Socio-economic adversity and child development: Mechanisms and Processes

1. Ingrid Schoon Institute of Education, London Socio-economic adversity and child development: Mechanisms and Processes A recent report by the United Nations suggests that poverty rates for children in the UK are several times higher than those in most Western industrialised countries. Indeed the UK is bottom of the league of the 21 most developed countries regarding rankings on poverty and associated well-being among children and young people - followed by the United States which are second to last (UNICEF, 2007). Given the persistence of extreme poverty even in highly developed countries it is essential to learn more about the impact of material hardship and socio-economic adversity on families and their children - and about the processes and mechanisms by which socio-economic adversity is transmitted across generations. Here I will examine the processes linking the experience of socio-economic adversity to family functioning and child development – with a particular emphasis on processes enabling positive adjustment in the face of adversity, a phenomenon also described as resilience. A recent report by the United Nations suggests that poverty rates for children in the UK are several times higher than those in most Western industrialised countries. Indeed the UK is bottom of the league of the 21 most developed countries regarding rankings on poverty and associated well-being among children and young people - followed by the United States which are second to last (UNICEF, 2007). Given the persistence of extreme poverty even in highly developed countries it is essential to learn more about the impact of material hardship and socio-economic adversity on families and their children - and about the processes and mechanisms by which socio-economic adversity is transmitted across generations. Here I will examine the processes linking the experience of socio-economic adversity to family functioning and child development – with a particular emphasis on processes enabling positive adjustment in the face of adversity, a phenomenon also described as resilience.

2. The recognition of resilience in human development – has emerged following consistent evidence of individuals bouncing back or continuing to function reasonably well, despite the experience of risk or adversity – The notion of resilience has been used to refer to: positive outcomes despite the experience of adversity continued positive or effective functioning in adverse circumstances recovery after a significant trauma The instigation of the notion of resilience in the 1970s brought with it a paradigmatic shift in how scientists began to view the causes and course of development, moving away from a constancy model which portrays human development as deterministic either through the effects of genetics or by the predetermining role of early experiences. The new focus on resilience implied a shift away from a deficit model to a focus on strengths, assets and resources that might counteract the effects of adversity Questions were raised about what can be done to promote adjustment in the face of adversity with the aim to create opportunities for positive development, possibly before maladjustment had even occurred. The recognition of resilience in human development – has emerged following consistent evidence of individuals bouncing back or continuing to function reasonably well, despite the experience of risk or adversity – The notion of resilience has been used to refer to: positive outcomes despite the experience of adversity continued positive or effective functioning in adverse circumstances recovery after a significant trauma The instigation of the notion of resilience in the 1970s brought with it a paradigmatic shift in how scientists began to view the causes and course of development, moving away from a constancy model which portrays human development as deterministic either through the effects of genetics or by the predetermining role of early experiences. The new focus on resilience implied a shift away from a deficit model to a focus on strengths, assets and resources that might counteract the effects of adversity Questions were raised about what can be done to promote adjustment in the face of adversity with the aim to create opportunities for positive development, possibly before maladjustment had even occurred.

3. Identifying resilience Not directly measured Inferred on the basis of constellations of exposure to adversity and the manifestation of positive adjustment in the face of that adversity Based on two judgements: is the person doing ok? is there now or has their been any significant risk or adversity to be overcome? The notion of resilience differs from other terms such as general positive adjustment or coping, insofar as it takes into consideration the circumstances and processes under which development takes place One cannot talk about resilience in the absence of adversity – and the construct of resilience is inferred on the basis of constellations of exposure to adversity and the manifestation of positive adjustment in the face of that adversity. Instead of focusing on maladjustment or deficits, the focus on resilience aims to identify strengths and resources that facilitate adjustment The identification of resilience is based on two fundamental judgements: is a person doing ok? And is there now, or has their been any significant risk or adversity to be overcome A central assumption in the study of resilience is that some individuals are doing alright, despite being exposed to an adverse risk situation, while others fail to do so. The very definition of resilience is therefore based on an expectation of successful or problematic adjustment in response to risk factors that are assumed to affect adaptations: In order to identify resilience it has to be established whether the circumstances experienced by individuals do in fact affect their chances in life. If there would be no association between the experience of adversity, access to resources and opportunities, and consequent adjustment, the phenomenon of resilience would be a mere chance event, a random occurrence. Positive adjustment occurring with, versus without conditions of adversity often have different correlates and thus reflect different constructs. Furthermore, several studies have found varying antecedents of positive adjustment in general versus resilience in the face of adversity (Rutter, 1990). The two-dimensional definition of resilience brings with it the challenge of defining our understanding of risk and adversity, as well as positive adjustment. The notion of resilience differs from other terms such as general positive adjustment or coping, insofar as it takes into consideration the circumstances and processes under which development takes place One cannot talk about resilience in the absence of adversity – and the construct of resilience is inferred on the basis of constellations of exposure to adversity and the manifestation of positive adjustment in the face of that adversity. Instead of focusing on maladjustment or deficits, the focus on resilience aims to identify strengths and resources that facilitate adjustment The identification of resilience is based on two fundamental judgements: is a person doing ok? And is there now, or has their been any significant risk or adversity to be overcome A central assumption in the study of resilience is that some individuals are doing alright, despite being exposed to an adverse risk situation, while others fail to do so. The very definition of resilience is therefore based on an expectation of successful or problematic adjustment in response to risk factors that are assumed to affect adaptations: In order to identify resilience it has to be established whether the circumstances experienced by individuals do in fact affect their chances in life. If there would be no association between the experience of adversity, access to resources and opportunities, and consequent adjustment, the phenomenon of resilience would be a mere chance event, a random occurrence. Positive adjustment occurring with, versus without conditions of adversity often have different correlates and thus reflect different constructs. Furthermore, several studies have found varying antecedents of positive adjustment in general versus resilience in the face of adversity (Rutter, 1990). The two-dimensional definition of resilience brings with it the challenge of defining our understanding of risk and adversity, as well as positive adjustment.

4. Paradigmatic Shifts Move from constancy model to dynamic models of development From pathogenic to transactive systems models From deficit models to asset models Constancy model: developmental course and risk trajectory determined either through the effects of genetics or the predetermining role of early experiences Dynamic model: recognises continuity and change across the life course Historically most studies of development of at-risk individuals tried to understand adjustment problems, as reflected in ill health or mental disorder, academic failure, behavioural problems, or motivational deficits. These pathogenic models, failed to recognise the larger social system in which development takes place – the transactive interactions between a changing individual and a changing context. Trying to understand the processes and mechanisms that enable individuals to beat the odds, to strive against adversity also let to a change of focus on problems and needs of populations at risk, to possible assets and strengths within individual and communities. from a deficit model defining individuals and communities in negative terms, disregarding what is positive and what works well – needs imply dependence on welfare services, on external actors Asset-based approach: understanding of processes for creating, strengthening and protecting assets in individuals and communities that protect and promote adaptation Constancy model: developmental course and risk trajectory determined either through the effects of genetics or the predetermining role of early experiences Dynamic model: recognises continuity and change across the life course Historically most studies of development of at-risk individuals tried to understand adjustment problems, as reflected in ill health or mental disorder, academic failure, behavioural problems, or motivational deficits. These pathogenic models, failed to recognise the larger social system in which development takes place – the transactive interactions between a changing individual and a changing context. Trying to understand the processes and mechanisms that enable individuals to beat the odds, to strive against adversity also let to a change of focus on problems and needs of populations at risk, to possible assets and strengths within individual and communities. from a deficit model defining individuals and communities in negative terms, disregarding what is positive and what works well – needs imply dependence on welfare services, on external actors Asset-based approach: understanding of processes for creating, strengthening and protecting assets in individuals and communities that protect and promote adaptation

5. Defining Risk Epidemiological research Expected probabilities of life chances and adjustment based on earlier or current experiences Can encompass genetic, biological, psychological, environmental, or socio-economic factors The notion of risk used in the study of resilience research stems from epidemiological research, identifying expected probabilities of (mal) adjustment. Fundamental to the idea of risk is the predictability of life chances from earlier circumstances and/or current experiences. Risk or adversity can comprise genetic, biological, psychological, environmental or socio-economic factors that are associated with an increased probability of maladjustment The notion of risk used in the study of resilience research stems from epidemiological research, identifying expected probabilities of (mal) adjustment. Fundamental to the idea of risk is the predictability of life chances from earlier circumstances and/or current experiences. Risk or adversity can comprise genetic, biological, psychological, environmental or socio-economic factors that are associated with an increased probability of maladjustment

