The Chinese Socio-Cultural Sustainability Approach: The Impact of Conservation Planning on Local Population and Residential Mobility

Abstract

Retaining indigenous populations is vital to the sustainable development and conservation of historic urban areas. However, little attention has been paid to Chinese conservation planning in an effort to safeguard indigenous people. This paper investigates population change and residential mobility in Chinese historic urban areas by applying demographic analysis and regression models to survey data collected in the Ping-Jiang Historic Quarter. Results indicate that few relocation behaviors are a result of the welfare housing policy and property ownership. However, residents’ intentions to move have increased, due to declining living conditions and tourism development in recent years. Classified by property ownership, public housing tenants and rented housing migrants were more willing to move, while private housing owners preferred to stay. Accordingly, there have been increasing trends of aging, poverty growth and population displacement, epitomized by the public housing population. Assessing planning impacts, welfare policy reduced residential mobility while undermining residents’ self-reliance to maintain their own houses. Without substantial social participation and community support, top-down conservation planning could only slow, rather than reverse, the trend of socio-cultural decline.

1. Introduction

Historic urban areas are among the most abundant and diverse manifestations of cultural heritage [1]. Their tangible and intangible values are comprised of not only material assets, but also the successive and existing cultures of local people. However, in the present global economy, cities with similar advantages increasingly use their cultural heritage as a tool for intercity competition [2,3]. Cultural policy functions as de facto economic policy in the general trend of gentrification as a global urban strategy [4,5]. In this context, the influx of tourists, replacement of traditional activities, and crumbling sense of identity all make historic urban areas unsuitable for the daily life of local people [6]. The sustainability of historic urban areas is confronted with the serious dilemmas of loss of indigenous people and loss of identity, especially in fast-developing countries that are currently embracing economic and social transitions [7].

In China, historic areas are the major locations for the practice of historic conservation. As of 2017, there had been 2700 “Historic and Cultural Quarters” (hereafter historic quarters) designated by the State Council, provincial governments and local authorities (incomplete statistics). Among these listed heritage sites, a number still functioned as a “living place” for indigenous people. In the past, the protection of historic quarters and the improvement of residents’ living conditions were considered a burden to local finance [8,9]. With the rapid growth of tourism since the beginning of the new century, historic residential areas have become hot tourist attractions. The new concept that “conservation and reuse are also (economic) development” was gradually accepted by decision makers [10], which then triggered the nation-wide commercialization of historic residential areas [11]. In this process, many indigenous people were mandatorily relocated for tourism development, leading to the loss of heritage values and social inequality problems [12,13]. However, in recent years a growing number of local governors have realized the importance of retaining indigenous populations, in view of the homogenization of tourism products, the increase of intercity competition, and the “governance transition” of state policy [14,15,16]. Retaining indigenous populations has become a focal issue in the practice of historic conservation [17,18].

In the existing literature of Chinese historic conservation, studies on the indigenous population mainly concern the demographic structure, living conditions, social participation and gentrification process. Many studies have found that the populations living in Chinese historic quarters were old and poor, and living in dilapidated houses with a lack of public facilities and modern infrastructure [19,20]. Their needs were seldom considered, due to a lack of social empowerment and community participation [21,22]. In fact, from the very beginning of strategy formulation, the gentrification process in Chinese historic quarters has been characterized in policy guidance [23,24,25]. Given the common perception of state dominance, mandatory relocation projects led by the “pro-growth coalition” were blamed for the loss of heritage values, as well as indigenous people [26,27]. However, after the promulgation of Property Law in 2007 and the Regulations on the Conservation of Historic Cities, Towns and Villages in 2008, the number of mandatory relocation projects saw a substantial decline. Meanwhile, large-scale commercial and high-end real estate developments have been subject to the strict supervision of conservation experts with the establishment of the Planning Inspector System [28,29]. The social force in urban affairs is also growing by virtue of the increasing number of conservation NGOs (Non-Governmental Organizations) and the promotion of community participation in planning processes [30,31]. In this context, as many studies have observed, voluntary resident relocation has become the main cause of population loss in Chinese historic quarters [32,33,34]. Nevertheless, residential mobility of indigenous people has seldom been empirically and systematically investigated.

