The independence and determinants of outcomes: The relevance for Policy Bilal Nasim, Researcher CMPO 9 th February 2010 DCSF 2010 Research Conference
Dec 31, 2015
The independence and determinants of outcomes: The relevance for Policy
Bilal Nasim, Researcher CMPO
9th February 2010
DCSF 2010 Research Conference
2
We still lag behind other countries in terms of labour productivity…
International comparisons of output per hour worked (UK=100)
Productivity is the main determinant of national living
standards. It refers to how well an economy uses the
resources it has.
The UK has relatively poor productivity compared to other major economies, although we have made progress in recent years.
ONS: International Comparisons of Productivity OMahoney and de Boer (2002) Britain’s Relative Productivity Performance: Updates to 1999
Up to one fifth of the UK's productivity gap with France and Germany is a result of the lack of skills of
workers in the UK.
Canada
France
Germany
Italy
Japan
UK
USA
G7 exc. UK
80
85
90
95
100
105
110
115
120
125
130
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
3
0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
0.400
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
-94
1994
-95
1995
-96
1996
-97
1997
-98
1998
-99
1999
-00
2000
-01
2001
-02
2002
-03
2003
-04
2004
-05
2005
-06
2006
-07
2007
-08
Gin
i
…income inequality is at its highest since records began …
The higher the gini coefficient the greater the
income inequality
Gini coefficient, UK 1970-2008• Income inequality rose
significantly over the past 30 years, particularly during the 1980s
• Levels of inequality have stabilised since the late 1990s, though have started to tick upwards in the past few years
• Inequality is now statistically significantly higher than the level of inequality observed in 1997
• Inequality in the UK is above the OECD average, but below that of the USA and Mexico
Sources: Households Below Average income / IFS (2009) Poverty and inequality in the UK 2009
4
…social mobility remains an issue…
38%31%
1958 Cohort 1970 Cohort
42%35%
1958 Cohort 1970 Cohort
Percentage of sons in lowest income quartile whose parents
are also in lowest quartile
Percentage of sons in highest income quartile whose parents
are also in highest quartile
Income mobility fell between cohorts born in 1958 and 1970 (and was high by international standards)…
…but recent evidence suggests this trend may be reversing
Relationship between family income and GCSE attainment
Sources: Blanden, Gregg and Machin (2005), Intergenerational Mobility in Europe and North America, Sutton Trust, LSEGregg and Macmillan (2008) Intergenerational Mobility and Education in the Next Generation, mimeo
0
0.2
0.4
0.6
0.8
1
1.2
1955 1960 1965 1970 1975 1980 1985 1990
bi r th c ohor ts
1 st
adar
d d
evia
tio
n c
han
ge
inco
me
wrt
to
GC
SE
sco
res
5
-30
-25
-20
-15
-10
-5
0
Material well-being
Housing andenvironment
Educationalwell-being
Health andsafety
Riskybehaviours
Quality ofschool life
ran
k
…and there are concerns about children’s overall well-being…
Comparative policy-focused child well-being in 30 OECD countries
12th 15th 22nd 20th 28th
4th
No. of countries in each ranking
Source: OECD (2009) Doing Better for Children
6
Progress is being made in a number of areas…
Attainment on most measures continues to rise……and fewer schools are missing floor targets
The recession is helping to push up participation in education and training
Attainment at age 19 is improving, whilst at the same time the FSM gap is slightly narrowing
Fewer looked after children are moved frequently, and more are in stable placements… …and more care-leavers are having positive outcomes – both in terms of
accommodation and employment, education or training
Breastfeeding is on a general upward trend
Infant mortality continues to fall and the rate of hospital admissions has fallen from its 2006-07 peak
Substance misuse and abuse amongst young people is on a general downward trend
The number of 1st time entrants to the criminal justice system has turned a corner……and rates of reoffending are down
Enjoy and
Achieve
Achieve Economic Well-being
Stay Safe / Economic Well-being
Be Healthy
Be Healthy / Stay Safe
Make a positive
contribution
Achieve Economic Well-being
Be Healthy
7
…but there are significant challenges in other areas:
Attainment gaps by FSM and SEN continue to persist, though there are signs of progress
Numbers of young people NEET are rising
…and whilst we’re making improvements in the attainment of looked after children, the 2011 targets will be challenging
We are unlikely to meet our 2010 Child Poverty target
Reductions in conception rates of teenage girls since 1998have only been gradual
Enjoy and achieve
Achieve Economic Well-being
Enjoy and achieve
Achieve Economic Well-being
Be Healthy
8
Outline
Introduction: Scene-setting
A. The interdependence and determinants of outcomes of childhood
B. The role of Schools and Pre-schools
C. Inter- and Intra-generational associations
9
The ‘outcomes’ of childhood are not mutually exclusive goods. The latest academic research has enabled us to understand these links better
Individual
Family
Institutional
Social
Early outcomes
Adult outcomes
Intergenerational impacts
Enjoy and Achieve
Make a positive
contributionBe healthy
Achieve economic well-being
Stay Safe
ECM Outcomes Later outcomesDeterminants of
outcomes
?