6. Variability of risk exposure Focus on single risk factors Accumulation of risk effects ? multiple risk models Statistical versus actual risk Plurality of meaning Duration and timing of risk effects Recognition of chronicity and severity of risk exposure: While early studies on resilience focused on a single risk factor, such as maternal psychopathology or experience of a stressful life event such as divorce, it soon became apparent that individual risk factors do not exert their effect in isolation, but in interaction with other influences. The relationship between any single risk factor and subsequent outcomes tends to be weak, and usually many variables are involved in determining an outcome. What distinguishes a high-risk individual from others is not so much exposure to a specific risk factor, but rather a life history characterised by multiple disadvantages. Serious risk emanates from the accumulation of risk effects, and it has been suggested that it is the number of these factors and their combined effect that exert a deleterious impact on developmental outcomes. Generally the use of multiple risk models has shown to be bring better predictions of individual outcomes than focus on single risk factors only. Another issue to be considered here is the differentiation between statistical and actual risk. Risks describe probabilities and not certainties. Even in circumstances where significant associations have been established between risk exposure and adjustment problems, questions may remain about the specific living conditions encountered by the individual. Individuals exposed to particular adverse life circumstances are treated as homogenous groups, despite possible variations in the degree to which their lives are actually shaped by the influence of the particular risk factors in question. Social class, for example, has been widely used as a risk indicator, although social class conveys little information about specific experiences to which children within a given level of social class are exposed . A child raised by working class parents will not necessarily experience poor quality care-giving. Neither would all individuals identified as being ‘at-risk’ consider this label appropriate for him or herself. There is a serious concern regarding stigmatisation and exclusion, predetermining failure of individuals exposed to severe hardship. Moreover, the experience of adversity might only be temporary and not long-lasting. Without more comprehensive information about the risk situation, it cannot be assumed that there actually is a significant risk exposure. Variability in risk exposure does, however, not necessarily invalidate research based on global risk indices such as social class. Knowledge about potential risk factors has been helpful in stimulating research into the processes and mechanisms by which these global risks influence individual adjustment, trying to clarify conditions in which they show their effect and where they don’t. Recognition of chronicity and severity of risk exposure: While early studies on resilience focused on a single risk factor, such as maternal psychopathology or experience of a stressful life event such as divorce, it soon became apparent that individual risk factors do not exert their effect in isolation, but in interaction with other influences. The relationship between any single risk factor and subsequent outcomes tends to be weak, and usually many variables are involved in determining an outcome. What distinguishes a high-risk individual from others is not so much exposure to a specific risk factor, but rather a life history characterised by multiple disadvantages. Serious risk emanates from the accumulation of risk effects, and it has been suggested that it is the number of these factors and their combined effect that exert a deleterious impact on developmental outcomes. Generally the use of multiple risk models has shown to be bring better predictions of individual outcomes than focus on single risk factors only. Another issue to be considered here is the differentiation between statistical and actual risk. Risks describe probabilities and not certainties. Even in circumstances where significant associations have been established between risk exposure and adjustment problems, questions may remain about the specific living conditions encountered by the individual. Individuals exposed to particular adverse life circumstances are treated as homogenous groups, despite possible variations in the degree to which their lives are actually shaped by the influence of the particular risk factors in question. Social class, for example, has been widely used as a risk indicator, although social class conveys little information about specific experiences to which children within a given level of social class are exposed . A child raised by working class parents will not necessarily experience poor quality care-giving. Neither would all individuals identified as being ‘at-risk’ consider this label appropriate for him or herself. There is a serious concern regarding stigmatisation and exclusion, predetermining failure of individuals exposed to severe hardship. Moreover, the experience of adversity might only be temporary and not long-lasting. Without more comprehensive information about the risk situation, it cannot be assumed that there actually is a significant risk exposure. Variability in risk exposure does, however, not necessarily invalidate research based on global risk indices such as social class. Knowledge about potential risk factors has been helpful in stimulating research into the processes and mechanisms by which these global risks influence individual adjustment, trying to clarify conditions in which they show their effect and where they don’t.

7. Positive adjustment Subjective evaluation Normative outcomes Who decides? Heterogeneity of adjustment: * Context dependency * Multiple domains of adjustment A major criticism concerning the study of resilience is the conceptualisation of positive or successful adjustment which involves value judgments about differences between expected and observed outcomes, as well as the causes of success and failure The identification of positive adjustment is based on normative or subjective judgements relating to a particular outcome or process of adjustment. What does it mean to be successful? And who defines success? It has been argued that the identification of positive adaptation does not require outstanding achievements, but should be based on behaviours within or above the expected average for a normative cohort (Masten, 2001). Describing resilience as ‘ordinary magic’ YET, also the criteria used to identify ‘average’ behaviour or achievements, are themselves culturally determined, reflecting biases grounded in the internalisation of values within the predominant culture we are participating in The outcomes used to define resilience may change as a function of age and can vary across contexts and cultures. Age-dependency, for example, is reflected in measures of children’s general competence that include assessments of academic attainment or behaviour adjustment. Academic attainment is the focus of this investigation, and positive academic attainment is defined by achievements at or above the average within the cohort populations – i.e. not on extraordinary achievements. Educational attainment is vitally important in our culture, and today it matters more than ever before in laying the foundations for future careers and well-being. Educational attainment is , for example, considered as an indicator of early developmental health. Yet, academic achievement is not necessarily a strong indicator of high competence or mastery among disadvantaged young people at all ages. While it is important to gain basic qualifications – young people themselves, especially when reaching adolescence, might be more concerned about coping with violence, drug abuse, sexual activities, peer pressure, marginalisation and disrespect Teenagers growing up in persisting and severe poverty, in households where the day-to-day needs of the family for additional resources is strongly present, the wish to leave school and gain full-time paid employment as soon as possible might be the resilient strategy rather than staying on in further education Concerns regarding the question of who decides the criteria of positive adaptation may become more and more pressing as the number of cross-cultural studies across and within nations increases (Ogbu, 1985). It is however also an issue in a society where generally prolonged participation in further education becomes ‘the norm’ for some subgroups of the population, especially the more privileged ones, while others are left behind. Researchers must become aware of the normative values underlying the identification of successful or unsuccessful outcomes, and learn to distinguish between their own values and interests and those of others who may have different or even opposing values (Kaplan, 1999). the choice of criteria for identifying positive adjustment has to be made explicit, as it has important implications for the definition of resilience. Another issue to be considered in defining resilience is the wide variety of outcomes across domains. The range of outcomes encompasses academic, emotional, behavioural, or physical adjustment. Not all children respond to adversity in the same way, and there are variations in adjustment. For example, it is possible that a child exposed to economic hardship shows good academic performance, but at the same time has behavioural problems (Luthar, 1991). Positive adaptation should be defined across multiple spheres, to avoid overly narrow definitions that can convey a misleading picture – and it should be noted that success in a particular domain cannot be assumed to generalise to other spheres – as resilience is not an all-or-nothing phenomenon. It has therefore been suggested to differentiate between specific domains of adjustment, i.e. academic, emotional or social resilience and to avoid notions of overall resilience (Luthar, 1993). A major criticism concerning the study of resilience is the conceptualisation of positive or successful adjustment which involves value judgments about differences between expected and observed outcomes, as well as the causes of success and failure The identification of positive adjustment is based on normative or subjective judgements relating to a particular outcome or process of adjustment. What does it mean to be successful? And who defines success? It has been argued that the identification of positive adaptation does not require outstanding achievements, but should be based on behaviours within or above the expected average for a normative cohort (Masten, 2001). Describing resilience as ‘ordinary magic’ YET, also the criteria used to identify ‘average’ behaviour or achievements, are themselves culturally determined, reflecting biases grounded in the internalisation of values within the predominant culture we are participating in The outcomes used to define resilience may change as a function of age and can vary across contexts and cultures. Age-dependency, for example, is reflected in measures of children’s general competence that include assessments of academic attainment or behaviour adjustment. Academic attainment is the focus of this investigation, and positive academic attainment is defined by achievements at or above the average within the cohort populations – i.e. not on extraordinary achievements. Educational attainment is vitally important in our culture, and today it matters more than ever before in laying the foundations for future careers and well-being. Educational attainment is , for example, considered as an indicator of early developmental health. Yet, academic achievement is not necessarily a strong indicator of high competence or mastery among disadvantaged young people at all ages. While it is important to gain basic qualifications – young people themselves, especially when reaching adolescence, might be more concerned about coping with violence, drug abuse, sexual activities, peer pressure, marginalisation and disrespect Teenagers growing up in persisting and severe poverty, in households where the day-to-day needs of the family for additional resources is strongly present, the wish to leave school and gain full-time paid employment as soon as possible might be the resilient strategy rather than staying on in further education Concerns regarding the question of who decides the criteria of positive adaptation may become more and more pressing as the number of cross-cultural studies across and within nations increases (Ogbu, 1985). It is however also an issue in a society where generally prolonged participation in further education becomes ‘the norm’ for some subgroups of the population, especially the more privileged ones, while others are left behind. Researchers must become aware of the normative values underlying the identification of successful or unsuccessful outcomes, and learn to distinguish between their own values and interests and those of others who may have different or even opposing values (Kaplan, 1999). the choice of criteria for identifying positive adjustment has to be made explicit, as it has important implications for the definition of resilience. Another issue to be considered in defining resilience is the wide variety of outcomes across domains. The range of outcomes encompasses academic, emotional, behavioural, or physical adjustment. Not all children respond to adversity in the same way, and there are variations in adjustment. For example, it is possible that a child exposed to economic hardship shows good academic performance, but at the same time has behavioural problems (Luthar, 1991). Positive adaptation should be defined across multiple spheres, to avoid overly narrow definitions that can convey a misleading picture – and it should be noted that success in a particular domain cannot be assumed to generalise to other spheres – as resilience is not an all-or-nothing phenomenon. It has therefore been suggested to differentiate between specific domains of adjustment, i.e. academic, emotional or social resilience and to avoid notions of overall resilience (Luthar, 1993).

