Edinburgh Research Explorer Socioeconomic deprivation and social capital in kinship carers using a helpline service

Kinship care is the preferred alternative for children who cannot remain with their birth family, maintaining birth family links and continuity in other parts of their life. However, kinship care has also been associated with risk factors including lack of support, difficult contact with biological parents, adverse childhood experiences for the child, carer stress and financial difficulties. Using routine data from a kinship care helpline service, this study employed a mixed-method analysis of the association between socioeconomic deprivation and risk factors reported by kinship carers and explored social capital in kinship families. Findings indicated common risk factors experienced by kinship families regardless of deprivation level. However, certain risk factors were reported more in areas of high or low socioeconomic deprivation. Mapping the social capital of kinship families indicated that kinship families connect more with relatives than other types of informal, semi-formal or formal support services. However, these relationships are often problematic. The most prominent obstacle to social capital growth in kinship carers was financial difficulties. Our findings suggest that kinship carers may use support services differentially according to deprivation level, and socioeconomic deprivation may influence building social capital in kinship families.

The demographics of kinship carers suggest wide heterogeneity (Rubin, Springer, Zlotnik, & Kang-Yi, 2017). Inchaurrondo, Bailón, Vicente, Tió, and Bolós (2015) established a categorical risk framework describing the presence of risk factors in kinship care across three domains: child, biological family and kinship family. The framework reflects the ecological and interactive nature of risk factors associated with kinship care, with their cumulative effect indicating worse outcomes (Raviv, Taussig, Culhane, & Garrido, 2010).
UK-based foster carers typically actively choose to pursue foster care as a vocation; are assessed, trained and supervised by social services or contracted private providers; and are financially compensated for their caring responsibilities. Kinship carers by contrast are more likely to become carers because of the immediate needs of a family member, and despite government legislation (e.g. Scottish Government, 2007), generally receive little if any professional or financial support (Zuchowski, Gair, Henderson, & Thorpe, 2019). It is estimated that in Scotland specifically, 76% of kinship care arrangements are informal, with no oversight or support by statutory services (Kidner, 2016). In this context, kinship carers are often living in poverty (Aldgate & McIntosh, 2006), non-married, less educated, female and older (often grandparents; Ehrle & Geen, 2002;Wu et al., 2015). Furthermore, poor health, lack of support, ceasing employment and isolation are common (Bachman & Chase-Lansdale, 2005;Selwyn, Farmer, Meakings, & Vaisey, 2013).
Research using UK Census data indicated that 1 in 43 children were living in kinship care in the poorest 20% of neighbourhoods, compared with 1 in 200 children in the wealthiest 20% of neighbourhoods (Nandy & Selwyn, 2013). Research to date has treated socioeconomic deprivation as a confounding variable, controlling for its effect to demonstrate the consequences of other factors such as being in foster care (Turney & Wildeman, 2017) or physical health outcomes (Bellis et al., 2014). Given the high convergence of socioeconomic deprivation and kinship care, more contextualized research may be needed to understand their respective roles in child and carer outcomes.
These issues may be influenced by the strains felt to fulfil parenting responsibilities unexpectedly (Jones et al., 2011) and meeting the financial costs of additional dependants (Broad, Hayes, & Rushforth, 2001). Kinship carers often face challenges that foster families may not necessarily experience, including legal issues around guardianship, disputes with birth parents, other caring responsibilities, resigning from jobs and financial hardship (Boetto, 2010;Selwyn et al., 2013). Managing relationships with birth parents can cause increased stress to the kinship carer (Sen & Broadhurst, 2011), which may subsequently affect the well-being of the kinship child. Whilst some findings indicate that more contact with birth parents increased the child's well-being (Metzger, 2008), higher rates of informal and unpredictable contact (Aldgate & McIntosh, 2006) may adversely affect the child via prolonged exposure to adverse experiences or failed contact arrangements (Burgess, Rossvoll, Wallace, & Daniel, 2010). Contradictory findings for the effects of birth family contact exist in foster child samples as well (McWey, Acock, and Porter, 2010;Rich, 2011), reflecting the need for further research to understand the complex dynamics that exist between kin and non-kin alternative caregivers, children and birth families.
