Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Sep 1.
Published in final edited form as: J Community Psychol. 2022 Mar 23;50(7):3280–3299. doi: 10.1002/jcop.22837

Residential Mobility, Neighborhood Cohesion, and Depressive Symptoms among Urban-Dwelling African American Adolescents

Andrew A Gepty 1, Sharon F Lambert 1, Adam J Milam 2, Nicholas S Ialongo 2
PMCID: PMC9378346  NIHMSID: NIHMS1785337  PMID: 35332553

Abstract

Mechanisms linking residential mobility and depressive symptoms among urban-dwelling African American adolescents have received little attention. This research examined neighborhood cohesion as a possible mechanism. Participants were 358 urban-dwelling African American adolescents (Mage = 14.78; SD = .34) who reported their neighborhood cohesion in Grade 10 and depressive symptoms in Grades 9 and 11, and for whom residential address information was available. There was a significant indirect effect of past moves in middle school and depressive symptoms one year later through reduced neighborhood cohesion. However, the indirect effect was not significant in a propensity score matched sample. Results from the full sample of adolescents suggest that neighborhood cohesion may play a role in the experience of depressive symptoms following past moves in middle school. Different findings for the propensity score matched sample highlight the need for future studies of residential mobility to employ strategies to correct for possible selection bias.

Keywords: Residential Mobility, Neighborhood Cohesion, African American, Depression, Adolescence

The United States population is fairly mobile, with estimates of about 10% of individuals and families changing homes and residences in the past year (United States Census Bureau, 2018). Residential mobility is particularly common among low income urban African American youth and their families (Lawrence et al., 2015). The United States has a long history of racial segregation which has resulted in African Americans being pushed into racially homogeneous neighborhoods and contributed to racial disparities in household wealth due to the denial of home equity through inequality in mortgages prices and racist practices of real estate agents (Quillian et al., 2020). Despite the passing of the Fair Housing Act and the Equal Credit Opportunity Act by the US Federal Government, housing discrimination against the African American individuals are still present, as housing segregation, income disparities, and high eviction rates continue to persist and even widen (Hepburn et al., 2021; Owens, 2019; Quillian et al., 2020). The lack of affordable and adequate housing and lower incomes stemming from continuing racial discrimination may contribute to the higher rates of residential mobility among low income African Americans (Bruch & Swait, 2019; DeLuca et al., 2019). For the youth in highly mobile families, residential mobility has been associated with poor behavioral and emotional health including delinquency and low academic achievement (Adam & Chase-Lansdale, 2002; Anderson et al., 2014; Wolff et al., 2017). Furthermore, the negative effects of residential mobility are more pronounced for youth of color compared to White youth (Perkins, 2017). Among the few studies examining the association between residential mobility and depressive symptoms (Jelleyman & Spencer, 2008), few have examined possible causal mechanisms that explain the associations. The current study examined neighborhood cohesion as a possible mechanism that links residential mobility and depressive symptoms. As youth change residences, social ties with community members decrease, possibly resulting in increased depressive symptoms.

Residential Mobility and Depressive Symptoms

Transitions and changes characteristic of adolescence have been proposed to drive the increase of depressive symptoms during this developmental period (Cicchetti & Toth, 1998). Prior research has suggested that neighborhood rates of residential mobility are associated with depressive symptoms among adolescents (e.g., Jelleyman & Spencer, 2008). In addition, residential moves reported by adolescents have also been linked with depressive symptomatology. For example, urban adolescents with high levels of mobility in childhood reported higher levels of internalizing symptoms than those with less residential mobility (Coley et al., 2012). Similarly, in a large multicity study, more residential mobility in childhood and adolescence was associated with increased internalizing symptoms (Anderson & Leventhal, 2016). Residential mobility may be a reflection of other important factors facing adolescents such as poverty and familial stressors (Anderson, Leventhal, & Dupéré, 2014). However, evidence suggests that residential mobility is associated with depressive symptoms even when accounting for other stressors and may be more influential during the adolescent period (Anderson & Leventhal, 2016; Anderson, Leventhal, & Dupéré, 2014; Simoni & Bauldry, 2020).

Two primary challenges characterize the existing literature on residential mobility and depressive symptoms. First, most research on residential mobility has not examined the effect of an individual residential move during adolescence on depressive symptomatology. Instead, residential mobility has been characterized by different conceptualizations such as rates of mobility based on decennial US census data (Hurd et al., 2013), self-reported mobility of lifetime moves (Rumbold et al., 2012), or self-reported mobility in a specific number of years (Adam & Chase-Lansdale, 2002). Thus, there is a paucity of studies on the effects of discrete individual moves (e.g., Porter & Vogel, 2014; Anderson & Leventhal, 2016) or comparisons between aggregate moves and discrete moves (e.g., Coley et al., 2013). While there is a body of literature examining individual residential moves experienced by military youth (e.g., Bradshaw et al., 2010), there is less on community samples of urban-dwelling youth.

A second challenge in the mobility literature is that although direct effects of residential moves during adolescence have been demonstrated, much of the literature has not explored why residential moves can lead adolescents to experience depressive symptoms (Anderson et al., 2014). Understanding mechanisms can facilitate the development of interventions to mitigate the effect residential mobility has on depressive symptoms. One possible mechanism that may account for the negative effects of residential mobility is the change in the adolescent interactions with the neighborhood social environment resulting from a residential move. The adolescent experience of the social environment in their neighborhood has been found to be an important link between aspects of the neighborhood and psychological outcomes (e.g., Hurd et al., 2013). This study will examine the associations between residential mobility conceptualized as individual residential moves and adolescent depressive symptoms, and whether perceived neighborhood social cohesion is a mechanism explaining the association.

Social Disorganization Theory and Residential Mobility

Social Disorganization Theory describes how residential mobility may influence low income urban youth within their communities (Sampson & Groves, 1989). This framework posits that neighborhood disorganization, including neighborhood-level residential mobility, disrupts positive social processes in communities. According to the social disorganization framework, reduced collective efficacy, i.e., the combination of social cohesion, perceptions of how one can rely on and trust other community members for support (Buckner, 1988), among neighbors and the willingness to intervene for the common good of the community, is one means through which neighborhood disorganization can affect youth outcomes (Sampson & Groves, 1989; Sampson et al., 1997). One aspect of collective efficacy, cohesion among members of a neighborhood, is a proximate mechanism through which neighborhood disorganization may affect individuals (Sampson & Groves, 1989; Sampson et al., 1997).

Connections with neighbors and community members may promote healthy development and behaviors among urban youth (Smith et al., 2016); in addition, without a sense of social cohesion within a community, youth may not have the necessary support that would otherwise lessen potential maladaptive outcomes (Sampson & Groves, 1989). Residential moves may disrupt adolescents’ ability to build and maintain ties to peers and neighbors (Leventhal & Newman, 2010), and lead to disruptions in social ties established in their previous neighborhood. Social disorganization theory suggests that the undermining of social processes from neighborhood-level residential mobility may result in negative outcomes such as crime and delinquency (Sampson & Groves, 1989). Similarly, although not explored directly, this theory suggests that neighborhood cohesion may be one mechanism linking residential mobility and adolescents’ depressive symptoms.

Neighborhood Perceptions and Individual Mobility Among Residents

The research on the associations between residential mobility and adolescent development and outcomes within the Social Disorganization approach focused on deficits and structural/neighborhood level definitions of neighborhood and social cohesion. Research using objective indicators, which dominated neighborhood past research, have yielded valuable insights on the influence of neighborhoods on residents and provided the groundwork for understanding the influence of neighborhood contexts (Arcaya et al., 2016). However, there remain gaps in knowledge about individual level experiences in changing residences and their neighborhoods. Defining neighborhood by residential location may have contributed to the often weak and mixed findings on neighborhoods and health outcomes (Diez Roux, 2001; Inagami et al., 2007). Kwan (2012) described the mixed results from prior research using residential place as the proxy definition of neighborhood, as the Uncertain Geographic Context Problem (UGCoP). Structural measures of neighborhood may not reflect actual exposure to residents and may not be concordant with individuals’ actual experiences living in their neighborhoods.

