Relocation preference and settlement: Lessons from the Poverty Alleviation Relocation Program in China

Shuai Zhou; Guangqing Chi; Huanguang Qiu; Zhen Lei; Erica A. H. Smithwick; Jiquan Chen

doi:10.1371/journal.pone.0309534

Abstract

We explored the linkages between socioeconomic and demographic factors, relocation preference, and settlement associated with China’s Poverty Alleviation Relocation Program. Using multivariate ordinal logistic regression, panel data modeling, and multilevel methods, we found that outdated infrastructure at places of origin, such as long distances to paved roads or elementary schools, increased the probability of relocation, and the presence of left-behind household members at the origin compromised re-settlement. This study sheds light on the community- and household-level factors that influence relocation preference and settlement, offering valuable insights for future research and informing the design and implementation of relocation projects.

Citation: Zhou S, Chi G, Qiu H, Lei Z, Smithwick EAH, Chen J (2024) Relocation preference and settlement: Lessons from the Poverty Alleviation Relocation Program in China. PLoS ONE 19(10): e0309534. https://doi.org/10.1371/journal.pone.0309534

Editor: Dingde Xu, Sichuan Agricultural University, CHINA

Received: January 18, 2023; Accepted: August 14, 2024; Published: October 4, 2024

Copyright: © 2024 Zhou et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data and code for reproducing this study is available at https://github.com/shuai-zhou/RelocationPaper.

Funding: This research was supported in part by the USDA National Institute of Food and Agriculture (#PEN04796, #PEN04965, Accession #7003407, Accession #7006628) to GC, The Eunice Kennedy Shriver National Institute of Child Health and Human Development (Award #P2C HD041025) to GC, and The National Science Foundation (Award #1313761) to JC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Relocation refers to a form of mobility characterized by the geographic displacement of individuals from one location to another. The term is often used interchangeably with migration, displacement, evacuation, and planned relocation [1, 2]. In recent decades, relocation has been employed as a strategy to facilitate infrastructure development [3, 4] and address challenges such as resource scarcity [5], poverty [6, 7], access to education and health care [8, 9], and climate-induced disasters [10, 11]. Development and problem solving—oriented relocations involving large populations are usually led by governments. One notable example of this is the relocation associated with the Three Gorges Dam—a government-led project in China designed to control floods and transform the environment—which resulted in the displacement of approximately six million people [12, 13].

Similar to migration, relocation can manifest in various forms. Depending on voluntariness and whether the movements occur across borders, relocation can be categorized as self-driven, forced, internal, and international relocation. Previous studies on relocation have primarily focused on self-driven relocation, particularly employment-related moves [14, 15]. They have identified various factors that influence willingness to relocate, including individual characteristics such as gender and marital status [16], place-level factors such as the similarity between places of origin and destination [17, 18], and the availability of financial incentives [19]. It is worthwhile to distinguish self-driven relocation and internal migration, two common geographical movements in human societies. Although both are forms of human mobility and share common drivers such as economic opportunities, improved living conditions, access to services, and social networks, they exhibit notable differences. Self-driven relocation involves individual agency and choices in the decision-making process, placing a strong emphasis on the active role of individuals in shaping their moves. In contrast, internal migration is triggered by a broader range of motivations, including economic, social, and environmental factors. Additionally, self-driven relocation is typically a voluntary action taken by individuals, reflecting their personal preferences and aspirations, while internal migration can encompass both voluntary and involuntary or forced movements that are influenced by factors such as conflict, environmental disasters, or government policies.

Other research has focused on government-led relocation projects, primarily aiming to evaluate the efficacy of such relocations and to assess their impacts on the involved populations. For example, HOPE VI and Moving to Opportunity are two well-known federal-led relocation projects in the United States. Studies of these two projects have generally suggested that relocation has had limited impacts on the participants’ financial, educational, and social well-being [6, 20, 21]. In recent decades, relocation has been increasingly driven by environmental disasters such as floods [22] and hurricanes [23, 24]. However, disaster-related relocation research often focuses on relocation and settlement separately. To the best of our knowledge, studies exploring the driving forces in relocation preference following government-led subsidized relocation projects, as well as settlement—the state of remaining residentially stable at destination communities—in one research design from socioeconomic and demographic perspectives are limited. The requisite data for a comprehensive analysis of relocation preference and settlement, including pre- and post-move records, are rarely available.

We attempt to fill this knowledge gap by systemically investigating the socioeconomic and demographic characteristics associated with relocation preference and settlement within the Poverty Alleviation Relocation Program (PARP), a part of China’s Targeted Poverty Alleviation Project (TPAP). The investigation intso the underlying determinants that shape relocation preferences and settlement patterns within the PARP offers a valuable opportunity to gain an understanding of the factors influencing individuals’ decisions to engage in government-led relocation projects and their subsequent settlement in the new communities. This research is of particular relevance in the context of global environmental change and escalating regional conflicts (e.g., the Ukrainian crisis), which are expected to cause a substantial increase in the frequency of relocations. Moreover, this research provides policymakers with critical insights for the effective formulation and implementation of relocation policies, thereby promoting the overall well-being of individuals involved in the PARP and offering guidance for similar relocation programs in the future.

First announced in 2014 and officially initiated in 2016, the TPAP aimed to lift about 70 million people out of extreme poverty by 2020. This involved the relocation of approximately ten million people from 1,400 targeted counties in 22 provinces that were characterized by high poverty rates and environmental vulnerability according to the PARP, the largest relocation project in recent history anywhere in the world in terms of the population involved [25, 26]. The first step of the PARP was the identification of impoverished populations. The government used an annual net individual income of 2,300 RMB (approximately 362 USD at the 2014 exchange rate) as the poverty threshold [25]. Populations earning below this threshold were registered, and their information was archived in the National Poverty Alleviation Information System for monitoring their poverty status based on the pre-defined threshold. One approach to relocation aimed to preserve the original social networks and support systems of the relocatees to enhance their participation. This involved resettling all the poor households from a village in a single location. However, it was not always feasible to accommodate an entire targeted population in one settlement, so some cohorts were dispersed across multiple settlements. During the relocation process, the government provided subsidies to targeted populations as incentives, usually in the form of government-built apartments (without property rights) proportional to household size and/or in-kind benefits in the relocation destination. Relocations usually happened within the same jurisdiction to reduce moving costs and social integration issues. Government officials visited targeted populations to disseminate detailed information about the PARP’s policies, aiming to increase their understanding and encourage participation. After relocation, participants were provided with community service work opportunities and training to bolster their employment prospects in the new community. Ultimately, households had the right to decide whether or not to relocate. In sum, PARP was a government-led and subsidized voluntary relocation program that enables us to examine the driving forces of the targeted population’s relocation preferences and settlement at the destination.

In the literature on relocation, prior studies have shown that factors such as community similarities, financial incentives, and demographic characteristics including age, gender, and income significantly influence relocation willingness. For example, Noe and Barber [17] found that people were more likely to relocate to communities that were similar to their original locations. Wagner and Westaby [18] found similar results, with cultural similarities between pre- and post-relocation communities being associated with higher levels of relocation willingness. They also found that financial incentives were the most important factor in explaining the differences in relocation willingness. Abraham, Bähr, and Trappmann [16] found that gender and marital status affected relocation willingness: married women had a lower willingness to relocate than married men; however, there was no difference in relocation willingness between single women and single men. Konopaske, Robie, and Ivancevich [27] also found that spouses’/significant others’ willingness to relocate played an important role in deciding whether or not to relocate. Lo and Wang [28] investigated the level of voluntary participation among TPAP participants and found a strong willingness for relocation, particularly among younger individuals, those with higher wealth, and those with off-farm employment opportunities.

In recent decades, environmental disasters such as hurricanes and floods have increasingly influenced relocation or evacuation. Studies have identified a bundle of socioeconomic and demographic factors that influence disaster-affected populations’ relocation or evacuation practices. For example, Bukvic and Owen [23] surveyed 46 households from highly affected coastal communities five months after Hurricane Sandy and explored their attitudes toward relocation. Their results suggested that respondents were aware of climate-related coastal risks such as hurricanes and sea-level rise and were willing to relocate to other places, mainly for health and safety reasons. In a relocation study from disaster-threatened areas in rural China, Xu et al. [29] found that sense of place significantly decreased relocation willingness, while risk perception significantly increased the probability of relocation. Not all disaster-threatened populations are relocatable. For instance, Mavhura et al. [22] conducted interviews with residents in a flood-prone community in Zimbabwe and found that, although relocation was considered a viable strategy to cope with floods, the lack of compensation for their lost land and livelihoods, along with the inadequate basic infrastructure in relocation sites such as schools and clinics, posed significant barriers to the relocation of the flood-affected population. Thiede and Brown [24] examined relocation behavior prior to Hurricane Katrina and found that Black and less-educated populations were least likely to relocate due to a lack of necessary resources such as transportation and financial means to initiate relocation.

