scispace - formally typeset
Search or ask a question
Journal ArticleDOI

Parental Resources and Child Abuse and Neglect

01 May 1999-The American Economic Review (American Economic Association)-Vol. 89, Iss: 2, pp 239-244
TL;DR: This article examined whether child maltreatment is affected by the socioeconomic circumstances of parents and found that children are more likely to be maltreated if their parents have fewer resources, and that socioeconomic circumstances do matter.
Abstract: A child’s welfare is affected not only by the wealth of her parents, but also by the quality of care her parents provide. Physical abuse, neglect, and other forms of child maltreatment impose severe hardships on children and may adversely affect them as adults (Cathy Widom, 1989) . We examine whether child maltreatment is affected by the socioeconomic circumstances of parents. Our hypothesis is that children are more likely to be maltreated if their parents have fewer resources. We use a broad conception of ‘‘resources.’’ It encompasses not only income, but also parental time and the quality of parental time. For example, a low-income working single mother may be short on resources needed to parent not only because she earns a low income, but also because she may not have the physical or emotional reserves to care for her children properly at the end of the day. Likewise, an unemployed father may provide less than adequate parenting not only because his income has been reduced, but also because of the depression and loss of self-esteem that may accompany unemployment (Arthur Goldsmith et al., 1996). We use state-level panel data to analyze the impact that socioeconomic circumstances ( in particular, parental work status and single parenthood) have on the incidence of child maltreatment. We find that socioeconomic circumstances do matter. States with higher fractions of children with absent fathers, and especially absent fathers and working mothers, have higher rates of child

Summary (1 min read)

I. Background

  • Child maltreatment is a large and growing problem in the United States.
  • The most common type of maltreatment reported to CPS is neglect, which constitutes about 58 percent of all reports.
  • None of the previous studies has analyzed family structure and parental ernployment in the kind of detail that the authors use here, and none has analyzed a panel of national data.

II. Data, Methods, and Results

  • Measuring child maltreatment is not a simple task.
  • Some states collect information on the number of families reported for child maltreatment, and some collect information on the number of children reported to be victims of abuse or neglect.
  • The authors definitions of "mother" and "father" require discussion.
  • As shown in Angus Deaton (1985) , the bias in the parameter estimates is a function of the variances and covariances of the state-level means constructed from the CPS.
  • The results in Table 1 indicate that higher fractions of children living with working mothers, no fathers, or nonworking fathers are associated with more reports of maltreatment, more substantiated reports, and more cases of physical abuse and neglect.

III. Conclusion

  • Using state-level panel data, the authors find that socioecononmic circumstances (in particular, parental work status and single parenthood) affect the incidence of child maltreatment.
  • States with higher fractions of children with absent fathers, and especially those with absent fathers and working mothers, have higher rates of child maltreatment, as do states with higher shares of nonworking fathers.
  • The authors also find some evidence that states with higher incomes have lower rates of child abuse and neglect, although this result was sensitive to specification.
  • The authors worK adds to the growing literature that relates economic circumstances to child well-being.
  • The authors results also have implications for the effects on children of welfare reforms that move single parents into the labor force without substantially increasing their incomes.

