglected the intrinsic differences of cities

• f different tiers. Second, some dimensions of inequality that migrants experienced in cities are a

combination of both social and spatial disparities. Further study is needed to identify the relationship between the scale of a city (measured as residing population size of cities and the administrative ranks

• f cities) and the types and severity of social inequality in it. Third, hukou policies implemented

across different tiered cities are different (Li et al. 2016; PRC-SCC 2014a, 2014b). There remains unknown how current hukou policy and reforms will impact on migrants’ intention to move or not, and where and when to move. This paper seeks to address three research questions. First, what is the spatial pattern of social inequalities across different tiered cities? Second, how do those social-spatial inequalities influence rural migrants’ future migration intention? Third, how do urbanisation policies, especially hierarchical hukou policies influence rural migrants’ further intention to migrate or not?

1 Biqing Li, PhD student, Hugo Centre for Migration and Population Research, Department of Geography, Environment

and Population, School of Social Sciences, Faculty of Arts, University of Adelaide, Australia

2 Dr Yan Tan, Associate Professor, Hugo Centre for Migration and Population Research, Department of Geography,

Environment and Population, School of Social Sciences, Faculty of Arts, University of Adelaide, Australia

3 Dr Dianne Rudd, coordinator, Department of Geography, Environment and Population, School of Social Sciences,

Faculty of Arts, University of Adelaide, Australia

4 Dr Susan Oakley, Associate Professor, Head of School of Social Sciences, Faculty of Arts, University of Adelaide,

Australia

• 2. Literature review

2.1 Migration theories Since Ravenstein’s (1885) pioneering work on “the laws of migration”, a number of theories have been developed in various social science disciplines to explain the reasons of migration. There are three levels of theoretical knowledge base of migration: individual or household decision making at the micro level, the migrant’s social context at the miso level, and the structural causes and the direction of migration at the macro level (Faist & Faist 2000). Neoclassical economics theory and new economics theory provide micro level explanations to

• migration. Neoclassical economics theory argues that migration is an individual level rational

decision depending on the expected migration return of places of destination and individuals will seek greatest absolute return(Massey 1999). New economics theory of migration, on the other hand, consider migration as a decision made by households instead of individuals and claims that greatest relative return, rather than greatest absolute return is the decisive factor behind migration (Massey 1999). Historical-structural theories and segmented labour market theory center their attention on macro context. Historical-structural theories argue that individuals or households are fundamentally constrained by structural forces. Segmented labour market theory proposed that migration is driven by the intrinsic labour demand of modern industrial societies. World system theory deems migration as the structural consequence of the expansion of markets within a global political hierarchy. Migrant network theory, institutional theory and cumulative theory emphasize the meso-structure mechanisms. Those theories are not contradictory. They supplement each other. All the miso and macro factors have to work through influencing the micro factors to form a migration intention. 2.2 Migration intention Massive migration research has demonstrated that people with intentions to move in an earlier time are more likely to migrate later on than those showing no intentions (Böheim & Taylor 2002; Gordon & Molho 1995). In migration research, intention studies have afforded valuable insight into the influence of perceptions of place utility and the role of individual, familial, and macroeconomic factors in determining destinations chosen (Agadjanian et al. 2008; De Jong et al. 1996; De Jong et

• al. 1985).

2.3 Migration and inequality Among all the determinates of migration intention, the impact of inequalities (both of economy and amenities; of individual/household, social context and structure) were underlined (Castles 2014; Massey 1999). There are mainly two types of inequalities that concern rural migrants in China: social inequality and spatial inequality. Tang et al. (2015) and Zi-cheng et al. (2013) explored the influence

• f inequalities in employment situation to settlement intention in urban areas; Cai et al.(2008), Tang

et al. (2015) and Fan (2011) analysed the relationship between income inequality and rural migrants’ intention to stay in urban areas or to leave; Cai and Wang (2008) also explained the impact of social status disparities on rural migrants’ migration intention. Fan (2011) took the inequalities of social network into account to explain migration intention; Zang et al. (2015) distinguished western, central and eastern part of China when discussing migration intention; Hao et al. (2015) distinguished the

migration intention (to settle down in current city or keep floating) of rural migrants in different level

• f urban unites (administrative level).

2.4 Distribution of inequality in China Unlike many other countries, China’s urbanisation processes is closely linked to the governmental system (Chan 2010). Therefore government policy is responsible for many inequalities during the urbanisation process. The social inequality migrants experienced can be largely attributed to hukou (Afridi et al. 2015). Spatial inequality concerning migrants also derives from “hukou” system. Before implementation of the “reform and opening up” policy in 1978, spatial inequality in China existed mainly between rural and urban areas due to the dual hukou divide between rural and urban areas. At that time, social welfare disparities between different administrative levels of cities were not significant (Li et al. 2016). But since 1978, due to development of China’s administrative system and the reform of hukou system, the pattern of inequality has changed. The disparities among cities, especially between different administrative levels of cities began to grow (Li et al. 2016). China has built a hierarchical city system in accordance to its administrative level and with reference to the formal system of fiscal resources (budget revenue allocation) (Chan 2010). In the meantime, a series

• f hukou reform measures fuzzed up the line between agricultural “hukou” and non-agricultural

“hukou” and reinforced the disparities among different levels of cities. In July 2014, the State Council

• f China (2014a) issued a policy to abolish the functional divisions between agricultural “hukou”

and urban “hukou” and set up a unified rural-urban residential registration system by 2020. In the same year, China’s National policy on urbanization—the New Style Urbanization Plan (2014-2020) proposed to relax “hukou” first and most rapidly in small cities and towns (cities with a population under 0.5 million and official towns). Stringency of qualification for “hukou” registration grow as the scale of the city grow (SCC 2014b). Thus, China is working towards weakening the rural-urban dualism, but has gradually built up a hierarchical rural-urban system to replace it. Thus, it is no longer valid to see all cities as identical when addressing migration intentions in China. Chan (2010) asserted that there was an almost perfect correlation between the administrative ranks and the average population growth rate of cities. The higher the rank, the faster the growth rate of

• cities. Li et al. (2016) stated that large cities placed at the top tier were allocated with substantial fiscal

resources so that they can provide more social services to their residents. But large cities are also most resistant to hukou reforms, and they usually set up strict standards to limit the number of migrants to be qualified for urban hukou status (Li et al. 2016, p. 60). In comparison, smaller cities with lower ranks do not have as much fiscal allocation, and couldn’t provide the same social services to their citizens as those large cities can do. Rural migrants, who are not entitled urban citizenship can’t claim equal social services as urbanities in the same city (Meng & Manning 2010, p. 8; White 2015, p. 288). Carrillo (2011) argued that county level city is more open and flexible in the incorporation of rural migrants. It is assumed that rural migrants residing in large cities placed at the top tier would experience more social inequality, compared to urban citizens, than rural migrants who live in small and low-ranked cities.

