1. Introduction
Research based upon official data reveals a steep social gradient in child abuse and neglect, with individuals of lower socioeconomic status (measured by indicators such as income, education, and employment) being more likely to come into contact with child protection services (
Pelton 1978,
2015;
Sedlak et al. 2010). Broadly, two categories of explanations have been suggested to explain this pattern.
On the one hand, scholars who view the phenomenon through the lens of social strain theories (e.g.,
Agnew 2006) argue that lower SES predicts the likelihood of maltreatment through parenting stress and parents’ ability to provide sufficient care for their children (
Lindo and Schaller 2014). The family stress model also suggests that the emotions and behaviors of parents are affected by financial concerns, which disrupt their child-rearing practice (
Conger et al. 1992;
Masarik and Conger 2017). In other words, child abuse can be a spillover effect of financial pressures, which explains its higher prevalence among the lower social classes.
Pelton (
1978) asserted that there is a common belief that the frequency and severity of child abuse and neglect are not related to socioeconomic status and can be observed among families from all social classes. Pelton calls this common belief ‘the myth of classlessness’ and argues that there is compelling evidence of the association between socioeconomic status and child maltreatment. Statistics demonstrate that the lower social classes are disproportionately represented in child abuse cases known to public agencies (
USHHS 2025). However, the maintenance of this myth serves the tendency to view child abuse as a psychodynamic problem in the context of a medical model of disease, treatment, and cure rather than as a sociological and poverty-related problem.
The relationship between socioeconomic factors and child maltreatment has also been demonstrated in more recent studies (e.g.,
Bywaters et al. 2015,
2016b;
Pelton 2015;
Wynd 2013). Evidence suggests that economic resources, particularly low income, play an important role in the risk of child abuse and neglect (
Berger and Waldfogel 2011). Low income predicts the likelihood of maltreatment through parents’ ability to provide sufficient care for their children. This mechanism is particularly relevant to the study of child neglect, which is commonly defined as the failure of a caregiver to provide for a child’s basic physical, medical, educational, or emotional needs (
Lindo and Schaller 2014). There is also compelling evidence about the link between deprivation and child maltreatment. In New Zealand, for example, research indicates that children living in the most deprived areas in 2014 had, on average, 13 times higher rates of substantiation, 18 times higher rates of Family Group Conferences (FGCs), and a 6 times higher chance of entering foster care, compared to children living in the least deprived areas (
Keddell et al. 2019).
On the other hand, some scholars who view social problems from the perspective of labeling theory (e.g.,
Becker 1963) scrutinize the overrepresentation of the lower social class in the official statistics of social problems. They emphasize the significance of surveillance in society and the disproportionate targeting of lower social classes for monitoring and control by those in positions of power (e.g.,
Garland 2001;
Wacquant 2009). This phenomenon has contributed to the overrepresentation of people from lower socioeconomic backgrounds in official data on social problems, including child maltreatment.
We all live in surveillance societies (
Hopkins Burke 2021). Today, surveillance societies operate through extensive gathering, recording, storing, analyzing, and utilizing information about individuals and groups as they conduct their daily activities (
Hopkins Burke 2021). One of these pieces of information (predictive factors) is the socioeconomic status of individuals, families, and/or neighborhoods. For instance, predictive policing models direct more police to certain neighborhoods based on historical crime data. Increased police presence results in more arrests in those areas, which feeds back into the model, justifying continued or escalated policing (
O’Neil 2016). A similar pattern has been observed in the area of child protection and is called ‘policing the family’ (
Roberts 2014,
2021,
2023), a system that leads to families from lower social classes and/or minority groups being disproportionately reported to child protection services (e.g.,
Holland et al. 2024;
Hyslop 2022).
Thus, the question is whether the overrepresentation of social problems such as child maltreatment in the lower social classes is because they commit more abusive behaviors than upper social classes or because policymakers tend to monitor and surveil the behavior of the lower social classes, a phenomenon that leads to the overrepresentation of particular groups of people in official statistics. It is worth noting that scholars who offer this argument do not dispute the evidence for the impact of poverty, unemployment, and material hardship on child maltreatment. Instead, they suggest that the observed trends might be explained by economically disadvantaged people being constantly targeted by public scrutiny and more likely to be known to social and law enforcement agencies (
Drake and Zuravin 1998).
The question of whether the relationship between SES and child maltreatment reflects a genuine association or results from hyper-surveillance and systemic racial and class biases remains a critical area of inquiry. Several key questions regarding this relationship remain underexplored. For instance, there is limited research on how poverty correlates with different types of maltreatment (physical, psychological, sexual abuse, and neglect) as well as how other SES dimensions, such as employment status and parental education, influence child maltreatment risks. Additionally, the literature lacks comprehensive comparisons of findings derived from different data sources, such as administrative records versus self-report data, and of how variations in the unit of analysis impact research outcomes. Knowledge about the causal relationship between SES and child maltreatment is still developing. Another crucial but understudied area is the interaction between poverty, income inequality, and child maltreatment rates, which has significant implications for social policy.
