Emerging divided cities in China: Socioeconomic segregation in Shanghai,2000

Abstract

Chinese

Compared with North America and Western Europe, Chinese cities used to feature a low extent of socioeconomic segregation. However, systematic analysis of the changes in socioeconomic segregation after the end of the provision of welfare housing is needed. Using residential-committee-level data from the fifth and sixth censuses of Shanghai, for the first time, this article systematically charts changes in socioeconomic segregation in Chinese cities over the period 2000–2010. Along with the emergence of high-status neighbourhoods and migrant neighbourhoods, Shanghai has grown more divided based on individual socioeconomic status. The extent of socioeconomic segregation in Shanghai was comparable to that of large US and European cities. While patterns of sociospatial divisions are different across central and suburban areas, the level of educational segregation becomes greater than that of hukou segregation. The crucial role of housing commodification in driving these changes highlights the importance of contextual and institutional factors in understanding the dynamics of segregation.

Keywords

socioeconomic segregation housing commodification Chinese cities Shanghai

Introduction

Socioeconomic segregation, that is, the residential sorting of socioeconomic groups in cities, has long been an important issue in urban studies (Nightingale, 2012). In recent years, however, there has been a renewed debate on the topic. While the classic ecological approach explains residential segregation as an outcome of the competition for space among socioeconomic groups, increasing attention is being paid to the role of contextual institutional factors (Maloutas and Fujita, 2012; Van Kempen and Murie, 2009). Against this backdrop, cities experiencing major changes in economic and political systems, such as former socialist cities, are important for appreciating the universal and particular forces in shaping segregation (Marcińczak et al., 2015; Tammaru et al., 2016a).

Residential segregation in urban China is essentially a post-reform phenomenon (Gu and Shen, 2003). Empirical studies, however, have shown that the extent of residential segregation based on socioeconomic status (SES) had remained relatively lower than that of western cities by the end of the 20th century (Li and Wu, 2008). While sociospatial differentiation is unfolding in multiple dimensions, the role of institutional legacies remains influential. For example, residential segregation has remained largely tenure-based, where work units and public rental housing are still important in shaping the social space of cities (Li and Wu, 2008; Madrazo and Van Kempen, 2012), while the confluence of state and market forces has led to a complex combination of segregation and mixing (Wu et al., 2017).

These studies have usefully revealed the hybrid dynamics of segregation in Chinese cities. Nonetheless, along with the abolishment of the welfare housing system in 1998, the new market-based residential sorting mechanism may reinforce the socioeconomic segregation in Chinese cities. As Huang and Clark (2002) stated, the housing reform has not only released the pent-up demand for housing, but has also allowed people to make residential decisions based on their preferences and affordability. The force of the market in housing differentiation is increasingly apparent (He et al., 2017; Yi and Huang, 2014). A number of studies have examined emerging sociospatial divisions and the underlying processes, such as gentrification and jiaoyufication (He, 2010; Wu et al., 2018), suburbanisation (Shen and Wu, 2013), the proliferation of gated communities (Pow, 2009; Zhang, 2012), high-poverty neighbourhoods (Wu et al., 2010) and migrant enclaves (Wang et al., 2010), as well as the potential effects of increasing segregation on social inequality (Xiao et al., 2017a, 2017b, 2019). Therefore, systematic analysis of the changes in socioeconomic segregation after the end of the provision of welfare housing is needed.

Against this backdrop, this article examines the patterns, levels and dynamics of socioeconomic segregation in Shanghai. It contributes to the research literature in four ways. First, while recognising sociospatial changes in Chinese cities as a result of a combination of different factors, the study highlights the increasing importance of market mechanisms in residential sorting. Second, previous studies on citywide patterns of segregation at the neighbourhood level were based on a single-year, cross-sectional analysis of the census data in 2000 (Li and Wu, 2008; Wu et al., 2014). However, drawing on a block-level analysis of travel survey data in Beijing, one recent study has indicated dramatic neighbourhood changes during the 2000s (Liu et al., 2019). This study draws on residential-committee-level data from both the 2000 and the 2010 census, and provides an updated, longitudinal empirical study on residential segregation in China. Third, it is among the first attempts to identify the relative importance of human capital and hukou status in shaping contemporary residential patterns. Finally, by examining the changing patterns of socioeconomic segregation after housing reform in China, this study also engages with current debate on the impacts of institutional changes in global segregation studies.

The article is organised as follows. The next section reviews the literature of socioeconomic segregation and highlights the need to explain segregation patterns based on contextual institutional contexts. Then we develop the analytical framework on the changing dynamics of segregation in China after housing reform and discuss the hypotheses derived from the framework. The following two sections present the methods and results of the empirical study based on neighbourhood-level census data in Shanghai. The article concludes with the main findings and broader implications for global segregation studies.

Socioeconomic segregation: institutional and contextual perspectives

Studies on residential segregation can be traced back to the classic theory of urban ecology. Noting the residential sorting and re-sorting of people into different neighbourhoods, Robert Park and his colleagues developed a model of urban residential structure based on biological analogies, that is, the ecological model (Burgess, 1928; Park, 1915). They recognised the causal relationship between social distance and physical distance as the basis for residential segregation. Moreover, the model suggested that sociospatial differentiation resulted from economic competition between groups of people for space. Empirical evidence supporting this notion was later found in both North America and Western Europe (Duncan and Duncan, 1955; Fischer et al., 2004; Musterd, 2005).

While the sources of sociospatial segregation lie in a variety of interrelated dimensions, social position, which reflects individuals’ opportunities in the housing market, is one of the determinants of patterns of residential segregation. Encapsulated in the concept of social class, the factors of income, education and occupation are usually used as indicators of a person’s social status (Farley, 1977; Ljunggren and Andersen, 2015; Reardon and Bischoff, 2011). Recently, along with radical economic restructuring under neoliberalisation and globalisation, the role of household type and lifestyle in shaping urban sociospatial differentiation has become increasingly apparent as well (Fischer et al., 2004; Pattaroni et al., 2012).

