Residential Segregation Research Articles

Racial differences in residential mobility after the Flint Water Crisis: A survival analysis.

Racial differences in residential mobility after the Flint Water Crisis: A survival analysis.

1267-P: Prenatal Racial/Ethnic Residential Segregation and Postpartum Diabetes in a 10-Year Cohort Study

Introduction and Objective: Postpartum diabetes increases cardiometabolic risk and may occur among individuals with and without gestational diabetes (GDM). The association of racial/ethnic residential segregation with postpartum diabetes is unclear. Methods: This cohort study followed 431,104 pregnant individuals for up to 10 years (mean ± SD 5.2 ± 3.4) at Kaiser Permanente Northern California. Cox regression estimated adjusted hazard ratio (aHR) of postpartum diabetes by prenatal racial/ethnic segregation via the Getis-Ord Gi* statistic. Results: Postpartum diabetes incidence was 2.5% in Asian/Pacific Islander (API), 2.6% in Black, 2.6% in Hispanic, and 0.8% in White. Black segregation was associated with a higher risk of postpartum diabetes in Black (aHR 1.37 [95% CI 1.08, 1.74]), Hispanic (1.12 [1.01, 1.24]), and White (1.20 [1.01, 1.42]) (Figure). Hispanic segregation was associated with a higher risk of postpartum diabetes in API (1.27 [1.12, 1.44]) and Hispanic (1.13 [1.01, 1.26]), while White segregation was associated with a lower risk. Associations were stronger in individuals without vs. with GDM. Conclusion: The higher risk of postpartum diabetes in Black and Hispanic segregated neighborhoods and lower risk in White segregated neighborhoods may inform structural targets for advancing maternal health equity. The stronger associations in individuals without GDM need investigation. Disclosure R.F. Chehab: None. A. Ferrara: None. L. Chen: None. A. Ngo: None. M. Greenberg: None. Y. Zhu: None. Funding NICHD (K99HD115836); NIMHD (R01MD018459)

Indoor thermal comfort in informal settlements: a case study of Mafalala, Maputo

ABSTRACT The rapid urbanization of African cities has intensified economic disparities, leading to residential segregation and inadequate housing conditions. This study examines indoor thermal comfort in the Mafalala neighborhood of Maputo, Mozambique, highlighting the severe impact of poor housing on residents’ well-being. For the first time, indoor temperature and humidity were measured and thermal comfort was assessed based on data collected from five dwellings during two campaigns in 2023. Results reveal significant thermal discomfort, particularly during the hot season. The widespread use of galvanized steel sheets for roofing and some walls exacerbates indoor temperatures, with some dwellings reaching up to 49°C during hot spells. While concrete block dwellings provide higher thermal inertia, they remain vulnerable due to poor insulation and ventilation. Using the Adaptive Comfort Model and Physiological Equivalent Temperature to assess thermal comfort and physiological stress, the analysis shows that dwellings are uncomfortable most of the time, exposing residents to moderate to extreme heat stress. Findings highlight the need for materials with higher thermal inertia, enhancing insulation, adequate ventilation, and shading solutions.

School segregation and native flight: evidence from school catchment area borders

Abstract Although these have been extensively theorized, determining the mechanisms that produce ethnic or racial segregation in schools and neighborhoods has proven difficult. We investigate one potential mechanism behind ethnic segregation: native flight from schools. In contexts where school enrollment is determined primarily by geographic proximity to schools, native or White parents with a preference for schools with low minority concentrations may choose to move away from schools with higher minority concentrations among students, contributing to both residential and school segregation. Using detailed, population-wide, geocoded register data on families and school catchment areas for elementary schools in Oslo, Norway, we investigate whether native parents move away from schools with higher shares of students with non-Western immigrant backgrounds. We first show that native-origin families systematically move away from schools with high shares of students with non-Western immigrant backgrounds. We employ a geographic regression discontinuity design, exploiting the fact that catchment area borders sort neighboring children into different schools; our results indicate that such moves may be causally linked to local school characteristics, not just their neighborhoods. This may contribute to segregation in schools and neighborhoods. However, the results are ambiguous and sensitive to model specifications, and more research is needed to draw firm conclusions.