Like migration, relocations are often coupled with counterstream return flows that undermine settlement. As mentioned earlier, settlement generally refers to the state of remaining residentially stable at a relocation destination. Previous studies have found two main mechanisms through which settlement can be undermined in government-led relocation projects. First, insufficient support at the relocation destination can destabilize government-led relocation. In an early study of the Employment Transfer Scheme in Scotland, a government-led labor-mobility project, Beaumont [30] found that 24.8% of the sampled population returned. The primary reasons for returns were insecure financial situations and the unavailability of satisfactory housing in the destination community. Second, the failure to achieve relocation goals may affect a settlement. In a comparative analysis of the HOPE VI and the Moving to Opportunity programs, Goetz [6] found limited evidence to support the notion that relocation positively impacted participants’ living conditions, neighborhood satisfaction, and employment security. It was also suggested that these limited improvements can potentially have an indirect destabilizing effect on the overall relocation process. Regarding disaster relocation, Groen and Polivka [31] found that returns vary across demographic groups, with Blacks, young adults, and single individuals being less likely to return to their place of origin after Hurricane Katrina, likely because their residences and neighborhoods experienced greater physical damage than those of other demographic groups.

Conceptual framework

In essence, relocation is a geographic movement of populations, making it analogous to migration. Previous studies have theoretically framed and empirically investigated relocation from a migration perspective [1, 2]. We followed this research convention and conceptualized relocation as a migratory response to unfavorable living and environmental conditions. Classical migration theory attributes initial, repeat, and return migrations to different driving factors at places of origin and destinations and a series of obstacles such as distance between the two ends [32, 33]. In the twentieth century, the theoretical investigation of migration from economic perspectives gained popularity, with neoclassical economics (NE) and the new economics of labor migration (NELM) as the two main theoretical frameworks. NE theory suggests that the markets at both the sending and the destination community are well-functioning but in different stages of development; migrants flow to geographically advantaged and well-developed labor markets for higher wages to maximize their incomes and well-being [34]. NE attributes return migration to financial failure at the destination. In summary, NE considers migration to be an income- or utility-maximizing behavior [35]. In response to such narrow foci, NELM contends that migration is driven by the interest of not only increasing income but also minimizing risk through diversifying sources of income [36, 37]. Massey and Espinosa [34] espoused that migration occurs not only to seek higher earnings but also to avoid market failure at places of origin and to accumulate human capital at places of destination. In a comprehensive examination of theoretical perspectives on migration and development, De Haas [38] also emphasized the significance of social and migrant networks in facilitating additional migration flows. Empirical studies have consistently demonstrated the influential role of these networks on migration dynamics both at the global level [39] and within specific regions, such as the migrations from Mexico to the United States [40].

While the aforementioned theoretical frameworks focused on either structural factors or the agency of actors, they neglected the interrelationship and feedback mechanisms between them [41–43]. The migration system theory was developed to incorporate structural factors and actors’ agencies within the migration process [44]. Specifically, the migration system theory takes a systematic approach and uses all the elements from the places of origin, destinations, and origin—destination combinations to study migration decision-making instead of separating places of origin and destination [45]. In doing so, the migration system theory bridges the micro—macro gap in the migration process by filling it up with the interactions and feedback mechanisms between individual actors and the broader environment [46]. For example, information about migrants’ situations at the destination can be transmitted to the place of origin, affecting further migrations, depending on whether the feedback is positive or negative [41].

Building on De Haas’s [38] theoretical approach, we developed a conceptual framework (Fig 1) that delineates the relationships and feedback mechanisms among various factors within the relocation—settlement system. Household characteristics, landholding, infrastructure, and social networks at both origin and destination locations individually influenced the relocation and settlement decision-making of the targeted population in the PARP project. Simultaneously, the act of relocation itself impacted the cohesion of households at both places of origin and destination, subsequently influencing the settlement outcomes of the relocatees.

Download:

Fig 1. The conceptual framework for exploring relocation preference and settlement in the Poverty Alleviation Relocation Program (PARP) in China.

Solid arrows show the direct impacts of covariates on relocation and settlement, respectively, while dashed arrows suggest a feedback mechanism that affects relocation and its subsequent settlement.

https://doi.org/10.1371/journal.pone.0309534.g001

Hypotheses

Drawing on theoretical and empirical evidence and our conceptual framework, we developed three specific hypotheses regarding the factors associated with relocation preference and settlement.

Hypothesis 1: Basic infrastructure at the place of origin affects relocation preference. Specifically, the distance to transportation and educational infrastructure increases relocation preference. The classic push—pull theory and the migration system approach emphasize the effect of community characteristics in deciding where to move [32, 41]. This implies that the basic infrastructure of the community is one of the major concerns that affect migrants’ relocation preference—people prefer places with higher transportation accessibility, more educational resources, and other easily accessible advantages. As such, we expected a positive association between relocation preference and the distance to basic infrastructure such as paved roads and schools.
Hypothesis 2: Landholding size negatively affects relocation preference; the more the landholding size, the less likely the household will relocate to a new settlement. In the Chinese context, landholding size has been found to be related to people’s movement [47]. Agriculture has a long-standing history in China, and it remains a crucial livelihood strategy for rural households. Research indicates that larger landholdings are associated with a reduced likelihood of rural households migrating to urban areas in China [48]. Despite the recent increase in rural—urban migration and the diversification of income sources through off-farm activities, challenges persist because of the existing population registration system, which prevents rural residents from obtaining urban citizenship and equal access to social welfare benefits [49, 50]. Consequently, the connection between rural residents and their farmland may remain strong because farmland and associated agricultural earnings provide a degree of financial stability. Therefore, we expected that landholding size would decrease relocation preference.
Hypothesis 3: Relocation completeness plays an important role in determining settlement at the destinations. Compared with those who have relocated their entire household, households with left-behind members are more likely to visit their places of origin after the relocation. As for the counterstream of migratory flow embedded in the migration system, return migration can be attributed to similar micro- and meso-level characteristics as the initial migration. However, in this specific relocation project, the importance of left-behind members is particularly emphasized in determining settlement. The PARP encourages the movement of the entire household. In reality, both internal and international migration sometimes start with young adults migrating first, then bringing children and the elderly, eventually finishing with a reunion of the household at the place of destination [51, 52]. Similar patterns were observed in the PARP: fieldwork observations during the survey indicated that young couples and their children were the first to move, while some elderly populations were temporarily left behind at their places of origin. The separation of household members and the interruption of family ties may lead to relocated households temporarily revisiting their places of origin. Consequently, we hypothesized that households with left-behind household members at their places of origin would experience greater instability following relocations.

Materials and methods

Data

We conducted three waves of surveys (2016, 2017, and 2019). The 2016 survey examined the driving forces of relocation willingness because relocation had not yet commenced in that year. In the 2017 and 2019 surveys, given that relocation had started and some households had arrived at their destinations, our objective was to investigate the driving forces behind settlement after relocation. Using a multistage sampling strategy, we selected our baseline survey sample from the National Poverty Alleviation Information System in 2016, the official poverty dataset maintained by the central government of China. First, we purposely selected eight provinces (Gansu, Shaanxi, Sichuan, Yunnan, Guizhou, Guangxi, Hunan, and Hubei) from the contiguous destitute areas (CDAs) [53] where poverty is clustered. These selected provinces cover the most southwestern parts of the CDAs (see Fig 2). Second, we randomly selected two government-identified poor counties from each of the eight provinces. Finally, we randomly selected 144 households from the selected counties. This led to a total of 2,304 households being selected for participation. These 2,304 households were interviewed using a household survey (see S1 File for an excerpt of the survey instruments). We used an additional questionnaire to collect contextual information at the community level, which was usually filled out by village committee members most familiar with their communities. We hired trained college and graduate students to assist with the face-to-face data collection.

Download:

Fig 2. Distribution of the contiguous destitute areas (CDAs), including the eight sampled provinces and 16 sampled counties.

The data for this figure comes from Natural Earth. This figure is for illustrative purposes only.

https://doi.org/10.1371/journal.pone.0309534.g002

We collected 2,185 responses from the 2016 wave (95% of the selected households). In the second wave of the survey in 2017, we tracked down 1,898 households (about 87% of the households surveyed in 2016). Among those 1,898 households, only 475 households (25%) had relocated, likely because the relocation program was in its infancy in 2017. In the third wave of the survey in 2019, using both face-to-face and phone interviews, we tracked down 2,034 of the households surveyed in 2016, among which 1,147 had relocated (56%). We used the first wave of the study for the analysis of relocation preference and combined the second and third waves of the study to explore the driving forces of the relocatees’ settlement. Our questionnaire and survey procedures for data collection from the sampled households were approved by the ethics committee at Renmin University of China (Grant #71861147002). Prior to the interviews, informed oral consent was obtained from each participant. Trained interviewers ensured that all participants were fully informed about the study’s objectives and the protection of their privacy, confidentiality, and anonymity. To safeguard participant confidentiality, all personal information that could potentially identify individuals was anonymized to prevent any traceability.

