Introduction
A few decades ago the focus in poverty research was based on the assumption that exploring family poverty would also cover children's situations (Grødem, Reference Grødem2008a). Today an explicit focus on how poverty affects children has a high priority on the research as well as the political agenda. An increasing number of studies collect data on various aspects of children's lives, often seen from children's own perspectives. This is particularly true for ethnographic studies within the tradition of sociology of childhood, but one can also trace a growing number of surveys with children as informants and the unit of analysis.
Another shift worth noticing is how the attention is no longer directed exclusively to risk, but also includes protective factors and children's well-being. Using child poverty as a case study, Barrett (Reference Barrett2003) underlines that although chronic poverty does increase the likelihood of a number of problems, not all children growing up in poor families will experience negative outcomes. Adverse circumstances increase the chances for negative outcomes, in particular if there are more of them, but in order to understand a child's development, protective factors must also be taken into consideration (Ghate and Hazel, Reference Ghate and Hazel2002; Chasse et al., Reference Chasse, Zander and Rasch2003; Sandbæk, Reference Sandbæk and Daly2007a).
Dimensions of children's well-being cover a wide range of issues, such as economic and material situation, housing and environment, health, education and well-being, social and civic participation, family forms, care and environment, exposure to risk and safety (UNICEF, 2007; European Parliament, 2007; Bradshaw and Richardson, Reference Bradshaw and Richardson2008; Richardson et al., Reference Richardson, Hoelscher and Bradshaw2008). Family relations have for many years been acknowledged as vital for children's development, having a protective impact if positive and representing a risk factor if negative. A positive mutual interaction between a child and at least one caregiver, coherence, warmth and the absence of conflicts have also been identified as enhancing children's resilience when growing up in poor and disorganised environments (Rutter, Reference Rutter1971, Reference Rutter, Rolf, Masten, Chicchetti, Nuechterlein and Weintraub1990; Armsden and Greenberg, Reference Armsden and Greenberg1987; Werner and Smith, Reference Werner and Smith1982, Reference Werner and Smith1992; European Commission, 2008).
However, there is also a certain ambiguity associated with drawing attention to family relations in poor families. Such a focus may contribute to portray negative outcomes of child poverty as a family failing (Ridge, Reference Ridge2007; Sandbæk, Reference Sandbæk, Wintersberger, Alanen, Olk and Qvortrup2007b). But it can also serve the opposite purpose and put an end to or nuance myths about negative family relations in poor families. Further, it can draw attention to what kind of support poor families need in order to cope with a difficult economic situation and take proper care of their children. Present models of parenting move away from emphasising particular parenting styles towards focusing on how parenting emerges from the interaction between several factors in the broader cultural environment, including social and community support (Pećnik, Reference Pećnik and Daly2007).
The aim of this article is to explore certain aspects of family relations in poor versus non- poor families, namely children's perception of parental support and monitoring. The article examines: (1) family income and children's perception of their parents; (2) children's evaluation of the family economy and the child–parent relationship; and (3) possible predictors of received support and monitoring. We start with introducing recent research in the field, followed by a presentation of the study providing the data for the article.
Resent research on family relations in poor families
Earlier studies have demonstrated an overrepresentation of a wide range of family risk factors associated with growing up in a poor family, such as family turmoil and violence, less social support and less responsive and more authoritarian parents (Bradshaw, Reference Bradshaw2002; Seecombe, Reference Seecombe2002; Evans, Reference Evans2004). Does this image change when children themselves are asked to be informants on child–parent relationships?
The Young in Norway survey (Bakken, Reference Bakken1998) was carried out in the school setting with a response rate of 94.3 per cent; 11,425 young people, mainly in the age group 14–17, answered questions about their relationship with their parents, school, friends and leisure activities as well as problem behaviour, criminality and drug use. 84 per cent agreed with the statement that their parents accepted them; 60 per cent reported that their parents took note of it if they were sad. Three out of four children felt that it was important for their parents to be informed about their leisure activities and close to 80 per cent believed that their parents usually knew where they were, while a little more than 70 per cent believed that their parents knew most of their friends. The general impression was that the children received a lot of support and acceptance from their parents (Bakken, Reference Bakken1998).