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

Parental
Resources
and
Child Abuse
and
Neglect
By
CHRISTINA
PAXSON
AND JANE
WALDFOGEL
*
A
child's
welfare is affected
not only by
the
wealth of her
parents, but also
by the qual-
ity
of care her
parents
provide. Physical
abuse, neglect, and other forms of child
mal-
treatment impose severe
hardships on chil-
dren and
may
adversely
affect
them
as adults
(Cathy Widom,
1989).
We examine
whether
child maltreatment is
affected
by the socio-
economic
circumstances of
parents. Our hy-
pothesis
is that
children are more
likely
to be
maltreated
if
their
parents
have
fewer re-
sources. We use a broad
conception
of "re-
sources."
It
encompasses
not
only
income,
but
also
parental time and the
quality
of
parental
time.
For example,
a
low-income
working sin-
gle mother
may
be short on
resources
needed
to
parent not
only
because
she
earns
a
low in-
come,
but also because she
may
not
have the
physical or
emotional
reserves to care for
her
children
properly
at the end of the
day.
Like-
wise,
an
unemployed
father
may
provide
less
than
adequate
parenting
not
only
because his
income has been
reduced,
but also
because of
the
depression
and loss of self-esteem
that
may
accompany
unemployment (Arthur
Goldsniith
et
al., 1996).
We use state-level
panel
data to
analyze
the
impact
that
socioeconomic
circum-
stances
(in
particular, parental
work
status
and
single
parenthood)
have
on
the inci-
dence of child
maltreatment. We find that
so-
cioeconomic circumstances
do matter.
States
with
higher
fractions
of children with
absent
fathers,
and
especially
absent
fathers and
working mothers,
have
higher
rates
of child
maltreatment.
Nonworking
fathers
are also
as-
sociated with
higher
rates of
maltreatment.
I.
Background
Child
maltreatment is a
large and
growing
problem
in
the United
States.
In
1996, over 3
million
cases of child
abuse and
neglect,
nearly 50
cases per
thousand
children, were
reported to
state
child-protective
services
(CPS)
agencies,
about a
fivefold
increase over
the
number
and
rate just 20
years
earlier
(Waldfogel,
1998).
The
most
common
type
of
maltreatment
reported to
CPS
is
neglect,
which
constitutes about 58
percent
of
all
re-
ports.
Physical abuse
makes
up
22
percent
of
reports.
Sexual
abuse,
emotional
maltreat-
ment,
and
other
categories
together
account
for the
remaining
20
percent (U.S.
Department
of
Health
and
Human
Services,
1998).
About
40
percent of
reports
are
substantiated
upon
investigation
by
CPS, and
just under
30 per-
cent are
kept
open
for
ongoing
intervention,
which
may
involve
removing
the
child from
home or
monitoring
the child's
safety
at
home
(Waldfogel,
1998).
It
has
long
been
noted
in
the
child-abuse
literature that children
who are
poor,
have un-
employed
fathers,
or
live with
single mothers
are more
likely
than
others to be
reported
to
CPS
(see
e.g., David
Gil,
1970;
Duncan
Lindsey,
1994),
although
it
has not been clear
whether
such children
really are more
likely
to
be maltreated or
are
simply
more
likely
to
be
reported.
There is also
evidence from
com-
munity studies that
children
living
in
poor
areas are more
likely
to be
identified as mal-
treated,
as
are children from
communities with
higher
levels of
unemployment
or
lone
par-
enthood
(see Lawrence
Steinberg
et
al., 1981;
James
Spearly
and Michael
Lauderdale, 1983;
Sheila
Ards, 1989;
James Garbarino
and
Kathleen
Kostelny,
1992;
and
Claudia Coulton
et
al., 1995).
However,
none of the
previous
studies has
analyzed
family
structure and
pa-
rental
ernployment
in the kind
of detail that
we
*
Paxson:
Woodrow Wilson
School, Princeton
Univer-
sity, Princeton
NJ
08544, and
NBER; Waldfogel:
School
of Social
Work, Columbia
University, New York,
NY
10027,
and Center for
Analysis
of Social
Exclusion,
Lon-
don School of
Economics.
We
are grateful to the
MacArthur Foundation for
financial
support,
to
Anne
Case,
Joanne
Gowa, Julie
Nelson,
and
Robert
Pollak for
useful comments, and to Amanda
Lockshin and Yan
Jiang
for research assistance.
239