The ecological model is premised upon assumptions of free economic competition for space. However, it has been demonstrated that the relationship between social inequality and socioeconomic segregation is not straightforward. Instead, market processes are always altered by diverse contextual institutions (Burgers and Musterd, 2002; Maloutas and Fujita, 2012; Van Kempen and Murie, 2009). Among others, welfare regimes and governance traditions have a great influence on socioeconomic segregation. In many developed countries in Western Europe and North America, recent globalisation and deindustrialisation have led to an increase in social inequality through social polarisation or professionalisation (Hamnett, 1996; Préteceille, 2000; Sassen, 1991). However, due to differences between welfare regimes, the extent of socioeconomic segregation in Western Europe has been relatively lower than that of the US (Musterd, 2005; Musterd and Ostendorf, 1998). The welfare state is a critical factor in explaining such differences. While the liberal welfare regime with minimal state intervention in the US usually leads to greater social inequalities and more unequal spatial outcomes, stronger state interventions in wealth redistribution in Europe have greatly moderated the impacts of economic restructuring (Arbaci, 2007; Van Kempen and Murie, 2009).

One of the significant welfare state arrangements that is closely associated with socioeconomic segregation is the housing system. Generally speaking, a more market-oriented approach to housing provision leads to higher levels of segregation (Kemeny, 1995). In the US, income-based inequality has led to high levels of income segregation (Reardon and Bischoff, 2011). Similarly, many European cities have witnessed an increase in socioeconomic segregation along with the recent weakening of the welfare state and liberalisation of the housing market (Boterman and Van Gent, 2014; Tammaru et al., 2016a). Moreover, housing policies also determine the locations of different types of housing in cities. In many European cities, for example, the concentration of public housing can generate a concentration of poverty (Lee and Murie, 1997; Van Ham and Manley, 2009).

Notably, from the institutional perspective discussed above, it is also important to go beyond the broad welfare regime types or housing systems and to identify unique local institutional arrangements and policies (Van Kempen and Murie, 2009). Research on socioeconomic segregation in post-socialist cities is particularly pertinent here. In the socialist era, the state directly controlled the production and redistribution of goods (including housing) to achieve an egalitarian society, while cities were characterised by homogenous and standardised physical and social landscapes. However, since a market mechanism was introduced in the 1980s, an important question is whether the introduction of a market economy will result in an increase in residential segregation. Given the increasing role of socioeconomic status in the spatial sorting of populations, some scholars predicted a growing degree of convergence with the patterns of the United States and Western Europe (Brade et al., 2009; Szelenyi, 1996; Węcławowicz, 2002). However, recent evidence from a wide range of post-socialist countries in Central and Eastern Europe has shown that the rising income inequality has not led to clear-cut sociospatial divisions (Marcińczak et al., 2015). The levels of segregation have remained unchanged or have even slightly decreased (Marcińczak et al., 2014; Sýkora, 2009). While some scholars recognised this phenomenon as the sociospatial legacy of state socialism in transition to capitalism (Gentile et al., 2012; Sýkora and Bouzarovski, 2012), Golubchikov et al. (2014) argued for an understanding of the legacy as an integral part of the capitalist spatiality.

Different from the ‘big bang’ reforms in Central and Eastern European countries, China adopted a gradual and incremental approach. In the sphere of the housing system, which has direct impacts on urban residential segregation, a series of experiments for establishing a housing market were carried out starting in the late 1970s (Wang and Murie, 1999). However, it was not until 1998 that the welfare housing system was abolished entirely, and residents were allowed to choose freely where to live in cities. Since then, in contrast to the cities of Central and Eastern European countries, where spatial changes seem to lag behind institutional and socioeconomic ones (Sýkora and Bouzarovski, 2012), Chinese cities have undergone tremendous changes. The difference is the increasing role of a market mechanism in residential sorting in China. In Central and Eastern European countries, the housing market in the 1990s was established through privatisation and the restitution of properties to pre-Second World War owners. The socialist housing estates retained their socioeconomic status and continued accommodating mixed populations (Kovács and Herfert, 2012). Segregation was consequently reduced as the social status of pre-Second World War neighbourhoods was upgraded and that of socialist housing estates downgraded (Tammaru et al., 2016b). Moreover, the relatively low level of new residential development and the underdeveloped mortgage market greatly constrained people’s residential mobility (Marcińczak et al., 2015; Roy, 2008). In China, institutional changes on both the supply and demand sides created a prosperous housing market. The next section further elaborates on the dynamics of residential segregation under the market transition in China.

Residential segregation after housing reform in urban China

Chinese cities used to be less segregated and more spatially homogenous than their counterparts in market economies. The spatial organisation of the population was in accordance with the Danwei-based housing allocation. Individual traits and preferences had rarely been translated into sociospatial divisions. Instead, urban residents were spatially mixed in self-contained Danwei (work-unit) compounds where they were employed. It was common for employees of different ranks and occupations to live together (Bray, 2005). The urban structure was characterised by a Danwei-based cellular structure and shaped by land use patterns (Lo, 1994; Yeh et al., 1995).

Different from that of many Central and Eastern European post-socialist countries, China’s economic reform has been characterised by radical marketisation, which, however, is orchestrated by the state (Wu, 2018). As Nolan (1995: 4) noted, there is a ‘paradox that in the transition from a “planned” economy, a central condition of success is the ability of the state to plan effectively’. Market mechanisms have been introduced to achieve the state’s strategic goals. Among others, housing reform was particularly useful for boosting domestic demand and hence for ensuring economic growth (Wu, 2015). Along with the full development of the housing market, sociospatial differences are greatly reinforced by market mechanisms. This argument is further elaborated from the temporal, spatial and social perspectives as follows.

Establishing a housing market was a major part of China’s market reform. After several phases of gradualist reform in the 1980s and 1990s, a housing market began to emerge. As housing development was increasingly transferred to specialised real estate companies, commodity housing became more important in housing provision. However, the early stage of housing reform was dominated, first, by housing privatisation through the selling off of public housing to sitting tenants, and second, by the direct involvement of Danwei in housing investment (Wang and Murie, 1999). Consequently, Danwei remained central to the spatial sorting of urban residents (Logan et al., 1999). By the late 1990s, although socioeconomic inequalities based on market mechanisms had become prominent (Bian and Zhang, 2002), Chinese cities did not see a sharp rise in sociospatial differentiation (Li and Wu, 2008).

