• Arab children;
  • Jewish children;
  • child maltreatment;
  • poverty;
  • child welfare services;
  • protective service responses, public disclosure of abuse;
  • voluntary family intervention


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

The authors examined the relationship between community characteristics and child maltreatment reports in Israel, comparing Jewish and Arab localities both in terms of maltreatment reports and the responses of the social services to these reports. Administrative data were obtained from the protective services and the Central Bureau of Statistics for 231 local authorities in Israel (covering 98% of children in Israel). Jewish communities showed significantly more reports of sexual abuse and physical abuse than Arab communities. Reporting rates also showed different associations with socioeconomic and demographic variables. In Jewish localities, demographic, economic, and educational factors were all correlated with the reporting of child maltreatment, whereas in Arab localities, only median age (young) of the local population was associated with a greater rate of child maltreatment reporting. No differences in the responses of the protective services were found. Implications of the results for future policy are discussed.

Current ecological theories posit that child maltreatment results from complex factors across multiple ecological levels, including individuals, families, communities, and culture (Belsky, 1993; Coulton, Crampton, Irwin, Spilsbury, & Korbin, 2007). Rates of maltreatment are associated with indications of a community's social and economic characteristics (Deccio, Horner, & Wilson, 1994; Hyde, 1999; Swanson, 2001), such as poverty, unemployment, access to health care, fragmented social services, social isolation, and neighborhood violence (Limber & Nation, 1998). Indications of community disorganization are also correlated with child maltreatment rates (e.g., residential instability; Deccio et al., 1994; Hyde, 1999; Swanson, 2001), vacant housing (Deccio et al., 1994; Zuravin, 1989), lower participation of women in the labor force (Swanson, 2001), crowding (Garbarino & Kostelny, 1992), and the child-care burden (Coulton, Korbin, & Su, 1999; Korbin, Coulton, Chard, Platt-Houston, & Su, 1998).

In addition, although some research indicates that ethnic differences affect the prevalence and severity of various types of abuse and neglect (Brenner, Fischer, & Mann-Gray, 1989; Connelly & Straus, 1992; Jones & McCurdy, 1992), other studies have not found ethnic or race differences in child abuse (e.g., Sledjeski, Dierker, Bird, & Canino, 2009). Inconsistent results were also obtained when the associations between nationality and ethnicity and child maltreatment were examined in Israel. Although some studies found that Arab localities showed lower rates of reporting child maltreatment (Ben-Arieh, 2010; Ben-Arieh & Haj-Yahia, 2006), others found a prevalence of child maltreatment in Arab localities similar to or higher than the rates estimated for Jewish children (Dwairy, 1991; Haj-Yahia & Ben-Arieh, 2000; Haj-Yahia & Dawud-Noursi, 1998). However, these studies neither included all the children (until the age of 18) nor did they include all the local authorities. They also failed to differentiate among the various manifestations of maltreatment or the differences in welfare responses to the reports. This is important, because various community characteristics like poverty may be more associated with certain types of maltreatment, such as physical and medical neglect, rather than, for example, sexual abuse.

In the present study, we deepen the examination of the relationship between community socioeconomic characteristics and the reporting of multiple types of child maltreatment, neglect, physical abuse, and sexual abuse. We further explore to what extent these relationships differ between Arab and Jewish communities in Israel. In addition, we examine for the first time whether the local social services respond differently to child maltreatment reports in Jewish and Arab communities.

Israel is a multicultural country composed of many different social, ethnic, and religious groups. Here, we focus on two major social groups: the Jewish majority and the Arab minority, mostly the Muslim citizens living within Israel (and not in the Palestinian Authority or the Occupied Territories). These two ethnic groups often live in separate communities, the Arab communities being more rural and most of them having lower socioeconomic status than the Jewish population. Given the higher poverty and crime rates in the Arab communities in Israel (Kop, 2007), we would expect higher rates of child maltreatment than in Jewish communities.