• 3. Conceptual Framework

As discussed by Simon (1980), the underlying framework for migration decision includes four general components: 1) background factors (micro, miso and macro), 2) perception of place utility,

3) intention to move, and 4) migration behaviour (De Jong et al. 1985). In this paper, only the relation among 1), 2) and 3) is discussed. 1) Background factors The social-spatial inequalities to be addressed exist in micro, miso and macro levels. But the miso and macro inequalities can be captured in the measures of socioeconomic status and individual perceptions (Du & Li 2012). This paper also looks into one of the sources of China’s inequalities concerning rural migrants: hukou related policies. Hukou and hukou reform policies are a crucial institutional factor influencing migrants’ intention to migrate or not migrate and where to migrate. It influences migration intention through mediating spatial inequality and social inequality. Urban hukou in higher ranked cities is linked to more benefits, which has created enormous “upward” pressure system wide in the last two decades, pushing migrants to move to a higher level city (Chan 2010). The standards of hukou transfer from rural to urban also grow in higher ranked cities, which pushes migrants down to lower ranked units. These information, as well as demographic characteristics serve as the background factors of migration. 2) Perception of place utility To link those inequalities of different levels to individual’s migration intention, this paper brings in the concept of “perceived utility of destination” as an intermediate variable. "Utility" is an economic term introduced by Daniel Bernoulli. “Individuals choose in such circumstances as if they were seeking to maximize the expected value of some quantity. The hypothetical quantity thus defined has been called utility” (Friedman & Savage 1952, p. 463). Utility of a destination refers to the utility that somebody perceives a certain destination. A destination can be seen as an accumulation of certain

• attributes. For every individual rural migrant, there is a certain index system of place attributes to

measure a destination’s utility. The background factors cultivate rural migrants’ index system to evaluate utility of different destinations and this index system will further influence their migration intention and the choice of destination. 3) Migration intention Migration intention in this paper is not limited to whether rural migrants want to leave the city or stay in the city, and whether they plan to return to hometown or go to another city. It distinguishes the administrative rank (and size) of destination unites. As shown in figure 3-1, this paper will specifically look into hukou system, social and spatial inequality, and rural migrants’ destination choices among two tiers of this rural-urban system: county level cities (the lowest level of formally established cities) and Municipalities (the highest level of formally established cities); Mega city with a residing population of more than ten million and small city with a residing population of less than half million.

Figure 3-1 Conceptual Framework: Determinants of Migration Intention

• 4. Study area

Figure 4-1 Location of Jinzhou county level city and Beijing city

There are two study areas (Figure 4-1) in this research: Beijing city and Jinzhou (晋州) county level city (referred as Jinzhou city in the following content). Beijing is the capital city of China with a total residing population of about 21.71 million at the end of year 2015 (National Bureau of Statistics of China 2016b). Jinzhou city is a county level city in Shijiazhuang prefecture, Hebei province of China. It holds a total residing population of about 0.13 million in the urban area at the end of year 2015 (National Bureau of Statistics of China 2016b). Beijing and Jinzhou both belong to the “Beijing-Tianjin-Hebei” urban cluster. They are geographically close and culturally similar. Yet they are highly distinctive from each other in terms of economic development level, population size and administrative rank.

• 5. Data collection

Data used in this paper include primary and secondary data. Secondary data comes mostly from the “China City Statistical Yearbook” published in 2016 and the “China Urban Construction Statistical Yearbook” in the same year. Primary data collection was undertaken in Beijing city and Jinzhou city. The total sample size of rural migrants was 643, with 436 collected in Beijing and 207 collected in Jinzhou. All responders are rural migrants working or living in urban areas5. Participants are selected through simple random sampling. The average age of respondents in Beijing is 37, in Jinzhou is 30. About 31 percent of respondents in Jinzhou falls in the age group of 25 to 29. While about 1/4 of respondents in Beijing falls in the age group of 35 to 39. Male accounts for 61 percent of the total sample in Beijing and 45 percent of the total sample in Jinzhou. As shown in figure 5-1, females with an age between 25 and 29 were most populous among the sample in Jinzhou and males with an age between 35 and 39 were most populous among the sample in Beijing. About 84 percent of respondents in Beijing and about 78 percent of respondents in Jinzhou are married at the time of survey. More than 60 percent of rural migrants responded in Jinzhou hold a degree of junior high or lower. However, more than half of respondents in Beijing hold a degree of high school

• r higher (see from figure 5-2).

5 This paper adopted the definition of rural migrants from the sixth national population census of China in 2010. Rural

migrants refer to those rural labours who migrated to urban areas, but they still hold rural “hukou” status at the time

• f interview. Those rural labourers who went to urban areas for a business trip, study, travel, visit or medical reason

and are expected to return to their registered rural area, are excluded.

• 0.2
• 0.1

0.1 0.2 0.3 18~19 20~24 25~29 30~34 35~39 40~44 45~49 50~54 55~59 60~64 65~

Age-Gender Pyramid in Jinzhou

Male Female 0.20 0.10 0.0 0.10 0.20 18~19 20~24 25~29 30~34 35~39 40~44 45~49 50~54 55~59 60~64 65~

Age-Gender Pyramid in Beijing

Male Female

Figure 5-1 Age-Gender Pyramid of respondents Figure 5-2 Distribution of education level of respondents

• 6. Models and variables

6.1 Descriptive statistics Descriptive statistics were employed to address issues related to spatial inequality, the distribution of social inequalities against rural migrants and among rural migrants in the urban system of China. A further comparison between county level cities and municipalities, small cities and mega cities will be given. 6.2 Regression models According to the conceptual framework, the macro and miso policy, social and spatial inequality factors are projected to individual’s perception of the utility of destinations. This study first tries to identify how hukou, demographic and social-spatial inequality factors influence rural migrants’ standard of choosing an ideal destination. Multiple linear regression model is used in this stage. Second, this study also aims to find out people with what demographic characteristics and utility standards intend to choose county level cities over municipalities, small cities over mega cities. Multinominal Logistic Regression Model is used in this stage. MLogit models are frequently used in situations of multiple choice of dependent variables. Please see Greene (2008:843-845) for technical details.

