Content uploaded by Steven Hope
Author content
All content in this area was uploaded by Steven Hope on Mar 11, 2015
Content may be subject to copyright.
Longitudinal and Life Course Studies 2010 Volume 1 Issue 3 pp 209 - 222
209
Family hardship and children’s development:
the early years
Ingrid Schoon*, Steven Hope#, Andy Ross#, and Kathryn Duckworth*
*Institute of Education, London
#National Centre for Social Research
I.Schoon@ioe.ac.uk
Received May 2010 Revised June 2010)
Abstract
Examining the factors and processes shaping school readiness provides important information
about how to enable young children to develop their cognitive potential and to succeed in
their school careers. The aim of this paper is to assess different mediating processes through
which family hardship affects children’s early development, both in terms of cognitive and
behavioural adjustment. Using data from the UK Millennium Cohort, we examine the
associations between persistent socio-economic hardship and young children’s development,
and investigate the role of maternal emotional distress, mother-child interactions, and
cognitive stimulation as potential mediators, in a sample of 14661 children, who were
followed from birth through age 3 years. Cognitive ability was assessed by standardized tests,
and child behaviour by maternal report, when the children were 3 years of age. The findings
suggest that persistent family hardship was significantly associated with child developmental
outcomes. The impact of hardship on cognitive and behavioural adjustment is partially
mediated by the level of maternal distress, which in turn shapes the quality of parent-child
interactions and the provision of a cognitively stimulating home environment. The findings
suggest differential pathways in the transmission of family disadvantage, where parenting
characteristics were more important in mediating the effect of hardship on behavioural
adjustment, than on early cognitive development. Findings are discussed in terms of their
policy implications.
Characteristics of children at school entry, provide
vital clues for predicting their performance during
their school careers and for later development.
Moving beyond a narrow view of school readiness
defined by measures of children’s cognitive
capacities, more holistic approaches, including
indicators of socio-emotional and behavioural
adjustment, have shown to be more useful indicators
of early functioning (Alexander 2009; Kagan 1992;
Meisels 1999), as both the possession and
implementation of skills are important. The
development of cognitive, behavioural, physical, and
socio-emotional capacities in the early years, forms
the foundation of wellbeing, learning and behaviour
across the life course and is crucial in shaping later
developmental adjustment (Duncan et al 2007;
Heckman 2006; Marmot 2010; McLoyd 1998; Rutter
1989). Previous research has shown that differences
in capabilities that exist at the beginning of school are
likely to perpetuate over time (Entwistle and
Alexander 1999; Feinstein and Bynner 2004; Schoon
2006). Indeed the early years have been identified as
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
210
a crucial window for interventions, a sensitive period
for skill formation (Heckman 2006), especially
regarding cognitive development (Sameroff, Seifer,
Baldwin and Baldwin 1993; Schuerger and Witt 1989).
Gaining a better understanding of early influences on
school readiness, is thus vitally important in enabling
young children to fully develop their potential. In the
following, we adopt a more holistic view of school
readiness, focusing on both cognitive and behavioural
outcomes, and examine the role of family hardship in
influencing the child’s development in the first three
years of life. Both academic and behavioural
adjustment are understood as markers of key
capabilities at school entry, enabling the child to meet
the demands of schooling (Janus and Duku 2007;
Lloyd and Hertzman 2009). We furthermore examine
the role of parent characteristics and parent-child
interactions as mediators impacting on the
association between hardship and child adjustment,
in order to identify potential protective mechanisms
and processes enabling children to strive against the
odds.
Socio-economic adversity and early adjustment
There is ample evidence of the association
between family hardship and children’s cognitive and
behavioural development (Bradley and Corwyn 2002;
Duncan and Brooks-Gunn 1997; Keating and
Hertzman 1999). Relative few studies, however, have
focused on early childhood (Linver, Brooks-Gunn and
Kohen 2002; Kiernan and Huerta 2008; Kiernan and
Mensah 2009; Robila and Krishnakumar 2006;
Waldfogel and Washbrook 2010), when the effects of
material hardship appear to be strongest (Brooks-
Gunn and Duncan 1997; Korenman, Miller, and
Sjaastad 1995; Plewis and Kallis 2008). There is
evidence to suggest that cognitive development in
the early years is malleable in response to
environmental conditions. For example, in a study
based on a sub-sample of the 1970 British cohort
study, Feinstein (2003) showed that differences in
cognitive development associated with income
inequalities, emerge as early as 22 months. The gap
appeared to widen as children aged, and around age
6, children in the highest achieving group, with
parents in the least privileged socio-economic group,
were overtaken by children from advantaged
backgrounds, who were in the low-achieving group at
age 22 months. Studies drawing on data collected for
the most recent UK Millennium Cohort, confirm the
corrosive effects of poverty on children’s cognitive
development, as well as their psycho-social
adjustment in early childhood (Blanden and Machin,
2010; George, Hansen, and Schoon, 2007; Kiernan
and Huerta, 2008; Kiernan and Mensah 2009;
Waldfogel and Washbrook 2010). Furthermore,
research findings based on the British Cohort Studies,
highlight that early disadvantage can have important
consequences and undermine later achievements.
