The Role of Residential Segregation in Racial Health Disparities During Childhood

The Role of Residential Segregation in Racial Health Disparities during Childhood

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Bethany L. Boettner, M.A.

Graduate Program in Sociology

The Ohio State University

2011

Dissertation Committee:

Cynthia G. Colen, Co-Advisor

Zhenchao Qian, Co-Advisor

John Casterline

Reanne Frank

Copyrighted by

Bethany Lynn Boettner

2011

Abstract

As individual-level explanations have generally proven insufficient for explaining racial differences in health, researchers have increasingly turned to residential contexts as a source of population health disparities. Residential segregation is a key defining characteristic of the American landscape and a powerful force for explaining racial inequality. Furthermore, segregation has been found to be a key explanatory mechanism for understand racial disparities in health through its impact on individual and neighborhood level socioeconomic status, discrimination and race-related stressors, and neighborhood quality. In this dissertation, I draw on life course perspectives and stress process models of health to explore how the distribution of social contexts along racial divisions are associated with physical health outcomes among children.

Using longitudinal data from the Early Childhood Longitudinal Survey

Kindergarten Class of 1998-1999 data (ECLS-K), I address three specific research questions regarding the role of residential segregation on child well-being. First, what is the relationship between residential segregation and child health outcomes such as obesity, asthma, and parent-rated health, and are the effects of segregation cumulative over the early life course? Second, how do neighborhood racial/ethnic tensions and social cohesion work in tandem to influence child health, and do these impacts vary by race?

Third, does residential segregation exacerbate the negative effects of family stressful events on health outcomes for school-aged children? ii

I find that the relationship between segregation and physical wellbeing varies by race, and cumulative measures of segregation are more powerful predictors of childhood health than indicators captured at a single point in time. Moreover, parental involvement in school programs, as a measure of social cohesion, is protective against negative health outcomes for White and Hispanic children, even in neighborhoods characterized by high levels of racial tensions, and is more predictive for some diseases (obesity) than others

(asthma). Finally, residential segregation measures do not exacerbate the negative effect of stress events on child health. Given these findings, it is apparent that the neighborhood racial context plays a key but complex role in producing and acerbating child health inequalities during the early part of the life course and may set the stage for the further entrenchment of these disparities as individuals age.

iii

Dedication

Dedicated to my parents, Richard and Diane, who taught me the value of education.

Acknowledgments

First and foremost I must acknowledge Cynthia Colen, my co-advisor, who pushed me to finish my degree and helped give me the confidence that I could do it.

Cindy also provided much needed advice and knowledge on the entire process of research, from beginning to end. Her guidance made this dissertation possible. I also wish to thank Cindy for the opportunity to work with her as a research assistant and colleague for the last few years of my studies, as I‟ve learned a great deal about research and collaboration.

I would also like to thank Zhenchao Qian, co-advisor, and committee members

Reanne Frank and John Casterline for their suggestions, thoughts, and time in helping mentor me through this project. Thank you also the Initiative for Population Research and the Department of Sociology for providing travel support, office space for secure data, and an environment of excellent scholarship and collegiality.

I wish to thank fellow graduate students Jamie Lynch, Lauri McCloud, Alexa

Trumpy, Dan Carlson, Tom Maher, and Lauren Pinkus for their advice and friendship through this process, and for encouraging my love of good food, trivia, and Ultimate.

Thank you to my family, who has been there for me in more ways than I can possibly count. And to Andy Fisher, my boyfriend, who taught me how to ride a bike and who motivates me to be a better scholar, worker, and partner.

Vita

2000...... Stow-Munroe Falls High School

2004...... B.A. Sociology and French, The Ohio State

University

2007...... M.A. Sociology, The Ohio State University

2004 to present ...... Graduate Teaching Associate and Lecturer,

Department Sociology, The Ohio State

University

Publications

Nicholson, Lisa M., Patricia M Schwirian, Elizabeth G Klein, Theresa Skybo, Lisa Murray-Johnson, Bethany Boettner, Gina French, Judith A Groner. 2011. “Recruitment and Retention Strategies in Longitudinal Clinical Studies with Low-Income Populations.” Contemporary Clinical Trials 32:3 353-362.

Tsai, Shane F., Mira Trivedi, Bethany Boettner, and Curt Daniels. 2011. “Usefulness of Cardiovascular Magnetic Resonance Imaging to Detect Aortic Abnormalities After Repair of Coarctation of the Aorta.” American Journal of Cardiology 102:2 297-201.

Egan, Matthew J., Sharon L. Hill, Bethany L. Boettner, Ralf J. Holzer, Alistair B. Phillips, Mark Galantowicz, John P. Cheatham, and John P. Kovalchin. 2011. “Predictors of Retrograde Aortic Arch Obstruction After Hybrid Palliation of Hypoplastic Left Heart Syndrome.” Pediatric Cardiology 32:1 67-75.

Luce, Wendy A., Randall M. Schwartz, Wendi Beauseau, Peter J. Giannone, Bethany Boettner, John P. Cheatham, Mark E. Galantowicz, Clifford L. Cua. 2011. “Necrotizing vi

Enterocolitis in Neonates Undergoing the Hybrid Approach to Complex Congenital Heart Disease.” Pediatric Critical Care Medicine 12:1 46-51.

Tsai, Shane F., David P. Chan, Pamela S. Ro, Bethany Boettner, Curt J. Daniels. 2010. “Rate of Inducible Ventricular Arrhythmia in Adults With Congenital Heart Disease.” American Journal of Cardiology 106: 730-36.

Birnbaum, B., Berger, G., Fenstermaker, B., Rowland, D. G., Boettner, B., Olshove, V., Galantowicz, M., Cheatham, J. P. and Cua, C. L. 2010. “Echocardiographic Parameters that Predict Outcome in Aortic Atresia Patients Undergoing Comprehensive Stage II Procedure.” Congenital Heart Disease 5: 409-415.

Fields of Study

Major Field: Sociology

vii

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... vi

Table of Contents ...... viii

List of Tables ...... ix

Chapter 1: Introduction ...... 1

Chapter 2: Exposure to Segregation and Health Outcomes in Childhood ...... 14

Chapter 3: Racial and Ethnic Tensions in the School and Neighborhood Environment:

Consequences for Child Physical Health ...... 46

Chapter 4: Segregation, Family Stress, and Racial Disparities in Childhood Physical

Health ...... 72

Chapter 5: Conclusion...... 97

References ...... 103

viii

List of Tables

Table 1: Weighted Descriptive Statistics by First and Last Interview and Race ...... 40

Table 2: Odds Ratios from Logistic Regression of Segregation on Poor Health ...... 41

Table 3: Odds Ratios from Logistic Regression of Segregation on Asthma ...... 42

Table 4: Odds Ratios from Logistic Regression of Segregation on Obesity ...... 43

Table 5: Cumulative Exposure to Segregation by Race in Last Interview ...... 44

Table 6: Odds Ratios from Logistic Regression of Chronic Exposure to Segregation on

Health ...... 45

Table 7: Weighted Neighborhood, School, and Individual Characteristics, by Race...... 67

Table 8: Odds Ratios from Logistic Regression of Neighborhood and School Contexts on

Obesity ...... 68

Table 9: Odds Ratios from Logistic Regression of Neighborhood and School Contexts on

Asthma ...... 69

Table 10: Odds Ratios from Logistic Regression of Neighborhood and School

Interactions on Obesity ...... 70

Table 11: Odds Ratios from Logistic Regression of Neighborhood and School

Interactions on Asthma ...... 71

Table 12: Weighted Sample Characteristics: Health, Segregation, and Stress by Racial

Identity ...... 93

Table 13: Odds Ratios from Logistic Regression of Family Stressful Events on Health . 94

Table 14: Odds Ratios from Logistic Regression of Family Stressful Events and

Dissimilarity Index Interactions on Health ...... 95

Table 15: Odds Ratios from Logistic Regression of Family Stressful Events and Racial

Isolation Interactions on Health ...... 96

Chapter 1: Introduction

Childhood health is an important component of health and well-being across the life course (Case, Fertig, and Paxson 2005; Palloni 2006; Jackson 2009). Racial inequalities in health are relatively small in childhood compared to later adulthood.

However, disparities in child wellbeing tend to translate into substantial disparities in adult wellbeing due to the cumulative exposure to systematic social, economic, and political exclusion of racial minorities over time. Racial and ethnic minority children, especially Black and Hispanic children, disproportionately experience increased rates of morbidity, stemming from chronic conditions such as asthma, obesity, and diabetes

(Flores et al. 2008; Singh, Siapush, and Kogen 2010). Although immigrants tend to fare better than expected with regard to health relative to Whites given their relatively low socioeconomic status (Hummer et al. 2007; Markides and Eschbach 2005), immigrant

Hispanic children still face elevated odds of developing obesity and poor physical health in early life (Crosnoe 2006). Furthermore, racial gaps in health have not been eliminated.

Disparities in obesity have widened in the last decade (Singh et al. 2010), and no appreciable improvements have been seen in Black-White infant mortality ratios (Heron et al. 2009; Schempf et al. 2007). Poor child health can set in motion a lifetime trajectory of health problems, both mental and physical (Haas 2008). Understanding the causes of racial inequality in health is paramount to designing effective programs and policies

1 aimed at reducing these disparities, a key goal of the U.S. Department of Health and

Human Services Healthy People 2010 initiatives.

Furthermore, segregation has been hypothesized to be a key explanatory mechanism for understand racial disparities in health through its impact on individual and neighborhood level socioeconomic status, discrimination and stress, and neighborhood quality

(Acevedo-Garcia et al. 2003; Williams and Sternthal 2010). Previous studies have explored the links between segregation and health in adults and in birth outcomes, yet few studies address the role of residential segregation on health outcomes during childhood, and especially among Hispanics.

In this dissertation, I draw on life course perspectives and stress process models of health to explore how social contexts and race are associated with child physical health. I address three specific research questions regarding the role of residential contexts on child health. First, what is the relationship between residential segregation and child health outcomes such as obesity, asthma, and parent-rated health, and are the effects of segregation are cumulative over the early life course? Second, what are the effects of neighborhood racial and ethnic tensions and social cohesion on child health, and do they vary by race? Third, does residential segregation exacerbate the negative effects of family

2 stressful events on child health? In answering these questions, this dissertation contributes to a better understanding of how structural discrimination and individual characteristics interact to influence child health.

Existing models of racial disparities in health

Socioeconomic status (SES) differences across racial and ethnic groups are substantial. Although 21% of all children under the age of 18 were living in poverty in

2009, this estimate masks significant racial differences (U.S. Census Bureau 2009). A full one third of African American and Hispanic children live in poor families, but just 12% of White children living below the poverty line. However, even at similar levels of income Blacks and Hispanics have lower levels of wealth accumulation, homeownership, and are more vulnerable to foreclosure (Krivo and Kaufman 2004; Shapiro 2004; Flippen

2010). The fundamental cause perspective argues that socioeconomic status influences health through in the use of money, knowledge, resources, and social connections to protect health through a variety of mechanisms (Phelan et al. 2004). Therefore, individual level SES is thought to be a major determinant of racial disparities in health through the differential access to and resources that benefit health. Research has confirmed that socioeconomic status explains much of the Black-White disparity in health, especially in middle age (Hayward et al. 2000; Sudano and Baker 2005).

However, SES differentials do not fully account for racial disparities in birth weight, obesity, or asthma during childhood (Crosnoe 2006; Kimbro, Brooks-Gunn, and

McLanahan 2007; Teitler et al. 2007).

Two key theoretical perspectives focus on understanding health and well-being at the individual level that are fruitful for understanding racial and ethnic health disparities in context, stress process and life course perspectives. Stress process models of health have developed over time to explain how social status and social contexts converge to influence individuals‟ lives and their exposure to stressors, how they respond to those stressors, and how both stressors and the coping response influence physiological functioning (Pearlin 1999). Additionally, disadvantaged social statuses have been shown to decrease access to or efficacy of mediators and moderators of stress, such as social support and coping mechanisms, which result in suboptimal mental and physical health.

Thus, it is thought that differential exposure to stressors is a primary mechanism though which race, class, and gender inequality is transformed into health disparities (Turner

2003; Thoits 2010). Although stress research originally focused on only negative life events, over time stress process researchers have also highlighted the important role of everyday, ongoing conditions (Wheaton 1994). Both negative life events and chronic strains negatively impact health; furthermore, their combination exerts a cumulative effect over the life course (Pearlin 1999; Pearlin et al. 2005).

Life course perspectives on health stress the importance of cumulative risk processes in determining well-being (Kuh and Ben-Schlomo 1997). The weathering hypothesis asserts that political and social marginalization accelerates ageing, and as a result, African Americans experience earlier declines in health than White Americans that also accumulate over time to increase racial inequality in health over the life course

(Geronimus 2001). Accelerated ageing may be a response to the prolonged exposure to

4 and coping with stressors that results from racism and discrimination against Blacks

(Walsemann, Geronimus, and Gee 2008). This weathering process disproportionately affects minorities but especially Black women, who experience multiple dimensions of disadvantage in social and economic attainment. Black women experience higher levels of hypertension and allostatic load, markers of increased stress and worse health

(Geronimus et al. 2006; Geronimus et al. 2007). Thus, research focused on explaining racial disparities in health needs to account for the ways in which stress and coping cumulatively affect minorities across the life course, and how differences emerge in the early stages of the life course.

Perceived racial and ethnic discrimination is one such marker of stress that varies greatly across race and is increasingly receiving the attention of sociologists and public health researchers of health. Negative social interactions, such as unfair treatment stemming from racist attitudes, contribute to racial disparities in health by elevating stress and taxing coping resources available to deal with these interactions (Schulz et al. 2000;

Williams and Sternthal 2010). Perceived discrimination has been linked to increased psychological distress among African American adults (Williams et al. 1997; Jackson et al. 1996; Schulz et al. 2006) and both African American and Hispanic children (Coker et al. 2009; Wong, Eccles, and Sameroff 2003). Discrimination experiences are also associated with increases in obesity, mortality, and unhealthy behaviors for African

American adults (Hunte and Williams 2009; Borrell et a. 2010).

However, individual experiences of racism cannot capture the wider effects of structural level institutional racism on health (Gee and Ford 2011). Williams and

5 colleagues (Williams and Collins 2001; Williams and Jackson 2005) argue that racial segregation is a fundamental cause of racial disparities in health in part because it represents the vast racial inequality across multiple dimensions of social life that is not directly rooted in individual attitudes and stereotypes. Although White avoidance of

African-American neighbors contributed to the emergence and to the persistence of

Black-White segregation levels (Zubrinsky and Bobo 1996; Quillian 2002), as Blacks moved in large numbers out of the South into the North after World War I during the

Great Migration in search of economic opportunities, housing discrimination also kept

Blacks out of predominantly White neighborhoods (Tolnay 2003). Effective institutional barriers to entry into these communities included restrictive covenants that denied deed transfers to non-White families as well as financial policies that restricted the ability of

Blacks to qualify for loans in White neighborhoods (Massey and Denton 1993; Pager and

Shepherd 2005). These forces resulted in urban segregated ghettos that experience high rates of unemployment, social disorganization, and crime (Wilson 1996). Therefore, the pervasiveness of African American residential segregation represent multiple levels of exclusion and discrimination, and wide ranging structural inequality that continues to have detrimental consequences for African American individuals.

In contrast to the negative underpinnings of residential segregation for African

Americans, research on ethnic enclaves suggest that living ethnic and immigrant neighborhoods may be beneficial to new immigrants because of cohesive social networks and the social support they are likely to provide to same-race and co-ethnic neighbors, suggesting Hispanic communities may be beneficial to Hispanic residents (Portes and

Zhou 1993). Ethnic communities can provide new immigrants with employment opportunities, reduced language barriers, knowledge about affordable housing and health care, and cultural and familial ties, all of which may help new immigrants adapt and acculturate to their new communities. Selective acculturation and strong ties to an ethnic community help buffer the second generation from downward socioeconomic trajectories and lead to improved chances of socioeconomic mobility (Zhou and Bankston 1998;

Hirschman 2001; Portes, Fernandez-Kelly, and Haller 2009). However, poor immigrants are more likely than their nonpoor counterparts to reside in socially disorganized and economically disadvantaged ethnic enclaves that are detrimental, not beneficial, to socioeconomic attainment in the next generation (Portes and Zhou 1993; Zhou 1997).

Furthermore in 2000, 27% of all Hispanic children lived in poverty, and rates are even higher among the first and second generations (36% and 28%, respectively) compared to third generation Hispanic children (23%) (Lichter, Qian, and Crowley 2005). For these particularly vulnerable children, immigrant and co-ethnic communities may not be as beneficial to health as has been previously expected.

Segregation has been associated with a wide range of negative health outcomes among African-Americans. Black-White dissimilarity indices are associated with increased African American infant and adult mortality (Cooper et al. 2001; Polednak

1996), and Black isolation is positively associated with low birth weight and poor self- rated health and obesity among adults (Subramanian et al. 2005; Chang 2006; Bell et al.

2006; Grady and Ramirez 2008). Black mothers in hypersegregated metropolitan areas, those that are segregated along multiple dimensions of residential patterns, experience

7 higher rates of pre-term birth than Blacks in hypersegregated areas, and in addition,

Black-White gaps in preterm birth are also larger in hypersegregated areas (Osypuk and

Acevedo-Garcia 2008).

However, the association between segregation and well-being has not been entirely negative among African Americans, and may partly reflect differences in which segregation dimension measures are used. For example, racial clustering was protective of birth outcomes among Blacks (Bell et al. 2006). Moreover, measures of residential evenness such as the dissimilarity index have not been as closely related to health outcomes as racial isolation; several studies do not find a link between Black-White dissimilarity and health among African Americans (Subramanian et al. 2005; Walton

2009). The dissimilarity index may have weaker theoretical link between evenness than isolation as it only measures the local residential patterns in relation to the wider MSA racial composition (Acevedo-Garcia et al. 2003; Bell et al. 2006 Kramer and Hogue

2009). Isolation may be a better representation of unequal experiences and be particularly detrimental to the extent that although it requires uneven residential patterns, high isolation represents many neighborhoods with high minority concentrations. Thus, although generally segregation is considered to be detrimental to the well being of

African Americans, existing research is not entirely consistent in these findings, and understanding how different measures of residential segregation are related to health is important for understanding the complex ways in which residential contexts affect well- being.

Furthermore, current evidence is inconclusive regarding the relationship between residential segregation and health among Hispanics. The association between segregation and health varies across racial/ethnic identity, immigrant status, and outcome measures, and most studies have used measures of racial composition rather than residential segregation indices. Although racial composition measures tap some of the complex residential patterns of Hispanics, they cannot account for wider spatial patterning that is shaped by housing market discrimination, socioeconomic status, and immigration.

Several studies find positive health effects of concentrated Hispanic neighborhoods, as expected by ethnic immigrant enclave perspectives. For example, higher Hispanic concentration is related to lower depression and mortality rates and better self-rated health among elderly Mexicans living in the Southwest (Ostir et al. 2003; Patel et al.

2003; Eschbach et al. 2004). Among children and adolescents, Hispanic concentration is positively associated with dietary habits overall (Lee and Cubbin 2002). However, the health effects of living among co-ethnics are not always beneficial for Hispanics, as expected by segmented or downward assimilation theories. Residential isolation is associated with worse physical health for Puerto Rican adults and increases depressive symptoms for Mexican adults (Lee and Ferraro 2007; Lee 2009). Hispanics are also at an increased risk of substance use and delinquency in neighborhoods with high concentrations of poor Hispanics, especially for native-born youths (Frank, Cerda, and

Rendon 2007). In sum, the findings regarding residential segregation and health among

Hispanics are not yet conclusive, and more research is needed to untangle the role of

9 residential segregation using indices of segregation while examining health across the life course.

Project Summary

Much of the work on residential segregation and health outcomes documents the magnitude of the association. The studies that work to explain the relationship most often use mediating models, whereby neighborhood socioeconomic status, neighborhood stressors, and individual level exposures mediate, or explain the link between residential segregation and health (Kramer and Hogue 2009). For example, Chang, Hillier, and

Mehta (2009) find that neighborhood disorder mediates in part the relationship between

Black racial isolation and women‟s weight and obesity. While it is theoretically important to consider the ways in which these factors are pathways through which residential segregation affects health, segregation is also likely to exacerbate the effects of individual stressors to health. Racially isolated neighborhoods lead to community disinvestment and suboptimal socioeconomic and social resources available to residents, key coping mechanisms that buffer the effect of stressors. To this end, then, segregation may in fact act through a stress proliferation model that, due to these reduced community resources, increases the detrimental effects of individual level characteristics and stressors.

In this dissertation, I unite the literature on structural discrimination and individual stress process models while assessing the interactive effects of neighborhoods and individuals on health outcomes. Although much research has been dedicated to how neighborhood characteristics such as residential segregation influence health, few have assessed how these social contexts interact with and modify individual level

10 characteristics such as race and family stress to influence child health. In doing so, I evaluate three distinct but related research questions.

In the first paper (Chapter 2), I address the relationship between residential segregation and health for children, including Hispanics, and the effects of cumulative exposure to segregation. I investigate whether exposure to residential segregation is predictive of racial/ethnic inequalities in childhood health using three health measures that unequally affect minority children, obesity, poor parent-rated health, and asthma diagnosis, and two measures of segregation, dissimilarity and isolation. I test how the effects of residential segregation vary across health measures and by racial/ethnic identity. Finally, I also assess whether cumulative exposure to segregation during the early life course magnifies the effects of segregation on health among African American and Hispanic children.

The second paper (Chapter 3) explores how racial/ethnic tensions in a neighborhood may explain the link between residential segregation and health.

Racial/ethnics tensions in neighborhoods may be an indicator of race-related stressors, excessive levels of community mistrust, and/or discrimination, all of which have been separately linked to both residential segregation and negative health outcomes. I investigate the extent to which residential racial/ethnic tensions in the neighborhood are detrimental to children‟s overall health and likelihood of obesity, especially for African

American and Hispanic children. I also evaluate whether social support and community social cohesion buffer the health effects of racial/ethnic tensions for children.

The third paper (Chapter 4) explores another potential pathway by which residential segregation affects childhood physical health, through the added effects of family stress on the likelihood of obesity. Family life events such as residential moves, parental marital status changes, unemployment or illness create stress for the whole family, and exposure to family stress is higher for minority children, and especially in segregated neighborhoods. I explore whether segregation exacerbates children‟s vulnerability to family stress, and whether the effect differs across dimensions of residential segregation.