One of the drawbacks in the literature on child maltreatment is that race and ethnicity are not consistently disentangled from socioeconomic status. Ethnic minority groups are at greater risk for poverty; therefore, they appear to have higher rates of child maltreatment. Also, because neglect is more likely to occur in families with fewer resources to provide for their children, neglect allegations are frequently made when the children's basic needs, such as for food, shelter, or clothing, are not adequately met (Korbin et al., 1998). This may explain why some racial and ethnic minority groups with higher rates of poverty are overrepresented in the child welfare system, particularly in the child protection system (Lindsey & Shlonsky, 2008). Nonetheless, cultural values and social organization such as strong social and family ties may make certain minority groups experiencing high levels of poverty less vulnerable to child maltreatment. In contrast, cultural values accepting corporal punishment as legitimate may result in higher levels of child maltreatment, regardless of the parents' material resources (Gracia & Herrero, 2008a, 2008b; Liao, Lee, Roberts-Lewis, Hong, & Jiao, 2011). For example, children in Arab schools in Israel report higher levels of physical violence and corporal punishment by their (Arab) teachers than do their peers in Jewish communities (Benbenishty, Zeira, Astor, & Khoury-Kassabri, 2002; Khoury-Kassabri, Astor, & Benbenishty, 2008). Therefore, it is important to separate the effects of belonging to an ethnic or cultural minority group and the effect of poverty.

Differences in reporting maltreatment among various social and ethnic groups may reflect not only the actual prevalence of this phenomenon, but it may also intersect with the degree of adherence to conservative and traditional values in society (Haj-Yahia, 1995, 2000). Ethnic groups may differ in their willingness to approach social services and thus be identified and reported to child protection agencies (Sue & Sue, 1990; Sue, Zane, & Young, 1994). Arabs, as a minority in Israel, may report less maltreatment to avoid stigmatization of their families or their communities. This may be especially relevant in small rural localities where there are close relationships between families and child welfare workers, who may be part of the same extended families. Also, despite the fact that protective service workers in Arab communities are mostly Arabs, they may be perceived by community members not only as “outsiders,” but also as representatives of the Jewish state (Ben-Arieh & Haj-Yahia, 2006; Haj-Yahia, 1995). Studies among the African American families in poor neighborhoods in the United States also found lower levels of maltreatment reporting, probably associated with mistrust of child welfare agencies (Freisthler, Bruce, & Needell, 2007).

Variations across communities in reporting of child maltreatment may also be related to how the local social service agency responds to maltreatment reports. Few studies have addressed this issue. Notably, Ben-Arieh and Haj-Yahia (2006) have investigated the association between local child maltreatment reports and the availability of social workers in the locality. Because social services in small rural Arab communities with lower socioeconomic status are weaker, employ fewer social workers, and have far fewer intervention resources than Jewish communities, they may be able to handle far fewer reports than better resourced social services. Indeed, Ben-Arieh (2010) found that the number of social workers and protective service workers in the local social services was significantly and positively correlated with the reported rate of child maltreatment in Arab communities.

The current study goes beyond the issue of level of staffing of social services and explores the associations between the socioeconomic characteristics of the community and the types of responses they tend to provide to maltreatment reports. In Israel, child protection is carried out by workers in local welfare departments who are specifically assigned, trained, and licensed to carry out child protection. The national child protection system provides consultation and training, little clinical support and supervision, and no resources. Thus, local child protective services in Israel have significant professional discretion to choose from a wide range of options when responding to maltreatment reports. They are also often hampered by the lack of local resources and organizational pressures. In Israel, involving the court is optional, and mandatory reports to the police can be waived in most cases. Furthermore, local protective services may involve the family in a nonpunitive voluntary family intervention in lieu of working with the court to force solutions on the family (such as removal from home). These choices are expected to be made on a case by case professional judgment. Nevertheless, given the lack of binding legal or policy directives, one might expect that the types of responses selected by professionals in a local authority may reflect factors such as cultural preferences and availability of resources. In the present study, for the first time, we compare the types of response of the protective services to reports of child maltreatment in Arab and Jewish communities in Israel. Although this study does not purport to identify causal relationships between maltreatment reports and differences in the types of response by the protective services, this comparison may add to our understanding of the complex relationships between socioeconomic and cultural factors and protective services patterns of response among two different ethnic groups.

To test our hypothesis that socioeconomic and demographic characteristics of localities are associated with the rate of reports of maltreatment, the present study focuses on the relationship between community characteristics and child maltreatment reports. We compare the rate and type of child maltreatment reports in Jewish and Arab localities in Israel and the associations between community characteristics and maltreatment reports. In addition, we compare responses to these reports by the protective services in Jewish and Arab localities.