The experience of family hardship in the early years,
undermines early cognitive development and
psychosocial adjustment, which in turn influences
later attainments, as shown in follow-up studies of
the 1958 and the1970 cohort (Bynner and Joshi 2002;
Bynner, Schuller and Feinstein 2003; Feinstein 2004;
Feinstein and Vignoles 2008; Schoon 2006; Schoon et
al 2002).
Beyond income
Much of the research to date on poverty effects on
child development has focused on the effects of
income (Blanden and Gregg 2004; Blanden and
Machin 2010; Waldfogel and Washbrook 2010); (see
also Gregg and Macmillan in this Special Issue). This
is not surprising, given that income poverty rates in
the UK remain high. At the turn of the Millennium
about 26 per cent of children lived in households with
an equivalised houshold income below 60 per cent of
the national median. Between 1998/9 and 2004, this
rate fell to 21 per cent, but has increased to 23 per
cent in 2009 (Hills, Sefton and Steward,2009;
MacInnes, Kenway and Parekh 2009). Given the
persistence of extreme poverty even in highly
developed countries, it is essential for developmental
scientists to learn more about the impact of poverty
and material hardship on families and children living
today.
In analysing effects of poverty on children’s
development, one should however not forget about
the families whose income is considered as ‘low
income’ just above the threshold of the poverty line.
Children in these low income families experience
many of the same hardships as children in families
defined as income poor, such as housing insecurity,
overcrowding, lack of amenities, or dependence on
state benefits to make ends meet. Consideration of
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
211
the linked contributions of family income and
material hardship has thus been recommended to
gain a better understanding of the corrosive effects of
family poverty on children’s development (Gershoff,
Aber, Raver and Lennon 2007; Plewis and Kallis 2008;
Yeung, Linver and Brooks-Gunn 2002).
Mediating processes
Although the association between family hardship
and child development is well documented, the ways
in which the experience of socio-economic hardship
influences children’s development have been less
well studied. Family interactions, neighbourhood
processes, and child-care quality have been shown to
mediate the effect of family hardship on child
development, illustrating the contextualized nature of
early child development (Brooks-Gunn and Duncan
1997; McLoyd 1994; McLoyd and Flanagan 1990).
There is evidence of promising effects of early
intervention programs, such as Sure Start in the UK,
which can improve the life chances of young children
(Melhuish, Belsky, Leyland, and Barnes, 2008).
Indeed, there is persistent research evidence to
suggest that early developmental trajectories can
change over time (Feinstein 2003; Rutter 1989;
Schoon 2006). It is thus vital to learn more about the
factors and processes that can potentially ameliorate
the negative impact of poverty on children’s early
development. For example, economic hardship has a
differential effect on specific child outcomes,
generally exhibiting a stronger effect on children’s
cognitive development than on behaviour (Conger
and Elder 1994; Kiernan and Huerta 2008; Kiernan
and Mensah 2009; Linver et al 2002; Plewis and Kallis
2008; Schoon, Cheng and Jones 2010). We thus will
examine the pathways linking family hardship to
cognitive and behavioural adjustment separately. The
lack of understanding of how the experience of
hardship influences child development, has greatly
hampered the ability of policy makers to design
effective interventions to improve children’s
development and wellbeing.
Theoretical models linking the experience of
material hardship to child outcomes have focused in
particular on the mediating role of family
psychological stress (Conger and Elder 1994; Mistry,
Biesanz, Taylor, Burchinal and Cox 2004; Yeung et al
2002), parent’s investments of time or money in their
children (Guo and Harris 2000), or a combination of
these factors (Gershoff et al 2007; Kiernan and
Mensah 2009; Linver, Brooks-Gunn and Kohen 2002;
Yeung et al 2002). The family investment model
asserts that income is associated with children’s
development, because it limits the amount of
resources, including money, time, energy, and
support, they have available for their children (Becker
and Thomes 1986; Haveman and Wolfe 1994; Mayer
1997). It does however, not specify how economic
circumstances might impact the quality of parent-
child interactions. The family stress model, on the
other hand, postulates that family hardship influences
children’s cognitive development and behaviour
through parental emotional distress resulting from
financial strain, which negatively influences parenting
practices, which are in turn associated with poorer
child outcomes (Conger, Ge, Elder, Lorenz and Simons
1994; Conger et al 1992, 1993; Elder and Caspi 1988;
McLoyd 1989; McLoyd 1994). It has also been shown
that parental psychological distress impacts on
parent’s abilities or willingness to invest in their
children, suggesting the appropriateness of
combining both models (Gershoff et al 2007; Kiernan
and Huerta 2008; Linver et al 2002; Yeung et al 2002).
However, there is also evidence to suggest that
different components of the family environment may
have differential effects on child outcomes. While the
provision of stimulating experiences in the home
environment is shown to be more strongly associated
with children’s cognitive development than with
behavioural adjustment, parent-child relations were
more strongly associated with children’s behaviour
(Linver et al 2002; Kiernan and Huerta 2008). It is thus
important to differentiate between cognitive and
emotional components of parenting, and to assess
their relative impact on child adjustment.
