►
From YouTube: Census Updates #2
Description
The census has been in the news with an announced rollback of field operations, a return to the statutorily set release date for reapportionment numbers and a promised change in how those numbers will be calculated (with the president’s memorandum signaling that undocumented immigrants will not be included).
A
A
huge
amount
of
attention
recently
and
the
headlines.
I've
read
such
as
this
court
fight
over
census,
tied
to
plan
for
excluding
illegal
immigrants
that
came
from
the
wall
street
journal
and
two
huge
questions
loom
as
census
winds
down.
That's
the
new
york
times
supreme
court
rules.
Trump
administration
may
end
census
early.
That
came
from
national
review,
so
lots
of
headlines,
and
we
will
talk
about
some
of
those
things
that
are
causing
headlines.
But
mostly
what
we're
interested
in
today
is
on
the
processing.
A
A
We
are
delighted
that
we
have
with
us
two
experts
on
the
census.
One
of
them
is
very
familiar
to
the
ncsl
audience.
That's
james
whitehorn.
The
chief
of
the
census,
redistricting
and
voting
rights
data
office
he's
our
most
frequent
speaker
ever.
We
would
never
put
on
a
meeting
without
james
in
attendance
and
then
we'll
work.
Welcome.
A
For
the
first
time
stephen
romoleski
he's
the
director
of
the
city
university
of
new
york's
mapping
program
and
you'll
soon
find
out
that
he's
been
doing
a
lot
of
mapping
related
to
the
census
before
I
do
turn
it
over
to
you.
James
I'd
like
to
mention
to
the
group
that
this
meeting
is
open
to
the
public.
A
A
If
any
of
you
have
something
in
particular
that
you
want
to
be
sure
that
you
get
out
of
today's
conversation,
if
you
could
put
that
in
the
chat
box,
that
would
be
great.
It
could
easily
be
that
james
or
steven
will
address
it
as
they
go
along,
but
if
not
we'll
be
coming
back
to
that
for
the
q
a
towards
the
end
of
our
session.
A
We
do
have
60
minutes
planned
for
this,
and
if
we
haven't
gotten
all
the
questions,
we
might
hang
on
a
little
bit
longer
we'll
kind
of
see
what
that
looks
like,
and
I'm
not
sure
if
I've
asked
our
speakers
if
they
can
stay
on
longer
than
that,
so
we'll
find
out
kristy.
Is
there
anything
else
I
should
say
before
we
get
going?
A
A
B
All
right
thanks!
Well,
I
do
want
to
say
that
I
I'm
very
happy
to
be
here
with
you
today.
I
wish
we
could
be
doing
this
in
person.
I
do
miss
the
the
registering
seminars
and
and
the
annual
meetings
that
ncsl
puts
on
that
really
gives
us
a
chance
to
get
together
and
talk
face-to-face,
but
I
am
thankful
for
ncsl
continuing
to
give
us
this
forum
to
connect
and
share
information.
B
B
B
All
states
dc
and
puerto
rico
reached
at
least
a
99
total
response
rate,
and
this
is
what
was
the
stated
goal
of
the
census
bureau,
as
we
were,
were
bringing
the
non-response
operation
to
a
completion
all,
but
one
state
was
at
99.9
total
response
rate.
We
never
put
100
because
we
know
there's
always
possibly
something
out
there.
Louisiana
was
the
one
state
that
did
not
meet
that
goal,
but
they
did
meet
that
99
total
response
rate
for
the
self-response
half
of
the
program.
B
We
had
our
census
questionnaire
assistance,
centers
operating
without
interruption
from
march
12th
through
october
15.
That
meant
that
they
they
continued
to
operate
and
function
as
they
were
intended
and
those
were
the
areas
where
people
could
call
if
they
needed
help
assistance
with
responding
to
the
census
from
a
language
perspective
or
if
they
even
wanted
to
give
their
census
response
through
a
telephone
interview,
as
opposed
to
using
paper
or
internet,
our
internet
self-response
instrument-
and
I
know
going
into
the
census.
This
was
a
big
concern.
B
People
were
really
worried
that
you
know,
since
it's
going
to
adding
a
major
internet
component
as
part
of
the
the
census,
it
could
be
the
next
healthcare.gov.
It
was
sort
of
the
underlying
fear.