After data cleaning, we got 2,146 valid samples for 2016 and 1,662 for 2017 and 2019 in combination. It is important to note that some missing data were identified in our samples, particularly pertaining to household head information and household income. Specifically, in the 2016 samples, eight households (0.37%) had missing data on household head information and five households (0.23%) had missing data on household income. In the 2019 samples, 14 households (1.22%) had missing data on household head information and four households (0.35%) had missing data on household income. Missing values for household heads usually happened when a household consisted of a widow and her children and no household head was reported. In those cases, we designated the widow as the household head and input household head information accordingly. If the widow was considered too old to make decisions for the household (age > 60), we designated the oldest child as the household head and filled out the missing information accordingly. For missing data on household income, we used unconditional mean imputation to fill in the missing information. For instance, if the household did not report its household income, the mean household income in the village to which the household belonged was used instead. The mean-based imputation of less than 1% of the samples did not significantly affect the results, especially when considering that these households are government-certified poor households from homogeneous areas who share many common traits, including financial situations. Another thing worth noting is that the research design did not incorporate weighting; consequently, we did not apply weight to the analysis.

Variables

Dependent variables.

Our dependent variables are relocation preference and settlement. We defined relocation preference as the likelihood of relocating to another place. In the survey we conducted in 2016, we asked the question, “How much would you like to relocate to the new settlement?” As the targeted populations might not have information about their exact relocation sites at this stage, the question aimed to assess their general willingness towards relocation, despite the lack of specific details about the relocation sites. Answers were coded from 1 to 5, where 1 represents “not likely” to relocate and 5 represents “very likely” to relocate. We used the answers to this question to denote relocation preferences.

In the relocation context, settlement often means residential stability at the relocation destination. In this study, we used the self-reported frequency of visiting places of origin to measure settlement. In the post-relocation survey in 2019, we asked “How often did you visit the places of origin after your relocation?” These answers were coded from 1 to 3, where 1 represents “never,” 2 represents “sometimes,” and 3 represents “often.” Answers to this question were used to represent settlement status. An important note is that while we used “visit” and “return” interchangeably, neither a visit nor a temporary return to a place of origin is return migration, and it should not be interpreted as such in this context.

Independent variables.

Our independent variables are household socioeconomic and demographic factors and community characteristics. The measurements of some of the independent variables such as age, gender, marital status, education, and household size are straightforward. Others are more conceptual and/or constructed through combinations of a set of variables that need additional explanation.

Household income. We calculated household income by subtracting household expenditures from cash income and market values of non-cash assets (e.g., agricultural and forestry products waiting to be sold).

Apartment satisfaction. We constructed apartment satisfaction by adding up 12 5-point scale indicators representing relocatees’ attitudes toward their destination government-built apartments and basic facilities such as apartment location and quality, water, electricity, road infrastructure, and recreational facilities. A larger number indicates more satisfaction with the apartment at the destination.

Social networks. Social networks, coupled with the social capital inherent in those social relationships, can manifest through shared values, trust, and cooperation [54]. Social networks have been shown to have an impact on population migration [55]. In a recent study investigating the association between farm size and social network formation among rice producers in China, Simpson [56] utilized the concept of friendship as a measure of social networks. We adopted a similar approach and defined close friends as individuals to whom participants expressed a willingness to lend more than 5,000 RMB (equivalent to approximately 724 USD at the 2016 and 2019 exchange rates), which is considered a threshold indicative of close relationships within the Chinese cultural context. The count of close friends was subsequently employed as a proxy to represent the social networks of the targeted relocatees. It is important to acknowledge that while this approach respects the cultural dimension of social networks in the Chinese context, it may not comprehensively capture the entirety of social networks as a broader conceptual construct.

Modeling

We used the 2016 survey data to answer the research question concerning relocation willingness and combined the 2017 and 2019 survey data to explore the driving forces of settlement. Accordingly, we employed different modeling approaches for the relocation willingness model and the settlement model.

In the relocation model, we treated the dependent variables as ordinal variables and applied the multivariate ordinal logistic model (OLM) to the data we collected in 2016. The ordinal logistic regression model can be expressed as follows [57–59]: (1) where P(Y ≤ j│X = x) is the cumulative probability that the dependent variable Y is in category j or a category less than j; α_j is the threshold for category j; β₁, β₂, …, β_k are the coefficients for the independent variables x₁, x₂, …, x_k; and ln is the natural logarithm.

As previously mentioned, the survey data were collected hierarchically, with households nested within counties, and counties nested within provinces. To address the hierarchical structure and clustering effect in the data, we employed a three-level (household-, county-, and province-level) multilevel model (MLM) for the relocation model. The MLM follows the general form as follows [60]: (2) where Y_ijk is the settlement for the ith household in the jth county within the kth province; γ₀₀₀ is the overall intercept, representing the expected value of Y when all predictors are at 0; X_ijk represents a vector of covariates; μ_0jk and μ_00k are the random intercept for level 2 (county) and level 3 (province), capturing the deviation of each level 2 and level 3 unit’s average outcome from the overall average, respectively; and ϵ_ijk is the level 1 (household) residual term, representing the deviation of household observations from their expected value based on the level 2 and level 3 group averages. It should be noted that in the relocation model, the covariates in Eqs 1 and 2 include household head’s characteristics and household’s characteristics such as its economic conditions, characteristics related to infrastructure accessibility, social network at the origin community, and household’s familiarity with the PARP project.

For the settlement model, we employed a similar approach as in the relocation model, using OLM and MLM to fit the data. Moreover, as these modeling techniques do not explicitly consider the panel features of the settlement model, which comprises two waves of surveys, we combined the 2017 and 2019 survey data, creating a panel dataset. Subsequently, we applied fixed-effects (FE) models to explore the driving forces behind the settlement patterns of relocatees. The panel data fixed-effects model includes all the explanatory variables in the earlier analyses that are time-varying between the survey years 2017 and 2019, including household income, household head age, household size, left-behind family members, friends at the destination and origin, and apartment satisfaction. For the explanatory variables in the earlier analyses that are time-invariant (such as household head’s gender, education, marital status, and ethnicity, and certain household characteristics such as livestock values, landholding size, relocation preference, and commuting time between the origin and destination communities), we created interaction terms between these variables with a dummy variable for the 2019 survey year and included the interaction terms in the settlement model. Including those interaction terms provides two advantages: it allows time-invariant variables to be included in the analysis, as fixed-effects models typically drop them; and more importantly, it enables us to estimate how the impact of these time-invariant variables on settlement changed in 2019 relative to in 2017. For instance, a household with a married couple could be more likely to settle at the destination the longer they have been relocated. The panel models follow the general form as follows: (3) where Y_it represents settlement for household i at time t, α is the intercept, X_it is a vector of covariates, μ_i is the household-fixed effect, v_t is the time-fixed effect, and ε_it is the error term. We also employed FE models with county- and province-clustered standard errors, accounting for potential correlation in error terms at the county and province levels, respectively. As mentioned above, the covariates in the settlement model in Eq 3 include those time-varying explanatory variables and the interaction terms of those time-invariant explanatory variables and a dummy variable for the survey year 2019.

As stated in the hypotheses, our particular interest in the relocation model is to examine the effects of basic infrastructure accessibility and landholding size on relocation willingness. These two factors play important roles in influencing the households’ livelihoods, well-being, and relocation decision-making processes, and can therefore significantly impact relocation decisions. In the settlement model, we focused on assessing the influence of relocation completeness on the settlement of relocated households at the destination. Relocation completeness refers to whether the household completely moved all its members. Specifically, we measured relocation completeness using the presence of left-behind household members at the origin. The completeness of relocation determines the integrity of the household, therefore influencing resettlement at the destination. Household income and livestock values were log-transformed before being included in the models. This transformation was necessary due to the large standard variations and the potential presence of outliers in the original measures. Since China is regionally diverse [61], we included county dummy variables both in the relocation preference and settlement models to control for hard-to-measure characteristics such as cultural differences and other unobserved heterogeneities. Note that we used relocation preference differently in the two models. In the relocation preference model, relocation preference was treated as the outcome variable, while in the settlement model, we treated relocation preference as a control variable to reveal the net effects of the key factors in question. The objectives were to explore factors that influence relocation preference and settlement in the subsidized relocation program, not to examine the effect of the government subsidy itself on the relocation process. The analyses were conducted using Stata 16.1.