Shifting from the focus on structural emphasis, the Pluralistic Neighborhood Theory (Aber & Nieto, 2000) highlights the role of collective socialization among neighborhood residents. Contrary to past deficit-focused conceptualizations, Pluralistic Neighborhood Theory posits that neighborhoods can be sources of strength and protective factors for urban-dwelling individuals (Aber & Nieto, 2000; Witherspoon & Ennett, 2011). Positive neighborhood social factors such as collective efficacy among residents can promote positive outcomes, mitigate negative outcomes, and be independent of neighborhood conditions (Witherspoon & Ennett, 2011; Witherspoon & Hughes, 2014). Researchers using the Pluralistic Neighborhood Theory emphasize the role of individual perceptions of neighborhood rather than census or administrative information (Witherspoon & Ennett, 2011). One study found that youth-reported home neighborhoods did not match any administrative spatial definitions (Colburn et al., 2019). Although adolescent perceptions of their neighborhood may somewhat match administrative data (Bass & Lambert. 2004; Goldman-Mellor et al., 2016), neighborhood perceptions may explain emotional outcomes better than census data. For example, Goldman-Mellor and colleagues (2016) found that in comparing adolescent self-report of neighborhood safety and census tract crime data, adolescent perceptions had stronger associations with anxious and depressive symptoms. Perceptions themselves have been shown to be the link between neighborhood structural concerns measured through the census and adolescent outcomes (Plunkettnet al., 2007). In a recent study, Jones and Dantzler (2021) found that in the context of residential mobility, perceptions of neighborhoods matter more in moving than the objective neighborhood setting. These findings highlight the importance of actual reported experiences and perceptions of neighborhood rather than assumed exposures with administrative data.

The importance of individual reported residential mobility and neighborhood cohesion has been supported by research in both adults and youth. With few exceptions (e.g., Sun et al., 2004), research has demonstrated the link between census-level residential mobility and diminished sense of social cohesion among adult samples. For example, Kingston and colleagues (2009) found that census-level residential mobility was negatively associated with social cohesion among adults in Denver. Youth who lived in areas characterized by higher census-level residential mobility reported lower perceptions of neighborhood cohesion (Hurd et al., 2013). Diminished social ties, or formal and informal relationships between individuals, may lead to the reduction in social cohesion among those experiencing residential mobility. In a comparison of movers and non-movers following a government neighborhood desegregation program, parents of adolescents who moved reported reduced social ties with individuals in the new neighborhood (Fauth et al., 2004). Magdol and Bessell (2003) found that families that experienced recent moves to outside the city where they previously lived reported reduced social ties to their new neighborhood. The few studies examining the association between residential mobility and perceptions of neighborhood social cohesion among adolescents have yielded similar findings. Pribesh and Downey (1999) found that youth who moved reported a reduction in social ties as defined as participation in school, family, or community settings.

Neighborhood Social Cohesion and Depression

The importance of social processes, such as social support, in the development and maintenance of depressive symptoms among adolescents has been a robust finding throughout the literature, with less social support being associated with more depressive symptoms (Rueger et al., 2016). Neighborhood cohesion may operate similarly to social support in relation to depressive symptoms (Leventhal & Brooks-Gunn, 2000). Consistent with this proposition, neighborhood social cohesion has been associated with fewer depressive symptoms (Hurd et al., 2013). Urban youth who lived in low cohesion neighborhoods were more likely than youth living in high cohesion communities to report the experience of depressive symptoms (Kingsbury et al., 2015). Adolescents who perceived that their neighborhoods were sources of support reported fewer symptoms of depression, consistent with previous work on perceived individual social support and depressive symptoms (Hurd et al., 2013; Rueger et al., 2016).

Empirical studies have found evidence supporting the proposition that neighborhood-level social disruption associated with residential mobility is linked with maladaptive outcomes through lowered sense of social cohesion and collective efficacy (Kingston et al., 2009). The few studies that have examined the links between individual-level residential mobility, neighborhood cohesion, and depressive symptoms have provided evidence for the individual history of residential moves → lower neighborhood cohesion → increased depressive symptoms pathway among adolescents. This same pattern is found in studies examining neighborhood level residential mobility. One study found that neighborhood disorganization, which included poverty and residential instability, was associated with increased depressive symptoms through lower adolescent reports of collective efficacy (Xue et al., 2005). Hurd and colleagues (2013) found a similar pattern with adolescents in neighborhoods with high residential mobility reporting lower levels of neighborhood cohesion; in turn, lower neighborhood cohesion was associated with increased depressive symptoms.

Current Study

Prior research has documented an association between neighborhood-level residential mobility and depressive symptoms among low income urban youth (Jelleyman & Spencer, 2008). Propositions put forth in Social Disorganization Theory suggest that the disruption of neighborhood social cohesion accounts for the association between residential mobility and adolescent depressive symptoms (Hurd et al., 2013; Sampson & Groves, 1989). However, while prior research has supported the individual pathways between residential mobility and depressive symptoms (Jelleyman & Spencer, 2008), residential mobility and neighborhood social cohesion (Pribesh & Downey, 1999), and neighborhood cohesion and depressive symptoms (Xue et al., 2005), few have examined the constructs together as part of a sequential pathway linking residential mobility to less cohesion, and less cohesion leading to depressive symptoms (see, Hurd et al., 2013 for an exception). Pluralistic Neighborhood Theory (Aber & Nieto, 2000) posits that positive neighborhood aspects and residents’ perceptions and experiences are important in understanding the role of the neighborhoods. Past research on residential mobility has focused on structural levels of mobility and neighborhood cohesion rather than individual perceptions and individual experiences of residential mobility. The primary aims of this study were to examine prospective associations between individual residential mobility and depressive symptoms with perceived neighborhood cohesion as a mechanism in a sample of low income urban African American adolescents. Another aim was to address methodological and conceptual limitations in the existing residential mobility literature.

To address the limitations in and expand on the existing residential mobility literature, the current study examined the associations between recent mobility of a past year move and aggregate history of past moves during middle school (6th – 9th Grade) on depressive symptoms in 11th Grade. Neighborhood cohesion in 10th Grade was examined as a mechanism linking residential mobility and depressive symptoms. Using statistical corrections for selection bias, the current study addressed possible issues with using dichotomous presence or absence of past year moves. We hypothesized that a move in the past year or a history of past moves and would be associated with an increase in depressive symptoms. Additionally, we hypothesized that a residential mobility would be associated with lower perceived neighborhood cohesion, which in turn would be associated with greater depressive symptoms. We hypothesized that the effect of a residential move on depressive symptoms would be accounted for by perceptions of neighborhood cohesion.

Method

Participants

Participants were 358 African American middle school students who were first assessed in the fall of first grade as part of a longitudinal evaluation of two school-based preventive interventions targeting early learning and aggressive behavior in 1st Grade (Ialongo et al., 1999b). The original sample consisted of 678 children entering 1st Grade at nine public elementary schools who had parental consent and provided assent for enrollment. Of the 678 children who participated in the intervention trial, 585 (86.2%) were African American. Of the 585 African American participants, 358 (61.2%) participated in follow-up assessments at Grades 9, 10, and 11; had valid residential addresses in each of Grades 5 to 10; and completed measures of neighborhood cohesion in Grade 10, depressive symptoms in Grade 9 and 11, and neighborhood disorder in Grade 9. These 358 participants comprised the sample of interest (i.e. analytic sample). Prior research suggests that this adolescent period is one in which residential mobility has an impact on depressive symptoms (e.g., Anderson, Leventhal, & Dupéré, 2014). Approximately half of the analytic sample was female (47.8%). Approximately three quarters of the sample (70.4%) received free- and reduced-priced meals (FARMs), an indicator of socioeconomic status. At the 9th grade assessment, youth ranged in age from 14.18 to 15.98 (M = 14.78, SD = .34). Based on t tests and chi square comparisons, the 358 African American students participating in this study did not differ from the 227 African American students not included in this study on past residential moves, distance moved, neighborhood cohesion, depressive symptoms in Grade 9, demographic characteristics (i.e., age, gender, FARM), neighborhood disorder, or intervention status (p > .05 for all variables). There was a significant difference in depressive symptoms in Grade 11 (t(450) = 2.15, p = .03); those not included and had non-missing data (N = 98) reported more depressive symptoms (M = 9.01, SD = 6.39) than those included in the sample (M = 7.24, SD = 7.03).

