The protective effect of neighborhood social cohesion in child abuse and neglect

Child Abuse & Neglect 52 (2016) 29–37

Contents lists available at ScienceDirect

Child Abuse & Neglect

Research article

The protective effect of neighborhood social cohesion in child abuse and neglect Kathryn Maguire-Jack a,∗ , Kathryn Showalter b a b

325B Stillman Hall, 1947N College Rd Columbus, OH 43210, USA 1947N College Rd, Columbus, OH 43210, USA

a r t i c l e

i n f o

Article history: Received 25 September 2015 Received in revised form 18 December 2015 Accepted 24 December 2015 Keywords: Child abuse Child neglect Neighborhoods Protective factors Social cohesion

a b s t r a c t Relations between parents within a neighborhood have the potential to provide a supportive environment for healthy and positive parenting. Neighborhood social cohesion, or the mutual trust and support among neighbors, is one process through which parenting may be improved. The current study investigates the association between neighborhood social cohesion and abuse and neglect, as well as specific types of abuse and neglect. The sample for the study is comprised of 896 parents in one urban Midwestern County in the United States. Participants were recruited from Women, Infants, and Children clinics. Negative binomial regression is used to examine the association between neighborhood social cohesion and child maltreatment behaviors, as measured by the Conflict Tactics Scale, Parent-to-Child Version (Straus et al., 1998). In this sample of families, neighborhood social cohesion is associated with child neglect, but not abuse. In examining the relationship with specific types of abuse and neglect, it was found that neighborhood social cohesion may have a protective role in some acts of neglect, such as meeting a child’s basic needs, but not potentially more complex needs like parental substance abuse. © 2016 Elsevier Ltd. All rights reserved.

Introduction Child maltreatment or the physical, sexual, and emotional abuse and neglect of children is reported to state and local investigators in the United States at a rate of 6 referrals per minute or 3 million referrals annually (U.S. Department of Health and Human Services, 2012). Certain neighborhood characteristics, including high rates of poverty, instability, and childcare burden, have been found to be associated with higher levels of maltreatment (Maguire-Jack, 2014). Although neighborhoods may put families at risk for maltreatment, strong bonds with neighbors may serve as a protective factor against such behaviors. The current study examines the potentially protective role of support between neighbors in preventing child maltreatment. Types of Maltreatment Child maltreatment is broadly defined as acts of commission (abuse) or omission (neglect) by a caregiver that results in harm, potential for harm, or threat of harm to a child (Leeb, Paulozzi, Melanson, Simon, & Arias, 2008). In recent years, maltreatment research has shifted from focusing on individuals to different levels of the social ecology, such as the ways

∗ Corresponding author. http://dx.doi.org/10.1016/j.chiabu.2015.12.011 0145-2134/© 2016 Elsevier Ltd. All rights reserved.

30

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

in which neighborhoods influence families. In examining neighborhood risks for maltreatment, several studies have found poverty to be strongly related to neglect (Drake & Pandey, 1996; Freisthler, Midanik, & Gruenewald, 2004; Garbarino & Crouter, 1978; Kim, 2004). Similarly, a random sample of parents at risk for maltreatment showed poverty to be related to abuse (Coulton, Korbin, & Su, 1999). Although there are similarities between the two types of maltreatment, this study focuses on the differences in parental behavior types including subtypes of neglect and abuse that are largely missing from the current literature. The behaviors that are encompassed by child abuse and neglect are very different and warrant individual exploration of the literature to fully understand their etiology. In a meta-analysis of the risks for child abuse and neglect, it was found that physical abuse was strongly associated with parent anger/hyper-reactivity, family conflict, and family cohesion (Stith et al., 2009). However, in the same analysis, neglect was associated with separate risk factors like parent–child relationship, perception of the child as a problem, parent’s level of stress, and parent self-esteem (Stith et al., 2009). Further, Black and colleagues (2001) found that caregivers who engage in physical abuse are emotionally reactive and distressed, avoid focused problem-solving, feel incompetent as parents, and have low levels of social support. While it seems physical abuse occurs in isolated incidents, neglect is more closely associated with risk factors that are prolonged, like caregiving deficits, poor prenatal and postnatal care, and homelessness (Hildyard & Wolfe, 2002). It is likely that the two types of maltreatment occur under different conditions but the circumstances of abuse and neglect subtypes warrant further investigation. Researchers have found some evidence of relationships between caregivers’ social relationships and maltreatment. One recent study examined the risk factors for three neglectful acts within Latino families: (1) not enough food for a child, (2) inability to take a child to the doctor due to cost, and (3) leaving a child in a questionably safe place (Johnson-Motoyama, 2013). This study found differences across these types of neglectful behaviors, including that smaller social networks are associated with a greater risk of leaving a child in an unsafe place, but not the other types of neglect. Other research compliments this in that neglectful parenting is found to be highest among individuals who report low social support from relatives, friends, and neighbors (Gauda, Polansky, Kilpatrick, & Shilton, 1993). Geographic regions are one organizing structure for social networks, and therefore the influence that neighbors have on one another could determine the presence of child maltreatment in families. To further the conversation of maltreatment subtypes, the current study examined subtypes of abuse and neglect and their relation to neighborhood social cohesion. For physical abuse, we considered corporal punishment and severe assault. Prior research has found that corporal punishment is more common among parents who are younger, less educated, single, have lower income, and those who have higher levels of stress and depression (Day, Peterson, & McCracken, 1998; Pinderhughes, Dodge, Bates, Pettit, & Zelli, 2000; Regalado, Sareen, Inkelas, Wissow, & Halfon, 2004; Straus & Stewart, 1999). In addition to these risk factors, more severe physical abuse is correlated with alcohol and drug abuse problems, domestic violence, having experienced abuse as a child, having conflict with the child or with the spouse, and child behavioral problems (Stith et al., 2009). For neglect, we considered basic needs (supervision, food, medical attention) and failure to provide adequate care due to mental health or substance abuse problems. There is general agreement in the field that child neglect and poverty are strongly correlated (Jonson-Reid, Drake, & Zhou, 2012). Basic needs neglect is more directly related to poverty, as parents with no intention of putting their child in risk of harm who are struggling to make ends meet would still be more likely to be unable to meet these basic needs. On the other hand, neglect due to caregiver mental health or substance abuse problems may differ from basic needs, in that such concerns may be more complex and therefore more difficult for neighbors to address.