?
??
? ?
?
?
??
?
?
??
??
10
A number of ECM outcomes are associated with achievement, though prior attainment still dominates
10
Achieve
Make a positive
contribution
Be healthy
Self-rated health and extracurricular activities are positively and significantly associated with academic achievement.
There is a very strong, significant – and causal – negative relationship between previous unauthorised absence levels and academic achievement.
Enjoy and achieve
(attendance)
Stay Safe(bullying)
Determinants of achievement at age 16
Source: Vignoles and Meschi, Forthcoming, “The determinants of non-cognitive and cognitive schooling outcomes”.
Achieve economic well-being
-10% 0% 10% 20% 30% 40% 50% 60%
Key Stage 2
Truancy
Age 14 SchoolEnjoyment
Bullied
Extra-curricularactivities
Health
effect size (% of standard deviation)
11
Happy children tend to be more successful, enjoy school and participate more
11
AchieveEnjoy
Stay Safe(bullying)
Pupils with high life satisfaction report significantly more positive school experiences, a greater frequency of extracurricular activities
participation and higher academic achievement…
…in contrast, being a victim of bullying is significantly related to worse levels of well-
being, behaviour and academic achievement, though the reverse is not true
Make a positive
contribution
Source: Gilman and Huebner, 2006, “Characteristics of adolescents who report very high life satisfaction”Gutman and Feinstein, 2008, “Children’s Well-Being in Primary School: Pupil and School Effects”
12
-40% -20% 0% 20% 40% 60% 80% 100% 120% 140%
Key Stage 2
Truancy
Age 14 SchoolEnjoyment
Bullied
Extra-curricularactivities
Health
% standard deviation effect
Children’s health is a key determinant of school enjoyment…
12
Enjoy
Stay Safe(bullying)
Be healthy
Bullied children at 14 are significantly less likely to enjoy school at 16 while the opposite is seen to be true for pupils with higher self-rated health.
Perhaps unsurprisingly, truanting behaviour is negatively associated with school enjoyment, though causality is likely to run in both directions
Enjoy and achieve
(attendance)
Determinants of school enjoyment at age 16
Source: Vignoles and Meschi, forthcoming, “The determinants of non-cognitive and cognitive schooling outcomes”.
13
…and is strongly associated with the likelihood of being bullied, though the direction of causation is yet to be established…
13
Stay Safe(bullying)
Be healthy
Pupils experiencing bullying at 14 are more likely to experience a greater degree of bullying at 16.
Pupils reporting health problems at 14 are significantly more likely to report being bullied at 16...
…but academic achievement (Key Stage 2) and attitude toward school (age 14) appear to be not associated with being bullied at 16.
Source: Vignoles and Meschi, forthcoming, “The determinants of non-cognitive and cognitive schooling outcomes”.