0.00 0.09 0.54 0.23 0.09 0.03 0.00 0.00 0.08 0.37 0.36 0.11 0.07 0.01 0.1 0.2 0.3 0.4 0.5 0.6 not educated primary school junior high school high school or equivallent college bachelor' degree master's degree Percentage Education Level Beijing Jinzhou

1) Dependent variables Perceived utility standards are the dependent variables in the first stage (Multiple Linear Regression). How to measure the utility of a destination? Rural migrants might think quite differently than urban

• citizens. Different discipline applies different foci as well. By browsing existing place branding

literature (Zenker et al. 2013) and by interviewing 10 rural migrants respectively in Beijing and Jinzhou, this study finally complied a pool of 26 different items to measure the utility of a destination for rural migrants (shown in Table 6-2-1, from A to Z). By asking “how much will the following city attribute influence your choice of destination in five years” and measuring it with a scale from 0 (“no influence at all”) to 100 (“very important”), this study was able to collect rural migrants’ opinions about “how to measure the utility of a destination”. Migration intentions are the dependent variables in the second stage (Multinominal Logistic Regression). Migration intentions are measured through two variables: The administrative rank of rural migrants’ destination choice in five years and the population scale of rural migrants’ destination choice in five years (shown in Table 6-2-1). 2) Independent variables Demographic and social-spatial inequality variables are considered as independent variables in the first stage; Demographic variables and 26 variables of standards of ideal destination are included as independent variables in the second stage (see from table 6-2-1). Demographic variables include current residing city, age, gender, income6, education level and marriage status. Inequality variables include quality of social network, perceived income level in residing city, perceived income level among rural migrants in residing city, perceived social status in residing city and perceived social status among rural migrants in residing city. Those five inequality variables reflect both social and spatial inequality. The value is reflection of both where a rural migrant is currently residing and the disparities comparing to other rural migrants or urban citizens.

6 Income in the model refers to the logarithm of real income in 2015.

Table 6-2-1 Definitions and statistics of variables

Variables Definition Model Ⅰ Model Ⅱ N Mean SD Min Max Migration destination intention variables m2a The administrative rank of rural migrants’ destination choice in five years. 0 = rural village 1 = official towns 2 = county-level city 3 = prefecture-level city 4 = vice-provincial city 5 = Municipality (base outcome) √ 599 3.419 1.650 5 m2b The population scale of rural migrants’ destination choice in five years. 0 = rural village 1 = city with a population under half million 2 = city with a population of half million to one million 3 = city with a population of one million to five million 4 = city with a population of five million to ten million 5 = city with a population of more than ten million (base outcome) √ 595 3.118 1.894 5 Standard of Utility of destinations A The general level of wage √ ○ 470 78.657 25.137 100 B Good job opportunities ○ 469 77.959 24.516 100 C The general price level/ Costs of living √ ○ 468 73.603 26.382 100 D Housing price/ Cost of renting ○ 467 74.535 28.237 100 E Living conditions ○ 470 72.919 27.329 100 F

• ver-all layout and appearance

√ ○ 469 65.365 29.052 100 G A variety of shopping opportunities √ ○ 469 67.405 28.971 100 H Number and distribution of factories √ ○ 469 57.038 29.957 100 I A lot of nature and public green area ○ 470 66.043 28.588 100 J water and grid accessibility √ ○ 470 75.438 25.942 100 K A wide range of cultural and recreational facilities and activities √ ○ 470 67.566 27.328 100

L Availability and quality of medical service √ ○ 469 78.186 26.059 100 M Quality of education √ ○ 468 80.481 24.427 100 N Accessibility to social insurance √ ○ 467 76.122 27.484 100 O Good traffic condition √ ○ 471 74.096 27.430 100 P Environment quality (low pollution) √ ○ 469 73.855 29.826 100 Q Social networks √ ○ 468 65.662 28.800 100 R Distance to hometown √ ○ 469 66.612 28.895 100 S Openness and tolerance to migrants ○ 468 70.767 27.751 100 T The energy and atmosphere √ ○ 466 65.423 29.408 100 U Life pace √ ○ 466 65.785 27.991 100 V Benefit attached to hukou √ ○ 460 60.589 31.248 100 W Standard of getting local urban hukou √ ○ 461 60.026 34.355 100 X The position in the rural-urban system √ ○ 462 62.117 30.883 100 Y General economic status of the particular region √ ○ 463 69.600 27.896 100 Z Potential of future development ○ 461 71.900 28.349 100 Demographic variables Region 1 = Beijing; 0 = Jinzhou ○ ○ 643 0.678 0.468 1 Gender 1 = Female; 0 = Male ○ ○ 643 0.442 0.497 1 Age Age in the year of 2017 ○ ○ 643 35.081 8.893 18 67 Rincome Income of last year (“Income2015”): the exact income that rural migrants earn in year 2016, measured in “Yuan” (5 Yuan is roughly equal to 1 AUD). 585 46497.040 39352.250 0 300000 Income Log (income2015) ○ ○ 573 10.527 0.728 7.438384 12.61154 Education “University or above” is used as the comparing base. Noeduorpri 1 = Primary school or not educated; 0 = Otherwise ○ ○ 643 0.089 0.284 1

Juhigh 1 = Junior high school; 0 = Otherwise ○ ○ 643 0.417 0.493 1 Highs 1 = High school; 0 = Otherwise ○ ○ 643 0.313 0.464 1 College 1 = College; 0 = Otherwise ○ ○ 643 0.103 0.304 1 Marriage 1 = Currently married; 0 = Currently not married (including single, divorced and widowed) ○ ○ 643 0.807 0.395 1 Social-Spatial Inequality Variables NHposition Number of people in high position you know (company or factory owners; managers; government officials or public institution employees) ○ 428 10.619 70.543 1019 PincomeC Perceived income level in residing city: the relative income that rural migrants perceive comparing to all other people in the residing city, including migrants and urban citizens. The range of this variable is from 0 to 57. The bigger the number, the higher level of perceived income. ○ 631 1.780 1.158 5 PincomeM Perceived income level among other rural migrants in residing city: the relative income that rural migrants perceive comparing to all other rural migrants in the residing city. The range of this variable is from 0 to 5. The bigger the number, the higher level of perceived income. ○ 631 2.133 1.222 5 PstatusC Perceived social status in residing city: the perceived social status comparing to all other people in the residing city, including migrants and urban citizens. The range of this variable is from 0 to 5. The bigger the number, the higher level of perceived social status. ○ 636 1.936 1.269 5 PstatusM Perceived social status among other rural migrants in residing city: the perceived social status comparing to all other rural migrants in the residing city, including migrants and urban citizens. The range of this variable is from 0 to 5. The bigger the number, the higher level of perceived social status. ○ 629 2.396 1.242 5