In other disciplines, aside from historic conservation, there are abundant studies on residential mobility in urban China, classified by different types of residential blocks and different social groups [35,36]. However, little attention has been paid to residential mobility in historic quarters. This is because conventional research about residential mobility is geared toward non-regulated or semi-regulated housing estates [37], which is the opposite of the situation for Chinese historic quarters. As a result, residential mobility in historic areas was only examined as a secondary consideration of studies of old persons’ migration and low-income groups. Residential mobility in Chinese traditional inner-city neighborhoods was found to be relatively low [38,39,40]. Moreover, the stability of the population was explained as the result of aging and poverty [41], which only provides a partial understanding of Chinese historic quarters, which used to be (and still are, in some cases) home to some of the city’s richest families. A few scholars have explored relocation intentions in Chinese historic quarters, but only considered the universal factors of urban renewal, housing conditions, and socio-demographics, which are applicable to all residential blocks [37]. The special housing policy and tourism development, which are essential parts of Chinese conservation planning, have been ignored when analyzing the determining factors of residential mobility of indigenous people.

Based on empirical research, this paper investigates population change and residential mobility in Chinese historic urban areas by applying demographic analysis and regression models to survey data collected in the Ping-Jiang Historic Quarter. Below, we first review the practices of historic conservation in both western countries and China. Planning measures in relation to residential mobility will be classified in preparation for regression analysis. Then, drawing on Bourne’s relocation model, we divide residential mobility into relocation behavior and relocation desire to explore the decisive factors inhibiting residential relocation, and the determining factors affecting residents’ relocation desire. Finally, the empirical findings will be discussed in comparison with the results presented in previous studies. These findings could help planners and decision makers to realize the limitation of top-down interventions, the importance of social participation, and the clarity of property rights as a key to solving China’s conservation dilemmas.

2. Literature Review

It is not questioned that the conservation of historic areas must take measures to protect local residents. The Valletta Principle clearly indicates that the key to sustainable conservation entails multidisciplinary cooperation and joint management of the “quantity, coherence… and time of change” [6]. No identity can be created in a short time. The spirit of a place and its sense of identity are based on the compact relationship between space and time, which is consolidated in the individual and collective memories of the people [42]. In this sense, the conservation of indigenous people not only needs to avoid sudden population displacement, but also rapid population migration, because high mobility is indicative of a lack of social stability and place attachment [43].

According to residential mobility studies, numerous factors are influential in driving population migration. For instance, the distribution of public facilities, housing prices, and job opportunities are decisive in the residential choice of a person or family [44,45]. However, before the act of relocation, the living quality of present housing, sense of community, and changes in age, income, marriage status, or family structure might affect residential satisfaction and fuel a subjective desire to move [46,47,48].

In developed countries, there is a tradition of multidisciplinary cooperation in historic conservation, especially when it comes to the indigenous community [49]. Conservationists, planners, sociologists, economic geographers and urban administrators provide comprehensive measures to retain indigenous people while attracting new ones [50]. For those cities which have undergone tremendous population loss [51], cultural-led urban regeneration was a primary measure to attract population and capital “back to the city” [52,53]. However, with the increase of housing prices, gentrification arises as an undesired side-effect of historic conservation [54,55]. To protect poor indigenous people, public housing, repair subsidies and other forms of social welfare have been provided. Meanwhile, students, the creative class, and the middle class are welcomed to settle in historic areas in an attempt to improve social vitality [56,57,58]. Recently, public policy to promote social sustainability has transformed from poverty alleviation into the promotion of community governance via capacity building and cross-sectional partnerships [59,60].

In comparison to European practice, Chinese planning measures to protect indigenous people are authoritatively top-down [61], with a lack of subjective assessment and feedback from citizens [62]. Under the present conservation system, top-down interventions applying to the indigenous population are conducted on three levels [63]. At the city level, a major strategy is to decentralize inner-city populations to new towns in the outskirts, as a solution to the overcrowded living conditions which prevail in most historic cities and urban areas [64]. At the neighborhood level, improvement plans for infrastructure, public service facilities, roading and traffic management are compulsory for conservation planning according to Chinese conservation laws [65]. At the building level, ranking and classification of built heritage are conventional methods to allocate limited conservation funds [66]. Thus, less valued traditional houses, which are homes for most indigenous people, have fallen into disrepair due to a lack of attention and conservation funds [67]. In order to retain indigenous people, local governments have to provide extra housing subsidies on the basis of the old public housing system.

In the context of this paper, it is necessary to elaborate on the special housing policy in Chinese historic urban areas. As a result of historic conservation, historic quarters in China have more or less retained the old housing system of the planned economy. This consists of three kinds of properties. The first is private housing belonging to local families, who have been living here for hundreds of years. The second is State-owned Public Rental Housing (hereafter public housing), most of which is abandoned or confiscated buildings belonging to big private housing. The third is Danwei-owned Public Rental Housing (hereafter Danwei housing). However, all Danwei housing had been sold to employees with full ownership (not including land ownership) before 1998, when the policy of the in-kind housing subsidy was terminated nation-wide. As a legacy of the planned economy, these properties are highly disordered. The complexity and ambiguity of property rights boundaries became the major barrier to urban renewal, which was seldom considered in the formal conservation system [68,69].