14
…and ceasing to enjoy school is associated with a range ofnegative outcomes
Children that enjoy school perform better at KS4, even when accounting for prior attainment and are significantly less likely to engage in risky and anti-social behaviour
children who are bullied perform worse than children who are not bullied and are more likely to experience behavioural problems…
…but are no more likely to truant…
Impact of school enjoyment on outcomes
Source: IFS (2009) Drivers and Barriers to Educational Success: Evidence from LSYPE
-8%
-6%
-4%
-2%
0%
2%
4%
6%
8%
10%
Key Stage 4 Smokefrequently
Drinkfrequently
TriedCannabis
Anti-socialbehaviour
Truancy
Find School worthwhile (Age 14) Stop liking school (Age 14 - 16)
Make a positive
contribution
AchieveEnjoy
Stay Safe(bullying)
15
The relative importance of individual, family, and social factors in determining child outcomes
Individual
Family
Institutional
Social
Early outcomes
Adult outcomes
Intergenerational impacts
Enjoy and Achieve
Make a positive
contributionBe healthy
Achieve economic well-being
Stay Safe
ECM Outcomes Later outcomesDeterminants of
outcomes
Source: Adapted from Crawford et al (2009) Drivers and Barriers to Educational Success Evidence from the Longitudinal Study of Young People in England DCSF RR 102
16
Family income is associated with a wide range of cognitive and non-cognitive outcomes...
Family
Paul Gregg, Carol Propper and Elizabeth Washbrook (2008) Understanding the Relationship between Parental Income and Multiple Child Outcomes: a decomposition analysis CMPO Working Paper No. 08/193
The steeper the curve, the stronger the relationship between income and the
outcome in questionSignificant Income gradients are observed in mid-childhood outcomes
These income gradients appear to be largest for cognitive outcomes…
….and smallest for obesity, self esteem and behaviour
17
0%
1%
2%
3%
4%
5%
6%
7%
poorest 2 3 4 richest
quintiles of socio-economic position
per
cen
t
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
poorest 2 3 4 richest
quintiles of socio-economic position
per
cen
t
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
poorest 2 3 4 richest
quintiles of socio-economic position
…and well as for a number of “behaviours”, though not always in the expected direction
17
Frequent smoker, age 14Participates in positive activities
Frequent drinker, age 14Anti-social behaviour
Source: Crawford et al (2009) Drivers and Barriers to Educational Success Evidence from the Longitudinal Study of Young People in England DCSF RR 102
Family
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
poorest 2 3 4 richest
quintiles of socio-economic position
per
cen
t
1818
No. These gradients can’t be considered causal.
Much – but not all – of the observed income gradients in childhood outcomes are transmitted via parental
characteristics and the home environment – which is partly a result of their parents’ socio-economic background.
FamilyDoes this mean income is all that matters?
However children living in poverty experience restricted access to many factors which might otherwise provide ‘protection’ against the
negative impact of low-income, including: friendships; time with family; good relationships with teachers; ability to concentrate on school work; health; stability of neighbourhood and school; and
membership of social networks.
19
Family income does seem to have a direct causal impact on childhood outcomes, but the effects are fairly modest…
• UK evidence suggest that a one-third reduction in family income increases the propensity to achieve no A-C GCSEs by between 1 and 3 percentage points…
• Canadian evidence indicates that a $1000 increase in child benefits causes approximately 7% of a standard deviation increases in Maths and PPVT scores… – …and 3% and 5% of a standard deviation reductions in childhood
hyperactivity and physical aggression respectively.
• …and American research has shown a $1000 increase in family income raises Maths and Reading scores by 6% of a standard deviation.
