Page
1
of
4
October
24,
2005
 
8:
00
AM
ET
Estimates
of
Concentration
of
Lead
in
Paint
by
Age
of
Housing
Estimates
of
the
concentration
of
lead
in
paint
by
the
age
of
housing
are
presented
in
Table
1.
Specifically
Table
1
presents
estimates
of
1)
average
lead
concentration
in
those
painted
surfaces
that
contain
lead­
based
paint,
These
average
lead
concentrations
represent
housing
categories
specified
in
the
column
headings
(
i.
e.,
housing
with
LBP
and
children
under
age
6,
within
the
specified
age
category)
and
the
specific
building
component
categories
specified
in
the
row
headings.
These
averages
are
accompanied
by
estimates
of
their
standard
errors.
and
2)
estimates
of
the
number
of
housing
units
in
the
U.
S.
housing
stock
within
specified
categories
as
noted
in
the
row
and
column
headings
of
the
table.
In
particular,
numbers
of
units
are
presented
within
three
age
categories
of
housing:
°
Housing
built
prior
to
1950
°
Housing
built
from
1950
to
1960
°
Housing
built
from
1961
to
1978
The
housing
categories
included
in
Table
1
focus
on
the
presence
of
lead­
based
paint
(
LBP)
on
specific
types
of
painted
surfaces
in
the
interior
and
exterior
of
the
housing
unit
that
could
be
disturbed
by
renovation
and
remodeling
(
R&
R)
activities
in
the
unit.
Therefore,
housing
built
after
1978
were
not
represented
among
the
age
categories
as
the
Federal
government
banned
use
of
lead­
based
paint
in
housing
after
1978.
The
housing
categories
also
focus
on
the
presence
of
children
aged
6
years
or
younger
residing
in
the
unit,
as
this
age
group
is
highly
vulnerable
to
experiencing
immediate
and
lasting
adverse
health
effects
from
exposure
to
lead­
based
paint
hazards.

The
source
of
data
represented
in
Table
1
is
the
National
Survey
of
Lead
and
Allergens
in
Housing
(
NSLAH;
Jacobs
et
al.,
2002;
Westat,
2001),
a
survey
sponsored
by
the
U.
S.
Department
of
Housing
and
Urban
Development
(
HUD)
and
the
National
Institute
of
Environmental
Health
Sciences
(
NIEHS).
The
NSLAH
is
the
most
recent
nationallyrepresentative
survey
on
the
extent
of
lead­
based
paint
in
target
housing,
with
the
main
field
component
of
the
survey
being
conducted
in
1998
and
1999.
The
target
population
in
this
survey
consisted
of
all
permanently­
occupied,
noninstitutional
housing
units
in
the
U.
S.
housing
stock,
including
single­
and
multi­
family
buildings,
that
had
the
potential
for
housing
young
children.
This
corresponded
to
approximately
96
million
units.
Within
the
831
units
selected
for
the
survey,
x­
ray
fluorescence
(
XRF)
technology
was
used
to
measure
the
concentration
of
lead
on
each
painted
surface
in
each
of
four
randomly­
selected
rooms,
with
one
room
selected
within
each
of
four
room
types:
kitchen,
common
living
areas,
bedrooms
(
preferably
those
containing
children),
and
other
rooms..
Lead­
based
paint
was
deemed
present
on
a
given
painted
surface
if
the
XRF
measure
taken
on
the
surface
was
1.0
mg/
cm2
or
higher.

The
NSLAH
featured
a
complex
probabilistic
survey
design,
where
eligible
units
were
allowed
to
differ
in
their
selection
probabilities.
As
a
result,
survey
weights
were
assigned
to
each
of
the
831
selected
housing
units.
Proper
use
of
these
weights
allows
data
collected
at
the
Page
2
of
4
October
24,
2005
 
8:
00
AM
ET
housing
unit
level
to
be
used
in
calculating
representative,
unbiased
estimates
of
unknown
parameters
within
the
survey's
target
population.
In
particular,
the
numbers
of
housing
units
Table
1.
Number
of
U.
S.
Housing
Units
in
Specified
Categories
by
Age,
Presence
of
Lead­
Based
Paint
(
LBP),
and
Presence
of
Children
Under
Age
6
Years,
and
Average
Lead
Content
(
mg/
cm2)
of
Surfaces
Containing
LBP
by
Component
Type
Source
of
data:
National
Survey
of
Lead
and
Allergens
in
Housing
Pre
1950
housing
1950
­
1960
housing
1961­
1978
housing
Total
#
of
units
24,561
14,377
29,200
#
of
units
with
LBP
20,302
9,271
6,464
#
of
units
w/
one
or
more
children
under
age
6
years
3,186
2,730
4,640
Pre
1950
housing
units
with
LBP
&
children
under
6
1950­
1960
housing
units
with
LBP
&
children
under
6
1961­
1978
housing
units
with
LBP
&
children
under
6
Total
#
of
units
2,916
1,523
688
#
of
rental
units
912
418
398
#
of
owner
occupied
units
2,004
1,105
290
#
of
units
w/
interior
LBP
2,524
1,144
464
#
of
units
w/
exterior
LBP
2,550
1,315
375
#
of
units
w/
LBP
on
interior
walls,
ceiling
or
floors
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,438
5.96
mg/
cm2
(
1.14)
485
4.28
mg/
cm2
(
0.85)
151
1.19
mg/
cm2
(
0.03)