The Young Voices poll was carried out in Europe and Central Asia among a representative sample of 15,200 boys and girls between the ages of nine to 17, from urban and rural, rich and poor backgrounds (UNICEF, 2001). The children were interviewed in their homes and in their own language with the permission of their parents or guardians, but not in their presence. The poll covered a wide range of topics, such as children's views on government and politics, how they felt in society, at school and at home, their contact with harmful or illegal substances and how they see their future.
Strong family ties were evident in all countries. An average of nine in ten children rated their relationship with their mothers as very good or good. Between seven and eight in ten gave their fathers an equal rating, with the Western CIS and Baltic States obtaining the lowest scores. More children from higher-income families saw their relationships with parents as positive compared to children from poorer families, but the differences were not big: 85 per cent from higher socio-economic groups rated their relationship with their mothers to be very good or good, compared to 83 per cent from middle and 72 per cent from lower socio-economic groups.
Although they described their relationships with parents in such positive terms, only four in ten reported that there was no violence at home: 46 per cent from Western European countries and 39 per cent from CEE/CIS and the Baltic States. There were hardly any differences between the socio-economic groups: 42 per cent from higher levels reported no violence at home, compared to 41 per cent among middle and lower socio-economic groups (UNICEF, 2001).
Redmond (Reference Redmond2008) has reviewed nine recent qualitative studies on children's perspectives on economic disadvantage, with children between five and 17 years old. The studies did not provide in-depth analysis of the connections between economic adversity and family functioning, but they still offer some insight into relationships in disadvantaged families. Families were central to children's lives and children from low-income families were not always severely disadvantaged in this respect. Most of them enjoyed close relations with at least one parent. Positive relations seemed to protect children and soften the impact of economic adversity. However, many children also felt the pressure of economic disadvantage and lack of economic resources caused worries, arguments and disagreements. Children contributed in various ways, by not asking for items and activities, by helping the parents to cope with the lack of material resources and by supporting them in their return to the labour market. This duality is well captured in Redmond's (Reference Redmond2008: 4) formulation that the children ‘both contributed to and drew on family strength as source of resilience’.
Children's own reporting reveals a more varied picture of the child–parent relationship than the traditional statistical analysis, including constraints and worries but also positive feelings and mutual protection. The study presented in this article can add valuable information to the recent literature, having both parents and children as informants and following them over several years. Compared to other European countries, the child poverty rate in Norway is low, between 4 and 8 per cent, depending on which measurements are being used (Whiteford and Adema, Reference Whiteford and Adema2007; Epland and Kirkeberg, Reference Epland and Kirkeberg2007). Studying child poverty in an affluent country, and also within a well-developed welfare state, like Norway can give important insights into the groups that are hard to reach and to explore how growing up in poverty in such wealthy environments affects children.
Data
How do Norwegian children experience life on low income? This is the focus in ‘Children's Level of Living – the Impact of Family Incomes’,Footnote 1 a representative national survey of children living in families with incomes below 60 per cent of the national median, using the modified OECD equivalence scale. The survey consists of two samples: the low-income sample, made up of families with incomes below the EU poverty line in 2000, and the control sample, consisting of families from across the income distribution. Children and parents are informants about various aspects of their lives in this panel study with three rounds of quantitative interviews over ten years, the last in 2009. The interviewers use computer-assisted interviewing (CAI), bringing laptop computers to the family homes, and reading the questions from the screen. The first data wave also contained qualitative interviews with 26 children and parents from low-income families (Thorød, Reference Thorød2006, Reference Thorød and Sandbæk2008). Both parents and the child had to consent to the child's participation. See Sandbæk (Reference Sandbæk2004: 29) for further discussions of ethical challenges related to interviewing children.
This article is based on the results from the second round of interviewing, which was carried out in 2006. In the first data wave in 2003, parents answered on behalf of children aged 6–9, while in 2006 all of the children were interviewed themselves. Statistics Norway interviewed 1,303 children aged nine to 16 years and a corresponding number of parents the second time. 1,068 informants belonged to the ‘low-income sample’, while 235 belonged to the ‘control sample’.
