Chapter 1. Introduction
The study "Children's level of living - the impact of family incomes" ("Barns levekår - betydningen av familiens inntekt") is a collaboration between NOVA and Norske Kvinners Sanitetsforening (The Norwegian Women's Public Health Association). It is a national representative survey focusing on the lives of children living in families with incomes below 60 per cent of the national median. Children and parents are informants about various aspects of their situation. The survey is longitudinal, with three rounds of interviews over 10 years, containing both quantitative and qualitative information. The panel design allows us to follow families over time. In this report we present and discuss the results from the second round of interviewing, which was carried out in 2006. Statistics Norway interviewed 1303 families and a corresponding number of children aged 9-15 years for the second time. 1068 informants belonged to the "low-income sample", while 235 belonged to the "control sample". A crucial question in this report is to what extent circumstances have changed in the course of the three years that have passed since the first round of interviews (Sandbæk (ed.) 2004).
Chapter 2. Methodological issues
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. The poverty line is defined as incomes below 60 per cent of the national median income, using the modified OECD equivalence scale. The interviews were carried out in spring, 2006. We interviewed both parents and children, the children were between 9 and 15 years old at the time of the interview. The response rates in 2006 were 65,6 per cent in the low-income sample and 75,8 per cent in the control sample, which is as expected in this type of survey research. 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.
Chapter 3. Family forms and employment
The chapter gives an overview of central background characteristics of the families in the survey. As in 2003, we find that the proportion of children living with both biological parents is ten percentage points lower in the low-income sample than in the control sample (63 v. 73 per cent). More children in the low-income sample have experienced family change in the period 2003-2006. This is partly because many of the 2003 lone parents have established new partnerships in 2006. Looking at employment, we find that 70 per cent of the control sample is made up of two-parent families with two incomes. The corresponding figure for Norwegian respondents the low-income sample is 48 per cent, and for non-western immigrants in the low-income sample, 26 per cent. Two-parent families where neither parent is in employment are virtually non-existent (below one per cent) in the control sample, while they make up three per cent of the Norwegian low-income families and 24 per cent of non-western immigrants in the low-income sample. The most common family constellation among non-western immigrants in the low-income sample are two-parent families where one is in employment. There is however a clear tendency towards higher rates of employment among the families in the low-income sample between 2003 and 2006. The education level is highest in the control sample, lowest among non-western immigrants in the low-income sample. With regard to receipt of social assistance, we find that 14 per cent of the families in the low-income sample received such assistance, compared to only two per cent in the control sample. All the categories studies have increased their (nominal) incomes in the period, but the differences between families inthecontrol sample, Norwegians in the low-income sample and non-western immigrants in the low-income sample are still highly visible. 63 per cent of non-western immigrants in the low-income sample, and 46 per cent of the Norwegians, have had incomes below the poverty line in at least three out of the six years we have information about.
Chapter 4. Housing
More families in the low-income sample than in the control sample have moved house in the period 2003-2006 (27 v. 14 per cent). The families that were most likely to move were those who reported an unsatisfactory housing situation in 2003. More families have established themselves as home owners: 64 per cent of the families in the low-income sample, and 87 per cent in the control sample, were home owners by 2006. Dwelling sizes, as measured by the number of rooms per household member, has increased little in the period. Non-western immigrants are more likely to live in cramped conditions than any other category - as many as 63 per cent of non-western immigrants have less than one room per household member, compared to 12 per cent of Norwegians in the low-income sample. The proportions who report housing problems is lower in 2006 than it was in 2003, but the families in the low-income sample still report more housing problems than families in the control sample. In 2003 48 per cent of the low-income families, and 62 per cent of the families in the control sample, reported none of the housing problems we had specified. The corresponding figures in 2006 were 57 and 69 respectively. The number of housing problems depends to some extent on whether or not the family has increased its income: there are more families with very poor housing conditions among those who had incomes below the poverty line in 2005 than among those who had increased their incomes above this level.