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References


The sample consisted of 231 local authorities in Israel (of a total of 256, 98% of all children in Israel). Of these, 170 localities were Jewish (66.9%) and 84 Arab (33.1%).

Data Collection

The present study was based on secondary analysis of the database of a national annual reporting system that gathers statistics on all maltreatment reports in every local authority in Israel and the distribution of protective services responses to these reports. These data were merged with the census information, which includes sociodemographic data provided by Israel's Central Bureau of Statistics on each of these local authorities.


Child maltreatment rates and types

For each locality, the protective services database provided information on the number of children reported to the protective services during the calendar year 2008–2009. These reports were classified into three categories: physical abuse, sexual abuse, and neglect (7.1% of the reports were not classified).

Protective services responses

For each locality, the protective service database contained the distribution of the number of responses to the reports, including court orders for supervision orders, requests for emergency orders, court-declared children in need, court orders for removal from home, reports to police, request to waive mandatory reports to police, and voluntary family intervention (with no court involvement).


Each locality in the study was defined as Jewish or Arab based on its ethnic composition. Most localities had close to 100% of one ethnicity, except for a few large Jewish cities, considered “mixed,” in which the proportion of Arab minority reached 24.5%.

Community demographic and socioeconomic characteristics

The sociodemographic characteristics included in the present study were selected based on the literature on child maltreatment and previous findings (Ben-Arieh, 2010; Ben-Arieh & Haj-Yahia, 2006; Garbarino & Kostelny, 1992; Korbin et al., 1998; Swanson, 2001). They were extracted from the Central Bureau of Statistics database on each of the local communities in Israel. Three sets of variables were chosen to reflect the demographic and socioeconomic status of each locality: demographic and age factors, economic factors, and education level.

Demographic and age factors

The size of the locality was computed as the total population-averaged number from the data from 2004 to 2006. The proportion of children aged 0–4 was computed as the percentage of the overall population. The proportion of children aged 0–17 was computed as the percentage of the overall population.

Locality socioeconomic status

This index was developed by the Israeli Central Bureau of Statistics based on the characterization and ranking of all local authorities in Israel by the socioeconomic level of the resident population. The index summarizes a number of socioeconomic variables and classifies the localities into 10 homogeneous clusters. Each locality is marked on a scale of 1–10 with 1 being the lowest socioeconomic status. Income refers to the average income in the locality. Unemployment is the percentage of adults receiving unemployment benefits based on all adults of working age in the locality.

Education level

High school diploma refers to the percent of residents eligible for a high school diploma, necessary for postsecondary education in Israel. Higher education refers to the percentage of residents who have participated in higher education. Table 1 presents the sample characteristics of Jewish and Arab localities.

Table 1. Sample Characteristics of Jewish and Arab Localities
  1. All comparisons (t test for independent samples) were significant. <.001.

Total population (thousands)36.6275.7512.4110.86
Median age27.535.3119.933.72
Children age 0–4 (%)9.833.6512.762.90
Elderly (more than 75 years old) (%)3.862.431.28.73
Mean income (Z score).39.91−.87.42
Unemployment (Z score)−.
High school diploma (%)55.0715.7240.1310.53
Higher education students (%)18.609.116.793.99
Overall ranking (1–100)94.1270.8546.2332.01


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

Reporting Child Maltreatment

In 2009, 33,751 new maltreatment reports were filed in Israel (17.74 reports per 1000 children, SD = 11.61): 37.2% were reports of neglect, 36.4% physical abuse, and 15.6% of sexual abuse (7.1% were not classified). Comparisons of the reports in Jewish and Arab localities (Table 2) revealed a significantly higher rate of reporting in Jewish localities (Jewish = 17.52, SD = 11.26, Arab = 11.92, SD = 10.81), t(235) = 3.67, < .001. This difference was mainly the result of considerably higher rate of reports of sexual abuse in Jewish localities (= 3.87, SD = 4.25, vs. = 1.59, SD = 1.86 in Arab localities) t(213) = 4.30, < .001. Reports of physical abuse were also slightly and significantly higher in Jewish than in Arab localities (Jewish = 5.75, SD = 4.00; Arab = 4.58, SD = 4.12), t(227) = 2.07, < .05. There were no significant differences in reports of neglect.