We
were
put
on
a
gao
watch
list
for
technology
projects
and
I'm
very
very
pleased
to
say
that
our
internet
self-response
operated
with
zero
downtime
from
march
12th
through
october
15th.
We
did
not
have
a
single
moment
of
down
time
for
that,
so
we're
very
proud
of
that.
B
The
non-response
follow-up,
where
we
really
go
out
and
we
get
the
capture
the
the
folks
who
are
reluctant
to
spawn
to
respond
to
hard
to
count
populations,
and
I
think
we're
going
to
hear
a
little
bit
more
about
that
from
our
second
speaker,
the
the
non-response
follow-up
rate
accounted
for
32.9
percent
of
our
total
responses
and
we
had
a
lot
of
challenges.
During
this
non-response
follow-up.
We
had
repeated
hurricanes
in
the
gulf
coast
states.
We
had
the
global
pandemic,
which
required
the
census
bureau
to
scramble
and
purchase
and
distribute
42
million
pieces
of
ppe.
B
The
way
we
were
able
to
successfully
complete
this
non-response
follow-up
can
trace
back
to
one
of
the
innovations
that
was
part
of
the
original
2020
census
plan,
which
was
to
use
technology
to
really
streamline
and
update
our
field
operations.
We
implemented
things
like
optimized
routing
for
casework.
We
had
nightly
shifting
of
cases
to
different
enumerators.
B
We
automated
a
lot
of
the
back
end
work
that
enumerators
would
have
had
to
do
when
they
were
in
the
office,
their
payroll,
their
timesheets,
their
trainings.
All
of
those
things
were
advanced
with
technology
to
streamline
the
operations
and
it
really
allowed
us
to
increase
our
enumerator
productivity
to
1.92
cases
per
hour.
Instead
of
the
1.1
cases
per
hour,
we
had
in
2010
when
we
were
working
in
a
fully
paper
environment.
B
We
were
also
very
lucky
in
that
a
lot
of
folks,
9
million
to
be
exact
or
9
million
households
self-responded
after
our
enumerators
visited
their
home
and
left
to
notice
a
visit.
They
didn't
capture
them
for
a
non-response
follow-up
enumeration,
but
they
did
leave
a
notice
that
we
had
been
there,
and
it
was
enough
of
a
reminder
that
these
folks
jumped
in
and
self-responded
on
their
own.
B
So,
as
with
the
the
the
self-response,
the
census
bureau
is
more
transparent
than
we
have
been
in
the
past
about
how
our
enumeration
was
going.
We've
posted
this
table
that
I've
got
up
on
the
screen
where
you
can
go
in
and
see
the
breakdown
between
self-response
and
non-response
follow-up
for
each
state
across
the
country.
But
in
addition
to
this,
we
also
posted
maps
that
people
could
go.
Look
at
for
their
area.
B
B
So
as
we
shift
from
collection
to
processing
and
then
the
eventual
data
dissemination,
a
big
concern
we
hear
is
regarding
data
quality,
considering
the
unprecedented
changes
with
the
2020
census.
I
want
to
touch
on
a
series
of
actions
we're
taking
to
ensure
that
the
operational
changes,
delays
and
phased
activities
do
not
negatively
impact
the
quality
of
the
census,
special
teams,
with
expertise
from
within
the
bureau,
in
the
field
of
operations,
statistical
methodology,
acquisition
and
utilization
of
administrative
records
and
then
our
subject
matter.
B
Multiple
working
groups
have
been
assembled
to
help
identify
new
and
emerging
ways
to
assess
or
ensure
quality.
This
was
going
on
while
we
were
collecting
data
as
well
as
now
when
we're
in
post
data
collection
and
they're,
also
working
with
existing
operational
teams
to
to
make
sure
that
we're
making
the
adjustment
and
changes
that
are
needed,
and
then
we've
also
added
this
data
quality
documentation
team.
That
will
be
ensuring
that
we're
we're
putting
out
the
correct
reports
that
are
needed
for
folks
to
help
understand
the
quality
of
the
2020
census.
B
So
some
of
the
first
and
earliest
data
quality
measures
that
we're
able
to
talk
about
are
about
our
response
rates
and
and
digging
in
just
a
little
bit
deeper.
So
the
2020
census,
non-response
follow-up,
metrics
that
we
have
available
tell
us
that
5.6
percent
of
the
addresses
nationwide
were
resolved
by
high
quality
administrative
records.
