How might changes to family income affect the likelihood of children being in out-of-home care? Evidence from a realist and qualitative rapid evidence assessment of interventions

Interventions that change family income include any policy or practice that directly or indirectly changes the amount of money a family have. Although theory regarding the relationship between poverty and child maltreatment is well established, theories of how family income change affects the likelihood of children being in out-of-home are not well developed. This realist rapid evidence assessment provides an overview of the process of how interventions that change family income affect the rate of children in out-of-home care. The study population is families at risk of their children entering care and families whose children are in care and are pursuing reunification. Ten studies were identified from an earlier scoping review. Intervention effect results are described and qualitative evidence about mechanisms and moderators is presented as an initial “ programme theory ” . The review makes an initial suggestion of four pathways through which a change in family income can alter the risk of child abuse and neglect and thus affect the rate of children in out-of-home care. These are: 1) the impact of employment; 2) changes to the home environment; 3) risk/prevention of homelessness; 4) building trusting relationships with social workers. National or local policies which increase a family ’ s income, for instance through tax and benefits regimes or the provision of free childcare, could potentially reduce the rate of children in out-of-home care. There is also a role for social workers in providing direct material help to families. work is to develop intervention theories and better understanding of the process of using material resources to help families at risk of their children being in out-of-home care.


Introduction
A relationship between poverty and child maltreatment is well established in research evidence (see Berger & Waldfogel, 2011;Bywaters et al., 2016;Hunter & Flores, 2021). Child maltreatment is in turn a key reason for children being removed from their birth families into out-of-home care (Perlman & Fantuzzo 2013). The current focus of this paper is on how changes to family income can affect the rate of children in out-of-home care.
Individual-level studies have reported higher rates of maltreatment in families facing economic disadvantage (Sidebotham et al., 2002;Doidge et al., 2017). Area-level studies, also known as aggregated studies or ecological studies, have found a similar association. A review by Coulton et al. (2007), based on mostly North American studies, found that child maltreatment reports concentrated in comparatively disadvantaged neighbourhoods. In a recent UK study, child protection intervention rates in the most deprived ten per cent of neighbourhoods were at least eight times higher than those in the least deprived ten per cent, with each step increase in deprivation accompanied by an increase in child protection interventions. This study also found a similar social gradient for a child's chances of being in out-of-home care (Bywaters et al., 2020). Yet, estimating the causal effect of specific economic determinants on child maltreatment is challenging, due to the interrelatedness of economic factors (Bullinger et al., 2021;Slack et al., 2017).
A recent study in England found that between 2015 and 2020, a 1 % increase in child poverty in a neighbourhood was associated with 5 additional children entering out-of-home care per 100,000. They estimated that over the study period 8.1 % of care entries were linked to rising child poverty, equivalent to 10,356 additional children (Bennett et al. 2021). Both Rostad et al. (2020) and Biehl and Hill (2018) in the USA, found that providing income support to families was associated with lower rates of children entering foster care. Bywaters et al., (2016, see also Bunting et al., 2018 identify-two very broad mechanisms for what they describe as a causal link between poverty and child maltreatment, which could contribute to more children being placed in out-of-home care. First, there are direct effects of poverty, whereby a lack of money and associated material hardship may impede a parent's ability to meet a child's needs, health, and development or invest in material or social support and better environmental conditions (also known as the investment modelsee Duncan et al., 2014). The legal definitions of neglect typically include inadequate shelter, food, and clothing (Fong, 2017). Second, there are indirect effects, whereby living in poverty causes parental stress, shame, stigma, and family conflict (also known as the family stress modelsee Conger et al., 2000), which could result in harsher or less supportive parenting practices, which are risk factors for child maltreatment (Stith et al., 2009). Other indirect effects include poor neighbourhood conditions which put child well-being at risk (Maguire-Jack & Font, 2017). For instance Shuey and Leventhal (2017) found that greater neighbourhood affluence is indirectly associated with mothers' lower reports of physical aggression with their children, due to more available neighbourhood services helping to manage children's time and activities. Webb and Bywaters (2018) found that area deprivation affected the way preventive family support and early intervention services are funded, with the most deprived local authorities, receiving the least funding for services