8. Resilience in context Multidimensionality Developmental perspective Processes and mechanisms Holistic approach Values and meaning ? A developmental-contextual systems approach ? based in life course research

9. This Figure illustrates the different spheres of influence shaping individual adjustment. Notably the varied spheres do not exist in isolation, they are inter-related and mutually interdependent. Furthermore the model advocates a systems view to resilience (postulating a holistic approach) incorporating multi-level person-environment interactions. The model also captures the transactional nature of development over time, focusing on the reciprocal interactions between risk experiences and individual adjustment, which are embedded in the wider socio-historical context. The model is an explicit developmental model, allowing us to assess the dynamic interactions between individual and context. Any point in the life span has to be understood as the consequence of past experience and as the launch pad for subsequent experiences and conditions. For example, early adjustment patterns influence later adjustment, and early risk experiences are linked to the experience of risk at later life stages. Early experiences and the meanings attached to them are carried forward into consequent situations. Yet, life-long development may also involve processes that do not originate at birth or early childhood but in later periods.This Figure illustrates the different spheres of influence shaping individual adjustment. Notably the varied spheres do not exist in isolation, they are inter-related and mutually interdependent. Furthermore the model advocates a systems view to resilience (postulating a holistic approach) incorporating multi-level person-environment interactions. The model also captures the transactional nature of development over time, focusing on the reciprocal interactions between risk experiences and individual adjustment, which are embedded in the wider socio-historical context. The model is an explicit developmental model, allowing us to assess the dynamic interactions between individual and context. Any point in the life span has to be understood as the consequence of past experience and as the launch pad for subsequent experiences and conditions. For example, early adjustment patterns influence later adjustment, and early risk experiences are linked to the experience of risk at later life stages. Early experiences and the meanings attached to them are carried forward into consequent situations. Yet, life-long development may also involve processes that do not originate at birth or early childhood but in later periods.

10. Three National British Birth CohortsAge of Cohort Members by Historical EventsBoom Economy Recession Economy Knowledge Economy This chart illustrates the age of cohort members by when they experienced the historical events mentioned. The 1958 cohort was born just before the end of the baby boom in an era of liberalisation and extraordinary economic growth. This ‘Golden Age’ came to an end with the early 1970’s with the 1973 oil crisis, followed by two major recessions. The recessions staring in the late 1970s were the most serious since the past fifty years, and brought with them unprecedented levels of unemployment, especially youth unemployment rates soared to record levels. There has been a marked decline in manual and non-skilled work opportunities (the usual entry point for young workers) and an increase in white-collar and professional jobs. New technologies brought changes in labour market opportunities and demands for a more highly skilled labour force. While in the 1070s most young people could move straight from school at minimum school leaving age into a job – while in the 1990s most young people continued in further education. Increasing qualification levels have been observed in most industrialised countries – and the age when young people enter employment has been effectively delayed. In most Western societies, more education and skill development is required to meet the increasingly complex and challenging demands of employment. Thus, there has been increasing participation in further education – as well as a rapid and continuous rise of women entering the labour market (Gallie, 2000). There has also been a general increase in living standards – more people own their own homes and fewer families live in overcrowded houses with shared ameneties The cohort studies offer the unique opportunity to gain a better understanding of the context dependency of adjustment in changing times, and to assess the impact of social change on individual lives. The studies provide data on educational and occupational aspirations assessed during adolescence as well as adult outcomes. It it thus possible to identify precursors and concurrent factors influencing aspirations, as well as adult outcomes in work and family related roles. This chart illustrates the age of cohort members by when they experienced the historical events mentioned. The 1958 cohort was born just before the end of the baby boom in an era of liberalisation and extraordinary economic growth. This ‘Golden Age’ came to an end with the early 1970’s with the 1973 oil crisis, followed by two major recessions. The recessions staring in the late 1970s were the most serious since the past fifty years, and brought with them unprecedented levels of unemployment, especially youth unemployment rates soared to record levels. There has been a marked decline in manual and non-skilled work opportunities (the usual entry point for young workers) and an increase in white-collar and professional jobs. New technologies brought changes in labour market opportunities and demands for a more highly skilled labour force. While in the 1070s most young people could move straight from school at minimum school leaving age into a job – while in the 1990s most young people continued in further education. Increasing qualification levels have been observed in most industrialised countries – and the age when young people enter employment has been effectively delayed. In most Western societies, more education and skill development is required to meet the increasingly complex and challenging demands of employment. Thus, there has been increasing participation in further education – as well as a rapid and continuous rise of women entering the labour market (Gallie, 2000). There has also been a general increase in living standards – more people own their own homes and fewer families live in overcrowded houses with shared ameneties The cohort studies offer the unique opportunity to gain a better understanding of the context dependency of adjustment in changing times, and to assess the impact of social change on individual lives. The studies provide data on educational and occupational aspirations assessed during adolescence as well as adult outcomes. It it thus possible to identify precursors and concurrent factors influencing aspirations, as well as adult outcomes in work and family related roles.

11. Developmental Focus:Causation, Selection, and Cumulative Risk Effects

12. Figure 2 show the pathways between the latent or unobserved variables, which represent continuities and interactions of social risk and academic adjustment, and give the standardised coefficients for the structural model. The variables shown are all latent or unobserved variables. We find a stark chaining, or continuity of risk factors. Uneven life chances start at birth. Parental social class at birth predicts the experience of risk at subsequent ages, and the experience of risk at one time point increases the probability that risk will also be encountered at a later time point. Parental social class has a moderate influence on academic adjustment. The experience of social risk at birth influences the level of later academic attainment. We can also observe continuities in academic adjustment level over time. Academic attainment at one time point is a significant predictor of academic attainment at a later time point. Continuities occur because current adjustment encompasses previous adjustment as well as earlier structural and functional change. The detrimental effect of experiencing disadvantage at one measurement point is carried forward into the future via decreased individual adjustment levels. There are small time-lagged risk effects indicating the added negative influence of social risk on subsequent attainment not accounted for by the risk carried forward in time. These time-lagged risk effects are only of small size, yet in both cohorts they are greatest at the transition from late childhood to adolescence. The experience of early disadvantage weakens individual adjustment, and this detrimental effect is then carried forward into the future. Subsequent experiences of adversity add to the deterioration of already reduced adjustment. A general premise of life course studies postulates that adaptations to change are influenced by what people bring to the new situation. If individual adjustment is already weakened at a very early age, it becomes more and more difficult to fully develop one’s potential. This negative chain effect undermines the academic adjustment of the young person, and ultimately the individual attainments in adulthood. Generally the results imply that cumulative adversity has effects beyond those associated with current or early adversity Figure 2 show the pathways between the latent or unobserved variables, which represent continuities and interactions of social risk and academic adjustment, and give the standardised coefficients for the structural model. The variables shown are all latent or unobserved variables. We find a stark chaining, or continuity of risk factors. Uneven life chances start at birth. Parental social class at birth predicts the experience of risk at subsequent ages, and the experience of risk at one time point increases the probability that risk will also be encountered at a later time point. Parental social class has a moderate influence on academic adjustment. The experience of social risk at birth influences the level of later academic attainment. We can also observe continuities in academic adjustment level over time. Academic attainment at one time point is a significant predictor of academic attainment at a later time point. Continuities occur because current adjustment encompasses previous adjustment as well as earlier structural and functional change. The detrimental effect of experiencing disadvantage at one measurement point is carried forward into the future via decreased individual adjustment levels. There are small time-lagged risk effects indicating the added negative influence of social risk on subsequent attainment not accounted for by the risk carried forward in time. These time-lagged risk effects are only of small size, yet in both cohorts they are greatest at the transition from late childhood to adolescence. The experience of early disadvantage weakens individual adjustment, and this detrimental effect is then carried forward into the future. Subsequent experiences of adversity add to the deterioration of already reduced adjustment. A general premise of life course studies postulates that adaptations to change are influenced by what people bring to the new situation. If individual adjustment is already weakened at a very early age, it becomes more and more difficult to fully develop one’s potential. This negative chain effect undermines the academic adjustment of the young person, and ultimately the individual attainments in adulthood. Generally the results imply that cumulative adversity has effects beyond those associated with current or early adversity