In the following we will test the usefulness of
combining the family stress and investment model to
explain variations in early adjustment, by drawing on
data collected for the UK Millennium Cohort. We will
assess the relative sizes of associations when
considering several aspects simultaneously, as well as
in their separate effects. Adding to the existing few
studies examining the mediating processes by which
family economic hardship influences cognitive and
behavioural development of young children, we will
take into account the effects of persistent hardship
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
212
and persistent maternal stress during early childhood,
and their impact on school readiness and behavioural
adjustment by age 3. In addition, instead of focusing
on household income (Kiernan and Mensah 2009), we
take into account material resources available to the
family at age 9 months and 3 years, to account for
persistence of family hardship. Furthermore, we
conceptually differentiate between proximal and
distal processes (Bronfenbrenner 1979), following the
assumption that the strongest influence on children’s
development are processes and interactions directly
experienced by the child, such as parenting
behaviour. Distal characteristics, such as family
income and material hardship, impact on children
insofar as they shape these proximal processes.
Furthermore, we take account of a number of
background, or control variables to ensure that the
predicted findings were not spurious. It has been
argued that economic hardship has little, or no, direct
effect on children’s outcomes, which are largely due
to other characteristics of the parents, such as
parental education (Rowe and Roger 1997). We thus
control for maternal education, as well as maternal
age, ethnicity, marital status at birth, total number of
siblings living in the household, sex of the child, low
birthweight (<2500 grams) and prematurity. Including
these basic demographic characteristics in the model
gives greater confidence in the links between the
variables included in the model.
Following from previous research, we tested the
following hypotheses: a.) family hardship is directly
associated with child outcomes, although the
association may vary across different outcomes, i.e.
cognitive and behavioural adjustment; b.) the
association between family hardship and child
outcomes persists after controlling for socio-
demographic characteristics of the family and
biological risk factors; c.) the association between
family hardship and child outcomes is mediated by i.)
constructs of the family stress model; ii.) the family
investment model; iii.) by the combination of both
models.
Method
Sample
The study draws on data collected for the
Millennium Cohort Study (MCS), a survey of 18,819
babies born between September 2000 and January
2002 into 18,553 families living in the UK (Dex and
Joshi 2005). The first sweep of the Millennium Cohort
Study was carried out during 2001 and 2002 when
most babies were 9-months old. The sample design
allowed for disproportionate representation of
families living in areas of child poverty. Electoral
wards based on 1998 geography were used as the
sampling frame and information about child poverty
was incorporated as provided in the Index of
Deprivation (Noble et al 2000). Due to
disproportionate sampling, special weights have to be
applied in analyzing the data (Plewis, Calderwood,
Hawkes, Hughes, and Joshi 2004).
Data were collected from the parents of the babies
via personal interview and self-completion
questionnaire, as well as direct assessment of
children’s cognitive abilities (Plewis et al 2004;
Shepherd, Smith, Joshi and Dex 2004). The following
analyses are based on 14,661 singleton babies, with
data from the 9 and 36 month data collection sweeps.
Measures
Focal variable: material hardship
An index of family material hardship was created,
based on five indicators assessed at both age 9
months and 3 years, including low net household
income [less than £10,400 per annum, which
represents the bottom quartile], receipt of income
support (a means-tested benefit), access to a car or
van, as well as housing-based measures including
home ownership and overcrowding [more than one
person per room]. The individual items were
dichotomised to create a summary hardship index for
each time point, with an observed range of 0 to 5.
Scale values of four and five were combined due to
small numbers, and indicate severe levels of hardship,
while a score of 0 indicates lack of hardship. There
was a strong correlation between reported hardship
at the two ages (r=0.81).
Mediating variables
Maternal emotional distress was assessed when the
child was 9 months and 3 years old. At age 9 months
(in 2000/1) a shortened 9-item version of the Rutter
Malaise Inventory (Rutter, Tizard and Whitmore
1970) was used. The Malaise Inventory is a self-
completion measure that has been widely used as a
measure of depression, anxiety and psychosomatic
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
213
illness in general population studies (McGee, Williams
and Silva 1986; Rodgers, Pickles, Power, Collishaw
and Maughan 1999) as well as in investigations of
high-risk groups, notably informal carers (Grant,
Nolan, and Ellis 1990). The shortened scale ranges
from 0 to 9, has acceptable internal consistency
(Cronbach’s alpha =.73), and correlates significantly
with previously diagnosed and currently treated
depression.
At age 3 (in 2003/4) the 6 item Kessler
psychological distress scale (K6) was used for the
identification of maternal anxiety and depression. The
K6 is a widely used screening instrument, which has
been especially developed for use in population
surveys (Kessler et al 2002). Responses are given on a
four-point Likert scale and are summed to produce a
uni-dimensional scale (alpha for the MCS=0.86) with a
range of 0 to 24. The two measures of psychological
distress are moderately correlated (r=0.47).