Measures

Residential mobility.

All participant addresses were converted to geocoded X,Y (longitude, latitude) coordinates. A change in residence for an adolescent was the difference in the coordinates between two consecutive grades, and was coded as a Move (1) for the latter grade year. No change in residence was coded as No Move (0) for the latter grade year. Two types of residential moves were calculated. Past Year Move Grade 10 was calculated as a Move or No Move between Grades 9 and 10. The sum of moves for grades 6, 7, 8, and 9 also was calculated; the possible range for Past Moves in Grades 6–9 was 0–4. Past Year Move Grade 10 and the sum of moves for Grades 6 – 9 were analyzed in separate models due to differences in outcomes of moving in the short term and in the long term (Garboden et al., 2017).

Neighborhood cohesion.

Adolescent perceptions of connections and support in their neighborhoods were assessed using seven items from the Neighborhood Cohesion Index (Buckner, 1988). The items included questions such as, “I fit in with my neighbors” and “I believe my neighbors would help me in an emergency.” Youth reported the extent that they agreed with statements about their neighborhood and neighbors on a four-point Likert scale (1 = “strongly disagree, 4 = “strong agree”). The seven items were aggregated to create a sum score, where higher scores indicated stronger feelings of cohesion in the neighborhood. Cronbach’s alpha for this scale was .81 in 10th Grade.

Depressive Symptoms.

Depressive symptoms were assessed using the Baltimore How I Feel-Adolescent Version, Youth Report (BHIF-AY; Ialongo et al., 1999), a youth self-report measure of depressive and anxious symptoms. Items for the depression scale were generated directly from DSM-IV (American Psychiatric Association 1994) criteria or drawn from existing child self-report measures, including the Children’s Depression Inventory (Kovacs, 1983, unpublished manuscript), the Depression Self-Rating Scale (Asarnow & Carlson, 1985), and the Hopelessness Scale for Children (Kazdin et al., 1986). Youth reported the frequency of depressive symptoms over the last 2 weeks on a four-point scale (1 = “never,” 4 = “most times”). A sum score was created by aggregating the items, and higher scores indicate more depressive symptoms. Cronbach’s alpha for the Depression scale was .87 in 9th Grade and .88 in 11th Grade.

Covariates

Distance moved.

The linear distance between residences was calculated using QGIS 2.18 Las Palmas (QGIS Development Team, 2016). Past year distance moved was the linear distance between 9th and 10th grade residences. Grade 6 to 9 distance moved was the sum of linear distances moved for 6th, 7th, 8th, and 9th grade Moves.

Neighborhood disorder.

Neighborhood disorder was measured using 10 items from the Neighborhood Environment Scale (NES; Elliott et al., 1985), a measure of neighborhood disorganization, including questions about crime (e.g., “ Every few weeks, some kid in my neighborhood gets beat up or mugged.”) and drug use and sales (e.g., “I have seen people using or selling drugs in my neighborhood.”). Youth rated each item on a 4-point Likert scale (1 = not at all, 4 = very much), and higher scores indicated greater neighborhood disorganization. Cronbach’s alpha for the scale was .86 for 9th grade. Neighborhood disorder was included to account for adolescents’ perception of their neighborhood of their residence at the time of the study.

Baseline symptomatology.

To account for the possibility that a prospective association between a residential mobility and subsequent depressive symptoms was due to the association between depressive symptoms prior to a move, the same measure of depressive symptoms in 9th grade were included as a control variable. Depressive symptoms in 9th Grade were controlled in order to account for depressive symptoms prior to a past year move in Grade 10 and to capture depressive symptomatology following a history of residential moves in Grades 6–9.

Demographic Information.

Participants’ age, gender, and receipt of FARMs were used as covariates to account for the possible influence of these demographic characteristics on the mobility variables, neighborhood cohesion, and depressive symptoms in 11th Grade.

Analytic Strategy

Propensity Score Matching.

Propensity score matching was used to account for selection bias by creating a balanced sample across covariates between adolescents who moved and did not move in the past year. Propensity scores are values assigned to each individual and reflect the probability for a given individual of experiencing or not experiencing an event or condition based on observed baseline covariates (Austin, 2011; Rosenbaum & Rubin, 1983). The propensity scores are created through a series of logistic regressions of adolescents moving or not moving in the past year on covariates that are theoretically related to the moving and depressive symptoms. Propensity scores for each individual in both groups reflect the characteristics of relevant factors for selection into a specified event or condition. With the calculation of the propensity scores, individuals in the event group are matched to individuals in the non-event group and balanced across covariates, and the matched sample would provide an unbiased estimate of effects of moving in the past year (Austin, 2011; Wolff, et al., 2017).

The Propensity Score Matching procedure in SPSS 25 was used to estimate probabilities of movers and non-movers in Grade 10 based on selected covariates. Propensity scores were generated using logistic regression of moving in the past year on selected covariates which included demographic variables (gender, age, FARMs), neighborhood disorder and depressive symptoms at Grade 9, and the history of residential mobility with Past Moves in Grades 6–9, and Distance Moved Grades 6–9. These covariates were selected for the propensity score model because they were associated with whether or not adolescents had moved, and in order to approach balance on demographic and background variables between the two groups. Once propensity scores were generated, matching of Past Year Movers and Non-movers was achieved using a one-to-one nearest neighbor matching without replacement. In order to ensure optimal and close matches, a .10 caliper of the standard deviation of the logit of the propensity score was applied (Austin, 2011). A new dataset was populated with only movers and their respective matched non-movers.

Mediation Analysis.

To examine whether adolescents who moved in the past year or had a history of moving in middle school (Grades 6–9) would experience an increase in depressive symptoms through perception of neighborhood cohesion, mediation analyses were conducted using the PROCESS macro in SPSS 25 (Hayes, 2017). Two regressions were performed; Past Year Move and Past Moves in Grades 6–9 were examined as predictor variables in separate models similar to past research finding different findings for each type of mobility (e.g., Brown & Orthner, 1990). Neighborhood cohesion in Grade 10 was regressed on Past Year Move, Past Moves in Grades 6–9, and covariates. Next, depressive symptoms were regressed on neighborhood cohesion, the residential mobility variable, and covariates. Indirect effects of residential mobility on depressive symptoms through neighborhood cohesion were tested using 5000 bootstrap samples with Davidson-MacKinnon heteroskedasticity-consistent standard errors to produce 95% bootstrapped confidence intervals (Hayes, 2017; Hayes & Cai, 2007).

Results

Descriptives

Sixty-nine adolescents moved in the past year between Grades 9 and 10. A majority of adolescents (n = 240; 67.0%) moved at least once between Grades 6 and 9. Of those with a history of moving between Grades 6 and 9, the majority had moved once (n = 144; 60.0%) or twice (n = 78; 32.5%) during those Grades (see Table 1). Descriptive statistics and bivariate correlations of all study variables are reported in Table 2. Past Year Move Grade 10 and Past Moves in Grades 6–9 were not associated with depressive symptoms in grade 11. Both mobility variables were significantly negatively correlated with neighborhood cohesion (rPast Year =−.16, p < .001; rPast Moves 6th-9th = −.15, p < .001). Neighborhood cohesion in Grade 10 was correlated negatively with depressive symptoms in grade 11 (r = −.17, p < .001).