Source: Stabile and Milligan, 2008 “Do child benefits affect the well-being of children? Evidence from Canadian child benefit expansions”, Gregg and Blanden, 2004 “Family Income and Educational Attainment: A Review of Approaches and Evidence for Britain”, Dahl and Lochner, 2008 “The impact of Family Income on child achievement”
20
0%
10%
20%
30%
40%
50%
60%
FamilyStructure and
lifecycle
Parental labourmarket status
ParentalEducation
LocalNeighbourhood
% e
xpla
inin
g t
he
inco
me
gra
die
nt
0%
5%
10%
15%
20%
25%
30%
35%
40%
Family Structureand lifecycle
Parental labourmarket status
ParentalEducation
LocalNeighbourhood
% e
xpla
inin
g t
he
inco
me
gra
die
nt
Total explained by socio-economic charateristics
Residual income effect
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
KS1 scores IQ Self Esteem Locus
% o
f th
e i
nc
om
e g
rad
ien
t
Parental education explains a good deal of the association between income and outcomes…
Family
KS1 further decompositionLocus of control further
decomposition
% of income gradient explained by socio-economic characteristics
Paul Gregg, Carol Propper and Elizabeth Washbrook (2008) Understanding the Relationship between Parental Income and Multiple Child Outcomes: a decomposition analysis CMPO Working Paper No. 08/193
21
0%
5%
10%
15%
20%
25%
30%
Family Structureand lifecycle
Parental labourmarket status
ParentalEducation
LocalNeighbourhood
% e
xpla
inin
g t
he
inco
me
gra
die
nt
-20%
-10%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Family Structureand lifecycle
Parental labourmarket status
ParentalEducation
LocalNeighbourhood
% e
xp
lain
ing
th
e in
co
me
gra
die
nt
Total explained by socio-economic charateristics
Residual income effect
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Fat mass Behaviour
% o
f th
e i
nc
om
e g
rad
ien
t
…and parental education also accounts for three-quarters of the association between income and fat mass…
Family
% of income gradient explained by socio-economic characteristics
Fat mass further decomposition
Behaviour further decomposition
Paul Gregg, Carol Propper and Elizabeth Washbrook (2008) Understanding the Relationship between Parental Income and Multiple Child Outcomes: a decomposition analysis CMPO Working Paper No. 08/193
22
0%
10%
20%
30%
40%
50%
60%
70%
IQ Key Stage 1 Locus ofcontrol
Self-esteem Behaviour Fat mass
Differences in the environments and behaviours of children from low and high income households predict a substantial proportion of the income gradients…
Mothers’ psychological functioning and access to social networks are particularly important for behavioural problems.
Health-related behaviours of low income parents are as important for cognitive outcomes and child mental and physical health.
If smoking rates among low income mothers were to fall to the rates of their higher income counterparts, the income gradients for fat mass and child behaviour would fall by one fifth
Percentage of the income gradient explained by differences in environments and behaviours of low
and high income households
Family
Parental characteristics account for one third of the cognitive income gradients, but over 50% of the shallower mental and physical
health outcome gradients
Paul Gregg, Carol Propper and Elizabeth Washbrook (2008) Understanding the Relationship between Parental Income and Multiple Child Outcomes: a decomposition analysis CMPO Working Paper No. 08/193
23
…but there also may be adverse consequences of higher-income lifestyles…
23
Maternal employment
Physical Activity
Long hours of maternal employment in pre-school period are associated with lower cognitive, non-cognitive and behavioural outcomes in children
Learning-focused activities and behaviours, car ownership and the temperature of the home in the pre- school period are all associated with greater fat mass in children at age 9
Family
Paul Gregg, Carol Propper and Elizabeth Washbrook (2008) Understanding the Relationship between Parental Income and Multiple Child Outcomes: a decomposition analysis CMPO Working Paper No. 08/193
24
…and we are only now starting to understand what behaviours and characteristics to be important in explaining parenting style.
Quality of interaction at
age 1
Educational communication
at age 1
Quality of interaction at
age 5**
Educational communication
at age 5**
x xBreastfeeding
x x xSocial Networks
x xMaternal Education
x Maternal Mental Health
Source: Adapted from Gutman et al (2009) Nurturing Parenting Capability: the early years
Beh
avio
urs
‘Cha
ract
eris
tics’
Predictors of parenting quality* at age 1 and 5
**Includes control for age 1 measure so the age 5 effects are additional effects
* Measure using the Thorpe Interaction Measure which involves a mother and child sharing a picture book at ages 1 and 5
x x xMarital Status
x x x Number of siblings
Family
25
0%
5%
10%
15%
20%
25%
30%
Key Stage 4 Smokefrequently
Drinkfrequently
Tried Cannabis Anti-socialbehaviour
Belief in ow n ability (scale) (Age 14) Lost belief in ow n ability (Age 14 - 16)
Individual child attitudes are critical. Levels of self-belief are related to attainment, whereas changes matter for engagement in risky behaviours.
25
Individual
There strong associations between children’s beliefs regarding their own ability and their academic attainment….
…but losing self-belief is also associated with increased likelihood in engagement in risky behaviours.