Table 2. Means and Standard Deviations of Rates of Maltreatment Reports in Jewish and Arab Localities
LocalitiesChildrenRate per 1,000 childrenLocalitiesChildrenRate per 1,000 children
  1. a

    Not all reports were classified into types. Therefore, the total number of reports is different from the total number of classified reports.

  2. *< .05. ***< .001. In t test for independent samples.

Total reportsa15728,498a17.5211.26825,253a11.9210.81 ***
Physical15310,3775.754.00781,9284.584.12 *
Sexual1444,6153.874.25736601.591.86 ***

Associations Between Locality Characteristic and Maltreatment Reports

Our hypothesis was that socioeconomic and demographic characteristics of the localities are associated with the number of maltreatment reports. The correlations in Table 3 are presented separately for Jewish and Arab localities. Of the 11 locality characteristics we examined, seven were significantly associated with the number of maltreatment reports. Overall, in Jewish localities there were more maltreatment reports in localities that were larger (= .38, p < .01), with high rates of unemployment (= .51, < .01), and with younger population (e.g., = .26, < .05 with median age). Community characteristics were also significantly associated with the various types of reports. Reports of physical abuse and of neglect both showed highest correlations with unemployment (= −.49 and −.52, respectively, both < .01). Sexual abuse showed weaker association with unemployment (= −.28, < .01) and with size of locality (= .31, < .01).

Table 3. Pearson Correlations Between Socioeconomic Characteristics, Total Reports of Maltreatment, and Types of Maltreatment in Jewish and Arab Localities
  1. *< .05. **< .01.

Demographic and age factors
Total population.26**.31**.31**.34**.
Median age.29**.07.31**.25**.34−.15.47**.47**
No. of children ages 0–4.32**.03.23**.16*.37*.11.56**.56**
Rate of children ages 0–17 (%).37**.05.32**.26**.30.14.46**.46**
Economic factors
Average income.08−.05.00−.02−.13−.10−.16−.16
Socio economic status (rank).19**−.−.07−.02.01
Education level
Higher education students−.11−.01−.10.14*−.12−.12−.11−.14
Eligibility for high school diploma.18**.07.14*.13.02−.10.09−.08

In Arab localities, the age pattern was also strongly associated with maltreatment reports (e.g., = .56, < .001 with the proportion of children aged 0–4 years in the population). Economic factors, on the other hand, were not associated with rates of maltreatment reporting. In both Arab and Jewish localities, community characteristics predicted sexual abuse less strongly than the reporting of other forms of maltreatment.

Protective Services Responses to Reports

Table 4 presents the responses of the protective services as a proportion of the number of children in the community to control for community size. For Jewish and Arab localities, the variance of most types of responses was very high. Voluntary family interventions were social workers' most frequent type of response to maltreatment among both Jewish and Arab localities (= 63.48, SD = 241.53), whereas court protective orders showed the lowest mean score (= 0.80, SD = 4.07). Jewish and Arab social workers showed similar use of court orders declaring children “in need” (= 29.47, SD = 100.50).

Table 4. Means and Standard Deviations of Responses of the Protective Services to Maltreatment in Jewish and Arab Localities
Reports to police22.8217.5019.4423.63
Request to override mandatory report to police11.3831.025.9315.39
Request of emergency orders2.737.635.0013.96
Court-declared children in need29.47100.5037.66122.94
Court orders for supervision orders6.4025.113.9911.97
Court orders for custody orders6.5318.152.254.45
Protective orders0.804.074.1115.81
Voluntary family interventions63.48241.5361.6397.69
Request of intermediate court orders7.2814.895.287.91

The absolute number of most response types differed between Jewish and Arab localities because of large differences in size between Jewish and Arab localities. When these responses were calculated as a proportion of the number of children in the community, however, there were no significant differences between the Jewish and Arab localities in the rates of the most common responses to maltreatment responses.