240 AEA PAPERS
AND PROCEEDINGS
MAY 1999
use here,
and none has
analyzed a panel
of
national data.
II. Data, Methods,
and Results
Measuring
child maltreatment is
not
a sim-
ple task. Direct measures
of child maltreat-
ment are difficult to
obtain,
since interviews
with
parents
or children are unlikely to yield
accurate information. Rather than direct
infor-
mation,
we use
annual information
on the
numbers
of
cases
of child maltreatment han-
dled
by
each state's child-protective service
agency. However,
state-level data have several
limitations that
one
should
keep in mind when
assessing
our results.
One
general
limitation
of
using
state-level
data
is
that one cannot
generalize
from
state-
level results to individual-level
behaviors (this
is
the "ecological fallacy"
problem).
If we
find, for instance, that states
with higher shares
of
unemployed
fathers have higher rates
of
maltreatment, we
cannot
conclude from
this
that
unemployed
fathers are more
likely
to
abuse
their children. This would be one
pos-
sible
explanation
for our
finding,
but not
the
only
one.
A more
specific problem
is
that
the state-
level
data
may
not
accurately
measure
the ac-
tual amount of child maltreatment. Not
all
cases of
maltreatment
are
reported,
and
some
reports
are not valid. The
process
of substan-
tiating reports may
also be
prone
to errors:
agencies may incorrectly
substantiate invalid
reports,
or not substantiate valid
reports.
Al-
though
in
theory
the
true
level of child mal-
treatment could be
greater
or less
than what
state
numbers
indicate,
the
general
consensus
among
scholars in this field is that
many
cases
of child maltreatment
go
unreported
and un-
substantiated
(Waldfogel,
1998).
Another
potential problem
is variation
across states
in
how
reports
of
child abuse
are
handled,
and
in how data
on
child abuse
are
reported.
The
operations
of state
child-
protective
service
agencies
vary along
several
dimensions.
First, reporting
requirements vary
across states. For
example,
states
differ
in
how
they
define maltreatment and
in
how
they
de-
fine "mandated
reporters"
(i.e.,
those
who
have a
legal responsibility
to
report suspected
maltreatment). Second,
states
have different
standards of evidence required to substantiate
a report
of child
maltreatment
as well as
dif
ferent classifications for substantiation deci
sions. Some states use
a
two-tier
system,
in
which
each report
is determined to be either
"6substantiated"
or
"unsubstantiated."
Other
states use a three-tier system which adds the
category "indicated," meaning that, although
there is good reason to suspect that maltreat-
ment has
occurred,
the
allegation
cannot
be
substantiated to the level of evidence required
by
state law.
Third, states
collect and
report
information on
child
maltreatment
in
different
ways.
For
example,
some
states
collect infor-
mation
on the number of
families reported
for
child maltreatment, and some collect
infor-
mation on the number of children reported to
be
victims of abuse or
neglect.
For all of
these
reasons,
it is
important that
our
analysis ade-
quately
accounts for
heterogeneity
across
states.
The quality of state-level data has improved
over time. Between 1976 and 1987, state-level
information on reports of child abuse was col-
lected by the Child Protection Division of the
American
Humane
Association. After
1987,
this information was collected
by
the National
Committee to Prevent Child Abuse.
However,
information on the number
of
substantiated
or
indicated victims was not uniformly collected
by
either
organization.
In
1988,
the National
Center
for Child Abuse and
Neglect
(NCCAN),
the federal
agency responsible
for
assisting
states
in
the
prevention
and treatment of child
maltreatment,
was
charged
with
establishing
a
national
clearinghouse
for
information
on
child maltreatment. Since 1990 NCCAN has
collected and
published
detailed
state-level
in-
formation
on
reports
of child maltreatment and
on numbers of substantiated and indicated
vic-
tims. Victims are classified
by type
of mal-
treatment
and
by
a
variety
of other variables
such as
age
and sex.
In this
paper
we use
only
the
1990-1996
NCCAN data.
Although
the
sample
is
smaller,
there are several
advantages
of
focusing
on
these
years. First, reports
of
maltreatment
are
more
likely
to be
accurately
measured after
NCCAN assumed control
of
data
collection.
Second,
we
can
distinguish
between
reports
of
maltreatment and the actual number
of sub-
stantiated
victims,
and we can
disaggregate