This article aims to review the existing literature on these aspects to provide a deeper understanding of the relationship between SES and child maltreatment. Addressing these gaps is essential, as the findings could have important implications for child protection policy. If a strong and direct relationship exists between SES and child maltreatment, policy efforts should shift toward addressing the underlying social determinants of child maltreatment rather than focusing solely on intervention at the individual or family level. However, if this relationship is largely driven by systemic biases or other factors, policymakers must confront a critical choice: either continue the neoliberal model of surveillance and control over populations deemed high-risk or fundamentally reevaluate the assumptions that shape policy responses to social issues such as child maltreatment.
While a strong theoretical foundation exists on the criminalization of poverty, there remains a notable gap in empirical research examining these dynamics. This article seeks to bridge that gap by critically analyzing existing studies and highlighting areas that require further investigation to inform child welfare policies. The purpose of this study, thus, is to critically examine the relationship between SES and child maltreatment in the current literature. Our study seeks to disentangle this relationship to explore how the aforementioned conceptual and methodological considerations may influence the strength and interpretation of this relationship.
2. Method and Materials
Given the substantial body of literature examining the relationship between SES and child maltreatment based on official and self-report data, we avoid replicating such analyses. Instead, the findings from existing studies are critically reviewed and synthesized. To achieve this, a narrative literature review was considered appropriate (
Oliver 2012;
Ridley 2012). We were interested in reviewing a broad range of literature and evidence sources without applying the search and standard quality appraisal criteria typically used in systematic reviews (such as excluding ‘gray literature’ from the review). Meta-analysis was not utilized as the method, as SES and child maltreatment are measured inconsistently across studies, making it unsuitable to combine association measures and calculate an effect size. Furthermore, producing a quantitative effect size or correlation coefficient was not the goal; instead, our objective was to develop a broad understanding of the relationship between SES and child maltreatment.
To identify the relevant literature, an extensive search of quantitative studies, government documents, dissertations, book chapters, and gray literature was conducted across multiple databases, including Google Scholar, Scopus, Science Direct, and ProQuest. A ‘snowballing approach’ was also employed to review reference lists from identified resources to capture additional literature. Using this method allowed us to identify more relevant resources as keywords may vary across disciplines or studies, making database searches incomplete. In addition, seminal or highly cited pieces may not be indexed with consistent terms. The search was limited to English-language publications with no restrictions on publication date. Key search terms included ‘child maltreatment’, ‘child abuse’, ‘child neglect’, ‘poverty’, ‘deprivation’, ‘socioeconomic status’, ‘income’, ‘education’, and ‘employment status’, along with lexical variations of these terms (e.g., maltreat*, depriv*, SES, etc.) to ensure comprehensive results.
Adopting a critical perspective, we reviewed and synthesized the literature to construct a narrative account of the relationship between SES and child maltreatment. This approach allowed us to acknowledge the complexities of the topic rather than reducing it to a mere measure that might quantify the strength and/or magnitude of the relationship. All reviewed studies were published in English, were conducted quantitatively, used official (administrative) and/or self-report data or were reviews of previous quantitative studies, and had income, education, (un)employment, and/or income inequality as an independent variable and child maltreatment rates as a dependent variable.
3. Results
There are numerous studies demonstrating a significant association between SES-related factors and child maltreatment (e.g.,
Bywaters et al. 2016b;
Pelton 2015). Systematic literature reviews suggest that higher income reduces the risk of child maltreatment in the United States (
Conrad-Hiebner and Byram 2020;
Kim et al. 2024b). Studies utilizing secondary analyses of administrative data yield similar findings (
Imran et al. 2019). Additionally, earnings instability (defined as a reduction in earnings by 30% or more) has been linked to increased involvement with child protection services (
Cai 2022;
Monahan 2020).
Various explanatory models have also been proposed to clarify the relationship between family socioeconomic circumstances and child maltreatment. Most studies propose either a direct effect, wherein material hardship or a lack of financial resources limits parent’s capacity to invest in support (such as childcare and education), or an indirect effect mediated by parental stress and neighborhood conditions (
Bywaters et al. 2016b).
While substantial evidence highlights the relationship between SES and child maltreatment rates, this article discusses six considerations in order to clarify the nuances of this relationship. These considerations are summarized in the following table (
Table 1).