To summarize, the conservation system in China is rooted in paternalism with a lack of social empowerment and community participation. In such an institutional environment, little effort has focused on the performance of top-down interventions, although plenty of “good” measures have been taken to protect indigenous people. Thus, this paper will provide an empirical study to evaluate the implementation performance of Chinese top-down conservation planning from a bottom-up perspective of residential mobility. The planning interventions mentioned above will be classified and integrated into a regression analysis of residential mobility.

3. Material and Methods

3.1. Research Framework

In the literature relating to residential mobility, there are two main types of theories. The first type are macroscopic theories that focus on the structural resource layout, social system, and economic boundary or advantages that “push-and-pull” population migration [70,71]. The second type are microscopic theories that focus on socio-demographic factors of individuals and families, including age, family size, family structure, education level, job changes, marriage status, and so forth, based on Rossi’s (1955) seminal work of “life circle theory” [72,73,74].

Among various models, Bourne’s decision model of residential relocation offered innovative ideas regarding the “threshold” between relocation behavior and relocation desire [75]. According to Bourne’s theory, the driving force of home relocation comes from internal and external “pressures”, such as changes in housing needs and the physical environment. When the sum of pressures exceeds a certain “threshold”, a subjective “relocation desire” will be generated that increases the possibility of one person/family to take real actions. Here we call it “relocation behavior”. However, there are certain barriers that stop relocation behavior from actually happening, such as the gap between high housing prices and low family income. The discrepancy between relocation behavior and relocation desire is not questioned today [47,76]. Exploring them separately is a basic approach to investigate the comprehensive impacts of external and internal factors regarding the non-perfect relationship between objective behavior and subjective desire [77].

Drawing on Bourne’s relocation decision model, we divide residential mobility into two parts: the “relocation behavior” that actually took place and a subjective “relocation desire”, influenced by both external pressures and internal pressures (Figure 1). In terms of relocation behavior, historical analysis of planning documents and comparative works of community demographics was conducted first in order to exclude top-down relocation projects from bottom-up relocation behavior. Then, semi-structured interviews were conducted with householders to determine the decisive factor (or factors) behind residents’ moving decision. Regarding relocation desire, linear regression models were built to explore the determinants of moving intention of different house owner groups, who are classified by the analysis result of the decisive factor of relocation behavior. Next, a logistic regression model was built to predict residential mobility trends and community population change.

3.2. Data and Methods

Data used in this paper were collected from visiting government departments and a social survey in the Ping-Jiang Historic Quarter. The Ping-Jiang Historic Quarter is located in the inner-city of Suzhou (Figure 2). It is among the most representative cases of Chinese conservation practice. The study area for this paper is one of the four communities in the Ping-Jiang Historic Quarter. It is called the homonymous “Historic Quarter Community” with clear administrative boundaries. Meanwhile, the community scope is also the original protected zone in the 1983 City Master Plan (Figure 3). As of 2017, the study area still had more than 7000 local residents living in an area of 43.1 ha. The effectiveness of conservation planning has yielded a continuous existing case that is traceable in its historic layering of social change.

3.2.1. In-Situ Analysis

In September 2015, we visited four government departments, including the Planning Bureau of Suzhou City, the Housing Bureau of Ping-Jiang, the Historic Quarter Community Center, and the Heritage Conservation and Environmental Improvement of Ping-Jiang Historic Quarter Corporation (hereafter Ping-Jiang Corporation). We asked for the provision of all official documents regarding urban planning and historic conservation in relation to the study area. All available documents were organized in chronological order and the evolution of demographic change was calculated. These data provide an overview of local authority actions, and details of the exact measures undertaken to retain the indigenous population during the past 30 years. Six types of conservation measures were identified, including: welfare housing policy and periodic housing repair, four relocation projects from 2001 to 2015, the improvement of infrastructure (water, electricity, drainage) in 2002, road paving before 2009, and sporadic construction of new public service facilities. These measures have been integrated into the design of the survey questionnaire and semi-structured interview.

3.2.2. Social survey

In September 2015 and May 2016, we interviewed 300 households in the study area and received 272 effective questionnaires. Respondents were chosen through systematic sampling according to their street and house numbers. The questionnaire consists of four parts (

Table 1. A sample of the questionnaire.