13. Protective Factors and Processes Individual attributes Characteristics of the family Interactions with ‘significant others’ Aspects of the wider social context Positive social, emotional and educational experiences can partially offset the effects of material deprivation Although there is a strong relationship between exposure to cumulative adversity and developmental outcomes, the relationship is by no means deterministic. The search for potential protective or buffering factors that mitigate, exacerbate, or mediate the risk associated with adverse socio-economic conditions, has led to the identification of three broad set of variables operating as resource factors that may impede or halt the impact of adverse experiences and promote academic resilience in the face of socio-economic adversity. These factors include characteristics of the individual, as well as factors that are external to the individual such as the family environment or the wider social context individual attributes: individuals demonstrating early academic resilience despite the experience of socio-economic hardship generally also performed better in most other school tests, demonstrated fewer persistent behaviour problems, and had more hobbies and social contacts than their more vulnerable peers. They enjoyed school, showed a strong belief in their own ability, and were academically motivated, i.e. they wanted further education after the minimum school leaving age. They furthermore showed good planning in their partnership and career choices, and had a positive outlook on live. Characteristics of their families: factors associated with positive adjustment during childhood and adolescents included a stable and supportive family environment, parents who showed interest in their child’s education and wanted their child to continue with education after the minimum school leaving age. A supportive family environment is furthermore characterised by parents who read to their child, who took an active interest and involvement in their education and career planning, and who took the children out for joint activities. Another important factor was a supportive father who helped the mother with the household chores (Schoon and Parsons, 2002b). Aspects of the wider social context: not only parents but also significant others in the wider social context provide vital sources of support, as for example teachers who recognise the children’s capabilities and who encouraged and supported their educational and occupational strivings. There was also evidence to suggest the importance of positive community forces, such as support and cohesion among neighbours and a sense of belonging to the community, as well as the benefits of a positive school environment. This triarchic set of factors can be understood as psychosocial resources that support or promote adaptive development. Individuals with many, or high levels of personal and social resources are more effective in coping with adversity than individuals with fewer (or lower level) resources. Positive social, emotional and educational experiences can partially offset the effects of material deprivation Although there is a strong relationship between exposure to cumulative adversity and developmental outcomes, the relationship is by no means deterministic. The search for potential protective or buffering factors that mitigate, exacerbate, or mediate the risk associated with adverse socio-economic conditions, has led to the identification of three broad set of variables operating as resource factors that may impede or halt the impact of adverse experiences and promote academic resilience in the face of socio-economic adversity. These factors include characteristics of the individual, as well as factors that are external to the individual such as the family environment or the wider social context individual attributes: individuals demonstrating early academic resilience despite the experience of socio-economic hardship generally also performed better in most other school tests, demonstrated fewer persistent behaviour problems, and had more hobbies and social contacts than their more vulnerable peers. They enjoyed school, showed a strong belief in their own ability, and were academically motivated, i.e. they wanted further education after the minimum school leaving age. They furthermore showed good planning in their partnership and career choices, and had a positive outlook on live. Characteristics of their families: factors associated with positive adjustment during childhood and adolescents included a stable and supportive family environment, parents who showed interest in their child’s education and wanted their child to continue with education after the minimum school leaving age. A supportive family environment is furthermore characterised by parents who read to their child, who took an active interest and involvement in their education and career planning, and who took the children out for joint activities. Another important factor was a supportive father who helped the mother with the household chores (Schoon and Parsons, 2002b). Aspects of the wider social context: not only parents but also significant others in the wider social context provide vital sources of support, as for example teachers who recognise the children’s capabilities and who encouraged and supported their educational and occupational strivings. There was also evidence to suggest the importance of positive community forces, such as support and cohesion among neighbours and a sense of belonging to the community, as well as the benefits of a positive school environment. This triarchic set of factors can be understood as psychosocial resources that support or promote adaptive development. Individuals with many, or high levels of personal and social resources are more effective in coping with adversity than individuals with fewer (or lower level) resources.

15. Models of Resilience Challenge model Cumulative effect model Protection effect model How do these different factors operate in changing the negative trajectory associated with risk? Previous research has identified three models of resilience: the protective, cumulative, and challenge model, each describing possible links between risk, resource factors, and adjustment. The challenge model of resilience suggests that low levels of risk exposure may have beneficial or steeling effects, providing a chance to practice problem solving skills and to mobilise resources (Masten, 1999; Rutter, 1987). The risk exposure must be challenging enough to stimulate a response, yet must not be overpowering. The crux of the challenge model is that moderate levels of risk exposure open up the opportunity to learn how to overcome adversity. From a developmental perspective, the challenge model can also be considered as a model of inoculation preparing the developing person to overcome significant risks in the future In the cumulative effect model, the risk or protective effect depends on the number of available resources, which combine additively to compensate or counteract the effects of adversity. The cumulation of assets or resources will outweigh the risks. Increasing the protective resources in quality or number could theoretically offset the negative effects of risk or adversity, or improve positive adjustment in general. According to such a cumulative or main effect model, resource factors can have an equally beneficial effect on those exposed and those not exposed to adversity. And indeed, most of the well-established resource factors, such as parenting skills or self esteem show their effect in general (low-adversity) conditions as well as high-risk conditions. While some authors use the term protective factors in a wider sense, referring to all resource factors associated with positive adaptation among at-risk groups, regardless of their interaction with exposure to adversity, or whether there are benefits to not-at-risk groups as well. The term ‘protective factors’ has originally only be used to describe beneficial affects in the presence of risk condition but not in their absence. The exposure to a protective factor should have beneficial effects on those exposed to the risk factor, but not benefit those not exposed to the risk factor, i.e. there should be an interactive relationship between the protective factor, the risk exposure, and the outcome. Interaction effects are, however, usually small in magnitude and are thus difficult to detect in variable based analysis How do these different factors operate in changing the negative trajectory associated with risk? Previous research has identified three models of resilience: the protective, cumulative, and challenge model, each describing possible links between risk, resource factors, and adjustment. The challenge model of resilience suggests that low levels of risk exposure may have beneficial or steeling effects, providing a chance to practice problem solving skills and to mobilise resources (Masten, 1999; Rutter, 1987). The risk exposure must be challenging enough to stimulate a response, yet must not be overpowering. The crux of the challenge model is that moderate levels of risk exposure open up the opportunity to learn how to overcome adversity. From a developmental perspective, the challenge model can also be considered as a model of inoculation preparing the developing person to overcome significant risks in the future In the cumulative effect model, the risk or protective effect depends on the number of available resources, which combine additively to compensate or counteract the effects of adversity. The cumulation of assets or resources will outweigh the risks. Increasing the protective resources in quality or number could theoretically offset the negative effects of risk or adversity, or improve positive adjustment in general. According to such a cumulative or main effect model, resource factors can have an equally beneficial effect on those exposed and those not exposed to adversity. And indeed, most of the well-established resource factors, such as parenting skills or self esteem show their effect in general (low-adversity) conditions as well as high-risk conditions. While some authors use the term protective factors in a wider sense, referring to all resource factors associated with positive adaptation among at-risk groups, regardless of their interaction with exposure to adversity, or whether there are benefits to not-at-risk groups as well. The term ‘protective factors’ has originally only be used to describe beneficial affects in the presence of risk condition but not in their absence. The exposure to a protective factor should have beneficial effects on those exposed to the risk factor, but not benefit those not exposed to the risk factor, i.e. there should be an interactive relationship between the protective factor, the risk exposure, and the outcome. Interaction effects are, however, usually small in magnitude and are thus difficult to detect in variable based analysis

16. Population versus Person centred approach Polulation/Variable centred approach: Main effects Interaction effects Person centred approach: How do variables combine in individuals Connection between risk and adjustment The link between risk and manifest competence can be analysed using either a variable- or a person based approach, each of which has its specific strengths and weaknesses. The variable-based approach focuses on general nomothetic probabilities, examining either main effect models or interaction effects (Luthar et al., 2000; Masten et al., 1988; Rutter, 1987). Variable-based approaches use a variety of multivariate strategies such as hierarchical regression, path analysis or structural equation modelling to identify linkages between risks, resources and adaptive functioning. A major problem with this approach is that it tells us nothing about how many individuals are actually facing high levels of risk or are manifesting high levels of competence (Luthar & Cushing, 1999). Furthermore interaction effects are notoriously unstable and difficult to establish because they are generally associated with small effect sizes (Rutter, 1990). The person–, or individual-based approach, on the other hand, involves the selection or grouping of specific individuals based on selected criteria (Block, 1980; Cairns et al., 1998; Magnusson & Bergman, 1988; Masten, 2001). There are various ways in which categories can be formed, based on theoretical considerations, classificatory analysis, empirical grouping or the identification of extreme cases (Hinde, 1998). Empirical grouping techniques for example encompass cluster analysis, latent profile analysis, log linear modelling, or configural frequency analysis (Magnusson, 1995). Here the grouping of individuals was based on extreme cases in order to assess interaction effects of protective factors in high and low risk conditions. The link between risk and manifest competence can be analysed using either a variable- or a person based approach, each of which has its specific strengths and weaknesses. The variable-based approach focuses on general nomothetic probabilities, examining either main effect models or interaction effects (Luthar et al., 2000; Masten et al., 1988; Rutter, 1987). Variable-based approaches use a variety of multivariate strategies such as hierarchical regression, path analysis or structural equation modelling to identify linkages between risks, resources and adaptive functioning. A major problem with this approach is that it tells us nothing about how many individuals are actually facing high levels of risk or are manifesting high levels of competence (Luthar & Cushing, 1999). Furthermore interaction effects are notoriously unstable and difficult to establish because they are generally associated with small effect sizes (Rutter, 1990). The person–, or individual-based approach, on the other hand, involves the selection or grouping of specific individuals based on selected criteria (Block, 1980; Cairns et al., 1998; Magnusson & Bergman, 1988; Masten, 2001). There are various ways in which categories can be formed, based on theoretical considerations, classificatory analysis, empirical grouping or the identification of extreme cases (Hinde, 1998). Empirical grouping techniques for example encompass cluster analysis, latent profile analysis, log linear modelling, or configural frequency analysis (Magnusson, 1995). Here the grouping of individuals was based on extreme cases in order to assess interaction effects of protective factors in high and low risk conditions.