This dissertation adds to the growing literature on racial disparities in health first by applying existing models of exposure to segregation for child health outcomes, as well as developing our understanding of how cumulative exposure to segregation shapes health at the beginning of adolescence. Second, this dissertation also assesses the interactive effects of segregation and discrimination by individual level characteristics and stressors, applying a newer model of stress proliferation to the study of residential segregation. This dissertation is motivated by the goals of the Healthy People programs to eliminate health disparities between racial and ethnic groups, yet substantial gaps remain (Sondik et al. 2010). In order to understand the mergence and persistence of racial disparities throughout the life course, it is important to understand the complex ways in which a child‟s individual risk factors interact with their social and neighborhood contexts. For example, if the negative health effects of family stressful events on child health are exacerbated within the context of residential segregation, then policies design

12 to reduce the impact of either stress or segregation will not be sufficient in reducing child health disparities.

Chapter 2: Exposure to Segregation and Health Outcomes in Childhood

Health disparities are a persistent component of racial inequality in the United

States. Racial and ethnic minority children, especially Black and Hispanic children, disproportionately experience increased rates of morbidity, especially stemming from chronic conditions such as asthma, obesity, and diabetes (McDaniel, Paxson, and

Waldfogel 2006; Von Hippel et al. 2007; Rosenbaum 2008). Additionally, disparities in child wellbeing tend to translate into substantial disparities in adult wellbeing due to the cumulative exposure to systematic social, economic, and political exclusion over time.

Research on health and wellbeing has increasingly focused on metropolitan areas and neighborhoods as influential contexts that create and maintain these disparities (Williams and Jackson 2005; Diez Roux and Mair 2010). Yet few studies assess how cumulative exposure to disadvantaged, segregated neighborhoods over the life course influences health, particularly among children.

Segregation has been associated with a wide range of negative health outcomes among African-Americans. Segregation increases African American adult mortality

(Jackson et al. 2000; LaVeist 2003), and racial isolation is positively associated with poor self-rated health, obesity, and low birth weight (Subramanian et al. 2005; Chang

2006; Bell et al. 2006; Grady and Ramirez 2008). Black mothers in hypersegregated metropolitan areas experience higher rates of pre-term birth than Blacks in 14 hypersegregated areas, and in addition, Black-White gaps in preterm birth are also larger in hypersegregated areas (Osypuk and Acevedo-Garcia 2008). However, the association between segregation and well-being has not been entirely negative among African

Americans. Racial clustering was protective of birth outcomes among Blacks (Bell et al.

2006), and in contrast to previous findings on mortality, Inagami et al (2006) find that living in predominantly Black neighborhoods reduces mortality levels relative to Blacks living in less isolated areas. Although generally segregation is considered to be detrimental to the well being of African Americans, existing research is not entirely consistent in these findings.

Current evidence is even more mixed regarding the role of segregation in determining the health of Hispanics. The relationship between segregation and Hispanic health varies across racial/ethnic identity, immigrant status, and health outcome measures. Several studies find positive health effects of living in Hispanic neighborhoods. Higher neighborhood Hispanic concentration was related to less depression and mortality and better self-rated health among elderly Mexicans living within the Southwest (Ostir et al. 2003; Patel et al. 2003; Eschbach et al. 2004). Among children and adolescents, neighborhood Hispanic concentration is associated with healthier dietary habits overall (Lee and Cubbin 2002). However, others have found negative effects of Hispanic segregation. Residential isolation reduces physical health for

Puerto Rican adults and increases depressive symptoms for Mexican adults (Lee and

Ferraro 2007; Lee 2009). Neighborhoods with high concentrations of poor Hispanics also are related to increased health risk behaviors such as substance use and delinquency

15 among Hispanic youth, particularly for U.S.-born Hispanics (Frank, Cerda, and Rendon

2007).

Research on links between segregation and health largely ignores childhood well being and cumulative exposure to segregated neighborhoods. Life course processes are important for determining health and wellbeing, as early life sets in motion health trajectories into adulthood and later life, especially chronic conditions (Kuh and Ben-

Schlomo 1997). Until recently, work on neighborhood effects did not adequately addressed early life neighborhood exposures in longitudinal research (Diez Roux and

Mair 2010). Children may spend more time within neighborhoods attending schools and playing in local backyards and parks than their parents, who may leave the neighborhood more often for work or other activities. Finally, although research on childhood outcomes has examined the cumulative effects of exposure to neighborhood poverty (Jackson and

Mare 2007; Crowder and South 2011), few have assessed the cumulative effects of exposure to segregated neighborhoods, either in children or adults (see Kramer and

Hogue 2009 for a review). Exposure to segregation is often a long-term neighborhood condition, particularly for urban and poor minority children, which may result in cumulative disadvantages to both socioeconomic and health outcomes. Understanding the effects of segregation on children‟s health and wellbeing is paramount in understanding the early determinants of health disparities.

I address the relationships between segregation and health for White, African

American, and Hispanic children, including the effects of cumulative exposure to segregation using data from five waves of the Early Childhood Longitudinal Survey

Kindergarten Class 1998-1999 (ECLS-K). These data spans 8 years, from 1998 to 2007. I investigate whether exposure to residential segregation is predictive of racial/ethnic inequalities in childhood health using three health measures that unequally affect minority children, obesity, poor parent-rated health, and asthma diagnosis.

Segregation

Residential segregation is a key defining characteristic in the American metropolis and a powerful force for explaining racial inequality across neighborhoods.

Hypersegregation is defined as areas that are highly segregated on five measures: evenness, exposure, clustering, concentration, and centralization (Massey and Denton

1993). Almost 40% of all African Americans live in metropolitan areas that are hypersegregated, and 60% live in areas highly segregated on the evenness dimension

(Iceland, Weinberg, and Steinmetz 2002). Overall, levels of residential segregation are lower for Hispanics than Blacks but have not decreased over time (Iceland and Nelson

2008). By 2000, Hispanics in both New York and Los Angeles were hypersegregated

(Wilkes and Iceland 2004). However, new immigrant destinations the Southeast and

Midwest are also becoming increasingly segregated, especially for foreign-born

Hispanics (Park and Iceland 2011).

Although segregation is less pervasive for Hispanics than African Americans, the social and economic consequences of segregation are observed in both groups. Among

African Americans, segregation contributes to suboptimal educational attainment and employment outcomes (Cutler and Glaeser 1997; Card and Rothstein 2007; Massey and

Fisher 2006). Among both African Americans and Hispanics, segregation exacerbates

17 levels of crime and homicide (Krivo et al. 2009; Feldmeyer 2010). Residential segregation is also associated with lower rates of homeownership, as well as lower housing quality among homeowners and high rates of foreclosure (Flippen 2010; Rugh and Massey 2010).

Segregation negatively influences African Americans in part because it is rooted in both interpersonal and institutional discrimination. White avoidance of African-

American neighbors contributed to the emergence and to the persistence of Black-White segregation levels, due in large part to negative stereotypes about Black residents and neighborhoods (Zubrinsky and Bobo 1996; Quillian 2002). Although Blacks are willing to move into balanced, integrated communities at rates significantly higher than Whites,

Blacks fear hostility from Whites in majority White neighborhoods that have no visible

Black presence (Krysan and Farley 2002). As Blacks moved in large numbers out of the

South into the North after World War I during the Great Migration in search of economic opportunities, housing discrimination also kept Blacks out of predominantly White neighborhoods (Tolnay 2003). Effective institutional barriers to entry into these communities include restrictive covenants that denied deed transfers to non-White families as well as financial policies that restricted the ability of Blacks to qualify for loans in White neighborhoods (Massey and Denton 1993; Pager and Shepherd 2005).

When these methods were less effective, Whites also used violence to drive out current minority residents and prevent potential minority residents from moving in. These forces resulted in urban segregated ghettos that experience high rates of unemployment, social disorganization, and crime (Wilson 1996).

In contrast to the negative underpinnings of residential segregation for African

2000, 27% of all Hispanic children lived in poverty, and rates are even higher among the first and second generations (36% and 28%, respectively) compared to third generation

Hispanic children (23%) (Lichter, Qian, and Crowley 2005). Residing in poor ethnic enclaves may be particularly detrimental to these children in the absence of strong organization and collective efficacy to ward off the associated consequences of poverty.

Patterns of acculturation are also strongly related to segregation. Hispanics that are native-born, have higher socioeconomic status, and better English language

19 proficiency are more likely to move into neighborhoods with fewer Hispanics and more

Whites (South, Crowder, and Chavez 2005; Iceland and Scopilliti 2008). Although higher socioeconomic attainment and acculturation may facilitate migration into more racially diverse neighborhoods, it may also prevent native born Hispanics from maintaining important social ties with co-ethnics. For Hispanics, the health consequences of migration to less segregated areas with fewer co-ethnics remain unclear.

New immigrant destinations, those with only a recent history of immigration, are growing increasingly segregated as the size of foreign-born and Hispanic population rises rapidly, and this segregation is not buffered by economic attainment among Hispanics

(Lichter et al. 2010). Children, particularly immigrant Hispanic children, are more vulnerable to school dropout in new immigrant destinations than in older, defined immigrant gateways, as schools and communities are unable to provide additional support to immigrant children for whom English is a second language (Fischer 2010). For

Hispanic children, the effects of place segregation may be highly complex, a mixture of economic and structural factors that vary by generation.

Segregation and Health

The major pathways that link segregation to health include restricted socioeconomic status and high concentrations of poverty, exposure to discrimination and subsequent elevated stress levels, and poor neighborhood environmental quality

(Acevedo-Garcia et al. 2003). Williams and colleagues (Williams and Collins 2001;

Williams and Jackson 2005) argue that racial segregation is a fundamental cause of racial disparities in health primarily through its impact on individual socioeconomic status and

20 neighborhood poverty. Extant research regarding residential segregation and health finds a significant negative association between segregation and wellbeing even after accounting for differential levels of SES at the individual and aggregate level

(Subramanian et al. 2005; Osypuk and Acevedo-Garcia 2008; LaVeist 2003). Although health disparities are reduced when accounting for socioeconomic differences between racial/ethnic groups, individual and aggregate differences in SES cannot fully explain the link between segregation and health.

As discrimination is a major force in creating and perpetuating residential segregation, it also serves as a mediating pathway between segregation and health. Racial discrimination has been found to be detrimental to both mental and physical health, although the relationship appears stronger for the former than the latter (Williams,

Neighbors, and Jackson 2003). Investigators have primarily explained this association by invoking the stress process model and arguing that exposure to racial discrimination predominantly impacts health by elevating levels of psychosocial stress to such an extent as to interfere an individual‟s ability to effectively cope with everyday life stressors such as financial hardship and parenting (Williams and Sternthal 2010). For African

Americans, both major life events of race-related mistreatment and negative everyday encounters can be upsetting, stressful, and cause cumulative harm over the life course

(Feagin and Sikes 1994). Among African Americans, personal experiences with racism result in poorer self-rated physical health and health behaviors (Borrell et al. 2006;

Borrell et al. 2010), and, more consistently, increased psychological distress (Williams et al. 1997; Jackson et al. 1996; Schulz et al. 2006). In Black adolescents, perceived

21 discrimination increases anger and depression, as well as decreases self-esteem, resiliency, academic performance, and perceived academic ability (Wong, Eccles, and

Sameroff 2003).

Overall discrimination helps to create and maintain segregation across neighborhoods; however, within neighborhoods, more racial integration may foster an atmosphere that encourages discrimination rather than tolerance (Welch et al. 2001; Hunt et al. 2007). Studying race-based stress and social support, Swaroop (2005) finds that the influence of exposure to other-race neighbors and isolation from same-race neighbors varies across racial groups. For Blacks, exposure to more White neighbors and fewer

Black neighbors increases perceived discrimination; for Latinos, exposure to more Black neighbors increases perceived discrimination and decreases socializing among neighbors.

This decrease in perceived discrimination may be responsible for findings that segregation has positive benefits to health, by reducing stress and increasing social support in segregated Black neighborhoods (e.g. Bell et al. 2006; Inagami et al. 2006).

Finally, the relationship between segregation and the built environment suggests that neighborhood effects on health are likely to be complex. Segregated neighborhoods have fewer grocery stores, as well as reduced availability of inexpensive healthy foods, recreational facilities, and parks (Zenk et al. 2005; Moore and Diez Roux 2006; Moore et al. 2008). Residents of segregated neighborhoods are also exposed to increased pollution, lead, and other environmental toxins (Morello-Froesch and Lopez 2006). Although low income, racial minority neighborhoods are generally poor environments for child wellbeing (Jackson and Mare 2007; Collins et al. 2008), some research suggests that this

22 relationship may not be straightforward. Among children in the Fragile Families study, those living in housing projects played outside more often than those living elsewhere, which may be a result of playgrounds that have been specifically built as part of these developments for the use of its residents. (Kimbro, Brooks-Gunn, and McLanahan 2011).

While increased levels of physical activity has been shown to be beneficial for a wide range of health outcomes among children, extensive time spent outside may increase the likelihood that a child is exposed to neighborhood physical and social disorder.

Cumulative Exposure

Life course perspectives on health stress the importance of cumulative risk processes in determining wellbeing (Kuh and Ben-Schlomo 1997). According to the weathering hypothesis, the chronic stress of economic and social disadvantage, racism, and political marginalization increase the psychological and physiological toll of everyday life on African Americans, which in turn triggers accelerated aging especially during prime childbearing years (Geronimus et al. 2006). African Americans experience faster rates of decline in health than whites, with increasingly higher rates of morbidity and mortality across the life course (Geronimus et al. 2001; Geronimus et al. 2010).

Evidence suggests this process starts early in adulthood, with worsening outcomes as age increases, especially for women‟s birth outcomes (Geronimus 1996). Additionally, upward social mobility does not offer a health benefit in birth outcomes to Black women as it does to White mothers (Colen et al. 2006). Given the cumulative and convincing evidence of weathering among Black women, it is important to understand why Black-

White health inequalities emerge at early ages and what factors cause them to grow more pronounced over time.

Although life course research has been focused on individual-level cumulative disadvantage (Hertzman and Power 2003), recent work on educational attainment and fertility behavior has begun to focus on cumulative exposure to neighborhood-level disadvantage (Clampet-Lundquist and Massey 2008; Crowder and South 2011). Racial differences in cumulative exposure to poverty are much larger than single point in time measures of poverty (Timberlake 2007). Both Black and Hispanic families living in poverty are more likely to live in segregated neighborhoods than their nonpoor counterparts (Iceland et al. 2010). In studying cumulative exposure to poverty, Jackson and Mare (2007) find that cumulative measures do not explain educational attainment and childhood health any better than poverty measures captured at single point in time.

However, they use data from the Los Angeles Family and Neighborhoods Survey (L.A.

FANS) that includes only 2 years residential data and one time measurement of outcomes, and just two waves of the Child Development Supplement to the Panel Study of Income Dynamics (PSID-CDS). Although the PSID-CDS data contains longer residential histories of focal children than L.A. FANS, it includes relatively few

Hispanics that cannot be analyzed separately. Additionally, Jackson and Mare only measure general overall health, and do not include any measures of specific physical health outcomes.

Even less consideration has been given to empirically examining the influence of cumulative exposure to segregation on health, despite the theoretical importance of

24 cumulative disadvantage of segregated neighborhoods (Massey 2004; Diez Roux and

Mair 2010). Using a retrospective recall of five social context measures throughout the life course, LaVeist (2003) finds that multidimensional, lifetime exposure to segregated schools, work places, child neighborhoods, and current neighborhoods accelerates mortality among African Americans. Survival rates for African Americans with high lifetime exposure was 20% lower than Blacks with low exposure, and only some of this relationship is explained by socioeconomic status, gender, and prior health status.

Similarly, among African Americans, childhood exposure to segregation is associated with smoking among African Americans independent of adult exposure (Landrine and

Klonoff 2000).

Finally, assessments of residential mobility programs such as Moving to

Opportunity (MTO) have generally found positive health outcomes for African

Americans moving into lower poverty and less segregated neighborhoods, particularly for mental health and obesity outcomes, suggesting that disrupting the cumulative exposure to poverty and segregation is beneficial (Kling et al. 2004; Leventhal and Brooks-Gunn

2003; Ludwig et al. 2008). However, the MTO program was not specifically designed to reduce participants exposure to residential segregation, and many African Americans moved to lower poverty neighborhoods that were still highly segregated. Moreover, recent critiques of the MTO program suggest that findings from this experiment may suffer from selection bias. (Clampet-Lundquist and Massey 2009; Ludwig et al. 2008).

Finally, beneficial outcomes from MTO are likely hampered by participants‟ initial poor health status (Osypuk and Glymour 2011).

I expand on these prior findings by critically assessing the role of residential segregation in child wellbeing among Whites, Blacks, and Hispanics. I examine the extent to which metropolitan level residential segregation impacts child wellbeing among

African American and Hispanic youth, and whether the impact of segregation on health is cumulative over time. Although I expect segregation generally to have negative effects on health, I expect that racial isolation may actually be beneficial, by reducing experiences of stress and discrimination, and increasing social support and networks.

1) Black-White segregation measured by the dissimilarity index will be detrimental to African American children’s health, but not to White and Hispanic children.

2) Hispanic-White dissimilarity will be detrimental to Hispanic children’s health, but not to White and African American children

3) Black residential isolation will be beneficial to African American children’s health, but not to White and Hispanic children.

4) Hispanic residential isolation will be beneficial to Hispanic children’s health, but not to White and African American children.

5) Due to cumulative exposure to segregation, chronic exposure to residential segregation will be detrimental to the health of African American children relative to White and Hispanic children.

Data and Methods

Data

The data for the subsequent analyses comes from the restricted version of the Early

Childhood Longitudinal Survey Kindergarten Class of 1998-1999 data (ECLS-K). The

ECLS-K is a nationally representative longitudinal study of approximately 20,000 kindergarteners in 1998-1999 that were re-interviewed in first grade (1999-2000), third grade (2002), fifth grade (2004), and eighth grade (2007). Data for the kindergarten wave are taken from the spring assessment unless otherwise noted. Pertinent information was collected from students (i.e. respondents), parents, teachers, school administrators, as well as directly assessed by specially trained interviewers. Respondents‟ home addresses were collected during each round and subsequently geocoded. The longitudinal design and wealth of sources make the ECLS-K ideal for assessing child health over the early life course while providing the ability to link geographic information to a nationally representative sample with large minority student samples.

Dependent Variables

I capture physical health using two measures of physical health (obesity and asthma), one measure of global health (parent-rated health). Obesity and parent-rated health are measured at each wave, but questions concerning whether or not the respondent has been diagnosed with asthma are only asked in waves during 3rd, 5th, and

8th grade, therefore, analyses where asthma is the dependent variable contain a subset of observations.

Body Mass Index (BMI) is calculated from interviewers‟ measurements of height and weight during each round of child assessments, as mass in kilograms divided by height in meters squared, is compared to gender-specific BMI-for-age growth charts from the Centers for Disease Control and Prevention (CDC). Based on recommendations from the CDC and Institute of Medicine, I categorize children as obese if their combined height and weight measurements put them at or above the 95th percentile for their gender and age. Previous research has established that children in this category experience an increased risk of secondary complications such as cardiovascular disease (Krebs et al.

2007; Kuczmarski et al. 2002). Parents report whether their child has ever been diagnosed with asthma by a doctor. They also report how healthy the child is on a scale from 1, which indicates excellent health, to 5 which indicates poor health. In multivariate models, parent-rated health has been dichotomously recoded, where 1 represents fair or poor health and 0 represents good, very good, or excellent health.

Independent Variables

The independent variables of interest capture the racial segregation of the metropolitan area in which the respondent resides during each wave of data collection. Respondent‟s addresses are matched to the corresponding Metropolitan Statistical Areas (MSA) or the

Primary Metropolitan Statistical Areas (PMSA). I use two dimensions of segregation, evenness and exposure, calculated by the Census Bureau using Census 2000 data

(Iceland, Weinberg, and Steinmetz 2002). Evenness, the degree to which the percentage of minority members in smaller residential areas matches the overall metropolitan area percentage, is measured by the dissimilarity index, which can be interpreted as the

28 percent of minority residents (Black or Hispanic, respectively) who would have to move census tracts for each census tract to match the percentage of that minority that live in the

MSA/PMSA as a whole. Exposure is the degree to which residents are exposed to members of other racial group by sharing the same neighborhoods, and is measured in this study by the isolation index, the weighted average percentage Black or Hispanic within census tracts averaged across the MSA/PMSA (Massey and Denton 1993). Both indices are on a scale of 0-100 with greater values indicating higher levels of residential segregation. For the models assessing chronic exposure, I count the number of rounds of data collection the child has experience levels of 60% or greater for each segregation measure, which is generally considered to be high exposure for ascertaining hypersegregation (Denton and Massey 1989; Wilkes and Iceland 2004). I categorize children as chronically exposed if they have experienced 4 or 5 rounds of exposure to high segregation, compared to children who have experienced fewer rounds. Most children experience either all rounds in high segregation or none or 1, so I dichotomize the chronic exposure variables comparing high exposure to low exposure, the reference category.

I control for a number of demographic and socioeconomic characteristics including child‟s gender (female is reference), age (in years), race (non-Hispanic White, non-Hispanic Black, or Hispanic), total household size, whether or not the child resides in a single parent household, and whether the child has no health insurance coverage, compared to either private or government insurance, because at the time of the

Kindergarten interview in 1998-1999 parents were only asked to report if the focal child

29 had any health insurance but were not asked to specify the type of coverage. I also incorporate a composite measure of socioeconomic status constructed by the ECLS-K that combines mother‟s and father‟s educational attainment, income, and occupational prestige and categorizes students into quintiles based on the combination of these measures.

Analytic Strategy

Analyses are restricted to respondents who have complete data on variables of interest including geocoded information, and I include respondents who have at least two waves of data. This provides information from a sample of 6,650 non-Hispanic Whites,

1,160 non-Hispanic Blacks, and 2,240 Hispanics, and a total of 39,110 observations. I employ multilevel logistic regression techniques to predict the likelihood of being obese, being diagnosed with asthma, and parent-reported poor health using population averaged models and robust standard errors by child. For the final set of chronic exposure models,

I restrict the sample to the last interview round in 8th grade in order to measure chronic exposure, and use Huber-White corrections to standard errors and cluster individual observations within MSA to account for heteroskedasticity and non-independence of observations. Sampling weights are applied to descriptive analyses to correct for both the survey non-response of those who are lost to follow up and the complex survey design of the ECLS-K. Unweighted sample sizes are rounded to the nearest 10 per restricted data agreement with the Institute of Education Services and the National Center for Education

Statistics.

Results

Weighted descriptive statistics by race are presented in Table 1 for respondent‟s first interview and last interview, except for asthma which is presented as lifetime diagnosis rates due to the more limited sample years in which this question was asked.