VOL.
89 NO. 2 CHILD
WELFARE
241
different
types
of maltreatment.
Third,
the
large
increase in reports
that characterized
the
1970's
and 1980's
was essentially
completed
by 1990.
Variations
in
the
measures of mal-
treatment
in the 1990's are
more
likely
to re-
flect actual
changes
in the
welfare
of children,
rather than
changes
in
the attention
that
states
focused
on the problem
of child maltreatment.
We
use four measures
of child maltreat-
ment.
The
first is the number
of
reports.
For
most states,
reports
are measured
as the num-
ber
of families
reported,
so
that
a
family
charged
with neglecting
three children
counts
as one report. For
some states
in
some years,
reports
are on a
child basis. We use informa-
tion
on
the average
ratio of child-based
to
family-based
reports
(available
for
a
large
subset
of
observations)
to convert
child-based
reports
to family-based
reports
when family-
based
reports
are unavailable.
This conversion
applied
to
20
percent
of our
sample.
Our second
measure
is
the
total number of
indicated
or substantiated
victims. Since
not
all
reports
are
valid,
the number of
indicated
or
substantiated
victims
may
provide
a
better
measure of maltreatment,
although
this
mea-
sure
may
also be affected by
a
state's willing-
ness to add
new cases to
the
caseload. Our
third and fourth
measures are
the
numbers of
children who were
victims of
physical
abuse
and
neglect.
In
principle,
family
socioeco-
nomic
circumstances
may have
different
ef-
fects
on
these
two
types
of
maltreatment.
In
practice,
since
some states allow children to be
coded
as victims
of both abuse
and neglect,
there
is
some
overlap
in
these
two categories
and
in
their models and some
double-counting
of children
in
the
victim
totals.
The
socioeconomic
variables were
con-
structed from the 1990-1996
rounds
of
the
March Current
Population
Survey (CPS).
These
variables reflect the living conditions
of
children
within each state and each
year,
rather
than the
living
conditions of the entire
popu-
lation.
For each
year,
we
selected records
for
all children
under the
age
of
18,
constructed
socioeconomic
variables
for
each
child,
and
then
computed
estimates of state-level
statis-
tics
(averages,
medians, etc.)
across
children,
using
the
appropriate
individual-level
survey
weights.
The state-level statistics
include the
median of the
logarithm
of
the child's
house-
hold per capita income, the fraction of children
living in urban areas, the fraction of children
who are white, black, or of another race, the
fraction
of
children
with
an employed mother,
the fraction with a nonworking
father,
and the
fraction with no father in the household. (We
restricted our sample
to
children with
a
mother
in the household; see the following discussion
of how "mother" is defined.) We also con-
structed more detailed measures of parental
presence
and work status. These show
the
frac-
tion
of children in each of six
categories,
which
represent
all the combinations
of
the
mother's
work status
with the father's
status:
two
working parents (48.1 percent
of
our sam-
ple), nonworking mother and working father
(23.5 percent), working
mother
and
no
father
( 10.9 percent), nonworking
mother
and
no
fa-
ther (7.9 percent), working
mother and
non-
working
father
(5.1
percent),
and two
nonworking parents (4.5 percent).
Our
definitions of
"mother"
and "father"
require
discussion.
Ideally,
we would
like to
distinguish between children who
live
with
bio-
logical and nonbiological parents. This is
potentially important, given
evidence that
non-
biological parents (stepparents
or cohabitants
of
the
parent)
are more
likely
to
abuse children
(Martini Daly
and
Margo Wilson, 1996).
How-
ever,
the
CPS
does not
permit
such fine
dis-
tinctions.
For our
purposes,
"father"
is
broadly
defined to
include biological fathers (who may
or
may
not be married to the
mother), stepfa-
thers, adoptive fathers,
and live-in
boyfriends
who
are not
biological
fathers.
Likewise,
"mothers"
can be
biological mothers, step-
mothers,
or
adoptive
mothers.
The state-level
aggregates
from
the
CPS
were
merged
with
the state-level
information
from
NCCAN
on
child maltreatment, and
in-
formation
on
the
number of
children
in each
state
was
obtained
from
Bureau
of
the
Census
publications.
Our final data set
consists
of
320
observations from
1990-1996, representing
all states
except
West
Virginia
and
Maryland
(which
did
not
report
all variables to
NCCAN), plus
the District of
Columbia,
for
an
average
of
6.5
years per
state.
Our
task is to estimate the
relationships
be-
tween
the
socioeconomic variables
and the
measures of
child
maltreatment.
One of our
primary
concerns is that the
state-level
socio-

