2024 年 12 巻 4 号 p. 253-278
Abstract Urban migration the expansion of informal settlements are pressing issues in fast-growing cities like Bangkok The Klong Toei community a significant informal settlement faces challenges related to relocation driven by urban development land reclamation initiatives from the Port Authority of Thailand This study examines how job location influences residents' willingness to relocate addressing a critical gap in urban planning literature concerning the socioeconomic factors impacting slum dwellers This research aims to explore the relationship between the proximity nature of employment the willingness of Klong Toei residents to relocate It seeks to inform more humane effective resettlement strategies that reflect the community's real needs preferences Employing a mixed-methods design this study integrates quantitative data from structured surveys with qualitative insights from interviews conducted in three distinct sub-communities of Klong Toei Analytical techniques including spatial autocorrelation cluster analysis are utilized to investigate the patterns influencing relocation decisions The findings indicate that job locations are randomly distributed not clustered with a statistically significant correlation between job location willingness to relocate Key factors affecting this willingness include job type distance from employment generational differences among residents The results emphasize the need for urban policymakers to consider employment factors in the planning of resettlement programs By aligning urban development strategies with the employment realities of slum dwellers policymakers can enhance the acceptance success of relocation efforts ensuring they meet the diverse needs of affected communities This study provides valuable insights for future policy-making urban planning processes
In fast-growing economies, informal settlement is a transient phenomenon brought about by urbanization and rapid economic growth when migrating to an urban city to obtain employment opportunities (Badmos, Callo-Concha et al., 2020; Marx, Stoker et al., 2013), which also triggers the disequilibrium supply of the housing market (Haddad, Ruel et al., 1999). According to the United Nations (2018), the number of people living in informal settlements is 370 million in East and South-East Asia. In Bangkok, in 2018, informal dwellers accounted for 24% of the total population in Thailand, and 579,603 people lived in informal settlements (Pongutta, Kantamaturapoj et al., 2021).
The informal settlement Klong Toei communityThe Klong Toei community is the largest informal settlement in Bangkok, comprising 26 sub-communities located on the banks of the Chao Phraya River in southern Bangkok, covering an area of 1.5 square kilometers. According to the history of the Klong Toei community in 1951 (DiNino, Garabedian et al., 2006), the government established the Port Authority of Thailand as a state-owned enterprise under the Ministry of Transportation (Port Authority of Thailand, 2018). The Khlong Toei Port is Thailand’s main port for sea cargo transportation, constructed in 1938 and completed after World War II. Nowadays, the informal settlement built on land belonging to the Port Authority of Thailand (PAT) is home to approximately 85,000 to 100,000 people. Figure 1 illustrates land owned by PAT in the Klong Toei district.
Source: Adapted from Ferrero, Castillo et al. (2018)
Since the 1970s, the policy to mitigate informal settlement in Thailand, Viratkapan and Perera (2006) indicate that the governmental implementation strategies primarily focus on land sharing, re-blocking, reconstruction, and relocation by giving security tenure for informal dwellers. The government sector National Housing Authority (NHA) was established in 1973 to address urban poverty and scarcity of housing demands and is mainly in charge of providing affordable apartments for low-income households. Unfortunately, it is doomed to failure because many projects are being built far away from the city, and those approaches need more support from politicians (Yap and De Wandeler, 2010). Since then, the Thai government launched the Baan Mankong program in 2003 (Boonyabancha, 2005b). It has successfully solved the issue of housing renovation in informal settlements across Thailand through government subsidies and a series of professional and systematic management. Furthermore, CODI, the government agency, proposes a housing environment improvement plan for the Klong Toei community. The final plan, however, has been shelved because of difficulties communicating with community leaders and the other planned commercial development goals.
Problem statementRecently, the PAT has requested land reclamation to expand Bangkok University’s educational campus according to the meeting with Duang Prateep Foundation (Personal communication, 2022) and economic development as the government’s plan to develop a smart community. Therefore, the PAT offers residents three options: 1). 33 sqm apartment on Soi Trimit, which is 2km away from the current location. 2). Eighty square meters of land in the Nong Chok district in the suburbs. 3). cash compensation (Ferrero, Castillo et al., 2018). In response to these compensations, residents express different concerns. For example, a 33-square-meter apartment cannot fulfil the living space requirements of a two-generation family. Even if the land can meet the space requirements, the prerequisite is that it is likely impractical for the residents to bear the cost of land preparation and house construction. Besides these issues, suburban employment opportunities are few, and educational institutions and living facilities need to be improved. Most importantly, people’s livelihood matters must be tackled and resolved. Likewise, monetary compensation depends on various individual economic backgrounds. Some residents stated that despite receiving funds, they still encounter housing problems due to heightened housing prices.
With the rapid economic development and the continuous expansion of the urban scale, many residents living in informal settlements illegally on public or private land often are distressed about eviction by landlords (Astuti, Rahayu et al., 2023). Given this, the assumption seems superficial from the governmental institutions’ viewpoint because the information received may be limited or from a unitary aspect. Willingness to relocate may vary according to diversity households have different concerns. From conversations with locals during the site survey, job location is likely to play a vital role in directing the dwellers’ willingness to relocate. Hence, this study aims to find the crux of the situation through face-to-face interviews by listening to residents’ voices, opinions, and notions to comprehend their willingness to relocate. Ultimately, the results obtained through the analysis intend to be transformed into a full regard for the relevant governmental institutions and stakeholders in the resettlement planning.