Table 1.

Descriptive Statistics for Past Year Move and Past Moves in Grades 6 by Sample Type

Full Sample Propensity Matched Sample
N = 358 (%) Distance Moved (km) N = 138 (%) Distance Moved (km)
Past Year Move Mean (SD) Range Mean (SD) Range
Did Not Move 289 (80.70) - - 69 (50.00) - -
Did Move 69 (19.30) 7.44 (71.60) 1.90 – 1272.71 69 (50.00_ 7.55 (71.60) 1.90 – 1272.71
Past Moves in Grade 6–9
Did Not Move 111 (33.00) - - 44 (31.90) - -
Did Move 240 (67.00) 7.77 (138.73) 0.10 – 2460.24 94 (68.10) 24.27 (210.50) 0.01 – 2460.24
One Move 144 (40.20) 5.55 (10.92) <0.01 – 107.73 56 (40.60) 3.70 (5.09) 0.01 – 21.12
Two Moves 78 (21.80) 64.44 (292.27) 0.18 – 2460.24 31 (22.50) 99.21 (489.78) 1.33 – 2460.24
Three Moves 15 (4.20) 31.77 (56.10) 1.51 – 232.45 5 (3.60) 59.30 (96.87) 10.76 – 232.45
Four Moves 3 (0.80) 19.97 (14.05) 10.36 – 36.09 2 (1.40) 23.23 (18.19) 10.36 – 36.09
Open in a new tab

Notes. km = kilometers. Distance Moved (km) is measurement of Euclidian distance moved by participants. Distance for Past Year Move reflects the distance moved in km for adolescents who moved in the past year. Distance Moved for Past Moves in Grade 6–9 is the total distances of all moves by adolescents in that time period.

Table 2.

Correlations, Means, and Standard Deviations of Primary Study Variables and Covariates

1 2 3 4 5 6 7 8 9 10 11
1. Moves Past Year - .17* −.26** −.04 .17* −.06 .12 −.13 .15 .06 −.06
2. Moves 6th −9th .10 - −.28** −.07 .11 .14 .15 −.01 .18* −.03 −.05
3. Perceived Cohesion −.16** −.15** - −.04 −.04 −.09 −.03 .09 −.04 −.10 −.10
4. Dep. Symptoms 11th .00 .01 −.17** - .23** −.09 −.19* .29** −.07 .11 .52**
Covariates
5. Distance MovedPast Yr .21 .07 −.03 .14** - −.01 −.04 .04 .08 −.03 .09
6. Distance Moved6th−9th  −.02 .14** −.09 −.06 .00 - .08 .08 .05 .06 .07
7. Age .09 .18** −.11* −.02 −.01 .10 - −.05 .06 −.14 −.06
8. Gender −.07 −.06 −.05 .20** .02 .02 −.07 - −.02 .08 .13
9. Lunch Status .10 .18** −.09 .02 .05 .06 .08 .01 - −.06 .01
10. Perceived Disorder .06 −.04 −.20** .23** −.01 .01 −.06 .07 −.01 - .30**
11. Dep. Symptoms 9th .02 −.04 −.12* .54** .07 .02 −.01 .14** .02 .31** -
Mean 0.19 1.00 19.60 7.24 7.45 17.77 14.78 1.48 0.70 17.65 8.42
SD 0.40 0.89 5.18 7.03 71.60 138.73 0.34 0.50 0.46 5.18 7.13
Matched Mean 0.50 1.02 19.29 7.54 19.33 26.27 14.80 1.47 0.73 17.96 9.17
Matched SD 0.50 0.91 5.27 6.84 114.58 210.50 0.32 0.50 0.45 6.87 7.82
Open in a new tab

Notes. Propensity Matched Sample above the diagonal. Moves 6th–9th = Past Moves in Grades 6 to 9. Dep. = Depressive.

*

p <.05.

**

p < .01

Mediation Analyses (Full Sample)

Past Year Moves and Past Moves in Grade 6–9 were examined in separate mediation models, with neighborhood cohesion at Grade 10 as the mediator and depressive symptoms at Grade 11 as the outcome. Gender, FARMs status, age, depressive symptoms in Grade 9, Past Year Distance Moved, Grades 6–9 Distance Moved, Neighborhood Disorder were covariates in each model (see Table 3).

Table 3.

Mediation Model of Past Year Move and Past Moves in Grade 6–9 Predicting to Depressive Symptoms in Grade 11 Through Neighborhood Cohesion

DV : Perceived Cohesion
Model Summary R2 = .10 MSE = 24.78 F (HC3) = 4.40 p < .01
Measure B SE t p
Intercept 42.23 12.71 3.32 <.01
Moves Past Year −1.71 0.79 −2.17 .03*
Moves 6th −9th −0.71 0.34 −2.13 .03*
Age −1.19 0.86 −1.38 .17
Gender −0.51 0.54 −0.94 .35
Lunch Status −0.51 0.58 −0.89 .38
Depressive 9th −0.04 0.04 −1.11 .27
Distance Moved6th−9th 0.00 0.01 −0.18 .86
Distance MovedPast Yr 0.00 0.01 0.07 .95
Perceived Disorder −0.14 0.04 −3.20 <.01**
DV: Depressive Symptoms
Model Summary R2 = .34 MSE = 33.56 F (HC3) = 12.16 p < .01
Measure B SE t p
Intercept 4.18 13.88 0.30 .76
Moves Past Year −0.81 0.89 −0.90 .36
Moves 6th −9th 0.29 0.38 0.77 .44
Perceived Cohesion −0.14 0.06 −2.19 .03*
Age −0.14 0.93 −0.15 .88
Gender 1.73 0.65 2.65 <.01**
Lunch Status −0.02 0.64 −0.04 .97
Depressive 9th 0.48 0.06 7.80 <.01**
Distance Moved6th−9th  0.00 0.00 −1.01 .31
Distance MovedPast Yr 0.01 0.02 0.56 .57
Perceived Disorder 0.06 0.05 1.17 .24
Indirect Effect of Residential Mobility on Depressive Symptoms
Effect Bootstrap Confidence Interval SE Lower Bootstrap Confidence Interval Upper Bootstrap Confidence Interval
Moves Past Year .23 .15 −.00 .58
Moves 6th – 9th .10 .06 .00 .25
Open in a new tab

Notes. Bootstrap Confidence Interval without zero indicates significant indirect effect.

*

p < .05.

**

p < .01

There was no simple association between moving in the past year and depressive symptoms the following year. Past Year Move Grade 10 was associated with less perceived neighborhood cohesion (B = −1.71, p = .03). In turn, less neighborhood cohesion predicted higher depressive symptoms (B = −.14, p = .03). However, the resulting bootstrap confidence intervals did not provide evidence of an indirect effect (indirect effect = .23, SE = .15, 95% CI = −.00, .08).

There was no simple association between Past Moves in Grades 6–9 and depressive symptoms the following year. Past Moves in Grades 6–9 were associated with less neighborhood cohesion Grade 10 (B = −71, p = .03). Furthermore, lower neighborhood cohesion was associated with higher depressive symptoms a year later (B = −.14, p = .03). There was a significant indirect effect of more Past Moves in Grades 6–9 on prospective depressive symptoms through lower perceived neighborhood cohesion (indirect effect = .09, SE = .06, 95% CI = >.00, .25).

Propensity Score Matched Sample

Matching Results.