3.1. Types of Child Maltreatment and Dimension of SES
The vast majority of studies examining the relationship between SES and child maltreatment focus primarily on the economic dimension of SES, often defining it as income poverty (
Kim et al. 2024b). These studies consistently demonstrate a strong correlation between income poverty and both physical abuse and child neglect (
Yang 2015). However, the relationship between income poverty and child sexual abuse appears to be statistically non-significant, indicating no notable differences in the risk of sexual abuse among children from varying socioeconomic backgrounds (
Zhang et al. 2022).
For instance,
Zhang et al. (
2022) analyzed the associations between county-level income inequality, poverty rates, and various forms of child maltreatment in the United States from 2009 to 2018. Their findings revealed that both county-level income inequality and poverty rates were significantly associated with higher overall rates of child maltreatment. Specifically, income inequality and poverty rates showed strong associations with physical and psychological abuse. Child neglect was also significantly linked to income inequality. In contrast, no significant relationship was observed between income inequality or poverty rates and sexual abuse. These findings align with a recent systematic review by
Kim et al. (
2024b), which similarly reported negative associations between family income and overall child maltreatment rates, physical abuse, and neglect, while identifying no significant relationship with sexual abuse. The association between income poverty and child neglect is particularly well supported in the literature. Given that neglect is often defined as a family’s inability to meet a child’s basic needs, including food, safe shelter, education, and adequate medical care (
Jonson-Reid et al. 2013), it often correlates with families’ economic circumstances (
Berger et al. 2017;
Maguire-Jack and Font 2017).
Furthermore, research on the effect of other dimensions of SES, such as parental education and employment status, on child maltreatment is limited and less consistent in establishing a clear association. For instance,
Wood et al. (
2012) analyzed hospital admissions for child abuse between 2000 and 2009 across 38 hospitals in 20 of the largest metropolitan areas. Their findings revealed no significant link between child abuse admission rates and local area unemployment rates.
Raissian (
2015) examined the relationship between county-level unemployment and child maltreatment reports in New York and suggested a negative relationship in metropolitan counties. Their research specifically showed that a one-percentage-point increase in the unemployment rate was associated with a 4.25% decrease in child maltreatment report rates. Conversely, other studies suggest a positive association between child maltreatment rates and male unemployment. It has been argued that changes in labor market conditions for fathers and mothers may have opposite effects on child maltreatment rates (
Lindo et al. 2018). While male unemployment is associated with an increase in child maltreatment rates, female employment has shown either no correlation (
Cherry and Wang 2016) or a negative relationship (
Lindo et al. 2018). Similarly,
Gillham et al. (
1998) found that male unemployment specifically increased rates of child physical abuse, whereas female unemployment showed no significant association with child maltreatment rates.
Even less is known about the relationship between parents’ level of education and child maltreatment. Some research shows that parents with higher education levels tend to have better knowledge of child development and effective parenting strategies, which can reduce the risk of maltreatment (
Merritt 2009). Similarly, other studies show that mothers with a high school diploma have a lower risk of committing child maltreatment compared to those without a high school diploma (e.g.,
Beimers and Coulton 2011;
Dubowitz et al. 2011). However, this relationship is influenced by various factors, including income, access to resources, and social support (
Mersky et al. 2009). Thus, education is only influential in interaction with employment opportunities and income level.
It seems that one can only confidently argue for the effect of income poverty and material hardship on a higher risk of physical abuse and child neglect. There is strong evidence to suggest that the economic dimension of SES, particularly income poverty, is closely tied to physical abuse and neglect. However, it should be noted that physical abuse and neglect in higher-SES families might be less visible to authorities (see also
Fussell 1992). In addition, the association between the economic circumstances of families and psychological abuse rates remains unclear, with some studies finding mixed results (
Zhang et al. 2021). There also remains a lack of evidence supporting an association between income poverty and sexual abuse.
3.2. The Impact of Data Source
Another important consideration in the relationship between SES and child maltreatment is the influence of data sources. Child maltreatment is only recorded in official data if an incident is reported to the child welfare system (
Kitsuse and Cicourel 2017). However, when children or their parents are surveyed through social research, they may disclose experiences of abusive behaviors that were not reported to officials. Consequently, such cases are documented in self-report data but are absent from official records. This discrepancy can be significant, given the well-documented gap between actual victimization rates and official reports (e.g.,
Lynch and Addington 2006). Studies examining the concordance between self-report data and official case records of child maltreatment highlight substantial discrepancies between these two data sources (
Negriff et al. 2017). Thus, it is worth questioning whether the observed relationship between SES and child maltreatment would differ based on the data source used.
