3.1. Case Study Area
In Lanzhou, UVs are mainly distributed in the four municipal districts, that is, Chengguan, Anning, Qilihe and Xigu, forming a dual-nuclear cluster center, Chengguan-Anning, and several secondary centers. With rapid urbanization and integration of construction land in the urban area of Lanzhou since 1990, a total of 83 UVs are divided into four types needing to be transformed: transformation on the original site, transformation on the relocated site, partial transformation, and comprehensive transformation. Among these, transformation on the original site is the most common, followed by transformation on the relocated site. Yantan (YT), subordinated to the Chengguan District, is one of the places where the land value-added benefits increased the most in Lanzhou, and has a superior location. Before 1990, it was only an important agricultural area in the suburbs of Lanzhou City, with growing of vegetables and apples being its main industries. Subsequently, with the establishment of the Lanzhou High-Tech Industrial Park in 1992, the land of Yantan was quickly expropriated as urban construction land, resulting in a rapid increase in land values. Meanwhile, the rapid urbanization of Yantan also produced 15 UVs, including Beimiantan New Village (BNV) and Xiaoyantan Village (XV). When the regional renewal development and UVs transformation are placed on the agenda, these UVs may face the dilemma of land expropriation, house demolition, and resettlement compensation of indigenous villagers. This study selects BNV and XV as case studies because of the two villages’ unique characteristics. First, they are both located in the core area of Lanzhou, where land value-added benefits are huge. The distribution of land value-added benefits has attracted the attention of different stakeholders, including the local government, real estate developers, the floating population, and indigenous villagers. Second, the two villages have similar characteristics in terms of population distribution, family structure and occupational composition, which makes residential sustainability comparable. Third, both villages have detailed data that can be obtained, and thus, they are relatively complete transformation samples.
In China, BNV is a rare “UV” that was experimentally transformed by adopting the new rural construction model in the city center. Located to the northeast of Yanbei Street, Yantan (
Figure 1), the village covers an area of 51.89 hm
2. In around 2004, the BNV village committee assigned some members to Xi’an and other places to investigate the experience of new rural construction, especially to learn how to activate collective land assets in new rural construction. Then, village committee cadres entrusted the Gansu Urban Rural Planning and Design Institute with planning the first and second new villages and signed house building agreements with the villagers. By 2006, the villagers were under uniform supervision to build new villages on the original site. In 2007, faced with disputes over the lack of homesteads for the married women in the village, the village committee established a third new village for them. This village is mainly inhabited by married women and their families, who are still registered as permanent residents in the BNV, so it is also called the Women Village. In 2007, the community village committee was transformed into the community resident committee, administrated by the Yanbei Street Office. The new village promotes community economic cooperation, the joint stock institution, village-owned enterprises such as the Yantan Pipeline Market, the Beimiantan Heating Station and Beimiantan Commercial and Trade Co., Ltd. (Lanzhou, China). These enterprises not only create employment opportunities for villagers, but their demand for space also further expands the market, renting the first floors of villagers’ houses as warehouses. The transformation parcels were divided into three parts. After some of the village’s land parcels had been expropriated by local government, villagers used land expropriation compensation as a new rural construction fund. The second part comprised villagers’ homesteads, which were used as the uniform new rural construction parcels. The surplus parts were used as reserved collective land parcels to promote collective economy development and to provide a guarantee for the villagers’ future life. This transformation model is greatly reliant on the willingness of landless villagers, namely, that they wish to be resettled on the original site and acquire the same residential area as before the transformation. Nowadays, villagers rely on informal economies such as self-owned house rental and self-employed commerce to create considerable profits. This economic structure has undergone a transformation in marketization towards “making money by the rental economy”.
XV is a community under the jurisdiction of Yanbei Street, Yantan, Lanzhou City (
Figure 1), with an area of 46.83 hm
2. Although XV has the same transformation background as BNV, local villagers made a living in a completely different way from the villagers in BNV before capital entry. The village underwent two rounds of farmland expropriation in 2005 and 2010 due to government road construction demand. By 2014, the village had become a village without agricultural production, and villagers earned money by renting self-owned houses, as well as the village collective storefront rental market. Once the economic rationality of villagers was awakened [
48], they rebuilt and renovated their self-owned houses to expand the rentable house areas and obtain greater benefits. The Lanzhou Yantan secondhand goods market is also a typical example of the village collective outsourcing land to outside contractors. The village committee established a market on this land, and then rented the market storefront to merchants everywhere, from which they collected rent. The village is one of the earlier urbanized villages; meanwhile, it is the most difficult urbanization area in Yantan. The transformation faces difficulties because the interests of the different stakeholders with respect to demolition compensation are hard to balance. The village’s developer is Lanzhou Jinhang Hongyan Real Estate Development Co., Ltd.(Chengguan district, Lanzhou, China), and the construction of the resettlement area is divided into two phases. At present, the Heshunyuan residence zone, which is the phase I resettlement area, has been completed, and 902 villagers whose houses had been demolished have been resettled. The phase II project is still in progress.