242
AEA PAPERS
AND
PROCEEDINGS
MAY 1999
economic
variables
may
be
correlated
with
unobserved
factors
that
also influence
child
maltreatment. An
obvious
problem
is that
the
upward
trend
in
reports of
child maltreatment
may
be
spuriously
correlated
with
trends in
other
variables, such
as
the
rate of labor-force
participation
of
mothers and
the general
increase in income
levels. To the
extent that
these
trends
are
common
across
states,
this
problem
is easily
handled
with the
addition
of
a set of
year
effects.
A
more
serious prob-
lem is
that there
may be other
factors
that af-
fect child
maltreatment, such as the resources
put into
a state's
child
protective-service
agency or the cost of child
care
in
a
state, that
are
unobserved but correlated with
other ob-
served
variables,
resulting
in
biased
parameter
estimates. To the
extent that
these factors
are
fixed over time within
states,
this
problem can
be remedied
through
the
introduction of a set
of
state-specific
fixed effects.
However,
the
use
of
fixed effects
comes at a
cost,
in
that
it
removes
the
cross-state
variation in the
data
that can
help us
identify
the
effects in which
we are
interested,
and it will
also exacerbate
attenuation bias due to
measurement error
in
the
independent
variables.
Attenuation bias is
especially likely to be a
problem,
since
the
independent
variables con-
structed from
the
CPS are estimates of state-
level
variables,
rather than
their true
values,
and thus are subject
to
sampling
error. Fortu-
nately,
this form of
measurement error
can be
corrected.
As
shown in
Angus
Deaton
(1985),
the
bias
in
the
parameter
estimates
is a func-
tion of
the variances
and covariances
of
the
state-level
means constructed from the
CPS.
These
variances
and
covariances
can be esti-
mated
from the micro data and
used to
adjust
the
parameter
estimates for
bias. We follow
the
methods
described
in
Deaton
(
1985),
treating
all variables
constructed
from the
CPS,
with
the
exception
of the median of
the
logarithm
of
income,
as
noisy estimates of
true
state values. This
correction
has
a
substantial
effect on the
parameter
estimates:
many
of
the
ordinary
least-squares
(OLS)
coefficients
in-
creased
(in
absolute
value)
by
25
percent;
the
fixed-effects estimates often
doubled.
Tables
1
and
2
present
regression
results.
The
dependent variables
in
both
tables are the
logarithms
of the
four
measures
of
child mal-
TABLE
1-DETERMINANTS OF CHILD
MALTREATMENT
A.
Independent
ln(Reports)
ln(Victims)
variable No FE FE
No
FE
F-E
ln(Kids
<
18)
0.951 0.322 0.933
2.382
(38.21)
(1.16) (21.85)
(4.02)
ln(Median income -0.521 0.182
-0.610 0.347
per
capita)
(3.46)
(1.43) (2.36)
(1.28)
Urban 0.237
-0.119 0.373
-0.141
(1.78)
(1.84) (1.64)
(1.02)
Black
-0.818
0.451
-0.901 -1.347
(2.60)
(0.66) (1.67)
(0.91)
Other race -1.470 -0.741
-1.005 -3.077
(4.83)
(0.83) (1.92)
(1.61)
Working
mom
0.801 0.019 1.639 0.088
(1.64)
(0.05) (1.96)
(0.11)
No dad
1.949 1.214
4.388
2.789
(2.65)
(2.40) (3.47)
(2.57)
Nonworking
dad 3.064
1.923
5.849
2.928
(3.87)
(3.08) (4.31)
(2.21)
B.
ln(Victims of
ln(Victims of
Independent
physical
abuse)
neglect)
variable No
Fe FE No
FE FE
ln(Kids
<
18) 0.934 3.293
0.896 4.005
(24.27)
(5.73) (14.89)
(4.57)
ln(Median
income
-0.500
0.606 -0.663
-0.200
per
capita)
(2.15)
(2.30) (1.82)
(0.50)
Urban
0.160 -0.024
0.159 0.045
(0.78)
(0.18) (0.49)
(0.22)
Black -1.741
1.403 -0.634
-1.624
(3.59)
(0.98) (0.84)
(0.74)
Other race -1.156
-4.142 -2.292
-3.078
(2.45)
(2.23) (3.11)
(1.09)
Working
mom 0.617
-0.358 3.329
0.986
(0.82)
(0.47) (2.82)
(0.85)
No
dad 3.230
2.332 5.811
1.963
(2.84)
(2.23) (3.26)
(1.24)
Nonworking
dad 5.809 1.979
7.278 1.853
(4.75)
(1.55) (3.81)
(0.96)
Notes:
Year
dummies included in all
regressions.
Absolute
t-
statistics
are
reported
in
parentheses.
The
columns labeled "no
FE"
do not include state fixed
effects,
and the
columns labeled
"FE"
do.
treatment. We show
specifications
with two
sets of
independent
variables. Both
include a
set
of
year
dummies,
the
logarithm
of
the
num-
ber of children
in
the
state,
the fraction of chil-
dren who
are
urban, black,
or other
nonwhite
race,
and the
logarithm
of
median household
income
per capita.
The
specifications
differ in
the level of
detail of the
measures of the home
environment. The
first,
in
Table
1,
includes the
fraction of
children with
mothers who
work,
the fraction of
children with
absent
fathers,