There are also strong associations between whether a child believes they have control over their own economic destiny and their academic attainment…
Impact of child self-belief on outcomes
Source: IFS (2009) Drivers and Barriers to Educational Success: Evidence from LSYPE
26
…and children with greater educational aspiration tend to perform better in school, and have fewer behavioural issues.
Impact of higher education aspirations on outcomes
Individual
Source: IFS (2009) Drivers and Barriers to Educational Success: Evidence from LSYPE
-40%
-30%
-20%
-10%
0%
10%
20%
30%
Key Stage 4 Smoke frequently Anti-social behaviour Truancy
Likely to apply to University (Age 14) Stopped thinking of applying to Uni (Age 14 - 16)
27
So-called “peer effects” have a small, albeit significant impact on attainment…
0
10
20
30
40
50
60
70
80
Level 1-2 Level 3 Level 4 Level 5+
Peer-group Quartile 1
Peer-group Quartile 2
Peer-group Quartile 3
Peer-group Quartile 4 pe
rcen
tile
of th
e ag
e-1
4 te
st s
core
re
lativ
e to
the
base
line
pupi
ls
Peer effect, strongest for medium & high attainers
Pupil’s prior attainmentTeacher assessment of age-11 ability
Peer groups could account for around 0.6% of the variance in pupils’ progress between the ages of 11 and 14. General differences between schools explain about 13% of the variance
Peer effects are dwarfed by the impact of pupils’ own prior attainment
Low ability pupils do not appear to benefit as much from mixing with high ability peers as intermediate and high ability pupils do
Peer group FSM status does not have a direct impact after controlling for attainment effects
Impact of peer group quality on attainment
Gibbons, S., and Telhaj, S. (2006) “Peer Effects and Pupil Attainment: Evidence from Secondary School Transition” CEE Discussion Paper 63
Social
28
…but neighbourhood characteristics in and of themselves appear to have little influence on outcomes, except NEET…
28
Social
Deprived individuals living in deprived areas are more likely to be NEET at age 17 than deprived individuals living innon-deprived areas.
However same study finds no evidence that neighbourhood deprivation consistently affects Key Stage 4 scores or any behavioural outcomes at age 16…
…though the literature is more mixed about the impact of neighbourhoods on behaviour outcomes
Impact of multiple deprivation on chances of being NEET (relative to 20% most deprived
neighbourhoods)
Source: IFS (2009) Drivers and Barriers to Educational Success: Evidence from LSYPE
-2.5%
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
2nd IMD quintile 3rd IMD quintile 4th IMD quintile Top IMD quintile
29
Policy Discussion: Section A
(i) There appear to be complementarities between ECM outcomes;
• Exploit interdependencies to improve and broaden the effect of policy.
• Proactive targeting of pupils at risk of future cognitive and non-cognitive difficulties.
• Supportive of the “Personalised Learning” approach.
30
(ii) The determinants of outcomes: Potential avenues to improve child outcomes;
Parents:
• Increasing educational capital of parents.
• Improving parental psychological health and health-related behaviours.
• Creating a better home and wider local environment for children.
31
Individual:
• Increasing the aspirations and self-belief of children
Income:
• Ensuring parents are aware of the potential adverse effects of aspects of high income life-styles.
• Continuation of the use of income transfers and benefits to disadvantaged families.
32
Outline
Introduction: Scene-setting
A. The interdependence and determinants of outcomes of childhood
B. The role of Schools and Pre-schools
C. Inter- and Intra-generational associations
33
24
24.5
25
25.5
26
26.5
27
27.5
28
28.5
CommunitySchool
VoluntaryAidedSchool
VoluntaryControlled
School
FoundationSchool
Academy K
S2
AP
S280
290
300
310
320
330
340
350
Ave
rag
e ca
pp
ed G
CS
E a
nd
eq
uiv
alen
ts p
oin
t sc
ore
p
er e
nd
of
KS
4 p
up
il
KS2 APS of Y7 intake 2003/4 Capped APS KS4 2007/08
0%
5%
10%
15%
20%
25%
Key Stage 2 English Key Stage 2 Maths Key Stage 4
Differences between schools account for between a tenth anda fifth of variation in academic attainment…
33
Institutional
Percentage of between-school variation in test scores
Around a tenth to a fifth of the variation in Key Stage 2 and Key Stage 4 is attributable to school differences.