Research objectivesThe first is to explore the cluster and map of job locations of Klong Toei residents.
The second is to study the relationship of job locations regarding Klong Toei residents' willingness to relocate.
The third is to identify the attributes of residents related to relocation decisions.
Research questionsThe willingness to relocate is complex due to the variety of considerations in each individual or depending on the household’s socioeconomic situation. Therefore, the research question arises:
First, what is the relationship between residents’ job locations and the factors affecting residents’ willingness to relocate?
Second, what kind of characteristics of residents led to the relocation decision?
Third, what types of jobs possess a higher willingness to relocate?
HypothesisFirst, residents’ job locations are likely spatially clustered, which is a significant factor affecting residents' willingness to relocate.
Second, generation disparity, job location, and type of job play influential roles in directing the relocation decision.
Third, residents who work mobile jobs are more willing to relocate due to flexibility in job locations.
Scope of the studyThis study is carried out in 50 districts in Bangkok to explore the cluster of job locations for residents in the Klong Toei community. The three selected Klong Toei communities have been deliberately chosen as a high density of residents that contain diverse cultures, walks of life, and economic sub-categories of informal settlement communities. Figure 2 illustrates the location of selected sub-communities including informal settlements (Lock 1-2-3 & 4-5-6), subdivisions (70 Rai), and four stories of height Flat as the survey area. Further description of each community: Lock is an unplanned settlement (Atitruangsiri, 2017). Conversely, 70 Rai is a planned area designed and cooperated by the PAT while relocating residents from a previous informal site in the early 90s (Goodwin, Mongkolsivaphorn et al., 2015). Moreover, PAT builds Flats for relocating slum dwellers and renting purposes.
Several pieces of literature are related to informal resettlement for various research purposes according to different socioeconomic statuses and cultural norms worldwide. A higher homogeneity in diverse societies and nations, in which this literature review emphasizes the factors commonly appearing to correspond to informal settlement mobility. Social capital significantly impacts informal dwellers’ relocation decisions, especially for disadvantaged groups in mid-low-income countries, with high social capital, strong bonding, and trust connect counterparts with similar socioeconomic status (Dewi, Eko et al., 2023; Kim, T.-K., Horner et al., 2005). Dwellers profoundly rely on solid relationships with relatives and friends, introducing job opportunities and finding dwellings. In Semarang, Indonesia, Manaf, Wahyono et al. (2018) point out that social environments, such as relationships with neighbors, relatives, family, and community activities, lead to the highest satisfaction and affect willingness to move out of the informal settlement. Likewise, Badmos, Callo-Concha et al. (2020) survey in Lagos, Africa, underlines the importance of neighborhood and family characteristics in residents’ decisions to relocate, while Arandel and Wetterberg (2013) discuss that informal resettlement disruptively stresses job opportunities, social networks, and livelihood.
In Bangkok and Ho Chi Minh, Carpenter, Daniere et al. (2004) surveyed five specific neighborhoods, showing that Bangkok’s informal settlements are closely linked to the city’s employment and transportation structure. However, due to tenure, informal settlements in Bangkok have a strong structured interaction over the right to live for an extended period. Regardless, social ties are more confined to community dwellers than ties outside the community. In a survey by Badmos, Callo-Concha et al. (2020), dwellers cite family ties, affordable housing, and proximity to work as significant factors in moving to an informal settlement. Moreover, Lall, Suri et al. (2006) revealing that moving to the city’s suburbs disrupts the social networks that dwellers rely on for livelihood resources. As a result, relocating informal dwellers without considering connections between family, relatives, neighbors, and ethnic enclaves, the new location without providing jobs and benefits, experiences dilemmas during the relocation process.
However, Bayrau and Bekele (2007) claim that relocating slums typically benefits urban development. The potential development of the private sector and land values increase employment opportunities and raise incomes. This study indicates that the main reason for the remaining reluctance to move out is that they are satisfied with the existing living environment and have a strong sense of security. The solid cultural value factor makes the residents prefer maintaining their neighborhood relations. Some works of literature study the sequence after relocation; for example, regarding resettlement, collaborative action between governmental authorities and private sectors plays a vital role in the successful implementation (Badhan and Siddika, 2019; Zuhdi, Rusli et al., 2023). Aziz, Ismail et al. (2014) describe the massive demolition of informal settlements and relocation to high-rise buildings in Kuala Lumpur during the 80s-90s and find that the previous social tie no longer existed after the resettlement. Ultimately, the collapse of the initial community structure, ethnic conflict, and lack of living environment led to a lack of sense of belonging among residents. Similarly, Samuel and Nisar (2021) point out that contrasts in ethnicity and cultural practices are likely to lead to conflicts after moving out of government housing because of religious disparities. Alternatively, depending on religious beliefs, family size is profoundly affected, with factors of solid social capital making them more inclined to stay in existing informal settlements.