Neighborhood Context of Maltreatment Children from families who experience substance abuse, depression, unemployment, and poverty are more likely to experience maltreatment (Stith et al., 2009). The ecological model suggests that factors exist at different levels of the social ecology including the individual, relationship, community, and societal level that influence child and family relations like child maltreatment (Belsky, 1980; Bronfenbrenner, 1979). Although much of the literature has focused on individual and family increased risk for maltreatment (Stith et al., 2009) there is a significant body of work dedicated to understanding influences at higher levels of the social ecology (see Coulton, Crampton, Irwin, Spilsbury, & Korbin, 2007; Freisthler, Merritt, & LaScala, 2006; Maguire-Jack, 2014 for reviews). Many studies have relied on social disorganization theory (Shaw & McKay, 1942) to provide a framework for understanding how neighborhoods might contribute to maltreatment (Ben-Arieh, 2010; Coulton et al. 1999; Ernst, 2001; Freisthler, Gruenewald, Ring, & LaScala, 2008; Fromm, 2004; Garbarino & Kostelny, 1992). The theory was first proposed to understand the relationship between community milieu and rates of juvenile delinquency (Shaw & McKay, 1942). As noted by Shaw and McKay (1942), “The communities with the highest rates of delinquents are occupied by those segments of the population whose position is most disadvantaged in relation to the distribution of economic, social, and cultural values” (p. 348). In other words, the theory proposes that the concentration of disadvantaged people in neighborhoods lends itself to criminal acts. Child maltreatment researchers have expanded this theory to understand neighborhood conditions for child abuse and neglect. Social disorganization theory applied to maltreatment suggests that distressed neighborhoods might put parents