It was not until the total abolishment of the welfare housing system in 1998 that the market became the dominant mechanism of residential sorting. In fact, previous reforms had already laid the foundation for a booming housing market. On the supply side, housing development was opened to private capital through the establishment of a land leasing system (Xu et al., 2009), while on the demand side, earlier privatisation emancipated people from the control of residential mobility. Subsidised sales of public housing made these people homeowners, giving them the possibility of gaining benefits from future sales and asset appreciation in the market. The creation of provident funds and other mortgage instruments further increased the affordability for better-off families (Li, 2010). Because the old system did not disappear at once, and because changes in space usually lag behind socioeconomic ones, earlier studies found a low level of residential segregation in the year 2000 (Li and Wu, 2008). But as marketisation accelerated, market purchases rose greatly from 9.5% to 31%, and market rentals from 6.9% to 23.0% during the first decade of the 21st century (Yi and Huang, 2014). Thus, individual residential choices are becoming important in shaping sociospatial differences (Li and Wu, 2004; Shen and Wu, 2013). This leads to our first working hypothesis:

Hypothesis 1: The extent of socioeconomic segregation increased in the first decade after housing reform.

In the meantime, the impacts of market transition on segregation are spatially uneven. The introduction of the market economy has brought about radical morphological and functional changes in cities. This, however, is perhaps most evident at the urban periphery. With few spatial legacies of the pre-reform era, the suburbs have been almost built up from scratch under the market transition. Because of the high cost of relocation, developers have tended to bypass the city centres and instead have undertaken construction of massive commodity housing estates in the suburbs (Shen and Wu, 2017). The newly built apartments and villas have enticed many middle- and upper-class families to move out of their previous public houses (Pow, 2009; Zhang 2012). Meanwhile, previously rural villages have been rebuilt and extended by farmers to make profits from the niche rental market of migrants. Urban villages represent the market provision of housing and public services for migrants. As Shen and Wu (2013) noted, social stratification and spatial unevenness are actually reinforced through the process of suburbanisation. However, the vast suburban areas have not been included in previous studies (Li and Wu, 2008). Based on these empirical facts, we suggest the following working hypothesis:

Hypothesis 2: The extent of socioeconomic segregation is greater in the suburbs than in the city centres.

Finally, with the establishment of the housing market after the 1998 housing reform, the sorting of populations between neighbourhoods becomes largely based on their market positions. Two individual-level attributes are particularly important in shaping socioeconomic segregation. First, to a great extent, human capital factors have replaced political factors as determinants of socioeconomic success. Due to the increasing economic returns from education over time (Jansen and Wu, 2012), people with a higher educational attainment were more likely to purchase commodity housing (Huang and Jiang, 2009). Consequently, this group was more likely to become homeowners and to enjoy larger and better housing (Song and Xie, 2014). In other words, market mechanisms have reinforced the effects of human capital on sociospatial division.

Second, hukou status has become another crucial factor leading to housing inequality in cities. In contrast to post-socialist cities, where urban restructuring was dominated by suburbanisation and de-urbanisation (Hirt, 2007; Szelenyi, 1996), Chinese cities have witnessed the arrival of large numbers of rural-to-urban migrants. But the previous household registration (hukou) system has been retained. Many studies have revealed the segregation of migrants from the local population (Liao and Wong, 2015; Shen, 2017). For example, most migrants are segregated from hukou holders in informal housing in urban villages (Hao, 2015).

Importantly, the concentration of migrants increasingly reflected their position in the housing market. In essence, the hukou system reflects the state’s retreat from housing and other social provision for migrants. Although Logan et al. (2009) demonstrated that the effects of hukou status were due to its relation to access to public housing, social housing programmes account for only a small proportion of housing provision. Moreover, because migrants have limited access to public housing, they have to resort to the market sector. Most migrants are disadvantaged in the housing market because they lag behind hukou holders in educational attainment, earnings and wealth. However, migrants with a higher level of education do have access to better housing and better neighbourhoods, as human capital is the key determinant of migrant income (Démurger et al., 2009). Moreover, migrants with a higher level of education are also more likely to obtain an urban hukou through merit-based selection (Wu and Zheng, 2018). Accordingly, regarding the relative importance of individual educational attainment and hukou status, we suggest the following hypothesis:

Hypothesis 3: Educational attainment has become more important than hukou status in driving segregation since the establishment of the housing market.

Data and methods

Data and geographic units

Shanghai is chosen as a case study to examine the empirical element of this research. As the ‘dragonhead’ to connect China’s economy with the global economy, the entrepreneurship of Shanghai’s municipal government is prominent. In particular, the city is at the leading edge of Chinese market-oriented housing reform. While welfare housing allocation still exists in some cities like Beijing, the system was completely abolished in Shanghai to establish a booming housing market. Thus, public housing schemes were almost absent until 2008. Meanwhile, real estate became a pillar of Shanghai’s economy. In 2009, when the sector reached its peak, its added value accounted for 8.2% of the city’s GDP, and it contributed 20.4% of the city’s economic growth (Sun, 2013). The situation in Shanghai has been replicated by many other cities across the country. Therefore, Shanghai is a fairly representative case to illustrate socioeconomic segregation after housing reform in China.

The data come from the 2000 and 2010 censuses. During that decade, the population of Shanghai increased by 37.53%, from 16.74 million in 2000 to 23.02 million in 2010. The growth largely resulted from the influx of large numbers of migrants, who accounted for 39% of Shanghai’s total population in 2010. This study covers the entire administrative area of Shanghai municipality. We include both urban administrative units –‘street’ (jiedao) – and rural administrative units –‘township’ (zhen). In 2000, suburban townships covered vast rural areas. But during the following decade most of them experienced massive industrial and residential development and can be included as urban areas. In addition, while many original farmers in rural villages have moved to township centres, the villages have witnessed the arrival of large numbers of migrants. Young people have moved from distant locations to the central and inner suburban areas and left the local elderly behind. Therefore, the inclusion of townships may reflect the role of (sub)urbanisation in driving sociospatial divisions.

Instead of large spatial units, ‘sub-districts’ (jiedao and zhen) – which are too large for it to be possible to identify differences between neighbourhoods – residential committees (juwei) and village committees (cunwei) are used as the accounting units. With an average population of 3000, which is comparable to the census tract in the US, the two kinds of units represent neighbourhoods in Shanghai and are suitable for measuring residential segregation. The 2000 census contained 329 sub-districts and 6256 residential committee units, and the 2010 census included 230 sub-districts and 5432 residential committee units. Notably, the residential committee boundaries between the two census years are inconsistent. However, because the analysis was based on an aggregate index rather than on changes in every unit, errors caused by this are tolerable and have little direct influence on the final results.