Clear racial disparities in all three outcome measures are apparent. Over 50% of White children are reported as being in Excellent health by parents at both time points, but only

40-45% of Blacks and Hispanics are rated to be in excellent health. Twice as many

Blacks and Hispanics are in good health as Whites, and almost 5% are rated as being in fair or poor health. Although all groups experience increases in rates of obesity between the first and last years, Hispanics start at much higher rates, almost 20% at the first interview, compared to only 9% of White and 12% of African American children. By the last interview, over 20% of African Americans are also obese. Finally, a quarter of

African American children are ever diagnosed with asthma by their last interview.

Residential patterning also varies by race. Black children experience much higher levels of residential segregation than either White or Hispanic children. White and

Hispanic children live in metropolitan areas that have, on average, Black-White dissimilarity indices of 59 and 57, respectively, in their last interviews, while their

African American counterparts reside in MSAs that have average Black-White dissimilarity indices of 64. Not surprisingly, Hispanics experience the highest levels of

Hispanic-White segregation, with average Hispanic-White dissimilarity of 51, compared to 44 among Whites and 45 among Blacks. Blacks also live in metropolitan areas where the average Black residents live in neighborhoods that are 59% Black, and Hispanics live

31 in metropolitan areas where the average Hispanic lives in a neighborhood that is 53%

Hispanic.

Half of African American children live in single parent households, compared to only about a quarter of Hispanics, and less than 20% of Whites. Racial differences in health insurance are also apparent. Although 5% or less of White children have no health insurance of any kind at both first and last interviews, 9% of Blacks and 20% of

Hispanics at the first interview do not have coverage. By their last interview, just 4% of

Blacks have no coverage, but 12% of Hispanics still lack health insurance coverage of any kind. Finally, racial disparities in socioeconomic status, a combined measure of education, income, and occupation, are quite pronounced. Less than 10% of Whites at both times fall in the bottom quintile of family SES, but a third of Blacks and 40% of

Hispanics do. In contrast, a third of White children fall into the top quintile of family

SES, yet less than 10% of Blacks and Hispanics do.

Multilevel logistic regression analyses are presented by health outcome and separately by Black and Hispanic segregation measures in Tables 2 (poor health), 3

(asthma), and 4 (obesity). Models 1 and 2 present the effect of Black-White Dissimilarity and Black Isolation measures, and Models 3 and 4 present the effects Hispanic-White

Dissimilarity and Hispanic Isolation, respectively. All models include interactions between segregation measures and each racial group to estimate whether the influences of segrgation on health differs by racial/ethnic group.

Table 2 presents the odds of being in fair or poor health for Whites, Blacks, and

Hispanics. In Model 1, Blacks are 2.5 times more like than Whites to be in poor health,

32 but the difference is not significant, whereas the odds of poor health are much higher for

Hispanics, who are 5.8 times more likely to be in poor health than Whites. The main effects of Black-White dissimilarity are not significant, therefore although the interaction between Black-White dissimilarity and Hispanic is significant, I do not interpret it. For

Model 2 presents the estimates of Black isolation on poor health, and Figure 2 presents the predicted probabilities. The main effects of Black isolation are positive, so as the average percentage Black across tracts in an MSA increases, the probability of poor health increases for Whites, although the effect size is small, a 1% in the odds of poor health for each 1% increase in Black isolation. The interaction is not significant for

Blacks, suggesting that the relationship between Black isolation and poor health is similar for Whites and Blacks. The relationship is negative for Hispanics, but the effect size is smaller, a 0.2% (=exp(ln(1.01)+ln(.99))) decrease in the odds of poor health with each

1% increase in Black isolation. In model 2, the odds of poor health among Hispanics is now 3.6 times higher, and 2.3 times higher for Blacks, relative to Whites.

Hispanic-White dissimilarity and Hispanic isolation, shown in models 3 and 4 of

Table 2, do not have significant main effects nor interaction effects with race on the odds of poor health. However, controlling for Hispanic segregation measures reduces the racial disparity in poor health between Hispanics and Whites, where the odds of poor health is just 1.5 times higher for Hispanics and not significantly different from Whites.

Table 3 presents the logistic regression models for asthma diagnosis. None of the main effects of Black-White residential dissimilarity, Black isolation, Hispanic-White dissimilarity, and Hispanic isolation are significantly associated with asthma diagnosis,

33 the only race interaction term that is significant is the Hispanic race by Hispanic isolation term. In that model, Model 4, both Black and Hispanics are 1.73 and 1.78 times more likely to be diagnosed with asthma. For Hispanics, Hispanic isolation is negatively associated with the odds of asthma, but again the effect size is small, a 1%

(=exp(ln(1.0)+ln(.99))) decrease in the odds of asthma for each 1% increase in Hispanic isolation.

As shown in Table 4, the association between residential segregation and obesity also varies by race. Black-White dissimilarity and Black isolation are not significantly associated with obesity for any group, but Hispanic-White dissimilarity as well as

Hispanic isolation are associated with obesity for Hispanics. Hispanic-White dissimilarity is positively associated with the odds of obesity, a 1% (=exp(ln1.0+ln1.01) increase in the odds of obesity for each 1% increase in Hispanic-White dissimilarity and Hispanic isolation.

Table 5 presents the level of chronic exposure by residential segregation measure and race, weighted to adjust for survey design. Two-thirds of African Americans live in highly segregation MSAs in at least 4 rounds of data collection. The level of exposure is lower for Whites, at 54%, and Hispanics, at 45%. The differences in exposure to Black

Isolation are more pronounced. Over half of African Americans but only about 30% of

Whites and Hispanics experience high levels of chronic exposure to Black Isolation.

Given the lower levels overall of Hispanic-White dissimilarity, only 14% of Whites and

Blacks experience chronic Hispanic segregation, but 21% of Hispanic so. Similar to

Black isolation, Hispanics experience Hispanic isolation at much higher rates than the

34 other groups; 36% of Hispanics but only 13% of Blacks and 5% of Whites experience chronic exposure to high levels of Hispanic isolation.

Table 6 presents the results from the logistic regressions of chronic exposures on health for segregation measures. Due to small sample sizes reporting poor health, stratification by race is not possible, although it would be preferable given the potential differences in the impact of chronic exposure. Exposure to Black-White segregation and

Black isolation does not have any significant cumulative effects on health for any of the three health measures. Cumulative exposure to Hispanic-White dissimilarity reduces the likelihood of poor health by 62%. In contrast, cumulative exposure to Hispanic isolation increases the odds of poor health by a factor of 2. Chronic Hispanic-White dissimilarity and Hispanic isolation do not significantly predict the odds of asthma diagnosis or obesity. However, these estimates should be read cautiously. Given the low rates of chronic exposure to Hispanic segregation, and the low rates overall of poor health, they are quite sensitive to small cell sizes.

Discussion

This paper examines the role of residential segregation in determining health and wellbeing across the early life course. Among African American adults, segregation has generally been detrimental to health, although Black isolation levels has been more consistently linked to negative outcomes than Black-White dissimilarity. Among

Hispanics, the relationship varies considerably across health outcomes and segregation measures. The goal of this research was to extend this conceptual model to children,

35 examining the role of segregation in determining racial disparities in child health as well as the added effects of cumulative exposure to residential segregation.

I find no evidence to support hypothesis 1, that Black-White dissimilarity would be detrimental to the health of Black children. Among African American children, Black-

White segregation is not associated with the likelihood of poor health, asthma, or obesity after accounting for other individual level characteristics. The results also do not support a protective effect of Black isolation against poor health for Black children. Similar to

Black-White dissimilarity, isolation is not significantly associated with any of the three health outcomes among African Americans. These findings are contrary to findings in the adult health literature that Black-White residential dissimilarity is related to worse health outcomes for Black adults as well as infants (Bell et al. 2006; Chang 2006; Subramanian et al. 2005). This divergent finding warrants more inquiries. Black-White segregation may not affect Black children adversely, either because exposure to segregation is cumulative across the life course and the effects do not contribute to racial inequality in health until later or because the processes of health inequality for young African

Americans is not tied to residential segregations on the metropolitan level. It is also possible that because children are less likely to be mobile around a city, the lower level neighborhood processes may be more important for racial inequalities in early life than indices at the metropolitan level can tease out.

As previous research has shown, the relationship between segregation and

Hispanic health varies across both segregation and health measures. I find that Black isolation levels are protective against poor health for Hispanic children, an unexpected

36 finding. In support of Hypothesis 2 that dissimilarity would be detrimental to the health of Hispanic children, Hispanic-White segregation is harmful but only for obesity levels.

Hispanic obesity increases with dissimilarity as well as with Hispanic isolation, which is in contrast to the hypothesized benefit of Hispanic isolation for Hispanic children.

However, Hispanic isolation does decrease the odds of asthma diagnosis among Hispanic children. Therefore, Hispanic children living in areas with greater exposure to other

Hispanics and more separation from other racial groups, neighborhoods that are generally considered to be positive environments because of ethnic social ties and support, have lower rates of asthma but higher rates of obesity. Living in more unequal metropolitan areas may expose Hispanics to more racism and discrimination, as well as providing fewer resources to help increase health and well-being in the face of socioeconomic and racial adversity, but it is unclear why this effect is only obesity and not poor health or asthma. Rates of exercise among Hispanic adults are also lower among those living in more segregated areas (Mellerson et al. 2010), so if this health behavior is passed on to children it may help explain why obesity is particularly affected by Hispanic-White segregation.

Although cumulative exposure to segregation was hypothesized to increase the likelihood of poor health, African American segregation is not related to the odds of poor health, obesity, or asthma. Only Hispanic-White dissimilarity and Hispanic isolation are associated with the odds of poor health, a decrease in the likelihood of poor health with higher levels of Hispanic-White dissimilarity, and an increase in the likelihood of poor health with high levels of Hispanic isolation. The lack of association between Black-

White segregation and health may reflect the shorter time spans in the early life course; by age 14, the cumulative nature of segregation may not be fully in effect. This may also reflect the limitations in the ECLS-K in calculating true duration measures of actual exposure, since only current residential place is geocoded at each round of data collection, and detailed residential histories in between rounds are not recorded. Also, while many children changes residences within metropolitan areas, few change residences across metropolitan areas, resulting in less variation in cumulative exposure to segregation. Further analyses are warranted to assess whether the cumulative nature of segregation varies across racial groups.

Finally, I note an important limitation with this study. Measuring residential segregation at the metropolitan level reports important information about how residential patterns vary across neighborhoods, and so is a relative measure of residential inequality.

However, measuring segregation at the MSA level clearly misses important variation across individuals that results in different exposure to residential patterns. Future research on the health effects of residential segregation may benefit from new measures designed to decompose regional variation into smaller spatial units that may better assess individual level experiences situated in the broader context of residential segregation patterns (Lee et al. 2008).

Overall, the results suggest that effect of segregation on child health is highly dependent upon segregation measure and health outcome, and not consistent across racial groups. This suggests that although segregation may be a determining factor in adult health, early life exposure to residential segregation is not a consistent determinant of

38 health disparities. The health of Hispanic children is more sensitive to these residential patterns than Whites and Blacks, even if the directions of the relationships vary by segregation and health measure, and may signify important variations within the

Hispanics as a group. By assessing the role of segregation among children and especially

Hispanic children, this research broadens our understanding of the social processes that create health disparities and under what circumstances residential segregation does influence child health. Research into the mechanisms that translate residential segregation into health outcomes needs to be sensitive to racial/ethnic differences in the effect of segregation measures, and pay particular attention to the reasons how and why dissimilarity and isolation measures differentially effect health.

White White Black Black Hispanic Hispanic First Last First Last First Last

Health Excellent 59% 55% 45% 44% 40% 45% Very Good 29% 33% 33% 34% 34% 30% Good 11% 10% 18% 18% 21% 19% Fair 2% 2% 4% 4% 5% 5% Poor 0% 0% 0% 1% 0% 0%

Obese 9% 16% 12% 21% 19% 25% Asthma, ever diagnosed 18% 24% 17%

Black -White Dissimilarity 59.82 59.28 64.07 63.67 58.22 57.39 Black Isolation 46.08 45.44 59.39 58.97 46.20 45.14

Hispanic -White Dissimilarity 44.03 44.08 45.51 45.42 50.57 49.99

Hispanic Isolation 26.77 26.87 29.2 29.02 53.31 52.76

Age, years 6.55 12.2 6.53 11.85 6.61 12.31

Male 51% 53% 52% 54% 53% 52% Single Parent Household 15% 19% 50% 49% 23% 26% No Health Insurance 5% 4% 9% 4% 20% 12% Household Size 4.42 4.38 4.52 4.57 4.86 4.86

Socioeconomic Status Quintile Lowest Quintile 7% 7% 30% 30% 41% 38% Second Quintile 17% 16% 23% 24% 22% 23% Third Quintile 20% 22% 21% 21% 16% 19% Fourth Quintile 25% 24% 17% 17% 14% 12% Highest Quintile 32% 31% 8% 8% 7% 8% Number of Persons Across all 6,650 1,660 2,440 years Table 1: Weighted Descriptive Statistics by First and Last Interview and Race

Poor Health Poor Health Poor Health Poor Health Model 1 Model 2 Model 3 Model 4 Black 2.58 2.31* 1.63 1.81** (1.29) (0.74) (0.58) (0.32) Hispanic 5.80** 3.57** 1.50 1.50 (2.30) (0.72) (0.54) (0.32) Male 1.27** 1.26** 1.26** 1.26** (0.10) (0.10) (0.10) (0.10) Age, years 1.01 1.01 1.01 1.01 (0.01) (0.01) (0.01) (0.01) Single Parent Household 1.22* 1.23* 1.22* 1.23* (0.11) (0.11) (0.11) (0.11) Household Size 1.05 1.05 1.05 1.05 (0.03) (0.03) (0.03) (0.03) No Health Insurance 0.91 0.92 0.92 0.92 (0.10) (0.10) (0.10) (0.10) Socioeconomic Status Quintiles 0.67** 0.67** 0.67** 0.68** (0.02) (0.02) (0.02) (0.02) Black -White Dissimilarity 1.01 (0.00) Black*Black -White Dissim 0.99 (0.01) Hispanic*Black -White Dissim 0.98* (0.01) Black Isolation 1.01* (0.00) Black*Black Isolation 1.00 (0.01) Hispanic*Black Isolation 0.99* (0.00) Hispanic -White Dissimilarity 0.99 (0.00) Black * Hispanic-White Dissim 1.00 (0.01) Hispanic * Hispanic-White Dissim 1.01 (0.01) Hispanic Isolation 1.00 (0.00) Black * Hispanic Isolation 1.00 (0.01) Hispanic * Hispanic Isolation 1.01 (0.00) N 39110 39110 39110 39110 χ2 408.19 413.64 401.75 403.43 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 2: Odds Ratios from Logistic Regression of Segregation on Poor Health

Asthma Asthma Asthma Asthma Model 1 Model 2 Model 3 Model 4 Black 1.37 2.18** 1.62 1.73** (0.52) (0.54) (0.44) (0.22) Hispanic 0.64 0.97 1.41 1.78** (0.19) (0.15) (0.41) (0.28) Male 1.67** 1.67** 1.68** 1.68** (0.10) (0.10) (0.10) (0.10) Age, years 1.05** 1.05** 1.05** 1.05** (0.01) (0.01) (0.01) (0.01) Single Parent Household 1.11 1.11 1.11 1.11 (0.06) (0.06) (0.06) (0.06) Household Size 0.97 0.97 0.97 0.97 (0.02) (0.02) (0.02) (0.02) No Health Insurance 0.84 0.84 0.84 0.84 (0.07) (0.07) (0.07) (0.07) Socioeconomic Status Quintiles 1.01 1.01 1.01 1.00 (0.02) (0.02) (0.02) (0.02) 1.00 Black-White Dissimilarity 0.00

1.00 Black*Black-White Dissim (0.01)

1.01 Hispanic*Black-White Dissim (0.00)

1.00 Black Isolation (0.00)

1.00 Black*Black Isolation (0.00)

1.00 Hispanic*Black Isolation (0.00)

1.00 Hispanic-White Dissimilarity (0.00)

Black * Hispanic-White Dissim 1.00

(0.01)

Hispanic * Hispanic-White Dissim 0.99

(0.01)

1.00 Hispanic Isolation (0.00)

1.00 Black * Hispanic Isolation (0.00)

Hispanic * Hispanic Isolation 0.99**

(0.00)

N 20180 20180 20180 20180 χ2 221.47 218.34 218.70 228.00 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 3: Odds Ratios from Logistic Regression of Segregation on Asthma

Obesity Obesity Obesity Obesity Model 1 Model 2 Model 3 Model 4 Black 0.99 1.17 1.11 1.33* (0.33) (0.27) (0.27) (0.15) Hispanic 1.43 1.75** 0.95 1.44** (0.31) (0.21) (0.21) (0.18) Male 1.22** 1.22** 1.22** 1.22** (0.06) (0.06) (0.06) (0.06) Age, years 1.08** 1.08** 1.08** 1.08** 0.00 0.00 0.00 0.00 Single Parent Household 1.07 1.08 1.08 1.08 (0.04) (0.04) (0.04) (0.04) Household Size 0.97 0.98 0.97 0.98 (0.01) (0.01) (0.01) (0.01) No Health Insurance 0.91 0.91* 0.91* 0.91* (0.04) (0.04) (0.04) (0.04) Socioeconomic Status Quintiles 0.92** 0.92** 0.92** 0.92** (0.01) (0.01) (0.01) (0.01) Black -White Dissimilarity 1.00 (0.00) Black*Black -White Dissim 1.01 (0.01) Hispanic*Black-White Dissim 1.00 (0.00) Black Isolation 1.00 (0.00) Black*Black Isolation 1.00 (0.00) Hispanic*Black Isolation 1.00 (0.00) Hispanic-White Dissimilarity 1.00 (0.00) Black * Hispanic-White Dissim 1.01 (0.01) Hispanic * Hispanic-White Dissim 1.01** (0.00) Hispanic Isolation 1.00 (0.00) Black * Hispanic Isolation 1.00 (0.00) Hispanic * Hispanic Isolation 1.01* (0.00) N 39110 39110 39110 39110 χ2 597.04 593.97 594.26 586.94 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 4: Odds Ratios from Logistic Regression of Segregation on Obesity

White Black Hispanic Chronic Black-White Dissimilarity >60 54% 65% 45%

Chronic Black Isolation >60 28% 56% 31%

Chronic Hispanic-White Dissimilarity >60 14% 14% 21%

Chronic Hispanic Isolation >60 5% 13% 36% Table 5: Cumulative Exposure to Segregation by Race in Last Interview

Poor Health Poor Health Asthma Asthma Obese Obese

Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 African American 1.91 1.8 1.49** 1.47** 1.32 1.37* (0.66) (0.62) (0.21) (0.19) (0.19) (0.21)

Hispanic 2.50** 2.13** 0.88 0.91 1.45** 1.50*** (0.67) (0.53) (0.11) (0.11) (0.17) (0.17)

Male 1.09 1.09 1.58** 1.58** 1.45** 1.46** (0.21) (0.21) (0.13) (0.13) (0.10) (0.10)

Age, years 1.16 1.13 0.97 0.97 0.86 0.85 (0.31) (0.29) (0.13) (0.12) (0.09) (0.09)

Single Parent Household 1.38 1.39 1.21 1.21 1 0.99 (0.33) (0.34) (0.13) (0.13) (0.09) (0.09)

Household Size 1.05 1.06 0.94 0.94 0.95 0.95 (0.07) (0.07) (0.03) (0.03) (0.03) (0.03)

No Health Insurance 0.87 0.81 0.9 0.92 1.35 1.37 (0.23) (0.21) (0.17) (0.17) (0.27) (0.27)

Socioeconomic Status Quintiles 0.64** 0.65** 1.03 1.03 0.79** 0.78** (0.06) (0.06) (0.04) (0.04) (0.03) (0.03)

0.93 0.94 0.82 Chronic Black-White Dissimilarity >60 (0.24) (0.10) (0.10)

Chronic Black Isolation >60 0.91 0.95 1.25

(0.24) (0.10) (0.15)

0.38*** 1.06 1.11 Chronic Hispanic-White Dissimilarity >60 (0.08) (0.14) (0.16)

Chronic Hispanic Isolation >60 2.00* 0.88 0.94

(0.60) (0.15) (0.14)

N 5140 5140 5140 5140 5140 5140 χ2 139.01 114.16 77.77 77.9 141.45 142.65 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 6: Odds Ratios from Logistic Regression of Chronic Exposure to Segregation on Health

Chapter 3: Racial and Ethnic Tensions in the School and Neighborhood Environment: Consequences for Child Physical Health

Despite improvements in population health and wellbeing for African Americans and Hispanics, racial disparities are still large and nontrivial, starting in childhood and persisting through adulthood. African American children are more likely to be born preterm and low birthweight, and are at increased risk of infant mortality (Schempf et al.

2007), obesity (Singh, Siahpush, and Korgan 2010), and asthma (McDaniel, Paxson, and

Waldfogel 2006). Rates of obesity are also higher among Hispanics, especially immigrants, compared to Whites (Singh, et al. 2010). In contrast, research regarding health outcomes for Hispanics has noted similar levels of low birth weight, infant mortality, and asthma compared to Whites, which is surprising given high rates of poverty and lower overall access to health care and insurance (Boardman, Finch, and

Hummer 2001; Hummer et al. 2007). At least part of this Hispanic paradox is explained by favorable mortality profiles for recent Mexican-origin immigrants, as healthy individuals are more likely to emigrate (Palloni and Arias 2004; Markides and Eschbach

2005; Akresh and Frank 2008).

Increasingly, researchers have turned to neighborhood contexts and perceived discrimination experiences as two key explanatory mechanisms driving these racial disparities (Williams and Jackson 2005; Williams and Sternthal 2010). The cumulative effects of poverty, racial segregation, and racial discrimination have led to noticeable 46 difference in the types and quality of neighborhoods that children experience, especially across racial groups. These differences in local social and physical environments impact health significantly, across a range of health measures (Diez Roux and Mair 2010). While neighborhood disorder, such as violence, crime, vacant housing, and drug problems, has been linked to worse health outcomes, social cohesion can greatly benefit health, buffering residents from the deleterious effects of neighborhood disorder (Cradock et al.

2009; Hill, Ross, and Angel 2005; Mair, Diez-Roux, and Morenoff 2010; Molnar et al.

2004). Additionally, the added benefits of social support and cohesion in immigrant neighborhoods also help to explain some of the health advantage of Hispanic immigrants

(Cagney, Browning, and Wallace 2007).

Perceived racism adds acute and chronic strain as well as stresses coping resources (Williams and Williams-Morris 2000; Thoits 2010). Evidence suggests the relationship between discrimination and health is strongest for mental health (Paradies

2006; Williams, Neighbors, and Jackson 2003). Findings are not fully conclusive about the effect of discrimination on physical health, but discrimination has also been linked to worse self-rated health, unhealthy behaviors, and cardiovascular disease markers such as hypertension (Borrell et al. 2006; Krieger and Sydney 1996). Although these relationships have been demonstrated in adults, the effect of racism on child health outcomes has not been fully explored. Results from several studies suggest discrimination experiences are detrimental to academic achievement and mental health among African

American and Hispanic adolescents (Caughy et al. 2004; Coker et al. 2009; Wong,

Eccles, and Sameroff 2003), but the effects of racism have not been assessed for childhood physical health.