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

31

at additional risk for maltreatment because of the multiple stressors they provide and the lack of social norms that provide a supportive environment for positive parenting (Sampson & Groves, 1989). The neighborhoods in which parents live have the ability to support or hinder positive parenting practices. Prior research on the community context of child maltreatment has found evidence of a relationship between maltreatment and rates of community poverty, unemployment, residential instability, childcare burden, alcohol and drug availability, and immigrant concentration (Coulton et al., 2007; Freisthler et al., 2006; Maguire-Jack, 2014). However, knowing that neighborhood risk factors can result in child maltreatment, we move to explore the way that neighbor social cohesion can, in contrast, protect against poor parenting practices. Neighborhood Social Cohesion Conceptually, it has been determined that neighborhood social cohesion (mutual trust among neighbors) and social control (norms regarding appropriate behaviors and willingness to intervene) together make up collective efficacy (Coulton et al., 1999; Guterman, Lee, Taylor, & Rathouz, 2009; Molnar, Buka, Brennan, Holton, & Earls, 2003; Sampson, Morenoff, & Gannon-Rowley, 2002). Studies of collective efficacy show that psychological aggression and physical assault toward children is higher among residents who have low perceptions of collective efficacy (Guterman et al., 2009). In the current study, we examine positive neighborhood relations in the form of neighborhood social cohesion or as defined by Kawachi and Berkman (2000), the “connectedness and solidarity among groups in society” or the two intertwined features of (1) the absence of latent social conflict. . .. and (2) the presence of strong social bonds – measured by levels of trust and norms of reciprocity” (p. 175). In our continuation of the neighborhood support conversation, we focus on one aspect of collective efficacy, neighborhood social cohesion, which relates to common bonds, trust, and feelings of support between neighbors. Neighborhood social cohesion might help parents avoid maltreating behaviors because of the ability to rely on one’s neighbors to help meet their physical and emotional needs of both themselves and their children. In contrast, neighborhood social control focuses on the norms of a community and the willingness to intervene when such norms are being violated. As such, one might expect parents to avoid engaging in maltreating behaviors in neighborhoods with high levels of social control due to the fear of retaliation. In this study, we examine the relation between neighborhood social cohesion and child abuse and neglect to focus on this potentially supportive component of neighborhoods. The research has been somewhat mixed on the importance of neighborhood social cohesion. In some of the early neighborhood work, Garbarino and Sherman (1980) found that residents of neighborhoods with higher rates of maltreatment had lower levels of neighborhood social cohesion, with reports that their neighbors were less likely to assist with childcare and engage in neighborhood exchanges compared to residents in neighborhoods with lower rates of maltreatment. Similarly, Vinson, Bladry, and Hargreaves (1996) found that areas with fewer interactions between familial networks, acquaintances, and neighborhood networks had higher levels of abuse. In more recent work, Guterman and colleagues (2009) found that perceptions of collective efficacy and neighborhood disorder were directly related to physical assault and indirectly related to neglect (poorer perceptions of neighborhood collective efficacy and disorder were associated with higher levels of maltreatment); while other researchers did not find evidence of such a relationship (Coulton et al., 1999; Molnar et al., 2003). One account for mixed neighborhood cohesion results could be that neighbor relations impact types and severity levels of maltreatment in different ways, a phenomenon we expect to find in the current study. Contributions of the Current Study and Research Aims Although a small body of work exists examining the relationship between neighborhood social cohesion and child maltreatment, to the authors’ knowledge, there are no studies investigating the relationship between neighborhood social cohesion and specific types of abusive and neglectful behavior. Relying on social disorganization theory, we hypothesize: (1) neighborhood social cohesion will be associated with lower levels of physical abuse; and (2) neighborhood social cohesion will be associated with lower levels of neglect. We also examine two additional research aims that are exploratory in nature and therefore do not offer a priori hypotheses: (1) To what extent is neighborhood social cohesion associated with lower levels of basic needs neglect and neglect due to a caregiver mental health or substance use problem; and (2) To what extent is neighborhood social cohesion associated with lower levels of corporal punishment versus severe assault? We are particularly interested in examining these different sub-types of neglect because of the wide variations of behaviors that fall under the category of “neglect.” We believe that parents may be more likely to reach out to their neighborhood networks for some subtypes of neglect than others. For example, the assistance that can be provided by a neighbor to help out when money is short and parents are unable to provide food for the family is less time intensive than preventing a parent from abusing drugs or alcohol. We are also interested in examining the severity of abusive behaviors. Parents who have positive and trusting relationships with their neighbors may reach out to ask for assistance in discipline techniques, but may be less likely to feel comfortable asking about their more severe abusive behaviors.

32

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

Methodology Sample and Procedure The data come from a study titled the Franklin County Neighborhood Services Study. A total of 1,326 parents across Franklin County, Ohio completed a survey related to the supports within their neighborhood and their individual risks for and use of child maltreatment behaviors. The full sample includes 1,053 parents from Women, Infants, and Children (WIC) clinics and 279 from four licensed childcare centers in Franklin County who participated in the survey. This convenience sample was selected to maximize geography across Franklin County, with WIC clinics selected in all regions of the County. The clinics were selected first and then all clinic supervisors who were approached agreed to be in the study. For the current study, we limit the sample to 896 respondents recruited from WIC clinics who had complete information on all study variables. WIC is a public benefit available to pregnant women and children age five and under with incomes under 185% of the federal poverty line. As a result, this convenience sample has lower incomes than the general population. To be eligible to complete the survey, respondents had to be at least 18 years of age, live in Franklin County, and serve in a parenting role to at least one child living in their home. All participants received an incentive in the form of a $10 gift card and were also given the opportunity to enter to win one of three $500 gift cards. The survey included questions regarding parenting stress, social support, neighborhood social cohesion and social control, mental health concerns, social service availability, economic hardship, and parenting behaviors including child maltreatment. To ensure confidentiality, the participants did not provide their name or contact information anywhere on the survey. Participants took the survey in paper-and-pencil format in the clinic waiting room, while they were waiting for their appointments. They were offered the opportunity to have the survey read to them in a private conference room if needed, but no participants selected this option. After the survey was completed, the participant put it into a sealed envelope, so that the survey proctors could not see participant responses. Participants who chose to partake in the raffle for a $500 gift card wrote their contact information on a 3 × 5 note card that was placed into a separate envelope with all other participants. The notecards could not be linked to the actual survey responses. The surveys were conducted during varying hours (for eight hours at a time) and on varying days at the different clinics, to maximize the variation of participants. At the end of each day, the surveys were transported to the University office and were locked in a filing cabinet in the researcher’s campus office, which was also locked. At the conclusion of data collection, all surveys were scanned into electronic format using scanning software. A member of the research team preformed quality assurance checks on all surveys. The response rate was very high, with approximately 94% of participants taking the survey. Aspects of the research design that we believe contributed to this high response rate were: (1) participants were an “active audience” in that they had to sit and wait for their WIC appointments anyway, so it was not as much of an imposition on their time; (2) the survey was relatively short (10 min) and they immediately received their compensation for participating; and (3) members of the research team were available to help with entertaining the participants’ children while the survey was completed. Measures Key Dependent Variables. Child neglect and physical child abuse were assessed using the Conflict Tactics Scale – Parent-toChild version (CTS-PC, Straus, Hamby, Finkelhor, Moore, & Runyan, 1998). This scale contains four subscales: physical assault, psychological aggression, neglect, and nonviolent discipline. Each subscale contains the following response options: “never,” “not in the past year, but it has happened,” “1 time,” “2 times,” “3–5 times,” “6–10 times,” “10–20 times,” and “more than 20 times.” The responses were recoded as recommended by Straus and colleagues (1998) using the midpoint of the ranges and a value of 25 for responses of “more than 20.” The value of 25 was selected for responses over 20 per the recommendation of the scale developer (Straus et al., 1998). Responses of “not in the past year but it happened” were recoded as 0. In the current study, we focus only on the physical abuse and neglect subscales. Examples of physical abuse questions include: “In the past year, how often have you hit him/her on the bottom with a belt, stick, or other hard object?” and “In the past year, how often have you hit him/her with a fist or kicked him/her?” Examples of neglect questions include: “In the past year, how often have you had to leave the child alone when you felt like some other adult should be with him/her?” and “In the past year, how often have you been so drunk or high that you had trouble caring for the child?” Cronbach’s alpha for the neglect scale in our sample was 0.67 and for the physical abuse scale was 0.80. We also investigate the individual neglect and abuse items to investigate the relationship between neighborhood social cohesion and different types of neglect and abuse. For neglect, we examine neglect related to basic needs and neglect related to caregiver mental health or substance use concerns. The basic needs items include: (1) leaving a child home alone when the caregiver felt an adult should be with him/her, (2) inability to provide food for the child, and (3) inability to take the child to the doctor for needed care. The neglect related to caregiver mental health or substance use problems include: (1) inability to show love because the caregiver was too caught up with own problems and (2) inability to provide appropriate care due to drugs or alcohol. For abuse, we examine corporal punishment items (e.g. spanking on the bottom) separately from severe assault items (e.g. burn child on purpose). The corporal punishment subscale is the “Minor Assault subscale” (Straus & Hamby, 1997) while the severe assault scale is the “Severe Physical Assault Subscale” (Straus & Hamby, 1997). The reliability of the corporal punishment scale was low (alpha = 0.57), while the reliability of the severe assault scale was