VOL. 89 NO.
2
CHILD
WELFARE 243
TABLE
2-DETAILED DETERMINANTS
OF
CHILD
MALTREATMENT
A.
Independent
ln(Reports)
ln(Victims)
variable No
FE FE No
FE FE
ln(Median
income -0.488
0.173 -0.513 0.459
per capita)
(3.09)
(1.31)
(1.89) (1.55)
Nonworking
mom,
0.315
1.161
2.804 0.807
no
dad (0.32)
(1.74)
(1.66) (0.54)
Mom
works,
no 4.525
1.170
8.108
4.259
dad
(3.19)
(1.64) (3.31)
(2.63)
Nonworking
mom, 2.078
1.367 8.222 6.344
nonworking
dad
(1.13)
(1.29)
(2.59) (2.58)
Mom
works,
4.247 2.287 5.063
0.142
nonworking dad (2.25)
(2.36)
(1.56) (0.07)
Mom works,
dad 0.058
-0.115 1.414 -0.420
works (0.10)
(0.25)
(1.49) (0.41)
B.
ln(Victims
of
ln(Victims of
Independent
physical
abuse)
neglect)
variable No
FE
FE
No FE
FE
ln(Median
income -0.386
0.753 -0.472 -0.041
per capita)
(1.58)
(2.46)
(1.23) (0.09)
Nonworking
mom,
2.057
-0.707 2.993 -1.285
no dad (1.35)
(0.45)
(1.25) (0.58)
Mom
works,
no
5.679 4.012
12.953 5.077
dad
(2.58)
(2.42)
(3.75) (2.16)
Nonworking
mom,
9.641 6.032
12.380
6.507
nonworking
dad
(3.37)
(2.34)
(2.77) (1.84)
Mom
works,
2.844 -2.066
5.380
-1.341
nonworking
dad
(0.97)
(0.92)
(1.17) (0.42)
Mom
works,
dad 0.751 -1.394
3.056 -0.018
works
(0.88)
(1.31)
(2.28) (0.01)
Notes: See
notes
to Table 1. Year dummies,
the
logarithm
of chil-
dren in the
state,
and
the fractions of children who are
urban, black,
and other
(nonwhite)
races
are
included in all
regressions.
and
the
fraction of children
with
present
but
nonworking fathers.
The
second,
in Table 2,
includes
the more detailed
set
of measures of
the mother's and father's
status.
(The
coef-
ficients for the
population,
urbanization,
and
race variables
are not
reported
in
Table 2.
They
are very
similar
to those shown in
Table
1.)
The main lesson to be drawn
from Tables
1
and
2 is that
parental
work status and single-
parenthood
affect
child maltreatment. The re-
sults
in
Table
1
indicate
that
higher
fractions
of children
living
with
working
mothers,
no
fathers,
or
nonworking
fathers are associated
with more
reports
of
maltreatment,
more sub-
stantiated
reports,
and
more cases of
physical
abuse
and
neglect.
The effect of
working
mothers on
maltreatment
becomes insignifi-
cant when fixed effects
are included, but the
effects of the status of
fathers stay significant
and large.
For
example,
the fixed effects im-
ply that, all else equal,
if the fraction of chil-
dren
living
with no
father were
to increase
from 0.10 (the average
value for Utah
in
1990-1996) to 0.30 (the
average for Missis-
sippi), the number of
victims of
child
mal-
treatment would rise
by 56 percent. The
effects of more children
with nonworking
fa-
thers are
extremely
similar to the effects of
more
children
with absent fathers.
It
is inter-
esting
that
the
fixed-effects
estimates of fa-
ther's
status
are
significant
for the total
number of victims and
the number of victims
of
physical abuse,
but insignificant for neglect.
The results in Table
2 provide a more de-
tailed
picture
of
the
relationship between
home
environment and
maltreatment,
and
they high-
light the idea
that there are interactions be-
tween
the mother's and father's status.
There
are several
key
results.
First,
the adverse
effect
of
absent fathers, seen
in
Table 1,
is
only pres-
ent
when
the
mother
works.
The
fraction of
children with nonworking
mothers
and
absent
fathers has
no effect on maltreatment
(relative
to the omitted
category
of
children with a
non-
working
mother and
a
working father).
How-
ever,
the combination
of
absent fathers
and
working
mothers
appears
to
be
particularly
harmful,
with
large
and significant effects
on
total maltreatment as
well as both physical
abuse and
neglect.
Second,
the adverse effects
of
two
nonworking parents
are
large
and
sig-
nificant, although the effects
of working
moth-
ers and
nonworking
fathers are
generally
insignificant. Third,
higher fractions
of
two
working parents
do not
appear
to be harmful.
In
general, higher
fractions
of
children
with
two
parents
in
the
household,
at least one
of
whom
is
working,
result
in
less maltreatment.
In
alternative
models
not shown
here,
we
added
controls
for the number of women
and
men arrested
for
drug
use
by
state and
year,
using
data from the
FBI's
UJniform
Crime Re-
ports.
We
found,
as have other researchers
(see e.g., Vicky
Albert and
Richard
Barth,
1996;
Sara
Markowitz and Michael
Grossman,
1998),
that there
is
a
positive relationship
between substance
abuse and child maltreat-
ment,
but this
relationship
did not hold
up
in