One commonly used source of self-report data on child maltreatment in the U.S. is the Fragile Families and Child Wellbeing Study (renamed the Future of Families and Child Wellbeing Study in 2023, FFCWS), a longitudinal birth-cohort study of 4898 children born between 1998 and 2000 across 20 large U.S. cities. Using this dataset,
Maguire-Jack and Sattler (
2023) explored the relationships between neighborhood poverty, family monetary wellbeing, and child maltreatment. Their findings revealed a sustained impact of neighborhood poverty on child neglect, fully mediated by family monetary wellbeing. However, they found no significant longitudinal association between neighborhood poverty and either physical or psychological abuse. In another study,
Maguire-Jack et al. (
2019) used childcare subsidy receipt as a proxy for low income and material hardship to assess child neglect based on five maternal self-report items. Their findings indicated that childcare subsidy receipt was only associated with supervisory neglect (i.e., leaving a child home alone despite recognizing the need for adult supervision) and not with other types of neglect, such as failing to provide food, medical care, or emotional support.
Contrastingly,
Thomas and Waldfogel (
2022), also using the FFCWS database, identified a significant association between income poverty, material hardship, and child maltreatment. This contradicts the findings of
Maguire-Jack and Sattler (
2023). The divergence can be attributed to differences in their methodologies. While
Thomas and Waldfogel (
2022) measured poverty and material hardship at the family level and child maltreatment as contact with Child Protective Services (CPSs hereafter),
Maguire-Jack and Sattler (
2023) focused on neighborhood-level poverty and self-report distinct forms of maltreatment, such as physical and psychological abuse and neglect. Notably, the approach used by
Thomas and Waldfogel (
2022) aligns with the methods often employed in studies relying on official data. Further research also supports the notion that, when child maltreatment is measured as involvement with CPSs, it tends to correlate strongly with income poverty and material hardship. Families with lower SES are not only more likely to come into contact with CPSs (
Yang 2015) but also face higher chances of being re-reported for child maltreatment (
Kahn and Schwalbe 2010).
The relationship between SES and child maltreatment in self-report data is complex, and the evidence is not monolithic. While official data indicates that children from economically deprived families are the majority of children involved in CPSs (
Feely et al. 2019), and income poverty and material hardship are consistently associated with certain forms of maltreatment, more details emerge when comparing findings from self-report data and official records. Studies relying on self-report data highlight associations between poverty and family monetary wellbeing and child neglect, whereas official data often reveal broader correlations between low SES and CPS involvement.
3.3. The Impact of Unit of Analysis
In the previous two sections, studies have employed various units of analysis with both child maltreatment and SES measured at the household, neighborhood, regional, or national levels. The choice of unit of analysis is a key methodological consideration that may influence the observed relationship between SES and child maltreatment. Unit analysis affects the observed relationship, as each scale reveals unique dynamics. Understanding these methodological differences is critical for forming accurate conclusions.
Research analyzing SES at the family/household level frequently highlights a strong relationship between income and child maltreatment. There are numerous examples that confirm this relationship.
Beimers and Coulton (
2011) showed that families involuntarily exiting Temporary Assistance for Needy Families (TANF) faced a 26% higher risk of child maltreatment compared to voluntary exits. Similarly, a negative correlation between household income and child abuse was found, with income shocks significantly increasing the likelihood of CPS involvement (
Berger 2004;
Cai 2022). As mentioned before, other studies also confirm this trend, linking material hardship and income instability to heightened risks of child abuse and neglect (
Pelton 2015;
Yang 2015), as well as contact with the CPSs (
Thomas and Waldfogel 2022). These findings emphasize family income as a determinant of child maltreatment.
Neighborhood-level analyses provide a broader perspective on SES and child maltreatment. Factors such as neighborhood poverty, inequality, and social cohesion play key roles.
Keddell et al. (
2019) found a marked relationship between small-area deprivation in New Zealand and child protection system contact. Another study showed that socioeconomic characteristics of neighborhoods, such as poverty and residential instability, are linked to higher child maltreatment reports (
Coulton et al. 2007).
Zhang et al. (
2022) highlighted complex interactions between neighborhood inequality and specific maltreatment types in the US. For example, higher neighborhood inequality was associated with a reduced likelihood of spanking in low-income families but increased physical abuse risks in higher-income families.
Maguire-Jack et al. (
2022) further demonstrated that neighborhood poverty during a child’s early age could indirectly lead to child neglect through family monetary wellbeing. However, there is no significant longitudinal relationship between neighborhood poverty and physical or psychological abuse. Earlier studies also suggest a strong correlation between neighborhood-level male unemployment rates and physical abuse but a less consistent relationship between child neglect and sexual abuse (
Gillham et al. 1998).
Regional- and state-level analyses reveal broader economic patterns influencing child maltreatment.