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

33

better (alpha = 0.79). While the overall neglect and abuse scales were measured continuously using the sum of the midpoints, as recommended by Straus and colleagues (1998), the smaller subscales of the types of abuse and neglect were measured dichotomously, to indicate whether each subtype had ever occurred. Key Independent Variables. The key independent variable is perceived neighborhood cohesion. This scale comes from the Social Cohesion and Trust Scale (Sampson, Raudenbush, & Earls, 1997). It is made up of 5 items intended to assess trust and cohesion among neighbors. The questions are assessed on a 5-point scale including the following response options “Strongly Disagree,” “Somewhat Disagree,” “Neither Agree nor Disagree,” “Somewhat Agree,” and “Strongly Agree.” Example questions include: “People in my neighborhood are willing to help their neighbors” and “People in my neighborhood generally don’t get along with each other.” The responses to these questions are averaged to provide an overall neighborhood cohesion score. The Cronbach’s alpha for this scale was 0.77. The covariates in this study included parenting stress, mental health concerns, economic hardship, race, marital status, education level, and number of children residing in the parent’s home. Parenting stress (alpha = 0.78) was measured using 31 questions from the Parent Stress Index – Short Form (PSI-SF, Abidin, 1990, 1995) which were assessed on a 4-point scale (“Strongly Disagree,” “Disagree,” “Agree,” and “Strongly Agree), including questions like “I feel trapped by my responsibilities as a parent,” and “Since having children, I have been unable to do new and different things.” Economic hardship was used in lieu of an income measure, because all families within the sample were relatively low-income due to sampling from WIC clinics. Experiences of economic hardships provided a better measure of the extent to which their low income status affected their daily lives. It was measured as a count of the total number of hardships out of eight that the parent had experienced at least once in the past year; for example, “In the past 12 months, was your telephone service, electricity, heating, or gas ever disconnected because there wasn’t enough money to pay the bill?” In terms of demographic characteristics, we include a series of dichotomous variables denoting whether the respondent was black, Hispanic, or other; whether the respondent married, and whether the respondent had a high school education or less. Number of children was measured continuously. Analysis All analyses were performed using STATA Statistical Software release 12 (StataCorp, 2011). For the full abuse and neglect scales, we used negative binomial regression on all models because the child maltreatment variables were of a count nature and were overdispersed at the left tail of the distribution. We did not use Poisson regression because our data did not meet the Poisson requirement of equal mean and variance. For the abuse and neglect subtypes, we ran four logistic regression models predicting the prevalence of any instance of basic needs neglect, neglect due to caregiver mental health or substance use concerns, corporal punishment, and severe assault. We began by examining the relationship between neighborhood cohesion and physical child neglect and physical abuse (Models 1 and 2 respectively). In Models 3 and 4, we examine basic needs neglect and neglect due to caregiver mental health or substance abuse concerns to understand the potentially protective effect of neighborhood social cohesion on different types of neglect. Finally, in Models 5 and 6, we examine corporal punishment separately from severe assault, to determine whether there is an association between neighborhood social cohesion and these different behaviors. Results Table 1 provides the descriptive statistics for the sample. Physically abusive behavior had occurred about six times in the past year and the average number of times neglectful behaviors had occurred in the past year was about two times. In terms of the neglect subtypes, 14% of parents reported leaving a child unattended, not providing food, or not taking a child to a doctor when s/he was sick at least one time in the past year. Additionally, 11% of parents reported neglect due to mental health or substance abuse concerns at least one time in the past year. In terms of the abuse subtypes, 41% of parents reported using corporal punishment and 16% reported severely assaulting their child at least one time in the past year. In terms of our key independent variable, respondents reported a mean level of neighborhood cohesion of 2.99, on a 5-point scale, indicating a neutral view of cohesion on average (1 was lowest level of cohesion, 5 was highest, and 3 was neutral). Of the covariates, the average level of parenting stress, respondents reported an average score of 1.88 (on a 4-point scale), suggesting a relatively low level of parenting stress. On this scale, a “2” would have indicated that they “disagreed” that the various concerns listed were a problem for them. On the economic hardship scale, the average number of hardships experienced in the past year was 2.16 of a possible 8, indicating that respondents were, on average, experiencing about two significant hardships in the past year. Of the demographic characteristics, 41% of the sample was white, 50% was black, 5% was Hispanic, and 4% was another race; 26% was married, and 61% had a high school education or less, and the average number of children residing in the home was 2.36. Table 2 shows the results from our primary models of interest, neighborhood cohesion and neglect (Model 1) and neighborhood cohesion and physical abuse (Model 2). In Model 1, we find that neighborhood cohesion is associated with lower levels of neglect. We also find that parenting stress and being black was associated with higher levels. Model 2 shows that there is no association between neighborhood cohesion and physical abuse in our sample. Parenting stress, being black, and having a greater number of children are associated with higher levels of physical abuse.