Chapter 5. Material living conditions
79 per cent of the families in the low-income sample own all the consumer durables we have identified as "central" (freezer, washing machine, dishwasher) in 2006, compared to 71 per cent in 2003. The corresponding figure in the control sample is 95 per cent (2006). The proportion in the low-income sample who say they lack a cash margin (could not handle an unexpected bill of NOK 10.000) has fallen from 8 to 4 per cent in the period. No-one in the control sample is in this situation. The proportion who report that money is insufficient is stable, and remains higher in the low-income sample than in the control sample. Those who had incomes below the poverty line in 2005 were more likely to give this response than families with higher incomes. Non-western immigrants, lone parents and families with weak connections to the labour market are overrepresented among the families with poor living conditions within the low-income sample. 62 per cent of the families in the low-income sample report none of the forms of deprivation we have asked about (poor housing conditions, lacking consumer durables, feel that money is insufficient). 26 per cent report one aspect of deprivation, 9 per cent report two, while 3 per cent of the low-income sample report deprivation in all three aspects. There has been an improvement in this respect in the period: in 2003 "only" 50 per cent of the families in the low-income sample reported no form of deprivation. Among those who reported all three problems in 2003, almost one-third report no problems in 2006. We thus see a general improvement, and at the same time we find that few families report very high levels of deprivation over time.
Chapter 6. Family life and social networks
The most important differences with regard to family life and social networks run between non-western immigrant families on the one hand, and Norwegian families on the other. Immigrant background influences these aspects more strongly than the family'sfinancialsituation. 33 per cent of non-western immigrants in the low-income sample do not go on annual holidays, compared to 11 per cent of the Norwegians in the low-income sample and 7 per cent of the control sample. 28 per cent of non-western immigrants do not have a restaurant meal or go to the movies every six months, compared to 16 per cent of Norwegians in the low-income sample and 11 per cent in the control sample. Still, the proportion of immigrants who invites friends or family to their home at least monthly is 85 per cent, which is higher than among Norwegians in the low-income sample (77 per cent) and among the control sample (80 per cent). There are no differences between the low-income sample and the control sample when it comes to contact with grandparents, but non-western immigrant children have significantly less contact with grandparents than other groups. 81 per cent of the children in the control sample see their grandparents at lest once a month, as do 71 per cent of Norwegian children in the low-income sample - and 14 per cent of non-western immigrant children. Non-western immigrants in the low-income sample have more vulnerable social networks than Norwegian low-income families and families in the control sample. 25 per cent of non-western immigrants report that they have no-one they can talk to in confidence within the family, 30 per cent say they have no-one to talk to outside the family. The corresponding figures for Norwegians are 7 and 9 per cent. 26 per cent lack someone in the family who can help them out in cases of illness, and 27 say they have no-one outside the family, compared to 9 and16 per cent of the Norwegians. About 10 per cent of non-western immigrants have no-one to turn to when need to talk in confidence or when they are ill, neither within or outside the family.
Chapter 7. Health and health-related issues among parents in 2006
As in 2003, parents in the low-income sample generally have a far poorer health situation than parents in the control sample. In the low-income sample 23 per cent of the mothers, and 22 per cent of the fathers, report that their self-perceived general health is mediocre or poor, compared to 13 and 15 per cent of mothers and fathers respectively in the control sample. The proportion who reports psychological problems is also much higher among mothers in the low-income sample than among mothers in the control sample. The change over time, from 2003 to 2006, is however quite similar for parents in the two samples, and the parents' health situation does not vary in any systematic way depending on the development of family incomes or the duration of income poverty. The analyses suggest that health problems created an obstacle to employment among parents in the low-income sample, thus health problems are probably one of the reasons why their incomes remain low. Among the parents in the low-income sample, non-western immigrants and lone parents in particular give the impression of a relatively poor health situation, while parents in the low-income sample whose incomes had increased above the poverty line by 2006 are less likely to report poor (self-perceived) general health and psychological problems. The feeling of having a poor family economy in particular seems to influence parents' subjective health situation.