Note: “○” refers to independent variables in significant models; “√” refers to dependent variables in significant models

7 This and the following six-scale variables are treated as continuous variables in the regressions in this article.

• 7. Findings

7.1 Spatial inequality in China’s hierarchical city system According to the “China City Statistical Yearbook” published in 2016, there are 656 formally established cities in China at the end of year 2015, including 4 municipalities, 15 vice-provincial cities, 276 prefecture-level cities and 361 county-level cities (National Bureau of Statistics of China, 2016a). Data in the “China Urban Construction Statistical Yearbook” shows that the administrative rank of a city is in correlation with the fund from all levels of government (see from table 7-1-1). The higher the rank, the more the fiscal budget(National Bureau of Statistics of China, 2016b). Table 7-1-1 National Fiscal Budget for Urban Maintenance and Construction by administrative rank

• f cities (2015)

Administrative rank Total (mean) National Level Budgetary Fund (mean) Provincial Level Budgetary Fund (mean) City-Level Budgetary Fund (mean) County-level city 55544.82 11937.84 4658.32 47259.48 Prefecture-level city 279280.57 13076.12 13315.60 253540.66 Vice-provincial city 2119890.13 24732.29 22744.67 1791127.87 Municipality 3749553.00 28321.00 951651.00 2946487.00 Note: Measurement Unit: 10,000 RMB Source: Calculated from data in “China Urban Construction Statistical Yearbook” 2016

In average, cities with higher ranks also possess higher Gross Regional Production value (see from table 7-1-2). Table 7-1-2 Gross Regional Production by administrative rank of cities (2015)

Administrative rank GRP (mean) County-level city 3595090.79 Prefecture-level city 8731007.36 Vice-provincial city 75518688.80 Municipality 193993525.00

Note: Measurement Unit: 10,000 RMB Source: Calculated from data in “China Urban Construction Statistical Yearbook” 2016

Besides the fiscal budget, the administrative rank of a city is also an indicator for construction land area of various facilities (see from table 7-1-3). On the average, high ranked cities are geographically “bigger” and obtain more land for many facilities. Table 7-1-3 Construction Land by Administrative Rank of Cities (2015)

Administrative rank Area of city (mean) Area of land for road traffic facilities (mean) Area of land for public facilities (mean) Area for green land and Square (mean) County-level city 122.89 3.48 1.23 3.12 Prefecture-level city 352.13 14.85 3.81 11.73 Vice-provincial city 1499.95 72.10 12.64 49.61 Municipality 7015.04 258.64 54.20 125.47

Note: Area Measurement Unit: Square Kilometre; Population Measurement Unit: 10,000persons Source: Calculated from data in “China Urban Construction Statistical Yearbook” 2016

Except for geographical implications, the administrative rank also echoes with residing population size of a city. Higher rank usually implies a bigger population of migrants as well as a bigger population with legal hukou in that city (see from table 7-1-4). Table 7-1-4 Urban Population by Administrative Rank of Cities (2015)

Administrative rank Population with local urban hukou in city area (mean) Population without local urban hukou in city area (mean) County-level city 19.42 2.39 Prefecture-level city 72.60 13.90 Vice-provincial city 433.09 104.71 Municipality 1500.61 251.86

Note: Population Measurement Unit: 10,000persons Source: Calculated from data in “China Urban Construction Statistical Yearbook” 2016

According to the “Notice of the State Council on Adjusting the Standards for Categorising City Sizes” issued by the Chinese State Council (2014), cities in China can be categorized into 5 scales and 7 levels in terms of resident population in the city. See from table 7-1-5. Table 7-1-5 Scales of cities in China

City levels Mega City Large City Big City Medium City Small City Ⅰ Ⅱ Ⅰ Ⅱ Resident Population Size >10 million 5~10 million 3~5 million 1~3 million 0.5~01 million 0.2~0.5 million <0.2 million

Source: the Chinese State Council (2014)

Even though the administrative rank of a city is in correlation with lots of indicators of the quality of a city, not all indicators rise as the rank rises. Table 7-1-6 shows that sometimes lower ranked cities possess better amenities than high ranked ones. For example, the mean of “Road Surface Area Per Capita” in low ranked cities is higher than that in high ranked cities. Table 7-1-6 Level of National Urban Service Facilities by Administrative Rank of Cities (2015)

Administrative rank Population Density (person/ square kilometre) (mean) Water Coverage Rate (%) (mean) Gas Coverage Rate (%) (mean) Road Surface Area Per Capita (m2) (mean) Wastewater Treatment Rate (%) (mean) Public Recreationa l Green Space Per Capita (m2) (mean) County-level city 3187.53 95.35 88.11 18.19 87.62 12.41 Prefecture-level city 3607.31 97.11 92.11 17.31 89.54 13.66 Vice-provincial city 4979.60 99.92 99.41 16.41 94.98 13.44 Municipality 2686.50 99.22 98.84 9.99 91.90 12.69

Source: Calculated from data in “China Urban Construction Statistical Yearbook” 2016

In conclusion, the administrative rank is a meaningful standard to classify cities. It echoes with the population size, the fiscal budget allocation, and the Gross Regional Production value of a city. Higher ranked cities are usually better, if only the economic indicators are considered. But if some amenities and the preference of a certain group are considered, this might not be the case.