B
We
use
administrative
records
when
census
taker
efforts
to
contact
the
housing
unit
are
unsuccessful,
and
our
analysis
shows
that
very
high
confidence
that
the
administrative
records
that
we
have
are
complete
and
correct.
Completing
13.9
percent
of
the
non-response
follow-up
workload
using
administrative
records
is
lower
than
we
expected.
We
expected
22.5
percent
of
cases
to
have
high
quality
administrative
records.
B
B
If
census
takers
can't
get
a
response
directly
from
a
household
after
three
visits,
except
where,
of
course,
the
high
quality
administrative
records
are
available
that
I
just
mentioned,
they
try
to
get
information
about
the
address
from
a
proxy
such
as
a
neighbor,
a
landlord
or
a
building
manager,
because
covet
19
delayed
the
start
of
sensor
census
taker
visits.
We
anticipated
that
we
might
need
significantly
more
proxies
this
decade.
However,
we're
pleased
to
report
that
the
proxy
rate
is
actually
similar
to
the
2010
proxy
rate.
You
can
see
the
24.1
and
the
23.8
on
the
slide.
B
So,
in
addition
to
these
metrics
and
and
the
groups
I
mentioned
previously,
the
census
traditionally
has
two
primary
methods
for
evaluating
quality.
Demographic
analysis
is
a
technique
used
to
develop
an
understanding
of
the
age,
sex
and
racial
composition
of
population
and
how
it's
changed
over
time
through
the
basic
demographic
processes
of
births,
deaths
and
migration.
B
B
The
second
method
is
the
post-enumeration
survey.
That's
a
short
survey.
That's
run
after
each
census
to
determine
how
to
the
it's
conducted
as
its
own
survey,
to
determine
how
many
people
were
were
captured
or
left
behind,
how
many
people
we
should
have
counted
and
how
many
people
we
did
count-
and
this
is
how
we
calculate
the
differential
under
count,
which
is
one
of
the
most
highly
followed
pieces
of
information
around
data
quality
for
the
census
that
people
expect
to
see.
B
So
now
I'm
going
to
switch
gears
a
little
bit
and
talk
about
the
registering
data
program
and
the
registration
data
products.
Some
a
little
bit
of
this
is
going
to
be
repetitive
for
folks
who
usually
tune
into
what
I
say,
but
I
think
I
have
a
little
bit
more
information
for
you
now
when
we
scheduled
this
meeting.
I
was
very
hopeful
that
we
would
have
a
new,
concrete
schedule
that
we'd
be
able
to
share
with
you.
B
I
also
included
on
this
slide,
which
I'll
be
providing
to
wendy
and
christy
so
that
they
can
distribute
it
to
the
attendees
later
the
link
to
those
products.
So
that's
on
this
slide
here
for
you.
I
don't
expect
people
to
be
able
to
grab
such
a
lengthy
url
during
our
presentation,
but
it
is
there
for
for
later
once
those
materials
come
out
now,
as
we've
mentioned
publicly
before
the
census
bureau,
that
is
in
order
to
deliver
the
apportionment
accounts
as
soon
as
possible.
B
B
Although
it's
not
yet
official,
I
am
hopeful
that
we'll
be
able
to
complete
the
delivery
of
the
geographic
products.
All
states
dc
puerto
rico
by
more
early
march,
we'll
be
working
to
provide
the
registration
tabulations
as
close
to
the
statutory
deadline
as
possible,
while
assuring
the
high
quality
of
that
data.
B
It
is
likely,
but
also
not
yet
determined
that
we'll
move
to
a
single
data
delivery
rather
than
a
tiered
weekly
delivery,
as
we
had
originally
planned
now.
As
soon
as
I
have
all
the
dates
and
timelines
and
the
final
expectations
for
recipients.
Lockdown
I'll,
be
communicating
that
directly
to
the
official,
liaisons
and
stakeholders
of
the
registration
data
program
to
groups
like
ncsl
through
the
redistributing
listserv
and,
of
course,
to
the
public.
B
So
now
that
we've
discussed
the
timing
I'll
admittedly
I
know
there
is
some
detail
missing
there.
I
want
to
touch
on
the
products
that
can
be
expected.
This
decade
we
have
the
four
primary
geographic
products,
the
shape
files,
the
maps,
the
block
assignment
files
and
the
block
to
block
relationship
files.