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References

We compared Jewish and Arab localities in terms of both rates of reporting and the types of responses of protective services and examined the relationships between community characteristics and reports of child maltreatment, separately for Jewish and Arab localities in Israel. Overall, Arab communities, despite their lower socioeconomic status, reported fewer cases of maltreatment. Rates were lower in Arab communities mainly as a result of slightly but significantly lower rates of reporting sexual abuse compared with Jewish localities. Despite the large differences in economic status between these two groups, rates of neglect reports were quite similar. These findings agree with those of Ben-Arieh (2010) and Ben-Arieh and Haj-Yahia (2006), but contrast with other studies that have found that Arab localities show similar or higher rates of child maltreatment reports than the estimated rates among Jewish children in Israel (Dwairy, 1991; Haj-Yahia & Ben-Arieh, 2000; Haj-Yahia & Dawud-Noursi, 1998).

We further hypothesized that socioeconomic and demographic characteristics of localities are associated with the number of reports of maltreatment. This hypothesis received only partial support and appeared to fit with Jewish better than with Arab communities. Among the Jewish localities, seven of 11 community characteristics were significantly associated with the number of maltreatment reports. More maltreatment was reported in larger Jewish localities with high rates of unemployment and a younger population. Community characteristics in the Jewish communities were also significantly associated with the forms of maltreatment, reports of sexual abuse, physical abuse, and neglect showing similar patterns of relationships and correlating highly with unemployment. These results are in line with earlier studies (Deccio et al., 1994; Freisthler et al., 2007; Gillham et al., 1998). In contrast, in Arab localities, socioeconomic indicators were not predictive of rates of maltreatment. Instead, only the young age of the population was strongly associated with maltreatment reports.

The largest difference between Jewish and Arab localities was that Arab communities showed lower rates of reporting child sexual abuse. Shalhoub-Kevorkian (2005) has argued that sociopolitical factors affect both helpers' and children's perceptions of disclosure and intervention in cases of sexual abuse. In her study of sexually abused Arab Palestinian girls, she found that girls often do not disclose their abuse because of feelings of embarrassment, guilt, shame, and fear of retribution. Children may also remain silent because they are unable to verbalize their abuse, fail to comprehend the act of sexual abuse, and have difficulty understanding the role of the criminal justice system. These fears are also affirmed and increased by the perceptions and attitudes of the various helpers. The helpers' interventions are colored both by their fear that the girls could be ostracized or killed for disclosing their abuse and their fear of the reactions of the formal and informal systems to public disclosure of the abuse (Shalhoub-Kevorkian, 2000).

Despite major differences in economic status, Jewish and Arab localities showed similar rates of neglect, even though neglect in Jewish localities was highly associated with un-employment, and there was virtually no such relationship in Arab localities. Possibly, there are different expectations of what constitutes neglectful parenting. What is considered neglect in Jewish localities may not be interpreted as neglect in Arab localities. Similarly, given the different patterns of childrearing in more traditional Arab families than in more Western Jewish families, some behaviors considered physical abuse by one group may not be interpreted as such by the other cultural group (Ferrari, 2002).

The relatively low levels of maltreatment reports in Arab communities, despite their higher levels of community distress, the lack of association in Arab communities between maltreatment rates and socioeconomic characteristics of the locality, as well as the inconsistencies between the present findings and other estimates carried out using other research methods in Israel, could be interpreted along two different lines: biased reports in Arab communities or the impact of cultural values and practices.

The literature frequently suggests that poor families are over-reported because of their more frequent contacts with social service agencies and because of workers' biases (Drake & Zuravin, 2010). In the present study, however, we found lower reporting levels among Arab minority families, who have significantly higher poverty rates, compared with the Jewish majority. This may be the result of attempts by Arab families, as a minority group, to rely less on social services and protect themselves against the interventions of the authorities and the formal system they distrust (Fenison, Popper, & Handelsman, 1990; Sue & Sue, 1990; Sue et al., 1994). Solving domestic problems without reporting to authorities has been noted by Arab Israeli researchers, such as Haj-Yahia (2000). The political conflict between Palestinians and the Israeli State may add an additional dimension to this problem (Shalhoub-Kevorkian, 2005). As a result, many Israeli Arabs may feel oppressed and discriminated against and prefer not to collaborate with the authorities against their own people.