In the study of residential relocation reasons and duration of living, Eluru, Sener et al. (2009) indicate that demographic, socioeconomic, and commute-related variables significantly influence moving or staying. In particular, the effect of family size on residence living period shows heterogeneity, which further explains that the main reason for moving is to cope with changes in family composition. Regarding dwellers’ mobility, Satu and Juthi (2019) surveyed three informal settlements in Bangladesh and revealed that family size is a critical factor affecting the mobility of informal dwellers. Likewise, Kim, T.-K., Horner et al. (2005) explore environmental and neighborhood factors at different life stages in people’s choices of jobs and places to live. The survey finds that the importance of age at home parenting stage is positively correlated with job availability. Conversely, young families without children or low-income individuals value job availability. In other words, relocation decisions are made by the stage in the household life cycle.
Theoretical Relocation DecisionThe behavior of the decision to migrate in demography, social science, and related disciplines has been widely studied. Wolpert (1965) proposes the theory of the behavior of the migration decision, known as the stress-threshold model. An extension of the model, Speare (1974) suggests residential satisfaction to witness how mobility influenced individual and residence variables. The cost-benefit model guides most migrants to move due to economic considerations. At the same time, job information by word of mouth and proximity to friends and relatives play a critical role in deciding where to move. The model is divided into three dimensions: individual or household characteristics, location characteristics, and social bonds that affect residential satisfaction. In the study, variables such as age, income, and duration of living may not direct the consideration of moving; however, it is likely to affect satisfaction as individuals accumulate wealth to live in a better environment. Simultaneously, living in the same neighborhood implies acquiring more friends and solid social bonds, and the proximity of local facilities increases satisfaction.
In the traditional urban economic model, the labor market relies on assumptions about information availability and employment mobility. Rouwendal (1998) argues that the employees are homogeneous and analyses the relationship between the two employment centers. When an employee commutes to work at another employment center, the counted commute costs could be 8% of income. It says the closer the employee lives to an employment center, the higher the possibility of high unemployment. It also states that the sub-center’s role in urbanized areas affects the spatial mismatch of urban labor markets. Following the spatial mismatch hypothesis, Bunel and Tovar (2014) explain the circumstances posed by labor and land markets due to residential segregation, employment fragmentation, and spatial friction in job search and commuting. Thus, it carries the phenomenon of unemployment, low income, and low house rent.
In urban areas, obtaining employment opportunities is complex and far more intricate than finding housing. Boschmann (2011) points out that low-income workers face significant location constraints, such as the geographic reach of job searches, skills-related job opportunities, commuting mobility, and affordable housing options. Regarding urban residents’ job accessibility, Reingold (1999) indicates that downtown residents are less likely to find jobs in the suburbs because of inaccessibility, limited information, and discrimination by suburban employers, leading residents in deprived inner-city areas to rely on seeking jobs within their perimeter. Likewise, Ommeren, Rietveld et al. (2000) examine the dependency on labor and explain the behavior of employees in job mobility and residential relocation. The search theory proposes factors that lead to an imperfect housing market, causing employees to be likely to reduce employment opportunities in resettlement. Moreover, commuting distance is vital when choosing jobs and residential mobility. The consideration of traffic congestion increases commuting costs in an urban megacity; when the cost of residence movement increases, it offsets the employee’s acceptance of the job opportunity.
Cities are more productive than surrounding hinterlands because urban economies include affluent labor markets, input-output linkages, and knowledge spillovers. Transportation is integral to residential decision-making and reduces the spatial mismatch between housing, work, and other activities (Kim, J. H., Pagliara et al., 2005). Furthermore, So, Orazem et al. (2001) analyse the relationship between wages, commuting costs, housing prices, housing choices, and job location. Transportation improvements reduce commute times, which increases the number of non-metropolitan population commuters to the metropolitan labor market. Likewise, Boschmann (2011) states that the convenience of bus routes is the most crucial factor in the choice of housing for disadvantaged working groups.
On the other hand, Prashker, Shiftan et al. (2008) sought to understand commuting distance and gender as factors in choosing where to live in Tel Aviv, Israel. The main factors influencing residential location decisions are residential units, location, accessibility, and individual characteristics. Household and neighborhood characteristics are considered explanatory variables in which household variables include age, children, income, and education, while neighborhood characteristics include density, crime rate, safety, and school quality.
Summary of literature reviewIn research on residential relocation behavior, many works of literature categorize moving determinants into demographic characteristics, individual and household socioeconomic, environmental attributes, housing supply, commuting means, and drivers of moving (Eluru, Sener et al., 2009). The literature on different aspects of the microeconomy, employment market, and relocation intentions shows that transportation and commuting choices of low-income households play an essential role in many high-income countries. In contrast, household characteristics, neighborhood relationships, employment opportunities, and livelihoods are often the primary considerations for informal dwellers in mid-low-income countries.
Many social and economic studies on employment attributes such as job accessibility, labor market, and industrial distribution show that spatial autocorrelation methods articulate straightforward and intuitive results in research related to geographic data. Moreover, traditional exploratory data analysis deals with relationships between variables and how they affect each other. In contrast, spatial autocorrelation associates a specific variable with location while considering the same variable’s values in the neighborhood. For example, Debnath and Naznin (2011) examine spatial dependencies between informal settlements and industrial development. Hess (2005) employs GIS to map residence and employment locations and calculate employment and transport access measures.