Chapter 8. Children's lives at school: school results, social relations and well-being
In the first round of interviews, most children in the low-income sample gave a positive impression regarding school as an arena for learning and for social relations. Still a higher proportion of children in the low-income sample than in the control sample replied that they believed they did poorly at school compared to "the others". A somewhat larger proportion of children in the low-income sample received special education. After the second round of interviews we found that this imagewassomewhat, but not considerably, changed. All participating children and young people replied on their own behalf in the second round of interviewing.
Eight out of ten children both in the low-income sample and in the control sample say that they enjoy school. Nine out of ten feel that they have good contact with the teacher and that they can talk to him or her. More than half agree that the teacher encourages them to express their own opinions and treats everyone fairly. Children and young people with non-western immigrant backgrounds express more positive attitudes on these dimensions than others. Almost nine out of ten in both sub-samples report that they have four or more good friends in class, and about 10 per cent say that they have two or three good friends.
Young people in the low-income sample report lower degrees in maths than young people in the control sample. The difference is small, but statistically significant. Apart from this, there are only minor variations with regard to reported degrees. The biggest differences are found between boys and girls in the control sample.
A higher proportion of respondents in the low-income sample than in the control sample report being bullied in schools, but this finding is not statistically significant. Those who reported bullying in 2003 have a higher risk of also being bullied in 2006.
Chapter 9. Leisure and social participation
As in 2003, we find significantly higher participation in organised leisure activities among Norwegian 9-12-year olds in the low-income sample than among non-western immigrant children in the same sample and age group. Among the 13-15-year olds, the proportion of non-western immigrants who do not participate in organised leisure is even higher. This is now the situation for 46 per cent of this group. In the control sample and among Norwegians in the low-income sample, the corresponding figures are 33 and 22 respectively. It is mainly the non-western immigrants girls who create the differences between the samples in this respect, the differences are smaller among boys.
In the youngest age group, we find that non-western immigrant children are less likely than children in either of the other categories to bring friends home. This is true for both boys and girls. We find no similar differences among the 13-15-year olds.
Children and young people in the control sample are somewhat more likely than their peers in the low-income sample to own CD-players or similar equipment, mobile phones, ski/ snowboard and bikes. If the low-income sample is split by immigrant background, we find that such equipment is more commonly owned by Norwegians in the low-income sample than among children with non-western immigrant backgrounds. The difference is particularly striking with regard to skis and snowboard, and considerably smaller with regard to ownership of bikes. All these differences are statistically significant.
Chapter 10. Children's views on the family's financial situation, and their own financial resources
In 2003 only children in the oldest age group answered questions about their own views of the family's financial situation. The majority of children then replied that their family's financial situation was "neither good nor bad", and that the family's incomes had not restricted their own social participation. Also in 2006, we observe little variation in children and young people's subjective assessment of their family's financial situation. This may be because parents act as "buffers" and shield their children from the realities of their financial situation, or, alternatively, because children hesitate to talk about the family's financial difficulties. Still, we find a somewhat stronger correspondence in 2006 than in 2003 between how parents assess the family's situation, and how children assessit.In families where parents say that money is insufficient, children are significantly more likely to say the same. Among the families in the low-income sample whose incomes had been low over time and parents say that money is insufficient, a considerably higher proportion of children and young people report that the family is "badly" or "very badly" off, compared to families in the low-income sample whose financial situation is better.
The indicators we have on children's own financial resources show only small differences between the two sections of the low-income sample (the long-term poor and the transiently poor) and the control sample. There were small differences in average weekly allowances, and also a certain difference in the proportion of 13-15-year olds who work for pay outside school. 27 per cent of children in the control sample work outside school, compared to just above 15 per cent in the low-income sample.