Basic Information
Gender	male	female
Hukou	local	non-local
Age	<18	18–25	25–40	40–65	>65
Employment	Retired	Employed	Unemployed	other
Income level * (RMB/month)	<1680	1680–2620	2620~4200	4200–8000	>8000
Family size (person)	1\2\3\4\5\6\7 … other numbers: ____;
Family structure	Live alone or with partner	Core family	Stem family	Joint family and other
Property ownership	Public housing	Private housing	Danwei housing	Rented and others
Housing area	_____m2
Length of residence (year)	<1	1–3	3–10	10–18	>18
Sense of community	1\2\3\4\5 (1 = weak, 5 = strong)
Neighborhood communication	1\2\3\4\5, (1 = none, high frequency)
Subjective Evaluation
Housing condition	1\2\3\4\5\6\7 (1 = very bad, 7 = very good)
Infrastructure	1\2\3\4\5\6\7 (1 = very bad, 7 = very good)
Road and traffic	1\2\3\4\5\6\7 (1 = very bad, 7 = very good)
Public facility	1\2\3\4\5\6\7 (1 = very bad, 7 = very good)
Tourism disturbance	1\2\3\4\5 (1 = high disturbance, 5 = none disturbance)
Expected rent	1\2\3\4\5 (1 = very high; 5 = very low)
Have tourism income?	yes	no
Relocation Desire
Do you want to move?	yes	no
How much do you want to move?	1\2\3\4\5 (1 = strong intention to move, 5 = not at all)
What matters in relocation decision?	Please list five most import factors1.________; 2. ________; 3. ________; 4. ________; 5. ________;
Comments on conservation plan and suggestions for community development?

* In Suzhou city in 2015, 1680 RMB/month is the minimum wage, 2620 RMB/month is the average retirement pension, 4200 RMB/month is the per capita income, and more than 8000 RMB/month is the top 15% income class.

). The first part concerns demographic information about the respondent and his or her family. The second part consists of structured self-ratings about “external pressures” (living condition factors and tourism development factors) and “internal pressures” (socio-demographic and emotional factors). Respondents were asked to rate each variable according to their subjective feelings and expectations. The third part is a semi-structured interview that asks respondents to list the major factors of the relocation decision in order of importance. The last part is a non-structured interview about respondents’ life experience, attitudes toward historic conservation, and suggestions for community development. All unstructured interviews were recorded by the interviewers.

The preliminary data showed (

Table 2. Length of residence and relocation desire.

		Householder		Family Members
		Number	Proportion	Number	Proportion
In total		272	100%	693	100%
Length of residence (year)	<3 *	28	10.3%	--	--
	3–18	39	14.3%	--	--
	>18	205	75.4%
Relocation desire (binary)	move	141	51.8%	487	70.3%
	Not move	131	48.2%	206	29.7%

* Non-local residents could apply for a local ID after they have been living here for at least 3 years.

) that 75.4% families had been living there for more than 18 years, before the termination of the “in-kind” housing subsidy in 1998. In this paper, these families are defined as “indigenous people”. At first glance, the local society was quite stable, as most residents rarely moved. However, 51.8% of house owners and 70.3% of family members expressed a strong desire to leave (the intentions of teenagers and infants are stated by their guardians). The community was at great risk of losing indigenous people.

3.2.3. Regression Analysis

In this article, we use both linear regression and logistic regression models. The purpose of linear regression analysis is to determine the most influential factors among all variables, which are determinant in increasing residents’ relocation desire among different householders. The logistic regression analysis focuses only on socio-demographic variables, in an attempt to determine the gap between different social groups and to predict social change trends; for instance, between low-income and middle-income families, old residents and young generations, local people and non-local migrants, and so forth.

In designing linear regression models, four groups of explanatory variables are proposed, including living condition-related attributes, tourism impact on local residents, socio-demographics, and emotional factors (

Table 3. Variables used in regression models and primary results.