7.2 Hierarchical distribution of social inequality against rural migrants in the urban system of China In the survey, there are two questions asked to collect rural migrants’ perception of the spatial distribution of social inequality against them. About 74 percent of responders think the benefit attached to a city’s urban hukou rises as its administrative rank rises (see from Figure 7-2-1). Figure 7-2-1 Relation between Administrative Rank of a City and the Benefit Attached to Hukou Besides, about 59 percent of responders think there are more differentiations against rural migrants in high ranked cities (see from Figure 7-2-2). Figure 7-2-2 Relation between Administrative Rank of a City and the Differentiation against Migrants

0.74 0.08 0.18 The higher the rank the higher level of benifit The higher the rank the lower level of benifit the rank is not related to the benefit 0.59 0.15 0.26 The higher the rank the more differentiations The higher the rank the less differentiations The administrative rank is not related to the differentiations

7.3 Comparison of Social-spatial inequality directed to rural migrants between a county level city and a municipality As mentioned before, as a rural migrant, living or working in municipalities can feel quite different from living or working in county level cities. First, as shown in graph 7-3-1, about 71 percent of respondents in Beijing think the differentiation level against rural migrants is relatively high, very high or the highest, however about 60 percent of respondents in Jinzhou think the differentiation level against rural migrants is relatively low, very low or the lowest. Second, even though the actual income

• f rural migrants in Beijing (with a mean of ￥53866 for year 2016) is generally higher than that in

Jinzhou (with a mean of ￥30310 for year 2016), 17 percent more rural migrants in Beijing, the municipality, feel their income is very low or lowest in that city, comparing to those in Jinzhou, the county level city (see from Graph 7-3-2). Even about 36 percent of respondents in Jinzhou think their income is relatively high in that city. Figure 7-3-1 Perceived differentiation level in current city Figure 7-3-2 Rural Migrants' Perception of Income in Current City

0.26 0.12 0.22 0.31 0.06 0.03 0.10 0.08 0.13 0.26 0.16 0.29 0.05 0.1 0.15 0.2 0.25 0.3 0.35 Lowest Very low Relatively low Relatively high Very high Highest Percentage Differentiation Level Jinzhou Beijing 0.14 0.14 0.33 0.36 0.01 0.00 0.18 0.27 0.26 0.25 0.01 0.02 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Lowest Very low Relatively low Relatively high Very high Highest Percentage Perception of own income in current city Jinzhou Beijing

About 47 percent of rural migrants in Jinzhou think their social status in that city is relatively high, very high or the highest, while correspondent proportion in Beijing is only 38 percent (see form Figure 7-3-3). Figure 7-3-3 Rural Migrants' Perceived Social Status in Current City The perceived crucial elements (choosing between ability and social network) to success of rural migrants in two cities are quite different as well. The largest data set (about 38 percent) of respondents in Jinzhou falls in “social network is more important than ability”, while about 41 percent of rural migrants responded in Beijing, the largest data set, think social network and ability are equivalent in pursuit of success (see from graph 7-3-4). Figure 7-3-4 Crucial Element of Success in Current city The perceived life pace is also quite different between rural migrants in Jinzhou and in Beijing. About 37 percent of respondents in Beijing think the life pace there is fastest, while only 7 percent of respondents in Jinzhou fall in the corresponding group (see from Figure 7-3-5).

0.12 0.13 0.27 0.40 0.06 0.01 0.22 0.21 0.21 0.31 0.05 0.02 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 Lowest Very low Relatively low Relatively high Very high Highest

Percentage Perceived Social Status

Jinzhou Beijing 0.12 0.38 0.16 0.20 0.13 0.11 0.15 0.41 0.14 0.19 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 Social network is most important Social network is more important than ability Social network and ability are equivalent Ability is more important than social network Ability is most important Percentage Element Comparison Beijing Jinzhou

Figure 7-3-5 Life Pace of Current City In conclusion, rural migrants in Jinzhou, as a group, are not experiencing same level of inequality because of their hukou status as those rural migrants in Beijing: Generally speaking, rural migrants in Jinzhou feel their income, social status are higher than those in Beijing. The living style are also quite different between two types of cities: the county level city Jinzhou is characterized with a slower life pace and rather higher appreciation of social networks comparing to the municipality Beijing. Studies addressing social inequalities between rural migrants and urban citizens are the mainstream, and the inequalities between those two groups are for sure the most significant. But the differentiations among rural migrants also exist, like shown in the data. Social inequality indicators such as income, social status, education level etc. are varied among rural migrants, just as they are varied among urban citizens. Only is it reasonable to take into account the social inequality between rural migrants and urban citizens, among rural migrants and spatial inequalities when explaining rural migrants’ future choice of destination. 7.4 Regression results: inequalities and destination preference Multiple linear regression model is used to predict how demographic characteristics and social-spatial inequalities influence the perceived importance of 26 standards of utility of destinations. Among all 26 models, 20 are significant. Table 7-4-1 presents the result of multiple linear regression. All dimensions of inequality and demographic variations play a significant role in influencing the perceived importance of 20 standards when choosing a future destination. In items of demographic variables, the evidence supported our expectation that people with different demographic characteristics value different things of destinations: Those rural migrants who currently reside in Beijing are more likely to pay less attention to “general level of age”, “environment quality”, “distance to hometown”, “benefit attached to hukou” and “standard of getting local urban hukou” than rural migrants who currently reside in Jinzhou; Older migrants are less likely to consider “water and grid accessibility” but more likely to consider “the position of chosen destination in China’s rural- urban system” comparing to younger ones; Those who only hold a degree from Junior high or high school are much more likely to consider “the general price level/cost of living” comparing to those who holds a bachelor or higher degree; Married respondents are less likely to care about the “the position of chosen destination in China’s rural-urban system”.

0.05 0.04 0.26 0.48 0.10 0.07 0.02 0.02 0.09 0.28 0.23 0.37 0.1 0.2 0.3 0.4 0.5 0.6 Slowest Very Slow Relatively Slow Relatively Fast Very Fast Fatest Percentage Life Pace Jinzhou Beijing

The results for social-spatial inequality indicate that, rural migrants in different social status possess different standards of an ideal migration destination: Those who own a better social network (Measured by the number of company or factory owners, managers, government officials or public institution employees known) are less likely to worry about “general level of age”, “the general price level/cost of living” and “distance to hometown”; Rural migrants with higher perceived income in residing city are more likely to pay attention to “the number and distribution of factories”; On the contrary, rural migrants with higher perceived income among rural migrants in residing city are less likely to pay attention to “the number and distribution of factories”; Respondents with higher perceived social status in residing city are less likely to care about “environment quality”; Respondents with higher perceived social status among rural migrants in residing city are more likely to value “over-all layout and appearance of the destination”, “a variety of shopping opportunities”, “cultural and recreational facilities and activities”, “availability and quality of medical service”, “accessibility to social insurance”, “traffic condition”, “environment quality”, “social networks”, “the energy and atmosphere”, “life pace” and “benefit attached to hukou”.