relative to their needs. Both direct and indirect effects, although logically separable, are likely to be at play at the same time (Bywaters et al, 2022) and can interact with personal and social problems such as parental mental illness, domestic violence, substance misuse and criminal justice involvement, which are also associated with harm to children (Cleaver et al., 2011;Brandon et al., 2008;Phillips et al., 2004). These pathways have been invoked in other literature in similar ways, for instance Shook (1999) theorised three pathways between income loss and child welfare risk. They grouped these pathways as: psychosocial theories, where increased financial pressure leads to parental stress and depression, thus reducing the quality of care given to children; deterioration in the home environment, leaving families without adequate resources for basic necessities; and a combination of multiple stressful events leading to child maltreatment, with income loss being just one of these events.
Children removed from the care of their parents represent the most severe cases of child maltreatment and decisions about the threshold for out-of-home care placement are likely to be affected by other factors such as local practice culture (Forrester et al., 2021). There is, however, a dearth of theory development in connection with interventions in this domainthat is, how increasing or decreasing family poverty might affect the rate of children in out-of-home care. The focus of this paper is therefore on out-of-home care rather than all types of child maltreatment or child welfare system involvement, as the mechanisms and system level contexts that influence whether a child enters state care, if family income changes, may differ from other levels of child maltreatment. However, untangling the differences in mechanisms between poverty and out-of-home care and poverty and child maltreatment may be difficult, since most children are placed in out-of-home care due to maltreatment (Bywaters et al., 2022).
It is familiar territory for social workers to try and meet the material and practical needs of families. Assessing for the provision of material help was one of the founding elements of social work, stretching back to the 19th century (Pierson, 2011). In England and Wales, where the authors of this review are based, section 17 of the 1989 Children Act empowers local authorities to provide financial assistance to improve child welfareeither cash-in-hand or other kinds of payments to meet short-term needs for housing, food and other necessities. Yet it remains the case that many of the social work practice models developed in the UK and elsewhere over recent years have focused more on the importance of building relationships and of therapeutic forms of supportespecially those designed to change the behaviour of parents -than they have on poverty and material need . Morris et al.'s (2018) research in England and Scotland found that poverty had become "the wallpaper of practice: too big to tackle and too familiar to notice" (p.370) and rarely the target of social workers' interventions. There is, however, some renewed and recent interest in reclaiming a role for poverty focused social work interventions. For example, Krumer-Nevo (2020) has developed a poverty-aware social work approach, designed for implementation by individual practitioners, albeit with strong support at an organisational and policy level.
At a national level, tax and benefit systems directly affect family income, with possible implications for child welfare (Cancian et al., 2013;Pelton, 2015). Various countries including Canada, Finland and Spain have trialled variations of universal or basic incomes (Forget, 2011;Kangas et al., 2019;Arnold, 2020), whereby citizens receive guaranteed and regular sums of money, irrespective of personal circumstances. In contrast to this, welfare reforms that reduce benefit entitlement reduce the income of many families in poverty, which could impact on child welfare, including the likelihood of children being in out-of-home care (Wildeman & Fallesen, 2017).
The purpose of this paper is to progress the theorising of the connection between changes to family income and the rate of children in out-of-home care. To this end, this rapid evidence assessment reviews ten papers identified in an earlier scoping review  as concerning the effect on out-of-home care rates of changing a family's income. The focus of this review is on the evidence of the mechanisms and moderators of family income change interventions from these papers, in other words, to help understand the ways in which direct and indirect changes in family income may interact with personal and social factors to reduce or increase the rate of children in out-of-home care.
Interventions that change a family's income are defined here as any policy or practice which directly or indirectly increases or decreases the amount of money available to a family. These include approaches which increase family income via housing subsidies, direct cash assistance (including as part of wider programmes of wrap-around support) or the provision of goods that increase a family's material wealth. There were also examples of welfare benefit reductions which resulted in less money being available to families. The paper's main aim is to articulate an outline of the most important components of interventions that change family income, the key ways that these components work (mechanisms), and the contexts in which they work best (moderators).