Results

Descriptive statistics

In 2016 a significant proportion of households (75.68%) showed a high likelihood of relocating, while less than 1% expressed a disinclination to voluntarily leave their place of origin (Table 1). The sampled households were predominantly headed by males (93.01%) in their 50s, with a majority being married (78.80%) and having attained less than a middle school education (70.60%). The largest ethnic group in China—the Han people—made up the majority of household heads (69.90%). On average, households comprise approximately four individuals. The log-transformed average household income and livestock values were 6.33 and 1.52, with the original values before transformation being 1,313.24 and 257.33, respectively. The average landholding size was 0.31 hectares. In terms of basic infrastructure, most households had access to electricity (94.36%), while a smaller proportion had access to running water (52.84%). The average distance to the nearest paved road and local market was 2.18 km and 10.78 km, respectively. As for educational facilities, the average distance was 7.15 km to the nearest elementary school, 16.00 km to the nearest middle school, and 54.25 km to the nearest high school. On average, households reported having 27.77 friends and had been visited approximately 4.47 times by government officials regarding the relocation policy at the time of the 2016 survey.

Download:

Table 1. Descriptive statistics for variables used in the relocation preference and settlement models in the Poverty Alleviation Relocation Program (PARP), 2016–2019.

https://doi.org/10.1371/journal.pone.0309534.t001

Regarding settlement patterns, a substantial proportion (78.32% in 2017 and 66.18% in 2019) of relocated households reported always or sometimes visiting their place of origin after the relocation. The primary reasons for their visits were farming (57.73%), socializing (26.28%), herding (7.66%), and other reasons (6.74%), with a minimal percentage (1.59%) using former residence in their place of origin as a temporary residence while visiting. The characteristics of the households, their economic situations, and their social networks remained similar for relocated households in 2017 and 2019 compared with their status prior to relocation in 2016. The only notable difference was observed in landholding size. At places of origin, the average landholding size was 0.48 hectares in 2017 and 0.58 hectares in 2019, while at relocation destinations, the average landholding size was 0.01 hectares in 2017 and 0.04 hectares in 2019. Among the relocated households, 9.47% in 2017 and 5.84% in 2019 had left-behind household members at their place of origin. The average satisfaction with apartments in the destination community was 51.49 in 2017, which decreased to 48.05 in 2019. Typically, the relocation destinations were within walking distance of approximately 48.47 and 55.94 minutes from their places of origin in 2017 and 2019, respectively.

Figs 3 and 4 present the correlation matrices of the covariates in the relocation preference and settlement models. Except for the correlation between the distance to the local market and elementary school, which is 0.5, all other correlations show small magnitudes. These findings indicate that there is no significant multicollinearity issue among the covariates in the models.

Download:

Fig 3. Correlation matrix of the covariates in the relocation preference model in the Poverty Alleviation Relocation Program (PARP) in China, 2016.

https://doi.org/10.1371/journal.pone.0309534.g003

Download:

Fig 4. Correlation matrix of the covariates in the settlement model in the Poverty Alleviation Relocation Program (PARP) in China, 2017–2019.

https://doi.org/10.1371/journal.pone.0309534.g004

Relocation preference model

We employed the stepwise estimation approach by adding household head demographic, household socioeconomic factors, and community characteristics in the models at each step using OLM and three-level MLM (Table 2).

Download:

Table 2. Relocation preference model in the Poverty Alleviation Relocation Program (PARP) in 2016.

https://doi.org/10.1371/journal.pone.0309534.t002

Models 1 and 2 include only the household head demographics and household characteristics, respectively. We found that household head demographics and household characteristics were not significant predictors of relocation preference at the household level. Model 3 is the full model with all covariates. The results showed that marital status and basic infrastructure were positively associated with relocation preference. On the one hand, married couples are more likely to relocate, a finding that may be explained by life course considerations that suggest married couples are more likely to seek out places with promising economic prospects [62, 63]. On the other hand, basic infrastructure, particularly longer distances to paved roads and elementary schools at the place of origin, increased the probability of relocation. This echoes the classic push—pull factors that influence migration decision-making, as adverse living conditions in a place of origin can trigger out-migration. Landholding size decreased relocation preference, however, such effects were insignificant.

Models 4 through 6 present the results from the multilevel model, which are generally consistent with the findings from the full model using OLM. This indicates the robustness of our findings. In sum, our results provide evidence supporting Hypothesis 1—that basic infrastructure increases the probability of relocation. Nevertheless, no strong evidence supporting Hypothesis 2—that landholding size at places of origin will negatively affect relocation preference.

Settlement model

We first employed OLM and MLM to analyze the settlement patterns of relocatees after their relocation. The OLM considers the ordered nature of the dependent variable, and the adoption of MLM was motivated by the need to address clustering effects arising from the hierarchical structure of the dataset (Table 3).

Download:

Table 3. Settlement model in the Poverty Alleviation Relocation Program (PARP), 2017–2019.

https://doi.org/10.1371/journal.pone.0309534.t003

Regardless of the modeling approaches and specifications, we consistently found that households with left-behind members were more likely to revisit their place of origin compared with households without left-behind members. Furthermore, the results also revealed that higher livestock values at the place of origin were associated with an increased frequency of relocatees’ visits. This evidence supports Hypothesis 3—that the completeness of the relocation process is an important factor affecting the settlement of relocatees at their destination.

To leverage the panel data structure while acknowledging potential correlation within geographical units, we then estimated panel data FE models with county- and province-clustered standard errors (Table 4). Consistent with results in Table 3, left-behind members play a crucial role in settlement (Models 10 through 12), with households with left-behind members being more likely to temporarily visit their places of origin after relocation. However, another variable of interest, livestock values, became less important in 2019 than in 2017, as reflected in the insignificant results of the variable. Collectively, these findings further support Hypothesis 3, that the presence of left-behind members was a primary concern for relocated households, significantly influencing their settlement and prompting relocatees to temporarily return to care for the left-behind members. The consistent results regarding the directions and significances of the interaction terms suggested robustness, irrespective of the various clustering options employed in the FE models. In addition, we found that the coefficient of the interaction term between household head’s marital status and survey year 2019 is significantly positive, indicating that a household with a married couple is more likely to settle at destination in 2019 than in 2017.

Download:

Table 4. Settlement model using panel data fixed-effects regression with different clustering options in the Poverty Alleviation Relocation Program (PARP), 2017–2019.

https://doi.org/10.1371/journal.pone.0309534.t004

Discussion and conclusions

Using data collected before and during the PARP in China, currently the largest contemporary government-led relocation project worldwide, this study explored factors associated with people’s attitudes and preferences about subsidized relocation and settlement at the relocation destinations. The results revealed that while the majority of relocatees demonstrated a high level of relocation willingness, there was variation in the levels of relocation preference and settlement among them. These differences were influenced by community-level characteristics such as distance to basic infrastructure, as well as household-level factors such as relocation completeness status. Overall, we provided empirical evidence supporting the notion that inadequate basic infrastructure serves as a push factor for individuals to leave their original locations within the PARP, while incomplete relocation and the resulting disrupted family ties act as a pull factor, leading to a return to their place of origin and undermining the process of settlement within the destination community.

Our results support previous theoretical and empirical understandings of relocation while revealing context-specific dynamics of relocation and settlement within the PARP. The high level of relocation preference observed in the PARP aligns well with previous research demonstrating the positive influence of financial incentives on relocation participation [19]. Our findings also highlight the significant impact of community-level factors, particularly the distance to the nearest paved road, on relocation preference. This finding aligns with the classic push—pull theory and is consistent with previous empirical evidence that limited transportation accessibility encourages outward migration [64].

Contrary to Hypothesis 2, we did not find strong evidence supporting the hypothesized relationship between relocation preference and landholding size in the sending communities. This finding suggests that in a rapidly developing world with increasing income sources and diverse livelihood strategies such as off-farm employment [65], farmland does not necessarily serve as a strong anchor tying individuals to their place of origin. Moreover, our findings indicate that the completeness of the relocation process significantly influences settlement outcomes. This underscores the significance of strong family ties in Chinese society, emphasizing the integral nature of familial relationships [66]. It further highlights that the stability of settlement may be compromised when these ties are disrupted, as observed in the PARP, where household members were separated between their places of origin and relocation destinations.

Our analyses have limitations. First, relocation disrupts the sense of attachment, familiarity, and identity of the relocatees, potentially decreasing the study participation rate among certain demographic groups; previous studies have shown that elderly people and households with children are less likely to relocate [27, 67]. However, the PARP witnessed a high rate of relocation preference (75.68% of the sampled households in 2016 reported they were very likely to relocate), likely because of the universally available government subsidies and the strong desire to move out of poverty-stricken and environmentally vulnerable regions. This can possibly bias the estimates of the relocation preference model upward.