34

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37 Table 1 Descriptive statistics. N = 896. Variable Dependent variables Physical abuse Any corporal punishment Any severe assault Neglect Any basic needs neglect Any neglect due to MHSA Key independent variable Neighborhood cohesion Covariates Parenting stress Economic hardship White Black Hispanic Other race Married More than high school education High school education or less Number of children

Mean or %

Std. deviation

Range

5.53 0.41 0.16 1.89 0.14 0.11

16.43

0–230 0–1 0–1 0–105 0–1 0–1

2.99

0.67

1–5

1.88 2.16 0.41 0.50 0.05 0.04 0.26 0.39 0.61 2.36

0.33 2.10

1–4 0–8 0–1 0–1 0–1 0–1 0–1 0–1 0–1 1–9

7.73

1.47

Note: MHSA = caregiver mental health or substance abuse concern. Table 2 Negative binomial regression results for neighborhood cohesion and child maltreatment. N = 896.

Neighborhood cohesion Parenting stress Economic hardship Black Hispanic Other race Married HS or less Number of children Model fit AIC BIC

Model 1: Neglect IRR (SE)

Model 2: Physical abuse IRR (SE)

0.65 (0.09)* 21.81 (9.92)* 1.02 (0.06) 2.42 (0.71)* 0.63 (0.39) 2.34 (1.54) 1.30 (0.40) 1.41 (0.38) 1.12 (0.12)

1.00 (0.10) 3.31 (0.83)* 1.07 (0.42) 1.99 (0.36)* 0.59 (0.24) 1.9 (0.48) 0.98 (0.17) 0.98 (0.17) 1.13 (0.11)*

2.013 −4234.60

4.15 −2323.08

* p < 0.05. Note: Incidence rate ratios reported.

Table 3 shows the results of our logistic regression analyses to examine the relationship between neighborhood cohesion and the subtypes of abuse and neglect. Neighborhood cohesion was associated with lower levels of basic needs neglect (Model 3), but not neglect due to caregiver mental health or substance abuse concerns (Model 4). There were no significant associations between neighborhood social cohesion and corporal punishment (Model 5) or severe assault (Model 6). Discussion The current study sought to elucidate the potentially protective role of neighborhood social cohesion and child maltreatment. Relying on social disorganization theory, we hypothesized that higher levels of perceived neighborhood social cohesion would be associated with lower levels of physical abuse and neglect. We found that neighborhood social cohesion was associated with lower levels of neglect, but not abuse. This finding provides some support for the theory in that residents who perceive their communities to be cohesive have lower levels of neglect. This finding differs somewhat from the finding of Guterman and colleagues (2009) that a negative perception of neighborhood processes (including social cohesion) was associated with a higher level of abuse (directly) and neglect (indirectly). However, there are important differences between the two studies. First, the samples between the two studies varied considerably in terms of geography – with the study by Guterman and colleagues (2009) using a sample of nearly 5,000 parents in 20 large, urban, U.S. cities, and our study using a sample of just over 1,000 parents in one county in Ohio. Additionally, we specifically examined neighborhood social cohesion, while Guterman and colleagues (2009) examined social cohesion, social control, and neighborhood disorder jointly. Therefore, the differential findings between the studies suggest that neighborhood social