#
of
units
w/
LBP
on
exterior
walls
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,280
7.47
mg/
cm2
(
2.64)
955
2.57
mg/
cm2
(
0.39)
202
10.34
mg/
cm2
(
6.86)

#
of
units
w/
LBP
on
interior
windows
 
Average
(
S.
E.)
lead
conc.
in
LBP
2,030
13.04
mg/
cm2
(
3.98)
644
1.63
mg/
cm2
(
0.23)
151
3.07
mg/
cm2
(
2.65)

#
of
units
w/
LBP
on
exterior
windows
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,615
14.18
mg/
cm2
(
3.45)
857
3.72
mg/
cm2
(
0.98)
296
5.11
mg/
cm2
(
2.81)

#
of
units
w/
LBP
on
interior
doors
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,784
9.66
mg/
cm2
(
5.39)
949
6.09
mg/
cm2
(
3.23)
250
1.66
mg/
cm2
(
0.65)

#
of
units
w/
LBP
on
exterior
doors
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,499
7.70
mg/
cm2
(
1.99)
1,107
6.41
mg/
cm2
(
1.91)
161
9.45
mg/
cm2
(
4.40)

#
of
units
w/
LBP
on
interior
trim
 
Average
(
S.
E.)
lead
conc.
in
LBP
1,449
11.08
mg/
cm2
(
2.76)
295
6.11
mg/
cm2
(
5.70)
32
2.46
mg/
cm2
(
0.01)

#
of
units
w/
LBP
on
exterior
trim
 
Average
(
S.
E.)
lead
conc.
in
LBP
634
22.60
mg/
cm2
(
7.49)
316
4.79
mg/
cm2
(
3.06)
120
8.71
mg/
cm2
(
6.32)

Notes:
1.
All
estimates
on
numbers
of
units
presented
in
this
table
are
specified
in
thousands
and
take
into
account
the
survey
design
and
the
sample
weights
assigned
to
each
surveyed
unit.
Data
for
only
those
surveyed
housing
units
that
could
be
successfully
placed
in
the
housing
category
specified
by
the
row
and
column
headings
were
considered
in
calculating
a
particular
estimate.
2.
The
last
eight
rows
of
this
table
contain
numbers
of
units
(
in
thousands),
followed
by
a
weighted
average
of
reported
XRF
measurements
across
all
painted
surfaces
for
the
component
type
specified
in
the
row
heading,
where
only
XRF
measurements
at
or
above
1.0
mg/
cm2
were
included
in
the
average.
Each
measurement
was
weighted
by
the
full­
survey
weight
assigned
at
the
building
component
level.
The
standard
error
(
S.
E.)
of
this
Page
3
of
4
October
24,
2005
 
8:
00
AM
ET
weighted
average
is
given
in
parentheses
and
also
takes
into
account
the
survey
design.
Only
XRF
measurements
taken
within
surveyed
housing
units
that
could
be
successfully
placed
in
the
housing
category
specified
by
the
row
and
column
headings
were
included
in
these
calculations.

specified
in
each
cell
of
Table
1
are
sums
of
the
assigned
full­
survey
weights
across
all
surveyed
units
that
fall
within
the
housing
categories
represented
by
that
cell
(
i.
e.,
as
specified
in
the
row
and
column
headings).
Thus,
these
numbers
are
estimates
of
the
total
number
of
units
within
the
target
population
that
fall
within
the
specified
housing
category
at
approximately
the
time
at
which
the
survey
was
conducted
(
1998­
99).

In
addition
to
survey
weights
assigned
at
the
housing
unit
level,
the
NSLAH
also
assigned
a
second
set
of
full­
survey
weights
at
the
room/
component
level.
These
weights
are
meant
to
be
used
when
analyzing
data
at
the
room/
component
level
in
order
to
obtain
results
that
are
representative
of
all
such
rooms
or
building
components
within
the
survey's
target
population.
Therefore,
those
XRF
measures
encountered
in
the
NSLAH
database
that
were
at
or
above
1.0
mg/
cm2
were
included
in
the
average
lead
measures
presented
in
the
last
eight
rows
of
Table
1,
and
each
measure
was
weighted
by
the
component­
specific
survey
weight
assigned
to
the
given
building
component
within
the
NSLAH
database
when
calculating
these
averages.
In
addition,
each
average
is
accompanied
by
its
estimated
standard
error,
which
was
calculated
using
the
set
of
99
jackknife
replicate
survey
weights
assigned
to
each
room
and
surface
in
the
NSLAH
database.