According to the literature review analogous to this study, Badmos, Callo-Concha et al. (2020), Satu and Juthi (2019) and Kim, T.-K., Horner et al. (2005) employ descriptive statistics and a chi-square test to examine determinants of slum dwellers and households in residential choices. Prashker, Shiftan et al. (2008) utilize descriptive statistics to estimate various characteristics and factors influencing individuals’ living choices. Regarding application selection, Debnath and Naznin (2011) apply Geoda to construct a spatial autoregressive model and perform the statistical analysis. Kim, T.-K., Horner et al. (2005) use GIS modeling to build a database of spatial variables to predict the importance of proximity to work. Manaf, Wahyono et al. (2018) analyse the satisfaction and willingness to move slum dwellers on a five-point Likert scale. Subsing, Tanpichai et al. (2021) utilize the SPSS application to analyze the factors affecting the quality of life of the elderly in Bangkok slums. Moreover, Pryer, Rogers et al. (2002) suggest that cluster analysis identifies subsistence groups among households to classify the homogeneity of groups.
The factors of job location are likely to influence residents’ decision-making in the Klong Toei community. As times changes and economic circumstances improve, early literature focuses on subjects such as malnutrition and the health of slum dwellers (Haddad, Ruel et al., 1999; Subsing, Tanpichai et al., 2021), environmental upgrading, land tenure (DiNino, Garabedian et al., 2006), and adaptation to relocation problems encountered later (Aziz, Ismail et al., 2014). Hence, there is no relevant research on whether the factors related to the residents’ workplace affect residents’ willingness to relocate. Nevertheless, relevant literature suggests that offering compensation schemes (Viratkapan and Perera, 2006) is the most effective approach to obtaining residents’ consent to relocation. However, these early findings and methods are likely only applicable to some residents of the Khlong Toei community nowadays.
The literature review mapping (Figure 3) represents the linkage from the theory of decision-making to migration by Wolpert (1965) to the model of mobility decision-making proposed by Speare (1974). Many empirical literature reviews on informal resettlement mainly emphasize the aspect of social capital and life cycle. On the other hand, regarding location characteristics, it has wildly studied the aspects of job accessibility and transportation for disadvantaged groups in high-income countries; in turn, job locations in mid-low-income countries have seldom been addressed when relocating informal dwellers. Hence, it leaves much room for further research after the prevailing conditions of the informal settlement state have altered.
This study employs a mixed methodology with a convergent parallel design to overlap the quantitative results and qualitative findings. Quantitative analytic methods contain spatial autocorrelation, chi-square test, and cluster analysis. Conversely, the qualitative context analysis seeks to be conducted independently, and the results of both approaches are combined in the interpretation results (Busetto, Wick et al., 2020).
Spatial autocorrelation has been crucial among spatial modelers since the early 1970s. According to Tobler's first law, measurements at locations close to each other will be more similar than measurements at distant locations. In other words, if features closer to each other are more similar than those farther apart, they are spatially correlated and form a cluster (Birch, Chikukwa et al., 2009; Peeters, Zude et al., 2015). Moran's I statistic is the most widely used measure and test of spatial autocorrelation (Getis, 2008) which tests the overall spatial interdependence between a region and its neighbors. However, spatial autocorrelation measures require a weight matrix that defines the local neighborhood around each geographic cell (Debnath and Naznin, 2011). The formula of Moran's I is as follows:
(1)
where: N is the number of observations (points or polygons);
The Chi-Square test is a statistical procedure for determining the difference between observed and expected values. It is also used to determine whether it correlates to the categorical variables and whether there is a difference or a relationship between two or more categorical independent variables. Therefore, Pearson's Chi-square (χ2) probability distributions typically test the distribution difference and aid in analyzing categorical variables (Hazra and Gogtay, 2016).
Cluster analysis is a technique to sort observations into similar sets or groups. Hierarchical algorithms build a tree-like structure by adding or deleting individual elements from a cluster (Ketchen and Shook, 1996). Furthermore, Ward's minimum variance and iterative partitioning methods outperform other methods and share the total sum of squares error criterion with K-means partitioning, which is used to cluster observations directly in Euclidean space (Murtagh and Legendre, 2014). The K-means procedure performs less decrement than hierarchical methods (Punj and Stewart, 1983), and divide samples into K clusters so that the within-cluster sum of squares is minimized (Hartigan and Wong, 1979). The formula of the K-means cluster is as follows:
(2)
where: S is sets of observations; k is the number of sets of predictors; x is the observation data point; μi is the mean of points in Si.
Thus, this study first computes Global Moran’s I to evaluate whether the relationship of residents’ job location is clustered or spatial independent. Meanwhile, spatial data utilizes the Geoda and QGIS 3.20 application to create a job location map. In expansion, the Crimestat application has been employed to illustrate the distribution of job locations through a standard deviation ellipse, which represents the mean and median center of job locations in the Klong Toei district. Secondly, to answer the research question of which job location determines residents’ willingness to relocate, the chi-square test is employed to examine the relationship between factors and job location with the willingness to relocate. In addition, since the data collected has been coded into categorical variables, the chi-square test adequately answers the research question. Finally, cluster analysis identifies clusters of points in space. It has been employed as a classification tool to interpret the residents’ groups, which provides a perspective on the categorized groups with different degrees of willingness to relocate. Moreover, this study utilizes SPSS to examine cluster analysis. Figure 4 demonstrates the research procedure for conducting the survey research and data analysis procedure.