Methods

There are two approaches to segregation analysis. First, the index-based approach, which emerged in the 1950s, is an effective way to measure the level of segregation in a city or region based on single-number measures. However, it is inherently aspatial, that is, places could have the same ID yet have very different spatial distributions of social groups (Grannis, 2002; Reardon and O’Sullivan, 2004). Second, the raw data approach was developed by Johnston and his colleagues (e.g. Johnston et al., 2007, 2009) to incorporate more geography into segregation studies. Different from the index-based approach that measures the extent to which social groups live apart from one another, the raw data approach focuses on the extent to which particular groups share residential neighbourhoods. Accordingly, typologies of neighbourhoods are developed based on the degree of compositional mixing. This approach looks beyond averages and identifies multiple types of residential mixing and segregation. It is also useful to examine change over time and across national contexts, where the specific groups identified in national censuses differ.

This study therefore includes both approaches to examine different aspects of socioeconomic segregation in Shanghai. First, based on the raw data approach, the article examines the geographies of residential intermixing in Shanghai. We adopted the typology of neighbourhoods developed by Marcińczak et al. (2015) to investigate the patterns of socioeconomic composition. Tables 1 and 2 show neighbourhood types according to residential composition in terms of educational attainment and hukou status respectively. The results of the assortment of the neighbourhoods in 2000 and 2010 were then mapped to reveal the changes after housing reform.

Table 1.

Neighbourhood types according to educational composition.

Share of groups based on educational attainment*	High	Middle	Low	2000	2010
High SES neighbourhood	50% <	0~50%	0~50%	52 (0.82%)	420 (7.73%)
High-middle SES neighbourhood	25%~49%	25%~49%	< 25%	75 (1.18%)	491 (9.04%)
High-low SES neighbourhood	25%~49%	< 25%	25%~49%	0	0
Middle SES neighbourhood	0~50%	50% <	0~50%	4880 (77.01%)	4240 (78.06%)
Middle-low SES neighbourhood	< 25%	25%~49%	25%~49%	98 (1.15%)	125 (2.30%)
Low SES neighbourhood	< 25%	25% <	50% <	1218 (19.22%)	150 (2.76%)
Mixed neighbourhood	25%~49%	25%~49%	25%~49%	2 (0.03%)	6 (0.11%)

Notes: *High educational attainment refers to college and above; middle educational attainment refers to junior high school, senior high school and technical high school; low educational attainment refers to primary school and below.

Table 2.

Neighbourhood types according to hukou composition.

Share of groups	Migrants	Hukou holders	2000	2010
Migrant neighbourhood	75% <	< 25%	82 (1.29%)	560 (10.31%)
Hukou-holder neighbourhood	< 25%	75% <	4909 (77.47%)	2489 (45.82%)
Mixed neighbourhood	< 75%	25% <	1334 (21.05)	2383 (43.87%)

Second, we used the traditional index-based approach to measure the extent of socioeconomic segregation. The index of dissimilarity (ID) is calculated as follows:

D = \frac{1}{2} \sum_{i = 1}^{N} | \frac{b_{i}}{B} - \frac{ω_{i}}{W} |

where $b$ and $ω$ are the minority and majority population counts in spatial units i, and B and W are the total minority and majority population counts of the entire region. The value of the ID ranges from 0 to 1, indicating no segregation to complete segregation respectively. We are aware that there are criticisms of this measurement, such as the checkerboard problem (White, 1983) and the modifiable areal unit problem (Wong, 1997). However, it remains the most widely used segregation index. It also makes it possible to compare the results with previous studies in China and other contexts.

The index-based analysis comprises two stages. In the first stage, a descriptive analysis of the overall changes of education and hukou segregation in Shanghai from 2000 and 2010 was conducted. Using the residential committee and the village committee as the accounting units, indices of dissimilarity are calculated for the entire Shanghai region, the city centre and the suburbs respectively. The outer ring road of Shanghai is used as the boundary defining the city centre and the suburbs. Sub-districts across the road are identified as central places. In 2000, we identified 117 sub-districts and 2797 residential committees in the central city, and 254 sub-districts and 3540 residential committees in the suburbs; in 2010, there were 105 sub-districts and 2701 residential committees in the central city, and 125 sub-districts and 2731 residential committees in the suburbs. With regard to the impact of the changing population on the index, although the index is independent of the relative size of the two groups used in its computation, it is appropriate to use for when there were hardly any minority residents in the beginning. In general, scholars use unweighted estimates to indicate segregation in the average place; weighted estimates are more indicative of the segregation experiences of minority groups. Since we are not interested in the segregation of specific groups, we use unweighted estimates in the analysis (see Logan et al., 2004).

In the second stage, we examined the spatial unevenness of segregation. The indices of dissimilarity for each sub-district are calculated with committees as the accounting units. According to Lichter et al. (2015), competing places or political units, rather than neighbourhoods, are the key units that shape segregation through ongoing local economic and demographic processes. In the case of China, the sub-district (jiedao or zhen) is the smallest administrative unit underpinned by the expenditure budget of local government (Li and Wu, 2008). Using the segregation scores of sub-districts as the dependent variable, multivariate analysis is further conducted.

In particular, following convention (Iceland and Sharp, 2013; Massey and Denton, 1989), a pooled regression analysis is adopted to test the significant differences of segregation across different times, spaces and social groups. Hukou and education ID scores for the sub-districts of both census years are pooled. Dummy variables for year (i.e. 2000 and 2010) are used to capture the general trend in segregation over time, dummy variables for location (i.e. city centre and suburbs) are used to examine the spatial pattern of segregation, and dummy variables for dimension of segregation (i.e. hukou and education) are used to compare the extent of hukou and educational segregation. Regression models are conducted for the whole city, for central and suburban places, and for 2000 and 2010 respectively.

Results

Patterns of socioeconomic intermixing in Shanghai

The typology of neighbourhoods based on their residential composition in Shanghai indicates an increase in sociospatial division since the housing reform (Figures 1 and 2). In 2000, the educational attainment of the population was relatively low. The geography of neighbourhoods was characterised by neighbourhoods of native residents with middle educational attainment in the central areas and neighbourhoods of native residents with low educational attainment in the outer suburban areas. To a large extent, the distribution of neighbourhoods reflects the urban–rural division between the central and the peripheral areas. But along with the arrival of migrants in the 1990s, there emerged large numbers of mixed neighbourhoods where migrants and natives lived together in inner suburban locations.

Figure 1.

Neighbourhood types by educational composition in 2000 (left) and 2010 (right).

Figure 2.

Neighbourhood types by hukou composition in 2000 (left) and 2010 (right).