Residential Segregation and Prolapse Surgery Complications in Older Black Women.

Residential segregation influences health outcomes. The objective of this study was to examine the relationship between residential segregation and prolapse surgery complications and readmissions among older Black women. This retrospective study included non-Hispanic Black women who underwent prolapse surgery from 2011 to 2018 in the Medicare 5% Limited Dataset. The primary outcome was 90-day complications. Other outcomes of interest were 30- and 90-day readmissions. We calculated the Index of Concentration at the Extremes (ICE) for each beneficiary's U.S. county of residence to measure geographic segregation by race, income, and both combined. We stratified the cohort into quintiles based on the ICE measures. Descriptive and comparative analyses were used to compare the demographic and clinical characteristics for each group. Poisson regression models were used to test the association between ICE measures and complications and readmissions. There were 872 Black women included in the analysis. Black women living in counties with the highest concentrations of Black residents had a 44% and 55% increased relative risk of 90-day complications compared to those in counties with the highest concentrations of White residents. Conversely, Black women living in the least segregated counties by race and income combined experienced 70% and 57% decreased risk of 30- and 90-day readmissions, respectively. The findings support our hypothesis that structural racism (measured by levels of residential racial and economic segregation) is associated with poorer outcomes after pelvic organ prolapse surgery. Further research is needed to identify neighborhood-level factors that contribute to and protect against inequities in postoperative outcomes after prolapse surgery.

Chlamydia and Gonorrhea Incidence and Residential Segregation: US Spatiotemporal Patterns (2013-2021).

Investigate how residential segregation is associated with chlamydia and gonorrhea incidence between 2013 and 2021. National-level secondary US data from 2013-2021 from the Centers for Diseases Control and Prevention Sexually Transmitted Infection surveillance dataset, American Community Survey, and Racial Segregation Index were analyzed using the Generalized Estimating Equation, and spatial regression. Analysis was divided into two periods (2013-2019; 2020-2021) to account for COVID-19 disruptions. Residential segregation was measured by dissimilarity index categorized into reference (<□0.25), moderate (0.26-0.50), high (0.51-0.75), and extreme (>□0.75) levels. Primary outcome measures were chlamydia and gonorrhea incidence rates. Residential segregation was the key independent variable with other social determinants of health covariates. 3058 counties within the contiguous United States were included within this study. Counties with missing data, and not within the contiguous United States were excluded. For chlamydia, from 2013-2019, segregation coefficients (i.e.,13.77 and 15.84 for moderate and high segregation) indicate that greater residential segregation is associated with higher chlamydia incidence rates (P<.0001). From 2020-2021, these coefficients increased (from 13 and 15 to 28.25 and 34.16), suggesting growing segregation-driven disparities. Gonorrhea followed a similar trend, with the coefficients increasing from 0.47 and 0.55 (P < .001) to 1.53 and 1.62 (P < .05), respectively. Spatial variation in the association between segregation and chlamydia incidence remained consistent, with stronger associations in the Southeastern, Midwest, and Western regions. Spatial variation in the association between segregation and gonorrhea incidence was more pronounced in the South and parts of the Midwest, with weaker associations in some Northern and Western regions. Residential segregation remained a substantial driver of chlamydia and gonorrhea transmission. The spatial patterns varied over time for both diseases. Further research should extend post-COVID-19 analysis to assess evolving relationships between residential segregation and STI incidence across U.S. regions. What is already known on this topic: Chlamydia and gonorrhea in 2023 were the most reported sexually transmitted infections in the United States, disproportionately affecting Black Americans. What this study adds: This study found that racial residential segregation was a substantial driver of chlamydia and gonorrhea transmission, especially during the COVID-19 pandemic. How this study might affect research, practice or policy: Study findings suggest that interventions aiming to reduce chlamydia and gonorrhea incidence rates in the United States should also include intervention activities that address adversities associated with racial residential segregation.