Methods
A rapid evidence assessment method was used (Thomas et al., 2013) to synthesise evidence from ten studies identified in a previous scoping review as looking at the effect of family income change on out-of-home care. The full details of the scoping review searches can be seen in the report , but in summary, eighteen databases were searched, as well as grey literature, along with consultation with experts in the field to identify further relevant studies. Studies needed to include evidence of effect on rates of out-of-home care entry or reunification and evaluate an intervention which was defined as a disruption to the system that increased or reduced a family's income. However, the focus of this paper is not on the effect of the intervention, but how it works. The paper focuses on out-of-home care, which usually indicates more serious child abuse and neglect, yet is easier to measure than the contested construct of child maltreatment. Included studies of out-of-home care include those focusing on care entry and also reunification. Although we recognise that there are differences between entry to care and reunification, both relate to whether or not a child is living away from their family home. Fifteen studies were identified initially showing effect, but only ten studies also identified evidence of the intervention's mechanisms and/or moderators. Therefore, as the focus of this paper, only ten studies were included; these are listed in Table 1. A brief summary of the effect of the interventions on care entry and reunification is provided (for a more detailed summary see Brand et al., 2019). As this is not a full systematic review, it has not included quality appraisal of effect results.
The main focus of the paper is a qualitative exploration of the pathways through which interventions that change family income can change the rate of children in out-of-home care. The qualitative evidence was collected from all sections of the retrieved papers, which includes the individual studies' presentation of theory related to the income change intervention, and discussions of hypothesised mechanisms, as well as qualitative data collection, for example interview data. A qualitative approach offers a richer understanding not only of whether an intervention worked to achieve an outcome, but for which families and under which circumstances (e.g. Pawson & Tilley, 1997;Pawson, 2006). The interventions studied in the included papers are complex, with inter-related components, and operate in a complex context that will change how they work for different families.
The rapid evidence assessment was based on the realist EMMIE framework (Johnson et al., 2015). This method helps us to ask not only what works, but also for which children and families and in which circumstances. EMMIE provides a pragmatic framework to capture, analyse and disseminate the type of evidence that is essential to decision makers under the following dimensions (Johnson et al., 2015): Table 1 Table of interventions and effects of the intervention on out-of-home care in the include studies (n = 10). Policies in the USA which called for parents whose children are placed in foster care to be ordered to offset some of the costs of that care. One of the goals of this system was to support and enforce non-resident parents' contributions to their children's financial well-being, however when children are placed out-of-home, generally government costs were offset rather than directly benefiting children.

Natural experiment Decreased reunification
Reunification Wells & Guo 2003;2006; USA The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 in the USA was designed to promote paid employment. Two features of this legislation were the elimination of the entitlement to cash assistance that had been available under the prior welfare policy and the restriction of cash assistance up to 60 months.
Longitudinal study using administrative data Decreased reunification E The overall effect direction and size of the effect of interventions that change family income on children being in out-of-home care; M The mechanisms through which interventions that change family income affect out-of-home care rates; M The moderators/contexts that change whether those mechanisms will 'fire' or don't fire; I The key sources of success and failure in implementing interventions that change family income; E The economic costs (and benefits) associated with interventions that change family income.
Social interventions are complex and context is important in determining outcomes, and as such require evaluative approaches that take this complexity into account (Pawson & Tilley, 1997;Pawson, 2006). Realist approaches were developed specifically to produce meaningful evidence for complex social policy decision-making. Realist review focuses on exploring the underlying theory about not just whether an intervention works (effect), but also how it interacts with people and place to bring about outcomes or not (e.g. Dalkin et al., 2015;Pawson & Tilley, 1997;Johnson et al., 2015). Family context (for example employment status, parental response to stressors, previous experience of services, a parents own parents' approach to managing family finances) will have a significant impact on whether or not an intervention that reduces family financial standing will result in an increased risk of neglect or abuse of the children in the family. Where context is key in determining outcomes, it is fruitful to apply the realist EMMIE framework (Bowers et al., 2018) to produce actionable findings for policymakers. This paper aims to examine the specific family circumstances that mediate the impact of a change in a family's income on the rate of out-of-home care. Following the method of systematic mixed-method EMMIE review (Johnson et al., 2015), evidence was summarised on effect (E), mechanisms (M), moderators (M), implementation (I) and economic impact (E), although this paper focuses mostly on mechanisms and moderators. A realist mechanism is the interaction between an intervention (such as change in income) and people's thoughts, feelings, attitudes, beliefs, perceptions. This is turn changes whether outcomes of interest, such as behaviour change, occur. The term moderator refers to the contextual factors that enable or inhibit whether mechanisms produce outcomes. For example, for parents with a previous history of nonprescription drugs (moderator), increased anxiety about money in response to a reduced income (mechanism) can lead to a change in parents' non-prescription drug-taking behaviour (intermediate outcome), which can change their behaviour towards or in relation to caring for their child (outcome). Evidence of mechanisms and moderators are presented together, as the activation of mechanisms is contextually contingent.