An additional factor that may have influenced the results and conclusions is the absence of control for relocation distance. Previous research has established that relocation distance plays a crucial role in shaping relocation decision-making [33, 45]. The reason for this omission is that, during the initial stages of the relocation project, a considerable portion of the targeted population was unaware of their future relocation destination. Second, our analyses did not consider any selection bias that can potentially incur endogeneity, especially when considering the settlement for relocated households. During the three-year period of this study, 60% of the sampled households moved to the new settlements. There might be some characteristics—for example, their abilities and aspirations to move—that selected them into the relocation process and affected their settlement. Without considering selection bias and its effects, the interpretations of the results are confined to associations rather than causal relationships.

Third, the measure of the social network using the number of close friends may be regarded as a simplified representation, considering that the concept of social network encompasses multiple dimensions and complexities. Social networks involve not only the number of close friends but also the structure, composition, and quality of relationships, as well as the various types of social ties within an individual’s social sphere. Therefore, capturing the full richness of social network dynamics requires a more comprehensive assessment that accounts for the diverse dimensions and nuances inherent in social interactions and connections.

Our findings have profound implications for government-led relocation projects and future studies. First, relocation is by no means the end of the story but the beginning [30]. What should follow are studies to find ways to promote relocatees’ social integration into new neighborhoods, skill-building for surviving and thriving, and assessing the long-term impacts on households and their offspring’s poverty status, health, and educational achievement. These research agendas are as important as the study of relocation itself.

Second, the study of government-led relocation calls for interdisciplinary efforts among geography, demography, and sociology. For instance, geographers can contribute to the selection of suitable relocation destinations and facilitate the relocation processes by disentangling the effects of “the sense of place” in forming identity and connecting people with landscapes [68], as well as the political and ecological factors that influence governance and decision-making [69]. Demographers and sociologists can develop and assess ways to promote social integration and socioeconomic well-being after relocations.

Last but not least, it is important to acknowledge the significant influence of environmental change and variability on global migration patterns. These factors have played a crucial role in driving migration worldwide and are projected to lead to displacement of an estimated 200 million people by the year 2050 [70]. Taking this into account, more work should be done to investigate the impacts of environmental change and variability on livelihood strategies and migration patterns in regions affected by different types of environmental factors.

Supporting information

S1 File. Poverty Alleviation Relocation Program (PARP) survey instrument (excerpt).

https://doi.org/10.1371/journal.pone.0309534.s001

(PDF)