Table 7-4-1 Linear Regression Result

A C F G H J K L M N Demographic Variables Beijing

• 7.874*

4.866 2.017

• 4.814
• 7.2
• 6.241
• 0.557

0.083

• 3.379
• 4.747

Female 3.602 0.55

• 2.811

0.166 0.591 1.781 1.999 4.874 1.399 1.612 Age

• 0.024
• 0.015
• 0.009

0.193

• 0.07
• 0.428*
• 0.11
• 0.053

0.001 0.114 Income

• 1.175
• 1.675
• 1.321

0.954 0.876 1.788 4.119

• 0.468
• 2.108
• 4.19

Noeduorpri

• 11.93

5.613 8.933

• 4.737
• 8.956
• 12.745
• 4.354
• 5.919
• 12.367
• 8.24

Juhigh

• 7.721

13.136* 8.38 5.46 7.982 3.855

• 0.542

5.734

• 1.725
• 4.293

Highs

• 1.076

12.933* 10.747 2.205 2.101 1.5 2.418 7.964 3.701

• 0.995

College 2.464 11.591 11.166 2.975 4.191 6.187 3.228 10.196 6.079 2.697 Married

• 0.995
• 0.277
• 1.182
• 6.662
• 0.377

2.673

• 0.767

5.261 6.392 0.846 Social-Spatial Inequality Variables NHposition

• 0.063***
• 0.046*
• 0.006
• 0.009
• 0.035
• 0.015
• 0.018
• 0.02
• 0.018
• 0.019

PincomeC

• 2.03

2.08 2.005 0.351 6.790** 2.275 1.3 0.501 2.934 1.951 PincomeM 3.545

• 2.779
• 2.55
• 1.16
• 4.991*

0.053

• 0.636

1.914 0.363 0.621 PstatusC 1.688 1.794 2.163 2.897 2.114 1.956 0.752 0.441

• 0.418
• 0.284

PstatusM 0.548 1.055 5.007** 3.794* 0.656 1.837 4.218* 4.692** 2.092 4.692** _cons 90.958*** 73.803** 56.374* 41.858 44.918 56.224* 14.363 54.041* 87.197*** 104.009*** R2 0.107 0.071 0.073 0.076 0.109 0.123 0.074 0.102 0.077 0.085 Case numbers 312 311 311 312 312 312 312 312 311 310 P 0.002*** 0.077* 0.060* 0.048** 0.001*** 0.000*** 0.057* 0.003*** 0.045** 0.021**

Note: * p<.10; **p<.05; ***p<.01

Table 4-4-2 continued

O P Q R T U V W X Y Demographic Variables Beijing

• 6.895
• 10.212*
• 5.054
• 9.858*
• 0.154

0.769

• 10.143*
• 18.979***
• 4.77

3.328 Female 2.646 6.405 4.184 5.519 4.775 0.439 3.646

• 3.311

2.208 1.452 Age 0.183

• 0.038
• 0.204
• 0.216

0.159 0.26 0.194 0.279 0.566* 0.295 Income

• 4.761*
• 1.787

1.62 2.05

• 2.457
• 0.965

1.804 3.616 0.591 0.265 Noeduorpri

• 8.699
• 12.915
• 6.528
• 12.773

7.138 7.424 7.73

• 5.23

9.858 7.975 Juhigh

• 3.19

0.857

• 1.216
• 5.046

8.445 9.645 8.724 3.169 2.166 8.726 Highs 3.734 5.442 0.861

• 0.624

8.286 9.891 10.185 6.325 8.207 8.26 College 5.525 3.643 9.501 0.014 10.728 9.353 14.324 12.289 6.59 8.304 Married

• 6.126
• 3.171
• 3.153

3.063

• 2.963
• 5.47
• 5.914
• 4.148
• 11.039*
• 8.629

Social-Spatial Inequality Variables NHposition

• 0.028
• 0.03
• 0.014
• 0.053**
• 0.002
• 0.004
• 0.04
• 0.037

0.004

• 0.006

PincomeC 2.931

• 0.524

1.444

• 0.899

4.318 1.882 1.072 1.491 2.07 3.895 PincomeM

• 2.179

2.117 1.131 2.065

• 1.218
• 0.719

0.298

• 1.016
• 2.613
• 3.446

PstatusC

• 2.555
• 4.500*

1.446 0.677 0.305 2.21 0.483 2.777 4.574* 4.123* PstatusM 6.630*** 4.376* 4.504* 2.382 3.972* 3.776* 4.007* 0.966 0.758 1.811 _cons 112.423** * 91.860** 40.254 46.643 62.224* 46.187 21.984 17.44 30.893 39.778 R2 0.112 0.084 0.136 0.09 0.068 0.068 0.082 0.098 0.073 0.077 Case numbers 313 313 311 312 312 311 309 310 310 311 P 0.001*** 0.022** 0.00*** 0.011** 0.096* 0.092* 0.029** 0.005*** 0.061* 0.044**

Note: * p<.10; **p<.05; ***p<.01

7.5 Regression results: destination preference and migration intention As discussed before, the administrative rank is explanatory to inequalities among cities and it is in correlation with the population size of a city. Administrative rank is a powerful standard to classify cities but might be recondite to rural migrants. So, this paper uses both administrative ranks and population sizes to classify destinations. Using two migration destination intention variables as dependent variables, demographic factors and 26 standards of utility of destination variables as independent variables, by applying multi-nominal logit regression, we get the results presented in table 7-5-1. A rural migrant who values income significantly could choose a total different destination than another who values environment the most. This regression aims to link rural migrants’ personal utility standards for migration destination to tiered destinations (see from Table 7-5-1). The table below only shows the result of county level cities comparing to municipalities and of small cities (with a population below half million) and first class mega cities (with a population more than 10 million). Both demographic features and utility standards are quite explanatory to rural migrants’ future destination choice in five years. When it comes to the choice between county level cities and municipalities, female comparing to male, migrants residing in Jinzhou comparing to those residing in Beijing, married comparing to un- married, are more likely to choose county level cities over municipalities. Rural migrants who care more about “living conditions” and “distance to hometown” are more likely to choose county level cities over municipalities. On the contrary, rural migrants who pay more attention to “the general level of wage” are more likely to choose municipalities instead of county level cities. When it comes to the choice between small cities and mega cities, rural migrants residing in Jinzhou comparing to those residing in Beijing, married comparing to un-married, are more likely to choose small cities over mega cities. Migrants who care more about “quality of education” and “distance to hometown” are more likely to choose small cities instead of mega cities as their future destination. While rural migrants who pay more attention to “the general level of wage”, “cultural and recreational facilities and activities”, “availability and quality of medical service” and “standard of getting local urban hukou” are more likely to choose mega cities over small cities. Table 7-5-1 Multinomial Logistic Regression Result