Additionally, we know little about how tension and conflict between racial groups within a neighborhood affect health outcomes. Racial tensions in a neighborhood may reflect low social cohesion and trust, as well as increased discrimination among residents of different racial groups. Racial tensions are potentially a source of chronic stress, similar to other types of neighborhood disorder such as crime, drug use, and residential instability. Most often racial tensions have revolved around Black-White divides, but racial tensions can also arise from immigration and assimilation issues, resulting in prejudice and discrimination that particularly affects Hispanics (Ha 2008; Wilson and

Taub 2006). Although other measures of neighborhood stressors have been linked to health outcomes, the role of neighborhood racial tensions in determining health has not yet been explored.

The goal of present study is to combine these two literatures of detrimental neighborhood effects and racial discrimination to explain racial disparities in childhood physical health. Using data from the Early Childhood Longitudinal Survey Kindergarten

Class 1998-1999 (ECLS-K), I study the effect of neighborhood racial and ethnic tensions on physical health in childhood. Second, I assess whether racial/ethnic tensions are detrimental to the health of African American and Hispanic children relative to White children. Finally, I assess whether social cohesion moderates the negative health effects of racial/ethnic tensions.

Stress, Discrimination, and Health

Stress process models of health have developed over time to explain how social structure influences health through exposure to stressors, and the ways in which disadvantaged social status generates increased stress that is translated into decreased physical and mental well-being (Pearlin 1999). Disadvantaged status may also decrease access or efficacy coping resources such as social support, which results in higher vulnerability to stressors. Differential exposure to stressors is a primary way that race, class, and gender inequality is transferred to health (Turner 2003; Thoits 2010). Both life events and chronic strain act as stressors that impact health, and the combined effects are cumulative over the life course (Pearlin 1999; Pearlin et al. 2005). Although stress is most often related to poor mental health outcomes such as depression, it is also detrimental to physical health through increased hormone responses such as cortisol and inflammatory responses (Boardman 2004; McEwan 1998).

The weathering hypothesis proposes that political, social, and economic marginalization accelerates ageing, and as a result, African Americans experience earlier declines in health that accumulate over time and serve to increase racial disparities in well-being over the life course (Geronimus 2001). Accelerated ageing may be a response to the prolonged exposure to and coping with stressors that results from racism and discrimination against Blacks (Walsemann, Geronimus, and Gee 2008). This weathering process disproportionately affects minorities but especially Black women, who experience multiple dimensions of disadvantage in social and economic attainment.

Black women experience higher levels of hypertension and allostatic load, markers of

49 increased stress and worse health (Geronimus et al. 2006; Geronimus et al. 2007).

Research focused on explaining racial disparities in health need to account for exposure to stress and the ways in which stress and coping differentially affect minorities.

Racism, rooted in ideology of social inferiority of groups based on racial group membership, leads to the development of negative attitudes and unfair treatment towards racial minorities by individuals and social institutions (Williams and Mohammed 2009).

For African Americans, both major life events of race-related mistreatment and negative everyday encounters can be upsetting, stressful, and cause accumulative harm over time

(Feagin and Sikes 1994). These acute and chronic stressors accumulate over time, and research has explored the role of discrimination stress in explaining racial disparities in health because of the pervasiveness of discrimination experiences (Williams and

Williams-Morris 2000; Thoits 2010). Among African Americans, personal experiences with racism have consistently been associated with increased psychological distress

(Williams et al. 1997; Jackson et al. 1996; Schulz et al. 2006). Isolated discriminatory events are associated with distress and depression, while chronic, everyday discrimination is also associated with anxiety (Kessler et al. 1999). These trends are apparent not only in adults, but also in adolescents. Among African American and

Hispanic teens, perceived racial/ethnic discrimination is associated with increased anger and depression (Coker et al. 2009; Wong, Eccles, and Sameroff 2003).

Although the relationship between discrimination and physical health is weaker and is not always consistent compared the findings for mental health, perceived discrimination has been associated with poorer self-rated health, hypertension, and

50 unhealthy behaviors among African Americans and Hispanics (Borrell et al. 2006;

Borrell et al. 2010; Finch et al. 2001; Krieger and Sydney 1996). Race-related stress may impact physical health by promoting adoption of unhealthy behaviors to cope with those stressors (Pascoe and Smart Richman, 2009). For example, estimates from the Multi-

Ethnic Study of Atherosclerosis (MESA) report that 43% of African Americans and 19% of Hispanics experienced racial/ethnic discrimination, and those experiences led to increased odds of heavy drinking among both groups (Borrell et al. 2010). Acceptance of negative racial stereotypes of one‟s own group, internalized racism, has also been linked to increased risk of overweight and obesity among Black women in the Caribbean, especially among those who use denial coping styles that reflect a defeated attitude about racism (Tull et al. 1999; Tull et al. 2005).

Racial discrimination at both interpersonal and institutional levels has also created pervasive patterns of residential segregation. White avoidance of African-American neighbors contributed to the emergence and to the persistence of Black-White segregation levels, due in large part to negative stereotypes about Black residents and neighborhoods (Zubrinsky and Bobo 1996; Quillian 2002). Although Blacks are willing to move into balanced, integrated communities at rates significantly higher than Whites,

Blacks fear hostility from Whites in majority White neighborhoods that have no visible

Black preference (Krysan and Farley 2002). Additionally, housing discrimination against

Blacks created barriers to entry into White neighborhoods through financial policies that restrict the ability of Blacks to qualify for loans in White neighborhoods and hostility from landlords and neighborhood residents towards Black housing searchers (Massey and

Denton 1993; Pager and Shepherd 2005). These social forces converged to create systematic differences in the types of neighborhoods experienced by the poor and racial minorities in the U.S. compared to affluent and non-minorities. By age 18, African

American and Hispanic children have spent on average 50% and 40% of their lives, respectively, in neighborhoods with a poverty rate in excess of 20%, compared to only about 5% of the first 18 years of the average White child (Timberlake 2007). Although these structural features are important for health, they do not fully explain racial disparities in health (Farmer and Ferraro 2005; Hayward et al. 2000; Osypuk and

Acevedo-Garcia 2008; Subramanian et al. 2005). As a result, social and physical environments of neighborhoods have increasingly gained attention for explaining neighborhood effects on health (Sampson, Morenoff, and Gannon-Rowley 2002).

Disadvantaged neighborhood contexts can intensify exposure to stress in the form of physical disorder, such as vacant buildings, graffiti, and disrepair, as well as neighborhood social disorder such as crime, drug use, and conflicts (Aneshensel 2009).

High neighborhood poverty levels were associated with higher perceptions of neighborhood disorder in both Baltimore and Chicago; racial segregation was also associated with neighborhood disorder in Chicago (Franzini et al. 2008; Sampson &

Raudenbush 2004). Neighborhood disorder is distressing to residents even if they are not directly victims of crime themselves. Neighborhood disorder creates fear, hopeless, and powerlessness that increases stress and anxiety (Ross and Mirowsky 2001; Schieman and

Meersman 2004). As a result, disadvantaged, segregated neighborhoods are detrimental to health. Neighborhoods marked by neighborhood physical and social disorder have

52 been found to increase unhealthy coping behaviors such as smoking and drinking

(Echeverria et al. 2008; Hill and Angel 2005). Lower neighborhood safety is associated with less physical activity and increased risk of overweight in children and adolescents, as well and worse dietary habits (Lee and Cubbin 2002; Lumeng et al. 2006; Molnar et al.

2004). Finally, since more African Americans and Hispanics rate their neighborhood as unsafe or threatening (Jordan and Gabbidon 2010), neighborhood disorder may disproportionately affect the health of African American and Hispanic children.

Although prior research has focused on the health effects of interpersonal discrimination experiences and neighborhood disorder, little attention has been given to the role of race-related disorder in the form of racial and ethnic tensions in the neighborhood. Neighborhood racial tensions are often rooted in xenophobic attitudes and beliefs against outsiders, such as in the threat of racial or ethnic turnover when groups view one another as competitors for resources (Wilson and Taub 2006). This was apparent in the conflicts between immigrant Korean shop owners and Black residents in

Los Angeles, which disintegrated into protests and even violence under conditions of extreme inequality (Cheng and Espiritu 1989; Lee 2002). Although most day to day interactions are not violent, underlying neighborhood tensions can be a source of chronic stress to residents. For example, Mexican immigrant children who perceive racial tensions in their school are more likely to expect to be discriminated against in the future, and have lower average grades than students who perceive racial harmony in their school

(Stone and Han 2005). African Americans in living in integrated neighborhoods also perceive more social cohesion but also experience more perceived discrimination than

African Americans living in predominantly Black neighborhoods (Swaroop 2005).

Hispanics also experience more discrimination when living in areas with more African

Americans compared to White or Hispanic neighborhoods, as well as reporting less social support in diverse neighborhoods compared to Hispanic neighborhoods. Due to racial tensions, diverse racial/ethnic contexts may then results in detrimental effects for the health of Black and Hispanics.

Social cohesion, both the absence of conflict and presence of social bonds, is one aspect of neighborhood social environments that is theorized to impact health positively.

Proposed pathways include through the effects of social influences on health behaviors, access to and knowledge about health care services, as well as through psychosocial processes of stress buffering and meaningfulness attached to social ties (Kawachi and

Berman 2000). Support from friends and neighbors may discourage unhealthy behaviors and encourage healthy ones as well as provide assistance for health when health concerns arise. In a school environment, social cohesion also likely increases communication among parents and the school, which in turn may increase the adoption of health behaviors and enforce healthy norms among parents and children, as well as increasing the knowledge about health programs, activities, and insurance programs for their children (Kawachi, Kennedy, and Glass 1999).

Social cohesion also provides a buffering effect against neighborhood stressors.

Social cohesion and informal social control not only reduce homicide and violent behavior directly (Maimon and Browning 2010; Morenoff, Sampson, and Raudenbush

2001) but also buffer the chronic stress of disadvantaged neighborhoods through social

54 support (Morenoff 2003). Among adults, neighborhood social support reduces depression, especially among women, even after accounting for the harmful effects of stressors such as physical decay of buildings, presence of graffiti, litter, and abandoned cars, as well as perceived violence (Mair et al. 2010). Children living in neighborhoods with low social cohesion and high social disorder also have increased risk of psychological problems (Lima et al. 2010). Although results are more mixed for physical health outcomes, there is evidence that both adults and children living in cohesive neighborhoods get more physical activity, especially outdoor activity time (Echeverria et al. 2008; Kimbro, Brooks, and McLanahan 2011; Molnar et al. 2004).

Social cohesion among parents at a school is likely to facilitate collective action to increase the availability of school resources that benefit child well-being. Wilson and

Taub (2006; 184) document how social cohesion can also overcome racial tensions – at least temporarily – when common problems arise. For example, they report instances in two neighborhoods where parents of multiple races came together to solve busing and school council at the local public schools where they shared common concerns for their children, even if the underlying racial tensions did not recede altogether. However, the buffering effect of social support has been shown to vary across social statuses such as race, socioeconomic status, and gender. Social support may be less effective at buffering the deleterious outcomes typically associated with neighborhood disorder in the most disadvantaged neighborhoods where the level of stressors that need to be overcome tend to be greater (Wight, Botticello, and Aneshensel 2006). In racially integrated neighborhoods, community cohesion is not associated with better mental health outcomes

55 among African Americans (Gary, Stark, and LaVeist 2007). Therefore, the presence of both neighborhood stressors and community support jointly must be accounted for in studying neighborhood effects on health, with an emphasis on how these effects by race.

Although research on racial disparities in health has explored the role of perceived discrimination on health and found consistent links with mental health and to a lesser extent physical health, few have studied the effects of discrimination on health in children, especially for physical health outcomes. A second, separate line of inquiry has studied the effects of neighborhoods on health, focusing on the role of physical and social stressors such as crime, vacancy and disrepair, and lack of social cohesion. Although neighborhood racial tensions have been discussed in studies of neighborhood choice where Blacks report avoiding predominantly White neighborhood for fear of hostility

(e.g. Krysan and Farley 2002), racial tensions have not been studied as a potential source of neighborhood stress in relation to health outcomes. The goal of the present study is to combine these two lines of inquiry to measure how much racial tensions are a stressor on the physical health of children, and whether the health effects of racial tensions vary by race. Finally, because social cohesion may buffer the effects of neighborhood stressors, I also assess whether social cohesion buffers the effects of racial tensions on child health.

Using data from the Early Childhood Longitudinal Survey Kindergarten Class of 1998-

1999 data (ECLS-K), I test the following hypotheses.

1) As a marker of neighborhood stress and discrimination, I expect racial and

ethnic tensions in the neighborhood to be negatively associated with child

physical health overall.

2) Because Whites report much lower levels of any kind of discrimination

(Kessler et al. 1999; Schulz et al. 2000) and health out comes among Whites

have not been shown to be sensitive to racial segregation (Subramanian et al.

2005), I expect racial and ethnic tensions to be negatively associated with

physical health for Hispanic and African American children, but not White

children.

3) Finally, since social cohesion is a buffer against neighborhood stressors, I

expect racial and ethnic tensions to be less detrimental to child physical

health in neighborhoods with high social cohesion.

Data and Methods

Data

The data for the subsequent analyses comes from the restricted version of the

Early Childhood Longitudinal Survey Kindergarten Class of 1998-1999 data. The ECLS-

K is a nationally representative longitudinal study of approximately 20,000 kindergarteners in 1998-1999 that were re-interviewed in first grade (1999-2000), third grade (2002), fifth grade (2004), and eighth grade (2007). Pertinent information was collected from students (i.e. respondents), parents, teachers, school administrators, as well as directly assessed by specially trained interviewers. Only data from the fifth wave

(8th grade) is used in the analyses to allow respondents to be old enough to experience and understand neighborhood race relations as well as providing questions about asthma diagnosis as well as obesity that are not asked in the early waves of the ECLS-k.

Dependent Variables

I capture two dimensions of physical health, weight and respiratory health, to understand the ways neighborhood racial tensions may vary by health outcome. Racial disparities by obesity and asthma are significant, and both conditions been linked to neighborhood conditions and stress in past research (Cagney et al. 2007; Franzini et al.

2009; Rosenbaum 2008). Body Mass Index (BMI) is calculated from interviewers‟ measurements of height and weight during each round of child assessments, as mass in kilograms divided by height in meters squared, is compared to gender-specific BMI-for- age growth charts from the Centers for Disease Control and Prevention (CDC). I categorize children as obese if their combined height and weight measurements put them at or above the 95th percentile for their gender and age, where there is an increased risk of secondary complications such as cardiovascular disease, and based on recommendations from the CDC and Institute of Medicine (Krebs et al. 2007;

Kuczmarski et al. 2002). Parents also report whether their child has ever been diagnosed with asthma by a doctor, and are compared to those who do not report asthma diagnosis.

Independent Variables

The independent variables of interest capture neighborhood racial/ ethnic tensions and social cohesion. School administrators are asked to respond to a series of questions about the neighborhood in which the school is located, by indicating whether it is a big problem, somewhat of a problem or no problem. They are asked whether there are tensions based on racial, ethnic, or religious differences on a scale from. I compare neighborhoods where racial tensions are either a big problem or somewhat of a problem

58 to neighborhoods that have no problems since only a small percentage report big problems. Finally, as a measure of social cohesion, I account for whether the school administrator perceives that parents are actively involved in the school‟s programs

(1=agree/strong agree, 0= neither agree/disagree, disagree, or strongly disagree).

Although it would be ideal to measure racial tensions and social cohesion in the exact neighborhood of residence, being able to use school administrator ratings of neighborhood and school contexts also helps eliminate self-reporting bias in perceived conditions of neighborhood and school contexts from parental reports (Diez Roux and

Mair 2010).

I control for a number of demographic and socioeconomic characteristics including child‟s gender (male is reference), age (in years), race (non-Hispanic White, non-Hispanic Black, and Hispanic), total household size, whether or not the child resides in a single parent household, and whether the child has no health insurance coverage

(compared to either private or government insurance). I also incorporate a composite measure of socioeconomic status constructed by the ECLS-K that combines mother‟s and father‟s educational attainment, income, and occupational prestige and categorizes students into quintiles based on these measures with higher values indicating higher socioeconomic status.

Analytic Strategy

Analyses are restricted to respondents who have complete data on variables of interest including school administrator responses about neighborhood and school climate.

This provides information from a sample of 4,720 non-Hispanic Whites, 600 non-

Hispanic Blacks, and 1,150 Hispanics, and a total of 6,470 observations. I employ logistic regression techniques to predict the likelihood of being obese and asthma diagnosis. I employ the Huber-White correction to standard error calculations and cluster individual observations within schools to account for heteroskedasticity and non-independence of observations. Sampling weights are applied to descriptive analyses to correct for both the survey non-response of those who are lost to follow up and the complex survey design of the ECLS-K. Unweighted sample sizes are rounded to the nearest 10 per the restricted data agreement with the Institute of Education Services and the National Center for

Education Statistics

Results

Table 7 presents the weighted descriptive statistics by race. Clear racial disparities can be seen in obesity, as 17% of Whites, 23% of African Americans, and 25% of Hispanic children being categorized as obese. More Blacks have been diagnosed with asthma,

19%, than either Whites (16%) or Hispanics (13%), which reflects previous findings

(Boardman et al. 2001). Blacks and Hispanics experience more neighborhood racial/ethnic tensions than Whites, but surprisingly, more Hispanics, 37%, attend schools where neighborhood tensions are present, compared to just 34% of Blacks. Less than one-half of African American adolescents go to schools where the administrators

60 perceive high parental involvement in school programs; slightly more Hispanic children

(62%) and considerably more White children (74%) experience this type of social cohesion. Hispanics are also more likely to have no health insurance (11%) or government health insurance (25%) compared to Whites. Two-thirds of Blacks and

Hispanics but 90% of Whites have private or military health insurance coverage. Finally, socioeconomic disparities across racial groups are substantial. Blacks (30%) and especially Hispanics (40%) are more likely to be in the lowest socioeconomic quintile of education, occupation, and income than Whites (8%). Similar disparity is seen at the highest quintile, with 30% of Whites but just 5% of Blacks and 8% of Hispanics are situated in this category.

Table 8 presents the logistic regression results predicting the odds of being obese, stratified by race. Model 1 includes the full sample, where all three racial/ethnic groups are combined. In this model, Hispanic children are 35% more likely to be obese than

White children, and although Black children are 28% more likely to be obese, the coefficient fails to reach significance. As expected, the presence of neighborhood racial/ethnic tensions increases the odds of obesity by 20%. In contrast, parental involvement at the school reduces the odds of obesity by 24%.

In Model 2, which includes Whites only, the presence of neighborhood tensions increase the odds of obesity by 18%, but the coefficient is not significant. Parental involvement reduces the risk of obesity by 2%. Among African Americans, in Model 3 neighborhood tensions reduce the likelihood of obesity by 18% percent, the opposite direct as expected, but once again the coefficient is not significant. Parental involvement

61 also reduces the odds of obesity but is not a significant effect. Finally, in Model 4, among

Hispanics neighborhood racial/ethnic tensions increases the odds of obesity by 39%, and similar to Whites, parental involvement decreases the odds of obesity by 29%.

Table 9 presents the logistic regression results predicting the risk of asthma diagnosis. Model 1 illustrates results from the full model, pooling all three racial groups together. Black children are 38% more likely to be diagnosed with asthma than Whites.

Hispanics are 17% less likely to be diagnosed with asthma, but the difference is not statistically significant. In the full model and the stratified models by race, neither racial/ethnic tensions nor parental involvement are significant predictors of the likelihood of being diagnosed with asthma. The only other covariate found to consistently influence the odds of this disease is gender, with girls facing a 30% lower odds of receiving an asthma diagnosis.

Tables 10 and 11 add the interaction between racial/ethnic tensions and parental involvement to the models, to assess whether the effects of racial/ethnic tensions depend upon the levels of social cohesion. Adding the interaction term does not reduce the higher likelihood of obesity for Hispanics, as shown in Model 1 of Table 10. In this model, high parental involvement is still associated with a reduction in the odds of obesity by 31%, but the main effects of neighborhood tensions and the interaction term are not significant.

Among Whites, only the high parental involvement main effects are significant, so the odds of obesity are 34% lower among those with high parental involvement in school programs compared to those with low involvement, but only when racial tensions are not

62 present. Among Blacks and Hispanics, neither racial tensions nor parental involvement are significantly associated with the odds of obesity.

Table 11 presents the interaction models for asthma diagnosis. Adding the interaction term does not reduce the higher likelihood of asthma among Blacks, as shown in Model 1. In the full model, racial/ethnic tensions main effects are now negative, with a

28% decrease in the odds of asthma. However, the interaction term with parental involvement is positive and significant, suggesting that for youths exposed to both neighborhood tensions and high parental involvement, the odds of asthma are increased by 25% (=ln(.72)+ln(1.74)) relative to those exposure to tensions or high involvement, or neither. For all three groups the effects of tensions in the presence of involvement are positive, but only among Whites is the interaction term significant. There is an 87% increase in asthma in the presence of both racial tensions and high parental involvement compared to White children with only tensions or high involvement, or neither.

Discussion

The main findings of this study suggest that the role of racial and ethnic tensions in neighborhoods in understanding childhood obesity is complex. Tensions are associated with a significant increase in obesity, but only among Hispanic children. Hispanics who experience racial/ethnic neighborhood tensions are almost 40% more likely to become obese by eighth grade than Hispanic children who do not experience racial tensions.

Given the high rates of obesity among Hispanics as a whole, exposure to racial and ethnic tensions may represent a particular stressor that increases vulnerability to obesity that furthers the need for public health interventions designed to help alleviate child obesity.

Parental involvement in schools is an important buffer against obesity but only for

Whites and Hispanics; White and Hispanic children who attend high parental involvement schools face 26% and 29%, respectively lower odds of obesity than children who attend schools characterized by low parental involvement. Among White children, parental involvement is a partial buffer against obesity, but only in the absence of racial/ethnic tensions, suggesting that parental involvement is protective against obesity but does not work as a buffer in more stressful environments.

Surprisingly, neither racial tensions nor parental involvement levels predict obesity and asthma among African Americans. Although perceived discrimination is important to understanding health disparities in African Americans, the connection is less conclusive for physical health. Perceived discrimination has been linked to hypertension and obesity but less consistently than the link between discrimination and poor mental health outcomes (Williams and Mohammed 2009). The lack of relationship between racial tensions and physical health may be reflecting the weaker link between discrimination and physical health. It may also be that racial/ethnic tensions are not capturing similar dimensions of stress as perceived discrimination. Additionally, because of high rates of perceived discrimination among African Americans, neighborhood racial tensions may not be more stressing in addition to interpersonal discrimination experiences.