All 69 movers in the past year were matched with corresponding non-movers using the Propensity Score Matching procedure. There were no significant differences among covariates between adolescents matched and those not selected for the matched sample (p > .05 for all matching t tests). In ensuring for optimized matching between movers and non-movers, there was no significant difference between matched movers and non-movers on all included covariates (p > .05 for all matching t tests) except for Past Moves in Grades 6–9. Matched movers (M = 1.17, SD = .84) moved more than non-movers (M = .87, SD = .95), t(136) = −1.99, p = .05.

Matched Sample Descriptives.

Adolescents in the matched sample were 47.1% female. A majority of adolescents in the matched sample received FARMs (73.2%) and the mean age in Grade 9 was 14.80 (SD = .32).The majority of adolescents experienced at least one move during Grades 6 to 9 (n = 94). Of those with a history of moving, the majority had moved either once (n = 56) or twice (n = 33) (See Table 1). There was no correlation between Past Year Move Grade 10 or Past Moves in Grades 6–9 and depressive symptoms in Grade 11 (p > .05 for all correlations). Both mobility variables were negatively correlated with neighborhood cohesion (Past Year Move Grade 10, r = −.26, p < .01; Past Moves in Grades 6–9, r = −.28, p < .01). There was no significant correlation between neighborhood cohesion at Grade 10 and depressive symptoms in Grade 11 (See Table 2).

Mediation Analyses (Matched sample).

Past Year Move Grade 10 and Past Moves in Grades 6–9 were analyzed in separate mediation models (see Table 3 and 4). There was no simple association between moving in the past year and depressive symptoms in Grade 11. There was a negative association between moving in the past year and neighborhood cohesion in Grade 10 (B = −2.36, p = .01). However, there was no significant association between neighborhood cohesion in Grade 10 and depressive symptoms in Grade 11 (B = −.04, p > .10). The resulting bootstrap confidence intervals did not provide evidence of an indirect effect (indirect effect = .01, SE = .03, 95% CI = −.04, .09).

Table 4.

Mediation Model of Past Year Move and Past Moves in Grade 6–9 Predicting to Depressive Symptoms in Grade 11 Through Neighborhood Cohesion with Propensity Score Matched Sample

DV : Perceived Cohesion
Model Summary R2 = .16 MSE = 24.98 F (HC3) = 2.35 p = .02
Measure B SE t p
Intercept 18.41 23.33 0.79 .01
Moves Past Year −2.36 0.90 −2.62 .01*
Moves 6th −9th −1.45 0.69 −2.10 .04*
Age −0.30 1.55 −0.19 .85
Gender 0.60 0.88 0.68 .50
Lunch Status 0.34 0.96 0.36 .72
Depressive 9th −0.07 0.06 −1.10 .27
Distance Moved6th−9th  0.00 0.03 −0.05 .96
Distance MovedPast Yr 0.00 0.01 0.10 .92
Perceived Disorder −0.08 0.07 −1.07 .29
DV: Depressive Symptoms
Model Summary R2 = .40 MSE = 30.31 F (HC3) = 4.73 p < .01
Measure B SE t p
Intercept 46.77 21.51 2.17 .03*
Moves Past Year 0.12 1.08 0.11 .91
Moves 6th −9th −0.04 0.54 0.07 .94
Perceived Cohesion −0.04 0.09 −0.49 .63
Age −3.03 1.44 −2.10 .04*
Gender 3.01 1.08 2.89 <.01**
Lunch Status −1.16 1.16 −1.00 .32
Depressive 9th 0.43 0.10 4.54 <.01**
Distance Moved6th−9th  −0.00 0.01 −0.57 .57
Distance MovedPast Yr 0.01 0.01 1.06 .29
Perceived Disorder −0.07 0.08 0.96 .34
Indirect Effect of Residential Mobility on Depressive Symptoms
Effect Bootstrap Confidence Interval SE Lower Bootstrap Confidence Interval Upper Bootstrap Confidence Interval
Moves Past Year .01 .03 −.04 .09
Moves 6th – 9th .06 .13 −.17 .38
Open in a new tab

Notes. Bootstrap Confidence Interval without zero indicates significant indirect effect.

*

p < .05.

**

p < .01

There was no simple association between Past Moves in Grades 6 to 9 and depressive symptoms in Grade 11. Past Moves in Grade 6–9 were associated with less neighborhood cohesion in Grade 10 (B = −1.45, p = .04). However, there was no association between neighborhood cohesion in Grade 10 and depressive symptoms in Grade 11 (B = −.04, p > .10). There was no significant indirect effect (indirect effect = .06, SE = .13, 95% CI = −.17, .38).

Discussion

Inattention to possible mechanisms of change has precluded a full understanding of the association between residential mobility and depressive symptom among urban adolescents. Results from the present study provided some evidence for an indirect effect of a history of residential mobility on prospective depressive symptoms through perceptions of neighborhood cohesion. In this study, African American adolescents with more past moves during middle school perceived less neighborhood cohesion in high school. In turn, lower perceptions of neighborhood cohesion were associated with more depressive symptoms. These findings are consistent with past literature on residential mobility and neighborhood cohesion and their effects on depressive symptoms and highlight the effects of the timing of mobility during adolescence. However, there were no indirect effects for adolescents who moved in the past year. Moreover, when using propensity score matched sample to account for selection bias among recent movers, the association between neighborhood cohesion and depressive symptoms was no longer apparent. These propensity score sample findings throw into question past research that has not corrected for biases in measures of residential mobility.

Residential Mobility and Neighborhood Cohesion

The associations between history of residential mobility and a past year move with lower perceptions of neighborhood cohesion across the full sample and the propensity score matched sample demonstrate the significance of change in residence on how low income urban African American adolescents view their communities. Consistent with Social Disorganization Theory (Sampson & Groves, 1989), a change in residence disrupted adolescents’ sense of connection with their communities. Prior research has shown that residential mobility measured at the aggregate neighborhood level (census data) was associated with less neighborhood cohesion (Hurd et al., 2013). Similar associations were observed in the present study when adolescents’ individual residential moves were associated with lower perceptions of neighborhood cohesion, highlighting the effect that change in residence can have on adolescent perceptions of cohesion. A change in residence may be accompanied by severing of ties with peers, neighbors, and family members in the previous neighborhoods and difficulty making new ties (Leventhal & Newman, 2010), resulting in the lower perceptions of neighborhood cohesion.

Neighborhood Cohesion and Depressive Symptoms

Prior research has suggested the positive value of neighborhood cohesion for reducing depressive symptoms among African American youth (e.g., Hurd et al., 2013). In the full sample, perceived neighborhood cohesion was associated with depressive symptoms in the hypothesized manner: lower neighborhood cohesion was associated with higher depressive symptoms prospectively. As suggested by previous research, the decreased sense of support in one’s community could influence depressive symptoms through less engagement with protective relationships in the neighborhood (e.g., Hurd et al., 2013). However, among a balanced sample of past year movers and non-movers, perceived neighborhood cohesion was not significantly associated with depressive symptoms, casting doubt on the association between neighborhood cohesion and depressive symptoms among past year movers. It is possible that the matching of covariates accounted for variance that produced the positive finding in the full sample analyses.

Using a Full Sample Versus a Propensity Score Matched Sample

Researchers increasingly have used propensity score matching to account for the possible selection bias inherit in families’ residential mobility (Garboden et al., 2017). For example, Anderson and Leventhal (2016) and Wolff and colleagues (2017) used propensity matched samples to examine the associations between movers and behavioral outcomes compared to non-movers; they found that residential mobility was associated with increases in internalizing and externalizing symptoms. The current study attempted to correct for potential bias by employing propensity score matching and creating groups of past year movers and non-movers who were equal on selected mobility and depression covariates. The associations between neighborhood cohesion and depression may be related to the covariates which are balanced though matching (Garboden et al., 2017) and which may reduce the variance that accounts for the effects. The reduced sample size (358 to 69) following the propensity match procedures may have also contributed to the lack of replication of findings from the analysis using the full sample. The smaller sample size may have been underpowered, making it difficult for the associations between variables to be detected in the mediation models (Fritz & MacKinnon, 2007). Other research attempting to use a propensity score-matched sample to account for bias failed to replicate findings in the non-matched sample. For example, Porter and Vogel (2014), in examining the associations between residential mobility and delinquency, failed to replicate the significant findings in regression models after using propensity scores to reduce selection bias. The use of propensity scores may produce more stringent results than traditional regression analyses, and may be a more rigorous test of effects for events like residential mobility that are subject to selection bias. However, propensity score matched models may be limited by the covariates available in the model and the sample size as seen in the current study (Nuttall & Houle, 2008).