m4a m4b Number of ~s 401 398 Prob>chi2 0.000 0.000 Psedo R2 0.402 0.386 Baseoutcome Municipality Mega city Demographic Variables Region 0.008*** 0.009*** Gender 2.489* 2.151 Age 0.969 0.950 Income 1.335 1.636

Noeduorpri 0.555 0.493 Juhigh 1.513 0.698 Highs 0.806 0.497 College 0.397 0.280 Married 4.513* 4.367* Standards of Utility of Destinations A 0.967* 0.960** B 1.023 1.025 C 0.992 1.002 D 0.990 0.981 E 1.024* 1.018 F 0.993 0.996 G 1.009 0.999 H 1.008 1.014 I 0.993 0.986 J 0.987 0.989 K 0.979 0.978* L 0.977 0.966* M 1.010 1.035* N 1.023 1.011 O 1.008 1.014 P 1.005 1.014 Q 1.003 1.000 R 1.023* 1.025** S 0.992 0.994 T 1.004 1.010 U 1.002 0.998 V 1.009 1.017 W 0.977 0.969** X 1.006 1.002 Y 1.016 1.015 Z 0.972 0.978 _cons 0.282 0.102

Note: * p<.10; **p<.05; ***p<.01; Number attached to variable is RRR

• 8. Conclusion

The current paper presents an empirical tested conceptual framework to link demographic characteristics, social-spatial inequality factors and hierarchical hukou policies to rural migrants’ migration intentions (destination choices) through rural migrants’ “destination utility standards”. Unlike other studies which sort migration intention into “stay or “leave”, or into “stay in this city, go to other cities, go back to hometown”, this study classifies migration intentions according to the administrative rank and the population size of the intended destination. Main conclusions can be drawn from this study are as follow: First, social inequalities against rural migrants are not distributed evenly among all cities in China. Evidence shows there is a possibility that more differentiations against rural migrants exist in cities with a higher administrative rank. Second, the study provides evidence that both demographic characteristics and social-spatial inequality status have a statistically and quantitatively significant association with rural migrants’ “destination utility standards”. The coarse-gained statistics used to describe the migration intentions

• f China’s rural migrants as they are all identical in social status. But in fact, there are differentiations

in the rural migrants group as well. As revealed in the results of linear regression, the demographic and social inequality factors can result in different emphasis when choosing a destination. Rural migrants currently residing in Beijing exhibit a lower probability of considering about general wage, environment quality, distance to hometown, benefit attached to hukou and standard of getting local urban hukou. Older migrants exhibit a higher probability to pay less attention to water and grid accessibility but to pay more attention to the position of destination in the rural-urban system. Migrants with higher income are more likely to lower their requirements for traffic conditions. Those with qualifications from junior high or high school are more likely to stress the cost of living. Married

• nes are likely to ease their demand for the position of destination in the rural-urban system. Those

migrants with better social network exhibit lower expectation in general wage, less consideration about the cost of living and distance to hometown. Those who perceive their income higher in residing city or lower among rural migrants in residing city are more likely to stress the number and distribution of factories. Our research provides new evidence which suggests that rural migrants with higher perceived social status in residing city exhibit a higher probability to stress the position of destination in the rural-urban system and the general economic status of destination while neglect the quality of environment; rural migrants with higher perceived social status among rural migrants exhibits a higher probability of paying more attention to over-all layout of destination, shopping

• pportunities, cultural and recreational facilities and activities, availability and quality of medical

service, accessibility to social insurance and traffic conditions, environment quality, social networks, the energy and atmosphere, life pace, and benefit attached to hukou. Third, the variation of demographic and social-spatial inequality factors among rural migrants influences their emphasis of standards of an ideal destination significantly. Those varied emphasis of standards of destination utility further direct them to different destinations. Those rural migrants who are female, married or with higher requirement of living conditions or distance to hometown and lower demand for wage are more likely to choose county level cities instead of municipalities. Rural migrants who are married, who stresses the quality of education or distance to hometown or take less notice of general wage, cultural and recreational facilities and activities, availability and quality of medical service or standard of getting local urban hukou are more likely to choose small cities over mega cities.

Fourth, evidence show that the current hukou system are failing to direct rural migrants away from mega cities to small cities. Since the launch of the “New style Urbanization plan” in 2014, China has built a hierarchical hukou gaining system, with the easiest standard in small cities and the most stringent standard in Mega cities, in the hope of directing more rural migrants to small cities. But the evidence in this study shows otherwise. The influence of hukou policies are measured through rural migrants’ perception of the benefit attached to hukou and the standard of getting local urban hukou. Rural migrants in Jinzhou or rural migrants with higher perceived social status among rural migrants in residing city are more liking to care about benefit attached to hukou. Rural migrants in Jinzhou, comparing to those in Beijing, are more likely to stress the standard of getting local urban hukou. Those who pay more attention to the standard of getting local urban hukou are more likely to choose mega cities instead of small cities. A possible explanation is, to rural migrants, the standard of hukou gaining serves as an official label of how popular and developed a city is, if it was only individual perception before. Instead of severing as an efficient order to compel rural migrants away from mega cities, the hierarchical hukou gaining standard system might have served as perfect advertisement to attract more rural migrants to mega cities.

References:

Afridi, F, Li, SX & Ren, Y 2015, 'Social identity and inequality: The impact of China's hukou system', Journal

• f Public Economics, vol. 123, pp. 17-29.