The overall response rates in 2003 were 54.2 per cent: 53.0 per cent in the low-income sample and 63.8 per cent in the control sample. In 2006, we obtained a response rate of 65.6 per cent in the low-income sample and 75.8 per cent in the control sample, which is common in this type of survey research. Table 1 provides an overview of gross and new samples. An analysis of non-responses indicates that we have the highest rates of non-response among non-western immigrants, and among families with weak ties to the labour market. Still, these two groups in addition to single mothers are over-represented among the respondents in the survey and we find expected differences between the two income groups (Sandbæk, Reference Sandbæk2004, Reference Sandbæk2008).
Table 1 Gross – and net sample in 2003 and 2006
Note: The distinction between the low-income – and control group is based on income information from 2000 provided by Statistic Norway, which also provided income information from the following years, up to 2005.
Children with a non-western immigrant background are excluded from the analyses presented here. The aim is to examine the effect of low income on family relations, and the control sample does not contain sufficient numbers of non-western immigrants for such comparisons. 1,076 children are thus included in the analysis; 841in the low-income sample and 227 in the control sample. In some of the analysis, we make a distinction between the families who still had an income below the poverty line in 2005 (low income in 2005) and those who had achieved a higher income (non-low income in 2005) in order to measure the effect of long-term low income. 93 per cent of the families with low income in 2005 had lived on low income for more than three years (Grødem, Reference Grødem and Sandbæk2008b).
Measurements
We measure two aspects of children's subjective perceptions of their parents, parental positive support and parental monitoring. Questions concerning parental support stem from Armsden and Greenberg's, (Reference Armsden and Greenberg1987) scale ‘Inventory of Parent and Peer Attachment’, while those on parental monitoring stem from Alsaker et al.’s (Reference Alsaker, Dunas and Olweus1991) ‘monitoring’ scale. We have taken the Norwegian versions of these scales used in The Young Investigation (Bakken, Reference Bakken1998) as models to be able to compare them with a more comprehensive sample and thus validate the results. Due to economic constraints, we have shortened and merged the two scales into one. The reliability of the subscales was analysed with Cronbachs alpha. The parental support scales obtained 0.65, while the scale measuring parental monitoring obtained 0.48. The low scores are probably due to the fact that the monitoring scale only contains three items. Table 2 displays the results from a factor analysis. The eigenvalues show that a two-factor solution fits the data, in accordance with the original scales measuring support and monitoring.
Table 2 Factor analysis of parental support and monitoring – varimax rotated
Children's perceptions of parental support and monitoring
As demonstrated in Table 3,Footnote 2 the overwhelming majority of the children reported having a positive and confident relationship with their parents. More than 95 per cent of the children agreed to the statement, ‘My parents accept me as I am’ and ‘I trust my parents’. Around 90 per cent said that their parents noticed it when they were sad and also felt that their parents understood them, while around 80 per cent told their parents about their problems.
Table 3 Children's perception of parental support and monitoring
On the questions measuring parental monitoring, a little above 90 per cent of the children agreed to the statement: ‘It is important for my parents to know where I am and what I do in my leisure time.’ There was a discrepancy between the parents’ wishes to monitor and their real overview; from 82 to 86 per cent felt that the parents knew their whereabouts 90 per cent in the low income sample and 87 per cent in the control sample reported that their parents knew most of their friends.
No significant differences were found between children in low-income families and children in the control sample regarding the children's perception of parental support and monitoring. There were also no significant differences between children whose families still had and those who no longer had a low income in 2005. Long-term poverty did thus not seem to have any effect on the children's perception of their parents on the aspects we have measured when we compare item by item. Girls had a more positive evaluation than boys of their parents on some, but not all items.
Children's evaluation of the family economy
The next step is to explore if children's subjective evaluation of the family economy has any bearings on their views of their parents.Footnote 3 The majority of the children reported that their family was well off, 66 per cent in the low-income sample and 71 per cent in the control sample. Only 5 per cent in the low-income and 1 per cent in the control sample reported to be badly off. The differences were significant, p < 0.05. At the same, time these figures also reveal that a considerable number – 34 per cent in the low income and 29 per cent in the control sample – did not consider their family to be well off. The control sample consists of all income groups, also some families with low income in one or more years in the period studied. This may contribute to the high proportion of children in the control sample who did not find themselves well off. Another reason may be that the children compare themselves with the most affluent among their peer group.