Asthma diagnosis appears to be less sensitive to the neighborhood and school contexts examined here. Neither racial/ethnic tensions nor parental involvement are significant predictors of asthma diagnosis. Among Whites, the presence of high parental

64 involvement results in decreased odds of asthma diagnosis, but only in the absence of racial/ethnic tensions. High parental involvement may be beneficial to respiratory health among Whites, but may not be able to overcome the stressors involved in exposure to racial/ethnic tensions.

Two limitations of this study are important to consider. First, it would be better to have measures of neighborhood tensions and social cohesion in the respondents‟ neighborhoods instead of their school neighborhoods. Unfortunately, the ECLS-K does not ask parents about the racial climate or social cohesion in their own neighborhood contexts. The ECLS-K questions on neighborhood tensions also do not have any detail about the nature of racial and ethnic tensions, such as what particular groups are involved or whether they have been related to violence or crime. Second, small (less than 100) cell sizes once dividing each racial group into the four types of neighborhood and school contexts may be limiting statistical power to be able to detect interactions between racial tensions and parental involvement, and this may be especially important due to the smaller sample size of African American children relative to Hispanics and Whites.

Despite these limitations, this study contributions to the literature on racial discrimination and neighborhood effects on physical health during childhood. Obesity levels among Hispanic children are the most sensitive to racial/ethnic neighborhood tensions, and are at increased risk of obesity with exposure to these racial tensions, especially in the absence of parental involvement in schools. More research is needed to unpack the role of racial and ethnic tensions on health; using more detailed measures of the type and level of conflict may present a more nuanced view of the relationship

65 between racial/ethnic tensions and health. Racial disparities in health are significant despite large gains in public health knowledge and health services in the last century

(Phelan and Link 2005). How individual-level characteristics such as race interact with contextual-level characteristics such as neighborhood poverty or racial tensions continues to be an important line of inquiry in understanding and explaining racial disparities in health.

White Black Hispanic Obese, % 17% 23% 25% Asthma, % 16% 19% 13%

Neighborhood Tensions, % 22% 34% 37% High Parental Involvement at School, % 74% 47% 62%

Female, % 48% 49% 49% Mean Age, in years 14.32 14.27 14.26 Single Parent Household, % 17% 47% 25% Mean Household Size 4.37 4.51 4.75 No Health Insurance, % 3% 3% 11% Government Health Insurance, % 8% 31% 25% Private or Military Health Insurance, % 89% 67% 64%

Socioeconomic Status

Lowest Quintile, % 8% 30% 40% Quintile 2, % 17% 26% 23% Quintile 3, % 21% 21% 17% Quintile 4, % 24% 19% 13% Highest Quintile, % 30% 5% 8%

Percent of Sample 66% 15% 19% Number of Respondents 4760 620 1150 Table 7: Weighted Neighborhood, School, and Individual Characteristics, by Race

Full Sample White Black Hispanic African American 1.28

(0.15)

Hispanic 1.35**

(0.12)

Neighborhood Tensions 1.20* 1.18 0.86 1.39* (0.10) (0.13) (0.18) (0.20)

High Parental Involvement 0.76** 0.74* 0.86 0.71* (0.06) (0.07) (0.17) (0.10)

Female 0.66** 0.59** 1.60* 0.56** (0.04) (0.05) (0.29) (0.08)

Age, in years 0.97 1.02 1.13 0.77 (0.09) (0.11) (0.26) (0.16)

Single Parent Household 1.02 1.09 0.98 0.96 (0.10) (0.14) (0.20) (0.18)

Household Size 0.94* 0.94 0.86* 1.01 (0.03) (0.04) (0.06) (0.05)

No Health Insurance 1.43* 1.81** 1.1 1.13 (0.22) (0.38) (0.51) (0.27)

Government Health Insurance 1.12 1.07 1.21 1.08 (0.12) (0.18) (0.26) (0.20)

Socioeconomic Quintiles 0.80*** 0.78*** 0.91 0.86* (0.02) (0.03) (0.08) (0.05)

N 6530 4760 620 1150 χ2 253.86 157.38 15.71 50.25 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 8: Odds Ratios from Logistic Regression of Neighborhood and School Contexts on Obesity

Full Sample White Black Hispanic African American 1.38*

(0.18)

Hispanic 0.83

(0.10)

Neighborhood Tensions 1.01 1.05 1.13 0.84 (0.09) (0.12) (0.28) (0.18)

High Parental Involvement 0.97 1.01 0.99 0.8 (0.08) (0.10) (0.23) (0.17)

Female 0.68** 0.69** 0.63* 0.63* (0.05) (0.06) (0.13) (0.12)

Age, in years 1.02 1.01 0.68 1.32 (0.10) (0.12) (0.20) (0.36)

Single Parent Household 1.21 1.17 1.19 1.41 (0.12) (0.15) (0.27) (0.33)

Household Size 0.95 0.95 1.04 0.91 (0.03) (0.04) (0.07) (0.08)

No Health Insurance 1.08 0.94 0.97 1.39 (0.19) (0.25) (0.58) (0.40)

Government Health Insurance 1.24 1.27 1.62 1.07 (0.15) (0.22) (0.40) (0.27)

Socioeconomic Quintiles 1.05 1.00 1.14 1.23* (0.03) (0.04) (0.11) (0.09)

N 6530 4760 620 1150 χ2 64.00 28.04 11.41 22.53 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 9: Odds Ratios from Logistic Regression of Neighborhood and School Contexts on Asthma

Full Sample White Black Hispanic African American 1.27

(0.15)

Hispanic 1.35**

(0.12)

Neighborhood Tensions 1.02 0.9 0.79 1.32 (0.12) (0.15) (0.23) (0.26)

High Parental Involvement 0.69** 0.66** 0.81 0.68 (0.06) (0.07) (0.20) (0.12)

Tensions * Involvement 1.34 1.57 1.21 1.1 (0.21) (0.34) (0.49) (0.32)

Female 0.66** 0.59** 1.60* 0.56** (0.04) (0.05) (0.30) (0.08)

Age, in years 0.97 1.02 1.13 0.77 (0.09) (0.11) (0.26) (0.16)

Single Parent Household 1.02 1.09 0.99 0.96 (0.10) (0.14) (0.20) (0.18)

Household Size 0.94* 0.94 0.86* 1.01 (0.03) (0.04) (0.06) (0.05)

No health insurance 1.43* 1.82** 1.09 1.13 (0.22) (0.38) (0.50) (0.27)

Government health insurance 1.12 1.08 1.21 1.08 (0.12) (0.18) (0.26) (0.20)

Socioeconomic Quintiles 0.80** 0.78** 0.91 0.86* (0.02) (0.03) (0.08) (0.05)

N 6530 4760 620 1150 χ2 252.17 158.64 15.95 49.99 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 10: Odds Ratios from Logistic Regression of Neighborhood and School Interactions on Obesity

Full Sample White Black Hispanic African American 1.38*

(0.18)

Hispanic 0.83

(0.10)

Neighborhood Tensions 0.72* 0.69 0.78 0.74 (0.10) (0.13) (0.25) (0.23)

High Parental Involvement 0.83 0.86 0.75 0.73 (0.08) (0.09) (0.20) (0.20)

Tensions * Involvement 1.74** 1.87* 2.18 1.28 (0.31) (0.44) (1.05) (0.55)

Female 0.68** 0.69** 0.63* 0.62* (0.05) (0.06) (0.13) (0.12)

Age, in years 1.01 1 0.69 1.33 (0.10) (0.12) (0.21) (0.35)

Single Parent Household 1.21 1.16 1.22 1.41 (0.12) (0.15) (0.27) (0.33)

Household Size 0.95 0.95 1.04 0.91 (0.03) (0.04) (0.07) (0.08)

No Health Insurance 1.08 0.94 0.93 1.39 (0.19) (0.25) (0.55) (0.40)

Government Health Insurance 1.23 1.28 1.59 1.06 (0.15) (0.22) (0.39) (0.26)

Socioeconomic Quintiles 1.06 1.00 1.15 1.23* (0.03) (0.04) (0.11) (0.09)

N 6530 4760 620 1150 χ2 71.24 34.87 15.5 22.39 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 11: Odds Ratios from Logistic Regression of Neighborhood and School Interactions on Asthma

Chapter 4: Segregation, Family Stress, and Racial Disparities in Childhood Physical Health

Hispanics and non-Hispanic Blacks now comprise 23% and 14%, respectively, of children under the age of 18 (Mather, Pollard, and Jacobsen 2011). Racial disparities in physical well-being begin to development early on in the life course, as poor health disproportionately affects minority children relative to Whites (Kimro, Brooks-Gunn, and

McLanahan 2007; Von Hippel et al. 2007). In addition, the racial inequalities have widened in the past decade. By 2007, 24% of Hispanic and African American children were obese, compared to just 13% of White children (Singh, Siapush, and Kogan 2010).

Although immigrants tend to fare better than expected with regard to health relative to

Whites given their relatively low socioeconomic status (Singh and Siahpush 2001;

Hummer et al. 2007), immigrant Hispanic children still face elevated odds of developing obesity and poor physical health in early life (Crosnoe 2006). Although socioeconomic status (SES) differences across racial groups have been hypothesized to explain these racial gaps, racial differences in childhood obesity and overall physical health persist even after accounting for differential access to financial resources and exposure to poverty, as well as behavioral differences such as physical activity (Singh et al. 2010).

Neighborhood conditions have increasingly played a role in our understanding of individual health outcomes, especially in regards to physical activity and obesity

72 outcomes (Black and Macinko 2008; Diez Roux and Mair 2010). For example, measures of the neighborhood built environment, such as walkability, physical activity spaces, and access to healthy foods, have all been linked to reduced obesity levels (Papas et al. 2007).

Neighborhood stress in the form of physical and social disorder, such as vacant housing, crime, and physical decay, have also been associated with increased rates of obesity

(Burdette and Hill 2008; Fish et al. 2010). However, these resources are also spatially patterned by structural features of neighborhood contexts. Lower-SES and high-minority compositions are inversely related to physical activity resources, and consequently, positively related to overweight and obesity (Gordon-Larsen et al. 2006; Boardman et al.

2005). Moreover, Black and Hispanic children experience worse neighborhood conditions relative to whites, including higher rates of exposure to neighborhood poverty and crime (Timberlake 2007; Peterson and Krivo 2009). Differences in neighborhood contexts may help explain racial disparities in child obesity and poor health.

Residential segregation is a key component of the neighborhood context that has been shown to contribute to racial disparities in health (Williams and Sternthal 2010;

Kramer and Hogue 2009). Although Black segregation from Whites is more pervasive than for other minority groups, Hispanic-White segregation remains stable and does not show evidence that it is likely to decrease over time (Wilkes and Iceland 2004; Iceland and Wilkes 2006). Residents of segregated minority neighborhoods typically experience excessive rates of social as well as physical disorder, such as crime, vacant housing, and lack of economic development (Charles 2003; Krivo, Peterson, and Kuhl 2009; Massey and Denton 1993). Residents living in segregated neighborhoods are also more frequently

73 exposed to stressful life events and chronic strains stemming from underemployment, unemployment, family instability, and residential mobility, including the risk of foreclosure (Card and Rothstein 2007; Rugh and Massey 2010).

Given these circumstances, we would expect that minority children in segregated neighborhoods may be exposed to more stressful family circumstances, such as parental unemployment and single parenthood (Cutler and Glaser 1997). As a result, children experiencing stressful family events may be more vulnerable to the negative health effects of stress by living in segregated neighborhood with fewer family and neighborhood resources to cope with stressful events. This added vulnerability to stress in segregated neighborhoods has not been tested on child health outcomes. This study estimates the extent to which racial and ethnic disparities in child physical health are the result of the combined influences of residential segregation and family stress. Using data from the Early Childhood Longitudinal Study Kindergarten Class of 1998-1999 (ECLS-

K), I estimate whether the negative effects of family stressful events on likelihood of obesity are worse in the context of residential segregation.

Stress, Segregation, and Health

The stress process incorporates multiple domains of stress to explain how stress influences health outcomes: sources of stress, mediating mechanisms such as coping and social support, and how symptoms are manifested within individuals in psychological and physiological functioning (Pearlin 1999). Differential exposure to stressors is a primary mechanism though which race, class, and gender inequality is transformed into health disparities (Turner 2003; Thoits 2010). Social statuses such as race, class, and gender also

74 alter access to or efficacy social support and coping mechanisms, which result in suboptimal mental and physical health manifestations as a result of disadvantaged social statuses. Sources of stress include major life events as well as chronic strains of daily life, which both negatively impact health; furthermore, the combination of both types of stressors exerts a cumulative effect over the life course (Pearlin 1999; Pearlin et al. 2005).

Although stress was originally developed to study mental health outcomes such as depression, it is also detrimental to physical health through increased hormone responses that are measured by stress-related biomarkers such cortisol and inflammatory responses

(Boardman 2004; McEwan 1998).

For children, the family environment provides a critical dimension within which the stress process can unfold, especially at younger ages when children spend much of their non-school time at home (Avison 2010). Stressful family events such as separation, divorce, as well as chronic stressors that can stem from ongoing conditions like financial strain or chronic illness of parents negatively affect parental emotional control and parenting quality, which in turn detrimentally influence child development and well- being (Conger et al. 2010). Multiple family life events have been shown to significantly decrease health-related quality of life in children as young as 11 and 12, including physical, emotional, and social functioning (Coker et al. 2011). Family stressful events have also been linked to depression in adolescence and early adulthood (Boardman and

Alexander 2011; Natsuaki et al. 2007).

Racial differences in the exposure to family stress may be an important driver of child health disparities. In the case of family instability such as marital or relationship

75 status changes, a much studied family stressor in relation to child development, the health and behavioral effects vary by race. Using the Fragile Families and Child Wellbeing

Study, Osborne and McLanahan (2007) find that African American children are more likely to experience multiple parental union transitions than Whites. They also find that both minority and White children experience more behavioral problems as a result of family transitions, so although transitions are not worse for African Americans than

White, African American children still experience more transitions, and this, overall, more problems as a result.

Increasingly, health researchers have focused on the role of neighborhood contexts in examining racial differences in stress exposure, coping mechanisms, and health disparities (Aneshensel 2010; Williams and Sternthal 2010). The role of socioeconomic status and race in stress process models has previously been defined at the individual level, but recent conceptualizations also emphasize the importance of interpreting stress on an aggregate level, particularly that of the neighborhood (Wheaton and Clarke 2003; Pearlin 1999). Furthermore, individual-based explanations of health and well-being fail to capture important determinants of ill-health, and both physical and social attributes of neighborhood contexts are needed to fully understand health outcomes and disparities (Diez Roux and Mair 2010). Neighborhood effects on health are not only partially explained by individual exposure to stress, but also conditional upon individual social status, exposure to stress, and access to psychosocial resources (Aneshensel 2010).

Residential segregation is a key defining characteristic of urban landscape that has potential to help explain the existence of longstanding racial disparities in childhood

76 health because of vastly different residential experiences among Blacks, Hispanics, and

Whites (Williams and Jackson 2005; Acevedo-Garcia et al. 2003). As of the 2000 U.S.

Census, 29 metropolitan areas are classified as hypersegregated along White-Black lines, indicating high levels of segregation on at least four out of five dimensions of segregation: evenness, exposure, concentration, centralization, and clustering (Wilkes and

Iceland 2004). Although Black-White segregation is declining, particularly in cities with relatively small Black populations, 40% of all African Americans live in metropolitan areas that are hypersegregated (Iceland, Weinberg, and Steinmetz 2002; Massey 2004).

Using new data from the Census 2010, Logan and Stults (2011) report that the typical

African American still lives in a neighborhood that is 45% Black; similarly, the typical

Hispanic now lives in a neighborhood that is 46% Hispanic. Metropolitan-area residential segregation levels are lower for Hispanics, but unlike Black-White segregation, levels are not decreasing (Iceland et al. 2002; Iceland and Wilkes 2006).

Segregated neighborhoods have been created through discriminatory housing market practices that concentrate poverty and reduce viable economic and educational pathways available to residents, resulting in high levels of physical and social disorder, which further reduces economic investment in segregated neighborhoods (Massey and Denton

1993; Wilson 1996). High rates of both physical and social neighborhood disorder – for example, crime, unemployment, failing infrastructure, high levels of residential instability and resident mobility – are stressful to residents because they reduce the ability of local residents to maintain an orderly public space, foster mutual trust, and create social cohesion and social capital (Sampson, Raudenbush, and Earls 1997; Hill, Ross, and

Angel 2005). Additionally, both Black and White residents perceive more social and physical disorder in predominantly Black neighborhoods than those living in non-Black neighborhoods, although the relationship is stronger for Whites than Blacks (Quillian and

Pager 2001; Schulz et al. 2008). As a result, segregated neighborhoods may be particularly distressing to residents through the economic disinvestment and reduced social cohesion in segregated neighborhoods and metropolitan areas. However, research on immigrants and ethnic enclaves suggest that living in an ethnic enclave is likely to be beneficial because of cohesive social networks, reduced cultural and language barriers, and social support they provide to same-race and co-ethnic neighbors, suggestions a potentially beneficial results of segregation for Hispanics (Portes and Zhou 1993).

Although residential segregation levels are lower among Hispanics than African

Americans, the consequences of unequal residential contexts tend to be detrimental to the well-being of both groups, especially regarding socioeconomic attainment. Among

African Americans, segregation contributes to suboptimal educational and employment attainment (Cutler and Glaeser 1997; Card and Rothstein 2007; Massey and Fischer

2006). Among both African Americans and Hispanics, segregation exacerbates levels of crime, victimization, and homicide (Krivo et al. 2009; Feldmeyer 2010), and is associated with lower rates of homeownership as well as lower housing quality among homeowners and high rates of foreclosure (Flippen 2010; Rugh and Massey 2010).

Segregated neighborhoods also have structural features that may contribute to poor population health. Segregated neighborhoods have fewer grocery stores, as well as reduced availability of inexpensive healthy foods, recreational facilities, and parks which

78 may contribute to more unhealthy behaviors and (Zenk et al. 2005; Moore and Diez Roux

2006; Moore et al. 2008). Residents of segregated neighborhoods are also exposed to increased pollution, lead, and other environmental toxins that are detrimental to health

(Morello-Froesch and Lopez 2006). These consequences of residential segregation are hypothesized to be part of the link between segregation and health because of the concentration of disadvantage, resulting in both added stressors for residents but also fewer socioeconomic resources and social coping mechanisms.

1996), and Black isolation is positively associated with poor self-rated health, obesity, and low birth weight (Subramanian et al. 2005; Chang 2006; Bell et al. 2006; Grady and

Ramirez 2008). Black mothers in hypersegregated metropolitan areas experience higher rates of pre-term birth than Blacks in hypersegregated areas, and in addition, Black-White gaps in preterm birth are also larger in hypersegregated areas (Osypuk and Acevedo-

Garcia 2008).

However, the association between segregation and well-being has not been entirely negative among African Americans, and may partly reflect differences in which segregation dimension measures are used. For example, racial clustering has been shown to be predictive of better birth outcomes among Blacks, although this study only includes mothers living in New York City (Bell et al. 2006). More importantly, measures of residential evenness have not been as closely related to health outcomes as racial

79 isolation; several studies do not find a link between Black-White dissimilarity and health among African Americans (Subramanian et al. 2005; Walton 2009). The dissimilarity index may have weaker theoretical link between evenness than isolation as it only measures the local residential patterns in relation the wider MSA composition (Acevedo-

Garcia et al. 2003; Kramer and Hogue 2009). Isolation may be a better representation of unequal experiences and be particularly detrimental to the extent that although it requires uneven residential patterns, high isolation represents many neighborhoods with high minority concentrations. Although generally segregation is considered to be detrimental to the well being of African Americans, existing research is not entirely consistent in these findings, and understanding how different measures of residential segregation are related to health is important for understanding the complex ways in which residential contexts affect well-being.

Current evidence is more inconclusive regarding the relationship between residential segregation and health among Hispanics. The association between segregation and health varies across racial/ethnic identity, immigrant status, and outcome measures.

Furthermore, many studies rely on measures of racial composition rather than residential segregation. Several studies find positive health effects of concentrated Hispanic neighborhoods. For example, higher Hispanic concentration is associated with lower levels of depression, better self-rated health, and lower mortality rates and mortality rates and better self-rated health among elderly Mexicans living in the Southwest (Ostir et al.

2003; Patel et al. 2003; Eschbach et al. 2004). Among children and adolescents, Hispanic

80 neighborhood concentration significantly predicts with healthier dietary habits overall

(Lee and Cubbin 2002).

Although racial composition measures tap some of the complex residential patterns of Hispanics, they cannot account for wider spatial patterning that is shaped by housing market discrimination. Studies using measures of segregation based on wider regional or metropolitan residential patterns have found that the health effects of living among co-ethnics are not always beneficial for Hispanics. Residential isolation is associated with worse physical health for Puerto Rican adults and increases depressive symptoms for Mexican adults (Lee and Ferraro 2007; Lee 2009). Hispanics are also at an increased risk of substance use and delinquency in neighborhoods with high concentrations of poor Hispanics, especially for native-born youths (Frank, Cerda, and

Rendon 2007). In sum, the findings regarding residential segregation and health among

Hispanics are not yet conclusive, but suggestive of a positive relationship between segregation and poor health. More research is needed to untangle the role of residential segregation using indices of segregation while examining health across the life course.

In addition to the role of socioeconomic status and neighborhood quality, part of the relationship between residential segregation and health is thought to be due to racial discrimination (Williams and Sternthal 2010). Negative social interactions, such as unfair treatment stemming from racist attitudes, contribute to racial disparities in health by elevating stress and taxing coping resources available to deal with these interactions

(Schulz et al. 2000; Williams and Sternthal 2010). Perceived discrimination has been linked to increased psychological distress among African American adults (Williams et

81 al. 1997; Jackson et al. 1996; Schulz et al. 2006) and both African American and

Hispanic children (Coker et al. 2009; Wong, Eccles, and Sameroff 2003). Discrimination experiences are also associated with increased obesity, unhealthy behaviors, and mortality for African American adults (Hunte and Williams 2009; Borrell et al. 2010).