Data management, analysis, and synthesis
Evidence relating to mechanisms and moderators was brought together using a process of realist synthesis. A realist synthesis follows similar steps to a systematic review, but differs in that the search for evidence within systematically identified studies is purposive and theoretically driven; the process is iterative; multiple types of evidence can be used; and it focuses on why and in what ways the intervention works (Rycroft-Malone et al. 2012).
In the prior scoping review, included studies were read and relevant sections coded to E, MM, I and E using QSR International's NVivo 12 qualitative data analysis software (QSR International Pty Ltd, 2018). The pieces of coded mechanism and moderator information from the included studies were compiled in an Excel worksheet. Each coded section was re-formulated into "if-then" statements (see Pearson et al., 2015;Brand et al., 2019) to capture theories about how interventions that change family income can impact on out-of-home care rates. For example, IF homelessness was reduced due to housing subsidies, THEN this also led to a reduction in alcohol dependence, intimate partner violence and economic stressors (converted data exert from Shinn et al., 2017). Particular attention was paid to nuance in relation to which parents, families, and children are most likely to be affected by income-change interventions and why, and in which circumstances income-change interventions are most likely to affect children being in out-of-home care and why. From a realist perspective, an intervention that did not work as a whole for a significant portion of families to achieve a particular outcome (effect analysis) is still interacting with different families in different ways. Such effect analysis does not tell us about how the individual intervention components are or are not working towards the intended outcomes. Examining more qualitative findings from those studies, for example from process evaluations, can provide a fuller picture of what works about these types of interventions and for whom to help plan and deliver more effective services for families and children.
These if-then statements were then grouped into themes. These themes related to either key components of the intervention, key mechanisms through which the key components worked, or key moderators that affected whether and how the intervention worked for certain families. Themes were chosen by the two reviewers most familiar with the data extracted and coded with the resulting if-thens. The two reviewers identified themes independently and reached agreement via discussion. Final themes were those groupings subjectively considered to best capture what was most important in the evidence in relation to how the interventions work, for which families, and under which circumstances.
Groups of if-then statements were then consolidated ( Fig. 1; see also Pearson et al., 2015) into a smaller number of consolidated explanatory accounts, clarifying more about how the intervention works, for whom, and in which circumstances. Please note that the connections between the mechanisms are speculative and require further testing.
Consolidated explanatory accounts were expressed in diagrams and narratives, to present a programme theory of how interventions that change family income work. The consolidated accounts were shared with colleagues experienced in social work practice to identify gaps, add nuance, and support understanding of the most important "pieces" of the theory for practice and policy.

Study characteristics
In total, ten studies were included (see Table 1). Of these, six related to care entry and five to reunification (one study related to both). Nine studies were conducted in the USA, one in Denmark (Wildeman & Fallesen, 2017). All studies were published in peer-reviewed journals. Five studies evaluated interventions that increased a family's income and included: housing subsidies and multi-component interventions with an element of financial help, for example cash assistance or practical assistance (including clothing, sundries, material goods such as furniture) or access to other resources. The remaining five studies evaluated interventions that decreased a family's income and included welfare benefit reductions through policy changes and child support enforcement when a child is in foster care. See Table 1 for full list of interventions and their descriptions.