Voluntary-Aided schools have the best GCSE results, but they also have a
higher quality intake
Sources: Vignoles and Meschi, forthcoming, “The determinants of non-cognitive and cognitive schooling outcomes”.Guttman and Feinstein (2008) “Children’s Well-being in Primary School: Pupil and School Effects”; DCSF “(2008) GCSE and Equivalent Results in England, 2007/8 and DCSF (2008) The Composition of Schools in England
Key Stage 2 and 4 attainment by school type
LAs with schools who have more ‘selective’ admissions
systems have more between-school variation
34
…but good teachers do seem to matter...
Being taught by a high-quality (75th percentile) rather than low-quality (25th percentile) teacher adds 0.425 of a GCSE point per subject (where 1 point= 1 grade).
Rivkin et al (2005) find the gap in GCSE points between a poor and non-poor student is 6.08 GCSE points.
If the poor student had good teachers for all 8 subjects and the non-poor student had poor (25th percentile teachers) for all 8, this would make up 3.4 points (56% of the difference).
Impact of teacher quality on GCSE attainment
Source: Burgess et al (2009) Do teachers matter? Measuring the variation in teacher effectiveness in England
Institutional
0
0.2
0.4
0.6
0.8
1
1.2
Difference betw een 25th and75th percentile
Difference betw een 5th and 95thpercentile
GC
SE
po
ints
per
su
bje
ct
35
…as do pupil-teacher relations…
Impact of pupil-teacher relations on age 16 outcomes
Sources: IFS (2009) Drivers and Barriers to Educational Success: Evidence from LSYPE Guttman and Feinstein (2008) “Children’s Well-being in Primary School: Pupil and School Effects”
Institutional
Good teacher-child relations are negatively associated with engagement in a wide range of risky behaviours…
Children attending schools with a high level of parental involvement experience 5% and 4% advantages in Key Stage 2 Maths and English respectively…
…while conversely, schools with a high level of parent-teacher disputes experience 5% lower Key Stage 2 scores on average.
-12%
-10%
-8%
-6%
-4%
-2%
0%
2%
4%
6%
8%
KS4 value added Frequent smoker Frequent drinker Ever triedcannabis
Played truant inthe last year
36
There also is evidence that more resources at the margincan improve outcomes, especially for disadvantaged groups…
Key Stage 2
Key Stage 3
Key Stage 4
Other key issues
An increase of £1,000 in average expenditure per pupil leads to an implies an average improvement in attainment corresponding to 4.3%
of a level in English, 3.5% in maths and 1.9% in science (OLS)
£100 more on per pupil expenditure increases Maths and Science attainment at Key Stage 3 on average by 4% of a level (IV)
£100 per annum over 5 years of additional expenditure perpupil would be associated with an improvement of about 0.3 in the
capped GCSE points score. (IV)
Spending the same amount on reducing the pupil-teacher ratio has a greater effect
Effects are of greater magnitude for pupils from more disadvantaged backgrounds, especially those with high prior attainment.
Levacic et al (2005) Estimating the Relationship Between School Resources and Pupil Attainment at Key Stage 3; Levacic et al (2006) Estimating the Relationship between School Resources and Pupil Attainment at GCSE; Holmand et al (2008) Impact of School Resources on Attainment at Key Stage 2
Institutional
37
Social class gaps open early, and continue to widen…
Source: Feinstein (2003). “Inequality in the Early Cognitive Development of British Children in the 1970 Cohort,” Economica, p73-97.