Implications

The negative effects of a past year move and past moves in middle school on adolescent perceptions of neighborhood cohesion highlight the negative effect moving has on perceptions about their community. Thus, when considering the effects of residential mobility, neighborhood cohesion may be an important factor to examine among African American adolescent movers. For youth who have moved recently or have a history of chronic moves, interventions emphasizing community belongingness and connectedness may be beneficial to restoring the positive effect of a sense of cohesion. Without a sense that their neighborhoods are sources of support, adolescents, may have difficulties reaching out and building positive relationships during this critical developmental period that would normally promote positive outcomes. Programs, such as Learning Families Project (Shen et al., 2017) and Your Family, Your Neighborhood (Lechuga-Peña & Brisson, 2018), that promote connections between neighbors may foster increased perceptions of cohesion and protect youth against the effects of residential moves.

This research has important implications for research focused on residential mobility and other events which may be rare within a sample or population under study. In this study, using propensity score matching to correct for possible biases produced null findings. Because propensity score matching can correct for selection biases and emulate the effects of random assignment to condition, using this matching technique can be more accurate in estimating of the associations and increasing confidence in causal inference. In this study, this added precision made it difficult to see effects of dichotomous and infrequent events such as residential moves. The null findings observed in this study suggest the need to replicate the past findings on residential mobility effects. Effects reported in prior research may be in fact smaller or even non-significant as in the current findings. The integration of methods such as propensity score matching may add rigor to the evaluation of past findings that may be flawed due to selection biases. Along with the increase in rigor due to the accounting for biases, propensity score matching can strengthen the analysis of rare or uncommon events.

Strengths and Limitations

This current study expands the literature on residential mobility, neighborhood cohesion, and depressive symptoms. The findings from this community epidemiologically defined sample can be generalized to similar populations, namely low income urban African American adolescents. By focusing on individual-level variables measuring residential mobility, neighborhood cohesion, and depressive symptoms, the study extends the residential mobility literature to consider conceptualizations and measurement beyond traditional methods that often rely on neighborhood-level data. Thus, the study was able to capture a more proximal experience of an adolescent’s experience in their neighborhood following a change in residence than a neighborhood-level variable. By assessing the possible mediating effect of perceived neighborhood cohesion, the study adds needed exploration into possible mechanisms linking residential mobility to mental health outcomes. Although there have been studies using longitudinal designs to study neighborhood effects, most prior investigations of neighborhood effects and residential mobility have been cross-sectional in design, preventing any inferences of cause and effect (Arcaya et al., 2016; Garboden et al., 2017). The longitudinal multi-wave design employed in this study enables the estimation of causal inference among the variables. Inclusion of past depressive symptoms accounted for the possible stable trajectory of depressive symptoms during adolescence (Costello et al., 2008). Residential mobility has been defined in various ways in terms of time and frequency of occurrence with different findings based on those conceptualizations (Garboden et al., 2017), but few have evaluated different measures within the same study. By examining the effects of aggregate moves from Grades 6 to 9 and a proximate past move during the past year, it was possible to evaluate whether the timing of the move matters for outcomes such as neighborhood cohesion and depressive symptoms. By using propensity score matching, possible biases in the selection of families into being movers or non-movers were corrected and accounted for in the study. With the reduced biases, the matched sample yielded more confidence in our longitudinal findings.

These study strengths should be evaluated in the context of some limitations. One consideration in using propensity score matching is the identification and inclusion of relevant covariates that are related to the condition and outcome; propensity scores are limited by covariates included in the model (Garboden et al. 2017; Porter & Vogel, 2014). One of the major pitfalls of propensity score matching is the unbalance in unobserved variables and confounders (Nuttall & Houle, 2008). In this study, it was not possible to include all possible covariates that may affect selection into changing residence or experiencing depressive symptoms. Life transitions and changes in employment are frequent contributors to whether families change residences (Coulter & Scott, 2015), but were not included in the current analysis. In addition to neighborhood cohesion, other mechanisms such as familial and peer social support, and neighborhood quality may further explain the associations between mobility and depressive symptoms (Coley et al., 2013; Hurd et al., 2013). The time between assessments, one year, may not have been optimal for observing effects of residential moves on proximate outcomes of neighborhood cohesion or depression. For example, the effects of residential change may be more evident in the short term as previous research on residential mobility suggests (Leventhal & Newman, 2010; Voight et al., 2012). The yearly assessments may have failed to capture additional moves that may have occurred that would not have been reflected in the singular change in reported address between data collection waves. Finally, self-reports of depression and neighborhood cohesion may be biased by factors such as negative cognitive styles and social desirability (Podsakoff et al., 2003).

Future Directions

Information about neighborhoods of origin and destination neighborhoods should be incorporated in assessment and evaluation of the residential mobility. The inclusion of different data sources including census level data such as racial composition, municipal data such as crime rates and occurrences, and commercial data such as housing prices can enrich the understanding of both the context in which moves originate and the context of destination environment. Previous work on neighborhoods, using the social disorganization framework, has used census data along with individual mechanisms (e.g., Kingston et al., 2009). Integrating neighborhood level and individual level variables could provide more comprehensive models of how residential mobility is associated with adolescent depressive symptoms. Factors related to the experience of changing neighborhoods should be included in future models of residential mobility effects. Along with disruption of social ties in the neighborhood, residential moves may be associated with change in schools and peer networks and differences in availability and accessibility of social services that may affect adolescents’ outcomes (Leventhal & Newman, 2010). Schools may play a role in the risk for depressive symptoms (Dunn et al., 2015); thus, inclusion of possible changes in schools may provide evidence for other possible mechanisms linking moving and depression. Because mobility may be linked with problem outcomes that co-occur in adolescence (e.g., delinquency [Porter & Vogel, 2014] and poor academic performance [Voight et al., 2012]), future investigations should include multiple outcomes rather than a single outcome, as in the current study. Because prior research has indicated some effects of residential moves occur within a short time frame after the event (Leventhal & Newman, 2010), longitudinal studies with shorter time periods between data collection waves should be implemented to examine more proximal short-term effects. Future studies with longitudinal data collection waves over short time periods can be designed to explore both proximate and long-term effects of moving on outcomes.

Acknowledgments

This research was supported by grants from the National Institute on Drug Abuse (DA11796 to Ialongo) and the National Institute of Mental Health (MH057005 to Ialongo; MH078995 to Lambert). We thank the youth, parents, and teachers who made this work possible.

Footnotes

The authors declare no conflicts of interest.