Parent-child relationship was assessed at age 3 years
using the Pianta scale (Pianta 1992), a 15 item self-
administered rating scale with responses on a 5-point
Likert scale. A total score was derived, with a high
score reflecting an overall positive relationship. The
alpha coefficient for the Pianta scale in the MCS
sample was .77.
Cognitive stimulation at age 3 years was measured on
the basis of maternal report on whether the child was
read to, taught the alphabet, counting or songs, at
least once a week, and whether the parents took the
child on visits to the library. The five items were
summed to an index of cognitive stimulation ranging
from 0 to 5.
Child Outcomes at age 3
School Readiness was assessed with the Bracken
School Readiness Assessment (BSRA) which was
individually administered to each child. The BSRA
comprises six subtests measuring children’s
knowledge of colours, letters, numbers, sizes,
comparisons of objects, and shapes (Bracken 2002). It
is a developmentally sensitive measure of children's
basic concept acquisition and receptive language
skills, designed for children ages two and a half
through to age seven. The BSRA has strong
psychometric characteristics and good validity (Panter
and Bracken 2009). In the following analysis we use
age-standardised scores.
Behavioural adjustment is measured with the
Strength and Difficulties Questionnaire (SDQ), a
behavioural screening questionnaire for 3 to 16 years
olds (Goodman 1997, 2001). It consists of 25 items,
assessed via parental report, generating an overall
scale score as well as scores for five subscales:
conduct problems, hyperactivity, emotional
symptoms, peer problems and pro-social behaviour.
For the following analysis, an overall difficulties mean
score for the whole sample, was computed by
summing replies to the subscales indicating behaviour
problems, i.e. conduct problems, hyperactivity,
emotional symptoms, and peer problems.
Control variables
A number of control variables were included in the
analysis to make sure that the results are not
spurious:
Mother’s age at birth of child
Mother’s marital status at the birth of the child
Mother’s education (below GCSE versus GCSE and
above)
Mother’s ethnicity (white versus other)
Total number of children living in household
Baby gender (male, female)
Prematurity (gestation period less than 37 weeks)
Low birthweight (less than 2500 grams)
Analytic Strategy
Structural equation modeling was used to assess
the pathways linking family hardship to children’s
developmental outcomes at age 3 years. All analyses
were carried out using the statistical package Mplus
5 (Muthén and Muthén 2007). This method allows
analysis of cases with missing data under the
assumption that the data are missing at random
(Little and Rubin 2002). Probit regressions were used
based on robust weighted least squares estimation.
Because some of the dichotomised variables
functioned as both independent and dependent
variables in the conceptual model, the theta
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
214
parameterization was necessary. Regression
estimates convert probit estimates for ordinal
dependent variables to a common metric that allows
comparison with standardised linear regression
estimates for the continuous variables.
In line with current practice, several criteria were
used to assess the fit of the data to the model. The 2
statistic is overly sensitive to model mis-specification
when sample sizes are large, or the observed
variables are non-normally distributed. The root
mean square error of approximation (RMSEA) gives a
measure of the discrepancy in fit per degrees of
freedom (e.g. values less than .05 indicate a good fit).
The final index of choice is the comparative fit index
(CFI), indicating if the model provides significantly
better explanation of the relations between variables
than the null hypothesis-model with no relations
between variables. Values above .95 indicate an
acceptable fit (Bentler 1990).
In a first step, bivariate Pearson’s correlations
between variables were calculated. Table 1 shows
means, standard deviations and correlations between
variables under study. In the next step the authors
investigated the association between family adversity
and child outcomes as well as mediator effects for
cognitive and behavioural adjustment separately. To
decompose the relative impact of family hardship,
control variables, and mediating variables, we tested
separate models. Model a establishes the direct link
between hardship and child outcomes. Model b adds
the control variables, and model c, the mediating
variables. This analysis sequence allowed the
examination of whether the relationship between
family hardship is partially or fully mediated by the
addition of the control variables and the mediators.
Furthermore, different models were tested to assess
whether family stress constructs, a parental
investment measure (i.e. cognitive stimulation), or
both, act as mediators of the relations between
family hardship and child outcomes.