References

1. De Sherbinin A, Castro M, Gemenne F, Cernea MM, Adamo S, Fearnside PM, et al. Preparing for resettlement associated with climate change. Science. 2011; 334(6055):456–7. pmid:22034418
- View Article
- PubMed/NCBI
- Google Scholar
2. Marino E. The long history of environmental migration: Assessing vulnerability construction and obstacles to successful relocation in Shishmaref, Alaska. Glob Environ Chang. 2012; 22(2):374–81.
- View Article
- Google Scholar
3. Chen J, John R, Zhang Y, Shao C, Brown DG, Batkhishig O, et al. Divergences of two coupled human and natural systems on the Mongolian Plateau. Bioscience. 2015; 65(6):559–70.
- View Article
- Google Scholar
4. Heming L, Rees P. Population displacement in the three Georges reservoir area of the Yangtze River, Central China: Relocation policies and migrant views. Int J Popul Geogr. 2000; 6(6):439–62.
- View Article
- Google Scholar
5. Hurlimann A, Dolnicar S. Voluntary relocation—An exploration of Australian attitudes in the context of drought, recycled and desalinated water. Glob Environ Chang. 2011; 21(3):1084–94.
- View Article
- Google Scholar
6. Goetz EG. Forced relocation vs. voluntary mobility: The effects of dispersal programmes on households. Hous Stud. 2010; 17(1):107–23.
- View Article
- Google Scholar
7. Li Y, Yin H, Liu S. Relocation selection for poverty alleviation: Factor analysis and GIS modeling. J Mt Sci. 2011; 8(3):466–75.
- View Article
- Google Scholar
8. Hogg D, Kingham S, Wilson TM, Ardagh M. The effects of relocation and level of affectedness on mood and anxiety symptom treatments after the 2011 Christchurch earthquake. Soc Sci Med. 2016; 152:18–26. pmid:26826805
- View Article
- PubMed/NCBI
- Google Scholar
9. Ladd HF, Ludwig J. Federal Housing Assistance, Residential Relocation, and Educational Opportunities: Evidence from Baltimore. Am Econ Rev. 1997; 87(2):272–7. https://www.jstor.org/stable/2950930
- View Article
- Google Scholar
10. Maldonado J. A multiple knowledge approach for adaptation to environmental change: lessons learned from coastal Louisiana’s tribal communities. J Polit Ecol. 2014; 21:61–83.
- View Article
- Google Scholar
11. National Academies of Sciences, Engineering and M. Community-Driven Relocation: Recommendations for the U.S. Gulf Coast Region and Beyond. Washington, DC; 2024.
12. Gleick PH, Cohen MJ. The World’s Water, 2008–2009: The Biennial Report on Freshwater Resources. Washington, DC: Island Press; 2009.
13. Stone R. Three Gorges Dam: Into the Unknown. Science. 2008; 321:628–32. pmid:18669836
- View Article
- PubMed/NCBI
- Google Scholar
14. Xu Y, Zhang Y, Chen J. Migration under economic transition and changing climate in Mongolia. J Arid Environ. 2021; 185:104333.
- View Article
- Google Scholar
15. Amartuvshin A, Chen J, John R, Zhang Y, Lkhagvaa D. How does mining policy affect rural migration of Mongolia? Land use policy. 2021; 107:105474.
- View Article
- Google Scholar
16. Abraham M, Bähr S, Trappmann M. Gender differences in willingness to move for interregional job offers. Demogr Res. 2019; 40:1537–602.
- View Article
- Google Scholar
17. Noe RA, Barber AE. Willingness to accept mobility opportunities: Destination makes a difference. J Organ Behav. 1993; 14(2):159–75.
- View Article
- Google Scholar
18. Wagner MR, Westaby JD. The willingness to relocate to another country: The impact of cultural similarity, destination safety, and financial incentive. Int J Psychol. 2009; 44(4):257–65. pmid:22029554
- View Article
- PubMed/NCBI
- Google Scholar
19. Konopaske R, Werner S. US managers’ willingness to accept a global assignment: Do expatriate benefits and assignment length make a difference? Int J Hum Resour Manag. 2005; 16(7):1159–75.
- View Article
- Google Scholar
20. Chaskin RJ. Integration and Exclusion: Urban Poverty, Public Housing Reform, and the Dynamics of Neighborhood Restructuring. Ann Am Acad Pol Soc Sci. 2013; 647(1):237–67.
- View Article
- Google Scholar
21. Popkin SJ, Katz B, Cunningham MK, Brown KD, Gustafon J, Turner MA. A Decade of HOPE VI: Research Findings and Policy Challenges. The Urban Institute. Washington, DC; 2004.
22. Mavhura E, Collins A, Bongo PP. Flood vulnerability and relocation readiness in Zimbabwe. Disaster Prev Manag. 2017; 26(1):41–54.
- View Article
- Google Scholar
23. Bukvic A, Owen G. Attitudes towards relocation following Hurricane Sandy: should we stay or should we go? Disasters. 2017; 41(1):101–23. pmid:26988896
- View Article
- PubMed/NCBI
- Google Scholar
24. Thiede BC, Brown DL. Hurricane Katrina: Who Stayed and Why? Popul Res Policy Rev. 2013; 32(6):803–24.
- View Article
- Google Scholar
25. Li Y, Su B, Liu Y. Realizing targeted poverty alleviation in China: People’s voices, implementation challenges and policy implications. China Agric Econ Rev. 2016; 8(3):443–54.
- View Article
- Google Scholar
26. Zhou Y, Guo Y, Liu Y, Wu W, Li Y. Targeted poverty alleviation and land policy innovation: Some practice and policy implications from China. Land use policy. 2018; 74:53–65.
- View Article
- Google Scholar
27. Konopaske R, Robie C, Ivancevich JM. Managerial willingness to assume traveling, short-term and long-term global assignments. Manag Int Rev. 2009; 49(3):359–88.
- View Article
- Google Scholar
28. Lo K, Wang M. How voluntary is poverty alleviation resettlement in China? Habitat Int. 2018; 73:34–42.
- View Article
- Google Scholar
29. Xu D, Peng L, Liu S, Su C, Wang X, Chen T. Influences of Sense of Place on Farming Households’ Relocation Willingness in Areas Threatened by Geological Disasters: Evidence from China. Int J Disaster Risk Sci. 2017; 8(1):16–32.
- View Article
- Google Scholar
30. Beaumont PB. The Problem of Return Migration Under a Policy of Assisted Labor Mobility: An Examination of Some British Evidence. Br J Ind Relations. 1976; 14(1):82–8.
- View Article
- Google Scholar
31. Groen JA, Polivka AE. Hurricane Katrina evacuees: Who they are, where they are, and how they are faring. Mon Labor Rev. 2008; 131(3):32–51.
- View Article
- Google Scholar
32. Ravenstein EG. The Laws of Migration. J Stat Soc London. 1885; 48(2):167–235.
- View Article
- Google Scholar
33. Lee E. A Theory of Migration. Demography. 1966; 3(1):47–57.
- View Article
- Google Scholar
34. Massey DS, Espinosa KE. What’s Driving Mexico-U.S. Migration? A Theoretical, Empirical, and Policy Analysis. Am J Sociol. 1997; 102(4):939–99.
- View Article
- Google Scholar
35. De Haas H, Fokkema T, Fihri MF. Return Migration as Failure or Success?: The Determinants of Return Migration Intentions Among Moroccan Migrants in Europe. J Int Migr Integr. 2015; 16(2):415–29. pmid:26161043
- View Article
- PubMed/NCBI
- Google Scholar
36. Stark O, Taylor E. Relative Deprivation and International Migration. Demography. 1989; 26(1):1–14. pmid:2737350
- View Article
- PubMed/NCBI
- Google Scholar
37. Stark O, Levhari D. On Migration and Risk in LDCs. Econ Dev Cult Change. 1982; 31(1):191–6.
- View Article
- Google Scholar
38. De Haas H. Migration and development: A theoretical perspective. Int Migr Rev. 2010; 44(1):227–64. pmid:26900199
- View Article
- PubMed/NCBI
- Google Scholar
39. Mahajan P, Yang D. Taken by storm: Hurricanes, migrant networks, and US immigration. Am Econ J Appl Econ. 2020; 12(2):250–77. pmid:38562148
- View Article
- PubMed/NCBI
- Google Scholar
40. Hunter LM, Murray S, Riosmena F. Rainfall patterns and U.S. Migration from rural Mexico. Int Migr Rev. 2013; 47(4):874–909. pmid:25473143
- View Article
- PubMed/NCBI
- Google Scholar
41. Bakewell O. Relaunching migration systems. Migr Stud. 2014; 2(3):300–18.
- View Article
- Google Scholar
42. De Haas H, Fokkema T, Fihri MF. Return Migration as Failure or Success? J Int Migr Integr. 2015; 16(2):415–29. pmid:26161043
- View Article
- PubMed/NCBI
- Google Scholar
43. Massey DS. A Missing Element in Migration Theories. Migr Lett. 2015; 12(3):279–99. pmid:27398085
- View Article
- PubMed/NCBI
- Google Scholar
44. Mabogunje AL. Systems Approach to a Theory of Rural‐Urban Migration. Geogr Anal. 1970; 2(1):1–18.
- View Article
- Google Scholar
45. Hauer ME, Holloway SR, Oda T. Evacuees and Migrants Exhibit Different Migration Systems After the Great East Japan Earthquake and Tsunami. Demography. 2020; 57(4):1437–57. pmid:32430892
- View Article
- PubMed/NCBI
- Google Scholar
46. Cooke M, Bélanger D. Migration Theories and First Nations Mobility: Towards a Systems Perspective. Can Rev Sociol. 2006; 43(2):141–64.
- View Article
- Google Scholar
47. Fei X, Hamilton GG, Zheng W. From the Soil: The Foundations of Chinese Society. Berkeley, CA: University of California Press; 1992.
48. Zhao Y. Leaving the countryside: Rural-to-urban migration decisions in China. Am Econ Rev. 1999; 89(2):281–6.
- View Article
- Google Scholar
49. Liu T, Feng H, Brandon E. Would you like to leave Beijing, Shanghai, or Shenzhen? An empirical analysis of migration effect in China. PLoS One. 2018; 13(8):1–21. pmid:30114254
- View Article
- PubMed/NCBI
- Google Scholar
50. Zhao Y, Kang B, Liu Y, Li Y, Shi G, Shen T, et al. Health insurance coverage and its impact on medical cost: Observations from the floating population in China. PLoS One. 2014; 9(11):e111555. pmid:25386914
- View Article
- PubMed/NCBI
- Google Scholar
51. Kofman E. Family-related migration: A critial review of European studies. J Ethn Migr Stud. 2004; 30(2):243–62.
- View Article
- Google Scholar
52. Goldstein A, Goldstein S, Guo S. Temporary Migrants in Shanghai Households, 1984. Demography. 1991; 28(2):275–91. pmid:2070899
- View Article
- PubMed/NCBI
- Google Scholar
53. Tian Y, Wang Z, Zhao J, Jiang X, Guo R. A Geographical Analysis of the Poverty Causes in China’s Contiguous Destitute Areas. Sustainability. 2018; 10(6):1895.
- View Article
- Google Scholar
54. Lin N. Social Capital: A Theory of Social Structure and Action. Cambridge, UK: Cambridge University Press; 2002.
55. Nawrotzki RJ, Riosmena F, Hunter LM, Runfola DM. Amplification or suppression: Social networks and the climate change-migration association in rural Mexico. Glob Environ Chang. 2015; 35:463–74. pmid:26692656
- View Article
- PubMed/NCBI
- Google Scholar
56. Simpson CR. Farm size shapes friend choice amongst rice producers in China: Some evidence for the theory of network ecology. Soc Networks. 2020; 61:107–27.
- View Article
- Google Scholar
57. Hosmer DW, Lemeshow S, Sturdivant RX. Applied Logistic Regression. Third Edit. Hoboken, NJ: John Wiley & Sons, Inc.; 2013.
58. Allison PD. Logistic Regression Using the SAS System: Theory and Application. Second Edi. Cary, NC: SAS Institute Inc.; 2012.
59. Liu X, Koirala H. Ordinal regression analysis: Using generalized ordinal logistic regression models to estimate educational data. J Mod Appl Stat Methods. 2012; 11(1):242–54.
- View Article
- Google Scholar
60. Sommet N, Morselli D. Keep calm and learn multilevel linear modeling: A three-step procedure using SPSS, Stata, R, and Mplus. Int Rev Soc Psychol. 2021; 34(1):203–18.
- View Article
- Google Scholar
61. Xie Y, Hannum E. Regional Variation in Earnings Inequality in Reform-Era Urban China. Am J Sociol. 1996; 101(4):950–92.
- View Article
- Google Scholar
62. White N, Haan M. The demographic determinants of inter-provincial migration declines in Canada: A decomposition analysis. Migr Stud. 2021; 9(3):331–59.
- View Article
- Google Scholar
63. Voia M, Ci W, Haan M. Love, Money, Location: The Interconnectedness of Marital Status, Income, and Location Choice of Immigrants to British Columbia, Canada. J Demogr Econ. 2017; 83(4):421–44.
- View Article
- Google Scholar
64. Chi G. The Impacts of Transport Accessibility on Population Change across Rural, Suburban and Urban Areas: A Case Study of Wisconsin at Sub-county Levels. Urban Stud. 2012; 49(12):2711–31.
- View Article
- Google Scholar
65. Zhou S, Chi G. Farmland Rental: The Impacts of Household Demographics and Livelihood Strategies in China. Land. 2022; 11:1318.
- View Article
- Google Scholar
66. Barbalet J. The structure of guanxi: Resolving problems of network assurance. Theory Soc. 2014; 43:51–69.
- View Article
- Google Scholar
67. Wiles JL, Leibing A, Guberman N, Reeve J, Allen R. The meaning of “aging in place” to older people. Gerontologist. 2012; 52(3):357–66. pmid:21983126
- View Article
- PubMed/NCBI
- Google Scholar
68. Masterson VA, Stedman RC, Enqvist J, Tengö M, Giusti M, Wahl D, et al. The contribution of sense of place to social-ecological systems research: A review and research agenda. Ecol Soc. 2017; 22(1):49.
- View Article
- Google Scholar
69. Robbins P, Hintz JG, Moore SA. Environment and Society: A Critical Introduction. Hoboken, NJ: John Wiley & Sons; 2022.
70. Myers N. Environmental refugees: A growing phenomenon of the 21st century. Philos Trans R Soc B Biol Sci. 2002; 357:609–13. pmid:12028796
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. De Sherbinin A, Castro M, Gemenne F, Cernea MM, Adamo S, Fearnside PM, et al. Preparing for resettlement associated with climate change. Science. 2011; 334(6055):456–7. pmid:22034418
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Marino E. The long history of environmental migration: Assessing vulnerability construction and obstacles to successful relocation in Shishmaref, Alaska. Glob Environ Chang. 2012; 22(2):374–81.
View Article
Google Scholar

[6] View Article

[7] Google Scholar

[ref3] 3. Chen J, John R, Zhang Y, Shao C, Brown DG, Batkhishig O, et al. Divergences of two coupled human and natural systems on the Mongolian Plateau. Bioscience. 2015; 65(6):559–70.
View Article
Google Scholar