B
B
I
sometimes
date
myself
in
in
gis
world
by
calling
them
atomic
polygons
but
they're
simply
they're
areas
that
are
smaller
than
blocks
they're,
just
individual
polygons
in
our
geographic
database,
but
each
one
is
tagged
with
all
of
the
geography
that
is
associated
with
that
phase,
so
that
you
can
build
your
own
topologically
integrated
geographic
database.
If
you
need
it
for
the
work
that
you're
planning
to
do.
B
As
I
mentioned,
also
we're
going
to
have
these
pdf
maps,
this
is
an
example
of
a
block
map
just
to
show
you.
These
are
our
full
size,
large
plotter
maps,
so
that
you
can
print
them
at
scale
a
large
scale
so
that
you
can
work
with
them
as
needed.
This
will
be
for
state
legislative
and
voting
districts,
there'll
be
census,
block
maps
which
will
have
every
block
within
the
county
illustrated
track
maps
and
school
district
maps.
B
B
The
block
files
list
all
of
the
blocks
of
geography
within
a
state
and
the
associated
geography
with
them,
depending
on
which
file
you're.
Looking
at
this
example
here
is
a
minor
civil
division
or
township
file.
So
you
can
see
it
lists
the
blocks.
We
can
see
that
all
the
blocks
in
the
highlighted
column
of
yellow
belong
to
this.
B
This
township
of
five
nine
zero
zero
zero,
but
you
use
that
in
conjunction
with
the
name
table
you
see
above
where
the
the
geographic
code
of
5900
equates
to
providence
city.
So
it's
just
an
example
of
how
you
can
use
these
files,
and
this
is
available
for
multi-level
multiple
levels
of
geography
and
it's
particularly
helpful
for
folks
who
want
to
crunch
numbers
from
the
census
and
have
the
associated
geography
without
actually
having
to
go
through
a
gis.
B
That's
exactly
the
same
file
we
produced
for
2010,
it's
the
same
file
we
produced
in
the
prototype
data
products
and
that's
also
the
product
we
intend
to
produce
for
the
2020
census.
In
addition,
in
the
prototype
product
and
in
the
2020
census,
you'll
see
this
new
table.
The
group
quarters
population
by
group
quarters
type.
This
is
a
total
population
in
group
quarters
and,
if
you're
not
familiar
with
census,
lingo,
a
group
quarters
is
things
like
the
correctional
institutions
for
adults
or
adult
prisons,
nursing
facilities,
college
university,
student,
housing
or
military
quarters.
B
B
B
There
are
two
main
differences
for
this
one:
this
product,
it's
going
to
include
some
other
race
alone
as
a
category
to
make
it
more
compatible
with
table
p4
from
the
pl
94
171
redistricting
data,
some
other
race
alone
is
not
a
category,
that's
in
the
american
community
survey.
So
that's
why
it's
different
since
we're
using
this
to
align
it
with
the
pl
data.
B
So
how
are
we
going
to
create
this
second
version?
There's
been
an
internal
expert
panel
working
on
this
at
the
census,
bureau
and
they're
expected
to
publish
their
final
methodology
paper
by
the
end
of
this
month.
They
previewed
that
report
at
our
last
scientific
advisory
committee
meeting
and
that
can
be
viewed
online.
B
All
the
data
products
I
mentioned
today,
with
the
exception
of
apportionment,
will
have
disclosure
avoidance
techniques
applied.
We've
done
that
for
the
last
several
decades.
For
all
of
these,
the
important
difference
is
that
this
decade,
we're
moving
to
something
that's
called
formal
privacy
which
injects
noise
into
the
data
to
protect
respondents,
while
still
maintaining
the
data
is
fit
for
use
by
data
users,
and
it
does
mean
yes,
I
mask
this
quite
frequently
that
the
pl
94171
redistricting
data
will
have
differential
privacy
applied.
B
B
B
B
So
the
folks
at
ipums
and
the
nhgis,
the
national
historic
gis
converted
these
ppms
into
tables
that
data
users
would
recognize
and
make
it
to
make
it
easier
for
them
to
work
with
now.
Our
latest
version
was
published
on
september
17th,
to
show
not
only
the
progress
today,
but
also
our
change
to
focus
solely
on
the
pl
94
171
redistricting
data.
Initially
this
the
the
output
from
the
decennial
census
would
accommodate
both
the
pl
file
and
the
demographic
housing
and
characteristic
file,
which
would
be
the
summary
file
one
replacement
for
this
decade.