Racial and ethnic differences in biomarker testing for targetable alterations among patients with HR+ HER2- metastatic breast cancer (mBC).

1084 Background: In recent years novel therapies have been approved for patients (pts) with mBC and ESR1 , AKT1 , PTEN , PIK3CA , and gBRCA alterations. Given limited evidence on which patients are receiving standard of care, this study assessed racial and ethnic inequities in biomarker testing and the role of social determinants of health (SDOH) in explaining potential inequities. Methods: This study leveraged the US nationwide Flatiron Health electronic health record (EHR)-derived, deidentified database of &gt; 750 000 pts with BC. Adult female pts diagnosed (dx) with HR+ HER2- mBC between 1/1/2011, and 4/30/2024, with a geocodeable address were included. Testing rates for alterations in ESR1 , PIK3CA , AKT1 , PTEN , and gBRCA were measured over time from mBC dx using variables extracted from unstructured clinician documentation in the EHR using machine learning. Fine and Grey models accounting for competing risks were used to estimate subdistribution hazard ratios (HR) and 95% confidence intervals (CI) for biomarker access. Models were adjusted for covariates including age, stage, ECOG status, and dx year, followed by practice setting and area-level SDOH factors (ie, English language proficiency, residential segregation, vehicle ownership, urbanicity, and residence in medically underserved areas). Results: The cohort included 36 316 pts (61.5% non-Latinx [NL]-White, 6.1% Latinx, 9.7% NL-Black, 1.9% NL-Asian, and 20.8% NL-Other/Unknown). Overall, Asian, Black, and Latinx pts were less likely than White pts to undergo biomarker testing (adjusted HR [95% CI]: Latinx, 0.88 [0.82-0.95]; NL-Black, 0.87 [0.82-0.93]; NL-Asian, 0.87 [0.76-0.98]). Racial/ethnic inequities in overall biomarker testing were partially explained by SDOH factors. Specifically, the White-Latinx inequity in testing was mediated by residential segregation ie, association attenuated towards the null (mediated HR [95% CI], 0.94 [0.87-1.02]), limited English proficiency (0.92 [0.85-1.00]), and lack of vehicle ownership (0.91 [0.84-0.98]). Compared with White pts, NL-Black pts were less likely to be tested for ESR1 (HR [95% CI], 0.86 [0.77-0.95]) and PIK3CA (0.86 [0.80-0.92]). Latinx pts were less likely to be tested for PIK3CA (0.87 [0.80-0.95]) and this inequity was mediated by residential segregation (0.96 [0.87-1.05]) and limited English proficiency (0.92 [ 0.84-1.00]). Conclusions: Asian, Black, and Latinx pts were generally less likely than their White counterparts to receive biomarker testing after a mBC dx, especially for PIK3CA and ESR1 . SDOH factors explained some of these biomarker testing inequities. Equitable access to biomarker testing should be prioritized to ensure patients have access to the most effective therapies. Future research should examine whether racial/ethnic inequities in biomarker testing are associated with inequities in treatment and outcomes.

Association of allostatic load (AL) and residential segregation with breast biopsy outcomes after screening mammography.