3.2. Survey Design and Data Collection
This study examines residential sustainability in UV transformation through field research and factor analysis method. The sustainability scale design for the two models is an important research foundation. Although there is no flawless survey scale currently available for this topic, the related research provides references that could help us to advance this work [
49,
50,
51,
52,
53,
54,
55,
56,
57,
58,
59,
60,
61,
62,
63]. On the basis of relevant research and discussion, we propose a conceptual framework for the evaluation of residential sustainability based on housing conditions, the community environment, and livelihood. With respect to sub-indicators of housing conditions, we referred to relevant studies on the housing conditions of rural floating populations and indigenous villagers in UVs [
49,
50,
51,
52,
53,
54], supplementing the sustainability content with previously available indicator structures. As for sub-indicators of community environment and livelihood sustainability, we specify indicators on the basis of precedents related to the evaluation of satisfaction in UV transformation [
53,
54,
55,
56,
57,
58,
59,
60,
61,
62,
63]. A household questionnaire survey method was used, combined by further interviews with 13 villagers and 5 village committee cadres to gain in-depth understanding of the relevant background context. Beginning with a preliminary investigation in November 2018, we officially completed the formal research work in February and March 2019. It is a common phenomenon that many relatives with consanguinity and kinship live intensively in large families in China’s villages. Although large families underwent relocation, villagers disengaged from their original family and made up new families after UV transformation, still living in the same resettlement area. Against this background, in order to reduce the possibility of selected objects belonging to the same large family, we firstly obtained a list of household heads’ names with the help of village committees. Then, we conducted a household questionnaire survey and adopted an equidistant sampling method according to family address, and randomly selected households to ensure that there was an equal probability that every household could be selected. Meanwhile, we asked for the household heads’ names, and noted them on the list in case of repetition. The content and structure of the questionnaire included basic family situation and a residential sustainability survey of the villagers. After the transformation, BNV comprised 670 households, with a permanent population of 2280; XV comprised 634 households, with a permanent population of 2120. Therefore, 134 and 127 questionnaires, respectively, should be distributed in the two villages, in accordance with the standard ratio of 1:5. A total of 260 questionnaires were distributed and 252 were recovered, of which 240 questionnaires were valid, representing an overall effective recovery rate of 95.2%. The following is a detailed explanation for the residential sustainability scale indicators.
(1) Interpretation of indicators of housing condition sustainability
We selected five indicators to reflect the housing condition characteristics in this study, the specific weighting and assignment methods of each indicator are shown in
Table A1. (1) Housing ownership index (HOI). Property rights of houses and housing source are selected as its sub-indicators, wherein the property rights of houses define the owner’s right to possess, use, benefit and dispose of the property, with housing source also reflecting the property right relationship, to a certain extent. (2) Housing crowding index (HCI). We use per capita residential area (m
2) and housing structure to measure this. The division method of per capita living area refers to
GB 50180-93 standard of the Code for Planning and Design of Urban Residential Areas. The housing structure is measured based on the whole set rate of the building. Generally speaking, the higher the number of complete sets in newly built houses, the higher the quality of the houses. (3) Housing privacy Index (HPI). We adopted housing privacy and housing function to measure this index. In general, if the houses’ privacy is high and the houses only carry out a residential function, the villagers’ residence will be disturbed to a lesser degree. (4) Housing facilities index (HFI). On the basis of a method of measurement found in the literature [
54], we establish three sub-facility indexes to measure HFI. The overall level of HFI is the sum of following three sub-indices: basic facility index, intermediate facilities index and advanced facilities index (
Table A1). Notably, we updated facility items according to the difference in the housing conditions of the resettled communities. The 7 items belonging to the former index cover basic facilities (including tap water, power facilities, natural gas, heating, flush toilet, water heater, water supply and drainage facilities), and each item scores 1 point. Items 8–14 are intermediate facilities (including lighting, fire fighting, elevator, parking lot, garbage collection, gate guard and monitoring, broadband network), and each item scores 2 points. Items 15–21 are advanced facilities (including fitness and entertainment venues, cultural and sports facilities, educational facilities, community parks and squares, medical facilities, business services, landscape and sitting-out areas), and each item scores 3 points. Specifically, we listed the above 21 housing facility items in our questionnaire; respondents then selected the corresponding facilities possessed by their families, or with which they were equipped in the community. Afterwards, we calculated the total scores of each questionnaire based on the rules and the calculation method provided in
Table A1. All index values ranged from 0 to 1. The higher the number, the better the facility configuration. (5) Housing quality index (HQI). We measured HOI based on three sub-indicators: building quality, housing ventilation effects and lighting condition, and residential area. Then we asked the villagers their satisfaction with these three aspects. The degree of villagers’ satisfaction was expressed as a numerical value using a five-point Likert scale, the responses in which were classified as ‘not at all satisfied’, ‘slightly satisfied’, ‘moderately satisfied’, ‘very satisfied’ and ‘extremely satisfied’, with the corresponding values of 1, 2, 3, 4 and 5, respectively. The specific calculation method was as follows:
where: HQI is the sum of three sub-indicators, and its value ranges from 1 to 15. According to the equidistance method, housing quality is divided into three degrees: when HQI < 5, the quality of housing is “low quality”; when HQI ranges from 5 to 10, housing quality is ”medium quality”; when HQI > 10, housing quality is “advanced quality”. The method for assigning weights and classifying the hierarchical structure for indicators is based on the literature—the evaluation method of the socio-economic status of the floating population in UVs [
55].