Intervention effects
Of the ten included studies, two were randomised controlled trials, seven used a non-randomised comparative design (quasi-experimental, longitudinal studies using administrative data, or natural experiment), and one was a single-arm quantitative study. Although this paper's main purpose is to theorise how interventions could work and for whom, a brief summary is given here of the overall effect results. The reviewed papers suggest that welfare reforms and child support enforcement that decrease a family's income have either no effect or a negative effect on out-of-home care entry and reunification (i.e. an increase in care entry or decrease in reunification).
The multicomponent or housing subsidy interventions which all increased family incomes had either no effect or a positive effect on outof-home care entry and reunification (i.e. a decrease in care entry or increase in reunification). In summary, the included evidence suggests that interventions which directly or indirectly improve the financial situation of families may help reduce the rate of children in out-of-home care. See Table 1 for more detail on study design and intervention outcomes.

Mechanisms and moderators of interventions that change family income
Four key pathways were identified from the studies. Three of these have an impact on the rate of children in out-of-home care directly via changes in family income. The fourth pathway results from the provision of practical assistance leading to indirect changes in family income.
The first three of these pathways (A, B, C in Fig. 2) change the rate of children in out-of-home care mainly through the impact of a change in family income on the risk of neglect or abuse. The first is through impact on maternal employment, the second is through changes in the family home environment, including living standards, perceived economic stability and ability to sustain housing, and the third pathway relates to the risk of homelessness.
The pathways from the different intervention to the rate of children in out-of-home care mainly operate through their effect on the risk of neglect or abuse (key intermediate outcome; Fig. 2). The ways in which these pathways have their impact is described in detail below. Neglect or abuse (or the perceived risk of this) will significantly affect whether a child is taken into care or returned home. In addition to the pathways described below, the evidence suggests that a child's risk of neglect or abuse is mediated by whether a family is struggling at the economic margins of society and is therefore more easily overcome by illness or unemployment (Huebner et al., 2012). Periodic ongoing support can help these struggling families and reduce this risk (Huebner et al., 2012). Whether or not withdrawal of welfare payments is voluntary is also suggested by this evidence to have an impact on child risk of neglect or abuse (Fein & Lee, 2003).

Pathway A: Impact of a decrease in family income on maternal employment (especially for families struggling on the margins)
In the identified studies (Fein & Lee, 2003;Wells & Guo, 2003, 2006, only maternal employment was considered (and not, for example, paternal employment). Of course, not all families known to social services are maternal lone parent families, but the evidence we summarise here refers only to the specific challenges facing these households.
For families in economic hardship, especially single mothers, income reductions make it more likely that the mother will seek employment (Fein & Lee, 2003). Whether the mother wants to seek or maintain employment can help determine whether there are positive or negative effects on her ability to meet her child's needs and avoid the need for out-of-home care (Fig. 2).
A positive effect of maternal employment can be improvement in the psychological wellbeing of mothers, resulting in feelings of self-esteem, autonomy, or success, all of which can positively affect their ability to meet their child's emotional needs and improve chances of reunification or reduce risk of out-of-home care entry (Fein & Lee, 2003).
However, maternal employment can also have a negative impact on whether a mother is able to supervise and/or take care of her child adequately, due to being away from the home at work. This negative impact is more likely where a single mother secures employment that is low wage, unstable, and/or requires evening or early morning work (Wells & Guo, 2003, 2006. This impact is mediated if a mother is in a situation in which she is able to arrange alternative care for times when she is at work (Fein & Lee, 2003). This will depend on her individual circumstances, such as relationship with and proximity of family and friends. Being in circumstances in which she is able to safely supervise children or not will directly affect their risk of neglect and need for outof-home care.
Maternal employment can also have a negative impact on a mother's level of stress, depression, and coping capacity, which in turn will impact whether she is able to supervise and care for her child, and the child's risk of maltreatment (Fein & Lee, 2003). Both of these will affect the need for the child to be in out-of-home care.
The summarised evidence suggests that it is not only an actual change in income (material and financial) but also the perceived threat of a change in income that will impact whether a single mother will seek employment (Fein & Lee, 2003).