Blanden and Machin (2007) Recent Changes in Intergenerational Mobility
020406080
100
22 42 62 82 102 122Age (months)
Att
ain
men
t p
erce
nti
le
0
20
40
60
80
100
22 42 62 82 102 122Age (months)
Att
ain
men
t p
erce
nti
le
The famous “Feinstein Chart” of the 1970 cohort
demonstrates that social class gaps open early, and
continue to widen…
Attainment (percentile rank) by SES and early ability
Attainment (percentile rank) by income and early ability
Hses, Hab’ty
Lses, Hab’tyHses, Lab’ty
Lses, Lab’ty
Hinc, Hab’ty
Hinc, Lab’tyLinc, Hab’ty
Linc, Lab’ty
…but this doesn’t appear to be a historical anomaly – the Millennium Cohort Study is begging to reveal a similar
trend…
38
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0.05
Pre-reading Early Number Concepts Language
< 1 yr pre-school 1 - 2 yrs pre-school
2 - 3 yrs pre-school > 3 yrs pre-school
Pre-School has significant effect on a child’s early cognitive outcomes…
…and these effect persist into early primary school
Children who spend more time in pre-school, experience significantly higher cognitive
attainments…
Impact of pre school on cognitive ability (pre-school) – compared to home children
Impact of pre-school on primary school reading – compared to home children
EPPE: Tech Paper 8a “Measuring the Impact of Pre-School on Children’s Cognitive Progress over the Pre-School Period”
0%
10%
20%
30%
40%
50%
60%
Entry to School Year 1 Year 2
% o
f st
and
ard
dev
iati
on
< 1 yr pre-school 1 - 2 yrs pre-school
2 - 3 yrs pre-school > 3 yrs pre-school
39
…and social and behavioural development.
39
Results show that children who have spent more time in
pre-school have significantly better social behavioural development.
Impact of pre school on non-cognitive outcomes (pre school) – compared to
home children
A small time in preschool appears to be best for co-
operation and conformity…
…but only those who have spent a long time in pre-
school are significantly more anti-social / worried
EPPE: Tech Paper 8b “Measuring the Impact of Pre-School on Children’s Social/Behavioural Development over the Pre-School Period”
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
Independence andconcentration
Cooperation andconformaty
Peer sociability Anti-social / Worried
< 1 yr pre-school 1 - 2 yrs pre-school 2 - 3 yrs pre-school > 3 yrs pre-school
40
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Low Quality Medium Quality High Quality
effe
ct s
ize
English Maths
The positive effect of pre-school persists well into childhood
…and good quality pre-school can help to promote young people’s self-regulation
There is a clear pre-school quality gradient in Key Stage 2 Maths and
English.
Effect of pre-school quality on age 11 English and maths
Impact of pre-school on self regulation and pro-social behaviour age 11
EPPE (3-11): Final Report from the Primary Phase: Pre-school, School and Family Influences on Children’s Development During Key Stage 2 (Age 7-11)
0
0.05
0.1
0.15
0.2
0.25
0.3
Self-regulation Pro-social behaviour
effe
ct s
ize
Low quality Medium quality High quality
41
Policy Discussion: Section B
(i) The role of Schools
• Improving the quality of teachers and ensuring good teacher-child/parent relations
• Ensuring a good child-school “fit” and further support for the “Personalised Learning” program.
42
(ii) The role of pre-schools:
• Encouraging attendance and duration in pre-schools.
• Improving the quality of pre-schools.
43
Outline
Introduction: Scene-setting
A. The interdependence and determinants of outcomes of childhood
B. The role of Schools and Pre-schools
C. Inter- and Intra-generational associations
44
Age 11 social and cognitive skills matter for employment and adult wages
44
Impact of social and cognitive skills at age 11
on probability of employment
Impact of social and cognitive skills at age 11
on (log) wages
Pedro Carneiro , Claire Crawford , Alissa Goodman, 2007: The Impact of Early Cognitive and Non-Cognitive Skills on Later Outcomes
45
Age 11 well-being and cognitive skills also associated with fewer mental, physical and behavioural problems in adulthood…
45
Impact of age 11 social and cognitive skills on non-cognitive outcomes
Source: Carneiro, P. et al (2007) The Impact of Early Cognitive and Non-Cognitive Skills on Later Outcomes CEE Discussion Paper 92
-30%
-25%
-20%
-15%
-10%
-5%
0%
Crime Poor Health Depression Mental healthproblems
effe
ct s
ize
Well-being (Age 11) Cognitive Skills (Age 11)
46
…so it is not surprising that intergenerational income mobility has stayed stubbornly high (by international standards) over recent years…
1970
=0.291
1958
=0.205
Decomposing the change in the relationship between family income and sons’ income
1958-1970
• 3/4 of the rise in the intergenerational co-efficient can be explained, mainly by strengthening relationships between family income and:
• Access to HE;
• Attainment at 16;
• Unemployment
• Parental income had more impact on Non-cognitive skills in the 1970 cohort, but the effect of non-cognitive skills on income mainly operated through raising attainment.