References

  1. Aber MS, & Nieto M. (2000). Suggestions for the investigation of psychological wellness in the neighborhood context: Toward a pluarlistic neighborhood theory. In Cicchetti D, Rappaport J, Sandler I, & Weissberg R. (Eds.), The promotion of wellness in children and adolescents (pp. 185–220). Washington, DC: CWLA Press. [Google Scholar]
  2. Adam EK, & Chase-Lansdale PL (2002). Home sweet home(s): Parental separations, residential moves, and adjustment problems in low-income adolescent girls. Developmental Psychology, 38(5), 792–805. 10.1037/0012-1649.38.5.792 [DOI] [PubMed] [Google Scholar]
  3. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. [Google Scholar]
  4. Anderson S, & Leventhal T. (2016). Residential mobility and adolescent achievement and behavior: Understanding timing and extent of mobility. Journal of Research on Adolescence, 27(2), 328–343. 10.1111/jora.12288 [DOI] [PubMed] [Google Scholar]
  5. Anderson S, Leventhal T, & Dupéré V. (2014). Residential mobility and the family context: A developmental approach. Journal of Applied Developmental Psychology, 35, 70–78. 10.1016/j.appdev.2013.11.004 [DOI] [Google Scholar]
  6. Anderson S, Leventhal T, Newman S, & Dupéré V. (2014). Residential mobility among children: A framework for child and family policy. Cityscape: A Journal of Policy Development and Research, 16(1), 5–36. Retrieved April 9, 2021, from http://www.jstor.org/stable/26326856 [Google Scholar]
  7. Arcaya MC, Tucker-Seeley RD, Kim R, Schnake-Mahl A, So M, & Subramanian SV (2016). Research on neighborhood effects on health in the United States: A systematic review of study characteristics. Social Science & Medicine, 168, 16–29. 10.1016/j.socscimed.2016.08.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Asarnow JR, & Carlson GA (1985). Depression self-rating scale: Utility with child psychiatric inpatients. Journal of Consulting and Clinical Psychology, 53, 491–499. 10.1037/0022-006X.53.4.491 [DOI] [PubMed] [Google Scholar]
  9. Austin PC (2011). An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behavioral Research, 46, 399–424. 10.1080/00273171.2011.568786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bass JK, & Lambert SF (2004). Urban adolescents’ perceptions of their neighborhood: An examination of spatial dependence. Journal of Community Psychology, 32(3), 277–293. 10.1002/jcop.20005 [DOI] [Google Scholar]
  11. Bradshaw CP, Sudhinaraset M, Mmari K, & Blum RW (2010) School transitions Among military adolescents: A qualitative study of stress and coping. School Psychology Review, 39(1), 84–105. 10.1080/02796015.2010.12087792 [DOI] [Google Scholar]
  12. Brown AC, & Orthner DK (1990). Relocation and Personal Well-Being among Early Adolescents. The Journal of Early Adolescence, 10(3), 366–381. 10.1177/0272431690103008 [DOI] [Google Scholar]
  13. Bruch E, & Swait J. (2019). Choice set formation in residential mobility and its implication for segregation dynamics. Demography, 56, 1665–1692. 10.1007/s13524-019-00810-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Buckner J. (1988). The development of an instrument to measure neighborhood cohesion. American Journal of Community Psychology, 16, 771–791. 10.1007/BF00930892 [DOI] [Google Scholar]
  15. Cicchetti D, & Toth SL (1998). The development of depression in children and adolescents. American Psychologist, 53(2), 221–241. 10.1037/0003-066X.53.2.221 [DOI] [PubMed] [Google Scholar]
  16. Colburn S, Pratt M, Mueller C, & Tompett CJ (2019). How adolescents define their home neighborhoods conceptually and spatially. Journal of Community Psychology, 48, 709–725. 10.1002/jcop.22290 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Coley RL, Leventhal T, Lynch AD, & Kull M. (2013). Relations between housing characteristics and well-being of low-income children and adolescents. Developmental Psychology, 49(9), 1775–1789. 10.1037/a0031033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Costello DM, Swendsen J, Rose JS, & Dierker LC (2008). Risk and protective factors associated with trajectories of depressed mood from adolescence to early adulthood. Journal of Consulting and Clinical Psychology, 76(2), 173–183. 10.1037/0022-006X.76.2.173 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Coulter R, & Scott J. (2015). What motivates residential mobility? Reexamining self-reported reasons for desiring and making residential moves. Population, Space and Place, 21, 354–371. 10.1002/psp.1863 [DOI] [Google Scholar]
  20. DeLuca S, Wood H, & Rosenblatt P. (2019). Why poor families move (and there they go): Reactive mobility and residential decisions. City and Community, 18(2), 556–593. 10.1111/cico.12386 [DOI] [Google Scholar]
  21. Diez Roux AV (2001). Investigating neighborhood and area effects on health. American Journal of Public Health, 91, 1783–1789. 10.2105/AJPH.91.11.1783 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Dunn EC, Milliren CE, Evans CR, Subramanian SV, & Richmond TK (2015). Disentangling the relative influence of schools and neighborhoods on adolescents’ risk for depressive symptoms. American Journal of Public Health, 105(4), 732–740. 10.2105/AJPH.2014.302374 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Elliot DS, Huizinga D, & Ageton SS (1985). Explaining delinquency and drug use. Beverly Hills, CA: Sage. [Google Scholar]
  24. Fauth RC, Leventhal T, & Brooks-Gunn J. (2004). Short-term effects of moving from public housing in poor to middle-class neighborhoods on low-income, minority adults’ outcomes. Social Science & Medicine, 59(11), 2271–2284. 10.1016/j.socscimed.2004.03.020 [DOI] [PubMed] [Google Scholar]
  25. Fritz MS, & MacKinnon DP (2007). Required sample size to detect the mediated effect. Psycholoical Science, 18(3), 233–239. 10.1111/j.1467-9280.2007.01882. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Garboden PME, Leventhal T, & Newman S. (2017). Estimated the effects of residential mobility: A methodological note. Journal of Social Service Research, 43(2), 246–261. 10.1080/01488376.2017.1282392 [DOI] [Google Scholar]
  27. Goldman-Mellor S, Margerison-Zilko C, Allen K, & Cerda M. (2016). Perceived and objectively-measured neighborhood violence and adolescent psychological distress. Journal of Urban Health, 93(5), 758–769. 10.1007/s11524-016-0079-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hayes AF (2017). Introduction to mediation, moderation, and conditional process analysis. New York, NY: Guilford Press. [Google Scholar]
  29. Hayes AF & Cai L. (2007). Using heteroskedastic-consistent standard error estimators in OLS regression: An introduction and software implementation. Behavior Research Methods, 39(4), 709–722. 10.3758/BF03192961 [DOI] [PubMed] [Google Scholar]
  30. Hepburn P, Louis R, Fish J, Lemmerman E, Alexander AK, Thomas TA, Koehler R, Benfer E, & Desmond M. (2021). U.S. eviction filing patterns in 2020. Socius. 10.1177/23780231211009983 [DOI] [Google Scholar]
  31. Hurd NM, Stoddard SA, & Zimmerman MA (2013). Neighborhoods, social support, and African American adolescents’ mental health outcomes: A multilevel path analysis. Child Development, 84(3), 858–874. 10.1111/cdev.12018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Ialongo NS, Kellam SG, & Poduska J. (1999a). Manual for the Baltimore how I feel (Tech. Rep. No. 2). Baltimore, MD: Johns Hopkins University. [Google Scholar]
  33. Ialongo NS, Werthamer L, Kellam SG, Brown CH, Wang S, & Lin Y. (1999b). Proximal impact of two-first-grade preventive interventions on the early risk behaviors for later substance abuse, depression, and antisocial behavior. American Journal of Community Psychology, 27, 599–641. 10.1023/A:1022137920532 [DOI] [PubMed] [Google Scholar]