By 2010, a decade after the housing reform, a spatial pattern featuring socioeconomic status-based spatial divisions began to emerge. While the average socioeconomic status of the population had improved, there were many more high-status neighbourhoods (see Table 1). These neighbourhoods were mostly distributed in the city centre, particularly on both sides of the inner ring road and along the radial traffic arteries that extended south-west from the urban core. The concentration of migrants at outer suburban locations gave rise not only to mixed neighbourhoods, but also to migrant enclaves (see Table 2).

Trends in socioeconomic segregation from 2000 to 2010

Table 3 describes the overall changes in socioeconomic segregation in Shanghai from 2000 to 2010. First and foremost, unlike the conventional perception that the extent of segregation in Chinese cities is moderate, the results indicate that different social groups are also greatly segregated at the neighbourhood level. In 2010, the isolation of people with higher education from those with secondary and primary education had reached as high as 52.2. The scores of hukou segregation were 39.3. The extent of segregation either reached or was even higher than that of large cities in both the US and European countries. For instance, even though the levels of socioeconomic segregation in US and European cities had increased in the last decade, none had an index larger than 40 (Marcińczak et al., 2015; Musterd, 2005). Indeed, one recent study based on directly comparable measures and data has demonstrated that the average index of dissimilarity of educational segregation for large US and French cities was 31 and 33 respectively (Quillian and Lagrange, 2016).

Table 3.

Dissimilarity indices for Shanghai in 2000 and 2010.

			Shanghai	Central city	Suburbs
Hukou segregation	Migrants–local hukou holders	2000	0.422	0.4083	0.4346
		2010	0.3933	0.2817	0.4672
		Changes	–6.80%	–31.01%	7.50%
Educational segregation	High–low educational attainment	2000	0.5472	0.3769	0.6529
		2010	0.5762	0.4861	0.5848
		Changes	5.30%	28.97%	–10.43%
	High–middle educational attainment	2000	0.3733	0.3121	0.4881
		2010	0.5222	0.4475	0.5165
		Changes	39.89%	43.38%	5.82%
	Middle–low educational attainment	2000	0.2487	0.1113	0.2247
		2010	0.1967	0.1686	0.2103
		Changes	–20.91%	51.48%	–6.41%

Second, the extent of segregation differs in central and suburban locations. Overall, different social groups are more segregated from each other in the suburbs, as the levels of segregation across both dimensions are higher on average in suburban areas than in city centres. But changes in hukou and educational segregation are spatially variegated. The city centre experienced a decline in hukou segregation, but a rapid increase in educational segregation. This was largely associated with the emergence of high-end estates at inner suburban locations. Meanwhile, along with the arrival of migrants and urban residents, the suburbs had become more heterogeneous than they were when local farmers dominated the region. But because they were spatially sorted through the housing market, migrants became more segregated in the suburbs, as hukou segregation increased by 7.5 percent.

Finally, hukou segregation declined while educational segregation greatly increased over the 2000s. The overall hukou segregation decreased by 6.8 per cent, which means hukou status became less important in shaping sociospatial differences. On the contrary, the population with a higher educational attainment became highly segregated from that with a lower educational attainment. In particular, segregation of the population with secondary education from the population with higher education increased by almost 40 per cent. These changes indicated that human capital is likely to become more important in driving segregation than is hukou status.

Multivariate analysis: Changing patterns of socioeconomic segregation

Table 4 presents the segregation scores for sub-districts in Shanghai in 2000 to 2010, where residential committees were used as basic accounting units. It is found that the index of dissimilarity was lower at the sub-district scale than at the city scale. Such differences may reflect a case of macro segregation (segregation between sub-districts) rather than a case of micro segregation (segregation within sub-districts) in Shanghai (Reardon et al., 2008). But the overall trends at both scales are similar. While the mean index of dissimilarity for migrants from hukou holders decreased, educational segregation across all pairwise comparisons experienced significant increases. In particular, residential segregation of groups with low educational attainment from those with middle educational attainment increased most.

Table 4.

Segregation scores for sub-districts in Shanghai, in 2000 and 2010.

Index of dissimilarity	Migrants–local hukou holders		Low–high educational attainment		Low–middle educational attainment		Middle–high educational attainment
Index of dissimilarity	Mean	SD	Mean	SD	Mean	SD	Mean	SD
2000	0.269	0.120	0.360	0.155	0.115	0.067	0.279	0.134
2010	0.237	0.134	0.413	0.182	0.174	0.070	0.367	0.181

Table 5 reports the results of pooled regression analysis on changes in segregation from 2000 to 2010. Model I confirms that, irrespective of the dimension, residential segregation greatly increased after the first decade of housing reform. This is true in both the city centre and the suburbs, as shown by Model II and Model III. This confirms the first hypothesis with regard to the increases in residential segregation over the decade.

Table 5.

Pooled regression models of segregation for sub-districts in Shanghai, 2000–2010.

	Model I:Entire city		Model II:Central city		Model III:Suburb
	Coef.	t	Coef.	t	Coef.	t
Year (year 2010 = 1)	0.042	5.96***	0.029	3.23***	0.069	6.79***
Constant	0.257	57.31***	0.222	36.4***	0.279	45.91***
	Adj R-squared = 0.0152		Adj R-squared = 0.0094		Adj R-squared = 0.0352
	F = 35.49		F = 10.41		F = 46.15
	P = 0.0000		P = 0.0013		P = 0.0000
	N = 2236		N = 996		N = 1240

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001.

Tables 6 and 7 present segregation patterns for 2000 and 2010 respectively. In terms of spatial pattern, the coefficients of suburb are significantly positive for both years, indicating that segregation is higher in suburban locations. With regard to the dimension of segregation, in 2000, while it is significantly negative in Model V for the city centre, the coefficients in Model IV for the entire city and Model VI for the suburbs are not significant. In other words, hukou segregation used to be dominant, particularly in the city centre. The extents of hukou segregation and educational segregation were almost the same. In 2010, the coefficients in all three models are positive and significant. This means that the effect of educational attainment significantly exceeded that of hukou. Human capital played a larger role in shaping segregation across the entire city.

Table 6.

OLS regression models of segregation for sub-districts in Shanghai, 2000.

	Model IV:Entire city		Model V:Central city		Model VI:Suburb
	Coef.	t	Coef.	t	Coef.	t
Dimension of segregation(education = 1)	–0.018	–1.93	–0.064	–4.63***	0.012	0.92
Location (suburb = 1)	0.057	6.84***
Constant	0.236	24.50***	0.206	30.05***	0.282	44.31***
	Adj R-squared = 0.0356		Adj R-squared = 0.0377		Adj R-squared = –0.0002
	F = 25.27		F = 21.47		F = 0.84
	P = 0.0000		P = 0.0000		P = 0.3584
	N = 1316		N = 524		N = 792

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 7.