17. Identification of Resilience(example: NCDS at age 7) Findings presented In the previous part were based on a variable-based approach of analysis focusing on general nomothetic probabilities. A major problem with this approach is that it tells us nothing about how many individuals are actually facing high levels of risk or are manifesting high levels of competence (Luthar & Cushing, 1999). Furthermore interaction effects are notoriously unstable and difficult to establish because they are generally associated with small effect sizes (Rutter, 1983). The person–, or individual-based approach, on the other hand, involves the selection or grouping of specific individuals based on selected criteria. There are various ways in which categories can be formed, including theoretical considerations, classificatory analysis, empirical grouping or the identification of extreme cases Has been argued that individual-based approaches may be less prone to statistically artifactual insights inherent in variable-based approaches, because a particular subset of high risk, high competence individuals has been identified (Luther & Cushing, 1999). Four groups describing different patterns of early academic adjustment were identified on the basis of experiences of socio-economic adversity and concurrent reading ability. Socio-economic adversity was defined by the social risk index, and academic adjustment was measured by early reading ability assessed at age 5 in BCS70 and age 7 in NCDS. Children who despite experiencing high socio-economic risk (4 or more risk indicators) showed good reading ability (i.e. above the median) in early childhood were identified as educationally resilient. Children with low reading ability (below the median) in similar circumstances were defined as vulnerable. The experiences of these two socially disadvantaged groups were compared and contrasted to a control group of individuals who grew up in circumstances characterised by low socio-economic risk. Among these privileged children a differentiation was made between individuals with a reading ability above and below the median. The first group was defined as the ‘multiple advantaged’, the latter had been labelled as ‘under achievers’. The categorisation of sample members into four groups aimed to identify high and low risk children with high or low levels of educational adjustment. Comparisons between the groups allows for person by situation interactions to be explored more fully (Hinde, 1998). Findings presented In the previous part were based on a variable-based approach of analysis focusing on general nomothetic probabilities. A major problem with this approach is that it tells us nothing about how many individuals are actually facing high levels of risk or are manifesting high levels of competence (Luthar & Cushing, 1999). Furthermore interaction effects are notoriously unstable and difficult to establish because they are generally associated with small effect sizes (Rutter, 1983). The person–, or individual-based approach, on the other hand, involves the selection or grouping of specific individuals based on selected criteria. There are various ways in which categories can be formed, including theoretical considerations, classificatory analysis, empirical grouping or the identification of extreme cases Has been argued that individual-based approaches may be less prone to statistically artifactual insights inherent in variable-based approaches, because a particular subset of high risk, high competence individuals has been identified (Luther & Cushing, 1999). Four groups describing different patterns of early academic adjustment were identified on the basis of experiences of socio-economic adversity and concurrent reading ability. Socio-economic adversity was defined by the social risk index, and academic adjustment was measured by early reading ability assessed at age 5 in BCS70 and age 7 in NCDS. Children who despite experiencing high socio-economic risk (4 or more risk indicators) showed good reading ability (i.e. above the median) in early childhood were identified as educationally resilient. Children with low reading ability (below the median) in similar circumstances were defined as vulnerable. The experiences of these two socially disadvantaged groups were compared and contrasted to a control group of individuals who grew up in circumstances characterised by low socio-economic risk. Among these privileged children a differentiation was made between individuals with a reading ability above and below the median. The first group was defined as the ‘multiple advantaged’, the latter had been labelled as ‘under achievers’. The categorisation of sample members into four groups aimed to identify high and low risk children with high or low levels of educational adjustment. Comparisons between the groups allows for person by situation interactions to be explored more fully (Hinde, 1998).

18. What are the long-term academic attainments of the four different groups? Figure shows the mean z-scores, standardised for the whole sample, during early and mid childhood and adolescence – the lines differentiate the attainment of children classified as ‘normative’ (or multiple advantaged), privileged low achievers, resilient and vulnerable In NCDS – Early academic resilience is to some extent maintained throughout childchool and adolescents – yet children with good reading skills from relatively disadvantaged family backgrounds did not maintain their performance to the same level as their more privileged peers – in both cohorts there is a drop at age 11 – which marks the transition from primary to secondary school. Another hurdle at which early resilient stumble is the exam at age 16 . This finding highlights again the need for continued support for disadvantaged young people. It is never too early and never too late for interventions Advantaged early low-achievers in both cohorts were able to improve their performance – early resilient children from disadvantaged backgrounds were falling behind – leading to an increasing divide between children from privileged and less privileged backgrounds – especially so in the later born cohort Changes in test scores among the selected extreme groups may only represent a regression towards the mean – yet there were not only changes in level but also in position The improved performance of advantaged early under-achievers as well as that of vulnerable individuals at age 10/11 and 16 could possibly also be explained by increased attention and support aimed at and provided for low achievers in general. Young people from disadvantaged backgrounds who appear to be getting by in school, receive however less attention and support Findings also suggest an increasing polarisation of achievement, with young people from relative privileged backgrounds who demonstrated below average academic attainment in early childhood benefitting most from changes in educational system, and the increasing expansion of further education. Comparing highest qualifications in adulthood among the four groups, the data indicate that early academic resilience showed some degree of stability throughout the transition from childhood to adulthood, and appears to be associated with continued participation in further education. Yet manifest early resilience did not enable young people to completely overcome the adversity of disadvantage. Privileged early under-achievers who showed below average reading ability in early childhood were outperforming their less privileged peers, who demonstrated above average academic potential in early childhood. What are the long-term academic attainments of the four different groups? Figure shows the mean z-scores, standardised for the whole sample, during early and mid childhood and adolescence – the lines differentiate the attainment of children classified as ‘normative’ (or multiple advantaged), privileged low achievers, resilient and vulnerable In NCDS – Early academic resilience is to some extent maintained throughout childchool and adolescents – yet children with good reading skills from relatively disadvantaged family backgrounds did not maintain their performance to the same level as their more privileged peers – in both cohorts there is a drop at age 11 – which marks the transition from primary to secondary school. Another hurdle at which early resilient stumble is the exam at age 16 . This finding highlights again the need for continued support for disadvantaged young people. It is never too early and never too late for interventions Advantaged early low-achievers in both cohorts were able to improve their performance – early resilient children from disadvantaged backgrounds were falling behind – leading to an increasing divide between children from privileged and less privileged backgrounds – especially so in the later born cohort Changes in test scores among the selected extreme groups may only represent a regression towards the mean – yet there were not only changes in level but also in position The improved performance of advantaged early under-achievers as well as that of vulnerable individuals at age 10/11 and 16 could possibly also be explained by increased attention and support aimed at and provided for low achievers in general. Young people from disadvantaged backgrounds who appear to be getting by in school, receive however less attention and support Findings also suggest an increasing polarisation of achievement, with young people from relative privileged backgrounds who demonstrated below average academic attainment in early childhood benefitting most from changes in educational system, and the increasing expansion of further education. Comparing highest qualifications in adulthood among the four groups, the data indicate that early academic resilience showed some degree of stability throughout the transition from childhood to adulthood, and appears to be associated with continued participation in further education. Yet manifest early resilience did not enable young people to completely overcome the adversity of disadvantage. Privileged early under-achievers who showed below average reading ability in early childhood were outperforming their less privileged peers, who demonstrated above average academic potential in early childhood.