Cherry and Wang (
2016) found that state-level male employment rates among young adults were inversely correlated with child maltreatment rates, with a 10% employment decline resulting in a 9.62% rise in maltreatment. County-level studies, such as those by
Kim and Drake (
2018), show that child poverty rates strongly correlate with maltreatment across racial and ethnic groups. Income inequality, measured by the Gini index, is also consistently linked to higher maltreatment rates (
Zhang et al. 2022). These studies highlight how systemic economic disparities and policies, such as minimum wage increases, can significantly affect maltreatment rates (
Raissian and Bullinger 2017).
There is very limited research on the relationship between SES and child maltreatment rates at a national level. While some studies suggest that income inequality at the national level is associated with weaker child wellbeing outcomes (
UNICEF 2020), establishing a direct relationship between poverty and child maltreatment is challenging.
Stoltenborgh et al. (
2015) found no significant global differences in physical and emotional abuse rates between developing and developed countries, though sexual abuse among boys was more prevalent in developing regions. These findings suggest that differences in reporting practices, cultural attitudes, and availability of resources for prevention and intervention complicate the relationship between SES and maltreatment at a global scale.
Put briefly, the unit of analysis can influence how the relationship between SES and child maltreatment is understood. Household-level studies emphasize the direct impact of income and material hardship, while neighborhood and regional analyses show mixed evidence. Cross-country studies introduce additional complexities. While child maltreatment correlates with income inequalities across the countries, it is hard to find evidence for the association between poverty and child maltreatment. A comprehensive understanding of child maltreatment and SES requires integrating insights across these levels to form a cohesive picture.
3.4. Causal Relationship or Correlation
Most studies examining the relationship between SES and child maltreatment rely on correlational and cross-sectional designs. These studies demonstrate statistical associations, showing that changes in one variable tend to coincide with changes in the other. However, such designs do not establish that SES causes child maltreatment. To move beyond simple correlations and account for potential confounding variables, some researchers have employed statistical techniques such as partial correlation and multivariate regression analysis. These methods help isolate the unique contribution of factors such as income poverty or child poverty by controlling for other independent variables that might influence child maltreatment rates (
Drake et al. 2022). For instance, using a multilevel linear model,
Kim et al. (
2024a) analyzed the relationship between child poverty and child maltreatment reports at the county level. Accounting for various control variables such as demographic characteristics and residential stability, they found that there is an association between child poverty and child maltreatment and concluded that policy changes aimed at reducing child poverty can indirectly reduce child maltreatment rates. Nevertheless, the study acknowledges limitations in establishing causality and the potential for confounding factors.
While such findings can help account for confounding variables and isolate the association between an independent variable and an outcome, they do not, by themselves, establish causality. The key limitation of correlational designs is that the co-occurrence of two phenomena does not necessarily mean that one causes the other. Establishing a causal relationship requires more rigorous methodological approaches, such as randomized controlled trials, quasi-experimental designs, or cohort designs with robust theoretical and empirical frameworks (
Shadish et al. 2002). A few studies have adopted these designs to examine the causal influence of SES on child maltreatment. For example, using a quasi-experimental design, a study by
Johnson-Motoyama et al. (
2022) aims to determine whether state Supplemental Nutrition Assistance Program (SNAP) policies that improve or stabilize household resources are associated with reductions in Child Protection Services (CPSs) involvement in the US. The study examines data at the state level for each year from 2004 to 2016 and suggests that each additional SNAP income generosity policy was associated with a reduction in CPS reports accepted for investigation, fewer substantiated reports, and fewer total foster care placements.
Even though there are some studies that bring us closer to the ability to infer a causal relationship, to date, only a limited number of them have investigated this relationship. Consequently, the empirical evidence remains insufficient to establish a definitive causal link.
Kim et al. (
2024a) explicitly mention that they have made bold assumptions regarding causality in order to provide preliminary estimates of possible indirect effects; they used terms like ‘effect’ not to claim causality but to describe coefficients and possible effects under their bold assumptions.
Johnson-Motoyama et al. (
2022) also repeatedly use the term ‘association’ in spite of controlling the relationship for several factors. In fact, they did not even use the term ‘cause’ once when reporting their findings.
While randomized controlled trials (RCTs) are widely regarded as the gold standard for evaluating causal effects, some studies employing RCT designs have demonstrated that socioeconomic interventions (such as income supplementation) can lead to reductions in child maltreatment (
Courtin et al. 2019). These findings provide important evidence that improving household socioeconomic conditions may reduce maltreatment risk. However, it would be logically fallacious to infer from such studies that low SES causes child maltreatment. This inference risks committing the fallacy of inversion of cause and effect, wherein a reduction in maltreatment following a socioeconomic intervention is interpreted as proof of a causal relationship between SES and maltreatment. Although experimental evidence supports the protective effects of improved SES, this does not necessarily imply a symmetrical or linear causal relationship in the opposite direction. Thus, caution should be exercised in drawing causal conclusions from intervention studies to the general relationship between SES and maltreatment.