1620 Background: Allostatic Load (AL) and residential segregation have been associated with the risk of breast cancer (BC). However, the independent effects of AL and measures of residential segregation (MRSs) on cancer detection and the false positive (FP) biopsy rate in a screening mammography population have not yet been assessed. Methods: We retrospectively identified women aged ³40 who underwent screening mammography between 1/1/2021-12/31/2021 and subsequent breast biopsy from the Mass General Brigham Biobank. We collected age and self-reported race/ethnicity. Each participant's zip code was geocoded to the corresponding census tract. We computed five MRS indices: Dissimilarity (DD), Isolation (BI), Delta (D), Absolute Centralization (AC), Spatial Proximity (SP). We collected the following biomarkers obtained within two years before the index screen: cardiovascular, metabolic, immunologic, renal lab values. AL was assigned one point for each lab value in the worst quartile and summed (continuous). We collected diagnostic breast imaging and biopsy encounters within 12 months after the index screen. Multiple imputation accounted for missing data. Multivariable logistic regression assessed age, race, AL and each of our MRSs association with cancer detection and FP rates. We applied Rubin's rules to estimate overall odds ratios (OR), confidence intervals (CI), and p-values for all covariates. Results: Of the 418 eligible women, 59.6% (N=249) had an FP biopsy, and 66.3% (N=277) had breast cancer, including cases of ductal carcinoma in situ. On average, women were 62 years old (SD=13); 85.6% White. DD was associated with a reduced risk of benign high-risk lesions (OR=0.69, 95% CI:[0.49-0.95]; p=0.025), and homogeneous, affluent census tracts—whether predominantly Black or White—were similarly protective (OR=0.85, 95% Cl:[0.72, 0.99]; p=0.041). MRS indices were linked to lower benign high-risk outcomes (e.g. SP, OR=0.53, 95% CI:[0.31, 0.91]; p=0.021). Age and race significantly predicted adverse events (AEs). Older age was consistently associated with increased AEs across all models (OR = 1.09, 95% CI:[1.00, 1.18]). Cancer detection also increased with age (OR = 1.20, 95% CI: [1.10, 1.30]; p &lt; 0.001). AL was significantly linked to cancer detection (OR = 1.13, 95% CI:[1.00, 1.28]). Conclusions: Some MRSs are associated with cancer detection and high-risk FP. AL remains associated with these cancer and high-risk FP, even accounting for these segregation measures. These factors may contribute to an increased risk of cancer, highlighting the significance of spatial and socioeconomic influences on screening outcomes. Clinical Relevance Statement: AL may serve as a biomarker to enhance biopsy selection following screen-detected mammographic abnormalities, potentially improving cancer detection rates.

Effect of allostatic load and measures of segregation on cancer detection and false positive rate after screening mammography.

565 Background: Allostatic load (AL), a cumulative stress measure and residential segregation (MRSs) have been associated with breast cancer (BC) outcomes. We assess the association of AL and MRSs on breast screening outcomes. Methods: From the Mass General Brigham Biobank, we retrospectively identify women aged ≥40, who underwent screening mammography from Jan 1, 2021 to Dec 31, 2021. We collected age, self-reported race/ethnicity, and zip code. Five MRSs were computed: Dissimilarity, Isolation (BI), Delta Index, Absolute Centralization, and Spatial Proximity. To compute AL, we assessed cardiovascular, metabolic, immune and renal lab values. AL was assigned one point for each lab value in the worst quartile and summed. We assessed any false positive (FP) and cancer diagnoses within 12 months after the index screen. Multiple imputation was performed for missing covariates. Multivariable logistic regression models were constructed to assess age, race, AL, and each MRSs association with cancer detection and FP rates. Rubin's rules were applied to estimate overall odds ratios (OR), confidence intervals (CI), and p-values for all covariates. Results: Of 13,754 women assessed, 1.2% (n=169) women received a cancer diagnosis. Most of the women were White (87.6%), 2.6% Asian, 5.3% Black; 1.5% self-identified as Hispanic; mean age, 64.4±11.3SD. Each point increase in AL increased the risk of cancer detection after screening mammography by 15% (OR=1.15,95% CI[1.06, 1.25],p=0.001). No association was detected between each MRS and cancer detection (all p&gt;0.22). BI, the expected proportion of neighbors belonging to the same group, was associated with FP rate, with BI &gt; 0.6 increasing the odds of FPs (OR=2.80[1.13-6.92],p=0.026). After adjusting for AL and MRSs, age and race were not significantly associated with cancer detection, but 5-year changes in age were associated with lower FP rate (OR=0.82[0.80, 0.84],p&lt;0.001). Conclusions: AL was associated with an increased risk of cancer detection after screening mammography after adjusting for age, race, and MRSs. The MRS BI above 0.6 was associated with an increased false positive rate. Further work is needed to confirm these observations. Clinical Relevance Statement: AL and MRS (i.e., BIS) may represent potential biomarkers for personalized mammographic screening for BC.