Variables	Min/Max	Frequency/Coding	Sample (N = 272)
			Mean	s.d.
Relocation desire	1/5	1 = strong, 5 = no desire	3.02	1.61
Living condition factors
Housing area (m2)	7/300	--	56.43	47.03
Per capita housing area (m2)	3.5/60	--	18.54	11.52
Housing condition	1/7	1 = very bad, 7 = very good	3.25	1.99
Infrastructure	1/7	1 = very bad, 7 = very good	3.51	1.65
Road and traffic	1/7	1 = very bad, 7 = very good	3.51	1.69
Public facility	1/7	1 = very bad, 7 = very good	5.65	1.24
Tourism development factors
Tourism disturbance	1/5	1 = high disturbance, 5 = none	3.69	1.02
Expected rent	1/5	1 = very high, 5 = very low	2.47	1.05
Having tourism income or not	--	1 = yes, 0 = no	0.10	0.30
Socio-demographic factors
Gender	--	1 = male, 0 = female	0.42	0.49
Hukou	--	1 = local, 0 = non-local	0.85	0.36
Age
<40	--	11.8%	--	--
40~65	--	22.8%	--	--
>65	--	65.4%	--	--
Employment
Retired	--	65.4%	--	--
Employed	--	30.1%	--	--
Unemployed and other	--	4.4%	--	--
Income level (RMB/month)
<2620	--	38.2%	--	--
2620~4200	--	39.3%	--	--
>4200	--	22.4%	--	--
Family size (person)	1/25		3.55	2.52
Family structure
Live alone or with partner	--	35.3%	--	--
Core family	--	21.3%	--	--
Stem family	--	35.3%	--	--
Joint family and other	--	8.1%	--	--
Emotional factors
Length of residence (year)
<3	--	10.3%	--	--
3~18	--	14.3%	--	--
>18	--	75.4%	--	--
Sense of community	1/5	1 = weak, 5 = strong	3.46	1.30
Neighborhood communication	1/5	1 = none, 5 = high frequency	3.91	1.17

). The selection of variables draws on existing studies about residential mobility in urban China, as well as the conservation measures mentioned above: (1) living quality is the basis of residential satisfaction, in which housing conditions, the quality of infrastructure, and access to public service facilities are influential factors [78]; (2) tourism development usually has negative impacts on daily life [79], but the level of local economic dependence on tourism could change residents’ attitudes [80]; (3) among various socio-demographic groups, age, family size and structure, income level, and the Chinese Hukou system have significant influence on residential satisfaction [46,48]; and (4) the strength of the local social network and the sense of community and place attachment could contribute to residential satisfaction and therefore decrease relocation desire [81].

4. Results

4.1. Stable Population and Housing Policy

By looking into the planning history and demographic change, the study area had a stable population in past decades, but has seen an increasing number of relocation behaviors in recent years. From 1987 to 2015, four relocation projects were initiated in the study area, with 844 households relocated. In comparison, the official registered number of households decreased from 3564 in 1987 to 2693 in 2015, a change of 871. These two data points differ by only 27 households. Thus, top-down relocation projects were the only cause of population decrease.

Further exploring relocation behavior, 272 respondents were divided into indigenous people and non-indigenous people who had lived there for less than 18 years. Most respondents turned out to be indigenous people (75.4%) with an additional 9.6% who inherited the house from their parents or grandparents and lived there for less than 18 years. Of the remaining 15%, 28 respondents were short-term tenants who had lived there for less than three years, and 10 cases were long-term tenants who had lived there for more than three years, of which two cases had been living there for more than 18 years but without a local ID. The remaining three were new householders who purchased private housing from former householders. In total, only 14.3% of householders who were indigenous people, moved or passed away in the past 18 years, and their houses were sold or rented out instead of being used to house the inheriting generation.

Seeing the contradiction between stable population and high relocation desire, we undertook a semi-structured interview with householders. Respondents were asked to list major considerations of their home relocation decision in order of importance. Results showed the decisive effects of housing policy and property ownership that prevented residents from moving away. According to housing bureau documents, there were three kinds of property in the study area: public housing (63.0%), private housing (30.5%), and Danwei housing (6.5%). No commodity housing was found, as a result of strict conservation planning.

For the majority of public housing tenants, welfare housing is a vital social security. The portion of householders with a monthly income below the city’s average level was 87.3%. The monthly rent of public housing was 4.1 RMB/m² in 2017 (2.5 RMB from 2006 to 2015), which was only one fifth of the price of the same type of houses in the market. Additionally, the local government of Suzhou city also empowers tenants with the permanent and inheritable “using right” of public rental housing, in an attempt to improve their economic status and, meanwhile, to decrease their intention to move.

For private housing owners, their house represents not only economic value but also the spiritual inheritance of their family. Moreover, local law says private housing owners have “full ownership of the property” (not including land ownership). Private housing owners could sell, rent out or vacate their house without fear of confiscation. As a result, the relocation behavior of private housing residents ranked the lowest, demonstrating once again the decisive role of property ownership in increasing residential satisfaction for Chinese people [82,83]. In addition, due to the strict requirements on repairing historic houses and the transaction routine of full payment, there were only three cases of transactions of private housing.

For Danwei housing owners, the Danwei had been closed for decades. All Danwei housing was sold to employees with full ownership (not including land ownership) before 1998. However, as time passed, Danwei housing has fallen into disrepair and is at risk of collapse. Nearly 30% of the original families have moved away (incomplete statistics), leaving houses vacant or rented out to rural-urban migrants. Due to the small proportion of Danwei housing in the study area, however, its influence on the integral frequency of relocation behavior was limited.