Pathway B: Impact of an increase or decrease in family income on family home environment
The identified evidence suggests that increases or decreases in a family's income can change the family home environment, including living standards and economic stability (Fig. 2). The home environment and economic stability of a family can affect child safety and whether a child's material needs are met (Fein & Lee, 2003). Both of these can increase the need to enter out-of-home care (Thleman & Dail, 1992) and reduce rates of reunification (Cancian et al., 2017).
A family home environment is complex and involves many potential mechanisms and moderators that will have an impact on the need for a child to enter care. The papers summarised here offer limited detail as to what these mechanisms and moderators are, beyond general descriptions of how the home environment is affected by income changes, and how this in turn can change the risk of child abuse or neglect.
The studies suggest that an intervention which increases a family's income can improve parental mental health. Improved mental health can make it easier for a parent to care for their child and meet their needs, and may reduce their substance misuse (Wells & Guo, 2006). These changes will improve the chance that a parent can safely care for their child at home.
The evidence suggests that negative influences on the family home environment, such as a decrease in the family's income, can affect the mental health of a parent (Fein & Lee, 2003;Wells & Guo, 2003), which can influence whether they are able to meet their child's material and emotional needs (Huebner et al., 2012;Wildeman & Fallesen, 2017;Fein & Lee, 2003;Wells & Guo, 2003). Recognition of a child's unmet needs can further increase stress and depression in a parent, and the resulting reduced coping capacity may lead to substance misuse (Fein & Lee, 2003). Substance use is a factor which can further reduce whether a parent can meet their child's emotional and basic needs and/or whether they can safely supervise their children (Fein & Lee, 2003).

Pathway C: How a decrease in family income can lead to family losing their home and an increased risk of abuse or neglect
A more substantial impact of interventions on family material and financial wealth is the risk of homelessness (Fig. 2). If an intervention reduces a family's income to the extent that they lose their home, this can lead to family separation, parental substance misuse, and the need for a child to go into out-of-home care (Shinn et al., 2017). To avoid homelessness, a parent might co-habit. This can ease financial difficulties and increase the ability of the household to provide adequate supervision of a child (Fein & Lee, 2003). However, it can in some cases also increase the risk of intimate partner violence (where co-habitation is with an intimate partner) or potentially introduce new sources of maltreatment or abuse by another adult sharing the child's home.

Pathway D: The indirect effects of practical/financial assistance
If an intervention meets a family's basic needs through providing practical assistance, families can focus on other aspects of life (Fig. 3). This can be particularly useful in multi-component interventions as it allows families to engage and benefit from other aspects of the intervention, such as training (Fraser et al., 1996). Providing practical assistance increases the time parents spend with workers and discussing basic needs demonstrates a more "human side" to social workers and that such tasks are not beneath them (Ryan & Schuerman, 2004). Both these mechanisms enable increased trust between the worker and the family and enable discussions about money management (Ryan & Schuerman, 2004). This helps the family improve their home environment, living situation, and economic stability (Fraser et al., 1996). These changes will improve their ability to care for their child safely at home.
If changes to welfare benefits policy increase the interaction between welfare agencies and child protection services then improvements in the detection of child maltreatment may also be possible (Fein & Lee, 2003). Parenting behaviour itself can also have an impact on a family's wealth in a welfare regime where sanctions are placed on benefits due to parents not meeting agreed targets set in court (Fein & Lee, 2003). In theory, if they encourage positive behaviour change, these financial measures can also result in parents being more likely to meet a child's needs and reduce their need for out-of-home care.