Whilst recent reports show that children in kinship care may present with fewer behavioural problems and better well-being in comparison to non-kin foster care (Winokur et al., 2018), findings highlight that placement into care itself has been associated with higher rates of depression, anxiety, attention-deficit hyperactivity disorder and conduct problems (Turney & Wildeman, 2016). Additionally, despite some reports indicating that kinship care is less stigmatizing than foster care (Messing, 2006), children can be taunted for not living with their parents . Like children in foster care, those in kinship care may have experienced adverse events in their lives, such as parental substance abuse, domestic violence and mental illness Turney & Wildeman, 2017), often the driving factors in kinship care arrangements.
Despite the clear difficulties faced by kinship carers, they often receive less support and access to services compared with foster carers including social worker supervision, respite care, training or support groups (Berrick et al., 1994). Moreover, when carers do receive support from formal services, perceptions of said services have been far from satisfactory . Studies exploring kinship carers' support systems indicated that families were "just getting by" (p.14, Blair & Taylor, 2006) with limited help from statutory services (Blair & Taylor, 2006;Selwyn et al., 2013). Carers may feel wary of contacting social services for support in case they are deemed unsuitable carers, or if struggling with depressive symptoms may lack the drive to seek help . Kinship carers often report having a small network of friends and family, including a loss of contact with their existing networks due to undertaking the kinship carer role , thus missing out on the potential benefits of informal social support on parenting stress (Gleeson, Hsieh, & Cryer-Coupet, 2016;Goodman, Potts, & Pasztor, 2007).
Formal and informal social support networks are central to social capital, which refers to connections between individuals with shared norms and values that facilitate trust and co-operation within or among groups (Healy & Côté, 2001). Bourdieu (1986) posited that social capital plays a role in social inequality by giving the upper and middle classes exclusionary power through better access to resources.
The importance of financial and material resources should not be overlooked; without an adequate level of resources to meet basic needs, carers may not be able to tap into their networks (Winkworth, McArthur, Layton, & Thompson, 2010). Social capital may act as a buffer against socioeconomic inequalities in health (Uphoff, Pickett, Cabieses, Small, & Wright, 2013), of relevance given the prevalence of physical and mental ill-health in kinship carers (Bachman & Chase-Lansdale, 2005).
Higher social capital is thought to benefit the whole family. In intact families, parental social capital has been found to be positively associated with parenting efficacy (Jang, Hessel, & Dworkin, 2017), which in turn predicts socio-emotional development (Shumow & Lomax, 2002), reduced mental health difficulties and positive educational outcomes in adolescents (Rothon, Goodwin, & Stansfeld, 2012). It is also positively associated with improved health and reduced risk of child abuse (Healy & Côté, 2001;Scrivens & Smith, 2013).
Current research presents a relatively linear perspective of kinship care, studying a series of independent issues and their subsequent influence on child outcomes. Enhancing the understanding of interactions between influential variables such as socioeconomic deprivation with cumulative risk and adverse experiences has been neglected (Nandy & Selwyn, 2013;Turney & Wildeman, 2017). Furthermore, social capital theory has not been applied to children in care, except within the context of education (Perez, 2010).
One way in which kinship families can receive support and facilitate social capital growth is through third-party organizations and charities. Such organizations provide a variety of services such as befriending, practical support (e.g. food banks) and family support at varying levels of intensity through individual, group-based or remote interventions. Helpline services fall into this latter category. Helpline services can offer crucial support and advice for families. Helplines increase accessibility by reaching larger populations that may not otherwise access services (McKenzie, Williamson, & Roberts, 2016), as they are often perceived as less stigmatizing (Boddy, Smith, & Simon, 2005). As issues surrounding support are common in kinship care families, these services may hold particular relevance. The present study used helpline call logs to delineate differences surrounding

| METHODS
Service data routinely collected by call handlers for a kinship care helpline were analysed following a data-sharing agreement between the helpline service and the University of Edinburgh, ethical approval granted by the latter. The helpline provides a listening service, support and advice and signposting to other services. Users of the helpline service can choose to call, text or use a webchat to access the service.
A mixed-method inductive cohort design was used to explore risk factors and help-seeking in kinship families.