OLS regression models of segregation for sub-districts in Shanghai, 2010.

	Model VII:Entire city		Model VIII:Central city		Model IX:Suburb
	Coef.	t	Coef.	t	Coef.	t
Dimension of segregation (education = 1)	0.081	6.39***	0.096	6.75***	0.065	3.04**
Location (suburb = 1)	0.097	8.87***
Constant	0.190	15.55***	0.178	14.4***	0.300	16.29***
	Adj R-squared = 0.1134		Adj R-squared = 0.0864		Adj R-squared = 0.0182
	F = 59.76		F = 45.54		F = 9.27
	P = 0.0000		P = 0.0000		P = 0.0025
	N = 920		N = 472		N = 448

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001.

Conclusions

In the former state socialist systems, cities in pre-reform China consisted of socially mixed neighbourhoods. After the introduction of the market economy, despite a common perception of increased sociospatial differentiation, empirical studies documented a relatively low level of segregation by individual-level attributes (Li and Wu, 2008). However, due to the data availability problem, whether sociospatial differences have been reinforced after the 1998 housing reform remains in question. Based on a longitudinal analysis of Shanghai, for the first time, this study empirically examines the changing patterns of socioeconomic segregation as well as the underlying mechanisms.

Our findings demonstrate that urban space in China has become more partitioned. Analysis of residential intermixing within neighbourhoods shows that clusters of high-status neighbourhoods and clusters of migrant neighbourhoods have emerged in Shanghai. Meanwhile, the level of socioeconomic segregation has also increased since the 1998 housing reform. In 2010, the levels of segregation by both educational attainment and hukou status in Shanghai were comparable to the levels of socioeconomic segregation in large US and European cities (Marcińczak et al., 2015; Quillian and Lagrange, 2016).

Patterns of sociospatial divisions differ significantly across the metropolitan area. The central areas have witnessed the emergence of many mixed neighbourhoods where migrants and hukou holders live. But the development of high-status neighbourhoods along the inner ring road has led to a great increase in educational segregation. In the suburbs, along with the emergence of both high-status neighbourhoods and migrant neighbourhoods, the suburbs have become more heterogeneous. Accordingly, the level of segregation is greater in the suburbs than in central areas.

The underlying drivers of residential segregation have also changed. The market mechanism has become dominant in residential sorting. In particular, housing commodification is a key factor contributing to the sociospatial divisions.¹ The multivariate analysis confirms that while hukou segregation remains important for understanding sociospatial divisions, the level of socioeconomic segregation based on individual human capital has increased and exceeded the level of hukou segregation.

Notably, recent policies on migrants in large cities reflect a strengthening of state interventions in residential sorting. However, human capital may become even more important in shaping socioeconomic segregation. On the one hand, the lowering of thresholds of hukou attainment and the resurgence in public housing provision are targeted towards talented people. In Beijing and Shanghai, for example, migrants without hukou now have access to public housing if they meet the criteria for talent identification. On the other hand, urban villages have been demolished on a huge scale to drive away migrants with low educational attainment and in low-end jobs. Consequently, sociospatial exclusion of this group could be reinforced.

Broadly speaking, recent studies on socioeconomic residential segregation call for cross-national comparisons to understand the phenomenon in different contexts (Maloutas and Fujita, 2012; Quillian and Lagrange, 2016; Tammaru et al., 2016a). In particular, residential differentiation in post-socialist cities has drawn the attention of many scholars. The focus is on whether the systematic transition from central planning to market economy would lead to clear-cut socioeconomic spatial divisions. The findings of this study on China contribute to the debate by highlighting the importance of contextual factors for the trend in segregation in these cities. Different from Central and Eastern European cities, where increasing economic inequality has brought about not segregated but mixed neighbourhoods (Marcińczak et al., 2015), socioeconomic segregation has increased in Shanghai. This distinction is associated with two specificities in China: the development of a booming housing market and the influx of large numbers of migrants. Therefore, although structural factors such as globalisation and market transition are important, these effects are mediated through contextual factors.

Footnotes

Acknowledgements

We are grateful to the editors and the anonymous reviewers for their very helpful comments and suggestions on earlier versions of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

We would like to acknowledge funding support from the National Natural Science Foundation of China (0042098019834233; 41501170). The first author thanks the Urban Studies Foundation for funding her research as a visiting scholar at University College London, UK.

Notes

ORCID iD

Yang Xiao

References

Arbaci

(2007) Ethnic segregation, housing systems and welfare regimes in Europe. International Journal of Housing Policy 7(4): 401–433.

Bian

Zhang

(2002) Marketization and income distribution in urban China, 1988 and 1995. Research in Social Stratification and Mobility 19: 377–415.

Boterman

Van Gent

WPC

(2014) Housing liberalisation and gentrification: The social effects of tenure conversions in Amsterdam. Tijdschrift voor Economische en Sociale Geografie 105: 140–160.

Brade

Herfert

Wiest

(2009) Recent trends and future prospects of sociospatial differentiation in urban regions of Central and Eastern Europe: A lull before the storm? Cities 26: 233–244.

Bray

(2005) Social Space and Governance in Urban China: The Danwei System from Origins to Reform. Stanford, CA: Stanford University Press.

Burgers

Musterd

(2002) Understanding urban inequality: A model based on existing theories and an empirical illustration. International Journal of Urban and Regional Research 26(2): 403–413.

Burgess

(1928) Residential segregation in American cities. The Annals of the American Academy of Political and Social Science 140: 105–115.

Démurger

Gurgand

, et al. (2009) Migrants as second-class workers in urban China? A decomposition analysis. Journal of Comparative Economics 37(4): 610–628.

Duncan

(1955) A methodological analysis of segregation indexes. American Sociological Review 20: 210–217.

10.

Farley

(1977) Residential segregation in urbanized areas of the United States in 1970: An analysis of social class and racial differences. Demography 14: 497–518.

11.

Fischer

Stockmayer

Stiles

, et al. (2004) Distinguishing the geographic levels and social dimensions of US metropolitan segregation, 1960–2000. Demography 41: 37–59.

12.

Gentile

Tammaru

Van Kempen

(2012) Heteropolitanization: Social and spatial change in Central and Eastern Europe. Cities 29(5): 291–299.