B
I
do
have
to
point
out-
and
this
is
the
user
note-
that
I
was
mentioning-
that
there
was
an
hour-
an
error
found
in
this
latest
version
that
may
make
the
data
appear
more
accurate
in
the
person
data
file
that
was
created,
there's
a
person
data
and
a
housing
unit,
data
file,
two
separate
ones.
The
person
data
file
was
more
accurate
than
it
may
actually
be.
B
We
had
occupied
housing
units
at
the
block
level
was
held
unchanged,
and
so
we
have
corrected
that
error
in
our
code
and
we're
committed
to
ensuring
a
level
of
accuracy
and
future
releases
that
is
comparable
or
better
than
the
data
from
that
september.
17Th
release
and
you
can
see
what
those
what
that
commitment
means
by
looking
at
the
detailed
summary
metrics
that
accompany
that
release.
B
The
those
eligible
are
the
highest
elected
officials
of
affected
governmental
units,
tribal
areas.
This
includes
federally
recognized
american
indian
tribes,
with
reservation
and
or
off
reservation
trust
lands.
The
alaska
native
regional
corporations,
also
native
villages,
states
and
their
equivalents
counties
and
their
equivalents.
Minor
civil
divisions,
townships,
consolidated
cities
incorporated
places,
so
these
legally
functioning
governments
all
have
the
ability
to
make
these
appeals.
B
However,
it's
very
important
that
people
understand
that
these
appeals,
even
when
successful,
do
not
alter
the
apportionment
counts,
the
redistricting
data
or
the
demographic
housing
and
characteristic
data
sets
and
if
you're
not
familiar
with
the
demographic
housing
and
characteristics
datasets.
As
I
mentioned,
that
is
what
we're
calling
the
product
that
would
have
typically
been
called
summary
file,
one
in
a
previous
census.
It's
been
modified
enough
that
they
felt
that
the
name,
a
modification
to
the
name
was
warranted.
B
B
B
So
in
2010
there
were
three
kinds
of
cases:
boundary,
geocoding
and
coverage,
but
for
2020
they've.
Actually,
it's
the
same
sort
of
types
as
you
just
saw,
but
they've
reorganized
it
into
just
two:
they
have
a
boundary
change
case
or
that
living
quarter
count
which
includes
both
the
geocoding
and
coverage
issues,
because
I
thought
it
made
it
more
sense
for
participants.
No
one
really
knows
if
account
issues
due
to
geocoding
or
coverage.
Until
we
do
the
research
so
asking
the
jurisdiction
to
choose.
One
of
those
factors
up
front
didn't
really
make
sense.
B
As
far
as
the
materials
required
in
2010,
we
required
a
full
address
list
for
the
blocks,
the
blocks
and
jurisdictions
that
thought
they
had
fishy
numbers.
But
this
decade
for
the
living
quarters
count
cases
we're
just
asking
for
the
geoids
of
the
blocks
that
they
think
they
have
incorrect
living
quarter
counts
and
the
number
of
housing
unit
and
group
quarters
they
think
should
be
in
those
blocks
and,
as
I
mentioned
at
least,
one
address
is
required
for
a
boundary
case,
because
otherwise
it
has
to
go
through
the
traditional
boundary
and
annexation
survey
process.
B
The
timeline
for
this
is
on
september
1st.
The
letter
will
go
out
to
all
those
tribal
chairpersons
and
highest
elected
officials
who
are
eligible
to
file
cases
of
2021
and
october
1st
of
2021.
The
census
bureau
will
begin
processing
the
cases
that
they've
received
and
then
we'll
stop
accepting
cases
on
june
30th
of
2023
and
then
provide
the
results
for
all
impacted
users
no
later
than
september
30th
of
2023.
B
A
Well
james,
thank
you
very
much.
It's
always
a
pleasure
to
work
with
you.
I
I
can
say
how
about
a
little
extra
on
such
and
such
and
boom
there
it
happens
for
us,
but
so
I
really
appreciate
that.
Stephen
before
we
come
to
you,
I
want
to
ask
james
a
couple
of
sort
of
yes,
no
questions
and
I
had
them
teed
up
and
now
I've
lost
my
place
here.
The
first
is
whether
you
will
have
proxy
rates
available
below
the
national
level.