In one of the few studies assessing the combined impact of segregation and family stress on health, Charles, Dinwiddie, and Massey (2004) use data from the National

Longitudinal Survey of Freshmen (NLSF), which surveys a representative sample of the entering freshman cohort at selective colleges and universities in 1999 and follows them for two years, to examine the role of family stress and segregation on the health of college students. They measure the number of stressful events such as crime, unplanned pregnancy, and illness among immediate family members that have occurred during the previous year. They then calculate the minority concentration of respondents‟ neighborhood during high school as their measure of segregation, and create a scale of poor health that includes visits to a psychological counselor, the student health center, and frequency of loneliness, homesickness, and serious illnesses. They find that that growing up in a high minority concentration neighborhood increases the number of stressful family events experienced and decreases overall health among college students, especially among African Americans. However, to the extent that NLSF respondents, who are minority students at selective universities, are not representative of the experiences most Hispanic and African American youths encounter in their daily lives, these empirical findings may not accurately capture the impact of residential segregation on racial disparities in family stress and health outcomes.

This study aims to address the role of residential segregation and stressful family events on the health and well-being African American, Hispanic, and White children. I assess whether residential segregation affects how children cope with family stressful events, and how this, in turn, impacts physical health outcomes. Although it is well documented that segregation increases family stress in a global sense, it is unclear whether segregation exacerbates children‟s vulnerability to family stress, and whether the effect differs across dimensions of residential segregation. That is, is there an added effect of family stress on child health in the presence of segregation? Given the pervasiveness of residential segregation, for both Hispanic and African American children, the ability of economic and psychosocial resources to overcome family stressful events may be reduced in the context of segregation. Using data from the ECLS-K, I test the following hypotheses.

1) Family stress will be positively associated with poor health outcomes.

2) Black-White and Hispanic-White dissimilarity will not moderate the

relationship between family stress and health; the effect of stress on health

will be similar across levels of dissimilarity.

3) Black and Hispanic isolation will moderate the relationship between family

stress and health; the effect of stress on health will be higher as isolation

increases.

Data and Methods

Data

I use data from the restricted version of the Early Childhood Longitudinal Survey

Kindergarten Class of 1998-1999 data (ECLS-K). The ECLS-K is a nationally representative longitudinal study of approximately 20,000 kindergarteners in 1998-1999 that were re-interviewed in first grade (1999-2000), third grade (2002), fifth grade

(2004), and eighth grade (2007). Pertinent information was collected from students (i.e. respondents), parents, teachers, school administrators, as well as directly assessed by specially trained interviewers. Only data from the eighth grade wave is used for this analysis as it is the only wave that asks about critical family stressful events in the past year.

Dependent Variables

I capture health using one measure of physical health (obesity) and one measure of global health (parent-rated health). I include these measures because they comprise a comprehensive look at health, and especially racial disparities in child health. Racial disparities in obesity are well-documented, starting as young as three years old (Kimbro et al. 2007), and risk of obesity is influenced by neighborhood conditions (Franzini et al.

2009). Finally, self-rated health overall is a strong predictor of mortality later in life, and so early differences in parent-rated health are important to asses (Jylha 2009).

Interviewers measured height and weight twice for each respondent, from which composite averaged height and weight measures are constructed. Body mass index

(BMI), calculated as mass in kilograms divided by height in meters squared, is compared

84 to gender-specific BMI-for-age growth charts from the Centers for Disease Control and

Prevention (CDC). I categorize children as being obese if their combined height and weight measurements put them at or above the 95th percentile for their gender and age, where there is an increased risk of secondary complications such as cardiovascular disease, based on recommendations from the CDC and Institute of Medicine (Krebs et al.

2007; Kuczmarski et al. 2002). Parents report how healthy the child is on a scale from 1, which indicates excellent health, to 5 which indicates poor health. In multivariate models, parent-rated health has been dichotomously recoded into poor health, where 1 represents fair or poor health and 0 represents good, very good, or excellent health.

Independent Variables

The independent variables of interest capture the racial segregation of the metropolitan area in which the respondent resides at the time of data collection.

Respondent‟s addresses are matched to the corresponding Metropolitan Statistical Areas

(MSA) or the Primary Metropolitan Statistical Areas (PMSA). I use two dimensions of segregation, evenness and exposure, calculated by the Census Bureau using Census 2000 data (Iceland, Weinberg, and Steinmetz 2002). Evenness, the degree to which the percentage of minority members in smaller residential areas matches the overall metropolitan area percentage, is measured by the dissimilarity index, which can be interpreted as the percent of minority residents (Black or Hispanic, respectively) who would have to move census tracts for each census tract to match the percentage of that minority that live in the MSA/PMSA as a whole. Exposure is the degree to which residents are exposed to members of other racial group by sharing the same

85 neighborhoods, and is measured in this study by the isolation index, the weighted average percentage Black or Hispanic within census tracts averaged across the MSA/PMSA

(Massey and Denton 1993). Both indices are on a scale of 0-100 with greater values indicating more residential segregation.

I measure the following stressful life events in the previous year, reported by parent experiences: being mugged, robbed, or attacked, having phone or electricity service cutoff from non-payment of bills, losing a job, moving residence, big changes in income, serious hospitalization or illness, changes in marital status, and experiencing a death in the family. I sum the number of events to create a measure of exposure to family stressful events. Because the distribution of this key independent variable is positively skewed with a concentration of respondents on the negative end of the distribution, I categorize events as zero, one, two, or three or more events in the last year. No stressful events is the reference category in all models.

I include a measure of parental social support that may act as a buffer against family stress. Parental involvement is reported by parents of the frequency that either they or another adult family participated in the following activities with the child: working on homework or school projects, attending school activities, concerts, plays or movies outside of school, sporting events outside of school, religious services, family social functions, taking day trips or vacations, working on a hobby or playing sports, going shopping, to restaurants or eating out, spending time just talking, watching television together, or doing something else fun together. Parents reported the frequency on a scale from 1, never, to 4 frequently. Responses are them summed across the 13 items

86 to create a scale of parental involvement, which had high consistency across the variety of responses, with a relatively high alpha level of 0.73.

I control for a number of demographic and socioeconomic characteristics including child‟s gender (male is reference), age (in years), race (non-Hispanic White is the reference, non-Hispanic Black, or Hispanic), total household size, whether or not the child resides in a single parent household, and whether the child has no health insurance coverage, government insurance (compared to either govern or government insurance). I also incorporate a composite measure of socioeconomic status (SES) constructed by the

ECLS-K that combines mother‟s and father‟s educational attainment, income, and occupational prestige and categorizes respondents into quintiles based on these measures with higher values indicating higher socioeconomic status.

Analytic Strategy

Analyses are restricted to respondents who have complete data on variables of interest including geocoded information and parent interview data on stress and involvement.

These restrictions provide data for a sample of 3,470 non-Hispanic Whites, 540 non-

Hispanic Blacks, and 1,140 Hispanics, for a total of 5,160 observations. I employ the

Huber-White correction to standard error calculations and cluster individual observations within MSA to account for heteroskedasticity and non-independence of observations.

Logistic regressions predict the odds of being obese and in poor health. Sampling weights are applied to descriptive analyses to correct for both the survey non-response of those who are lost to follow up and the complex survey design of the ECLS-K. Unweighted

87 sample sizes are rounded to the nearest 10 per the restricted data agreement with the

Institute of Education Services and the National Center for Education Statistics.

Results

Table 12 presents weighted descriptive statistics by race. Racial disparities are evident in obesity and self-rated health. 23% of Blacks and 26% Hispanic children are categorized as obese, but only 16% of White children fall into this category. Black and

Hispanic children are reported as being in fair or poor health twice as often (4% and 5%, respectively) as White children (just 2%). Black respondents are most likely to experience stressful family events during the previous year. 48% of White children and

44% of Hispanics, but only 30% of Black, experience no stressful events in the past year.

One third of all groups experience just one event, whereas 8% of Whites, 10% of

Hispanics, and 19% of Blacks experience 3 or more events. All three groups report similar levels of parent involvement, with an average score of 41.55 for Hispanics, 43.33 for Blacks, and 43.78 for Whites.

Average levels of Black-White and Hispanic-White dissimilarity reflect overall higher levels of Black as opposed to Hispanic segregation. African American children reside in MSAs that have a dissimilarity index of 63%, higher than the average White or

Hispanic child. More pronounced racial differences are noticeable for Black isolation levels, where the average Black child lives in an MSA with 58% isolation, compared to

46% and 45% of Whites and Hispanics, respectively. Hispanic-White dissimilarity indices for Hispanic children are at 50%, on average, compared to just 44% and 45% for

White and Black children. The racial differences in exposure to Hispanic isolation are

88 more prominent. Hispanic children live in MSAs that have an average of 53% Hispanics in census tracts. Blacks and Whites are exposed to Hispanic isolation levels of just under

30%, on average.

Almost half (45%) of Black children reside in single parent households, but only

16% of Whites and 24% of Hispanics reside in similar households. The socioeconomic disparity across racial groups is substantial. Compared to just 6% of Whites, Blacks

(29%) and especially Hispanics (41%) are more likely to be categorized in the lowest socioeconomic quintile of education, occupation, and income. Similar patterning is seen within the highest quintile. 34% of Whites but just 6% of Blacks and 9% of Hispanics have access to socioeconomic resources that place in the highest quintile. Hispanics are also much more likely to have no insurance coverage, at 12%, compared to just 3% of

Whites and Blacks. While 90% of White children have private or military health insurance, just 71% of Blacks and 64% of Hispanics have that same level of coverage.

Table 13 presents odds ratios from logistic regression predicting obesity and poor health. Only individual-level variables are included in these models. Both two and three or more stressful events are associated with an increase the likelihood of obesity relative to no events. Two stressful events increases the odds of obesity by 32%, whereas three events increases the odds of obesity by 46%. Both African American and Hispanic children experience increased odds of obesity as well, by 36% and 45%, respectively.

Model 2 presents the odds ratios for predicting poor health. Only 3 or more stressful events are significantly related to the probability of poor health; it is associated with an increase in the odds by 2.5 relative to those experiencing no stressful events. Once

89 controlling for these individual characteristics, the odds of poor health are higher for

Blacks relative to Whites, but the difference is not statistically significant. However,

Hispanic children are 2.16 times more likely to be in poor health.

Table 14 presents the results from interaction models assessing the joint effects of residential segregation as measures by the dissimilarity index and stressful events. For both outcomes, Model 1 includes the interactions with Black-White dissimilarity, Model

2 the interactions with Hispanic-White dissimilarity. None of the stressful events and dissimilarity main effects is significant, nor are any of the interactions terms, for either obesity or poor health outcomes. Thus, Black-White and Hispanic-White dissimilarity do not exacerbate the effects of negative stressful events on either obesity or poor health.

Accounting for dissimilarity levels also do not change the racial disparities seen in health.

Black and Hispanics are still at higher odds of obesity, and Hispanics are still more likely to be in poor health relative to Whites.

Table 15 presents the results from interaction models assessing the joint effects of residential segregation as measures by the isolation index and stressful events. For both outcomes, Model 1 includes the interactions with Black isolation, Model 2 the interactions with Hispanic isolation. None of the main effects or interaction effects are significant for black isolation on obesity or poor health. In Model 2, the effect of three or more stressful events is significant, but since the main effects and interaction effects are not significant, this significant term is not meaningful. As with the models including dissimilarity, including these interaction terms do not change the racial differences in the likelihood of obesity and poor health.

Discussion

The goal of this study is to assess the role of family stressful events on the health and well-being of African American, Hispanic, and White children, and the extent to which residential segregation exacerbates the detrimental effects of stress on health. The main findings of this study highlight the detrimental role of family stressful events for the physical health of children. As expected in Hypothesis 1, as children experience two and more family stressful events, the likelihood of obesity increases. The effect of stress on poor health is felt only at high levels of stress, 3 or more events.

The findings presented here do support hypothesis 2 of this chapter that dissimilarity indices do not change the role of stress on child physical health. There is no evidence to suggest that residential dissimilarity levels influence child health by moderating the effect of family stress. The effects of family stress do not vary across segregation levels as measured by the dissimilarity index, for obesity and poor health.

Although highly dissimilar metropolitan areas are unequally distributed, the spatial patterning and interracial exposure of residents can vary greatly across dissimilar areas, and this dissimilarity may not capture experiences of segregation in a way that is meaningful to stress exposure, coping, and child health (Kramer and Hogue 2009;

Avecedo-Garcia et al. 2003).

Additionally, Hypothesis 3 is not supported, as isolation levels also do not moderate the relationship between family stress and health. Although in theory, measures of isolation may better capture overall community disinvestment than

91 dissimilarity, high levels of Black and Hispanic isolation are not acting as an exacerbating force on the role of family stress for child health outcomes. Furthermore, accounting segregation and stress levels does not account for racial differences in health.

Blacks and Hispanics are still more likely to be obese, and Hispanics are more likely to be in poor health.

Two limitations of this study are important to consider. First, it would ideal to have more detailed information about the duration and timing of family stressful events to have a better understanding of their relative importance to child health. Second, smaller sample sizes for African Americans in particular reduce the ability to estimate he effects of stress and health separately by race. The effects of segregation and stress are likely to differ, and future work should investigate how these relationships vary across race with larger minority samples.

Despite these limitations, this study contributes to the understanding of how individual level stressors such as negative life events influence health. Although the influence of family stressful events does not vary across residential segregation levels, family stress in all contexts negatively impacts health, resulting in a higher odds of obesity and poor health, especially at high levels of stress. Future research should consider the ways that stress exposure may further explain racial disparities in child health by exploring more detailed measure of stress as well as how other aspects of residential contexts may influence stress exposure.

White Black Hispanic Obese, % 16% 23% 26% Parent-Rated Poor Health, % 2% 4% 5%

No Stressful Events, % 48% 30% 44% 1 Stressful Event, % 31% 29% 31% 2 Stressful Events, % 14% 22% 16% 3 or More Stressful Events, % 8% 19% 10% Parental Involvement Scale, mean 43.78 43.33 41.55

Black -White Dissimilarity, mean 60% 63% 57% Black Isolation, mean 46% 58% 45% Hispanic-White Dissimilarity, mean 44% 45% 50% Hispanic Isolation, mean 26% 29% 53%

Female, % 48% 45% 48% Age, mean in years 14.31 14.25 14.24 Single Parent Household, % 16% 45% 24% Household Size, mean 4.36 4.53 4.80 No Health Insurance, % 3% 3% 12% Government Health Insurance, % 7% 27% 24% Private/Military Health Insurance, % 90% 71% 64%

Socioeconomic Status Lowest Quintile, % 6% 29% 41% Quintile 2, % 15% 25% 23% Quintile 3, % 20% 20% 16% Quintile 4, % 25% 20% 12% Highest Quintile, % 34% 6% 9%

Percent of Sample, Weighted 61% 17% 22% Number of Respondents, Unweighted 3,470 540 1,140 Table 12: Weighted Sample Characteristics: Health, Segregation, and Stress by Racial Identity

Model 1 Model 2 Obese Poor Health 1 Stressful Event 1.197 1.04 (0.10) (0.27)

2 Stressful Events 1.323* 1.71 (0.15) (0.49)

3 or More Stressful Events 1.457* 2.467** (0.20) (0.73)

Parental Involvement Scale 0.982* 0.947** (0.01) (0.02)

African American 1.358* 1.53 (0.18) (0.49)

Hispanic 1.451** 2.162** (0.14) (0.55)

Female 0.699** 1.03 (0.05) (0.20)

Age, years 0.84 1.31 (0.09) (0.37)

Single Parent 0.94 1.14 (0.10) (0.31)

Household Size 0.96 1.03 (0.03) (0.08)

No Health Insurance 1.30 0.90 (0.23) (0.35)

Government Health Insurance 1.07 1.53 (0.13) (0.43)

Socioeconomic Status Quintiles 0.809** 0.718** (0.03) (0.08)

N 5160 5160 χ2 201.04 135.94 Psuedo R2 0.04 0.12 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 13: Odds Ratios from Logistic Regression of Family Stressful Events on Health

Model 1 Model 2 Model 1 Model 2 Obese Obese Poor Health Poor Health 1 Stressful Event 1.26 1.08 0.18 0.69 (0.46) (0.35) (0.21) (0.59)

2 Stressful Events 0.95 1.28 1.13 0.42 (0.46) (0.53) (1.16) (0.40)

3 or More Stressful Events 1.78 1.14 0.31 0.52 (0.99) (0.57) (0.38) (0.57)

Black-White Dissimilarity 1.00 0.98

(0.00) (0.01)

1 Event * Black Dissimilarity 1.00 1.03

(0.01) (0.02)

2 Events * Black Dissimilarity 1.01 1.01

(0.01) (0.02)

3+ Events * Black Dissimilarity 1.00 1.04

(0.01) (0.02)

Hispanic-White Dissimilarity 1.00 0.99

(0.01) (0.01)

1 Event * Hispanic Dissimilarity 1.00 1.01

(0.01) (0.02)

2 Events * Hispanic Dissimilarity 1.00 1.03

(0.01) (0.02)

3+ Events * Hispanic Dissimilarity 1.01 1.03

(0.01) (0.02)

Parental Involvement Scale 0.982* 0.982* 0.947** 0.948** (0.01) (0.01) (0.02) (0.02)

Black 1.352* 1.371* 1.55 1.52 (0.18) (0.18) (0.50) (0.49)

Hispanic 1.456** 1.494** 2.166* 2.094* (0.14) (0.15) (0.54) (0.57)

Female 0.698** 0.699** 1.03 1.03 (0.05) (0.05) (0.20) (0.20)

Age, years 0.84 0.83 1.31 1.34 (0.09) (0.09) (0.36) (0.38)

Single Parent 0.94 0.94 1.16 1.14 (0.10) (0.10) (0.31) (0.31)

Household Size 0.96 0.96 1.03 1.03 (0.03) (0.03) (0.08) (0.08)

No Health Insurance 1.30 1.29 0.91 0.90 (0.23) (0.23) (0.36) (0.35)

Government Health Insurance 1.07 1.06 1.53 1.55 (0.13) (0.13) (0.43) (0.44)

Socioeconomic Status Quintiles 0.809** 0.811** 0.724** 0.718** (0.03) (0.03) (0.08) (0.08)

N 5160 5160 5160 5160 χ2 202.25 202.87 153.4 144.51 Psuedo R2 0.05 0.05 0.12 0.12 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 14: Odds Ratios from Logistic Regression of Family Stressful Events and Dissimilarity Index Interactions on Health

Model 1 Model 2 Model 1 Model 2 Obese Obese Poor Health Poor Health 1 Stressful Event 1.27 1.340 0.39 1.18 (0.25) (0.20) (0.21) (0.61)

2 Stressful Events 1.11 1.543 1.16 1.15 (0.31) (0.31) (0.69) (0.66)

3 or More Stressful Events 1.942 1.718* 0.73 1.68 (0.62) (0.39) (0.53) (0.98)

Black Isolation 1.00 0.987

(0.00) (0.01)

1 Event * Black Isolation 1.00 1.022

(0.00) (0.01)

2 Events * Black Isolation 1.00 1.01

(0.01) (0.01)

3+ Events * Black Isolation 0.99 1.026

(0.01) (0.01)

Hispanic Isolation 1.00 1.01

(0.00) (0.01)

1 Event * Hispanic Isolation 1.00 1.00

(0.00) (0.01)

2 Events * Hispanic Isolation 1.00 1.01

(0.01) (0.01)

3+ Events * Hispanic Isolation 1.00 1.01

(0.01) (0.01)

Parental Involvement Scale 0.982* 0.981* 0.947** 0.950** (0.01) (0.01) (0.02) (0.02)

Black 1.339* 1.377* 1.53 1.48 (0.17) (0.18) (0.50) (0.48)

Hispanic 1.454** 1.653** 2.197** 1.65 (0.14) (0.18) (0.56) (0.50)

Female 0.699** 0.701** 1.03 1.03 (0.05) (0.05) (0.20) (0.20)

Age, years 0.84 0.82 1.31 1.40 (0.09) (0.09) (0.36) (0.40)

Single Parent 0.94 0.94 1.16 1.15 (0.10) (0.10) (0.32) (0.32)

Household Size 0.96 0.96 1.03 1.02 (0.03) (0.03) (0.08) (0.08)

No Health Insurance 1.30 1.32 0.93 0.87 (0.23) (0.23) (0.36) (0.34)

Government Health Insurance 1.06 1.08 1.54 1.48 (0.13) (0.14) (0.43) (0.42)

Socioeconomic Status Quintiles 0.808** 0.815** 0.726** 0.710** (0.03) (0.03) (0.08) (0.08)

N 5160 5160 5160 5160 χ2 203.61 207.06 152.89 158.32 Psuedo R2 0.05 0.05 0.12 0.12 *p<.05, **p<.01, one-tailed. Robust standard errors are in parentheses. Table 15: Odds Ratios from Logistic Regression of Family Stressful Events and Racial Isolation Interactions on Health

Chapter 5: Conclusion

Racial and ethnic minority children, especially Black and Hispanic children, disproportionately experience family and neighborhood poverty and parental unemployment (DeNavas-Walt, Proctor, and Smith 2010; Timberlake 2007). Black and

Hispanic children also endure suboptimal health outcomes relative to White children that persist across time as well as across the life course (McDaniel, Paxson, and Waldfogel

2006; Kimbro et al. 2007). Although immigrants tend to fare better than expected with regard to health relative to Whites given their relatively low socioeconomic status and education levels (Hummer et al. 2007; Markides and Eschbach 2005), immigrant

Hispanic children still face elevated odds of developing obesity and poor physical health in early life (Crosnoe 2006). The portion of all children under the age of 18 that are

Hispanic and Black is increasing over time; new Census 2010 estimates report that 23% of children under 18 are Hispanic, and 14% are Black (Mather et al. 2011). As these larger groups of minority children age into adulthood, their relative health disadvantage may result in larger burdens on public health care systems. Thus, understanding how health disparities emerge in childhood now is key to reducing current disparities as well as preventing disparities in future generations.

Researchers have increasingly turned to residential contexts as a source of population health inequality (Diez Roux and Mair 2010). Residential segregation is a key defining characteristic in the American urban landscape and a powerful force for explaining racial inequality overall (Massey and Denton 1993). Furthermore, segregation has been hypothesized to be a key explanatory mechanism for understand racial disparities in health through its impact on individual and neighborhood level socioeconomic status, discrimination and stress, and neighborhood quality (Acevedo-

Garcia et al. 2003; Williams and Sternthal 2010). Previous studies have explored the links between segregation and health in adults and in birth outcomes, yet few studies address the role of residential segregation on health outcomes during childhood, and especially among Hispanics.

While it is theoretically important to consider the ways in which these factors are pathways through which residential segregation affects health, segregation is also likely to exacerbate the effects of individual stressors to health. Racially isolated neighborhoods lead to community disinvestment and suboptimal socioeconomic and social resources available to residents, key coping mechanisms that buffer the effect of stressors. To this end, then, segregation may in fact act through a stress proliferation model that increases the detrimental effects of individual level characteristics and stressors.