13.

Golubchikov

Badyina

Makhrova

(2014) The hybrid spatialities of transition: Capitalism, legacy and uneven urban economic restructuring. Urban Studies 51: 617–633.

14.

Grannis

(2002) Discussion: Segregation indices and their functional inputs. Sociological Methodology 32(1): 69–84.

15.

Shen

(2003) Transformation of urban sociospatial structure in socialist market economies: The case of Beijing. Habitat International 27: 107–122.

16.

Hamnett

(1996) Social polarization, economic restructuring and welfare state regimes. Urban Studies 33(8): 1407–1430.

17.

Hao

(2015) The effects of residential patterns and Chengzhongcun housing on segregation in Shenzhen. Eurasian Geography and Economics 56(3): 308–330.

18.

(2010) New-build gentrification in Central Shanghai: Demographic changes and socioeconomic implications. Population, Space and Place 16: 345–361.

19.

Liu

Yang

, et al. (2017) Housing differentiation and housing poverty in Chinese low-income urban neighborhoods under the confluence of state–market forces. Urban Geography 38: 729–751.

20.

Hirt

(2007) Suburbanizing Sofia: Characteristics of post-socialist peri-urban change. Urban Geography 28: 755–780.

21.

Huang

Clark

(2002) Housing tenure choice in transitional urban China: A multilevel analysis. Urban Studies 39: 7–32.

22.

Huang

Jiang

(2009) Housing inequality in transitional Beijing. International Journal of Urban and Regional Research 33(4): 936–956.

23.

Iceland

Sharp

(2013) White residential segregation in US metropolitan areas: Conceptual issues, patterns, and trends from the US census, 1980 to 2010. Population Research and Policy Review 32: 663–686.

24.

Jansen

(2012) Income inequality in urban China, 1978–2005. Chinese Sociological Review 45(1): 3–27.

25.

Johnston

Poulsen

Forrest

(2007) The geography of ethnic residential segregation: A comparative study of five countries. Annals of the Association of American Geographers 97(4): 713–738.

26.

Johnston

Poulsen

Forrest

(2009) Research note – Measuring ethnic residential segregation: Putting some more geography in. Urban Geography 30: 91–109.

27.

Kemeny

(1995) From Public Housing to the Social Market: Rental Policy Strategies in Comparative Perspective. London: Routledge.

28.

Kovács

Herfert

(2012) Development pathways of large housing estates in post-socialist cities: An international comparison. Housing Studies 27(3): 324–342.

29.

Lee

Murie

(1997) Poverty, Housing Tenure and Social Exclusion. Bristol: Policy Press.

30.

S-M

(2010) Mortgage loan as a means of home finance in urban China: A comparative study of Guangzhou and Shanghai. Housing Studies 25: 857–876.

31.

S-M

(2004) Contextualizing Residential Mobility and Housing Choice: Evidence from Urban China. London: SAGE Publications.

32.

(2008) Tenure-based residential segregation in post-reform Chinese cities: A case study of Shanghai. Transactions of the Institute of British Geographers 33: 404–419.

33.

Liao

Wong

(2015) Changing urban residential patterns of Chinese migrants: Shanghai, 2000–2010. Urban Geography 36: 109–126.

34.

Lichter

Parisi

Taquino

(2015) Toward a new macro-segregation? Decomposing segregation within and between metropolitan cities and suburbs. American Sociological Review 80: 843–873.

35.

Liu

Silva

Long

(2019) Block-level changes in the sociospatial landscape in Beijing: Trends and processes. Urban Studies 56(6): 1198–1214.

36.

Ljunggren

Andersen

(2015) Vertical and horizontal segregation: Spatial class divisions in Oslo, 1970–2003. International Journal of Urban and Regional Research 39: 305–322.

37.

(1994) Economic reforms and socialist city structure: A case study of Guangzhou, China. Urban Geography 15: 128–149.

38.

Logan

Bian

(1999) Housing inequality in urban China in the 1990s. International Journal of Urban and Regional Research 23: 7–25.

39.

Logan

Fang

Zhang

(2009) Access to housing in urban China. International Journal of Urban and Regional Research 33: 914–935.

40.

Logan

Stults

Farley

(2004) Segregation of minorities in the metropolis: Two decades of change. Demography 41: 1–22.

41.

Madrazo

Van Kempen

(2012) Explaining divided cities in China. Geoforum 43: 158–168.

42.

Maloutas

Fujita

(2012) Residential Segregation in Comparative Perspective. Farnham: Ashgate Publishing.

43.

Marcińczak

Gentile

Rufat

, et al. (2014) Urban geographies of hesitant transition: Tracing socioeconomic segregation in post-Ceauşescu Bucharest. International Journal of Urban and Regional Research 38: 1399–1417.

44.

Marcińczak

Tammaru

Novák

Gentile

(2015) Patterns of socioeconomic segregation in the capital cities of fast-track reforming postsocialist countries. Annals of the Association of American Geographers 105: 183–202.

45.

Massey

Denton

(1989) Hypersegregation in US metropolitan areas: Black and Hispanic segregation along five dimensions. Demography 26: 373–391.

46.

Musterd

(2005) Social and ethnic segregation in Europe: Levels, causes, and effects. Journal of Urban Affairs 27: 331–348.

47.

Musterd

Ostendorf

(1998) Urban Segregation and the Welfare State. London and New York: Routledge.

48.

Nightingale

(2012) Segregation: A Global History of Divided Cities. Chicago, IL: University of Chicago Press.

49.

Nolan

(1995) China’s Rise, Russia’s Fall: Politics, Economics and Planning in the Transition from Socialism. Basingstoke: Palgrave Macmillan.

50.

Park

(1915) The city: Suggestions for the investigation of human behavior in the city environment. American Journal of Sociology 20: 577–612.

51.

Pattaroni

Kaufmann

Thomas

(2012) The dynamics of multifaceted gentrification: A comparative analysis of the trajectories of six neighbourhoods in the Île-de-France region. International Journal of Urban and Regional Research 36: 1223–1241.

52.

Pow

(2009) Public intervention, private aspiration: Gated communities and the condominisation of housing landscapes in Singapore. Asia Pacific Viewpoint 50: 215–227.

53.

Préteceille

(2000) Segregation, class and politics in large cities. In: Bagnasco

LeGalès

(eds) Cities in Contemporary Europe. Cambridge: Cambridge University Press, pp. 74–97.

54.