  34. Inagami S, Cohen DA, & Finch BK (2007). Non-residential neighborhood exposures suppress neighborhood effects on self-rated health. Social Science and Medicine, 65, 1779–1791. 10.1016/j.socscimed.2007.05.051 [DOI] [PubMed] [Google Scholar]
  35. Jelleyman T, & Spencer N. (2008). Residential mobility in childhood and health outcomes: a systematic review. Journal of Epidemiology & Community Health, 62(7), 584–592. 10.1136/jech.2007.060103 [DOI] [PubMed] [Google Scholar]
  36. Jones A, & Dantzler P. (2021). Neighbourhood perceptions and residential mobilty. Urban Studies, 58(9), 1792–1810. 10.1177/0042098020916440 [DOI] [Google Scholar]
  37. Kazdin AE, Rodgers A, & Colbus D. (1986). The hopelessness scale for children: Psychometric characteristics and concurrent validity. Journal of Consulting and Clinical Psychology, 54, 242–245. 10.1037/0022-006X.54.2.241 [DOI] [PubMed] [Google Scholar]
  38. Kingsbury M, Kirkbride JB, McMartin SE, Wickham ME, Weeks M, & Colman I. (2015). Trajectories of childhood neighborhood cohesion and adolescent mental health: Evidence from a national Canadian cohort. Psychological Medicine, 45(14), 3239–3248. 10.1017/S0033291715001245 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Kingston B, Huizinga D, & Elliot DS (2009). A test of social disorganization theory in high-risk urban neighborhoods. Youth & Society, 41(1), 53–79. 10.1177/0044118X09338343 [DOI] [Google Scholar]
  40. Kwan M. (2012). The uncertain geographic context problem. Annals of the Associations of American Geographers, 102(5), 958–968. 10.1080/00045608.2012.687349 [DOI] [Google Scholar]
  41. Lawrence E, Root ED, & Mollborn S. (2015). Residential Mobility in Early Childhood: Household and Neighborhood Characteristics of Movers and Non-Movers. Demographic Research, 33, 939–950. 10.4054/demres.2015.33.32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Lechuga-Peña S, & Brisson D. (2018). Your Family, Your Neighborhood: Results from a feasibility and acceptability study of parent engagement in subsidized project-based housing. Journal of Community Practice, 17(6), 696–713. 10.1080/10705422.2018.1521353 [DOI] [Google Scholar]
  43. Leventhal T, & Brooks-Gunn J. (2000). The neighborhoods they live in: The effects of neighborhood residence on child and adolescent outcomes. Psychological Bulletin, 126(2), 309–337. 10.1037/0033-2909.126.2.309 [DOI] [PubMed] [Google Scholar]
  44. Leventhal T, & Newman S. (2010). Housing and child development. Children and Youth Services Review, 32, 1165–1174. 10.1016/j.childyouth.2010.03.008 [DOI] [Google Scholar]
  45. Magdol L, & Bessel DR (2003). Social capital, social currency, and portable assets: The impact of residential mobility on exchanges of social support. Personal Relationships, 10, 149–169. 10.1111/1475-6811.00043 [DOI] [Google Scholar]
  46. Nuttall GA, & Houle TT (2008). Liars, damn liars, and propsensity scores, Anesthesiology, 108(3–4), 10.1097/01.anes.0000296718.35703.20 [DOI] [PubMed] [Google Scholar]
  47. Owens A. (2019). Building inequaility: Housing segregation and income segregation. Sociological Science, 6, 497–525. 10.15195/v6.a19 [DOI] [Google Scholar]
  48. Perkins KL (2017). Reconsidering residential mobilty: Differential effects on child wellbeing by race and ethnicity, Social Science Research, 63, 124–137. 10.1016/j.ssresearch.2016.09.024 [DOI] [PubMed] [Google Scholar]
  49. Plunkett SW, Abarca-Mortensen S, Behnke AO, & Sands T. (2007). Neighborhood structural qualities, adolescents’ perceptions of neighborhoods, and Latino youth development. Hispanic Journal of Behavioral Sciences, 29(1), 19–34. 10.1177/0739986306295038 [DOI] [Google Scholar]
  50. Podsakoff PM, MacKenzie SB, Lee J, & Podsakoff NP (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. 10.1037/0021-9010.88.5.879 [DOI] [PubMed] [Google Scholar]
  51. Porter L, & Vogel M. (2014). Residential mobility and delinquency revisited: Causation or selection? Journal of Quantitative Criminology, 30, 187–214. 10.1007/s10940-013-9200-7 [DOI] [Google Scholar]
  52. Pribesh S, & Downey DB (1999). Why are residential and school moves associated with poor school performance? Demography, 36(4), 521–534. 10.2307/2648088 [DOI] [PubMed] [Google Scholar]
  53. QGIS Development Team. (2016). QGIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org [Google Scholar]
  54. Quillian L, Lee JJ, & Honoré B. (2020). Racial discrimination in the U. S. housing and mortage lending markets: A quantitative review of trends, 1976–2016. Race and Social Problems, 12, 13–28. 10.1007/s12552-019-09276-x [DOI] [Google Scholar]
  55. Rosenbaum PR, & Rubin DB (1983a). The central role of the propensity score in observational studies for causal effects. Biometrika, 70, 41–55. 10.1093/biomet/70.1.41 [DOI] [Google Scholar]
  56. Rueger SY, Malecki CK, Pyun Y, Aycock C, & Coyle S. (2016). A meta-analytic review of the association between perceived social support and depression in childhood and adolescence. Psychological Bulletin, 142(10), 1017–1067. 10.1037/bul0000058 [DOI] [PubMed] [Google Scholar]
  57. Rumbold AR, Giles LC, Whitrow MJ, Steele EJ, Davies CE, Davies MJ, & Moore VM (2012). The effects of house moves during early childhood on child mental health at age 9 years. BMC Public Health, 12(583). 10.1186/1471-2458-12-583 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Sampson RJ, & Groves WB (1989). Community structure and crime: Testing social disorganization theory. American Journal of Sociology, 94, 774–802. 10.1086/229068 [DOI] [Google Scholar]
  59. Sampson RJ, Raudenbush SW, and Earls F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 227(5328, 918–924. 10.1126/science.277.5328.918 [DOI] [PubMed] [Google Scholar]
  60. Shen C, Wan A, Kwok LT, Pang S, Wang X, Stewart SM, Lam TH, & Chan SS (2017). A community based intervention program at enhance neighborhood cohesion: The Learning Families Project in Hong Kong. PloS one. 10.1371/journal.pone.0182722 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Smith EP, Faulk M, & Sizer MA (2016). Exploring the meso-system: The roles of community, family, and peers in adolescent delinquency and positive youth development. Youth & Society, 48(3), 318–343. 10.1177/0044118X13491581 [DOI] [Google Scholar]
  62. Sun IY, Triplett R, & Gainey RR (2004). Neighborhood characteristics and crime: A test of Sampson and Groves’ model of social disorganization. Western Criminology Review, 5(1), 1–16. [Google Scholar]
  63. United States Census Bureau. (2018). Geographic mobility: 2016–2017. Retrieved January 7. 2019, from https://www.census.gov/data/tables/2017/demo/geographic-mobility/cps-2017.html
  64. Voight A, Shinn M, & Nation M. (2012). The longitudinal effects of residential mobility on the academic achievement of urban elementary and middle school students. Educational Researcher, 41(9), 385–392. 10.3102/0013189X12442239 [DOI] [Google Scholar]
  65. Witherspoon D, Ennett S. (2011).An examination of social disorganization and pluralistic neighborhood theories with rural mothers and their adolescents. Journal of Youth and Adolescence, 40, 1243–1253. 10.1007/s10964-009-9499-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Witherspoon DP, & Hughes DL (2014). Early adolescent perceptions of neighborhood: Strengths, structural disadvantage, and relations to outcomes. The Journal of Early Adolescence, 34(7), 866–895. 10.1177/0272431613510404 [DOI] [Google Scholar]
  67. Wolff KT, Baglivio MT, Intravia J, Greenwald MA, & Epps N. (2017). The mobility of youth in the justice system: Implications for recidivism. Journal of Youth and Adolescence, 46, 1371–1393. 10.1007/s10964-016-0498-y [DOI] [PubMed] [Google Scholar]
  68. Xue Y, Leventhal T, Brooks-Gunn J, & Earls FJ (2005). Neighborhood residence and mental health problems of 5- to 11-year-olds. Archives General Psychiatry, 62, 554–563. 10.1001/archpsyc.62.5.554 [DOI] [PubMed] [Google Scholar]

RESOURCES