This study employs the survey research method to collect primary data via face-to-face interviews in three selected Klong Toei sub-communities for gaining the individual, household, and occupational-related data for the quantitative approach by explanatory method to investigate the relationship between factors and job location on the willingness to relocate. At the same time, a semi-structured questionnaire is designed at the end of the questionnaire to discuss paramount factors while considering moving to a new place for the qualitative content analysis approach.
Sampling designCluster sampling is a statistical method used to divide the population into clusters or groups (Taherdoost, 2016), and each cluster provides a miniature representation of the entire population. Considering the implementation, this study adopts the cluster sampling technique suitable for large geographical areas in Lock, 70 rai, and Flat. In addition, randomization is operated to classify the population, refrain from the bias generated during the survey, and maximize the characteristics of the population through the sample. According to the central limit theorem, a minimum of 30 sampling requirements is sufficient and satisfies the population’s standard deviation (Henderson and Sundaresan, 1982; Islam, 2018). Therefore, this study examines primary data by interviewing 30 samplings of each community, a total of 90 residents.
Unit of analysisThe unit of analysis is individuals who live in Lock, 70 Rai, and Flat in Klong Toei communities.
VariablesBased on the framework extracted and adopted by Speare (1974), the questionnaire contents are organized according to variables identified from the theory and literature reviews. This study divides the determinants of relocation decisions into three categories: individual characteristics, household characteristics, and occupational attributes. The dependent variable (Y) applies a four-Likert scale to measure the degree of residents’ willingness to relocate, while independent variables are categorized. Figure 5 displays the conceptual framework of variables affecting the willingness to relocate into individual characteristics, including independent variables (X) of age, marital, gender, and educational attainment. Household characteristics consist of independent variables (X) of household members, the dependency ratio (children and elderly in the household), monthly household income, duration of living, relatives and close friends living in the community, and type of tenure. Core independent variables (X) occupational attributes include job type and location.
The median age of the respondents is 35, and 54.4% of them are male while 45.6% are female. Of the respondents, 48.9% are single, and 31.1% have attained a university or higher vocational degree. The median number of people living in each household is four. 53.3% of the respondents have no children while 63.3% have no elderly living in the household. 57.8% of households have a monthly income of 5000-20,000 Baht, and 71.1% of respondents work within 3 km. The median distance for job location is 1.2 km. The average job location near Lock is 1.78 km, 70 Rai 3.93 km, and Flat 2.64 km. The median duration of living in the community is 20 years, and the median number of relatives and close friends living in the community is three. Regarding the type of tenure, 45.6% are sub-renters, and 53.4% of the respondents work in the retail and services sector.
Spatial data Job location mapDuring the face-to-face interview, participants are requested to identify their job location using Google Maps on an iPad. Figure 6 illustrates job locations of respondents in three different communities. Based on the map, it can be inferred that residents in Lock community work within Klong Toei and surrounding districts, and are located close to the centroid. On the other hand, residents in 70 Rai are likely to work far away from the Klong Toei community. Flat has a higher percentage of residents working in the private office sector and their job location is in close proximity to the central business district of Bangkok. Additionally, the minimum job location distance is 0.08 km within the community, whereas the maximum job location distance is 34.7 km in the Nong Chok district.
The standard deviational ellipse outlines the geographical distribution trend by summarizing the dispersion and orientation of the observed data. It is determined by average location, dispersion or concentration, and orientation through delineating spatial point data via GIS (Wang, Shi et al., 2015). The locus of the standard deviation values yields the ellipse as the axis rotates around the mean center (Lefever, 1926). Figure 7 illustrates the distribution of Klong Toei residents’ job locations. The overlaid standard deviational ellipse represents an intuitive visualization of the spatial distribution of Klong Toei residents’ job locations, demonstrating that the concentration patterns of job locations are highly centralized in the Klong Toei district. More specifically, a considerable number of residents are working within the community.
In many studies, the social tie is likely a critical factor leading to residents preferring to stay. However, in the modern Klong Toei community, the social tie is less vital than in previous studies. On the contrary, age appears as a significant factor affecting the willingness to relocate, such as the young generation aged 17-25 in Flat being more willing to move than the age greater than 36 in Lock and 70 Rai. Furthermore, it is noteworthy that Generation Z is likely to be more willing to move due to environmental concerns such as a safer environment for children growing up. The findings from the residents’ quotes and distribution of different age groups indicate that generation disparity plays a vital role in determining the relocation decision. Table 1 summarises age groups by willingness to relocate.
Despite the assumption that residents born and raised in the community are likely improbable to move generally, the survey finds that 34 respondents were born and raised in the community. However, the qualitative findings reveal that another aspect of the second generation is keen to move because of the uncomfortable fighting status for compensation or with neighbours. Moreover, in Lock, there is a minority of senior residents with a longer duration of the living who express that they are likely to move due to better economic status and some friendship issues in the community.