Citations
More filters
Journal ArticleDOI
TL;DR: Results indicate that adults with documented histories of childhood abuse and/or neglect have lower levels of education, employment, earnings, and fewer assets as adults, compared to matched control children.
Abstract: Child abuse and neglect represent major threats to child health and well-being; however, little is known about consequences for adult economic outcomes. Using a prospective cohort design, court substantiated cases of childhood physical and sexual abuse and neglect during 1967-1971 were matched with nonabused and nonneglected children and followed into adulthood (mean age 41). Outcome measures of economic status and productivity were assessed in 2003-2004 (N 1/4 807). Results indicate that adults with documented histories of childhood abuse and/or neglect have lower levels of education, employment, earnings, and fewer assets as adults, compared to matched control children. There is a 14% gap between individuals with histories of abuse/neglect and controls in the probability of employment in middle age, controlling for background characteristics. Maltreatment appears to affect men and women differently, with larger effects for women than men. These new findings demonstrate that abused and neglected children experience large and enduring economic consequences.

503 citations


Cites background or result from "Parental Resources and Child Abuse ..."

  • ...Consistent with our results, Paxson and Waldfogel (1999, 2002) showed that abuse and neglect are more common in families of lower socioeconomic status, so that maltreatment may exacerbate differences in the prospects of rich and poor children....

    [...]

  • ...…it is theoretically plausible that any relationship between childhood events and subsequent outcomes is confounded with or explained by social class differences (Case, Fertig, & Paxson, 2005; Currie & Hyson, 1999; Heckman, 2007; MacMillan & Munn, 2001; Paxson & Waldfogel, 1999; Widom, 1989c)....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the effects of income, family structure, and public policies on several indicators of child maltreatment were explored using data from the National Longitudinal Survey of Youth (NLSS).

298 citations

Journal ArticleDOI
TL;DR: Findings reinforce earlier findings that demographic characteristics, maternal depression, maternal alcohol use, and intra-family patterns of violence may largely contribute to child abuse and suggest that income may play a substantially more important role in regard to parental violence in single- parent families than in two-parent families.

288 citations

Journal ArticleDOI
TL;DR: This paper examined how child maltreatment is affected by parental economic circumstances and found that increases in the fractions of children with absent fathers and working mothers are related to increases in many measures of maltreatment, as are increases in families with two nonworking parents and those with incomes below 75% of the poverty line.
Abstract: We examine how child maltreatment—including neglect, physical and sexual abuse, and other forms of maltreatment—is affected by parental economic circumstances. Using state‐level panel data on cases of maltreatment and numbers of children in foster care, we find that increases in the fractions of children with absent fathers and working mothers in a state are related to increases in many measures of maltreatment, as are increases in the share of families with two nonworking parents and those with incomes below 75% of the poverty line. Decreases in state welfare benefit levels are associated with increases in foster care placement.

210 citations


Cites background from "Parental Resources and Child Abuse ..."

  • ...In an earlier paper (Paxson and Waldfogel 1999), we presented preliminary evidence that fluctuations in these socioeconomic factors are associated with large changes in rates of child maltreatment....

    [...]

Journal ArticleDOI
TL;DR: In this paper, the authors examined the impact of welfare reforms on several measures of child maltreatment, including the incidence of reported and substantiated cases of maltreatment and the number of children living in out-of-home care.
Abstract: This paper examines the impact of welfare reforms on several measures of child maltreatment. The authors use state-level data from 1990 to 1998 to examine whether recent welfare reforms have increased or reduced the incidence of reported and substantiated cases of maltreatment, the incidence of specific types of substantiated maltreatment—physical abuse and neglect—and the number of children living in out-of-home care. The welfare reforms considered are the imposition of: family caps, lifetime limits, work requirements, sanctions for non-compliance, and the restriction of welfare benefits to immigrants. How welfare benefit levels and changes in state Earned Income Tax Credit programs affect reports and substantiated cases of maltreatment are also considered. Evidence strongly indicates that reductions in states' welfare benefit levels increase the number of children in out-of-home care, and some evidence indicates that strict lifetime welfare limits and tougher sanctions for noncompliance are related to higher levels of substantiated maltreatment. The evidence on family caps is mixed: family caps appear to be associated with fewer instances of substantiated maltreatment, but more children in out-of-home care. Because most of the welfare reforms examined have been in effect for only a short time, these results should be considered preliminary. Overall, however, they provide some evidence that the recent welfare reforms in the United States may have increased child maltreatment. © 2003 by the Association for Public Policy and Analysis and Management.