98 assessed how these social contexts interact with and modify individual level characteristics such as race and family stress to influence child health. In doing so, I evaluate three distinct but related research questions.

(asthma). Finally, residential segregation measures do not exacerbate the negative effect of stress events on child health.

Conclusions

Overall, the results suggest that relationship between residential segregation and child health is highly dependent upon segregation measure and health outcome, and not consistent across racial groups. Although segregation may be a determining factor in adult health, early life exposure to residential segregation is not a consistent determinant of health disparities. The health of Hispanic children is more sensitive to these residential patterns than Whites and Blacks, even if the directions of the relationships vary by segregation and health measure, and may signify important variations within Hispanics as a group. By assessing the role of segregation among children and especially Hispanic children, this research broadens our understanding of the social processes that create health disparities by expanding models of structural discrimination health effects to

Hispanics as well as children. Research into the mechanisms that translate residential segregation into health outcomes needs to be sensitive to racial/ethnic differences in the effect of segregation measures, and pay particular attention to the reasons how and why dissimilarity and isolation measures differentially effect health.

This dissertation also contributes to our understanding of the effects of racial/ethnic tensions in the neighborhood. Obesity levels among Hispanic children are most sensitive to racial/ethnic neighborhood tensions, especially in the absence of parental involvement in schools. More research is needed to unpack the role of racial and ethnic tensions on health; using more detailed measures of the type and level of conflict may present a more nuanced view of the relationship between racial/ethnic tensions and health.

Surprisingly, residential segregation and neighborhood racial/ethnic tensions are not related to obesity among African Americans. These findings are contrary to findings in the adult health literature that residential segregation is related to worse health outcomes for Black adults as well as infants (Chang 2006; Kramer and Hogue 2009;

Subramanian et al. 2005). This divergent finding warrants more inquiries. Black-White segregation may not affect Black children adversely, either because exposure to segregation is cumulative across the life course and the effects do not contribute to racial inequality in health until later or because the processes of health inequality for young

African Americans is not tied to residential segregation on the metropolitan level.

Additionally, although perceived discrimination is important to understanding health disparities in African Americans, the connection is less conclusive for physical health.

100

These findings may be reflecting the weaker link between discrimination and physical health. Finally, neighborhood racial tensions may not be more stressing in addition to interpersonal discrimination experiences, which are already elevated even among young

Black children entering adolescence.

The effects of family stress on obesity and poor health do not vary across segregation levels as measured by the dissimilarity index or the isolation index, suggesting that residential segregation patterns may not be a contributing factor to differential exposure and vulnerability to stressful events among children. Although highly uneven metropolitan areas are clearly unequally distributed, the spatial patterning and interracial exposure of residents can vary greatly, and may not capture experiences of segregation in a way that is meaningful to health (Kramer and Hogue 2009; Avecedo-

Garcia et al. 2003).

Although the ECLS-K provides a rich assessment of family, school, and neighborhood contexts, there are a few limitations of this study to be noted. First, it would ideal to have more detailed information about the duration and timing of residential moves, racial and ethnic tensions in neighborhoods, as well as family stressful events. Although repeated cross-sectional measures particularly of residential mobility and family events provide useful information, it is likely that timing and duration of exposure to these circumstances are key. Second, measuring residential segregation at the metropolitan level reports important information about how residential patterns vary across neighborhoods, and so is a relative measure of residential inequality. However, measuring segregation at the MSA level clearly misses important variation across

101 individuals that results in different exposure to residential patterns. Future research on the health effects of residential segregation may benefit from new measures designed to decompose regional variation into smaller spatial units that may better assess individual level experiences situated in the broader context of residential segregation patterns (Lee et al. 2008).

Despite these limitations, this study contributes to our understanding of the ways in which individual level stressors such as negative life events interact with community and metropolitan level contexts, to jointly affect health. Overall, the association between residential segregation and child health is strongest among Hispanics and for the most part the relationship is detrimental to health. Despite hypotheses that co-ethnic neighborhoods are beneficial to Hispanics and especially immigrants, these findings concur with work by Frank and colleagues (Frank et al. 2007; Frank and Bjornstrom

2011) that high levels of exposure to co-ethnics may in fact be detrimental to health in childhood.

Racial disparities in health are significant despite large gains in public health knowledge and health services in the last century (Phelan and Link 2005). African

Americans, and increasingly Hispanics, are spatially segregated within U.S. cities, and the negative consequences of segregation for minorities are apparent across varying measures of well-being. Although a wide range of explanatory mechanisms are responsible for creating racial disparities in health, the way in which residential segregation patterns exposure to stressors and responses to those stressors is one piece of that larger puzzle that warrants further examination.

102

References

Acevedo-Garcia, Dolores, Kimberly A. Lochner, Theresa L. Osypuk, and S. V. Subramanian. 2003. “Future Directions in Residential Segregation and Health Research: A Multilevel Approach.” American Journal of Public Health 93:215- 221.

Akresh, Ilana Redstone, and Reanne Frank. 2008. “Health Selection Among New Immigrants.” American Journal of Public Health 98:2058-2064.

Aneshensel, Carol S. 2009. “Toward Explaining Mental Health Disparities.” Journal of Health and Social Behavior 50:377-394.

Aneshensel, Carol. 2010. “Neighborhood as a Social Context of the Stress Process.” in Advances in the Conceptualization of the Stress Process, edited by William R. Avison, Carol S. Aneshensel, Scott Schieman, and Blair Wheaton. New York, NY: Springer New York.

Avison, William R. 2010. “Incorporating Children‟s Lives into a Life Course Perspective on Stress and Mental Health.” Journal of Health and Social Behavior 51:361 - 375.

Bell, Janice F., Frederick J. Zimmerman, Gunnar R. Almgren, Jonathan D. Mayer, and Colleen E. Huebner. 2006. “Birth outcomes among urban African-American women: A multilevel analysis of the role of racial residential segregation.” Social Science & Medicine 63:3030-3045.

Black, Jennifer L, and James Macinko. 2008. “Neighborhoods and obesity.” Nutrition Reviews 66:2-20.

Boardman, Jason D, Brian K Finch, and Robert A Hummer. 2001. “Race/ethnic differences in respiratory problems among a nationally-representative cohort of young children in the United States.” Population Research and Policy Review 20:187-206.

Boardman, Jason D. 2004. “Stress and physical health: the role of neighborhoods as mediating and moderating mechanisms.” Social Science & Medicine 58:2473- 2483. 103

Boardman, Jason D., Jarron M. Saint Onge, Richard G. Rogers, and Justin T. Denney. 2005. “Race Differentials in Obesity: The Impact of Place.” Journal of Health and Social Behavior 46:229-243.

Boardman, Jason D., and Kari B. Alexander. 2011. “Stress trajectories, health behaviors, and the mental health of black and white young adults.” Social Science & Medicine 72:1659-1666.

Borrell, Luisa N., Catarina I. Kiefe, David R. Williams, Ana V. Diez-Roux, and Penny Gordon-Larsen. 2006. “Self-reported health, perceived racial discrimination, and skin color in African Americans in the CARDIA study.” Social Science & Medicine 63:1415-1427.

Borrell, Luisa N., Ana V. Diez Roux, David R. Jacobs Jr., Steven Shea, Sharon A. Jackson, Sandi Shrager, and Roger S. Blumenthal. 2010. “Perceived racial/ethnic discrimination, smoking and alcohol consumption in the Multi-Ethnic Study of Atherosclerosis (MESA).” Preventive Medicine 51:307-312.

Burdette, Amy M., and Terrence D. Hill. 2008. “An examination of processes linking perceived neighborhood disorder and obesity.” Social Science & Medicine 67:38- 46.

Cagney, Kathleen A., Christopher R. Browning, and Danielle M. Wallace. 2007. “The Latino Paradox in Neighborhood Context: The Case of Asthma and Other Respiratory Conditions.” American Journal of Public Health 97:919-925.

Card, David, and Jesse Rothstein. 2007. “Racial segregation and the black-white test score gap.” Journal of Public Economics 91:2158-2184.

Case, Anne, Angela Fertig, and Christina Paxson. 2005. “The lasting impact of childhood health and circumstance.” Journal of Health Economics 24:365-389.

Caughy, Margaret O‟Brien, Patricia J. O‟Campo, and Carles Muntaner. 2004. “Experiences of Racism Among African American Parents and the Mental Health of Their Preschool-Aged Children.” American Journal of Public Health 94:2118- 2124.

Chang, Virginia W. 2006. “Racial residential segregation and weight status among US adults.” Social Science & Medicine 63:1289-1303.

Chang, Virginia W., Amy E. Hillier, and Neil K. Mehta. 2009. “Neighborhood Racial Isolation, Disorder and Obesity.” Social forces 87:2063-2092.

104

Charles, Camille Zubrinsky. 2003. “The Dynamics of Racial Residential Segregation.” Annual Review of Sociology 29:167-207.

Charles, Camille Z., Gniesha Dinwiddie, and Douglas S. Massey. 2004. “The Continuing Consequences of Segregation: Family Stress and College Academic Performance.” Social Science Quarterly 85:1353-1373.

Chau, Michelle, Kalyani Thampi, and Vanessa R. Wight. 2010. Basic Facts About Low- income Children, 2009: Children Under Age 18. New York: National Center for Children in Poverty, Columbia University, Mailman School of Public Health.

Cheng, Lucie, and Yen Espiritu. 1989. “Korean Businesses in Black and Hispanic Neighborhoods: A Study of Intergroup Relations.” Sociological Perspectives 32:521-534.

Clampet‐Lundquist, Susan, and Douglas S. Massey. 2008. “Neighborhood Effects on Economic Self‐Sufficiency: A Reconsideration of the Moving to Opportunity Experiment.” The American Journal of Sociology 114:107-143.

Coker, Tumaini R., Marc N. Elliott, David E. Kanouse, Jo Anne Grunbaum, David C. Schwebel, M. Janice Gilliland, Susan R. Tortolero, Melissa F. Peskin, and Mark A. Schuster. 2009. “Perceived Racial/Ethnic Discrimination Among Fifth-Grade Students and Its Association With Mental Health.” American Journal of Public Health 99:878-884.

Colen, Cynthia G., Arline T. Geronimus, John Bound, and Sherman A. James. 2006. “Maternal Upward Socioeconomic Mobility and Black-White Disparities in Infant Birthweight.” American Journal of Public Health 96:2032-2039.

Collins, James W., Jennifer Wambach, Richard J. David, and Kristin M. Rankin. 2008. “Women‟s Lifelong Exposure to Neighborhood Poverty and Low Birth Weight: A Population-Based Study.” Maternal and Child Health Journal 13:326-333.

Conger, Rand D, Katherine J Conger, and Monica J Martin. 2010. “Socioeconomic Status, Family Processes,and Individual Development.” Journal of Marriage and Family 72:685-704.

Cooper, Richard S. 2001. “Social inequality, ethnicity and cardiovascular disease.” International Journal of Epidemiology 30:S48.

Cooper, Richard S, Joan F Kennelly, Ramon Durazo-Arvizu, Hyun-Joo Oh, George Kaplan, and John Lynch. 2001. “Relationship between Premature Mortality and Socioeconomic Factors in Black and White Populations of US Metropolitan Areas.” Public Health Reports 116:464-473. 105

Cradock, Angie L., Ichiro Kawachi, Graham A. Colditz, Steven L. Gortmaker, and Stephen L. Buka. 2009. “Neighborhood social cohesion and youth participation in physical activity in Chicago.” Social Science & Medicine 68:427-435.

Crosnoe, Robert. 2006. “Health and the Education of Children from Racial/Ethnic Minority and Immigrant Families.” Journal of Health and Social Behavior 47:77 - 93.

Crowder, Kyle, and Scott J. South. 2011. “Spatial and temporal dimensions of neighborhood effects on high school graduation.” Social Science Research 40:87- 106.

Cutler, David M., and Edward L. Glaeser. 1997. “Are Ghettos Good or Bad?” The Quarterly Journal of Economics 112:827-872.

Dailey, Amy B., Stanislav V. Kasl, Theodore R. Holford, Tene T. Lewis, and Beth A. Jones. 2010. “Neighborhood- and individual-level socioeconomic variation in perceptions of racial discrimination.” Ethnicity & Health 15:145-163.

DeNavas-Walt, Carmen, Bernadette D. Proctor, and Jessica C. Smith. 2010. Income, Poverty, and Health Insurance Coverage in the United States: 2009. U.S. Census Bureau, Current Population Reports, P60-238. Washington, D.C.: U.S. Government Printing Office.

Diez Roux, Ana V., and Christina Mair. 2010. “Neighborhoods and health.” Annals of the New York Academy of Sciences 1186:125-145.

Echeverría, Sandra, Ana V. Diez-Roux, Steven Shea, Luisa N. Borrell, and Sharon Jackson. 2008. “Associations of neighborhood problems and neighborhood social cohesion with mental health and health behaviors: The Multi-Ethnic Study of Atherosclerosis.” Health & Place 14:853-865.

Eschbach, Karl, Glenn V. Ostir, Kushang V. Patel, Kyriakos S. Markides, and James S. Goodwin. 2004. “Neighborhood Context and Mortality Among Older Mexican Americans: Is There a Barrio Advantage?” American Journal of Public Health 94:1807-1812.

Farmer, Melissa M, and Kenneth F Ferraro. 2005. “Are racial disparities in health conditional on socioeconomic status?” Social Science & Medicine 60:191-204.

Feagin, Joe R., and Melvin P. Sikes. 1994. Living with racism: the black middle-class experience. Boston, MA: Beacon Press.

106

Feldmeyer, Ben. 2010. “The Effects of Racial/Ethnic Segregation on Latino and Black Homicide.” Sociological Quarterly 51:600-623.

Finch, Brian Karl, Robert A. Hummer, Bohdan Kol, and William A. Vega. 2001. “The Role of Discrimination and Acculturative Stress in the Physical Health of Mexican-Origin Adults.” Hispanic Journal of Behavioral Sciences 23:399 -429.

Fischer, Mary J. 2010. “Immigrant educational outcomes in new destinations: An exploration of high school attrition.” Social Science Research 39:627-641.

Fish, Jason S., Susan Ettner, Alfonso Ang, and Arleen F. Brown. 2010. “Association of Perceived Neighborhood Safety on Body Mass Index.” American Journal of Public Health 100:2296-2303.

Flippen, Chenoa Anne. 2010. “The Spatial Dynamics of Stratification: Metropolitan Context, Population Redistribution, and Black and Hispanic Homeownership.” Demography 47:845-868.

Flores, Glenn, Lynn Olson, and Sandra C. Tomany-Korman. 2005. “Racial and Ethnic Disparities in Early Childhood Health and Health Care.” Pediatrics 115:e183 - e193.

Frank, Reanne, and Eileen Bjornstrom. 2011. “A tale of two cities: Residential context and risky behavior among adolescents in Los Angeles and Chicago.” Health & Place 17:67-77.

Frank, Reanne, Magdalena Cerda, and Maria Rendon. 2007. “Barrios and Burbs: Residential Context and Health-Risk Behaviors among Angeleno Adolescents.” Journal of Health and Social Behavior 48:283-300.

Franzini, Luisa, Margaret O‟Brien Caughy, Saundra Murray Nettles, and Patricia O‟Campo. 2008. “Perceptions of disorder: Contributions of neighborhood characteristics to subjective perceptions of disorder.” Journal of Environmental Psychology 28:83-93.

Franzini, Luisa, Marc N. Elliott, Paula Cuccaro, Mark Schuster, M. Janice Gilliland, Jo Anne Grunbaum, Frank Franklin, and Susan R. Tortolero. 2009. “Influences of Physical and Social Neighborhood Environments on Children‟s Physical Activity and Obesity.” American Journal of Public Health 99:271-278.

Gary, Tiffany L., Sarah A. Stark, and Thomas A. LaVeist. 2007. “Neighborhood characteristics and mental health among African Americans and whites living in a racially integrated urban community.” Health & Place 13:569-575.

107

Gee, Gilbert C., and Chandra L. Ford. 2011. “Structural Racism and Health Inequities.” Du Bois Review: Social Science Research on Race 8:115-132.

Geronimus, Arline T. 1996. “Black/white differences in the relationship of maternal age to birthweight: A population-based test of the weathering hypothesis.” Social Science & Medicine 42:589-597.

Geronimus, Arline T. 2001. “Understanding and eliminating racial inequalities in women‟s health in the United States: the role of the weathering conceptual framework.” Journal of the American Medical Women’s Association 56:133-136.

Geronimus, Arline T., Margaret Hicken, Danya Keene, and John Bound. 2006. “Weathering and Age Patterns of Allostatic Load Scores Among Blacks and Whites in the United States.” American Journal of Public Health 96:826-833.

Geronimus, Arline T, John Bound, Danya Keene, and Margaret Hicken. 2007. “Black- white differences in age trajectories of hypertension prevalence among adult women and men, 1999-2002.” Ethnicity & Disease 17:40-48.

Geronimus, Arline T., Margaret Hicken, Jay Pearson, Sarah Seashols, Kelly Brown, and Tracey Cruz. 2010. “Do US Black Women Experience Stress-Related Accelerated Biological Aging?” Human Nature 21:19-38.

Gordon-Larsen, Penny, Melissa C. Nelson, Phil Page, and Barry M. Popkin. 2006. “Inequality in the Built Environment Underlies Key Health Disparities in Physical Activity and Obesity.” Pediatrics 117:417 -424.

Grady, Sue C, and Iván J Ramírez. 2008. “Mediating medical risk factors in the residential segregation and low birthweight relationship by race in New York City.” Health & Place 14:661-677.

Ha, Shang E. 2010. “The Consequences of Multiracial Contexts on Public Attitudes toward Immigration.” Political Research Quarterly 63:29 -42.

Haas, Steven. 2008. “Trajectories of functional health: The „long arm‟ of childhood health and socioeconomic factors.” Social Science & Medicine 66:849-861.

Hayward, Mark D., Toni P. Miles, Eileen M. Crimmins, and Yu Yang. 2000. “The Significance of Socioeconomic Status in Explaining the Racial Gap in Chronic Health Conditions.” American Sociological Review 65:910-930.

Hayward, Mark D., and Bridget K. Gorman. 2004. “The Long Arm of Childhood: The Influence of Early-Life Social Conditions on Men‟s Mortality.” Demography 41:87-107. 108

Heron, Melonie, Donna L Hoyert, Sherry L Murphy, Jiaquan Xu, Kenneth D Kochanek, and Betzaida Tejada-Vera. 2009. “Deaths: final data for 2006.” National Vital Statistics Reports 57:1-134.

Hertzman, Clyde, and Chris Power. 2003. “Health and human development: understandings from life-course research.” Developmental Neuropsychology 24:719-744.

Hill, Terrence D., Catherine E. Ross, and Ronald J. Angel. 2005. “Neighborhood Disorder, Psychophysiological Distress, and Health.” Journal of Health and Social Behavior 46:170 -186.

Hill, Terrence D., and Ronald J. Angel. 2005. “Neighborhood disorder, psychological distress, and heavy drinking.” Social Science & Medicine 61:965-975. von Hippel, Paul T., Brian Powell, Douglas B. Downey, and Nicholas J. Rowland. 2007. “The Effect of School on Overweight in Childhood: Gain in Body Mass Index During the School Year and During Summer Vacation.” American Journal of Public Health 97:696-702.

Hirschman, Charles. 2001. “The Educational Enrollment of Immigrant Youth: A Test of the Segmented-Assimilation Hypothesis.” Demography 38:317-336.

Hummer, Robert A, Daniel A Powers, Starling G Pullum, Ginger L Gossman, and W Parker Frisbie. 2007. “Paradox found (again): infant mortality among the Mexican-origin population in the United States.” Demography 44:441-457.

Hunt, Matthew O., Lauren A. Wise, Marie-Claude Jipguep, Yvette C. Cozier, and Lynn Rosenberg. 2007. “Neighborhood Racial Composition and Perceptions of Racial Discrimination: Evidence From the Black Women‟s Health Study.” Social Psychology Quarterly 70:272-289.

Hunte, Haslyn E. R., and David R. Williams. 2009. “The Association Between Perceived Discrimination and Obesity in a Population-Based Multiracial and Multiethnic Adult Sample.” American Journal of Public Health 99:1285-1292.

Iceland, John, Daniel Weinberg, and Erika Steinmetz. 2002. Racial and Ethnic Residential Segregation in the United States: 1980-2000. U.S. Government Printing Office: Washington, DC.

Iceland, John, and Rima Wilkes. 2006. “Does Socioeconomic Status Matter? Race, Class, and Residential Segregation.” Social Problems 53:248-273.

109

Iceland, John, and Kyle Anne Nelson. 2008. “Hispanic Segregation in Metropolitan America: Exploring the Multiple Forms of Spatial Assimilation.” American Sociological Review 73:741-765.

Iceland, John, and Melissa Scopiletti. 2008. “Immigrant Residential Segregation in U.S. Metropolitan Areas, 1990–2000.” 45:79-94.

Iceland, John, Kimberly A. Goyette, Kyle Anne Nelson, and Chaowen Chan. 2010. “Racial and ethnic residential segregation and household structure: A research note.” Social Science Research 39:39-47.

Inagami, Sanae, Luisa N. Borrell, Mitchell D. Wong, Jing Fang, Martin F. Shapiro, and Steven M. Asch. 2006. “Residential Segregation and Latino, Black and White Mortality in New York City.” Journal of Urban Health 83:406-420.

Jackson, James S., Tony Brown, David R. Williams, Myriam Torres, Sherrill L. Sellers, and Kendrick Brown. 1996. “Racism and the physical and mental health status of African Americans: a thirteen year national panel study.” Ethnicity & Disease 6:132-147.

Jackson, Margot. 2009. “Understanding Links Between Adolescent Health and Educational Attainment.” Demography 46:671-694.

Jackson, Margot I., and Robert D. Mare. 2007. “Cross-sectional and longitudinal measurements of neighborhood experience and their effects on children.” Social Science Research 36:590-610.

Jackson, Sharon A., Roger T. Anderson, Norman J. Johnson, and Paul D. Sorlie. 2000. “The relation of residential segregation to all-cause mortality: a study in black and white.” American Journal of Public Health 90:615-617.

Jordan, Kareem L., and Shaun L. Gabbidon. 2010. “Race/Ethnicity and Perceptions of Safety Among a National Sample of Americans.” Criminal Justice Review 35:281 -294.

Jylhä, Marja. 2009. “What is self-rated health and why does it predict mortality? Towards a unified conceptual model.” Social Science & Medicine 69:307-316.

Kawachi, Ichiro, and Bruce P Kennedy. 1997. “Socioeconomic determinants of health: Health and social cohesion: why care about income inequality?” BMJ 314:1037.

Kawachi, Ichiro, and Lisa F. Berkman. 2000. “Social Cohesion, Social Capital, and Health.” in Social Epidemiology. New York: Oxford University Press.