(2) Interpretation of indicators of community environment sustainability and livelihood sustainability
The residential environment and livelihood sustainability are essential factors affecting living quality. We selected 19 sub-indicators from 6 dimensions evaluating community environment and livelihood sustainability, and their operational definitions are as follows (
Table A2). Community planning and environment includes property management (
X1), support for public facilities (
X2), landscape greening (
X3), community security
(X4), and places of public activity and facilities in the community (
X5). The location condition of resettlement areas not only has an important impact on villagers’ daily commute, but plays a role in potential market thrust in terms of stimulating tenants’ rental behavior. Therefore, we selected the convenience of public transport around (
X6), distance between the resettlement area and the city center (
X7), and commuting status (
X8) as sub-indicators. The economic situation after the transformation is not only directly related to villager livelihood, but also indirectly reflects their occupational class. We selected stability of income source (
X9), work intensity and tiredness (
X10), and income increase or decrease situation after transformation (X
11) as sub-indicators. As many scholars have stressed the importance of social networks and their impact on the lives of villagers [
57,
58,
63], we selected friendliness and support of community members (
X12), scope for making friends in the community (X
13), the degree of reservation in the acquaintance society (
X14), and the impact of social networks on livelihood sustainability (
X15) to measure neighborhood attachment. Social guarantee situation reflects social equity, and is one of the most important ways for villagers to deal with emergency situations. Thus, we used coverage level (
X16) and guarantee level (
X17) of the villagers’ social insurance to evaluate this. Given that the villagers have turned to non-agriculture households since UV transformation, we only considered urban insurances that villagers had already enjoyed (i.e., old-age insurance, medical insurance, unemployment insurance, employment injury insurance and other social insurances), and exclude rural insurances. Additionally, psychological willingness is also an important aspect for evaluating residential sustainability. We choose degree of satisfaction with resettlement compensation (
X18) and willingness to reside and enter old-age care in the resettlement area for a long period (
X19) as its sub-indicators. The degree of residential sustainability for each of the above factors was obtained through investigation, and is expressed in numerical form by a five-point Likert scale. Villagers’ subjective feelings or views in response to these issues were classified as ‘completely unsustainable’, ‘somewhat unsustainable’, ‘neither sustainable or unsustainable’, ‘completely sustainable’, ‘somewhat sustainable’, with corresponding values of 1, 2, 3, 4 and 5, respectively.
To avoid there being too many indicators, the dimension structure was merged using the factor analysis method. These main factors can represent most of the original factor information, and also show the structural dimension of residential sustainability in UVs. They are converted into values between 1 and 100 points based on a standard score, using main factor variance contribution rate as the weight. The composite score was subsequently calculated. New factors were transformed in the same way to obtain the specific scores for each dimension. The main expression equation of factor analysis is:
where:
Xi represents the original indicator variable;
Fm (
m <
i) represents the unobservable common factor, which synthesizes the original i indicator information; and ε represents the information part that is not included as a special factor.
Among the 240 valid questionnaires received from the two villages, males and females in BNV accounted for 63.33% and 36.67%, respectively (
Table 1), while they accounted for 45.32% and 54.68%, respectively, in XV. In terms of age, samples from both villages were dominated by middle-aged and elderly people. The educational level of villagers in both villages was low, mainly at the level of primary or junior middle school. Respondents were from the following several types of family. In BNV, 43.33% and 31.67% were from extended families and joint families, respectively. In XV, respondents were predominantly from nuclear families and extended families (28.33%, 40.00%). The size of the families tended towards miniaturization. As for occupational types, the villagers mainly worked as general staff, or had left the village for jobs and self-employed jobs. In addition, employees of township organs, organizations and departments also accounted for a relatively high proportion of the villagers in BNV, mainly belonging to the staff of village collective organizations.