[9] View Article

[10] Google Scholar

[ref4] 4. Heming L, Rees P. Population displacement in the three Georges reservoir area of the Yangtze River, Central China: Relocation policies and migrant views. Int J Popul Geogr. 2000; 6(6):439–62.
View Article
Google Scholar

[12] View Article

[13] Google Scholar

[ref5] 5. Hurlimann A, Dolnicar S. Voluntary relocation—An exploration of Australian attitudes in the context of drought, recycled and desalinated water. Glob Environ Chang. 2011; 21(3):1084–94.
View Article
Google Scholar

[15] View Article

[16] Google Scholar

[ref6] 6. Goetz EG. Forced relocation vs. voluntary mobility: The effects of dispersal programmes on households. Hous Stud. 2010; 17(1):107–23.
View Article
Google Scholar

[18] View Article

[19] Google Scholar

[ref7] 7. Li Y, Yin H, Liu S. Relocation selection for poverty alleviation: Factor analysis and GIS modeling. J Mt Sci. 2011; 8(3):466–75.
View Article
Google Scholar

[21] View Article

[22] Google Scholar

[ref8] 8. Hogg D, Kingham S, Wilson TM, Ardagh M. The effects of relocation and level of affectedness on mood and anxiety symptom treatments after the 2011 Christchurch earthquake. Soc Sci Med. 2016; 152:18–26. pmid:26826805
View Article
PubMed/NCBI
Google Scholar

[24] View Article

[25] PubMed/NCBI

[26] Google Scholar

[ref9] 9. Ladd HF, Ludwig J. Federal Housing Assistance, Residential Relocation, and Educational Opportunities: Evidence from Baltimore. Am Econ Rev. 1997; 87(2):272–7. https://www.jstor.org/stable/2950930
View Article
Google Scholar

[28] View Article

[29] Google Scholar

[ref10] 10. Maldonado J. A multiple knowledge approach for adaptation to environmental change: lessons learned from coastal Louisiana’s tribal communities. J Polit Ecol. 2014; 21:61–83.
View Article
Google Scholar

[31] View Article

[32] Google Scholar

[ref11] 11. National Academies of Sciences, Engineering and M. Community-Driven Relocation: Recommendations for the U.S. Gulf Coast Region and Beyond. Washington, DC; 2024.

[ref12] 12. Gleick PH, Cohen MJ. The World’s Water, 2008–2009: The Biennial Report on Freshwater Resources. Washington, DC: Island Press; 2009.

[ref13] 13. Stone R. Three Gorges Dam: Into the Unknown. Science. 2008; 321:628–32. pmid:18669836
View Article
PubMed/NCBI
Google Scholar

[36] View Article

[37] PubMed/NCBI

[38] Google Scholar

[ref14] 14. Xu Y, Zhang Y, Chen J. Migration under economic transition and changing climate in Mongolia. J Arid Environ. 2021; 185:104333.
View Article
Google Scholar

[40] View Article

[41] Google Scholar

[ref15] 15. Amartuvshin A, Chen J, John R, Zhang Y, Lkhagvaa D. How does mining policy affect rural migration of Mongolia? Land use policy. 2021; 107:105474.
View Article
Google Scholar

[43] View Article

[44] Google Scholar

[ref16] 16. Abraham M, Bähr S, Trappmann M. Gender differences in willingness to move for interregional job offers. Demogr Res. 2019; 40:1537–602.
View Article
Google Scholar

[46] View Article

[47] Google Scholar

[ref17] 17. Noe RA, Barber AE. Willingness to accept mobility opportunities: Destination makes a difference. J Organ Behav. 1993; 14(2):159–75.
View Article
Google Scholar

[49] View Article

[50] Google Scholar

[ref18] 18. Wagner MR, Westaby JD. The willingness to relocate to another country: The impact of cultural similarity, destination safety, and financial incentive. Int J Psychol. 2009; 44(4):257–65. pmid:22029554
View Article
PubMed/NCBI
Google Scholar

[52] View Article

[53] PubMed/NCBI

[54] Google Scholar

[ref19] 19. Konopaske R, Werner S. US managers’ willingness to accept a global assignment: Do expatriate benefits and assignment length make a difference? Int J Hum Resour Manag. 2005; 16(7):1159–75.
View Article
Google Scholar

[56] View Article

[57] Google Scholar

[ref20] 20. Chaskin RJ. Integration and Exclusion: Urban Poverty, Public Housing Reform, and the Dynamics of Neighborhood Restructuring. Ann Am Acad Pol Soc Sci. 2013; 647(1):237–67.
View Article
Google Scholar

[59] View Article

[60] Google Scholar

[ref21] 21. Popkin SJ, Katz B, Cunningham MK, Brown KD, Gustafon J, Turner MA. A Decade of HOPE VI: Research Findings and Policy Challenges. The Urban Institute. Washington, DC; 2004.

[ref22] 22. Mavhura E, Collins A, Bongo PP. Flood vulnerability and relocation readiness in Zimbabwe. Disaster Prev Manag. 2017; 26(1):41–54.
View Article
Google Scholar

[63] View Article

[64] Google Scholar

[ref23] 23. Bukvic A, Owen G. Attitudes towards relocation following Hurricane Sandy: should we stay or should we go? Disasters. 2017; 41(1):101–23. pmid:26988896
View Article
PubMed/NCBI
Google Scholar

[66] View Article

[67] PubMed/NCBI

[68] Google Scholar

[ref24] 24. Thiede BC, Brown DL. Hurricane Katrina: Who Stayed and Why? Popul Res Policy Rev. 2013; 32(6):803–24.
View Article
Google Scholar

[70] View Article

[71] Google Scholar

[ref25] 25. Li Y, Su B, Liu Y. Realizing targeted poverty alleviation in China: People’s voices, implementation challenges and policy implications. China Agric Econ Rev. 2016; 8(3):443–54.
View Article
Google Scholar

[73] View Article

[74] Google Scholar

[ref26] 26. Zhou Y, Guo Y, Liu Y, Wu W, Li Y. Targeted poverty alleviation and land policy innovation: Some practice and policy implications from China. Land use policy. 2018; 74:53–65.
View Article
Google Scholar

[76] View Article

[77] Google Scholar

[ref27] 27. Konopaske R, Robie C, Ivancevich JM. Managerial willingness to assume traveling, short-term and long-term global assignments. Manag Int Rev. 2009; 49(3):359–88.
View Article
Google Scholar

[79] View Article

[80] Google Scholar

[ref28] 28. Lo K, Wang M. How voluntary is poverty alleviation resettlement in China? Habitat Int. 2018; 73:34–42.
View Article
Google Scholar

[82] View Article

[83] Google Scholar

[ref29] 29. Xu D, Peng L, Liu S, Su C, Wang X, Chen T. Influences of Sense of Place on Farming Households’ Relocation Willingness in Areas Threatened by Geological Disasters: Evidence from China. Int J Disaster Risk Sci. 2017; 8(1):16–32.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref30] 30. Beaumont PB. The Problem of Return Migration Under a Policy of Assisted Labor Mobility: An Examination of Some British Evidence. Br J Ind Relations. 1976; 14(1):82–8.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref31] 31. Groen JA, Polivka AE. Hurricane Katrina evacuees: Who they are, where they are, and how they are faring. Mon Labor Rev. 2008; 131(3):32–51.
View Article
Google Scholar

[91] View Article

[92] Google Scholar

[ref32] 32. Ravenstein EG. The Laws of Migration. J Stat Soc London. 1885; 48(2):167–235.
View Article
Google Scholar

[94] View Article

[95] Google Scholar

[ref33] 33. Lee E. A Theory of Migration. Demography. 1966; 3(1):47–57.
View Article
Google Scholar

[97] View Article

[98] Google Scholar

[ref34] 34. Massey DS, Espinosa KE. What’s Driving Mexico-U.S. Migration? A Theoretical, Empirical, and Policy Analysis. Am J Sociol. 1997; 102(4):939–99.
View Article
Google Scholar

[100] View Article

[101] Google Scholar

[ref35] 35. De Haas H, Fokkema T, Fihri MF. Return Migration as Failure or Success?: The Determinants of Return Migration Intentions Among Moroccan Migrants in Europe. J Int Migr Integr. 2015; 16(2):415–29. pmid:26161043
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref36] 36. Stark O, Taylor E. Relative Deprivation and International Migration. Demography. 1989; 26(1):1–14. pmid:2737350
View Article
PubMed/NCBI
Google Scholar

[107] View Article

[108] PubMed/NCBI

[109] Google Scholar

[ref37] 37. Stark O, Levhari D. On Migration and Risk in LDCs. Econ Dev Cult Change. 1982; 31(1):191–6.
View Article
Google Scholar

[111] View Article

[112] Google Scholar

[ref38] 38. De Haas H. Migration and development: A theoretical perspective. Int Migr Rev. 2010; 44(1):227–64. pmid:26900199
View Article
PubMed/NCBI
Google Scholar