SAS
®
Release
9
(
Version
9.1)
was
used
to
calculate
the
numbers
of
units
and
average
paint­
lead
levels
presented
in
Table
1,
while
a
SUDAAN
®
Release
9
software
module,
called
within
SAS
®
,
was
used
to
generate
standard
errors
on
the
average
paint­
lead
levels.
Note
that
the
NSLAH
database
is
accompanied
by
only
limited
guidance,
and
no
sample
programs,
on
proper
application
of
survey
weights
in
calculating
summary
statistics
for
subsets
of
the
survey's
target
population.
As
a
result,
the
accuracy
of
the
summary
statistics
presented
in
Table
1
depend
in
part
on
the
approach
used
to
incorporate
survey
weights
into
the
calculations.

The
most
recognized
source
of
nationally
representative
data
on
the
U.
S.
housing
stock
is
the
American
Housing
Survey
(
AHS),
a
periodic
national
survey
conducted
by
the
Bureau
of
the
Census
for
HUD.
The
2003
AHS
is
the
most
recent
nationally­
representative
source
of
household
data
available
on
the
U.
S.
housing
stock,
involving
approximately
55,000
sampled
housing
units.
The
target
population
for
the
AHS
is
the
entire
U.
S.
housing
stock,
including
vacant
housing
and
housing
that
may
not
necessarily
have
the
potential
for
containing
young
children.
While
the
broader
target
population
slightly
limits
the
usefulness
of
the
AHS
to
estimate
statistics
needed
for
Table
1,
its
primary
limitation
is
the
lack
of
information
on
the
lead
content
of
painted
surfaces
or
on
the
presence
of
lead­
based
paint
in
surveyed
housing.
However,
AHS
data
could
be
used
to
estimate
statistics
by
age
category
on
the
total
number
of
units
and
the
number
of
units
with
children
under
age
6
years,
which
are
specified
in
the
first
and
third
rows
of
Table
1.
Table
2
lists
estimates
of
these
statistics
based
on
2003
AHS
data
for
housing
units
that
are
meant
to
be
occupied
year­
round.
Page
4
of
4
October
24,
2005
 
8:
00
AM
ET
Estimated
numbers
of
housing
units
differ
between
Table
1
and
Table
2
for
a
number
of
reasons.
The
primary
reasons
include
differences
between
the
AHS
and
NSLAH
in
their
target
populations,
housing
category
definitions,
survey
methods,
and
time
periods
represented.
For
example,
the
AHS
considers
all
housing
units
within
its
target
population,
including
vacant
units,
while
the
NSLAH
focused
only
on
occupied
housing
units
with
the
potential
for
containing
children.
Also,
note
that
the
column
headings
in
Table
2
differ
slightly
from
those
in
Table
1,
as
the
AHS
recorded
the
year
in
which
a
housing
unit
was
built
only
within
pre­
specified
categories.
This
may
contribute
in
part
to
lower
estimates
for
the
middle
age
category
(
i.
e.,
housing
built
from
1950
to
1959/
1960)
in
Table
2
versus
Table
1,
while
estimates
in
the
other
two
age
categories
are
higher
in
Table
2
versus
Table
1.

Table
2.
Estimated
Number
of
U.
S.
Housing
Units
That
Could
Be
Occupied
Year­
Round,
by
Age
Category
and
Presence
of
Children
Under
Age
6
Years
Source
of
data:
2003
American
Housing
Survey
(
national
sample)

Pre
1950
housing
1950
­
1959
housing
1960­
1979
housing
Total
#
of
units
Occupied
units
only
25,584
11,933
34,255
%
difference
from
Table
1
+
4.2%
­
17.0%
+
17.3%

Occupied
+
vacant
units
28,928
13,033
37,668
#
of
units
w/
one
or
more
children
under
6
3,821
1,658
4,744
%
difference
from
Table
1
+
2.0%
­
3.9%
+
2.2%

References
Jacobs,
DE,
Clickner,
RP,
Zhou,
JV,
et
al.
(
2002)
The
prevalence
of
lead­
based
paint
hazards
in
U.
S.
housing.
Environmental
Health
Perspectives.
110(
10):
A599­
A606.

Westat
(
2001)
National
Survey
of
Lead
and
Allergens
in
Housing.
Final
Report.
Volume
I:
Analysis
of
Lead
Hazards.
Revision
6.0.
Final
report
prepared
for
the
Office
of
Lead
Hazard
Control,
U.
S.
Department
of
Housing
and
Urban
Development.
18
April
2001.