| Age groups | Respondent quotes | |
|---|---|---|
| Very improbable and improbable to move | 46-55 years old |
"The environment in the community has been improved much better compared to before that I feel improbable to move. But honestly to say as everyone knew it, the drag problems are serious." (70 rai respondent 26, Age 47) "I am living here 35 years with six children and one elderly household and work at Soi Kheha. I feel very improbable to move because the rental I pay is only 120 baht per month." (70 Rai respondent 3, Age 55) |
| 56-65 years old | "I am working in Nong Chok district by two tiao transportation. I do not feel probable to move because it is convenient to reach anywhere close to the city center. However, there will be transportation problems outside of the city." (70 Rai respondent 2, Age 56) | |
| Probable and very probable to move | 17-25 years old | "I work at PAT as a repairing technician and my family owns the house. However, I feel probable to move for a better living environment." (Flat respondent 29, Age 22) |
| 26-35 years old | "I was born and raised in the community. I am very probable to move that I consider providing the children a better environment not seeing drag users when they grow up." (Lock respondent 22, Age 27) | |
| 36-45 years old | "As a second generation born and raised here, I believe that moving away is beneficial for my 70 years old father practically, reduce the problems and make life easier." (70 rai respondent 1, Age 40) | |
A semi-structured questionnaire at the end of the survey aims to interact with residents and ask about essential factors while considering relocating. Most residents favour the housing environment and proximity to the facility. However, some residents voice their perception of which job location is the most critical factor when making relocation decisions. Table 2 summarizes respondents’ quote job location as the most significant factor.
| Job location impacts | Respondent quotes | |
|---|---|---|
| Very improbable and improbable to move | Proximity to job location |
"I am a street cleaner working for Gotomo in the morning and being a motorcycle taxi in the afternoon. I feel very improbable to move because this community is close to my job location." (70 Rai respondent 21) "Because of the cheaper rental and close to job location, I feel improbable to move. I appreciated getting much help from the neighbors in the community due to the limitation, I lost both hands." (70 Rai respondent 22) |
| Losing job location | "I have approximately 50 close friends since living in the community for 20 years. I think job location is most important. If the foundation is not here anymore, I cannot continue to teach those kids." (Lock respondent 3) | |
| Probable and very probable to move | Job opportunity | "I have lived in the community for three years and feel probable to relocate. From my point of view, job location, monthly income, inflation, and language skills create more opportunities and job choices." (70 Rai respondent 9) |
| Move in due to job location nearby |
"The reason why I moved here is that my job location is close to this community. I feel probable to move as I live here for two years." (70 Rai respondent 20) "I share the flat with my close friend and have lived in the community just one year because my job location is nearby." (Flat respondent 2) |
|
| Proximity to job location | "I am the house owner living in the community for 16 years. I feel probable to move, and if it is a must to choose compensation, I will relocate to a condo nearby, because it is easy to my workplace." (Flat respondent 10) | |
Nearly half of the respondents are working within 1 km of the community. The tendency of job types such as working at home and selling goods online is likewise blooming in the modern era of the Klong Toei community. Inevitably, residents working within the community are likely to have a greater proportion of saying they are improbable to relocate. However, on the other hand, some residents show a different degree of willingness to relocate because the flexibility of working at home means they can work elsewhere. Furthermore, the type of jobs in the retail and services sector is more improbable to relocate for the 70 Rai, followed by the Flat. Table 3 represents the job types of respondents within 1km of the community.
| Type of job | Respondent quotes | |
|---|---|---|
| Very improbable and improbable to move | Business | "I am doing online selling business. As a house owner, I feel very improbable to move because I just rebuild my house for a better living environment." (Lock respondent 27) |
| Retail and services |
"My job is work for hire and working in a motos company at soi 35. I feel very improbable to move because I live and work in the same community” (70 Rai respondent 8) "I think the market nearby makes life easier as I am selling food in the market so that I feel improbable to move.” (Flat respondent 12) |
|
| Probable and very probable to move | Retail and services | "I am selling goods on Facebook and will probably move back to my hometown. I suggest PAT offer more options for compensation due to some minority groups trying to negotiate a higher cash compensation." (Lock respondent 14) |
| Professional | ""I feel probable to move although I was born and raised in the community for 38 years. As a tattoo technician, my home is the workplace as a studio." (Lock respondent 16) | |
| Private office | "I am an outsource office worker related to PAT. Since I was born and living in the community for 22 years, I feel probable to relocate." (70 Rai respondent 28) | |
The spatial autocorrelation tool results suggest that the job location pattern at each feature location is spatial independent. In other words, there is no relationship between residents’ job locations. Figure 8 indicates that Moran’s Index is 0.085, close to zero, implying that the residents’ job locations are spatially independent. Moreover, the pseudo p-value is 0.001, indicating statistical significance.
As seen in Table 4, the value of the variable job location in the distance is 16.347, more significant than the critical value of 12.592. Additionally, asymptotic significance (2-sided) reveals the figure 0.012, smaller than the p-value 0.05, indicating job location in the distance has a statistically significant relationship with the dependent variable, the willingness to relocate at the 95% confidence interval. Moreover, the significance at the 90% confidence interval reveals that variables of elderly in the household, type of job, and monthly household income are statistically significant with the willingness to relocate.
| Variables | Pearson Chi-Square | ||
|---|---|---|---|
| Value | df | Asymptotic significance (2-sided) | |
| Age | 13.063 | 12 | 0.364 |
| Gender | 3.437 | 3 | 0.329 |
| Marital | 8.751 | 12 | 0.724 |
| Household member | 2.196 | 9 | 0.988 |
| Children in household | 2.753 | 6 | 0.839 |
| Elderly in household | 11.127 | 6 | 0.085* |
| Education attainment | 8.968 | 9 | 0.440 |
| Type of job | 18.688 | 12 | 0.096* |
| Monthly household income | 15.933 | 9 | 0.068* |
| Job location in distance | 16.347 | 6 | 0.012** |
| Type of tenure | 4.117 | 9 | 0.904 |
| Duration of living | 16.756 | 12 | 0.159 |
| Relatives in community | 1.163 | 3 | 0.762 |
| Close friends in community | 0.922 | 3 | 0.820 |
| Note. * significant at the 90% confidence interval. | |||
| ** significant at the 95% confidence interval. | |||
The hierarchical cluster analysis used Ward’s method to minimize the sum of square errors. At the same time, Euclidean works in lower dimensions to determine a better distribution of clusters to describe the phenomenon of different groups. Figure 9 illustrates the dendrogram generated by the hierarchical K-means clustering process defined into five branches.