110

Kessler, Ronald C., Kristin D. Mickelson, and David R. Williams. 1999. “The Prevalence, Distribution, and Mental Health Correlates of Perceived Discrimination in the United States.” Journal of Health and Social Behavior 40:208-230.

Kimbro, Rachel Tolbert, Jeanne Brooks-Gunn, and Sara McLanahan. 2007. “Racial and Ethnic Differentials in Overweight and Obesity Among 3-Year-Old Children.” American Journal of Public Health 97:298-305.

Kimbro, Rachel Tolbert, Jeanne Brooks-Gunn, and Sara McLanahan. 2011. “Young children in urban areas: Links among neighborhood characteristics, weight status, outdoor play, and television watching.” Social Science & Medicine 72:668-676.

Kling, Jeffrey R., Jeffrey B. Liebman, Lawrence F. Katz, and Lisa Sanbonmatsu. 2004. “Moving To Opportunity And Tranquility: Neighborhood Effects On Adult Economic Self-Sufficiency And Health From A Randomized Housing Voucher Experiment.”

Kramer, Michael R., and Carol R. Hogue. 2009. “Is Segregation Bad for Your Health?” Epidemiologic Reviews 31:178 -194.

Krebs, Nancy F., John H. Himes, Dawn Jacobson, Theresa A. Nicklas, Patricia Guilday, and Dennis Styne. 2007. “Assessment of Child and Adolescent Overweight and Obesity.” Pediatrics 120:S193-228.

Krieger, Nancy, and Stephen Sidney. 1996. “Racial discrimination and blood pressure: the CARDIA Study of young black and white adults.” American Journal of Public Health 86:1370-1378.

Krivo, Lauren J., and Robert L. Kaufman. 2004. “Housing and Wealth Inequality: Racial- Ethnic Differences in Home Equity in the United States.” Demography 41:585- 605.

Krivo, Lauren J., Ruth D. Peterson, and Danielle C. Kuhl. 2009. “Segregation, Racial Structure, and Neighborhood Violent Crime.” American Journal of Sociology 114:1765-1802.

Krysan, Maria, and Reynolds Farley. 2002. “The Residential Preferences of Blacks: Do They Explain Persistent Segregation?” Social Forces 80:937-980.

Kuczmarski, Robert J et al. 2002. “2000 CDC Growth Charts for the United States: methods and development.” Vital and Health Statistics. Series 11, Data from the National Health Survey 1-190.

111

Kuh, Diana, and Yoav Ben-Shlomo, eds. 1997. A Life Course Approach to Chronic Disease Epidemiology. New York: Oxford University Press.

Landrine, Hope, and Elizabeth A. Klonoff. 2000. “Racial Segregation and Cigarette Smoking Among Blacks: Findings at the Individual Level.” Journal of Health Psychology 5:211 -219.

LaVeist, Thomas A. 2003. “Racial Segregation and Longevity among African Americans: An Individual-Level Analysis.” Health Services Research 38:1719- 1734.

Lee, Barrett A., Sean F. Reardon, Glenn Firebaugh, Chad R. Farrell, Stephen A. Matthews, and David O‟Sullivan. 2008. “Beyond the Census Tract: Patterns and Determinants of Racial Segregation at Multiple Geographic Scales.” American Sociological Review 73:766-791.

Lee, Jennifer. 2002. “From Civil Relations to Racial Conflict: Merchant-Customer Interactions in Urban America.” American Sociological Review 67:77-98.

Lee, Min-Ah. 2009. “Neighborhood residential segregation and mental health: A multilevel analysis on Hispanic Americans in Chicago.” Social Science & Medicine 68:1975-1984.

Lee, Min-Ah, and Kenneth F. Ferraro. 2007. “Neighborhood Residential Segregation and Physical Health among Hispanic Americans: Good, Bad, or Benign?” Journal of Health and Social Behavior 48:131-148.

Lee, Rebecca E., and Catherine Cubbin. 2002. “Neighborhood Context and Youth Cardiovascular Health Behaviors.” American Journal of Public health 92:428- 436.

Leventhal, Tama, and Jeanne Brooks-Gunn. 2003. “Moving to Opportunity: an Experimental Study of Neighborhood Effects on Mental Health.” American Journal of Public Health 93:1576-1582.

Lichter, Daniel T., Domenico Parisi, Michael C. Taquino, and Steven Michael Grice. 2010. “Residential segregation in new Hispanic destinations: Cities, suburbs, and rural communities compared.” Social Science Research 39:215-230.

Lichter, Daniel T, Zhenchao Qian, and Martha L Crowley. 2005. “Child Poverty Among Racial Minorities and Immigrants: Explaining Trends and Differentials.” Social Science Quarterly 86:1037-1059.

112

Logan, John R., and Brian J. Stults. 2011. The Persistence of Segregation in the Metropolis: New Findings from the 2010 Census. Census Brief prepared for Project US2010. http://www.s4.brown.edu/us2010/Data/Report/report2.pdf.

Ludwig, Jens, Jeffrey B. Liebman, Jeffrey R. Kling, Greg J. Duncan, Lawrence F. Katz, Ronald C. Kessler, and Lisa Sanbonmatsu. 2008. “What Can We Learn about Neighborhood Effects from the Moving to Opportunity Experiment?” The American Journal of Sociology 114:144-188.

Lumeng, Julie C, Danielle Appugliese, Howard J Cabral, Robert H Bradley, and Barry Zuckerman. 2006. “Neighborhood safety and overweight status in children.” Archives of Pediatrics & Adolescent Medicine 160:25-31.

Maimon, David, and Christopher R. Browning. 2010. “Unstructured Socializing, Collective Efficacy, and Violent Behavior Among Urban Youth.” Criminology 48:443-474.

Mair, Christina, Ana V. Diez Roux, and Jeffrey D. Morenoff. 2010. “Neighborhood stressors and social support as predictors of depressive symptoms in the Chicago Community Adult Health Study.” Health & Place 16:811-819.

Markides, Kyriakos S., and Karl Eschbach. 2005. “Aging, Migration, and Mortality: Current Status of Research on the Hispanic Paradox.” The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 60:S68 -S75.

Massey, Douglas S. 2004. “Segregation and Stratification: A Biosocial Perspective.” Du Bois Review: Social Science Research on Race 1:7-25.

Massey, Douglas S., and Nancy A. Denton. 1989. “Hypersegregation in U.S. Metropolitan Areas: Black and Hispanic Segregation along Five Dimensions.” Demography 26:373-391.

Massey, Douglas, and Nancy Denton. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press.

Massey, Douglas S., and Mary J. Fischer. 2006. “The effect of childhood segregation on minority academic performance at selective colleges.” Ethnic and Racial Studies 29:1.

Mather, Mark, Kelvin Pollard, and Linda A. Jacobsen. 2011. First Results From the 2010 Census. Washington, D.C.: Population Reference Bureau.

113

McDaniel, Marla, Christina Paxson, and Jane Waldfogel. 2006. “Racial Disparities in Childhood Asthma in the United States: Evidence From the National Health Interview Survey, 1997 to 2003.” Pediatrics 117:e868-877.

McEwan, Bruce S. 1998. “Stress, Adaptation, and Disease: Allostasis and Allostatic Load.” Annals of the New York Academy of Sciences 840:33-44.

Molnar, Beth E, Steven L Gortmaker, Fiona C Bull, and Stephen L Buka. 2004. “Unsafe to play? Neighborhood disorder and lack of safety predict reduced physical activity among urban children and adolescents.” American Journal of Health Promotion 18:378-386.

Moore, Latetia V., and Ana V. Diez Roux. 2006. “Associations of Neighborhood Characteristics With the Location and Type of Food Stores.” American Journal of Public Health 96:325-331.

Moore, Latetia V., Ana V. Diez Roux, Kelly R. Evenson, Aileen P. McGinn, and Shannon J. Brines. 2008. “Availability of Recreational Resources in Minority and Low Socioeconomic Status Areas.” American Journal of Preventive Medicine 34:16-22.

Morello-Frosch, Rachel, and Russ Lopez. 2006. “The riskscape and the color line: Examining the role of segregation in environmental health disparities.” Environmental Research 102:181-196.

Morenoff, Jeffrey D. 2003. “Neighborhood Mechanisms and the Spatial Dynamics of Birth Weight.” The American Journal of Sociology 108:976-1017.

Morenoff, Jeffrey D, Robert J Sampson, and Stephen W Raudenbush. 2001. “Neighborhood Inequality, Collective Efficacy, and the Spatial Dynamics of Urban Violence.” Criminology 39:517.

Natsuaki, Misaki N., Xiaojia Ge, Gene H. Brody, Ronald L. Simons, Frederick X. Gibbons, and Carolyn E. Cutrona. 2007. “African American Children‟s Depressive Symptoms: The Prospective Effects of Neighborhood Disorder, Stressful Life Events, and Parenting.” American Journal of Community Psychology 39:163-176.

Osborne, Cynthia, and Sara McLanahan. 2007. “Partnership Instability and Child Well‐Being.” Journal of Marriage and Family 69:1065-1083.

Ostir, G V, K Eschbach, K S Markides, and J S Goodwin. 2003. “Neighbourhood composition and depressive symptoms among older Mexican Americans.” Journal of Epidemiology and Community Health 57:987-992. 114

Osypuk, Theresa L., and Dolores Acevedo-Garcia. 2008. “Are Racial Disparities in Preterm Birth Larger in Hypersegregated Areas?” American Journal of Epidemiology 167:1295 -1304.

Osypuk, Theresa L., and M. Maria Glymour. 2011. “Examining the mental health effects of a housing mobility program by vulnerable subgroups: who may benefit, and who may not, from moves to lowpoverty neighborhoods?” Presented at the annual meeting of the Population Association of America, Washington, D.C.

Pager, Devah, and Hana Shepherd. 2008. “The Sociology of Discrimination: Racial Discrimination in Employment, Housing, Credit, and Consumer Markets.” Annual Review of Sociology 34:181-209.

Palloni, Alberto. 2006. “Reproducing Inequalities: Luck, Wallets, and the Enduring Effects of Childhood Health.” Demography 43:587-615.

Palloni, Alberto, and Elizabeth Arias. 2004. “Paradox Lost: Explaining the Hispanic Adult Mortality Advantage.” Demography 41:385-415.

Papas, Mia A., Anthony J. Alberg, Reid Ewing, Kathy J. Helzlsouer, Tiffany L. Gary, and Ann C. Klassen. 2007. “The Built Environment and Obesity.” Epidemiologic Reviews 29:129 -143.

Paradies, Yin. 2006. “A systematic review of empirical research on self-reported racism and health.” International Journal of Epidemiology 35:888 -901.

Park, Julie, and John Iceland. 2011. “Residential segregation in metropolitan established immigrant gateways and new destinations, 1990-2000.” Social Science Research 40:811-821.

Pascoe, Elizabeth A., and Laura Smart Richman. 2009. “Perceived discrimination and health: A meta-analytic review.” Psychological Bulletin 135:531-554.

Patel, K, Karl Eschbach, Laura Rudkin, M. Kristen Peek, and Kyriakos S. Markides. 2003. “Neighborhood context and self-rated health in older Mexican Americans.” Annals of Epidemiology 13:620-628.

Pearlin, Leonard. 1999. “The Stress Process Revisited: Reflections on Concepts and Their Interrelationships.” Pp. 395-415 in Handbook of the Sociology of Mental Health, edited by Carol S. Aneshensel and Jo C. Phelan. New York: Kluwer Academic.

115

Pearlin, Leonard I., Scott Schieman, Elena M. Fazio, and Stephen C. Meersman. 2005. “Stress, Health, and the Life Course: Some Conceptual Perspectives.” Journal of Health and Social Behavior 46:205 -219.

Peterson, Ruth D., and Lauren J. Krivo. 2009. “Segregated Spatial Locations, Race- Ethnic Composition, and Neighborhood Violent Crime.” The ANNALS of the American Academy of Political and Social Science 623:93 -107.

Phelan, J. C., and B. G. Link. 2005. “Controlling Disease and Creating Disparities: A Fundamental Cause Perspective.” The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 60:S27-S33.

Phelan, Jo C., Bruce G. Link, Ana Diez-Roux, Ichiro Kawachi, and Bruce Levin. 2004. “„Fundamental Causes‟ of Social Inequalities in Mortality: A Test of the Theory.” Journal of Health and Social Behavior 45:265-285.

Polednak, A P. 1996. “Trends in US urban black infant mortality, by degree of residential segregation.” American Journal of Public Health 86:723-726.

Portes, Alejandro, and Min Zhou. 1993. “The New Second Generation: Segmented Assimilation and Its Variants.” Annals of the American Academy of Political and Social Science 530:74-96.

Portes, Alejandro, Patricia Fernandez-Kelly, and William Haller. 2009. “The Adaptation of the Immigrant Second Generation in America: A Theoretical Overview and Recent Evidence.” Journal of Ethnic & Migration Studies 35:1077-1104.

Quillian, Lincoln. 2002. “Why Is Black-White Residential Segregation So Persistent?: Evidence on Three Theories from Migration Data.” Social Science Research 31:197-229.

Rosenbaum, Emily. 2008. “Racial/Ethnic Differences in Asthma Prevalence: The Role of Housing and Neighborhood Environments.” Journal of Health and Social Behavior 49:131 -145.

Ross, Catherine E., and John Mirowsky. 2001. “Neighborhood Disadvantage, Disorder, and Health.” Journal of Health and Social Behavior 42:258-276.

Rugh, Jacob S., and Douglas S. Massey. 2010. “Racial Segregation and the American Foreclosure Crisis.” American Sociological Review 75:629 -651.

Sampson, Robert J., Jeffrey D. Morenoff, and Thomas Gannon-Rowley. 2002. “Assessing „Neighborhood Effects‟: Social Processes and New Directions in Research.” Annual Review of Sociology 28:443-478. 116

Sampson, R.J., and S.W. Raudenbush. 2004. “Seeing disorder: Neighborhood stigma and the social construction of „broken windows‟.” Social Psychology Quarterly 67:319-342.

Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277:918 - 924.

Schempf, Ashley H., Amy M. Branum, Susan L. Lukacs, and Kenneth C. Schoendorf. 2007. “The Contribution of Preterm Birth to the Black-White Infant Mortality Gap, 1990 and 2000.” American Journal of Public Health 97:1255-1260.

Schieman, Scott, and Stephen C. Meersman. 2004. “Neighborhood Problems and Health Among Older Adults: Received and Donated Social Support and the Sense of Mastery as Effect Modifiers.” The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 59:S89 -S97.

Schulz, Amy J., Clarence C. Gravlee, David R. Williams, Barbara A. Israel, Graciela Mentz, and Zachary Rowe. 2006. “Discrimination, Symptoms of Depression, and Self-Rated Health Among African American Women in Detroit: Results From a Longitudinal Analysis.” American Journal of Public Health 96:1265-1270.

Schulz, Amy, David Williams, Barbara Israel, Adam Becker, Edith Parker, Sherman A. James, and James Jackson. 2000. “Unfair Treatment, Neighborhood Effects, and Mental Health in the Detroit Metropolitan Area.” Journal of Health and Social Behavior 41:314-332.

Schulz, Amy J., Shannon N. Zenk, Barbara A. Israel, Graciela Mentz, Carmen Stokes, and Sandro Galea. 2008. “Do Neighborhood Economic Characteristics, Racial Composition, and Residential Stability Predict Perceptions of Stress Associated with the Physical and Social Environment? Findings from a Multilevel Analysis in Detroit.” Journal of Urban Health 85:642-661.

Shapiro, Thomas M. 2004. The hidden cost of being African American: how wealth perpetuates inequality. New York: Oxford University Press.

Singh, Gopal K., and Mohammad Siahpush. 2001. “All-Cause and Cause-Specific Mortality of Immigrants and Native Born in the United States.” American Journal of Public Health 91:392-399.

Singh, Gopal K., Mohammad Siahpush, and Michael D. Kogan. 2010. “Rising Social Inequalities in US Childhood Obesity, 2003-2007.” Annals of Epidemiology 20:40-52.

117

Sondik, Edward J., David T. Huang, Richard J. Klein, and David Satcher. 2010. “Progress Toward the Healthy People 2010 Goals and Objectives.” Annual Review of Public Health 31:271-281.

South, Scott J, Kyle Crowder, and Erick Chavez. 2005. “Geographic Mobility and Spatial Assimilation among U. S. Latino Immigrants.” International Migration Review 39:577-607.

Stone, Susan, and Meekyung Han. 2005. “Perceived school environments, perceived discrimination, and school performance among children of Mexican immigrants.” Children and Youth Services Review 27:51-66.

Strauss, Richard S., and Harold A. Pollack. 2001. “Epidemic Increase in Childhood Overweight, 1986-1998.” JAMA: The Journal of the American Medical Association 286:2845 -2848.

Subramanian, S.V., Dolores Acevedo-Garcia, and Theresa L. Osypuk. 2005. “Racial Residential Segregation and Geographic Heterogeneity in Black/White Disparity in Poor Self-Rated Health in the US: a Multilevel Statistical Analysis.” Social Science & Medicine 60:1667-1679.

Sudano, Joseph J., and David W. Baker. 2006. “Explaining US racial/ethnic disparities in health declines and mortality in late middle age: The roles of socioeconomic status, health behaviors, and health insurance.” Social Science & Medicine 62:909-922.

Swaroop, Sapna. 2005. The Social Consequences of Racial Residential Integration. Ph.D. dissertation, Department of Sociology, University of Michigan, Ann Arbor, MI.

Teitler, Julien O., Nancy E. Reichman, Lenna Nepomnyaschy, and Melissa Martinson. 2007. “A Cross-National Comparison of Racial and Ethnic Disparities in Low Birth Weight in the United States and England.” Pediatrics 120:e1182-1189.

Thoits, Peggy A. 2010. “Stress and Health: Major Findings and Policy Implications.” Journal of Health and Social Behavior 51:S41 -S53.

Thompson, Douglas R, Ronaldo Iachan, Mary Overpeck, James G Ross, and Lori A Gross. 2006. “School Connectedness in the Health Behavior in School‐Aged Children Study: The Role of Student, School, and School Neighborhood Characteristics.” Journal of School Health 76:379-386.

Timberlake, Jeffrey M. 2007. “Racial and Ethnic Inequality in the Duration of Children‟s Exposure to Neighborhood Poverty and Affluence.” Social Problems 54:319-342.

118

Tolnay, Stewart E. 2003. “The African American „Great Migration‟ and Beyond.” Annual Review of Sociology 29:209-232.

Tull, Eugene S., Yah-Tyng Sheu, Cleve Butler, and Karimah Cornelious. 2005. “Relationships between perceived stress, coping behavior and cortisol secretion in women with high and low levels of internalized racism.” Journal of the National Medical Association 97:206-212.

Tull, Eugene S. et al. 1999. “Relationship of internalized racism to abdominal obesity and blood pressure in Afro-Caribbean women.” Journal of the National Medical Association 91:447-452.

Turner, R. Jay. 2003. “The Pursuit of Socially Modifiable Contingencies in Mental Health.” Journal of Health and Social Behavior 44:1-17.

Walsemann, Katrina M., Arline T. Geronimus, and Gilbert C. Gee. 2008. “Accumulating Disadvantage Over the Life Course.” Research on Aging 30:169 -199.

Walton, Emily. 2009. “Residential Segregation and Birth Weight among Racial and Ethnic Minorities in the United States.” Journal of Health and Social Behavior 50:427 -442.

Welch, Susan, Lee Sigelman, Timothy Bledsoe, and Michael Combs. 2001. Race and Place: Race Relations in an American City. Cambridge University Press.

Wheaton, Blair, and Philippa Clarke. 2003. “Space Meets Time: Integrating Temporal and Contextual Influences on Mental Health in Early Adulthood.” American Sociological Review 68:680-706.

Wight, Richard, Amanda Botticello, and Carol Aneshensel. 2006. “Socioeconomic Context, Social Support, and Adolescent Mental Health: A Multilevel Investigation.” Journal of Youth and Adolescence 35:109-120.

Wilkes, Rima, and John Iceland. 2004. “Hypersegregation in the Twenty-First Century.” Demography 41:23-36.

Williams, David R., and Chiquita Collins. 2001. “Racial residential segregation: a fundamental cause of racial disparities in health.” Public Health Reports (Washington, D.C.: 1974) 116:404-416.

Williams, David R., and Pamela Braboy Jackson. 2005. “Social Sources Of Racial Disparities In Health.” Health Affairs 24:325-334.

119

Williams, David R., and Michelle Sternthal. 2010. “Understanding Racial-ethnic Disparities in Health.” Journal of Health and Social Behavior 51:S15 -S27.

Williams, David R., and R. Williams-Morris. 2000. “Racism and Mental Health: the African American experience.” Ethnicity and Health 5:243-268.

Williams, David R., Harold W. Neighbors, and James S. Jackson. 2003. “Racial/Ethnic Discrimination and Health: Findings From Community Studies.” American Journal of Public Health 93:200-208.

Williams, David R., Yan Yu, James S. Jackson, and Norman B. Anderson. 1997. “Racial Differences in Physical and Mental Health: Socio-economic Status, Stress and Discrimination.” Journal of Health Psychology 2:335-351.

Williams, David, and Selina Mohammed. 2009. “Discrimination and racial disparities in health: evidence and needed research.” Journal of Behavioral Medicine 32:20-47.

Wilson, William J. 1996. When Work Disappears: The World of the New Urban Poor. New York: Knopf.

Wilson, William Julius, and Richard P. Taub. 2006. There Goes the Neighborhood: Racial, Ethnic, and Class Tensions in Four Chicago Neighborhoods and Their Meaning for America. New York: Knopf.

Wong, Carol A., Jacquelynne S. Eccles, and Arnold Sameroff. 2003. “The Influence of Ethnic Discrimination and Ethnic Identification on African American adolescents‟ School and Socioemotional Adjustment.” Journal of Personality 71:1197-1232.

Zenk, Shannon N., Amy J. Schulz, Barbara A. Israel, Sherman A. James, Shuming Bao, and Mark L. Wilson. 2005. “Neighborhood Racial Composition, Neighborhood Poverty, and the Spatial Accessibility of Supermarkets in Metropolitan Detroit.” American Journal of Public Health 95:660-667.

Zhou, Min. 1997. “Segmented Assimilation: Issues, Controversies, and Recent Research on the New Second Generation.” International Migration Review 31:975-1008.

Zhou, Min, and Carl L. Bankston. 1998. Growing Up American: How Vietnamese Children Adapt to Life in the United States. New York: Russell Sage Foundation.

Zubrinsky, Camille L., and Lawrence Bobo. 1996. “Prismatic Metropolis: Race and Residential Segregation in the City of the Angels.” Social Science Research 25:335-374.

120