A
Great,
thank
you
and
we
have
a
new
jerseyite
who'd
like
to
ask
about
the
delivery
of
redistricting
data
for
new
jersey
and,
as
you
know,
and
most
people
on
the
call
probably
know
they
run
their
elections
in
the
eve
odd
numbered
years,
so
they
really
like
to
get
their
data
early.
So
what
can
you
tell
us
about
when
that
data
will
come,
knowing
that
their
primaries
are
in
june?.
B
So,
in
the
past
we've
delivered
data
a
couple
states
each
week
over
a
series
of
several
weeks
and
looking
at
our
schedules
over
and
over
to
try
to
get
as
close
to
our
statutory
deadline
as
possible
and
also
to
to
maintain
the
quality
of
the
data.
We've
found
that
we
can
gain
several
weeks
by
moving
in
that
direction,
where
every
state
gets
their
data
on
the
same
day.
A
Kevin
good
luck,
working
that
out
james!
Thank
you.
As
always.
Stick
with
us
there's
going
to
be
more
questions.
I
trust
you
can
see
that
they're
coming
in
in
the
chat
box
and
we'll
get
back
to
those
stephen
ruminski.
You
can
see
that
this
audience
is
largely
people
who
are
interested
in
redistricting.
I
know
your
interests
are
much
broader,
so
please
share
slides.
Take
it
away.
C
C
Good
thanks
so
much
so.
My
team
at
the
city
university
of
new
york
graduate
center
has
been
working
closely
with
a
coalition
of
civil
rights
organizations
and
census
stakeholders
in
general
and
philanthropic
foundations,
supporting
them
to
analyze
the
self-response
rates
throughout
the
course
of
the
2020
census.
So
groups
can
prioritize
where
they
should
focus
their
work,
and
this
actually
is
a
reprise
of
an
effort
that
we
a
similar
effort
that
we
did
in
the
leading
up
to
and
during
the
2010
census.
C
And
so
we
have
amassed
a
wealth
of
information
on
our
interactive
map.
I
just
shared
in
the
chat
box
the
link
to
that
map
and
we've
been
providing
weekly
updates
to
these
groups
during
the
self-response
operation
and
during
nerfu,
and
so
I'm
going
to
highlight
some
of
those
concerns
through
our
slides
today
and
information.
So
james
went
through
some
of
the
overall
statistics,
so
I
won't
go
into
great
detail
on
that.
C
The
self-response
rate
nationwide
is
27
by
67
percent,
which
exceeds
the
self-response
rate
from
2010,
and
I
should
point
out
also.
We
just
heard
that
the
census
bureau
will
be
publishing
next
week
next
wednesday.
An
update
of
the
self-response
rates
that
will
include
any
last-minute
responses,
basically
and
then
also
in
2021
they'll,
publish
the
final
self-response
rate,
so
these
numbers
might
change.
James
also
mentioned
the
importance
of
self-response
rates,
so
I'll
go
through
this
quickly.
C
Although
the
census
bureau
and
james
reiterated
this
has
provided
that
on
a
national
level,
we
have
the
self-response
data,
because
the
census
bureau
has
been
able
to
publish
it
down
to
census
tracts,
and
so
it
provides
a
really
good
way
of
evaluating,
not
just
what
the
national
picture
looks
like.
But
what
the
geographic
distribution
is
of
this
information,
and
it's
the
only
thing
we
have
at
the
local
level
at
this
point
to
gauge
anything
about
accuracy
and
quality
and
and
get
a
sense
of
that
going
forward.
So
it's
still
very
useful
information.
C
C
We
know
that,
of
course,
there
were
change
differences,
substantial
differences
in
how
the
census
was
conducted
and,
of
course,
there
was
no
pandemic
in
2010
compared
with
2020,
but
it
is
still
a
valid
metric
of
getting
gauging
what
the
how
well
the
response
rate
did
this
time
around.
Even
the
census,
bureau
and
james's
slide
highlights
that
the
u.s
rate
exceeded
the
2010
rate.
C
But
another
thing
to
point
out
about
this
chart
is
that
here's,
what
the
so-called
normal
self-response
period
would
have
been
from
basically
mid-march
through
the
end
of
april,
then,
of
course,
the
pandemic
hit
and
that
changed
everything
not
only
changed
the
census
schedule,
but
upended.