Table 5 displays K-Means clustering algorithm results. Four out of five live in 70 Rai, while one is from Flat. Despite cluster one, single males are likely willing to relocate, while the other four clusters show an improbable and very improbable willingness to relocate. The household members are between three to five persons, four out of five including one child, while the elderly appear in two clusters that indicate different perceptions of willingness to relocate. Regarding education attainment, two out of five graduated from senior school or vocational degree and high school, respectively, while one attained a university or higher vocational degree. Moreover, it is worth noting the similarity that four of the five clusters work in the logistics and transportation sector, earning a monthly household income between 20,000-50,000 Baht, and the type of tenure has no contract with the landowner. However, the duration of living varies in each cluster, and the number of relatives and close friends living in the community indicates a more outstanding range between clusters.
Hence, according to the results of algorism from the K-means cluster, five clusters have been defined. Namely, Cluster 1: Gen Z single male, Cluster 2: Sub-renter work within 3 km, Cluster 3: Rooted married female, Cluster 4: Educated male work within the community, Cluster 5: Second generation work in retail and services sector. Furthermore, in comparison of each cluster, a significant similarity between Gen Z single male and Sub-renter work within 3 km both represented most cluster members that showing community, gender, children in the household, educational attainment, monthly household income, and job location in the distance are highly consistent. However, age and marital status indicate a significant disparity in willingness to relocate.
Regarding rooted married females, the monthly household income is the lowest between 5000-20,000 Baht, but the job location distance is 4.69 km. Moreover, the significant extended period of 46.1 years also depicts the attribute of rooted married females who have lived for the most extended years. Lastly, educated males working within the community and second-generation work in the retail and services sector represented the minority of cluster members who have similar characteristics on monthly household income, type of tenue, the closest job location within 1 km, a considerable number of close friends and relatives in the community, and the degree of very improbable willingness to relocate. Furthermore, the different ages, marital status, household component, type of job, and contractive educational attainment distinguish the substantial differences between the two clusters.
By summarizing the characteristics of the five clusters, it is instinctive to discover the difference between one and another. For example, a Gen Z single male is young and single in a 5-household member family, showing a possible willingness to relocate. In contrast, a sub-renter working within 3km is a married male living in a 3-household member family for 20 years, showing an improbable willingness to relocate. Both represent most residents in the Klong Toei community. On the other hand, rooted married females and educated males work within the community representing the same 5-household member family. However, different components of living with two elderlies likely lead to the more vital improbable willingness to relocate. Finally, second-generation work in the retail and services sector indicates that a unique case of second-generation residents prefers to stay in the community where they were born, raised, and work within the community.
| Name of cluster | Gen Z single male | Sub-renter work within 3 km | Rooted married female | Educated male work within the community |
Second Gen work in retail & services sector |
|---|---|---|---|---|---|
|
Cluster 1 (N = 41) |
Cluster 2 (N = 31) |
Cluster 3 (N = 13) |
Cluster 4 (N = 4) |
Cluster 5 (N = 1) |
|
| Community | 70 Rai | 70 Rai | 70 Rai | 70 Rai | Flat |
| Gender | Male | Male | Female | Male | Female |
| Age | 26 | 44 | 50 | 43 | 22 |
| Marital status | Single | Married | Married | Married | Single |
| Household member | 5 | 3 | 5 | 5 | 4 |
| Children in household | 1 | 1 | 1 | 1 | 0 |
| Elderly in household | 1 | 0 | 0 | 2 | 0 |
| Education attainment | Senior school or vocational degree | Senior school or vocational degree | High school |
University or higher vocational degree |
High school |
| Type of job | Logistics and transportation | Logistics and transportation | Logistics and transportation | Logistics and transportation | Retail and services |
| Monthly Household income | 20,000-50,000 | 20,000-50,000 | 5000-20,000 | 20,000-50,000 | 20,000-50,000 |
| Job location in distance (km) | 2.45 | 2.24 | 4.69 | 0.66 | 0.19 |
| Type of tenure | No contract | Sub-renter | No contract | No contract | No contract |
| Duration of living | 11.7 | 22.9 | 46.1 | 31 | 20 |
| Relatives in community | 6 | 4 | 10 | 7 | 100 |
| Close friends in community | 5 | 6 | 6 | 63 | 50 |
| Willingness to relocate | Probable | Improbable | Improbable | Very improbable | Very improbable |
In addition, Figure 10 demonstrates the scatter plot of duration of living by age by clusters. Despite the single case cluster 5 being born and raised in the community, it generally displays the distribution of scatter plots on the relationship between age and duration of living by five clusters, which intuitively convey the older generation and longer duration of living, delivering an improbable willingness to relocate.