4.2. Differentiated Relocation Desire and Population Displacement

4.2.1. Determinants of Relocation Desire

Before the analysis was conducted, all “scale” type variables were tested by the “reliability analysis” of SPSS (PASW Statistics 18, SPSS (Hong Kong) Ltd., Hong Kong, China). All variables passed the test with Cronbach’s Alpha index of 0.844. All explanatory variables and the dependent variable were then tested using “bilateral correlation analysis” to exclude variables with a correlation coefficient higher than 0.75. Finally, linear regression models were built with a five-rating relocation desire as the dependent variable. Samples were divided into three groups according to respondents’ property ownership. Four models all passed the model test, with significance lower than 0.001. The results of the non-standardized coefficient and corresponding significances are presented in

Table 4. Linear regression models of householders’ relocation desire.

Variables	Model1 Whole Sample	Model2 Public Housing Tenants	Model3 Private Housing Owners	Model4 Migrants and Other
Relocation desire (1 = strong)	Mean = 3.02	Mean = 2.75	Mean = 4.19	Mean = 2.42
Living condition factors
Housing area	0.004	0.010	−0.001	0.000
Per capita living area	0.001	−0.004	0.020	−0.029
Housing condition	0.431 ***	0.406 ***	0.300 **	0.638 **
Infrastructure	0.106 *	0.142 *	−0.036	0.033
Road and traffic	0.031	−0.006	0.058	0.013
Public facility	0.090	0.139	−0.009	−0.183
Tourism development factors
Tourism disturbance	0.040	−0.009	−0.027	0.147
Expected rent	0.168 **	0.113	0.152	0.160
Having tourism income or not	0.546 *	0.060	0.222	1.785 *
Socio-demographic factors
Gender	0.135	0.120	0.106	0.442
Hukou	−0.463	−0.479	0.007	0.674
Age	0.003	0.335	−0.644 *	−0.121
Employment	0.007	0.225	0.005	−0.553
Income level (RMB/month)	−0.067	0.091	0.236	−0.777 **
Family size (person)	−0.072	−0.307 *	0.097	−0.098
Family structure	−0.090	0.044	−0.013	0.188
Emotional factors
Length of residence	0.319	0.056	1.061 **	0.221
Sense of community	0.114	0.082	0.612 **	0.083
Neighborhood communication	−0.040	0.002	0.278	−0.150

Notes: *** significant at 0.001; ** significant at 0.01; * significant at 0.05.

.

For all respondents, poor condition of housing, old infrastructure, and expected rent income were positively related to their relocation desire, while the factor of tourism income/job opportunities decreased residents’ intention to move. However, among the 272 samples, only 26 respondents were working in the tourism industry, of which 18 were non-local migrants, and 8 were indigenous people. The effect of tourism development in providing job opportunities for local people has been very limited.

Looking into the relocation desire of different householders, rural–urban migrants and Danwei housing owners showed the strongest relocation desire, public housing tenants took second place, and most private housing owners were reluctant to move. For public housing tenants, the factors of poor condition of housing and old infrastructure were dominant. The share of public housing tenants who rated their housing condition below or equal to “not good” was 60.7%. Surprisingly, no public housing tenants had paid to repair their house, while more than 70% of private housing owners (incomplete statistics) had fixed the house at their own expense. Thus, the condition of private housing was much better than public housing in this case. Moreover, the length of residence, strong sense of community, and age factors also contributed to the low intention to move for private housing owners. Non-local migrants and Danwei housing owners were living in dilapidated houses, with the shortest length of residence and lowest sense of community. For rural-urban migrants, in particular, the only reasons to live there were the cheap rent and job opportunities in the central city. Most felt it was difficult to integrate into the local community since the majority of indigenous residents are elderly with strong social networks.

4.2.2. Emerging Population Displacement

As relocation desire is positively related to relocation behavior, this section focuses on the relocation desire of different social groups in order to indicate the future trends of residential mobility and population change. A binary logistic regression model was built by simplifying the previous five-rating relocation desire into: “move” and “not move”. The results showed the most significant factor was “family structure”, which was followed by “income level” (

Table 5. Binary logistic regression of relocation desire.