3.5. Class, Ethnicity, and Racial Bias
The relationship between SES, ethnicity, and child maltreatment is complex, often raising concerns about potential racial bias within child protection systems. Some argue that systemic racism and biases in reporting, investigation, and intervention practices lead to the disproportionate identification of minority children as victims of maltreatment (e.g.,
Irwin 2009;
Owen and Statham 2009;
USHHS 2025). Given that minority children are overrepresented in lower-income groups, this could create a strong link between poverty and child maltreatment. However, some scholars have tested this hypothesis by examining whether racial disparities in child maltreatment persist after controlling for SES-related factors such as income poverty. If racial bias were the primary driver, one would expect child maltreatment rates to remain disproportionately high for minority children (e.g., Black children or Indigenous children) across all income levels compared to White children.
Kim and Drake (
2018) found that, as county-level child poverty rates rise in the United States, maltreatment rates increase for White, Black, and Hispanic children alike. However, when comparing children from similar economic backgrounds, White children had equal or even higher maltreatment rates than Black children. This challenges the assumption that racial disparities in child maltreatment stem from systemic bias in child welfare reporting. Instead, it suggests that poverty, rather than race, is the primary driver of child maltreatment disparities. Similar findings have been reported in other contexts.
Bywaters et al. (
2016a) found no significant racial differences in child welfare system involvement between Black and White children in England when SES was accounted for. Other studies further reinforce the argument that minority overrepresentation in child maltreatment statistics is a consequence of their higher likelihood of experiencing poverty rather than racial bias in CPSs (
Pelton 2015;
Putnam-Hornstein and Needell 2011;
Thomas and Waldfogel 2022).
While poverty is a major factor, the community context in which families live also contributes to disparities in child maltreatment reporting.
Jonson-Reid et al. (
2013) found that Black children in the U.S. are reported for severe and basic needs neglect more often than White children, even within similar income brackets. However, this discrepancy is best explained by the fact that Black children are more likely to live in high-poverty neighborhoods with fewer resources and social support systems. The lack of available assistance in these communities increases the likelihood of neglect cases being reported and investigated, rather than indicating racial bias in the CPS itself.
In post-colonial societies such as New Zealand, the issue of overrepresentation of minority children in child protection systems takes on a unique dimension. For instance, in New Zealand, some scholars argue that Māori (the indigenous people of Aotearoa New Zealand) children are disproportionately affected by racial bias in child protection practices (
Keddell and Hyslop 2019). Nevertheless, research indicates that deprivation plays a significant role in explaining these disparities.
Keddell et al. (
2019), in another study, found that placement rates into care for Māori children were substantially higher in the most deprived areas (851 per 100,000) than in the least deprived areas (350 per 100,000). The same pattern was observed among Pākehā (European New Zealanders) children, where placement rates rose from 78 per 100,000 in the least deprived areas to 616 per 100,000 in the most deprived areas. Although the social gradient was flatter for Māori, the findings suggest that deprivation is a significant driving force behind higher child protection involvement among both Māori and Pākehā children.
Some research suggests that disparities in child maltreatment statistics are more strongly linked to SES-related factors than racial bias within child protection systems. The overrepresentation of minority children in maltreatment statistics reflects their disproportionate presence in economically disadvantaged groups. While racial bias in individual cases cannot be entirely ruled out, broad statistical patterns indicate that poverty is the main determinant of disparities in child maltreatment rates. In effect, it can be argued that child maltreatment reports reflect broader societal inequalities rather than actual differences in maltreatment prevalence (see also
Drake and Zuravin 1998). Biases in child protection practice may be more closely related to classism and class bias rather than racism and racial bias. Nonetheless, it is worth emphasizing that the relationship between SES, ethnicity, and child maltreatment is complex, and concerns about racial bias in child protection systems remain valid, as there are other factors than income poverty that may contribute to racial disparities.
3.6. Poverty or Inequality
While income poverty and material hardship place families at higher risk of child maltreatment, recent studies suggest that inequality also plays a crucial role by weakening social cohesion and exacerbating socioeconomic stressors. Numerous studies have demonstrated the strong association between poverty and child maltreatment (
Kim et al. 2024b). Poverty and material hardship increase parental stress, reduce access to essential resources, and heighten the risk of neglect and abuse. However, research also suggests that social cohesion (i.e., mutual trust and shared expectations among neighbors) and informal social control mediate the association between poverty and abuse rates (
Maguire-Jack et al. 2022;
McLeigh et al. 2018).