Reporting biases may not be solely responsible for the pattern of findings in this study. Differences in culture and social organization between Jewish and Arab communities may result in different rates and patterns of maltreatment. Thus, the actual behaviors of parents toward their children, how such behaviors are interpreted, the extent to which such behaviors are reported, and how the protective services tend to respond to such reports may all vary in different groups. Certain minority groups may be less vulnerable to child maltreatment, despite their socioeconomic status, as their community's strong social ties may serve as a protective factor by reducing family stress and thus lowering levels of maltreatment.

Certain beliefs and perceptions shared by an ethnic or cultural group may also serve as a protective factor. In the context of the present study, familism may be relevant to offer an explanation for the findings. Familism is regarded as a cultural value and a commitment to provide an emotional support system for family members and to emphasize greater importance of the family in relation to the individual. It places great emphasis on dependency and reliance on others, as well as on the family unity, with a sense of obligation among the family members to care for all members, especially children (Ferrari, 2002). Familism characterizes many Hispanic and African American communities. African Americans who live in close proximity to their extended families report greater life satisfaction (Ellison, 1990), and kin play an important role in the care of young children and in the children's positive development (Garcia Coll, 1990). Studies on Hispanic Americans suggest that familism represents mechanisms that help immigrants hold on to their heritage culture and protect them from destructive outcomes, such as drug and alcohol misuse (Ramirez et al., 2004; Rodriguez, Mira, Paez, & Myers, 2007). Moreover, familism has been identified as a key process in Hispanic people's parenting that helps Hispanic parents protect their adolescents against problematic outcomes in the American society (Bush, Supple, & Lash, 2004). Coohey (2001) found that familism was protective against child abuse for both Hispanics and non-Hispanic Whites.

The concept of familism may also apply to Arab families. Several authors have emphasized the importance of the family institution in the Arab culture, thus supporting the possibility that these values serve as a protective factor against maltreatment. It should be noted, however, that the concept of familism and the centrality of the family in the Arab culture have also been associated with increased levels of maltreatment and attitudes that support using violence against women and children (Khoury-Kassabri & Straus, 2011; Pitner, Astor, Benbenishty, Haj-Yahia, & Zeira, 2011).


The present study examined, for the first time, whether Jewish and Arab localities differ in the relative use of social services as responses to maltreatment reports. Surprisingly, despite large socioeconomic and cultural differences between the two groups of localities, no significant differences were found. Thus, although Ben-Arieh (2009) found a positive correlation between the availability of social workers and the rate of reported maltreatment, our results indicated no significant difference to the maltreatment reports filed. This finding may reflect the similarity in socialization between Arab and Jewish social workers who are educated and trained in the same educational settings. It is also possible that structural constraints in the Israeli context shape all services, regardless of regional variations. This issue should be explored further, as the current study was the only one so far to address this issue and needs to be replicated and refined.

The present study adds to the previous studies in Israel by Ben-Arieh and Haj-Yahia (2006), yet its limitations should not be overlooked. The study was based on an administrative database aggregated at the locality level and, as such, does not permit analyses at the family–child level. Aggregating at the locality level may hide important variability within localities, especially for larger localities that contain several distinct communities possibly with different patterns of maltreatment. Finally, this secondary analysis is limited by the scope and quality of the data it contains.

Our data show a clear need to complement this large-scale national study with more focused research examining some of the interpretation suggested here. It is important to understand the cultural and social dynamics that lead from behaviors to interpretations to reporting and then to protective service responses. The present study indicates that such an exploration must be sensitive to culture and sociodemographic group, especially as models of child maltreatment tend to overlook cultural interpretations and the role of context in the responses to reports of maltreatment. In relation to policy, the present study indicates that, at least among the Jewish localities, there is a need to prioritize and focus on communities under economic stress, especially in large localities with many young children. Furthermore, the protective services need to understand better the dynamics of maltreatment in Arab localities. Lower rates of maltreatment in these communities should not be seen only at face value. Cultural and gender sensitivity are needed. Interventions should also be implemented on the basis of partnership and negotiation with the children and their families to ensure that the actions taken are in the best interests of the children in their own cultural context. Finally, more should be done to learn from Arab social workers as to how they understand child maltreatment and reporting patterns to identify the true needs of children in this minority group.


  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. References