Quillian

Lagrange

(2016) Socioeconomic segregation in large cities in France and the United States. Demography 53: 1051–1084.

55.

Reardon

Bischoff

(2011) Income inequality and income segregation. American Journal of Sociology 116: 1092–1153.

56.

Reardon

O’Sullivan

(2004) Measures of spatial segregation. Sociological Methodology 34: 121–162.

57.

Reardon

Matthews

O’Sullivan

, et al. (2008) The geographic scale of metropolitan segregation. Demography 45(3): 489–514.

58.

Roy

(2008) Mortgage markets in central and eastern Europe: A review of past experiences and future perspectives. International Journal of Housing Policy 8: 133–160.

59.

Sassen

(1991) The Global City: New York, London, Tokyo. Princeton, NJ: Princeton University Press.

60.

Shen

(2017) Stuck in the suburbs? Sociospatial exclusion of migrants in Shanghai. Cities 60: 428–435.

61.

Shen

(2013) Moving to the suburbs: Demand-side driving forces of suburban growth in China. Environment and Planning A 45: 1823–1844.

62.

Shen

(2017) The suburb as a space of capital accumulation: The development of new towns in Shanghai, China. Antipode 49(3): 761–780.

63.

Song

Xie

(2014) Market transition theory revisited: Changing regimes of housing inequality in China, 1988–2002. Sociological Science 1: 277–291.

64.

Sun

(2013) The dependence of Shanghai’s economy on real estate substantially weakened. People’s Daily. Available at: http://theory.people.com.cn/n/2013/0220/c40531-20538994.html (accessed 16 December 2017).

65.

Sýkora

(2009) New socio-spatial formations: Places of residential segregation and separation in Czechia. Tijdschrift voor Economische en Sociale Geografie 100: 417–435.

66.

Sýkora

Bouzarovski

(2012) Multiple transformations: Conceptualising the post-communist urban transition. Urban Studies 49: 43–60.

67.

Szelenyi

(1996) Cities under socialism – and after. In: Andrusz G, Harloe M and Szelenyi I (eds) Cities After Socialism: Urban and Regional Change and Conflict in Post-socialist Societies. Oxford: Blackwell, pp. 286–317.

68.

Tammaru

Marcińczak

Van Ham

, et al. (eds) (2016a) Socio-economic Segregation in European Capital Cities: East Meets West. London: Routledge.

69.

Tammaru

Musterd

Van Ham

, et al. (2016b) A multifactor-approach to understanding socio-economic segregation in European capital cities. In: Tammaru

Marcińczak

Van Ham

, et al. (eds) Socio-economic Segregation in European Capital Cities: East Meets West. London: Routledge, pp. 1–29.

70.

Van Ham

Manley

(2009) Social housing allocation, choice and neighbourhood ethnic mix in England. Journal of Housing and the Built Environment 24: 407–422.

71.

Van Kempen

Murie

(2009) The new divided city: Changing patterns in European cities. Tijdschrift voor economische en sociale geografie 100: 377–398.

72.

Wang

Murie

(1999) Commercial housing development in urban China. Urban Studies 36: 1475–1494.

73.

Wang

(2010) Housing migrant workers in rapidly urbanizing regions: A study of the Chinese model in Shenzhen. Housing Studies 25: 83–100.

74.

Węcławowicz

(2002) From egalitarian cities in theory to non-egalitarian cities in practice: The changing social and spatial patterns in Polish cities. In: Marcuse

Van Kempen

(eds) Of States and Cities: The Partitioning of Urban Space. Oxford: Oxford University Press, pp. 183–199.

75.

White

(1983) The measurement of spatial segregation. American Journal of Sociology 88: 1008–1018.

76.

Wong

DWS

(1997) Spatial dependency of segregation indices. Canadian Geographer 41: 128–136.

77.

(2015) Commodification and housing market cycles in Chinese cities. International Journal of Housing Policy 15(1): 6–26.

78.

(2018) Planning centrality, market instruments: Governing Chinese urban transformation under state entrepreneurialism. Urban Studies 55(7): 1383–1399.

79.

Webster

(2010) Path dependency and the neighbourhood effect: Urban poverty in impoverished neighbourhoods in Chinese cities. Environment and Planning A 42: 134–152.

80.

Cheng

Chen

(2014) Sociospatial differentiation and residential segregation in the Chinese city based on the 2000 community-level census data: A case study of the inner city of Nanjing. Cities 39: 109–119.

81.

Cheng

Young

(2017) Social differentiation and spatial mixture in a transitional city: Kunming in southwest China. Habitat International 64: 11–21.

82.

Edensor

Cheng

(2018) Beyond space: Spatial (re)production and middle-class remaking driven by Jiaoyufication in Nanjing City, China. International Journal of Urban and Regional Research 42(1): 1–19.

83.

Zheng

(2018) Household registration, urban status attainment, and social stratification in China. Research in Social Stratification and Mobility 53: 40–49.

84.

Xiao

Guo

, et al. (2017a) Estimating the willingness to pay for green space services in Shanghai: Implications for social equity in urban China. Urban Forestry & Urban Greening 26: 95–103.

85.

Xiao

Wang

Fang

(2019) Exploring the disparities in park access through mobile phone data: Evidence from Shanghai, China. Landscape and Urban Planning 181: 80–91.

86.

Xiao

Wang

, et al. (2017b) An assessment of urban park access in Shanghai – Implications for the social equity in urban China. Landscape And Urban Planning 157: 383–393.

87.

Yeh

(2009) Land commodification: New land development and politics in China since the late 1990s. International Journal of Urban and Regional Research 33: 890–913.

88.

Yeh

AG-O

(1995) The social space of Guangzhou city, China. Urban Geography 16: 595–621.

89.

Huang

(2014) Housing consumption and housing inequality in Chinese cities during the first decade of the twenty-first century. Housing Studies 29: 291–311.

90.

Zhang

(2012) In Search of Paradise: Middle-Class Living in a Chinese Metropolis. Ithaca, NY: Cornell University Press.

Emerging divided cities in China: Socioeconomic segregation in Shanghai,2000–2010

Abstract

Keywords

Introduction

Socioeconomic segregation: institutional and contextual perspectives

Residential segregation after housing reform in urban China

Data and methods

Data and geographic units

Methods

Results

Patterns of socioeconomic intermixing in Shanghai

Trends in socioeconomic segregation from 2000 to 2010

Multivariate analysis: Changing patterns of socioeconomic segregation

Conclusions

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

Notes

ORCID iD

References