| Sample
The initial sample included all calls through the helpline over a period of 1 month (1 April to 1 May 2017), including those provided through a call-back service to kinship carers. This resulted in 218 interactions with 146 callers over the identified period. For the sample in RQ1, repeat callers were combined into one case. Similar exclusion criteria for samples in RQ1 and RQ2 included if the caller did not provide a postcode or if callers did not identify their kinship carer role. Callers were further excluded from the sample in RQ1 if the child was below age 5 or over age 17 to create homogeneity in the sample in relation to access to universal educational system support. This avoided age becoming a false confounding variable through augmented supports routinely offered to families caring for pre-school children and the cessation of support once children leave school. In theory, access to education-based resources should be a stable variable for 5-17 year olds, and any differences that emerge would be noteworthy. Callers were excluded from the analysis if they were calling in a professional capacity (e.g. social workers, researchers and youth workers). The final sample in RQ1 consisted of 50 callers and 119 interactions with 70 callers in RQ2. See Figure 1 for a flow chart of sample inclusion.
For both samples, most callers were female (RQ1, 88%; RQ2, 87.1%). The callers were typically grandparents to the kinship child, F I G U R E 1 Flow chart of participant inclusion for RQ1 and RQ2 followed by aunts. Both samples also comprised a small number of uncles, siblings and non-relative carers. See Table 1 for full breakdown.

| Measures
Demographic data (caller gender, relationship to the child and postcode) were collated along with basic call details. These included the nature of each contact made, the guidance given, action plan agreed with the caller and which agency callers were referred or signposted to.

| Socioeconomic status
In order to measure socioeconomic status, postcodes were matched against the Scottish Index of Multiple Deprivation (SIMD; Scottish Government, 2016) to produce a ranking. This identified carers who live in different areas of socioeconomic deprivation, calculated into quintiles, each representing 20% of the data zones. Indicators of socioeconomic deprivation, according to SIMD, are ranked on seven domains: employment, income, crime, housing, health, education and access.

| Risk factors
Previously identified risk factors have been diverse, and there was concern in this study to consider these broadly and comprehensively but within an a priori framework, recognizing that callers may only disclose risk factors pertinent to the immediate issue. We selected two theoretical frameworks which measured environmental risks and those relating to the child and kinship carer specifically.

provides eight individual indicators of the presence of
ACEs and is widely used in research with populations identified as having disrupted upbringings. These indicators include parental separation, parental incarceration, domestic violence, household member mental illness, household member substance abuse, emotional abuse/neglect, physical abuse/neglect and sexual abuse. These were coded as present/absent in each record and were then summed to produce a total exposure rating of number of ACE types.
The Risk Factor Framework has previously been used in kinship care (Inchuarrondo et al., 2015). This consists of three main categories each with 5-6 sub-categories (see Table 2). Within each sub-category, Inchaurrondo et al. (2015) identified a number of individual risk items.
These were coded as present/absent and total scores for each category calculated from this. A "collateral" score was generated by summing the categories together. For the purposes of the current study, an additional risk factor found in previous research to be prevalent, "parental death," was also used in reference to biological parent/s risk factors.