A matter of measurement? A Swedish register-based study of migrant residential segregation and all-cause mortality.

A matter of measurement? A Swedish register-based study of migrant residential segregation and all-cause mortality.

Impact of neighborhood disadvantage and residential isolation on cognitive outcomes in cancer survivors.

12056 Background: The Social Vulnerability Index (SVI), a composite measure of socioeconomic deprivation, household composition, minority status, and housing type and transportation, is a reliable marker of neighborhood disadvantage. Residential Segregation measures the degree to which a minority group is distributed differently than the majority group across census tracts. The isolation index is a measure of segragation that captures the extent to which minority members are exposed only to one another. These measures have been associated with poor survival in patients with cancer, yet studies on adverse outcomes in surviors are limited. Cognitive impairment is an adverse outcome highly prevalent in up to 40% and persists up to 5y in hematologic cancer survivors treated with blood or marrow transplantation (BMT). We postulate that neighbohood disadvantage and residential isolation may have a negative effect on cognitive outcomes in BMT survivors. Methods: We included 71 patients treated with allogeneic BMT, enrolled in the the Cognitive Training and Genetics Attitudes (cTAG) study. Objective cognitive function was measured using a comprehensive in-person battery of standardized neuropsychological tests. Standardized T-scores were categorized as deficit scores (range 0 to 5), and averaged across all tests to estimate a global deficit score(GDS), which was used as a measure of cognitive impairment. Residential addresses, geocoded and joined to corresponding Census block group and tract, were used to match patients with their corresponding SVI and residential segregation scores. Multivariable logistic regression models adjusted for age, sex, race, and clustering at the tract level were used to estimate associations with GDS. Results: Median age at study participation was 58y (IQR: 46, 63), 57.8% were male, and 16.9% were non-Hispanic Black. Primary diagnosis was 66.2% acute leukemia and 22.5 myelodysplastic/myeloproliferative neoplasms. Average time since BMT was 1.5y (SD = 1.2). Prevalence of global cognitive impairment was 19.7% (95%CI: 11.2-30.9). A high overall SVI score was associated with GDS (aOR = 4.7, 95%CI: 1.1, 20.2, p = 0.035). Vulnerability related to socioeconomic status (aOR = 4.9, 95%CI: 1.2, 20.2, p = 0.03) and housing type and transportation (aOR = 4.0, 95%CI: 1.1, 15.2, p = 0.04) were significantly associated with GDS. Higher residential isolation index ( &gt; 0.6) was significantly associated with increase in GDS (aOR = 5.4, 95%CI: 1.3, 22.5, p = 0.02) adjusting for age, sex, race, and clustering at the tract level. Conclusions: Cancer survivors residing in areas with higher indicators of social vulnerability and isolation are at increased risk of cognitive decline post-BMT. These findings highlight the overall need for dedicating appropriate resources and care planning especially for individuals surrounded by others from their same group within their residential areas.

Impact of social determinants of health on mortality in diffuse large B-cell lymphoma (DLBCL) using real-world data.