19. Pathways linking socio-economic adversity and child development Family Stress Model (Conger et al., 1992, 93; Elder & Caspi, 1988; McLoyd, 1989): Links family economic stress to problematic adolescent development Postulates that economic stress affects adolescent adjustment indirectly through family processes (i.e. through parental mood, relationship difficulties, and parenting)  The family stress model (and its variants) argues that low household income influences children’s development through its impact on parental mental health which then influences parenting practices (Conger et al., 1992, 1993; Elder & Caspi, 1988; McLoyd, 1989). This model has been more effective in predicting behavioural adjustment than cognitive adjustment in children and adolescents (Conger & Elder, 1994; Linver, Brooks-Gunn & Kohen, 2002). For example, it has been found that economic pressure was indirectly related to poor parenting, through high parental depressed mood, and that poor parenting was related to adolescent externalising behaviour (Conger et al., 1993). the family stress model,, postulates that economic stress experienced in the family influences children’s development through its impact on parental mental health which then influences parenting practices. Although the family stress model has been widely used to explain the association between material or economic adversity and child development, the model and its variants have mostly been applied for the study of cognitive and behavioural development among children and especially among adolescents. Although the effects of material hardship appear to be strongest during the preschool and early school years rather than during adolescence (Brooks-Gunn, Duncan, 1997) there are very few studies examining the family stress model with younger children or infants (execpt Brody & Flor, 1997; Elder & Caspi, 1988; Jackson, Books-Gunn, Huang & Glassman, 2000; Linver, Brooks-Gunn & Kohen, 2002). The aim of the following paper is to apply the family stress perspective to the study of infants at 9 months of age, drawing on data collected for the UK Millennium Study. Another contribution of this paper is that it will extent the family stress model to examine the influence of fathers on early child development. The role of fathers is still a vastly understudied area, especially regarding the development of young children and infants, which have focused nearly exclusively on mothers and their children (Demo & Cox, 2000). The outcome variable is problem behaviour as reported by the mother. Although there is some evidence to suggest that maternal depression and stress could lead to biased ratings of problem behaviour – parental ratings have shown to be associated with observational assessments (Kochanska et al, 1996).  The family stress model (and its variants) argues that low household income influences children’s development through its impact on parental mental health which then influences parenting practices (Conger et al., 1992, 1993; Elder & Caspi, 1988; McLoyd, 1989). This model has been more effective in predicting behavioural adjustment than cognitive adjustment in children and adolescents (Conger & Elder, 1994; Linver, Brooks-Gunn & Kohen, 2002). For example, it has been found that economic pressure was indirectly related to poor parenting, through high parental depressed mood, and that poor parenting was related to adolescent externalising behaviour (Conger et al., 1993). the family stress model,, postulates that economic stress experienced in the family influences children’s development through its impact on parental mental health which then influences parenting practices. Although the family stress model has been widely used to explain the association between material or economic adversity and child development, the model and its variants have mostly been applied for the study of cognitive and behavioural development among children and especially among adolescents. Although the effects of material hardship appear to be strongest during the preschool and early school years rather than during adolescence (Brooks-Gunn, Duncan, 1997) there are very few studies examining the family stress model with younger children or infants (execpt Brody & Flor, 1997; Elder & Caspi, 1988; Jackson, Books-Gunn, Huang & Glassman, 2000; Linver, Brooks-Gunn & Kohen, 2002). The aim of the following paper is to apply the family stress perspective to the study of infants at 9 months of age, drawing on data collected for the UK Millennium Study. Another contribution of this paper is that it will extent the family stress model to examine the influence of fathers on early child development. The role of fathers is still a vastly understudied area, especially regarding the development of young children and infants, which have focused nearly exclusively on mothers and their children (Demo & Cox, 2000). The outcome variable is problem behaviour as reported by the mother. Although there is some evidence to suggest that maternal depression and stress could lead to biased ratings of problem behaviour – parental ratings have shown to be associated with observational assessments (Kochanska et al, 1996).

20. The Family Stress Model Figure provides a graphic illustration of the family stress model of economic hardship. The theoretical model proposes that the experience of material hardship, such as low income, influences the emotions, behaviours and relationships of family members. The model postulates that the effects of family economic pressures are felt by children at least in part through their influence on parental emotional state and behaviour (Conger, et al., 1992). Conditions of poverty increase parents’ psychological distress, which in turn leads to relationship conflict . Economic difficulties disrupt psychological well-being, and hardship related depression will increase the risk of relationship conflict. Parental depressed mood in conjunction with relationship conflict decrease effective parenting, which in turn influences child adjustment. Parents who are already irritable and in conflict with each other are less likely to show effective parenting – there is a ‘spillover effect’ from relationship conflict to behaviour towards children The original model postulates that economic hardship affects the degree of economic pressure experienced by the family which in turn influences coercive family processes. Economic hardship is translated through the experience of economic pressure into psychological adjustment. The model raises concerns regarding subjective versus objective experience of economic difficulties (Conger et al., 1994). While the construct of economic pressure represents a subjective reality, the direct experience of material hardship is a objective indicator of material resources available to the family, such as low income, poor housing, or car ownership.Figure provides a graphic illustration of the family stress model of economic hardship. The theoretical model proposes that the experience of material hardship, such as low income, influences the emotions, behaviours and relationships of family members. The model postulates that the effects of family economic pressures are felt by children at least in part through their influence on parental emotional state and behaviour (Conger, et al., 1992). Conditions of poverty increase parents’ psychological distress, which in turn leads to relationship conflict . Economic difficulties disrupt psychological well-being, and hardship related depression will increase the risk of relationship conflict. Parental depressed mood in conjunction with relationship conflict decrease effective parenting, which in turn influences child adjustment. Parents who are already irritable and in conflict with each other are less likely to show effective parenting – there is a ‘spillover effect’ from relationship conflict to behaviour towards children The original model postulates that economic hardship affects the degree of economic pressure experienced by the family which in turn influences coercive family processes. Economic hardship is translated through the experience of economic pressure into psychological adjustment. The model raises concerns regarding subjective versus objective experience of economic difficulties (Conger et al., 1994). While the construct of economic pressure represents a subjective reality, the direct experience of material hardship is a objective indicator of material resources available to the family, such as low income, poor housing, or car ownership.

21. Extension of Family Stress Model: Application of the model to young infants Representative UK sample (the Millennium Cohort) Longitudinal perspective Most variations of the family stress model have been applied to adolescents – there are very few studies on young children or infants, although the effects of poverty have been described to be stronger among young children than among adolescents (Duncan and Brooks-Gunn) Studies on young children, in turn, are mostly based on studies involving mothers. Much of what we know about families with young children are based on mothers’ perception of family relations, and there are relatively few studies that included the father’s perspective. By studying infants in two-parent families we will be able to examine the differential impact of material hardship on mothers and fathers. This aspect of the investigation is important, because earlier research suggests, that fathers may be more adversely affected than mothers by family financial difficulties (McLoyd, 1989). We will run separate analysis for mothers and fathers to determine whether parental gender differences in the stress process will be present for a contemporary UK sample.Most variations of the family stress model have been applied to adolescents – there are very few studies on young children or infants, although the effects of poverty have been described to be stronger among young children than among adolescents (Duncan and Brooks-Gunn) Studies on young children, in turn, are mostly based on studies involving mothers. Much of what we know about families with young children are based on mothers’ perception of family relations, and there are relatively few studies that included the father’s perspective. By studying infants in two-parent families we will be able to examine the differential impact of material hardship on mothers and fathers. This aspect of the investigation is important, because earlier research suggests, that fathers may be more adversely affected than mothers by family financial difficulties (McLoyd, 1989). We will run separate analysis for mothers and fathers to determine whether parental gender differences in the stress process will be present for a contemporary UK sample.

22. Millennium Cohort:Indicators of Child Development Behavioural adjustment age 9 mths: Child Temperament (9 items from the Carey Infant Temperament Scale: regularity and adaptability) age 3 years: Strength and Difficulties Questionnaire: SDQ (Goodman, 1997) age 5 years: SDQ Cognitive development age 3 years: Bracken School Readiness Scale; BAS naming vocabulary age 5 years: BAS (verbal and performance subtests: verbal comprehension, naming vocabulary, pattern construction, picture similarities) (N=18819 babies born into 18553 families)

23. Indicators of Material Hardship

24. Maternal psychological distress Age 9mths: shortened version (9 items) of the Malaise Inventory (Rutter et al., 1970) 36 mths: Kessler K6 (Kessler et al, 2002) Both tests are: ? Self completion instruments ? good reliability (? >.80) ? Good validity (correlates with previously diagnosed depression and currently treated depression) In terms of diagnosed and treated depression, 25% of the main respondent (usually the mother) and 9% of their partners have had at some point been diagnosed as being depressed by their GP, and of these 36% of main respondents and 25% of their partners were currently being treated for depression. In addition, general levels of psychological distress, or depression, among parents of the 9-month old babies were assessed using a shortened version of the Malaise Inventory (Rutter et al., 1970). This self-completion measure has been widely used in both general population studies (McGee et al., 1986; Rutter et al., 1976; Rodgers et al., 1999) and in investigations of high-risk groups, notably informal carers (Grant et al., 1990). In the Millennium Cohort Study, only 9 of the original 24 items of the Malaise were used. The internal consistency of the shortened scale is acceptable (r=.72 for women and r=.67 for men) and it correlates well with the information on previously diagnosed depression (r=.36), currently treated depression (r=.32), and general life satisfaction (r=-.38). The percentage of parents reporting individual symptoms is shown in Table 3. All but one item (often getting into a violent rage) are more commonly reported by mothers than by fathers. Generally mothers show higher levels of psychological distress than fathers. In comparison to other general population samples (Rodgers et al., 1999) parents of 9-months olds in the Millennium Cohort are more likely to feel tired most of the time, every little thing is more likely to get on their nerves, yet they are less likely to get into a violent rage. Furthermore fathers of 9 months old babies are more likely to often get worried about things than men in other population samples. 33% of mothers and 36% of fathers report no symptoms at all. Yet, 22% of mothers and 17% of fathers show 3 or more symptoms of psychological distress. It is likely that some of the symptoms are related to the fact of having a young baby in the house and that they will change as the child grows older. In terms of diagnosed and treated depression, 25% of the main respondent (usually the mother) and 9% of their partners have had at some point been diagnosed as being depressed by their GP, and of these 36% of main respondents and 25% of their partners were currently being treated for depression. In addition, general levels of psychological distress, or depression, among parents of the 9-month old babies were assessed using a shortened version of the Malaise Inventory (Rutter et al., 1970). This self-completion measure has been widely used in both general population studies (McGee et al., 1986; Rutter et al., 1976; Rodgers et al., 1999) and in investigations of high-risk groups, notably informal carers (Grant et al., 1990). In the Millennium Cohort Study, only 9 of the original 24 items of the Malaise were used. The internal consistency of the shortened scale is acceptable (r=.72 for women and r=.67 for men) and it correlates well with the information on previously diagnosed depression (r=.36), currently treated depression (r=.32), and general life satisfaction (r=-.38). The percentage of parents reporting individual symptoms is shown in Table 3. All but one item (often getting into a violent rage) are more commonly reported by mothers than by fathers. Generally mothers show higher levels of psychological distress than fathers. In comparison to other general population samples (Rodgers et al., 1999) parents of 9-months olds in the Millennium Cohort are more likely to feel tired most of the time, every little thing is more likely to get on their nerves, yet they are less likely to get into a violent rage. Furthermore fathers of 9 months old babies are more likely to often get worried about things than men in other population samples. 33% of mothers and 36% of fathers report no symptoms at all. Yet, 22% of mothers and 17% of fathers show 3 or more symptoms of psychological distress. It is likely that some of the symptoms are related to the fact of having a young baby in the house and that they will change as the child grows older.