Chapter 11. The health situation among low-income children in 2006
In analyses of health conditions among children it is important not only to look at disease and disability, but also at the child's physical, mental and social development, since such indicators represent children's potential for health later in life. Controlled for age and gender, children in the low-income sample are on average 2-3 centimetres shorter than children in the control sample. There is also a tendency towards higher rates of obesity among low-income children, while diseases and disabilities are not found more frequently among low-income children than among children in the control sample. Regarding well-being, psycho-somatic symptoms and social adjustment, the situation for children with a non-western immigrant background is similar to that of children in the control sample. As to psycho-social development and well-being, there are however clear differences between children in the control sample and Norwegian children in the low-income sample. For instance, among the Norwegian low-income children 24 per cent report at least one psycho-somatic symptom weekly, compared to 17 per cent among the control-sample children. 15 per cent of the Norwegian low-income children have been home from school 10 days or more "since the summer holiday" because of illness, compared to 9 per cent of the children in the control sample. Norwegian low-income children also report more contact with welfare services (children and young people's psychiatric services, child welfare services), more parents report that the child has ADHD (3,8 v. 0,9 per cent) and among the 13-15-year olds there are more children in the low-income sample than in the control sample who smoke tobacco daily. This negative profile is particularly evident for families who have had low incomes over time, and it is also more striking in lone parent families and in families where the parents report health problems. This suggests that low family incomes, and also other family-related circumstances that are more prevailing in low-income families, may be a stress factor for children that can be linked to a higher risk of low satisfaction, unfortunate psycho-social development and problem behaviour. This seems to occur most often among Norwegian children in the low-income sample, but - surprisingly - not so much among children with non-western immigrant backgrounds. It should be noted that the health situation among Norwegian children in the low-income sample is not generally poor, but there is an overrepresentation of problematic circumstances that may imply a poorer health development over time.
Chapter 12. Self-confidence and locus of control
The results of the analyses of Harter's (1982, 1988) two scales for self-perceived competence show that children from families with low incomes have a lower self-perceived competence score than children in the control sample. Also infamilieswhere the parents report that money is insufficient, children are more likely to have low self-perceived competence. Children living with both biological parents score significantly higher than others on Harter's children- and youth scale.
Nowicki and Strickland's (1973) scale for locus of control shows that children from families who have had low incomes for several years are more likely to explain events as coincidences or as caused by external circumstances. Correspondingly, they are less likely to emphasise the impact of their own skills and resources. The same result is found for families in which the parents report that money is insufficient. Finally, our results show that children from non-western immigrant families were also more likely to display an external locus of control, that is, explain events by external circumstances.
Chapter 13. Social exclusion
The chapter is based on the qualitative interviews carried out as part of the project "Children's level of living", that is, 26 interviews with children aged 10-13 (2004). It discusses to what extent children living in low-income families are more likely to be socially excluded, and, if so, what exclusion processes are at play. Social exclusion is linked to both structural and relational factors, and is analysed across the dimensions school, leisure, friends and consumption. The access to money has a strong impact on activities both in school, and in organised and spontaneous leisure activities. This creates a basis for social exclusion. The excluding processes operate along two tracks: from structure to relations, and from low to high degrees of exclusion. The children in the sample are found in various positions. Some children experience very low degrees of social exclusion, other face exclusion in many different areas. It seems that the children living in the poorest financial circumstances are most at risk for social exclusion.
Chapter 14. Children's level of living - conclusions
The concluding chapter provides an overall discussion of how the findings from the second round of interviews shed light on the research questions posed in the project. We look more closely at developments in the families' levels of living in the period 2003-2006 in the low-income sample compared to the control sample. Further, we look at the relationship between low incomes and levels of living: Can the families in our low-income sample be described as poor? The parents' responses show significant differences in all the areas we have researched, and the low-income sample are systematically worse off than the control sample. The image painted in the interviews with children is more mixed. As for the parents, wherever there are differences we find that the children in the low-income sample are worse off than children in the control sample, but in a number of areas we find no or only minor differences between children in the two samples. Can this be because parents succeed in protecting their children against the consequences of living with low incomes? Or is it more precise to say that they prioritise their children's needs, but do not fully succeed in protecting them? Several findings in this report suggest that children are aware of the family's actual financial situation, particularly when the situation is very difficult. Towards the end of this chapter, we discuss how we can maintain a focus on children in the political efforts to end child poverty. We discuss measures along two dimen¬sions: measures that build on the parents' wish to provide a good environ¬ment for their children, and measures that are directly targeted at the children themselves on the various arenas where they live their lives.