Ingrid Schoon, Stephen Hope, Andy Ross and Kathryn Duckworth Family hardship and children’s development
215
Table 1: Means, Standard Deviations, and bivariate Pearson Correlations between variables included in the model (* p<0.01; ** p<0.001)
Variables
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
M
SD
1. Material
hardship (9
mths)
1.00
1.06
1.38
2. Material
hardship
(36 mths)
.81**
1.00
0.96
1.33
3. Bracken
-.34**
-.34**
1.00
103.69
16.29
4. SDQ
.31**
.31**
-.29**
1.00
9.55
5.26
5. Malaise
(9 months)
.18**
.18**
-.10**
.28**
1.00
1.67
1.76
6. Kessler
(36 mths)
.25**
.27**
-.17**
.36**
.47**
1.00
3.54
3.86
7. Pianta
-.19**
-.18**
.17**
-.61**
-.27**
-.37**
1.00
64.51
6.86
8. Cogn
stimulation
-.15**
-.15**
.26**
-.17**
-.06**
-.09**
.11**
1.00
3.74
0.99
9. Maternal
ethnicity
-.18**
-.15**
.21**
-.10**
-.05**
-.12**
.02*
.13**
1.00
0.85
0.54
10.Maternal
age
-.42**
-.40**
.18**
-.23**
-.08**
-.12**
.15**
-.05*
.04**
1.00
28.66
5.87
11.Maternal
education
-.41**
-.40**
.29**
-.24**
-.09**
-.15**
.11**
.18**
.17**
.18**
1.00
0.77
0.42
12. Married
at birth
-.59**
-.53**
.17**
-.18**
-.09**
-.15**
.12**
.06**
.03**
.29**
.21**
1.00
0.83
0.37
13. Nr of
children
.13**
.12**
-.23**
.01
.06**
.06**
.06**
-.14**
-.12**
.32**
-.19**
.04**
1.00
0.94
1.06
14. Birth-
weight low)
.06**
.06**
-.07**
.06**
.04**
.05**
-.02*
.03**
-.09**
-.01
-.06**
-.03**
-.05
1.00
0.06
0.24
15.
Gestation
(premature)
.03*
.03**
-.04**
.04**
.03*
.04**
-.01
.00
-.01
-.00
-.01
-.02*
-.01
.50**
1.00
0.07
0.25
16. Gender
(female)
.00
.01
.12**
-.10**
-.02*
-.01
.06**
.07**
-.01
.00
.01
.00
.00
.02†
-.01
1.00
0.49
0.50
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
216
Results
Figure 1 shows the structural equation model
assessing the pathways linking family hardship to
school readiness. The usual structural equation
modeling conventions are used, depicting latent
variables as a circle, and manifest variables in
rectangles. The two latent variables comprise
indicators of family hardship on the one hand, and
maternal distress on the other, providing measures of
persistent hardship and persistent distress, averaging
the experiences at age 9 months and 3 years. Unique
and error variance for each manifest variable and
disturbance on the latent variables are included in the
model (not shown in the diagram). Path estimates are
given as standardised regression coefficients, that
may be squared to obtain the variance shared by
adjacent variables. All paths in the model were
significant at the 5% level (parameter estimates
divided by their standard errors), and the model
provides a good fit to the data.
Figure 1. Predicting school readiness at age 3: the full model
(a. no mediators no controls, b. no mediators with controls, c. with mediators and controls)
Cognitive
stimulation
Control
Battery
.34
-.47
.19
-.28 (c)
Parent-child
relationship
.08
-.15
-.30 (b)
-.38 (a)
Hardship
School
readiness
MODEL FIT
Χ2=112.86***, df=23
CFI=.995
RMSEA=.016
Maternal
emotional
distress
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
217
The association between family hardship and
cognitive ability was statistically significant, both
without (ß=-.38) and with controls (ß=-.30). With
controls the model explained 19% of the variance in
cognitive ability. After adding the mediators, the path
from family hardship to child cognitive development
reduced to ß=-.28, suggesting that parenting factors
only partially mediate the association between poverty
and cognitive development. Adding the parenting
characteristics enables us to explain an additional 4%
of variance in school readiness, in addition to that
explained by family poverty and controls.
Figure 2. Predicting behaviour adjustment at age 3: the full model
a. no mediators no controls, b. no mediators with controls, c. with mediators and controls)
In a next step we assessed the pathways linking
family hardship to behavioural adjustment (Figure 2).
Family hardship was significantly associated with
behaviour problems (ß=-.34). Adding the control
variables reduces the direct association to ß=-.26.
When the controls are included, the model explained
14 per cent of the variance in behaviour problems.
The full model, depicted in Figure 2 shows the
combination of both family stress and family
investment constructs. After adding all the mediators
we can explain an additional 32 per cent of the
variance in behavioural adjustment, and the path
from family hardship to child cognitive development
reduced to ß=-.15.
Cognitive
stimulation
Control
Battery
.39
-.48
-.09
.15 (c)
Parent-child
relationship
-.56
-.15
.26 (b)
.34 (a)
Hardship
Behaviour
problems
MODEL FIT
Χ2=320.74, df=19
CFI=.988
RMSEA=.033
Maternal
emotional
distress
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
218
Discussion
The study illustrates the corrosive effect of family
hardship on the cognitive development and
behavioural adjustment of young children. The
experience of hardship in the first three years of life
undermines the formation of skills that are necessary
for the child to succeed in their school careers. The
study furthermore identifies the role of
characteristics in the family environment as potential
mediators, differentiating between the impact of
constructs identified within the family stress and the
family investment models. In particular, the study
tests the viability of combining both models to gain a
better understanding of how family hardship is
associated with early developmental outcomes (see
also Linver et al 2002). Constructs of both models
mediated the association between family hardship
and child development. However, the provision of
stimulating experiences in the home appears to be
more important for the cognitive development of the
child, while family stress constructs emphasizing the
role of maternal distress and less-involved parenting,
appear to be especially important for behavioural
adjustment (see also Linver et al 2002; Kiernan and
Huerta 2008; Yeung, Linver, and Brooks-Gunn 2002).