| Data analyses
A conversion mixed-method analysis was employed using IBM SPSS v.21 and NVivo11. To test for associations between socioeconomic deprivation and risk factors, correlations and t tests were conducted between the SIMD score, call times, risk factors and number of ACEs. and α error probability = 0.05 and power = 0.8 was n = 34. A qualitative framework of thematic and mixed-method analysis, employing a deductive approach (Braun & Clarke, 2006), was used to identify and establish themes of risk factors and experiences of the kinship family in relation to SIMD. A small pool of data consisting of 20 cases was extracted, and the coding technique (a mixture of inductive and deductive coding) finalized main themes and sub-themes, whose occurrence was named "frequent" when they occurred over five times.
To test RQ2, after systematically identifying and coding the core themes relating to social capital using the procedure suggested by Braun and Clarke (2006), data were transformed into quantitative data, also using a frequency count of the common themes. Research interpretation or bias (Charmaz, 2006) was minimized through regular reflective group discussion to reach a shared understanding of the data, Themes were then grouped, and further codes were added or combined as the call logs were reviewed in an iterative process (Braun & Clarke, 2006) to avoid duplication. Definitions of social capital used by Bourdieu (1986) and Coleman (1988), that is, as the networks and relationships through which knowledge, support and resources are sourced, and which are in keeping with a broader consensus on the social capital construct (Enfield & Nathaniel, 2013) informed the analysis. Social capital was coded across four themes defined by Scrivens and Smith (2013): social and support networks, personal relationships, trust and cooperative norms and civic participation. Social and support networks were operationalized as people they know, networks they belong to, services they utilize and the outcomes, or potential outcomes, of these (i.e. information and advice and emotional, material, practical, financial, intellectual and professional resources and support).

| RESULTS
RQ1: Do levels of socioeconomic deprivation differentiate use of a helpline and reporting of risk factors in kinship carers?
One or more identified risk factor relating to the individual, biological parent/s and/or kinship family was identified in 82% (n = 41) of cases.
One or more ACEs were identified in 50% (n = 25) of total cases (M = .92, SD = 1.10). The mean SIMD raw score was 2,577 (SD = 1,928.96), with 36% of the sample categorized as residing in the 20% most deprived area of Scotland and 8% in the 20% least deprived areas (see Table 1). The correlations between total number of risk factors and SIMD score for the individual, biological and kinship are reported in Table 3 The analysis was slightly under-powered, and the findings should be interpreted cautiously, but is mitigated by the large effect sizes found.

| Thematic analysis of the risk factors in areas of low and high socioeconomic deprivation
From the total sample (n = 50), 58% (n = 29) of the sample was categorized as experiencing high socioeconomic deprivation, that is, were in the bottom 40% on SIMD rankings and 42% (n = 21) defined as experiencing low socioeconomic deprivation, that is, were in the top 40% on SIMD rankings. Using a deductive thematic approach with these two sub-groups, themes were identified using concepts derived from financial difficulties, the ACE and Risk Factor Frameworks. The qualitative differences in sub-themes reported by kinship carers living in areas of high or low socioeconomic deprivation are mapped in Figure 2.

| Financial difficulties
Financial difficulties were a prominent theme throughout the data.
Sub-themes of seeking financial advice/information were found.
Other sub-themes related to more serious concerns, including financial worries and immediate financial deficits, indicated actual financial strain on the kinship carer. Exploring the associations of these themes on differing levels of socioeconomic deprivation, it emerged that the two more severe financial-related themes were more frequently reported in areas of high socioeconomic deprivation.

| Kinship carer
Themes surrounding the kinship carer such as those related to kinship family characteristics (age-related differences, health and support) were identified. Risk factors surrounding support were highly prevalent across the sample. In several cases where lack of support was identified as a theme, callers appeared to be using the service to discuss their own problems, suggesting a lack of local alternative supports. Older callers described more difficulties relating to lack of social support. This and mental health problems were noted more frequently in those from areas of high socioeconomic deprivation.

| Child
Main themes of child's behaviour, emotional well-being, health and school emerged. Several sub-themes surrounding aggression and risk-taking behaviour emerged, more commonly reported in areas of high socioeconomic deprivation. A recurring connection between behavioural problems and financial struggles was noted.
Themes relating to the child's well-being were more frequently reported in areas of low socioeconomic deprivation, specifically relating to anxious and depressive symptoms. Other subthemes relating to the health of the child such as physical and learning disabilities were equally identified across high and low socioeconomically deprived areas, as were school truancy and difficulties at school. All connections were coded within the four themes. Callers were found to be connected with a broad range of formal, semi-formal and informal networks (Table 4). Carers were more connected with relatives than any other informal, semi-formal or formal support network.