[114] View Article

[115] PubMed/NCBI

[116] Google Scholar

[ref39] 39. Mahajan P, Yang D. Taken by storm: Hurricanes, migrant networks, and US immigration. Am Econ J Appl Econ. 2020; 12(2):250–77. pmid:38562148
View Article
PubMed/NCBI
Google Scholar

[118] View Article

[119] PubMed/NCBI

[120] Google Scholar

[ref40] 40. Hunter LM, Murray S, Riosmena F. Rainfall patterns and U.S. Migration from rural Mexico. Int Migr Rev. 2013; 47(4):874–909. pmid:25473143
View Article
PubMed/NCBI
Google Scholar

[122] View Article

[123] PubMed/NCBI

[124] Google Scholar

[ref41] 41. Bakewell O. Relaunching migration systems. Migr Stud. 2014; 2(3):300–18.
View Article
Google Scholar

[126] View Article

[127] Google Scholar

[ref42] 42. De Haas H, Fokkema T, Fihri MF. Return Migration as Failure or Success? J Int Migr Integr. 2015; 16(2):415–29. pmid:26161043
View Article
PubMed/NCBI
Google Scholar

[129] View Article

[130] PubMed/NCBI

[131] Google Scholar

[ref43] 43. Massey DS. A Missing Element in Migration Theories. Migr Lett. 2015; 12(3):279–99. pmid:27398085
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

[ref44] 44. Mabogunje AL. Systems Approach to a Theory of Rural‐Urban Migration. Geogr Anal. 1970; 2(1):1–18.
View Article
Google Scholar

[137] View Article

[138] Google Scholar

[ref45] 45. Hauer ME, Holloway SR, Oda T. Evacuees and Migrants Exhibit Different Migration Systems After the Great East Japan Earthquake and Tsunami. Demography. 2020; 57(4):1437–57. pmid:32430892
View Article
PubMed/NCBI
Google Scholar

[140] View Article

[141] PubMed/NCBI

[142] Google Scholar

[ref46] 46. Cooke M, Bélanger D. Migration Theories and First Nations Mobility: Towards a Systems Perspective. Can Rev Sociol. 2006; 43(2):141–64.
View Article
Google Scholar

[144] View Article

[145] Google Scholar

[ref47] 47. Fei X, Hamilton GG, Zheng W. From the Soil: The Foundations of Chinese Society. Berkeley, CA: University of California Press; 1992.

[ref48] 48. Zhao Y. Leaving the countryside: Rural-to-urban migration decisions in China. Am Econ Rev. 1999; 89(2):281–6.
View Article
Google Scholar

[148] View Article

[149] Google Scholar

[ref49] 49. Liu T, Feng H, Brandon E. Would you like to leave Beijing, Shanghai, or Shenzhen? An empirical analysis of migration effect in China. PLoS One. 2018; 13(8):1–21. pmid:30114254
View Article
PubMed/NCBI
Google Scholar

[151] View Article

[152] PubMed/NCBI

[153] Google Scholar

[ref50] 50. Zhao Y, Kang B, Liu Y, Li Y, Shi G, Shen T, et al. Health insurance coverage and its impact on medical cost: Observations from the floating population in China. PLoS One. 2014; 9(11):e111555. pmid:25386914
View Article
PubMed/NCBI
Google Scholar

[155] View Article

[156] PubMed/NCBI

[157] Google Scholar

[ref51] 51. Kofman E. Family-related migration: A critial review of European studies. J Ethn Migr Stud. 2004; 30(2):243–62.
View Article
Google Scholar

[159] View Article

[160] Google Scholar

[ref52] 52. Goldstein A, Goldstein S, Guo S. Temporary Migrants in Shanghai Households, 1984. Demography. 1991; 28(2):275–91. pmid:2070899
View Article
PubMed/NCBI
Google Scholar

[162] View Article

[163] PubMed/NCBI

[164] Google Scholar

[ref53] 53. Tian Y, Wang Z, Zhao J, Jiang X, Guo R. A Geographical Analysis of the Poverty Causes in China’s Contiguous Destitute Areas. Sustainability. 2018; 10(6):1895.
View Article
Google Scholar

[166] View Article

[167] Google Scholar

[ref54] 54. Lin N. Social Capital: A Theory of Social Structure and Action. Cambridge, UK: Cambridge University Press; 2002.

[ref55] 55. Nawrotzki RJ, Riosmena F, Hunter LM, Runfola DM. Amplification or suppression: Social networks and the climate change-migration association in rural Mexico. Glob Environ Chang. 2015; 35:463–74. pmid:26692656
View Article
PubMed/NCBI
Google Scholar

[170] View Article

[171] PubMed/NCBI

[172] Google Scholar

[ref56] 56. Simpson CR. Farm size shapes friend choice amongst rice producers in China: Some evidence for the theory of network ecology. Soc Networks. 2020; 61:107–27.
View Article
Google Scholar

[174] View Article

[175] Google Scholar

[ref57] 57. Hosmer DW, Lemeshow S, Sturdivant RX. Applied Logistic Regression. Third Edit. Hoboken, NJ: John Wiley & Sons, Inc.; 2013.

[ref58] 58. Allison PD. Logistic Regression Using the SAS System: Theory and Application. Second Edi. Cary, NC: SAS Institute Inc.; 2012.

[ref59] 59. Liu X, Koirala H. Ordinal regression analysis: Using generalized ordinal logistic regression models to estimate educational data. J Mod Appl Stat Methods. 2012; 11(1):242–54.
View Article
Google Scholar

[179] View Article

[180] Google Scholar

[ref60] 60. Sommet N, Morselli D. Keep calm and learn multilevel linear modeling: A three-step procedure using SPSS, Stata, R, and Mplus. Int Rev Soc Psychol. 2021; 34(1):203–18.
View Article
Google Scholar

[182] View Article

[183] Google Scholar

[ref61] 61. Xie Y, Hannum E. Regional Variation in Earnings Inequality in Reform-Era Urban China. Am J Sociol. 1996; 101(4):950–92.
View Article
Google Scholar

[185] View Article

[186] Google Scholar

[ref62] 62. White N, Haan M. The demographic determinants of inter-provincial migration declines in Canada: A decomposition analysis. Migr Stud. 2021; 9(3):331–59.
View Article
Google Scholar

[188] View Article

[189] Google Scholar

[ref63] 63. Voia M, Ci W, Haan M. Love, Money, Location: The Interconnectedness of Marital Status, Income, and Location Choice of Immigrants to British Columbia, Canada. J Demogr Econ. 2017; 83(4):421–44.
View Article
Google Scholar

[191] View Article

[192] Google Scholar

[ref64] 64. Chi G. The Impacts of Transport Accessibility on Population Change across Rural, Suburban and Urban Areas: A Case Study of Wisconsin at Sub-county Levels. Urban Stud. 2012; 49(12):2711–31.
View Article
Google Scholar

[194] View Article

[195] Google Scholar

[ref65] 65. Zhou S, Chi G. Farmland Rental: The Impacts of Household Demographics and Livelihood Strategies in China. Land. 2022; 11:1318.
View Article
Google Scholar

[197] View Article

[198] Google Scholar

[ref66] 66. Barbalet J. The structure of guanxi: Resolving problems of network assurance. Theory Soc. 2014; 43:51–69.
View Article
Google Scholar

[200] View Article

[201] Google Scholar

[ref67] 67. Wiles JL, Leibing A, Guberman N, Reeve J, Allen R. The meaning of “aging in place” to older people. Gerontologist. 2012; 52(3):357–66. pmid:21983126
View Article
PubMed/NCBI
Google Scholar

[203] View Article

[204] PubMed/NCBI

[205] Google Scholar

[ref68] 68. Masterson VA, Stedman RC, Enqvist J, Tengö M, Giusti M, Wahl D, et al. The contribution of sense of place to social-ecological systems research: A review and research agenda. Ecol Soc. 2017; 22(1):49.
View Article
Google Scholar

[207] View Article

[208] Google Scholar

[ref69] 69. Robbins P, Hintz JG, Moore SA. Environment and Society: A Critical Introduction. Hoboken, NJ: John Wiley & Sons; 2022.

[ref70] 70. Myers N. Environmental refugees: A growing phenomenon of the 21st century. Philos Trans R Soc B Biol Sci. 2002; 357:609–13. pmid:12028796
View Article
PubMed/NCBI
Google Scholar

[211] View Article

[212] PubMed/NCBI

[213] Google Scholar

Figures

Abstract

Introduction

Conceptual framework

Hypotheses

Materials and methods

Data

Variables

Dependent variables.

Independent variables.

Modeling

Results

Descriptive statistics

Relocation preference model

Settlement model

Discussion and conclusions

Supporting information

S1 File. Poverty Alleviation Relocation Program (PARP) survey instrument (excerpt).

References