25. Parenting at age 3 years Parent-child relationship scale (Pianta, 1994): maternal sensitivity 15 items (warm, affectionate relationship, comfort, praise, anger, discipline, being in tune) maternal report good internal consistency (alpha =.77) summary score

26. Cognitive Stimulation at age 3 years Maternal report on how frequently the child is taught: the alphabet counting songs Summary score (alpha = .64)

27. Control Variables Mother’s age at birth of child/at interview Mother’s education (below GCSE; GCSE and above) Parental employment (in paid employment or not) Mother’s ethnicity (white versus other) Sex of child Birthweight (< 2500 grams) Prematurity (gestation < 37 weeks) Total number of siblings living in the household

28. Family Hardship and Behaviour Problems

29. Family hardship and problem behaviour at 3 years (time-weighted)

30. Family Hardship and School Readiness

31. Family hardship and school readiness at 3 years (time-weighted)

32. Family Investment Model Theoretical models linking the experience of material hardship to child outcomes have been developed, focusing in particular on the mediating role of family psychological stress and family investments in children (Conger & Elder, 1994; Elder & Caspi, 1988; Haveman & Wolfe, 1994). The family stress model (FSM) postulates that socio-economic adversity, in particular family income, influences children’s cognitive development and behaviour through its impact on parental emotional distress, which influences parenting practices, which in turn are associated with child outcomes (Conger, Rueter, Conger, Crockett, & Silbereisen, 2000; Conger, Ge, Elder, Lorenz, & Simons, 1994; Conger et al., 1992, , 1993; Elder & Caspi, 1988; McLoyd, 1989; McLoyd, 1994). It is argued that perceptions of financial strain associated with low income, and/or job loss, have a negative impact on parental mental health and parenting practices (Conger & Elder, 1994; Elder & Caspi, 1988). A second explanation, the family investment model (FIM), posits that income is associated with children’s development because it enables families to purchase goods and services that are beneficial to the child’s development, i.e. that they invest in their children’s human capital (Becker & Thomes, 1986; Haveman & Wolfe, 1994; Mayer, 1997). While the FSM postulates that family hardship influences child development because poverty decreases the quality of parents non-monetary capacity, such as their mental well-being and interactions with their children, evidence in support of the FIM suggests that poor parents have less money to purchase or produce important inputs into their young children’s development, such as an enriched home learning environment, childcare settings outside the home, or high quality schooling. Theoretical models linking the experience of material hardship to child outcomes have been developed, focusing in particular on the mediating role of family psychological stress and family investments in children (Conger & Elder, 1994; Elder & Caspi, 1988; Haveman & Wolfe, 1994). The family stress model (FSM) postulates that socio-economic adversity, in particular family income, influences children’s cognitive development and behaviour through its impact on parental emotional distress, which influences parenting practices, which in turn are associated with child outcomes (Conger, Rueter, Conger, Crockett, & Silbereisen, 2000; Conger, Ge, Elder, Lorenz, & Simons, 1994; Conger et al., 1992, , 1993; Elder & Caspi, 1988; McLoyd, 1989; McLoyd, 1994). It is argued that perceptions of financial strain associated with low income, and/or job loss, have a negative impact on parental mental health and parenting practices (Conger & Elder, 1994; Elder & Caspi, 1988). A second explanation, the family investment model (FIM), posits that income is associated with children’s development because it enables families to purchase goods and services that are beneficial to the child’s development, i.e. that they invest in their children’s human capital (Becker & Thomes, 1986; Haveman & Wolfe, 1994; Mayer, 1997). While the FSM postulates that family hardship influences child development because poverty decreases the quality of parents non-monetary capacity, such as their mental well-being and interactions with their children, evidence in support of the FIM suggests that poor parents have less money to purchase or produce important inputs into their young children’s development, such as an enriched home learning environment, childcare settings outside the home, or high quality schooling.

33. Family hardship and school readiness at 3 years (time-weighted) Family investment model: Income is associated with child development because it enables families to invest in, purchase , or provide goods and services that are beneficial to the child’s development. Investment in the child’s human capital, i.e. education Family stess model suggests that family hardship influences child development because poverty decreases the quality of parents non-monetary capcity, ie their mental well-being and inteeractions with the child, the family investment model suggest that poor parents have less opportunity to purchase or produce important inputs into their children’s develoopment, usch as an enriched home learning environment, childcare settings, or high quality schooling. Both models explain the mediating mechanisms of family hardship on child developmetFamily investment model: Income is associated with child development because it enables families to invest in, purchase , or provide goods and services that are beneficial to the child’s development. Investment in the child’s human capital, i.e. education Family stess model suggests that family hardship influences child development because poverty decreases the quality of parents non-monetary capcity, ie their mental well-being and inteeractions with the child, the family investment model suggest that poor parents have less opportunity to purchase or produce important inputs into their children’s develoopment, usch as an enriched home learning environment, childcare settings, or high quality schooling. Both models explain the mediating mechanisms of family hardship on child developmet

34. Summary The experience of socio-economic hardship: is associated with cognitive and behavioural development can exacerbate maternal distress may undermine effective parenting Important to disentangle emotional and cognitive components of parenting and adjustment Important to assess specific risk effects Important to test for mediating processes Hardship exacerbate irritability and conflict – both men and women are affected in similar ways Maternal influences on child adjustment appear to be stronger than father’s.. The results underline the importance of early childhood as an important and sensitive phase in child development. Already at 9 months we can observe social inequalities in child health. The impact on socio-economic disadvantage is especially strong for markers of individual differences in behaviour and responsiveness, i.e. early child temperament. In comparison to their more privileged peers infants from low income families are twice as likely at risk of developing a difficult temperament in the first year of life. It may be that the negative effects from growing up in poor economic conditions during infancy are carried forward into the future and influence consequent adjustment patterns. Thus, the present study underscores the importance of investigating the potential association of family income and early child outcomes, as well as their long-term consequences.   Income may be a proxy for a range of problems created by disadvantage, such as poorer prenatial care, poorer medical services, greater social stressors that could impact on child development, generating social pathways to children’s longterm behaviour adjustment as well as cognitive functioning.   The path model tested in the present study indicates direct as well as indirect associations of income and child temperament. The indirect association demonstrated the importance of the family environment, as both maternal emotional distress and beliefs about parenting practices mediated the association between income and child temperament. Mothers with lower income may experience more emotional distress and are more likely to develop a laissez faire, uninvolved parenting style (not talking to, stimulating, or cuddeling their baby, or providing a regular feeding or sleeping routine), which in turn may contribute to more adjustment problems of their children. Associations of maternal distress and parenting practices were significant even after controlling for key maternal characteristics such as age at first birth, total number of siblings and level of education.   These differential findings suggest that it is vital to disentangle the emotional and cognitive components of parenting. The observed effects cannot easily be explained away by maternal characteristics, such as education level or age, although some background characteristics, especially mother’s education made independent contribution to parenting practices in particular. The effects we found for parenting and maternal distress were, however, over and above the effect of the control variables. Thus our data suggests that given appropriate parental support, addressing both emotional and cognitive aspects of parenting, poor children might experience better developmental outcomes. Hardship exacerbate irritability and conflict – both men and women are affected in similar ways Maternal influences on child adjustment appear to be stronger than father’s.. The results underline the importance of early childhood as an important and sensitive phase in child development. Already at 9 months we can observe social inequalities in child health. The impact on socio-economic disadvantage is especially strong for markers of individual differences in behaviour and responsiveness, i.e. early child temperament. In comparison to their more privileged peers infants from low income families are twice as likely at risk of developing a difficult temperament in the first year of life. It may be that the negative effects from growing up in poor economic conditions during infancy are carried forward into the future and influence consequent adjustment patterns. Thus, the present study underscores the importance of investigating the potential association of family income and early child outcomes, as well as their long-term consequences.   Income may be a proxy for a range of problems created by disadvantage, such as poorer prenatial care, poorer medical services, greater social stressors that could impact on child development, generating social pathways to children’s longterm behaviour adjustment as well as cognitive functioning.   The path model tested in the present study indicates direct as well as indirect associations of income and child temperament. The indirect association demonstrated the importance of the family environment, as both maternal emotional distress and beliefs about parenting practices mediated the association between income and child temperament. Mothers with lower income may experience more emotional distress and are more likely to develop a laissez faire, uninvolved parenting style (not talking to, stimulating, or cuddeling their baby, or providing a regular feeding or sleeping routine), which in turn may contribute to more adjustment problems of their children. Associations of maternal distress and parenting practices were significant even after controlling for key maternal characteristics such as age at first birth, total number of siblings and level of education.   These differential findings suggest that it is vital to disentangle the emotional and cognitive components of parenting. The observed effects cannot easily be explained away by maternal characteristics, such as education level or age, although some background characteristics, especially mother’s education made independent contribution to parenting practices in particular. The effects we found for parenting and maternal distress were, however, over and above the effect of the control variables. Thus our data suggests that given appropriate parental support, addressing both emotional and cognitive aspects of parenting, poor children might experience better developmental outcomes.