All
of
the
outreach
efforts
that
had
been
so
well
planned
and
thought
out
and
funded
and
the
self-response
operation
was
extended
substantially
through
the
end
of
september
early
august,
I'm
sorry
the
end
of
august
end
of
july,
early
august
and
unfortunately,
because
of
the
covid
crisis.
C
It
was
great
that
the
time
frame
was
extended,
but
that
made
it
extremely
difficult
to
do
any
kind
of
outreach
or
get
out
the
count
efforts.
Even
though
organizations
were
really
successful
in
pivoting
their
face-to-face
efforts
into
digital
and
virtual
efforts,
but
the
rates
went
up,
they
went
up,
but
you
know
relatively
slowly
and
then
you
could
still
self-respond
during
the
door
knocking
operation.
C
So
that's
the
box
on
the
right
there
and
the
rates
continue
to
go
up
there
as
well
and
again,
eventually,
nationwide
exceeding
the
2010
rate
and
so
I'll
just
go
through
some
of
these
slides
quick.
I
won't
even
go
into
the
detail
here.
This
is
just
for
reference
that
you
could
take
a
look
at
more
closely
later,
but
in
terms
of
time
I
want
to
get
to
so.
C
So
this
gives
us
you
a
sense
of
some
of
the
concerns
about
where
excessive
or
greater
amounts
of
door
knocking
were
needed
compared
with
2010
in
what
eventually
was
a
shorter
amount
of
time,
and
so
this
really
highlights
the
concern
about
really
wanting
to
know
those
metrics
about
the
door
knocking
operations.
So
we
can
gauge
better
what
the
quality
was
of
the
information
collected
through
the
door
knocking
operation.
C
You
can
see
how
it
changed
from
mid-july
through
mid-august,
mid-october
and
another
way
we
looked
at
the
concern
about
where
the
greatest
amount
of
door
knocking
would
be
needed
is
looking
at
the
census
tracts
nationwide
that
had
the
lowest
self-response
rates,
where
the
greatest
amount
of
door
knocking
would
be
required
and
just
for
analysis
purposes.
We
looked
at
the
bottom,
20
percent
census,
tracts
that
had
rates
in
the
bottom
20
percent.
C
In
this
case,
we
highlight
that
there
were
fewer
people
who
are
hispanic
based
on
the
american
community
survey
2014
through
2018
estimates
in
the
bottom
20
percent
of
tracks,
a
smaller
foreign-born
population
and
also
fewer
renter
households.
These
are
all
hard
to
count,
communities
and
population
groups
that
would
be
disproportionately
impacted
by
census.
Accuracy
concerns
people
of
color,
for
example,
had
net
under
counts
in
2010,
and
so
this
is.
C
These
statistics
show
that.
Well,
maybe
there
were
some
quality
improvements
because
of
increases
in
self-response
rates,
but
nonetheless,
when
you
look
overall
at
the
the
population
distribution
of
the
tracks
in
the
bottom
20
percent,
you
see
that
it's
disproportionately
non-hispanic
white
less
so
rather
disproportionately
higher
pop
percent
of
people
who
are
black
or
hispanic
about
the
same
in
terms
of
the
asian
population,
but
also
disproportionately
american
indian
people
with
lower
incomes
and
also
foreign-born
population.
C
So
that's
definitely
a
concern
just
some
key
takeaway
takeaways
that
I've
been
going
through
again.
The
overall
concern
is
that,
because
response
rates,
self-response
rates
tended
to
be
lower
in
communities
of
color,
they
needed
a
greater
share
of
door
knocking
and
because
the
door
knocking
operation
tends
to
provide
less
robust
data.
C
So
we're
very
concerned
this.
The
stakeholder
community,
I
think,
is
very
concerned
about
some
of
these
questions,
because
even
though
we
were
able
to
beat
the
2010
self-response
rate,
a
lot
of
these
numbers
indicate
that
there
may
be
other
data
quality
issues
locally
and
for
particular
populations
that
may
not
have
been
offset
by
those
improvements
in
the
nationwide
self-response
rate.
So
we'll
wait
to
see
what
some
of
these
post-enumeration
metrics
look
like
and
see
what
that
tells
us
I'll
end
it
there
thanks
very
much.
A
Thank
you
stephen.
I
hadn't
really
thought
of
it.
This
way
when
we
put
this
group
together,
but
I'm
now
thinking
that
for
the
last
several
years
when
I've
been
thinking
about
the
census-
and
this
is
probably
true
for
many
of
you
as
well-
I
was
thinking
about-
is
it
complete?