Coulton et al. (
2007) highlight how social integration within neighborhoods serves as a protective factor. In communities where residents maintain strong social ties, share resources, and offer mutual support, the impact of economic hardship is less severe. Similarly,
Manabe (
2004) found that increased use of public spaces and frequent interaction among neighbors correlate with lower child maltreatment rates. Social cohesion also plays a critical role in buffering the effects of poverty during social shocks such as the COVID-19 pandemic.
Wu and Xu (
2020) found that strong social bonds helped families manage economic hardship, reducing the likelihood of harsh parenting practices and child maltreatment (see also
Oleson and Costello 2019). These findings, thus, suggest that while poverty is a key driver of child maltreatment, the presence of supportive communities can reduce its impact.
The level of social cohesion is strongly related to income inequalities rather than poverty (
Wilkinson and Pickett 2011,
2018). Research shows that societies with greater income inequality experience higher levels of social fragmentation, distrust, and economic insecurity, all of which contribute to child maltreatment (
Putnam 2015;
Wilkinson and Pickett 2011,
2018). Several studies provide strong evidence linking income inequality to child maltreatment.
Eckenrode et al. (
2014) found that county-level income inequality, as measured by the Gini coefficient, was positively correlated with child maltreatment rates, even after controlling for child poverty, demographic, and economic variables, as well as state-level differences in maltreatment rates.
Zhang et al. (
2022) confirmed these findings in a study of 902 U.S. counties over a ten-year period (2009–2018), demonstrating that greater income inequality was associated with higher rates of overall child maltreatment, neglect, and physical abuse. Their study also found that the effects of inequality were exacerbated in counties with high poverty rates. At a global level, research suggests that countries with lower income inequality tend to perform better on child wellbeing indicators. Studies highlight that nations with more equitable income distribution generally have lower child maltreatment rates and better overall child wellbeing outcomes (
UNICEF 2020;
Wilkinson and Pickett 2018). This indicates that addressing material hardship alone may not be enough to significantly reduce child maltreatment.
While alleviating poverty is critical, focusing solely on improving economic conditions for disadvantaged families may have limited effects on reducing child maltreatment. There are two main reasons for this. First, failing to address the broader economic structures that create material hardship and social insecurities means that poverty remains a persistent risk factor. Second, inequality itself erodes social cohesion, removing an important protective factor that can buffer families from economic stress. In contrast, a better distribution of wealth and income alongside policies that improve social cohesion and community social bonds will be effective in addressing the root causes of child maltreatment and as crucial as addressing poverty and material hardship.
4. Discussion
A substantial body of literature supports social strain theory, emphasizing the impact of income poverty and material hardship on parental capacity and stress (
Conger et al. 1992). However, the findings of this article highlight several important considerations in the observed association between SES and child maltreatment. While the other reviews focus on summarizing and synthesizing the findings of the previous studies, this article aims to highlight some methodological and conceptual nuances about the SES–child maltreatment relationship omitted from other systematic reviews on this topic.
There is compelling evidence that income poverty and material hardship are associated with increased child maltreatment rates (
Drake et al. 2022;
Kim et al. 2024b). However, the relationship is less consistent when examining psychological abuse, and there is little evidence supporting an association between income poverty and sexual abuse. The source of data and the unit of analysis also play a critical role in shaping findings. Most studies investigating the relationship between SES and child maltreatment rely on administrative data, measuring poverty at the micro-level (i.e., family level) and child maltreatment as contact with CPSs. As a result, these studies tend to find a negative association between income poverty and child maltreatment risk. In contrast, studies utilizing self-report data and measuring poverty at the macro-level (i.e., neighborhood level) present a more complex and less definitive relationship. For example,
Maguire-Jack and Sattler (
2023) found that neighborhood poverty was associated with child neglect (mediated by family monetary wellbeing) but showed no significant relationship between neighborhood poverty and physical or psychological abuse.
Another key factor in understanding the SES–child maltreatment relationship is the mediation effect of social cohesion. Some studies suggest that social cohesion, characterized by mutual trust, shared expectations among neighbors, and informal social control, mediates the association between poverty and abuse rates (
Maguire-Jack et al. 2022;
McLeigh et al. 2018). Trust and social integration within neighborhoods serve as protective factors against the adverse effects of poverty (
Coulton et al. 2007). In communities where residents maintain strong social ties, share resources, and provide mutual support, the impact of economic hardship is less severe (
Manabe 2004). This raises the broader question of whether child maltreatment is primarily about lack of resources or is more closely tied to relative poverty, inequality, and perceived inferiority (
Wilkinson and Pickett 2018). As social beings, individuals perceive themselves in relation to those around them. While absolute poverty and the inability to meet basic needs can lead to frustration and distress, studies measuring poverty primarily rely on indicators such as household income below a set threshold (e.g., 60% of the median disposable household income) or material hardship. Thus, poverty in itself may not necessarily induce stress, frustration, or anger, particularly if it is a shared condition (e.g.,
Festinger 1954). However, inequality, which fosters a sense of inferiority, is a stronger predictor of distress (
Wilkinson and Pickett 2011). This perspective suggests that policies focused solely on financial assistance for the poor may not be as effective as anticipated. Such policies can create a new label (i.e., beneficiaries), which may become another source of stigma and psychological distress. Therefore, addressing inequality alongside poverty is crucial for addressing the problem of child maltreatment.