From the cluster analysis results, it is noteworthy to comprehend further the traits and characteristics of resident works in the logistics and transportation sector. Thus, twelve of 20 samplings are categorized in the logistics and transportation sector, while four respondents work for hire, two are professional, and two are in private office sectors, all work related to PAT. As shown in Figure 11, the individual characteristics by willingness to relocate, indicate that the age group between 17 and 25 is more willing to relocate. However, educational attainment indicates a considerable discrepancy, which can conclude that the lower the educational level, the higher the willingness to relocate.
Regarding household characteristics, Figure 12 depicts household characteristics by willingness to relocate. A household member of 8- 10 people is more likely to move, and a household with children is slightly more likely to move. In contrast, households having elderly deliver a significantly improbable relocation. Moreover, income does not suggest the difference in whether to relocate. As one crucial indicator, duration of living tells an impressive phenomenon: residents who live less than ten years and 21-30 years are likely to move. However, relatives and close friends in the community do not reveal a distinct position corresponding to the willingness to relocate. On the other hand, the type of tenure a sub-renter has more willingness to move compared to others, such as house owner or benefit house offered by the company toward a low willingness to move.
Regarding the occupational attribute, Figure 13 illustrates occupational attributes by a willingness to relocate. A compelling result is that residents’ job location between 3-6 km is higher probability of moving. Nevertheless, those in the logistics and transportation sectors are most likely to move, which implies that their job attributes are mobile, and the relocation plan may not affect their livelihood. In turn, residents working in the retail and services sector work for hire for PAT, indicating a profoundly improbable relocation.
To sum up, investigating the residents’ characteristics in logistics and transportation and jobs related to PAT (N = 20), consistent with content analysis findings, indicates that the age of 17-25 is most willing to relocate. On the contrary, household living with the elderly delivers an influential improbable willingness to relocate, which aligns with the statistical test indicating that residents living with the elderly have a statistically significant relationship with willingness to relocate. Moreover, the results depict education attainment in a discrepancy status that the lower the education level, the higher the willingness to move. In addition, the type of tenure as a sub-renter and job location between 3-6 km possesses more willingness to move.
The informal settlement is often doomed to symbolize the degree of social and economic development. Since the Thai government initiated to tickle informal settlement issues, the National Housing Authority (NHA) was established to support disadvantaged groups by building affordable housing and intervening to secure tenure and provide the infrastructure. Moreover, the government institution Community Organization Development Institute (CODI) purpose of the Bann Mankong project has been implemented nationwide and recognized as successful progress by providing the urban disadvantaged groups with access to housing security via subsidies and offering housing loans directly to the community management (Boonyabancha, 2005a). However, the financial subsidy technique may only work in some informal settlements because the landowners, mainly settlements occupying governmental land, have further goals to develop their land after the expiration of land-sharing agreements.
This study demonstrates the relationship between residents’ job locations and examines the factors influencing residents’ willingness to relocate from the Khlong Toei community. The results indicate that job locations are spatially independent, and in response to the research question, job location has a statistically significant relationship corresponding to residents’ willingness to relocate. Moreover, the content analysis finding supports the theory that generation disparity is momentous in relocation decisions, indicating that the younger the generation, the higher the probability of relocating. Most residents are likely willing to relocate when quoting the importance of job location. In contrast, the type of job within 1 km of the community, in 70 Rai, residents who work in the retail and services sector possess a lower willingness to relocate, which in turn, work in the private sector show a higher willingness to relocate.
Furthermore, the cluster analysis technique interprets that residents who work in logistics and transportation have a contradictive status. The outcome of the five clusters provides insight into comparing residents’ perceptions. It describes Gen Z single males who are likely willing to relocate. At the same time, sub-renter work within 3 km represents most residents delivering an improbable willingness to relocate.
Slum resettlement is a broad subject during the urban development process. The compensation scheme often faces controversial situations, such as providing houses, land, or cash. It seems challenging to meet the consensus between landlord and residents due to consideration of the size of the housing unit, proximity to the city, or fixed amount of money is likely not fit various socioeconomic residents. Hence, this study aims to examine the relationship between residents' relocation decisions and highlights that residents’ job location reflects their willingness to relocate. Based on the findings, consideration should be addressed to residents’ job locations to stimulate the motivation for relocation. Recommendations for future study can be addressed in evaluating compensation forms, leaving room for studying the failure of compensation schemes and how to formulate a practical resettlement strategy. Ultimately, the governmental agencies and institutions direct the successfulness of relocation plans that says redevelopment policy reviews can benefit the informal resettlement a helpful perspective when counting on governmental intervention.
Conceptualization, S.H.T. and S.A.; methodology, S.H.T, N.P.1, and S.A.; software, S.H.T. and S.A; investigation, S.H.T. and S.A.; resources, S.H.T.; data curation, S.H.T. and S.A.; writing—original draft preparation, S.H.T. and S.A.; writing—review and editing, S.H.T., N.T. and S.A.; supervision, N.P.1, N.P2, and S.A. All authors have read and agreed to the published version of the manuscript.
The authors declare that there are no conflicts of interest regarding the publication of the paper.
We thank the Klong Toei residents who cooperated with the questionnaire survey and interview.