While parenting characteristics explained relative
little of the variation in cognitive development, in
addition to the influence of family hardship and the
control variables, they were crucial in reducing the
negative impact of family hardship on behavioural
adjustment.
Furthermore, combining both models enabled us
to illustrate how distal influences impact on more
proximal experiences of the child, and to identify the
role of persistent maternal distress as a mediating
factor, linking family hardship to parenting behaviour
as well as cognitive stimulation, which in turn
influences children’s development. Maternal
depression is generally considered a risk factor for
poor socio-emotional and cognitive development
(Cummings and Davies 1994), although the
associations between maternal depression and child
outcomes are complex (Downey and Coyne 1990),
and not all studies have found a relationship between
maternal distress and cognitive development (Kiernan
and Huerta 2008; Linver, Brooks-Gunn and Kohen
2002). Variations in severity, chronicity, and timing of
depression (Campbell, Cohn and Meyers 1995), as
well as heterogeneity in sampling and other potential
risk factors such as low social support, can contribute
to differences in child outcomes (Sameroff et al
1993). In this study, the authors accounted for
persistence of maternal distress between ages 9
months and 3 years, and found that mothers exposed
to persistent hardship, with reduced access to
economic resources, are more likely to experience
continued stress, which in turn is associated with
reduced investment in their children (in terms of
cognitive stimulation) as well as less involved parent-
child interactions, which in turn are associated with
their children’s developmental outcomes. These
associations were significant even after controlling for
a number of background characteristics, such as
mother’s education, ethnicity, and marital status, as
well as indictors of early biological risk, to ensure that
the findings were not spurious. It should be noted
that in another study also using the Millennium
cohort, the association between persistent maternal
distress and cognitive functioning was also apparent
(Kiernan and Mensah 2009).
In interpreting the findings, a number of
limitations have to be considered: the hypothesized
pathways examined in the model, test specific
assumptions regarding the combination of the family
stress and the family investment model. The
observed associations do not imply causal
relationships between the factors, as there might be
other explanatory processes not included in the
model. For example, there might be a reciprocal
relationship between the child characteristics and
parenting behaviour (Bell and Chapman 1986; Rutter
2002), and parenting behaviour might change over
time. Also, the role of the father in supporting
positive development in the face of family hardship
has not been addressed. Furthermore, while family
hardship and maternal distress were assessed at two
time points, measures of family investment and
parent-child interactions were only available at age 3,
the same age when the outcome variables were
assessed. Another limitation is that, except for the
assessment of cognitive ability, all other measures
were obtained via maternal report, and the inclusion
of some objective or independent observational data
would have helped to improve the validity of the
findings. It is also likely that other mediators of the
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
219
association between family hardship and child
outcomes exist that are beyond the scope of this
study, as for example characteristics of the
neighbourhood, or availability of social support.
Given these limitations, the findings provide some
useful insights into the pathways linking family
hardship to early cognitive and behavioural
functioning. The findings suggest that economic
hardship has a slightly stronger association with
cognitive than with behavioural development,
confirming evidence from previous studies (Conger
and Elder 1994; Conger et al 1994; Kiernan and
Huearta 2008; Linver et al 2002, Plewis and Kallis
2008; Schoon et al 2010). The study furthermore
highlights the role of maternal depression as a
mediator between distal and proximal experiences,
and its association with good quality parent-child
interactions, as well as the provision of a stimulating
home environment. Of course, not all mothers
suffering from depression are affected in their ability
to provide a good enough, sensitive and caring
environment for their children (Cicchetti, Rogosch,
and Toth 1998), yet maternal depression appears to
be a risk for their children’s cognitive and especially
behavioural adjustment. About 17% of mothers in the
Millennium Cohort reported that they were
depressed when their child was 9 months old, as well
as at age 3 (Kiernan and Mensah 2009). Contextual
risk factors such as poverty, marital conflict, and
stressful life events may exacerbate maternal
depression, and consequently the child’s
development, suggesting the importance of being
vigilant in detecting or screening for maternal
depression, especially among highly disadvantaged
families. The findings furthermore suggest the
usefulness of disentangling the emotional and
cognitive components of parenting and the home
environment, to gain a better understanding of the
processes shaping cognitive and behavioural
development. While parenting processes are more
effective in mediating the influence of poverty on
behavioural adjustment, they play a relative small
role in mediating the effects of family hardship on
cognitive development. It is therefore not enough to
develop policies targeting the improvement of
parenting behaviour and parental health. What is
needed is a concentrated effort to reduce or
eradicate child poverty and to improve the living
conditions of families with young children.
The study has shown that family hardship has a
direct influence on children’s developmental
outcomes and plays a role in shaping maternal mental
health as well as parenting behaviours. Given the
long-term consequences of achievement gaps
emerging early in life, the fact that this gap widens
throughout the childhood years (Feinstein 2003;
Schoon 2006), and that children who fall behind in
early development are more likely to fall further
behind at subsequent stages, renders the reduction
of family hardship during the first years of life a
priority (Heckman 2006; Marmot 2010). The
possibility of correlated unobserved characteristics
and alternative mediating processes, opens the field
for further investigation into the mechanisms and
processes involved in the early inter-generational
transmission of disadvantage. These efforts should
focus their attention to both cognitive and
behavioural adjustment during the early years, as
both capabilities are the foundation for later
developmental adjustment, and are cross-fertilizing,
as shown in Duckworth and Schoon 2010 (this issue).