Discussion
This rapid evidence assessment summarises evidence from ten studies that evaluate interventions that change family income, in terms of the process of how they work and for which families, on the risk of children needing out-of-home care. The identified interventions that change family income included those which increased family incomee.g. housing subsidies, cash assistance and provision of goods (as part of a wider programme of wrap-around support); and policies which reduced family incomee.g. welfare benefit reduction. The evidence we have summarised indicates that interventions that directly or indirectly improve a family's financial situation may potentially help reduce the rate of children in out-of-home care. Evidence from the wider literature supports these findings. Four recent systematic reviews show low family income, income losses and economic inequality directly and indirectly effect child outcomes, such as child health and development, maternal mental health, parenting, the home environment (Cooper & Stewart, 2021), and child maltreatment (Conrad-Hiebner & Byram, 2020;Hunter & Flores, 2021;Bywaters et al., 2022), all of which can increase the likelihood of out-of-home care.
There are four main pathways through which a change in family income can alter the risk of out-of-home care. One is through the need for mothers to seek employment in response to a reduction or perceived likely reduction in the family income. The second is through changes to the home environment itself. Both of these pathways involve changes to a parent's level of stress (in the included studies, this was always related to mothers), mental health problems and substance misuse. This can change whether a parent is able to care for their child and can lead to neglect and abuse, and ultimately the need for out-of-home care. The third pathway is income reductions leading to homelessness, which increases the risk of child harm. The fourth pathway is practical assistance enabling a range of positive benefits, including familiarity and perception of helpful caring by the social worker or other professional, and building trusting relationships between parents and workers. A trusting relationship can form the basis for a family to engage with and access other services and other types of support from the worker providing practical assistance.
The evidence also suggested that families already on the margins of economic viability were more prone to the consequences of economic change. In interventions that change family income, relatively small changes in income or resources are likely to make little difference to families who are well resourced, but may be detrimental to families already on the edge of destitution. For instance, the amount of debt or savings a family has could change the effect of an intervention on a family's circumstances. This is particularly relevant given the increase in child poverty in the UK in recent years (Bennett et al. 2021). The nature of the income change is also likely to be more consequential for these families. As the programme theory eludes, the threat of a change in income or the risk of homelessness was just as pervasive to family behaviour and state of mind as actual changes to family circumstances. This suggests that it is not simply the amount of income that matters, but whether that income is secure and predictable or changing, and how long low income persists.
Child maltreatment was a key intermediate outcome between poverty and out-of-home care, suggesting that the mechanisms between poverty and out-of-home care and poverty and maltreatment are very similar. Bywaters et al. (2016) distinguish the direct effects of poverty on child maltreatment from the indirect effects. In this programme theory, the pathways A, C, and D outlined above could all be said to relate to indirect effects of family income change on out-of-home care, although D is more indirect than others. Pathway B is the most direct pathway, with changes to family income affecting the economic stability of a family which can affect whether a child's material needs are met. The impacts of income change on employment, home conditions and housing are again more immediate, although it is via the influence of these factors on parental well-being that rates of child maltreatment are affected. The fourth pathway is more indirect still, wherein parents come to trust child welfare professionals because they value the material help provided, leading to a consequent reduction in risk to children.
Some of the pathways outlined in this review are focused only on mothers. This is because the studies themselves referred only to mothers. It is of course important to consider the effect of income change on all relevant family members, and not to make the mistake of assuming that only mothers are or should be responsible for childcare-related issues. In fact, Paxon and Waldfogel (1999) found that fathers' employment status can affect levels of child maltreatment, with unemployed fathers having poorer self-esteem and mental health, therefore lacking the emotional reserves to care for their children. Lindo et al. (2018) also found that child maltreatment decreased with male employment, but increased for female employment. They hypothesise that one of the mechanisms responsible for this difference is time spent with children. Statistically children are more likely to be abused by males than females and if the father is unemployed, the child may spend more time with them, thus increasing their risk of abuse. Whereas if a mother finds employment the child may either spend more time with the father, with another care giver (who may or may not have a higher propensity to be abusive), or alone. Social work with children and families has a history of being heavily focused on work with mothers, with women in families often being expected to take primary responsibility for protecting children, regardless of whether or not a male partner is also present, and this can reinforce prejudices connecting father absence and mother blaming (Strega et al., 2008).