There may also be many reasons why children from low-income families report that their family is well off. Children may be hesitant to reveal their economic difficulties, due to the stigma attached to labelling themselves and their family as poor. The methods used in terms of standardised questions may also not be the best to make children share such worries. A third element may be that the children are not fully aware of the situation. Our results indicate that parents try to protect their children from the economic difficulties. While more than half of them reported that they cut down on their own expenses, only 12 per cent reported that they used less money on the children to make ends meet (Grødem, Reference Grødem and Sandbæk2008c). The finding that parents in hardship give priority to their children is in accordance with a number of other studies (Chasse et al. Reference Chasse, Zander and Rasch2003; Rysst, Reference Rysst2006; Redmond, Reference Redmond2008). However, it is important to underline that protecting the children from the consequences of the lack of economic resources will only be possible when the economic constraints are not too severe (Middleton et al., Reference Middleton, Ashworth and Braithwaite1997).
The majority of children from low-income families reported that they were well off, but we do find associations between economic problems as reported by parents and the children themselves. A lower proportion of children in families that still lived on low income in 2005 reported that their family was well off compared to children whose families no longer lived below the poverty line, 58 per cent compared to 67 per cent. While 8 per cent of those living in low-income families in 2005 found that their family was badly off, this was only true for 4 per cent of the low-income families who no longer lived on low income in 2005 and 1 per cent of the control sample. The differences were significant, p < 0.05 (the table is not shown).
The children's evaluation of the family economy and perceived support and monitoring
While the family income was not associated with a more negative perception of the support and acceptance offered by their parents, the analysis revealed that the parent–child relationships were clearly more troubled in the families where children felt the family economy was poor. Table 4 shows that these children were less likely than the children who were well off to tell their parents about their problems, 49 per cent compared to 84 per cent; they felt that their parents understood them more rarely, 74 per cent compared to 93 per cent; they had less trust in them, 84 per cent compared to 98 per cent; and, finally, also felt that their parents more rarely noticed when they were sad, 74 per cent compared to 94 per cent. Except for acceptance, all the differences were significant, p < 0.001.
Table 4 Children's perception of the family economy by their relationship with their parents
Note: *** p < 0.001, ** p < 0.01.
There were also differences between the groups on variables measuring parental monitoring. 77 per cent among those who felt the family was badly off compared to 84 per cent among those who felt their family was doing well confirmed that their parents knew where and how they spent their leisure time. The proportion of parents who knew their friends were reported to be respectively 86 per cent compared to 89 per cent and the proportion of children who felt it was important for the parents to be informed about their whereabouts were 88 compared to 93 per cent. Only the last result was significant p < 0.01. There were no significant differences between children living in families with low income in 2005 and children in families who were no longer living with low income.
Can children's perceptions of parental support and monitoring be predicted?
Moving from bivariate to multivariate associations, we use the same measures of parental support and monitoring as in the previous section. The scale is continuous, with values ranging from 1–5 on each item, with a possible variation from 8 to 40 scores. The actual values range from a minimum score of 19 to a maximum score of 40, with mean of 36.18 and standard deviation of 3.36. The independent variables used in the regression were entered in three blocks: (1) information from the children regarding their perception of their own health, general well-being and family economy; (2) information from the parents regarding their own health and psychological stress; and (3) background variables, including children's age, gender, family forms and family income.
Children's age is a continuous variable, ranging from nine to 16. Gender is a dichotomous variable, girls coded as 1, boys as 0. Family forms were coded as a dichotomous variable, where living with both parents was coded as 1 and all other family forms coded as 0. Family income was also coded as a dichotomous variable, where families who remained in the low income category also in 2005 were coded as 0 and other families 1.
The children's perception of family economy, health and general well-being were all coded into dichotomous variables, where positive responses such as “well” or “very well” were given the value 1 and all other responses given the value 0, including those who responded in-between. The parents report of their general health was entered as a dichotomous variable, where bad/very bad health is coded as 0, all other variables coded as 1. The parents’ psychological health was measured as an index, including the parents’ perception of being nervous and/or suffering from anxiety and/or being depressed. Having one or more of these symptoms is coded as 0, not having them coded as 1.
