Upload
rachel-egan
View
217
Download
1
Tags:
Embed Size (px)
Citation preview
1
Transmission of Socio-economic Inequalities
Paul Gregg, Carol Propper and Liz Waskbrook
ALSPAC User Group
2
Transmission of Socio-economic Inequalities
• Growing evidence that children of poorer families suffer penalties not just in terms of schooling but also health and behaviour (e.g. Propper and Rigg CP105, CP 125).
• How does low income translate into these poorer outcomes?
• Do the mediating influences differ across outcomes?
3
Overview
• Aim: to compare and contrast the importance of different risk factors including income in accounting for the gaps between low and higher income children
• Examine range of different middle childhood outcomes– IQ, School performance, locus of control, self esteem, behavioral
problems, risk of obesity (age 9)
• Use ALSPAC cohort (c 9000 children born in 1991/2 in Avon)
4
Research questions
• Which aspects of the environments of low income children matter for each outcome? [Proximal factors]
• Which types of low income family do worst in fostering each outcome? [Distal factors]
• Are the adverse environments that matter related to particular family characteristics?
• Estimate a model which decomposes the association of income with these factors on the outcomes
• Unified model across range of outcomes and large set of distal and proximal factors
5
Modelling framework
βα
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Distal factors or family
characteristics (C)
(e.g. household composition,
parental education)
γ
θ
λπ
6
Proximal factorsFactors that capture the environment, or lived experience, of the
child and that vary with family income:– Parental psychological functioning: Anxiety/ depression,
weighted life events, financial difficulties, parental relationship, frequency of smacking, social networks, locus of control
– Preschool childcare: Type and intensity, between birth and age 3, between age 3 and school entry
– Health at birth and health behaviours: Birth weight and gestation, parental smoking, breastfeeding, diet at age 3
– Home learning environment: Books and toys, maternal teaching, educational outings, mother’s and partner’s reading and singing with child
– Physical home environment: Car ownership, garden, noise, crowding, damp/mould
– School peer quality: Fixed effects
7
Family characteristicsFactors associated with income that independently
influence the proximate environment of the child, but that do not impact on children directly (distal factors):
• Family structure and life cycle: Single parenthood, siblings, mother’s age
• Parental labour market status: Mother’s and partner’s employment and occupational class
• Family education: Mother’s, partner’s and maternal grandparents’ qualifications
• Local environment: Local deprivation, social (public) housing
8
Decomposition I
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
9
Decomposition II
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
10
Decomposition III
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
11
The income gradient
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
δ
Oij = cons + δj Yi + eij
for the jth outcome of the ith child
12
The income gradient
• Low income children are cognitively, emotionally and physically disadvantaged compared with their better off counterparts
• Gradients steeper for cognitive than other outcomes
• Childs Locos of Control has an intermediate gradient
13
Outcomes standardized to mean 100, SD 10. Coefficients on adverse outcomes reversed, such that higher scores = more favourable outcomes.
p < 0.01 for all gradients
14
Decomposition I
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
15
Table 1: Income, proximal factors and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
16
Table 2: Income, health behaviors and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
17
Table 3: Income, preschool childcare and child outcomes
IQ KS1 Locus of control Self esteem Behavior Fat mass
Pre-school childcare 2.7 1.3 13.4 *** -9.2 -14.0 14.8 (Total) Birth to age 3 2.2 1.6 11.3 ** 0.7 2.2 20.5 ** (Sub-total) Age 3 and 4 0.5 -0.4 2.1 -8.4 -18.8 ** -6.0 (Sub-total) Partner or relative (3 & 4) 0.0 0.0 -0.2 0.1 0.4 0.3 Playgroup (3 & 4) 0.3 0.4 -0.2 0.7 -0.5 0.0 Childminder (3 & 4) 0.5 0.0 -0.7 -9.5 ** 2.7 -3.0 Nanny (3 & 4) -0.4 -0.3 1.5 -2.1 -7.9 * -2.6 Nursery (3 & 4) -0.2 -0.5 1.5 3.9 -11.7 ** -0.6 Other (3 & 4) 0.3 0.1 0.1 -1.6 * -1.7 0.2
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
18
Table 4: Income, the physical home environment and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
19
Decomposition II
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
20
Income, family education and child outcomes
21
Decomposition III
Proximal factors(P)
(e.g. parental psychological functioning,
health behaviors, home learning environment,
school choice)
Child outcome at age 7, 8 or
9(O)
(Ln) family income at age 3 & 4
(Y)
Family characteristics
(C) (e.g. household composition,
parental education)
22
Findings I
• Multiple aspects of the environments in which low income children are raised are associated differentially with poorer outcomes – no ‘magic bullet’
• Lack of income is only one of a number of disadvantages faced by poor children. The association of low education of parents with outcomes for low income children is particularly important
• Observed proximal factors are more strongly associated with the income gaps in socio-emotional and health outcomes than in cognitive outcomes
23
Findings II
• The role of non-home environments such as child care and schools in generating inequality is very minor compared with the family environment
• The psychological stresses associated with raising a child on a low income are important in generating greater behavioral problems among poor children, but are also associated with a higher risk of obesity and poorer scholastic outcomes
• Poorer health behaviors (smoking, diet fed to children) among low income parents are an important pathway through which poorer health, behaviour and scholastic outcomes are transmitted.