11097 Background: Treatment advances in DLBCL have led to remarkable improvements in patient outcomes. Social determinants of health (SDOH) can contribute to inequities in outcomes in multiple cancer types, and there is a paucity of studies evaluating their impact in DLBCL. Methods: We used the nationwide Flatiron Health electronic health record derived de-identified database and included adults with a confirmed diagnosis of DLBCL from 2011-2024 to evaluate the association between SDOH and real-world overall survival (from time of initial treatment). Area-level SDOH variables were derived from the Census Bureau’s American Community Survey and the 2019 AHRQ’s SDOH database. These census tract level measures were grouped into US population-weighted quartiles and evaluated across the following domains: economic, social (including racial segregation), neighborhood &amp; physical environment, and healthcare. We estimated adjusted hazard ratios (aHR) for the highest social deprivation quartile (least resourced areas) compared to the lowest quartile using Cox proportional hazard models, adjusting for age, sex, race/ethnicity, LDH, ECOG, presence of extranodal disease, stage, and cell of origin. Results: We included 6,855 patients in the analysis. After adjustment, residence in areas with the highest social deprivation was consistently associated with increased mortality across all SDOH domains compared with the lowest deprivation areas. A higher risk of death (&gt; 20%) was found for patients residing in predominantly Black vs. White neighborhoods, those residing in medically underserved areas, and areas with the lowest levels of private insurance. Similarly, residence in areas with the least access to the internet, computing devices, and cellular data plans was associated with increased mortality risk. Conclusions: Higher SDOH deprivation was significantly associated with increased mortality among patients with DLBCL, despite controlling for demographic and clinical factors. The SDOH influencing mortality ranged from socioeconomic and technological inequities to limited healthcare access and racial segregation. These factors may aid improved prognostication of DLBCL, and future studies should focus on developing interventions to mitigate SDOH-linked inequities in DLBCL. aHR 95% CI Domain Households that received food stamps/SNAP 1.15 1.02, 1.29 Economic Households with no internet access 1.15 1.02, 1.29 Neighborhood Households without a computing device 1.12 1.00, 1.26 Physical Households without cellular data plan 1.13 1.00, 1.26 Environment Medically underserved area 1.20 1.02, 1.40 Healthcare Population private health insurance (≤ 64) 1.24 1.10, 1.39 Population TRICARE/military/ VA insurance only (≤64) 1.17 1.05, 1.30 Population no health insurance (≤64) 1.18 1.06, 1.33 Residential Segregation Blacks (reference: Whites) 1.23 1.01, 1.51 Social

Longitudinal Associations From US State/Local Police and Social Service Expenditures to Suicides and Police-Perpetrated Killings Between Black and White Residents.

Despite documented inequities in suicide trends and police-perpetrated killing for Black US Americans, there is little research investigating how structural factors like government expenditures may drive these outcomes. This study examined associations from police and social services expenditures to later suicides and police-perpetrated killings for Black and White residents. This longitudinal study analyzed 2010-2020 US Census of Governments-tracked state and local government expenditures and Centers for Disease Control and Prevention (CDC)-tracked years of potential life lost (YPLL) to suicide and police-perpetrated killing. Dynamic structural equation models estimated 1- and 5-year lagged associations. Models adjusted for reverse associations (i.e., violent death to later expenditures) and state-level variables including Medicaid expansion, Black-White population, racial residential segregation, political representation, overall expenditures, state firearm policies, and firearm violence rates. For suicide, every $100 increase in per capita police expenditures was associated with 35 more YPLL 1 year later (γ=0.35, 95% credible interval [CI] 0.02-0.90) and 28 more YPLL 5 years later (γ=0.28, 95% CI 0.001-0.55) per 100,000 Black residents. For police-perpetrated killings, every $100 increase in per capita police expenditures was associated with 7 more YPLL 1 year later (γ=0.07, 95% CI 0.02-0.12) per 100,000 Black residents. As such, a $100 per capita increase in annual police expendiutres translated to 14,385 more YPLL to suicide, and 2,877 more YPLL to police-pepetrated killing, 1 year later for the United States' 41.1 million Black residents. There were no associations between police expenditures and outcomes for White residents. Conversely, every $100 increase in per capita housing and community development expenditures was associated with 29 fewer YPLL to suicide 5 years later per 100,000 Black residents (γ=-0.29, 95% CI -0.53 to -0.05). Every $100 increase in per capita kindergarten through 12th grade (K-12) education expenditures was associated with 4 fewer YPLL to suicide 1 year later per 100,000 White residents (γ=-0.04, 95% CI -0.07 to -0.01). Consistent with CDC recommendations to promote housing stability as suicide prevention, reducing police expenditures and increasing housing expenditures may decrease Black-White inequities in YPLL to suicide and police-perpetrated killing.