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

35

Table 3 Logistic regression results for neighborhood cohesion and maltreatment subtypes. N = 896. Model 3: Basic needs neglect OR (SE) Neighborhood cohesion Parenting stress Economic hardship Black Hispanic Other race Married HS or less Number of children Model fit AIC BIC

0.75 (0.11)* 6.08 (1.80)* 1.11 (0.05)* 1.71 (0.40)* 1.75 (0.80) 1.85 (0.94) 1.05 (0.26) 1.41 (0.31) 1.13 (0.07) 0.75 −5371.35

Model 4: Neglect due to MHSA OR (SE)

Model 5: Corporal punishment OR (SE)

Model 6: Severe assault OR (SE)

0.87 (0.14) 7.80 (2.61)* 1.10 (0.06) 2.24 (0.62)* 2.37 (1.18) 2.11 (1.19) 1.66 (0.46) 1.22 (0.30) 1.09 (0.08)

1.07 (0.11) 1.65 (0.36)* 1.09 (0.04)* 1.25 (0.19) 0.99 (0.34) 0.65 (0.25) 1.14 (0.19) 0.61 (0.09)* 1.14 (0.05)*

0.90 (0.13) 5.24 (1.54)* 1.11 (0.05)* 3.19 (0.76)* 2.02 (0.97) 1.29 (0.76) 1.34 (0.32) 1.07 (0.21) 1.11 (0.07)

0.62 −5491.86

1.33 −4852.74

0.78 −5342.36

* p < 0.05. Note: MHSA = caregiver mental health or substance abuse concern. Note: Odds ratios reported.

control and disorder may be driving the associations with physical abuse, since those are not considered in the current study. In the further analysis of subtypes of abuse and neglect, we found that neighborhood social cohesion was related to lower levels of basic needs neglect (leaving a child unattended, providing adequate food, and taking the child to the doctor when s/he is sick) but not related to neglect resulting from a caregiver’s mental health or substance use problems. These findings coincide with the finding from Garbarino and Sherman (1980) that residents in neighborhoods with higher rates of maltreatment reported that their neighbors were less likely to assist with childcare. The findings suggest that parents in our sample are willing to rely on their neighbors for immediate favors – helping out with babysitting or emergency cash or food assistance, but are less able to rely on them for more complex problems, such as assisting with substance abuse or mental health issues. We did not find a significant relationship between the abuse variables including corporal punishment and severe assault behaviors and the neighborhood social cohesion measure. There was also one unexpected finding in our covariates that is worth noting. Lower levels of education were associated with lower levels of corporal punishment. One hypothesis for this unexpected finding is related to the relatively low frequency of low education in our overwhelmingly low-income sample. In our sample, only 15% respondents reported having less than a high school education, which is lower than one might expect given our study population. There may have been some social desirability bias in reporting education level among our respondents. On the other hand, the higher level of education may have been due to a unique sample, and this uniqueness could be contributing to the unexpected finding. Taken together, these findings suggest that there are protective factors associated with having more cohesive communities but that there is a limit to the reach of such strategies. While increasing connections between neighbors may decrease some neglectful behaviors, this strategy alone may not be effective in reducing more complex family problems or violent behaviors. Implementing programs or campaigns intended to increase connections between neighbors must be done in conjunction with other strategies that have been shown to be effective at targeting these more extreme forms of maltreatment. Limitations There are several important limitations that must be considered when interpreting the findings from this study. First, the study relied on parents within one large, urban county in Ohio. Neighborhood processes may differ significantly in more rural or suburban areas, where the interactions between neighbors are likely different quantitatively and qualitatively. Additionally, the sample was largely a low-income sample, by virtue of the survey being conducted in WIC clinics. The convenience nature of the sample as well as the low-income status of all participants hinder the generalizability of the results. Third, the study was cross-sectional in nature. We surveyed parents at one time point, and therefore are unable to show how changes in neighborhood social cohesion over time relates to changes in maltreatment behaviors. Fourth, the study relied on self-reported child maltreatment behaviors, through the use of the CTS-PC (Straus et al., 1998). The data are not the same as officially reported child maltreatment to child welfare agencies and may be subject to social desirability bias. Fourth, because this sample was comprised of low-income families by design, there was limited variation at the neighborhood level in structural aspects of the neighborhood, such as poverty, unemployment, single-headed households, etc., and we were therefore unable to account for these characteristics in the study. This study would have been strengthened with the inclusion of these neighborhood-level characteristics and a multi-level examination. Unfortunately, with this sample, the intraclass correlation coefficient was almost zero. Future studies must strategically sample at the neighborhood level to assure variation at this level.