Gaps in the programme theory
Although pathways A to D provide some insight from intervention studies into potential mechanisms and moderators, it is important to acknowledge that these pathways do not explain every step of the process from income changes to children being in out-of-home care (or not being reunified home again). It is worth noting that these gaps may be addressed in other literature; these are gaps found based on our search, which is limited to the outcome of out-of-home care. Some gaps in the process are hereby noted, with numbered points represented on Fig. 4 Homelessness could add to stresses in a variety of other ways than those described, such as affecting children's physical, emotional health, development, and education; a lack of safety and security; shame and social stigma; and the impact on parents' mental health. Displacement from the family's local area may also remove sources of informal and formal support (Cross et al., 2021). 4. The studies lacked depth regarding what it is about the home environment that can lead to substance misuse and why substance misuse can reduce parenting capacity. For instance, are there moderators that make this dynamic better or worse for some people than others? 5. The included literature failed to delve further into why changes in parent mental health, wellbeing, and coping capacity could reduce the parent's ability to meet their child's material and emotional needs. They explain little about the internal processes that could contribute to this outcome and whether the effect of parent mental health is better or worse for some people or in different situations. 6. The studies suggested that if a child's emotional and material needs are not met this can increase the child's risk of abuse and neglect. However, the studies lack detail as to why and whether this risk is modified in different circumstances. For abuse in particular, it is likely that there are multiple factors at play (e.g. guilt, stress etc.). 7. There was a considerable gap regarding the risk of intimate partner violence as it is only addressed under pathway C. It could be argued that the risk of intimate partner violence may also fall under pathways A and B due to the risk of increased stress, mental health problems and substance misuse.

Gaps in pathway D
The included studies did not reflect on the complexity of providing practical/financial assistance to families. As Saar-Heiman and Krumer-Nevo (2021) discovered, providing material assistance is challenging for workers due to conflict and tension around transparency, decisionmaking, surveillance and deception, and the political nature of providing such assistance playing a key role in the ways it is implemented.
Overall, the included papers lacked depth about which children are particularly affected by the mechanisms identified, for example by age, ethnicity, gender, or disability. These are therefore areas which need further investigation.

Study limitations
This concise and rapid evidence assessment does not constitute a full systematic review of the literature about interventions that change family income and has not included quality appraisal of effect results. The evidence assessment theorises how interventions that change family income affect the rate of children in out-of-home care, but these links are not empirically tested. It should be noted that there are several multicomponent interventions where the particular impact of help with family income cannot be isolated from other elements. The scoping review from which studies were identified specifically considered interventions that change family income that are evaluated in terms of their effect on out-of-home care rates. Wider literature describing income change interventions more generally, and papers evaluating income change interventions regarding a different outcome, were not included in the review. We have been made aware that our search did not retrieve every single relevant article, possibly because of the search terms used (full details of the search can be found in the scoping review report -Stabler et al., 2022). The gaps identified in the programme theory may be addressed in other literature and could be due to the restrictions of the initial scoping review. Including wider literature could potentially enrich the programme theory.
Further reviews and original empirical studies are needed to further describe and test intervention pathways, so that the effects of material support for families are well evidenced, with the aim of maximising positive benefits for parents and children.

Conclusions
Although this review has not included a full synthesis of effectiveness results and quality appraisal of studies, there is some evidence to support a direct increase to family finances as an appropriate intervention to prevent maltreatment and thus, out-of-home care. There is certainly no evidence in favour of a reduction in family finances, with some studies showing this can increase rates of children in out-of-home care. The findings of this paper could be seen to support recommendations from others, for a universal child allowance (Slack & Berger, 2020;Pelton, 2015). It would be ideal to increase family income for all people in poverty through changes to welfare benefits, reducing economic inequality and bringing the social benefits found in more equal societies (Wilkinson & Pickett, 2010). However, there is nonetheless a more local role for children's services to provide extra income for families who are struggling, as a preventative measure. Given the cost of foster care and residential care, this could be money well spent if it reduces the need for children to be in out-of-home care. Westlake et al. (2022) conducted pilot evaluations in England which devolved budgets to social workers. They found that social workers were creative and diverse in the way they used the budget to meet families' needs and when done in a timely manner could lead to improvements in home conditions, relationships and behaviour, thus reducing the need for children to enter out-of-home care. Social workers have an important role in helping at-risk families in poverty to maximise the material help available to them. Since some evidence suggests that responding to poverty is not central to the culture of contemporary front-line practice, a renewed emphasis is needed on poverty-aware practice and the promotion of opportunities for direct material help to families.
To date, although theory regarding the relationship between poverty and child maltreatment is well established, the process of how to use material resources to help families to reduce the risk of out-of-home care has been under-theorised. This paper, which is based on a limited number of studies, provides an initial theory. More theorising, and intervention testing is needed, as part of the endeavour to better understand the link between changes to family income and out-of-home care.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.