| Biological family
Personal relationships with family, friends, neighbours and colleagues were explored according to feelings about personal relationships, the frequency of contact and the proximity of these networks to the family's residence. Callers reported overwhelmingly more negative (78.9%) than positive personal relationships (21.1%).
Trust and cooperative norms were operationalized as "trust and confidence in organisations, services, and institutions." There was an almost equal split in high and low trust, with 28 and 30 references, respectively ( Civic participation was defined as "contact with/writing to a member of parliament or local councillor, attending a community consultation or public council meeting." This form of social capital was very low in this sample, with only one caller who mentioned contacting their MP for support.

| What are the factors facilitating or impeding social capital growth in kinship families?
Outside of the facilitation provided by call-handlers, potential barriers to social capital growth were found to be wide-ranging. However, the most common barrier to social capital growth reported by kinship carers was financial stress or hardship (32 instances). This was followed by loss of an existing network (20 instances), or physical or mental ill-health (16 instances). Additionally, a lack of support received when the carers asked acted as a barrier to social capital growth (12 instances). Personal reasons (9 instances) and practical barriers such as older age, transport and lack of time due to caring responsibilities (9 instances) were identified. Finally, a lack of internet (2 instances) or an isolated local area (1 instance) further impeded social capital growth. seeking financial advice. However, there were also differences in risk factors according to socioeconomic grouping including acute or serious financial difficulties, perception of children's behaviour and wellbeing and abuse/neglect of the child, the latter being further supported through correlational analysis. This may indicate that the way in which kinship carers use the helpline service differs according to level of socioeconomic deprivation. Mapping of social capital indicated that carers' main social networks were relatives; however, relationships were often negative. Notably, financial difficulties were the most common barrier to social capital growth for carers. This highlights that socioeconomic deprivation not only influenced the type of risk factor reported but could also be a risk for impeding social capital growth in kinship carers.

| DISCUSSION
We corroborated previous research identifying similar factors surrounding the child, kinship carer and biological family Inchuarrondo et al., 2015;Jones et al., 2011). A main theme prevalent across areas of high or low socioeconomic deprivation was lack of support, reflecting previous research in Scotland (Children 1st, 2012), the United Kingdom and internationally (Boetto, 2010;Sakai, Lin, & Flores, 2011). The current findings reinforce that support remains a prominent issue for kinship carers, regardless of deprivation level (Bywaters et al., 2016).
Contact with biological parents was reported as a risk across the sample. Continued contact with biological parents in kinship care is encouraged, associated with greater understanding of a child's personal background, enhanced well-being, sense of belonging, identity and social support (Dubowitz et al., 1994;León, Jiménez-Morago, & Muñoz-Silva, 2017;McWey & Mullis, 2004;Messing, 2006). However, our findings align with research on the potentially problematic nature of contact relating to unpredictability, unsupervised visits and unplanned encounters (Kiraly & Humphreys, 2013;Selwyn et al., 2013), highlighting the need for support for kinship carers to preserve birth family relations.
Whilst seeking financial advice was common across the sample, more serious financial concerns and deficits were more common in areas of higher deprivation, as expected. Child behavioural problems were commonly reported alongside these serious financial difficulties.
These findings support previous research; socioeconomic status and socially disadvantaged areas have been linked to higher externalizing and aggressive behaviour (Dodge, Pettit, & Bates, 1994). Factors relating to social context, such as family life stressors and social support, may mediate this relationship (Dodge et al., 1994). Future research could investigate whether increasing social support for carers is related to better well-being and behavioural outcomes in kinship children.
One of the most notable findings is the greater reporting of child's histories of neglect and abuse and concerns about their mental health in areas of lower socioeconomic deprivation. There is a wellestablished link between ACEs such as neglect or abuse, and mental health problems in children and adults (Springer, Sheridan, Kuo, & Carnes, 2007). However, as children in care are more likely to experience ACEs in general, and the occurrence of adverse early experiences is higher in individuals both living in deprivation and those in care (Turney & Wildeman, 2017), we hypothesize an effect of prioritization of concerns, such as accessing money to feed children. Alternatively, there may be avoidance of disclosing information that might implicate family members or even their own parenting.
Use of helpline services is influenced by factors such as age, gender and most relevant to the present findings, deprivation (Cook, Randhawa, Large, Guppy, & Chater, 2013). The difference in call times might be an artefact of practical financial matters being responded to more efficiently than relational and behavioural issues. However, there may be more nuanced factors around entitlement with those from areas of lower socioeconomic deprivation feeling more confident in asking for help and pursuing a conversation until their needs are met. Kinship carers experiencing greater socioeconomic deprivation may have less time availability, due to busier households with more dependent children (Broad et al., 2001). Disentangling the causes of the difference could help organizations ensure that services meet the needs of kinship carers irrespective of socioeconomic status.
Social capital has not previously been investigated in kinship carers, who are reported to be more isolated and less formally supported than other types of foster carers (Ehrle & Geen, 2002;Selwyn et al., 2013). Consistent with previous literature (Berrick et al., 1994;Blair & Taylor, 2006;Selwyn et al., 2013), kinship families described being connected more with relatives than informal or formal supports. Despite this reliance on informal interpersonal relationships, the quality of these relationships was found to be overwhelmingly negative, reflecting extant literature (Boetto, 2010;Selwyn et al., 2013). However, this finding should be interpreted with caution within the context of the helpline's role in assisting with problems (i.e. it is less likely that positive experiences would be discussed). Nevertheless, these findings call into question how well supported kinship families feel, especially if formal service support is also lacking.
Schools were the most frequently described formal support network, as would be expected due to their universality. This reflects policy in Scotland, which situates school, as a universal provider, as the main contact point between a support team and the family (Scottish Government, n.d.). Contrary to previous findings, the next most accessed formal service was social services (Mason, Falloon, Gibbons, Spence, & Scott, 2002;Selwyn et al., 2013), followed by the National Health Service. Few third sector organizations (other than the helpline) were described, which may indicate a lack of awareness of other support organizations. This highlights that universal services such as schools and general practice surgeries are prominent in kinship family networks and may potentially act as entry routes to accessing further support.
Financial worries were reported as the major barrier to social capital growth in kinship carers. Bourdieu and Wacquant (1992) suggested that those from areas of high socioeconomic deprivation may lack social capital and miss out on the benefits of group cooperation (Horvat, Weininger, & Lareau, 2003). This study found some evidence in support of this; there was only one reference to kinship families approaching their local MP for support. It is possible that kinship families who are not aware of this route or do not feel empowered to access such networks. Whilst the current study was not able to directly investigate the link between deprivation and social capital in carers, findings indicate a possible association. Future research could examine this association, alongside the effect of this relationship on the differing risk factors reported by kinship carers.