36

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

Finally, it is important to note that perceptions of the supportiveness of neighbors is likely directly influenced by characteristics of the respondent, as this measure is subjective. Related to this issue, some studies have asked maltreating and nonmaltreating parents within the same neighborhood about their perceptions of the availability of social support within the neighborhood and found that the maltreating parents have much more negative perceptions about social support in their neighborhood (Gracia & Musitu, 2003; Polansky, Gaudin, Ammons, & Davis, 1985). As a result, some researchers have concluded that social support is purely a factor of the parent’s psychology (Polansky et al., 1985). The difference in perception could be related to maltreating parents being rejected by their neighbors or something endogenous to the parent. Implications Despite these limitations, this study has important implications for researchers, practitioners, and policy makers. As previously stated, the existing research on this topic is extremely limited making preventive practices for subtypes of abuse and neglect difficult to determine. The finding that perceptions of neighborhood social cohesion are differentially related to abuse versus neglect, and also subtypes within neglect, suggest an additional line of inquiry be conducted to understand the role of neighborhood context on specific parenting behaviors. By understanding community relations and their impact on individual’s behaviors, specific interventions to improve parenting practices can be created. Future research should examine the structural characteristics of neighborhoods as well as other neighborhood processes (e.g. social control) and their role in preventing specific subtypes of abuse and neglect. Additionally, future research should be longitudinal to measure changes in neighborhood characteristics, and processes should also include more geographically diverse populations so that the findings will be more generalizable. Knowing that trust and belongingness can prevent neglectful behaviors, service providers and policy makers should support programming that builds neighborhood social cohesion. Additional research is needed to understand what types of interventions and practices can improve social cohesion, we do offer some suggestions. Prevention programs that incorporate neighborhood relations into their services by utilizing parenting meetings where parents can connect with members of their community, and parent mentor programs, in which parents can learn by example from peers, may have a significant impact on child neglect. While providing funding to programs that specifically target social cohesion can make families feel a since of belongingness to their neighborhood, we also recognize the need for systematic change by policy makers to prevent child maltreatment. Parents who are isolated from informal social support structures like friends and neighbors are likely also isolated from formal social support structures like community programs and state and federally funded organizations. In order to avoid this isolation, policy makers should work to make such resources equally accessible to neighborhoods either by way of public transportation or creating and maintaining centrally located social welfare agencies. Acknowledgement The authors would like to thank Leah Gjertson for thoughtful feedback on this project. The research was funded by an internal grant from The Ohio State University, College of Social Work. References Abidin, R. R. (1990). Parenting Stress Index. Charlottesville, VA: Pediatric Psychology Press. Abidin, R. R. (1995). Parenting Stress Index. Professional manual (3rd ed.). Odessa, FL: Psychological Assessment Resources. Belsky, J. (1980). Child maltreatment: An ecological integration. American Psychologist, 35(4), 320–335. Ben-Arieh, A. (2010). Localities, social services and child abuse: The role of community characteristics in social services allocation and child abuse reporting. Children and Youth Services Review, 32(4), 536–543. Black, D. A., Heyman, R. E., & Smith Slep, A. M. (2001). Risk factors for child physical abuse. Aggression and Violent Behavior, 6(2–3), 121–188. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press. Coulton, C. J., Crampton, D. S., Irwin, M., Spilsbury, J. C., & Korbin, J. E. (2007). How neighborhoods influence child maltreatment: A review of the literature and alternative pathways. Child Abuse & Neglect, 31, 1117–1142. Coulton, C., Korbin, J., & Su, M. (1999). Neighborhoods and child maltreatment: A multi-level study. Child Abuse & Neglect, 23(11), 1019–1040. Day, R. D., Peterson, G. W., & McCracken, C. (1998). Predicting spanking of younger and older children by mothers and fathers. Journal of Marriage and the Family, 60, 79–94. Drake, B., & Pandey, S. (1996). Understanding the relationship between neighborhood poverty and specific types of child maltreatment. Child Abuse & Neglect, 20(11), 1003–1018. http://dx.doi.org/10.1016/0145-2134(96)00091-9 Ernst, J. S. (2001). Community-level factors and child maltreatment in a suburban county. Social Work Research, 25(3), 133–142. Freisthler, B., Gruenewald, P. J., Ring, L., & LaScala, E. A. (2008). An ecological assessment of the population and environmental correlates of childhood accident, assault, and child abuse injuries. Alcoholism: Clinical and Experimental Research, 32(11), 1969–1975. Freisthler, B., Merritt, D. H., & LaScala, E. A. (2006). Understanding the ecology of child maltreatment: A review of the literature and directions for future research. Child Maltreatment, 11(3), 263–280. Freisthler, B., Midanik, L. T., & Gruenewald, P. J. (2004). Alcohol outlets and child physical abuse and neglect: Applying routine activities theory to the study of child maltreatment. Journal of Studies on Alcohol, 65(5), 586–592. Fromm, S. (2004). The processes that moderate the effect of community structural factors on neighborhood child maltreatment rates Unpublished dissertation. Raleigh, NC: North Carolina State University. Garbarino, J., & Crouter, A. (1978). Defining the community context for parent–child relations: The correlates of child maltreatment. Child Development, 49(3), 604–616. Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a community problem. Child Abuse & Neglect, 16(4), 455–464. http://dx.doi.org/10.1016/ 0145-2134(92)90062-V