Residential Segregation and Epigenetic Age Acceleration Among Older-Age Black and White Americans.

Our study tests residential segregation as an explanation for biological aging disparities between Black and White Americans. We analyze data from 288 Black and White older-age adults who participated in Wave 6 (2019) of the Americans' Changing Lives study, a nationally representative cohort of adults in the contiguous United States. Our outcome of interest is epigenetic age acceleration assessed via five epigenetic clocks: GrimAge, PhenoAge, SkinBloodAge, HannumAge, and HorvathAge. Residential segregation is operationalized at the census tract level using the Getis-Ord Gi* statistic and multilevel modeling procedures that adjust for state-level clustering. We uncover three key findings. First, epigenetic age profiles are comparable among White respondents regardless of where they live. Second, Black respondents express roughly three years of accelerated epigenetic age (GrimAge), relative to White counterparts, regardless of where they live. Third, diminished education levels and homeownership rates, coupled with elevated levels of traumatic stress and smoking, explain why Black residents in segregated Black areas exhibit accelerated epigenetic age. However, these factors do not explain why Black respondents living outside segregated Black areas also exhibit epigenetic age acceleration. Our findings suggest residential segregation only partially explains why Black Americans tend to live shorter lives than White Americans.

Generative Multimodal Models for Social Science: An Application with Satellite and Streetscape Imagery

Although there is growing social science research examining how generative AI models can be effectively and systematically applied to text-based tasks, whether and how these models can be used to analyze images remain open questions. In this article, we introduce a framework for analyzing images with generative multimodal models, which consists of three core tasks: curation, discovery, and measurement and inference. We demonstrate this framework with an empirical application that uses OpenAI's GPT-4o model to analyze satellite and streetscape images ( n = 1,101) to identify built environment features that contribute to contemporary residential segregation in U.S. cities. We find that when GPT-4o is provided with well-defined image labels, the model labels images with high validity compared to expert labels. We conclude with thoughts for other use cases and discuss how social scientists can work collaboratively to ensure that image analysis with generative multimodal models is rigorous, reproducible, ethical, and sustainable.

(In) visible spaces as fractals of an ethnically inclusive city. Case of Siliguri- The corridor to India

Several cities have experienced residential segregation, and common interaction spaces have come to the rescue. However, little is known about ethnic (in) visible interaction spaces in India. By extracting data from the electoral roll assisted by primary survey, this study analyses ethnic interaction spaces across three differing wards in Siliguri City. The ethnic mapping reveals the relationship between spatial interaction features and settlement. To our awareness, this is the first spatial level study assessed in India that highlights planning ethnically fitting spaces grounded on residents’ lived practices.