Explanatory Variables		N = All
	Frequency	B	Exp(B)
Gender
0 = female	58.50%	--	--
1 = male	41.50%	−0.017	0.983
Hukou
0 = non-local	14.70%	--	--
1 = local	85.30%	0.199	1.22
Age
1 ≤ 40	11.80%	0.253	1.288
2 = 40~65	22.80%	0.758	2.134
3 ≥ 65 (reference)	65.40%	--	--
Employment
0 = retired	65.40%	--	--
1 = employed	30.10%	1.285	3.616
2 = unemployed and other	4.40%	0.871	2.388
Income level (RMB/month) **
0 ≤ 2620	38.20%	--	--
1 = 2620~4200	39.30%	1.240 **	3.457
2 ≥ 4200	22.40%	0.474	1.606
Family size (person)	--	0.102	1.107
Family structure ***
0 = Live alone or with partner	35.30%	--	--
1 = core family	21.30%	0.507	1.66
2 = stem family	35.30%	2.003 **	7.408
3 = Joint family and others	8.10%	1.390 *	4.013

*** significant at 0.001; ** significant at 0.01; * significant at 0.05.

).

Among all types of family structures, residents who “live alone or with partners” accounted for one-third of all respondents. Among these one-person households and husband-and-wife families, 81.2% residents were above 65 years old. According to residents’ descriptions, their children had moved to new towns on the outskirts at the beginning of the new century due, in large part, to the population decentralization strategy. Moreover, younger core families, stem families, and joint families that still have children in their houses showed higher intentions to move. As seen in recent years, most inheritors chose to sell or rent out their parents’ old houses, while others simply leave the house vacant. Based on this sample, the stability of the population has been undermined by the significant problem of a loss of people from younger generations.

Regarding income level, richer residents, most of whom are public housing tenants, showed higher intentions to move. Specifically, 60.0% of public housing tenants, who had an income above 4200 RMB/month, indicated they would like to move. In comparison, only 9.5% of private housing owners at the same income level were willing to move. Accordingly, more public housing tenants have begun to remodel their houses and sublet public housing to rural-urban migrants for extra income, which is illegal according to local regulations on public housing and historic conservation. In 2014, 200 such cases were identified, resulting in fines being issued by the local security department. However, the low fines, in comparison to sublet income, did not stop more public housing tenants from moving away or illegally subletting public housing. As time passed, an increasing number of rich public housing tenants, who are indigenous people, have been replaced by poor rural-urban migrants. The retained indigenous population living in public housing has become older and poorer than private housing families (Figure 4 and Figure 5).

5. Conclusions and Discussion

After 30 years of market-led reform, the urban society in China has undergone tremendous changes. Previously low, the mobility of the urban population has increased with the progress of housing commodification and other market-led socio-economic changes. In the context of fast urban renewal and development, many historic areas retain the majority of indigenous people via strict historic conservation. Whether the local governments adopt mandatory residential relocation is considered decisive in retaining indigenous populations. Unlike mainstream research into government-led relocation projects, this paper focused on bottom-up residential relocation on the basis of the empirical case study of Ping-Jiang Historic Quarter. A socio-culturally prioritized conservation strategy has been adopted in this case, which allows further discussion on voluntary population loss and the performance of conservation measures, rather than simply characterizing Chinese local government as entrepreneurial or pro-growth. The major results of the analysis showed property ownership and welfare housing policy were effective in retaining indigenous people, while nearly half of all residents showed a strong intention to move due to poor housing conditions and an expanding rent gap between current use and tourism development. Substantial differences in relocation desire were found between public housing tenants, private housing owners, and rural-urban migrants. Private housing owners were reluctant to move, while public housing tenants and rural-urban migrants showed a strong relocation desire, affected by differing levels of housing conditions, individual/family income, sense of community, and day-to-day neighborhood communication. Driven by differentiated relocation desires of different social groups, there have been increasing trends of aging, poverty growth and population displacement, epitomized by the replacement of richer public housing tenants by poorer rural-urban migrants.

As indicated in this sample, the demographic structure is generally old and poor, which is in line with the findings for other historic residential areas, where houses are also dilapidated and the local population is stable, with a strong social network [38,39]. However, the major cause of poverty in Chinese historic quarters is not dilapidated housing conditions or cheap rent. The influx of poor migrants in recent years only accounts for a small portion of all poor people. In fact, long before housing reform and historic conservation, the majority of poor residents had already settled in public housing provided by local governments. In this sense, public housing is a repository for poor people, rather than the cause of problem per se [84], which is different from the situation in other historic areas in non-regulated housing markets, in which poverty has increased [85]. In terms of the aging process, the elderly revealed lower intentions to move than young residents, as seen in most studies [48]. However, in the present study, the elderly prefer to live close to their elderly neighbors in comparison to their children, which differs from the findings of existing studies [86]. In addition, the decentralization policy has accelerated the aging process to a large extent, which reveals the negative impact of historic city conservation on the social vitality of old downtown areas.