K. Maguire-Jack, K. Showalter / Child Abuse & Neglect 52 (2016) 29–37

37

Garbarino, J., & Sherman, D. (1980). High-risk neighborhoods and high-risk families: The human ecology of child maltreatment. Child Development, 51(1), 188–198. Gauda, J., Polansky, N., Kilpatrick, A., & Shilton, P. (1993). Loneliness, depression, stress, and social supports in neglectful families. American Journal of Orthopsychiatry, 63(4), 597–605. Gracia, E., & Musitu, G. (2003). Social isolation from communities and child maltreatment: A cross-cultural comparison. Child Abuse & Neglect, 27(2), 153–168. http://dx.doi.org/10.1016/S0145-2134(02)00538-0 Guterman, N. B., Lee, S. J., Taylor, C. A., & Rathouz, P. J. (2009). Parental perceptions of neighborhood processes, stress, personal control, and risk for physical child abuse and neglect. Child Abuse & Neglect, 33, 897–906. Hildyard, K. L., & Wolfe, D. A. (2002). Child neglect: Development issues and outcomes. Child Abuse & Neglect, 26(6–7), 679–695. Johnson-Motoyama, M. (2013). Does a paradox exist in child well-being risks among foreign-born Latinos, U.S.-born Latinos, and whites? Findings from 50 California cities. Child Abuse & Neglect, 38(6), 1061–1072. Jonson-Reid, M., Drake, B., & Zhou, P. (2012). Neglect subtypes, race, and poverty: Individual, family, and services characteristics. Child Maltreatment, 8(1), 30–41. Kawachi, I., & Berkman, L. (2000). Social cohesion, social capital, and health. In L. F. Berkman, & I. Kawachi (Eds.), Social epidemiology (pp. 174–190). New York, NY: Oxford University Press. Kim, J. S. (2004). Neighborhood effects on the etiology of child maltreatment: A multi-level study Unpublished dissertation. Austin, TX: University of Texas at Austin. Leeb, R. T., Paulozzi, L. J., Melanson, C., Simon, T. R., & Arias, I. (2008). Child maltreatment surveillance: Uniform definitions for public health and recommended data elements. Atlanta: Centers for Disease Control and Prevention. Molnar, B. E., Buka, S. L., Brennan, R. T., Holton, J. K., & Earls, F. (2003). A multilevel study of neighborhoods and parent-to-child physical aggression: Results from the project on human development in Chicago neighborhoods. Child Maltreatment, 8(2), 84–97. Maguire-Jack, K. (2014). Multilevel investigation into the community context of child maltreatment. Journal of Aggression, Maltreatment & Trauma, 23(3), 229–248. http://dx.doi.org/10.1080/10926771.2014.881950 Pinderhughes, E. E., Dodge, K. A., Bates, J. E., Pettit, G. S., & Zelli, A. (2000). Discipline responses: Influences of Parents’ socioeconomic status, ethnicity, beliefs about parenting, stress, and cognitive-emotional processes. Journal of Family Psychology, 14, 380–400. Polansky, N. A., Gaudin, J. M., Jr., Ammons, P. W., & Davis, K. B. (1985). The psychological ecology of the neglectful mother. Child Abuse & Neglect, 9(2), 265–275. http://dx.doi.org/10.1016/0145-2134(85)90019-5 Regalado, M., Sareen, H., Inkelas, M., Wissow, L. S., & Halfon, N. (2004). Parents’ discipline of young children: Results from the National Survey of Early Childhood Health. Pediatrics, 113, 1952–1958. Sampson, R. J., & Groves, W. B. (1989). Community structure and crime: Testing social-disorganization theory. The American Journal of Sociology, 94(4), 774–802. Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing neighborhood effects: Social processes and new directions in research. Annual Review Sociology, 22, 443–478. Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277(5328), 918–924. Shaw, C., & McKay, H. (1942). Juvenile delinquency in urban areas. Chicago: University of Chicago Press. StataCorp. (2011). Stata statistical software: Release 12. College Station, TX: StataCorp LP. Stith, S. M., Liu, T., Davies, L. C., Boykin, E. L., Alder, M. C., Harris, J. M., Som, A., McPherson, M., & Dees, J. (2009). Risk factors in child maltreatment: A meta-analytic review of the literature. Aggression and Violent Behavior, 14(1), 13–29. http://dx.doi.org/10.1016/j.avb.2006.03.006 Straus, M., & Hamby, S. (1997). Measuring physical and psychological maltreatment of children with the conflict tactics scales. In G. Kaufman Kantor, & J. L. Jasinski (Eds.), Out of the darkness: Contemporary perspectives on family violence (pp. 119–135). Thousand Oaks, CA: Sage Publications. Straus, M., Hamby, D., Finkelhor, D., Moore, W., & Runyan, D. (1998). Identification of child maltreatment with the Parent–Child Conflict Tactics Scale: Development and psychometric data for a national sample of American parents. Child Abuse & Neglect, 22(4), 249–270. Straus, M. A., & Stewart, J. H. (1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration, in relation to child and family characteristics. Clinical Child and Family Psychology Review, 2, 55–70. U.S. Department of Health and Human Services. (2012). Child maltreatment 2012 Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau [online]. Washington, DC: Government Printing Office. Available from http://www.acf.hhs.gov/sites/default/files/cb/cm2012.pdf Vinson, T., Bladry, E., & Hargreaves, J. (1996). Neighbourhoods, networks, and child abuse. British Journal of Social Work, 26, 523–543.