• Determinants of IQ and school performance are not the same: IQ deficits are explained more by parental education (proxying inherited ability?), while local neighbourhood matters more for academic outcomes (peer effects?)
24
Findings III
• Not everything about higher income lifestyles is beneficial for children.
• Full time maternal employment/early child care are associated with poorer behavioural outcomes, car ownership and educationally-oriented home environments are associated with greater risk of obesity
25
Additional slides
26
Table 5: Income, family characteristics and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
27
Table 6: Income, family structure and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
28
Table 7: The residual income gradient, proximal factors and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
29
Table 8: Income, family education, proximal factors and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
30
Table 9: Income, the local environment, proximal factors and child outcomes
Numbers are % of total income gradient. Stars relate to test of significance of underlying coefficient.
31
Appendix
32
Income, maternal psychological functioning and child outcomes
33
Income, the home learning environment and child outcomes
34
Income, parental labour market status and child outcomes
35
Income, family education and child outcomes
36
Income, local environment and child outcomes
37
Income, family structure, proximal factors and child outcomes
38
Income, parental labour market status, proximal factors and child outcomes
39
Table A1: Items from the Nowicki-Strickland Internal-External Scale (Locus of Control) 1. Do you feel that wishing can make good things happen? 2. Are people nice to you no matter what you do? 3. Do you usually do badly in your school work even when you try hard? 4. When a friend is angry with you is it hard to make that friend like you again? 5. Are you surprised when your teacher praises you for your work? 6. When bad things happen to you is it usually someone else's fault? 7. Is doing well in your class-work just a matter of 'luck' for you? 8. Are you often blamed for things that just aren't your fault? 9. When you get into an argument or fight is it usually the other person's fault? 10. Do you think that preparing for tests is a waste of time? 11. When nice things happen to you is it usually because of 'luck'? 12. Does planning ahead make good things happen? Items require yes/no answers and are coded such that 1 indicates externality and 0 indicates internality. Reponses are summed to derive the final locus of control scale.
40
Table A2: Items from Harter’s Self Perception Profile for Children (Self esteem) No. Statement 1 Statement 2 1 Some children feel that they are very
good at their school work Other children worry about whether they can do the school work that they have been given
2 Some children are often unhappy with themselves
Other children are pretty pleased with themselves
3 Some children feel like they are just as clever as other children their age
Other children aren’t so sure and wonder if they are as clever
4 Some children don’t like the way they are living their life
Other children do like the way they are living their life
5 Some children are pretty slow in finishing their school work
Other children can do their school work quickly
6 Some children are happy with themselves as a person
Other children are often not happy with themselves as a person
7 Some children often forget what they learn
Other children can remember things easily
8 Some children like the kind of person they are
Other children often wish they were someone else
9 Some children do very well at their classwork
Other children don’t do very well at their classwork
10 Some children are very happy being the way they are
Other children wish they were different
11 Some children have trouble working out the answers in school
Other children can almost always work out the answer
12 Some children are not very happy with the way they do a lot of things
Other children think the way they do things is fine