Acknowledgements
The analysis and writing of this paper were supported by grants from the Nuffield Foundation and the UK
Economic and Social Research Council (ESRC): L326253061, RES-594-28-0001. Data from the Cohort Studies
were supplied by the ESRC Data Archive. Those who carried out the original collection of the data bear no
responsibility for its further analysis and interpretation.
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
220
References
Alexander RE. (2009) Children, their world, their education: final report and recommendations of the Cambridge
primary review. Routledge, London.
Becker GS and Thomes N. (1986) Human capital and the rise and fall of families. Journal of Labour Economics, 4,
S1-S139.
Bell RQ and Chapman M. (1986) Child effects in studies using experimental or brief longitudinal approaches to
socialization. Developmental Psychology, 22(5), 595-603.
Blanden J and Gregg P. (2004) Family income and educational attainment: a review of approaches and evidence
for Britain. Oxford Review of Economic Policy, 20, 245-236.
Blanden J and Machin S. (2010) Intergenerational inequality in early years assessments. In K Hansen, H Joshi and
S Dex. eds. Children of the 21st century. The first five years (pp. 153-168). Policy Press, Bristol.
Bollen KA. (1989) Structural Equations with Latent Variables. Wiley, New York.
Bradley R, Caldwell BM, Rock SL and Ramey CT. (1989) Home environment and cognitive development in the
first 3 years of life: a collaborative study involving six sites and three ethnic groups in North America.
Developmental Psychology, 25, 217-235.
Bradley R and Corwyn R. (2002) Socio-economic status and child development. Annual Review of Psychology, 53,
371-399.
Bronfenbrenner U. (1979) The ecology of human development : experiments by nature and design. Harvard
University Press, Cambridge, MA.
Brooks-Gunn J and Duncan GJ. (1997) The effects of poverty on children. Children and Poverty, 7(2), 55-71.
Bracken BA. (2002) Bracken school readiness assessment administration manual. The Psychological Corporation,
San Antonio, Texas.
Brooks-Gunn J and Duncan GJ. (1997) The effects of poverty on children. Children and Poverty, 7(2), 55-71.
Bynner J and Joshi H. (2002) Equality and opportunity in education: evidence from the 1958 and 1970 birth
cohort studies. Oxford Review of Education, 28(4), 405-425.
Bynner J, Schuller T and Feinstein S. (2003) Wider benefits of education: skills, higher education and civic
engagement. Zeitschrift für Pädagogik, 49(3), 341-361.
Campbell S, Cohn J and Meyers T. (1995) Depression in first-time mothers: mother-infant interaction and
depression chronicity. Developmental Psychology, 31, 349-357.
Cicchetti D, Rogosch FA, and Toth SL. (1998) Maternal depressive disorder and contextual risk: contributions to
the development of attachment insecurity and behavior problems in toddlerhood. Development and
Psychopathology, 10(2), 283-300.
Conger KJ and Elder GH. (1994) Families in troubled times: adapting to change in rural America (1st edn.) Aldine
De Gruyter, New York.
Conger R, Ge X, Elder GH, Lorenz F and Simons RL. (1994) Economic stress, coercive family process, and
developmental problems of adolescents. Child Development, 65, 541-561.
Cummings E and Davies P. (1994) Maternal depression and child development. Journal of Child Psychology and
Psychiatry, 35, 73-112.
Downey G and Coyne JC. (1990) Children of depressed parents - an integrative review. Psychological Bulletin,
108(1), 50-76.
Duckworth K and Schoon I. 2010 Progress and attainment during primary school: the roles of literacy, numeracy
and self-regulation. Longitudinal and Life Course Studies 1, 223-240.
Duncan GJ, Dowsett CJ, Claessens A, Magnuson K, Huston AC, Klebanov P, Pagani LS, Feinstein L, Engel M,
Brooks-Gunn J, Sexton H, Duckworth K and Japel C. (2007) School readiness and later achievement.
Developmental Psychology, 43(6), 1428-1446.
Entwistle DR and Alexander, KL. (1999) Early schooling and social stratification. In R Pianta and M Cox. eds. The
transition to kindergarten (pp. 13-38). Brookes, Baltimore.
Feinstein L. (2003) Inequality in the early cognitive development of British children in the 1970 cohort.
Economica, 70, 73-98.
Feinstein L. (2004) Mobility in pupils' cognitive attainment during school life. Oxford Review of Economic Policy,
20(2), 213-229.
Feinstein L and Bynner J. (2004) Mobility in pupils' cognitive attainment during school life. Oxford Review of
Economic Policy, No 20.
Feinstein L and Vignoles A. (2008) Individual differences in the pathways into and beyond higher education in
the UK: a life-course approach. Journal of Social Issues, 64(1), 115-133.