| Limitations
Whilst the SIMD has frequently been used to assess deprivation, no other information regarding covariates of deprivation such as income or house price was available. This resulted in an assumption being made that low SIMD equated to deprivation in all cases. Although a large proportion of those living in socioeconomically deprived areas reported greater financial difficulties, future research could incorporate a multivariate examination of socioeconomic deprivation delineating between deprivation in the birth and the kinship family.
Additionally, there are other important protective and resilient factors resulting from the heterogeneous nature of kinship care and the presence of individual strengths as identified in previous work (Blakely, Leon, Fuller, & Jhe Bai, 2017;Burgess et al., 2010) that were not explored. A more in-depth analysis with a larger sample could assess the presence of both risk and protective factors and would allow researchers to better distinguish between individual differences that may impact outcomes. It should be noted that our inclusion of schoolaged children provided some degree of homogeneity to allow comparisons to take place. However, the findings cannot necessarily be generalized to kinship families of pre-school aged children, especially given the dominant role that school had as a support in our data.
Finally, the findings from this study are limited due to the type of data used. Call summaries were based on needs-led conversations without standardized questions. They were written by the call handler, potentially providing inconsistency in reporting call outcomes. Full transcripts and a longitudinal approach using standardized assessments would strengthen the findings and allow for richer insights.

| Implications for policy and practice
Despite policy legislation introducing statutory support for kinship carers in Scotland (Scottish Government, 2007), risk factors surrounding a lack of support and financial difficulties were prevalent, indicating an urgent need to provide more comprehensive and accessible supports for kinship carers. Third-party organizations providing resources like helplines play a pivotal role and can strengthen joint working between carers and other agencies.