185 citations


Cites background from "Parental Resources and Child Abuse ..."

  • ...mothers who are working, but not when more children are living with single mothers who are not working ( Paxson and Waldfogel, 1999, 2002 )....

    [...]

  • ...In parallel, states whose rates of child poverty increase, experience increases in reported and substantiated cases of maltreatment ( Paxson and Waldfogel, 1999, 2002 )....

    [...]

References
More filters
Journal ArticleDOI
14 Apr 1989-Science
TL;DR: Findings from a cohort study show that being abused or neglected as a child increases one's risk for delinquency, adult criminal behavior, and violent criminal behavior; however, the majority of abused and neglected children do not become delinquent, criminal, or violent.
Abstract: Despite widespread belief that violence begets violence, methodological problems substantially restrict knowledge of the long-term consequences of childhood victimization. Empirical evidence for this cycle of violence has been examined. Findings from a cohort study show that being abused or neglected as a child increases one's risk for delinquency, adult criminal behavior, and violent criminal behavior. However, the majority of abused and neglected children do not become delinquent, criminal, or violent. Caveats in interpreting these findings and their implications are discussed in this article.

1,897 citations

Journal ArticleDOI
TL;DR: In this paper, the authors consider the possibility of tracking "cohorts" through household surveys, defined as a group with fixed membership, individuals of which can be identified as they show up in the surveys.

1,364 citations

Journal ArticleDOI
TL;DR: Using census and administrative agency data for 177 urban census tracts, variation in rates of officially reported child maltreatment is found to be related to structural determinants of community social organization: economic and family resources, residential instability, household and age structure, and geographic proximity of neighborhoods to concentrated poverty.
Abstract: Using census and administrative agency data for 177 urban census tracts, variation in rates of officially reported child maltreatment is found to be related to structural determinants of community social organization: economic and family resources, residential instability, household and age structure, and geographic proximity of neighborhoods to concentrated poverty. Furthermore, child maltreatment rates are found to be intercorrelated with other indicators of the breakdown of community social control and organization. These other indicators are similarly affected by the structural dimensions of neighborhood context. Children who live in neighborhoods that are characterized by poverty, excessive numbers of children per adult resident, population turnover, and the concentration of female-headed families are at highest risk of maltreatment. This analysis suggests that child maltreatment is but one manifestation of community social organization and that its occurrence is related to some of the same underlying macro-social conditions that foster other urban problems.

602 citations

Book
22 Nov 1989
TL;DR: Theoretical Approaches from the study of work, employment, and unemployment have been studied in this paper, including single-group studies, comparison group studies, and studies with older-age groups.
Abstract: Contents: Introduction and Overview.- Background Research.- Theoretical Approaches from the Study of Work, Employment, and Unemployment.- Theoretical Approaches: Some Wider Frameworks.- Scales and Measures.- Youth Unemployment: Single-Group Studies.- Youth Unemployment: Comparison-Group Studies.- Youth Unemployment: Longitudinal Studies.- Studies with Older-Age Groups.- Epilogue.- References.- Author Index.- Subject Index.

381 citations

Book
01 Jan 1970
TL;DR: In this paper, the authors developed a definition and conceptual model of violence against children on the basis of a series of nationwide epidemiologic studies, public opinion, and press surveys.
Abstract: This paper develops a definition and conceptual model of violence against children on the basis of a series of nationwide epidemiologic studies, public opinion, and press surveys. Culturally sanctioned use of physical force in child rearing, poverty and discrimination, deviance in bio-psycho-social functioning, and chance events are identified as causal dimensions of physical child abuse. The scope of the phenomenon and selected findings from the surveys are discussed and social policies aimed at primary, secondary, and tertiary prevention are suggested. Attention is drawn to massive societal abuse of children, which is a related but much more serious social problem.

341 citations