Given the enduring loss of young generations in this protected neighborhood, it is not hard to notice that young rural–urban migrants are a potential resource to reverse this trend. However, the migrant group has the strongest desire to move, which is opposite to existing findings [37]. In spite of poor living conditions, most migrants found it hard to be integrated into local society. As a legacy of the planned economy, the majority of elderly residents share similar life experiences that enhance the cohesion of local social networks, but only among indigenous people. With the lack of place attachment and sense of community, the residential satisfaction of rural-urban migrants in this sample is much lower than those in urban villages, unprotected old neighborhoods, and other migrant gathering places [35]. In this sense, before the introduction of any new classes or real estate projects aimed at housing mix, the promotion of social integration between old and new resident groups might be a pivotal issue.

As for public housing tenants, the permanent and inheritable usage rights given to them did not lessen their intention to move. On the contrary, an increasing number of original tenants, mostly rich, have remodeled their houses to create more rooms and sublet these rooms to rural–urban migrants, which is illegal and detrimental to the built heritage. This phenomenon has been found in other Chinese historic quarters [69]. However, under the pressure of historic conservation and public opinion, the local authorities, who are the true owners of public housing, have found it difficult to clear all illegal buildings [68]. This is because a number of truly poor tenants have also expanded kitchens and bathrooms illegally to improve their own living conditions. Behind this dilemma is an unclear system of property rights that provides room for speculation. Without clear regulations on the conditions to access and withdraw welfare housing, individuals naturally tend to maximize their own profit, which is evident as no public housing tenants paid to repair their houses. Consequently, unclear property rights and obligations lead to a “reverse monopoly” of public housing by original tenants, accompanied by the inherent problems of social inertia and path dependence in a planned and welfare system. As all conservation funds come from public finance and tourism income in situ, substantial improvement of all residents’ housing conditions will not transpire for a considerable period of time. Due to the lack of internal social vitality and active community participation to support conservation actions, top-down interventions can only slow, rather than reverse, the trend of socio-cultural decline.

The findings of this paper have several important implications for Chinese conservation planning. First, for improvement plans to be more consistent with residents’ demands, investment in indoor housing maintenance and infrastructure installation should be prioritized. Second, during the planning of public service facilities, more basic facilities, medical facilities and elderly care facilities are needed in historic areas, rather than in other normal residential blocks. In addition, preferential policies, such as cuts in tax or rents, should be provided for local shops, whose major clients are poor indigenous people. Third, the local government should also purchase or rent vacant houses from residents who are willing to move. After repair and reconstruction, these houses could be used as a new kind of social housing or commodity housing to attract young students, the creative class, and middle-income families. Finally, we hold the opinion that innovation in the housing system is not only a key to retain indigenous people, but also a catalyst to social participation. The local government could motivate residents, especially the majority of public housing tenants, to take part in heritage conservation and tourism management through home ownership and benefit sharing. A subsequent step would be to encourage those involved with tourism development to make the significance of the heritage site accessible to the whole neighborhood and, more importantly, to cultivate a spirit of collective responsibility to protect community assets. Over time, successful government–people cooperation and substantial tourism income will trigger a self-reinforcing process that generates trust and improves the experience and capability of community governance. Without doubt, the combination of government force and community force could provide more lasting vitality for the sustainable development and conservation of historic urban areas.

Although this article provides some insights into the mobility trends of indigenous people, more scientific inquiries are needed to improve the success rate of social sustainability targets, which constitute an increasingly important part of China’s historic conservation. It is recommended that further research could be undertaken in the following areas: (1) further research on different social groups, classifying social cohorts and interest groups in a more detailed and comprehensive way, and meanwhile identifying primary and secondary demands apart from residential satisfaction; (2) further research on the relationship between different social groups, as conflicting and common interests are of the highest importance in setting social development goals; (3) further research on multidisciplinary measures to alleviate conflicts and to highlight common interests as a catalyst to social participation and collective action; and (4) long-range studies and follow-ups are ultimately needed to adjust planning interventions to the ever-changing situations of social structures, social demands and residents’ attitudes toward historic conservation and tourism development. As for our study, cross-section data were used to illustrate the present situation of relocation behavior and relocation desire. However, residents’ intention to move is likely to change in view of the age structure of most households and the changing attitude of local government on public housing policy. The next task on our research agenda is to conduct a tracking study of the relocation behavior and desire of indigenous people, in an attempt to investigate to what extent and how the changes in determining factors could influence actual relocation behavior and subjective relocation desire.