Table 5 shows that the most comprehensive model with all the independent variables (block 3) predicts 12 percent of the variation. Children's subjective evaluation of their health, well-being and of the family economy appear to be strongly related to the outcome variable also when background characteristics are controlled for. Children's evaluation of their general well-being plays a particularly important role. The coefficients for background variables, age, gender and family forms, also come out as significant. The older children perceived less support from their parents, boys less support than girls and this is also the case for children living with one of the biological parents. Controlled for the other variables in the regression, low income does not add anything, which is also the case for the parents’ subjective assessment of their health and psychological stress.
Table 5 Children's perception of parental support and monitoring, OLS regression
Notes: ***p < 0.001, ** p < 0.01, *p < 0.05.
Discussion
This article has explored Norwegian children's perception of parental support and monitoring in disadvantaged versus more-affluent families. An overwhelming majority felt that they received support and acceptance from their parents and that they took an interest in their leisure activities and friends. How can such positive responses be interpreted? Firstly, the results demonstrate how important parents are to their children. Rating the relationship in such positive terms does, however, not imply that there are no conflicts or difficulties. As indicated in UNICEF's (2001) The Young Voices poll, many children rated their relationship as positive even if there was violence in the family. When these things happen, children often try to understand and excuse their parents (Council of Europe, 2007). Secondly, age probably contributes to the high figures. The children in our study are aged from nine to 16 years old. Older children seem to be more critical of their parents as could also be seen in The Young Investigation in Norway (Bakken, Reference Bakken1998). This does not necessarily imply that the parents are less important, but young people are moving towards more independence and self-determination. Thirdly, the methods may also play a role in producing the image of non-problematic relations. Our measures capture limited aspects of the relationship between children and parents. Further, standardised questions do not allow the respondents to share nuances and individual stories behind the figures. These limitations are however similar across the income groups. Our study finds that most parents, regardless of income, do their best to give priority to their children (Sandbæk, Reference Sandbæk2008). The results presented in this article, may indicate that children acknowledge and appreciate their efforts.
Poverty in itself does not seem to cause poor parent–child relationships, but several factors may play a role in how poverty affects family relations. The stress poverty puts on parents and children should not be underestimated. A majority of families living below 60 per cent of the median income in Norway are homeowners, have a car and equipment such as a washing machine, freezer and dishwasher. However, there are significant differences in disfavour of the low-income families’ possession of all these items (Grødem, Reference Grødem and Sandbæk2008c). Norwegian families living on longterm social assistance are struggling with a number of severe problems (Hjelmtveit, Reference Hjelmtveit, Harsløf and Seim2008). Poverty is also handed over from one generation to another. Children whose parents receive social assistance are more likely to receive such support as adults (Lorentzen and Nielsen, Reference Lorentzen and Nielsen2008; Wiborg and Hansen, Reference Wiborg and Hansen2008). Consequently, also in Norway there is a need for additional measures to break these negative circles. Given the number of recent studies indicating that parents living in poverty prioritise their children's needs (Rysst, Reference Rysst2006; Redmond, Reference Redmond2008; Sandbæk, Reference Sandbæk2008), there may be reason to strengthen universal measures to support their efforts to care for their children themselves. Further, measures must be channeled directly towards children and their arenas. The qualitative analysis (Thorød, Reference Thorød and Sandbæk2008) demonstrates how disadvantaged children try to protect their parents, while at the same time strive to avoid being excluded from peer groups at school and in leisure activities, due to not being able to pay for the same activities as their friends. Reduced costs at school and at least one free leisure activity can make it easier for children to keep up with their friends.
While the family's income was not associated with any differences in the children's perception of their parents’ acceptance and monitoring, the children's own subjective evaluation of a low-income family economy revealed poorer relations between children and parents. This finding is not well explained in our statistical analysis, but corresponds to children's own stories in qualitative interviews of how they do not ask for their needs if they fear the parents cannot afford to pay, how they try to find solutions and not to share their worries. Maybe it is difficult to trust a parent who cannot provide for you? Health may also play a role in how a child rates his or her relationship with their parents. Unhappy children may be more critical towards their parents and more easily label the economy as bad, but this may also work the other way around: a negative relationship with their parents is likely to make children unhappy. Children and young people themselves must play a more important role as informants in further explorations of the complex linkages between family income, intra-family distribution and parent–child relationships.