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
221
George A, Hansen K and Schoon I. (2007) Child development. In K Hansen and H Joshi eds. Millennium cohort
study. second survey. A user's guide to initial findings. Institute of Education, Centre for Longitudinal
Studies, London.
Gershoff ET, Aber JL, Raver CC and Lennon MC. (2007) Income is not enough: incorporating material hardship into
models of income associations with parenting and child development. Child Development, 78(1), 70-95.
Guo G and Harris KM. (2000) The mechanisms mediating the effects of poverty on children's intellectual
development. Demography, 37(4), 431-447.
Heckman JJ. (2006) Skill formation and the economics of investing in disadvantaged children. Science, 312, 1900-
1912.
Hills J, Sefton T and Steward K. (2009) Towards a more equal society. Poverty, inequality and policy since 1997.
Policy Press, Bristol.
Janus M and Duku E. (2007) The school entry gap: socio-economic, family, and health factors associated with
children's school readiness to learn. Early education and development, 18, 375-403.
Kagan SL. (1992) Readiness past, present and future. Shaping the agenda. Young Children, 48, 48-53.
Kiernan KE and Huerta MC. (2008) Economic deprivation, maternal depression, parenting and children's
cognitive and emotional development in early childhood. British Journal of Sociology, 59(4), 783-806.
Kiernan KE and Mensah FK. (2009) Poverty, maternal depression, family status and children's cognitive and
behavioural development in early childhood: a longitudinal study. Journal of Social Policy, 38, 569-588.
Korenman S, Miller JE and Sjaastad JE. (1995) Long-term poverty and child development: evidence from the
NLSY. Children and Youth Services Review, 17(1-2), 127-155.
Linver MR, Brooks-Gunn J and Kohen DE. (2002) Family processes as pathways from income to young children's
development. Developmental Psychology, 38(5), 719-734.
Lloyd JEV and Hertzman C. (2009) From kindergarten readiness to fourth-grade assessment: longitudinal analysis
with linked population data. Social Science and Medicine, 68(1), 111-123.
MacInnes T, Kenway P and Parekh A. (2009) Monitoring poverty and social exclusion. Joseph Rowntree
Foundation, York.
Marmot M. (2010) Fair society healthy lives. The Marmot review: Strategic Review of Health Inequalities in
England post 2010. http://www.ucl.ac.uk/gheg/marmotreview/FairSocietyHealthyLives.
McLoyd VC. (1998) Socioeconomic disadvantage and child development. American Psychologist, 53(2), 185-204.
Melhuish E, Belsky J, Leyland,AH and Barnes J. (2008) Effects of fully-established Sure Start local programmes on
3-year-old children and their families living in England: a quasi-experimental observational study. Lancet,
372(9650), 1641-1647.
Meisels SJ. (1999) Assessing readiness. In R Pianta and M Cox. eds. The transition to kindergarten (pp. 39-66).
Brookes, Baltimore.
Mistry RS, Biesanz JC, Taylor LC, Burchinal M and Cox MJ. (2004) Family income and its relation to preschool
children's adjustment for families in the NICHD study of early child care. Developmental Psychology,
40(5), 727-745.
Panter JE, and Bracken BA. (2009) Validity of the Bracken School Readiness Assessment for predicting first grade
readiness. Psychology in the Schools, 46(5), 397-409.
Plewis I and Kallis C. (2008) Changing economic circumstances in childhood and their effects on subsequent
educational and other outcomes. DWP Working Paper No. 49. HMSO, Norwich.
Rutter M. (1989) Pathways from childhood to adult life. Journal of Child Psychology and Psychiatry and Allied
Disciplines, 30(1), 23-51.
Rutter M. (2002) The interplay of nature, nurture, and developmental influences - the challenge ahead for
mental health. Archives of General Psychiatry, 59(11), 996-1000.
Sameroff AJ, Seifer R, Baldwin A and Baldwin C. (1993) Stability of intelligence from preschool to adolescence -
the influence of social and family risk-factors. Child Development, 64(1), 80-97.
Schoon I. (2006) Risk and resilience. Adaptations in changing times. Cambridge University Press, Cambridge.
Schoon I, Bynner J, Joshi H, Parsons S, Wiggins RD and Sacker A. (2002) The influence of context, timing, and
duration of risk experiences for the passage from childhood to mid-adulthood. Child Development,
73(5), 1486-1504.
Schuerger JM and Witt AC. (1989) The temporal stability of individually tested intelligence. Journal of Clinical
Psychology, 45, 294-302.
Waldfogel J and Washbrook E. (2010) Low income and early cognitive development in the UK. The Sutton Trust
http://www.suttontrust.com/reports/Sutton_Trust_Cognitive_Report.pdf
Yeung WJ, Linver MR and Brooks-Gunn J. (2002) How money matters for young children's development: parental
investment and family processes. Child Development, 73(6), 1861-1879.
Ingrid Schoon, Andy Ross, Stephen Hope and Kathryn Duckworth Family hardship and children’s development
222