From a labeling theory perspective, the overrepresentation of lower-SES families in official child maltreatment statistics may stem from hyper-surveillance as well as problematizing the behaviors of the lower social class (
Nazari et al. 2024). The findings of this article provide some support for this claim. Research has demonstrated that biases exist within child protection systems, aligning with labeling theory and hyper-surveillance arguments. However, evidence suggests that these biases are more strongly linked to social class than race, indicating that economically disadvantaged families, regardless of their racial background, are more likely to be scrutinized and subjected to intervention (
Kim and Drake 2018). These findings regarding class versus racial bias in CPSs suggest that cultural responses to child abuse and neglect can only be effective to a limited extent if the broader socioeconomic context remains unaddressed. Furthermore, they call for a critical reassessment of the hypothesis that racial bias in child protection practices is the primary driver of overrepresentation of minority groups.
Similar dynamics have been observed in studies on incarceration. For example,
Lewis (
2018) argues that the mass incarceration of Black individuals in the U.S. cannot be attributed solely to institutional racism; rather, it is a mechanism within a capitalist system to manage the poor. Lewis argues that once class is controlled for, race does not significantly impact the likelihood of incarceration. This does not imply that racism is irrelevant. While it must be acknowledged that the historical legacy of racism has contributed to Black (or more broadly minority groups) economic disadvantage, framing issues such as child maltreatment or incarceration through the lens of institutional racism assumes that minorities come into contact with the system because of their racial identity. In reality, however, data suggest that economic disadvantage is the primary factor driving interactions with state institutions.
It seems that, while there is robust evidence linking income poverty and material hardship to child maltreatment, research also suggests that social cohesion, relative poverty, and inequality play crucial roles. At the same time, biases in child protection systems disproportionately affect lower-SES families, lending support to labeling theory. The complexities of these relationships underscore the need for further research that integrates both perspectives, recognizing that child maltreatment is shaped by both structural socioeconomic conditions and systemic biases.
5. Conclusions
The complex interplay between inequality and child maltreatment as well as the intersection of racial bias and social class warrants more focused empirical research. Future studies should aim to clarify relationships, explore these dynamics across diverse contexts, and address methodological limitations in the existing literature. While further research is needed, current findings provide a basis for considering several important policy implications.
The overrepresentation of lower-SES families in child maltreatment statistics does not justify right-wing policy interventions that assume economically disadvantaged parents need to learn better parenting skills. Instead, it highlights the need for systemic support. Research indicates that low-SES parents are more likely to adopt an authoritative parenting style (
Schneider and Schenck-Fontaine 2022). Nonetheless,
Putnam (
2015) argues that poor families tend to adopt stricter parenting practices not due to a lack of understanding of permissive parenting but as a protective response to high-risk environments. In environments where children face greater external dangers, a more punitive approach becomes a strategy for safeguarding them from harm.
Several studies suggest that policies aimed at providing economic stability, such as adequate income, stable housing, and secure employment, can help reduce the risk of child maltreatment (e.g.,
Pelton 2015;
Raissian and Bullinger 2017). This aligns with research demonstrating that strengthening social cohesion and support networks can buffer the adverse effects of economic hardship and poverty (
Schneider et al. 2017). Social support systems play a crucial role in protecting needy families, particularly those facing multiple stressors such as poverty, social isolation, intimate partner violence, substance abuse, and previous interactions with child welfare services (
Wu and Xu 2020).
In addition, addressing child maltreatment effectively requires policies that go beyond alleviating material hardship. It necessitates reducing inequality and combating the stigma attached to poverty. Child maltreatment does not exist in isolation but is embedded within broader socioeconomic structures. In the neoliberal era where society is riven by inequality, attempts to address child maltreatment as a separate issue risk overlooking its systemic roots. As
Wilkinson and Pickett (
2011) argue, many social policies operate under the assumption that the poor must be taught how to live more responsibly. This perspective obscures the fundamental role of inequality and relative deprivation in shaping social problems. As a result, structural factors are often ignored, and the false notion persists that poverty is the result of individual failings rather than systemic disadvantages (
Nazari et al. 2024).
In conclusion, policies aimed at alleviating class disparities may provide more effective and sustainable solutions for families involved in child welfare systems and are perhaps better for everyone.