35. Protective processes Reduction of risk impact or sensitivity to risk Breaking negative chain reactions or increasing positive chain reactions Promoting positive experiences leading to greater self esteem and self-efficacy Opening up new opportunities Supporting planful competence and orientation to the future (Rutter, 1990) The findings presented here suggest that the experience of a stable and supportive relationship with one’s parents can reduce the impact of socio-economic adversity and consequently reduce negative chain reactions by influencing the level of early academic adjustment as well as planning for the future. The experience of successful coping with challenges might reduce the sensitivity to later risk. Positive response from significant others, such as a teacher, can increase the chance of positive chain reactions or open up positive opportunities. Affirmative school experiences can promote feelings of self-esteem and self-efficacy, which in turn might stimulate the individual to persevere in difficult circumstances, and to maintain a positive outlook on life. It has, however, to be noted that the evidence in support of these suggestions is still jagged and more research of the interactive processes that foster and promote positive adjustment in the face of adversity is needed. The role of these factors and processes varied with the context and the developmental stage under investigation, highlighting the importance of considering the context dependency of resilience and the changing needs at different developmental stages and transition periods when designing intervention programs. There are content- and context specific variations in response to adversity, and young people who demonstrate resilience in the face of one type of risk may not be able to overcome other types of risk. Some children may be resilient against the negative effects of poverty because they have supportive families, but he same children may be less successful in overcoming the effects of attending underfunded schools or violent neighbourhoods. It is also the case that different resource factors may be associated with different risks and outcomes, or with the same risk and outcome at different life stages. It is thus difficult to identify universal protective factors, raising concerns regarding any asset lists postulating similar effects for all groups of individuals, in all contexts, and for all outcomes (Fergus & Zimmerman, 2005). Findings from one context or population may not apply in another context or population. – and assets and resources are not equally distributedThe findings presented here suggest that the experience of a stable and supportive relationship with one’s parents can reduce the impact of socio-economic adversity and consequently reduce negative chain reactions by influencing the level of early academic adjustment as well as planning for the future. The experience of successful coping with challenges might reduce the sensitivity to later risk. Positive response from significant others, such as a teacher, can increase the chance of positive chain reactions or open up positive opportunities. Affirmative school experiences can promote feelings of self-esteem and self-efficacy, which in turn might stimulate the individual to persevere in difficult circumstances, and to maintain a positive outlook on life. It has, however, to be noted that the evidence in support of these suggestions is still jagged and more research of the interactive processes that foster and promote positive adjustment in the face of adversity is needed. The role of these factors and processes varied with the context and the developmental stage under investigation, highlighting the importance of considering the context dependency of resilience and the changing needs at different developmental stages and transition periods when designing intervention programs. There are content- and context specific variations in response to adversity, and young people who demonstrate resilience in the face of one type of risk may not be able to overcome other types of risk. Some children may be resilient against the negative effects of poverty because they have supportive families, but he same children may be less successful in overcoming the effects of attending underfunded schools or violent neighbourhoods. It is also the case that different resource factors may be associated with different risks and outcomes, or with the same risk and outcome at different life stages. It is thus difficult to identify universal protective factors, raising concerns regarding any asset lists postulating similar effects for all groups of individuals, in all contexts, and for all outcomes (Fergus & Zimmerman, 2005). Findings from one context or population may not apply in another context or population. – and assets and resources are not equally distributed

36. What can be done to improve life chances for children and young people Improve overall distribution of resources and opportunities Support children and their parents Invest in education – improve standards for all Recognise diversity in adjustment Holistic approach Developmental perspective Sustainability of programs Never too early, never to late for interventions It is unacceptable that children and young people being born into less advantaged families have systematically worse life chances than their more affluent peers. A commitment to improve the life chances of young people should be the priority of every developed nation aiming to harness and foster the potential of their citizens. Any researcher interested in positive development should avoid to argue in favour of individual explanations over structural or social causes. Both factors are important and developmental outcomes are based on the active interactions between individual and context. Social, cultural, and environmental factors shape who we are and how well we function. The aim must be to improve the overall distribution of resources and opportunities, leading to more positive experiences and outcomes for all. Childhood origins shape adult destinations, and there remain marked continuities in socio-economic circumstances both across generations and across the life course. Inequality and disadvantage are passed down generations. Children who grew up in poverty continue to experience poverty as they grow up, and when they become parents themselves. Recognising that uneven life chances start at birth, and that the ways in which individuals respond to adverse situations are shaped by their earlier experiences in life, it is vital that inequalities are addressed early in life, before children start school, or even earlier, before they are born. The key to giving young people a good start to life is to help their parents. The living standards of the worst off should be brought closer to the average, bringing with it greater equality of economic and social conditions and a fairer distribution of resources. Advancing the life chances of young people would also imply investment in education, especially in pre-school education, aiming to improve the standard of education for all. There should be appropriate support at key transition periods, notably during entry into school, the transition from primary to secondary education, as well as regarding further education, training and employment. Interventions in the early years are crucial for reducing inequalities, yet individuals require support at all life stages to alleviate the inter-linked effects of poverty. The findings underscore the importance of assessing developmental opportunities in socio-demographically diverse samples and in different socio-historical contexts. Life chances and opportunities are associated with social transformations. Conceptualising resilience in purely individualistic terms is misleading, as this would neglect the continuing importance of social structures and fail to address issues of social stratification and unequal opportunities. Children born into families in different socio-economic positions have widely differing prospects regarding opportunities for education and employment. Aiming to improve life chances of the most disadvantaged one would have to recognise the diversity of experiences, as well as the importance of social ties, or linked lives Identifying and building upon the strengths and ambitions of individuals and the families they are born into can promote their own feelings of competence and capability and can stimulate enduring positive changes. The identification of ‘strengths’ has however to be congruent with the experiences of marginalized populations. Adult outcomes are often determined not only according to individual capabilities and effort, but by factors beyond individual control. Persisting inequalities call for sustained investment and reform in tackling poverty and inequality, challenging the very existence of an exclusive and divided society. It is unacceptable that children and young people being born into less advantaged families have systematically worse life chances than their more affluent peers. A commitment to improve the life chances of young people should be the priority of every developed nation aiming to harness and foster the potential of their citizens. Any researcher interested in positive development should avoid to argue in favour of individual explanations over structural or social causes. Both factors are important and developmental outcomes are based on the active interactions between individual and context. Social, cultural, and environmental factors shape who we are and how well we function. The aim must be to improve the overall distribution of resources and opportunities, leading to more positive experiences and outcomes for all. Childhood origins shape adult destinations, and there remain marked continuities in socio-economic circumstances both across generations and across the life course. Inequality and disadvantage are passed down generations. Children who grew up in poverty continue to experience poverty as they grow up, and when they become parents themselves. Recognising that uneven life chances start at birth, and that the ways in which individuals respond to adverse situations are shaped by their earlier experiences in life, it is vital that inequalities are addressed early in life, before children start school, or even earlier, before they are born. The key to giving young people a good start to life is to help their parents. The living standards of the worst off should be brought closer to the average, bringing with it greater equality of economic and social conditions and a fairer distribution of resources. Advancing the life chances of young people would also imply investment in education, especially in pre-school education, aiming to improve the standard of education for all. There should be appropriate support at key transition periods, notably during entry into school, the transition from primary to secondary education, as well as regarding further education, training and employment. Interventions in the early years are crucial for reducing inequalities, yet individuals require support at all life stages to alleviate the inter-linked effects of poverty. The findings underscore the importance of assessing developmental opportunities in socio-demographically diverse samples and in different socio-historical contexts. Life chances and opportunities are associated with social transformations. Conceptualising resilience in purely individualistic terms is misleading, as this would neglect the continuing importance of social structures and fail to address issues of social stratification and unequal opportunities. Children born into families in different socio-economic positions have widely differing prospects regarding opportunities for education and employment. Aiming to improve life chances of the most disadvantaged one would have to recognise the diversity of experiences, as well as the importance of social ties, or linked lives Identifying and building upon the strengths and ambitions of individuals and the families they are born into can promote their own feelings of competence and capability and can stimulate enduring positive changes. The identification of ‘strengths’ has however to be congruent with the experiences of marginalized populations. Adult outcomes are often determined not only according to individual capabilities and effort, but by factors beyond individual control. Persisting inequalities call for sustained investment and reform in tackling poverty and inequality, challenging the very existence of an exclusive and divided society.

37. Thank you I.Schoon@city.ac.uk