Exploring the Impacts of Spatial Contexts on the Life Course Trajectory Status of Vulnerability

ABSTRACTThe aim of this study is to investigate in what ways the spatial context matters for the life course trajectory status in terms of vulnerability. In particular, it explores the impacts of spatial contexts aggregated from life course trajectories. It uses a longitudinal micro‐dataset, 1990–2019 from Statistics Sweden, to analyze the relationship between geographical context constructed by aggregated life course trajectories, and individual life course trajectories. A latent class analysis (LCA) is employed to identify life courses and examines how these trajectories are influenced by individualized neighborhoods. The findings show that spatial context plays a significant role in shaping individuals' life course trajectories of vulnerability: (1) being in any sort of trajectory of vulnerability, (2) four transitional categories of vulnerability. Residing in a context of Families in distress was associated with higher odds of experiencing vulnerability. Other factors such as country of birth, gender, earlier vulnerability also impact life course trajectory status. This study contributes to answering the question if poor neighborhoods make their residents poorer, posed by Jürgen Friedrichs in the late 1990s. The study goes beyond merely measuring cross‐sectional single variable residential segregation patterns as context, offering valuable insights into consequences, supporting planning for geographic equality of opportunity.

Mapping Urban Divides: Analyzing Residential Segregation and Housing Types in a Medium-Sized Romanian City

This study investigates residential segregation and housing types in Craiova, Romania, with a particular focus on the disparities shaped by historical and contemporary urban developments. Using collected data from former hostels built for young workers during the communist era, this research maps and analyzes the spatial distribution and living conditions of these housing types at a neighborhood level. Key metrics such as the number of inhabitants, the surface area of rooms, the current occupancy rates, and the number of unoccupied rooms were collected. Additionally, residential segregation is measured using indices of dissimilarity, isolation, exposure, concentration, and centralization, providing a comprehensive view of the socio-spatial divides within the city. The findings indicate significant disparities between these buildings with unsuitable living conditions and the newer residential developments, revealing a clear urban divide. No differences have been identified in terms of access to urban services like education, health, green areas, banks, or supermarkets, despite the appropriate location differences being noted in access to water and gas supply, and internet services. This study contributes to the understanding of how housing types and access to services in Craiova shape patterns of residential segregation, and it suggests policy interventions aimed at mitigating the negative impacts of these urban divides.

Socio-spatial inequalities in urban mobility: the immigrant-native travel time gap in German cities – a mixed method study

ABSTRACT The study ascertains to what extent immigrants in major German cities spend more time on daily mobility and attempts to identify the underlying reasons for this phenomenon. High travel time expenditures have detrimental impacts on health and well-being, employment, and civic engagement besides other areas of life. Daily mobility disadvantages can be linked to residential segregation, which indeed has increased considerably in German cities in recent decades. We hypothesise, that living in disadvantaged neighbourhoods that provide lacking access to public transportation and local services contributes to higher travel time expenditures of immigrants in German cities. We utilise the representative Mobility in Germany 2017 survey, which provides detailed information at the individual- and household-level (N individuals = 54 , 259 ), enriched with address-level neighbourhood data. Results of hierarchical regression models indicate that immigrants invest a greater amount of time in daily mobility related to commutes, errands, and care work for equivalent distances travelled. However, neither car-ownership, individual transport choices, living in disadvantaged neighbourhoods, nor socio-economic differences fully account for the travel time expenditure gap to natives. Semi-structured qualitative interviews (N = 29 ) suggest that immigrants make compromises in their daily mobility to protect themselves from perceived unsafe situations, which increases their travel time expenditures.

The Economic Urban Divide: A Detailed Study of Income Inequality and Segregation in Dutch Urban Areas (2011–2022)

Abstract Research on segregation and economic inequality is often limited to major capitals and conurbations, neglecting smaller cities. This oversight can lead to public policies based on insights that may not be universally applicable. Leveraging geo‐coded register data, this study addresses this problem in the case of the Netherlands by computing income inequality and residential segregation annually in all urban areas from 2011 to 2022. Contrary to most literature, this paper shows that inequality and segregation have remained stable or decreased in most cases. In addition, when looking at how income is distributed among social segments, how segregated they are, and at which geographical scale segregation occurs, we find significant variation between urban areas. More unequal urban areas also tend to be more segregated, but patterns vary, and the same segregation levels can coexist with diverse inequality metrics. Four groups of urban areas are identified through a cluster analysis.