Agadjanian, V, Nedoluzhko, L & Kumskov, G 2008, 'Eager to leave? Intentions to migrate abroad among young people in Kyrgyzstan', International Migration Review, vol. 42, no. 3, pp. 620-651. Black, R, Natali, C & Skinner, J 2005, Migration and inequality, World Bank Washington, DC. Böheim, R & Taylor, MP 2002, 'Tied down or room to move? Investigating the relationships between housing tenure, employment status and residential mobility in Britain', Scottish Journal of Political Economy, vol. 49, no. 4, pp. 369-392. Cai, H & Wang, J 2008, 'Factors influencing the migration intentions of rural workers in the Pearl River Delta', Social Sciences in China, vol. 29, no. 3, pp. 157-171. Cao, G, Li, M, Ma, Y & Tao, R 2015, 'Self-employment and intention of permanent urban settlement: Evidence from a survey of migrants in China's four major urbanising areas', Urban Studies, vol. 52, no. 4, pp. 639-664. Carrillo, B 2011, Small town China: rural labour and social inclusion, Taylor & Francis. Castles, Sa 2014, The age of migration : international population movements in the modern world, International population movements in the modern world, Fifth edition / Stephen Castles, Hein de Haas and Mark J. Miller. edn, eds Hda Haas & MJa Miller, Guilford Press. Chan, KW 2010, 'Fundamentals of China's Urbanization and Policy', China Review, vol. 10, no. 1, pp. 63-93. China, NBoSo 2016a, CHINA CITY STATISTICAL YEARBOOK. China, NBoSo 2016b, China Urban Construction Statistical Yearbook. De Jong, GF, Richter, K & Isarabhakdi, P 1996, 'Gender, values, and intentions to move in rural Thailand', International Migration Review, pp. 748-770. De Jong, GF, Root, BD, Gardner, RW, Fawcett, JT & Abad, RG 1985, 'Migration intentions and behavior: Decision making in a rural Philippine province', Population & Environment, vol. 8, no. 1, pp. 41-62. Du, H & Li, S-M 2012, 'Is It Really Just a Rational Choice? The Contribution of Emotional Attachment to Temporary Migrants' Intention to Stay in the Host City in Guangzhou', China Review, vol. 12, no. 1, pp. 73- 93. Faist, T & Faist, T 2000, The volume and dynamics of international migration and transnational social spaces.

Fan, CC 2011, 'Settlement Intention and Split Households: Findings from a Survey of Migrants in Beijing's Urban Villages', China Review, vol. 11, no. 2, pp. 11-41. Friedman, M & Savage, LJ 1952, 'The expected-utility hypothesis and the measurability of utility', Journal of Political Economy, vol. 60, no. 6, pp. 463-474. Gordon, I & Molho, I 1995, 'Duration dependence in migration behaviour: cumulative inertia versus stochastic change', Environment and Planning A, vol. 27, no. 12, pp. 1961-1975. Greene, W 2008, Econometric Analysis, Pearson Education Inc, New York. Hao, P, Tang, S, Muellerleile, C & Akers, J 2015, 'Floating or settling down: the effect of rural landholdings

• n the settlement intention of rural migrants in urban China', vol. 47, no. 9, pp. 1979-1999.

Li, Z, LeGates, R & Zhao, M 2016, Understanding China's Urbanization The Great Demographic, Spatial, Economic, and Social Transformation, eds R LeGates & M Zhao, Edward Elgar Publishing, Cheltenham, Gloucestershire. Lin, D, Li, X, Wang, B, Hong, Y, Fang, X, Qin, X & Stanton, B 2011, 'Discrimination, Perceived Social Inequity, and Mental Health Among Rural-to-Urban Migrants in China', Community Mental Health Journal, vol. 47,

• no. 2, pp. 171-180.

Lipton, M 1980, 'Migration from rural areas of poor countries: the impact on rural productivity and income distribution', World Development, vol. 8, no. 1, pp. 1-24. Massey, DS 1999, 'Why does immigration occur? a theoretical synthesis', in PK C Hirschman, J DeWind (ed.), The Handbook of International Migration: The American Experience, Russell Sage Found, New York, pp. 34- 52. Meng, X & Manning, C 2010, 'The great migration in China and Indonesia: trends and institutions', in X Meng (ed.), The great migration : rural-urban migration in China and Indonesia, Edward Elgar, Cheltenham. PRC-SCC 2014a, 'Instructions on furthering the hukou system reform (Guanyu Jinyibu Tuijin Huji Zhidu Gaige de Yijian)'. PRC-SCC 2014b, the “new style urbanization” plan (2014-2020), viewed 05, 25 2016, <http://politics.people.com.cn/n/2014/0317/c1001-24649809.html>. Ravenstein, EG 1885, 'The laws of migration', Journal of the Statistical Society of London, vol. 48, no. 2, pp. 167-235. Simon, HA 1980, 'The behavioral and social sciences', Science, vol. 209, no. 4452, pp. 72-78.

Tang, S & Feng, J 2015, 'Cohort differences in the urban settlement intentions of rural migrants: A case study in Jiangsu Province, China', Habitat International, vol. 49, pp. 357-365. the Chinese State Council 2014, Notice of the State Council on Adjusting the Standards for Categorizing City Sizes, viewed 0620 2016, <http://www.gov.cn/xinwen/2014-11/20/content_2781156.htm>. White, MJ 2015, International Handbook of Migration and Population Distribution, Dordrecht: Springer Netherlands, Dordrecht. Yue, Z, Li, S, Feldman, MW & Du, H 2010, 'Floating Choices: A Generational Perspective on Intentions of Rural–Urban Migrants in China', vol. 42, no. 3, pp. 545-562. Zang, B, Lv, P & Warren, CMJ 2015, 'Housing prices, rural-urban migrants' settlement decisions and their regional differences in China', Habitat International, vol. 50, p. 149. Zenker, S, Petersen, S & Aholt, A 2013, 'The Citizen Satisfaction Index (CSI): Evidence for a four basic factor model in a German sample', Cities, vol. 31, pp. 156-164. Zhu, Y 2003, 'The floating population's household strategies and the role of migration in China's regional development and integration', International Journal of Population Geography, vol. 9, no. 6, pp. 485-502. Zi-cheng, W & Wei-guo, Y 2013, 'Self-employment or wage-employment? On the occupational choice of return migration in rural China', China Agricultural Economic Review, vol. 5, no. 2, pp. 231-247.