Source: Blanden, Gregg and MacMillan (2006) “Explaining Intergenerational Income Persistence: Non-cognitive Skills, Ability and Education” CMPO Working Paper 06/146
where
iparentsi
sonsi YY lnln
-0.02 0 0.02 0.04 0.06 0.08 0.1
Health
Non-Cogntive 10/11
Cognitive tests 10/11
Age 16 Exams
Post 16 Education
Post 18 Education
Labour Market
Unexplained
Total
Change in intergerational persistance 1958-1970
47
…and recent evidence shows that intergenerational associations are evident for the whole range of ECM outcomes…
47
Stay safe
Bullying and discrimination
Safe from crime and ASB
Have security, stability and cared for
Intergenerational conditional correlations between parent and child outcomes
Source: Blanden et al (forthcoming) “Research on the Intergenerational Links in the Every Child Matters Outcomes” CEE
0 0.02 0.04 0.06 0.08 0.1 0.12
Bullied
Mother scared ofgoing out, child
scared
Mother's interest inchild age 10
Father's interest inchild age 10
Broken home
conditional correlation
48
0 0.05 0.1 0.15 0.2 0.25
Birth Weight
Measles
Mumps
Chicken Pox
MMR Immunisation
Internalising behaviour age 5
Externalising behaviour age 5
Externalising behaviour age 10
Externalising behaviour age 16
Father smoke age 16
Mother smoke age 16
Parent youth drinking (16) and child drinking(age13-16)
Parent adult drinking (30) and child drinking(age13-16)
conditional correlation
There are strong intergenerational associations between a number of health outcomes, for example birthweight
48
Be healthy
Physically Healthy
Mentally / Emotionally Healthy
Healthy Lifestyles
Intergenerational conditional correlations between parent and child outcomes
Source: Blanden et al (forthcoming) “Research on the Intergenerational Links in the Every Child Matters Outcomes” CEE
49
0 0.05 0.1 0.15 0.2 0.25
Vocabulary age 5
Spelling age 10
Maths age 10
Child not liked by otherchildren
conditional correlation
The link between early cognitive ability of parents and children is not as strong as later cognitive ability
49
Enjoy and
achieve
Ready for school
Achieve stretching national educational standards in
primary school
Achieve personal social development and enjoy
recreation
Intergenerational conditional correlations between parent and child outcomes
Source: Blanden et al (forthcoming) “Research on the Intergenerational Links in the Every Child Matters Outcomes” CEE
50
…and there is generally less of a link between “making a positive contribution” indicators…
50
Make a positive
contribution
Engage in decision making and support community and
environment
Develop positive relationships and choose not
to bully and discriminate
Intergenerational conditional correlations between parent and child outcomes
Source: Blanden et al (forthcoming) “Research on the Intergenerational Links in the Every Child Matters Outcomes” CEE
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
Social activity index
Bullies age 10
Bullies age 16
Parent discriminatoryview s and bullying
condtional correlation
51
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Father left school early,child w ants to leave early
Mother left school early,child w ants to leave early
Aspirations for child toleave at compulsory age
Parent expectations ofchild going on to university
Number of rooms inhousehold
Parent ow ns house
conditional correlation
…and the link between educational and economic well-being is particularly strong
51
Achieve economic well-being
Engage in further education
Live in decent homes and sustainable communities
Intergenerational conditional correlations between parent and child outcomes
Source: Blanden et al (forthcoming) “Research on the Intergenerational Links in the Every Child Matters Outcomes” CEE
…and similarly, children living in social housing are much more likely to live in social housing as adults…
52
Children born in 1970 living in social housing are over 3 times more likely to live in social housing in adulthood...
…and social housing is associated with almost every indicator of disadvantage
measurable.
Achieve economic well-being
53
Policy Discussion: Section C
Inter-generational persistence and intra-generational associations:
• Closing the gap in non-cognitive skills between rich and poor children.
• Continued action to raise attainment of poor children at 16 and beyond.
• Ensuring poor children do not experience long-term unemployment early in their careers.