Can
they
get
to
all
the
households?
A
I
also
want
to
point
out
that
one
of
the
slides
you
had
stephen
teaches
me
about
the
law
of
diminishing
returns.
That
was
the
day-by-day
increase
in
the
response
rates
and
it
was
getting
to
be
a
mighty
flat
line
there
at
the
end,
as
we
got
into
the
first
couple
of
weeks
of
october,
so
law
of
diminishing
returns
brought
back
to
me
from
my
high
school
days.
We
have
some
good
questions
here,
feel
free
everyone
to
put
some
additional
ones
in
the
box.
A
I
want
to
start
out
with
jeff's
question,
and
that
has
to
do
with
whether
the
bureau
is
likely
to
request
from
congress,
in
extension
on
the
april
first
deadline,
and
I
know
that
you
can
only
answer
with
what
you
know
right
now
and
you
can't
predict
what
other
court
cases
might
come
and
interfere.
But
do
you
know
if
there's
any
appetite
at
the
bureau
to
request
an
extension
on
that
deadline?.
B
B
A
C
I
was,
I
was
only
going
to
mention
that
just
keep
in
mind
that,
of
course,
through
most
of
that
time
frame,
the
covid
crisis
was
really
preventing
in
some
places,
largely
in
some
places,
completely
any
kind
of
face-to-face
contact
or
large
gatherings,
and
I
know
from
new
york
city's
perspective.
For
example,
there
was
a
substantial
amount
of
money
that
the
city
and
funders
put
in.
A
B
Far
so
those
products
will
be
part
of
the
the
2010
to
2020
block
crosswalk
is
going
to
be
part
of
the
redistricting
data,
geographic
support
products,
and
so
we've
done
testing
as
far
as
creating
the
file
to
seeing
the
file.
But
we
haven't
produced
the
actual
file
because
we
need
to
finish
the
completion
of
the
of
the
geographic
products
themselves,
which
there's
still
quite
a
few
steps
that
have
to
be
done.
B
I
can
say
that
our
initial
view
of
the
geography
that
we've
built
from
the
decennial
census
for
2020
from
the
folks
who
work
in
our
geography
division.
They
feel
that
it's
looking
in
really
good
shape,
and
we
also
believe
that
we
achieved
our
goal
of
driving.
The
number
of
census
blocks
back
down
away
from
the
11
plus
million
that
we
had
in
in
2010
back
down
towards
something
just
over
8
million
like
we
had
in
2000.
So
we
were
successful
there.
A
And
douglas,
if
that
didn't
answer
your
question
stick
something
else
in
the
inbox
and
then
we
have
andy
beveridge.
He's
got
a
comment
here,
there's
a
huge
problem
with
the
latest
release
of
for
testing
differential
privacy,
though
the
population
data
appears
to
be
better.
The
characteristics
are
worse,
for
one
case,
at
least
in
new
rochelle.
The
result
appears
to
dilute
the
population
of
african
americans
in
the
predominantly
african-american
district
and
the
latinos
in
the
predominantly
latino
district,
as
well
as
diluting
the
non-hispanic
white
population
in
their
for
predominantly
non-hispanic
white
districts.
B
It's
also
important
to
note
that
the
cqr
results
will
also
be
affected
by
differential
privacy,
whether
it's
a
record
that
was
already
privacy
protected
or
if
it's
a
record
that
wasn't
included
in
the
decennial
census,
it
will
likely
need
to
have
additional
disclosure
avoidance
applied
to
it
so
yeah.
I
think
that
distinction
will
actually
be
done
internally
and
then,
just
as
a
side
note
to
the
parade
to
address
the
prelude
we
we
are
very
aware,
at
least
through
my
office.
B
The
the
feedback
that
we've
been
getting
over
several
years
from
the
redistributing
communities
is
about
that
importance
of
of
populations
and
and
also
the
importance
of
not
diluting
homogeneous
populations.
Because
of
the
the
work
that's
done
under
the
voting
rights
act.
And
so
you
know
they're
continuing
to
make
improvements
to
the
system
as
as
as
we
approach
its
actual
use
and
we're
consistently
making
that
argument
to
them
about
that
critical
nature
of
as
a
component
for
the
registering
data
set.
A
B
So
that's
where
we,
we
don't
have
the
concrete
schedule,
but
I'm
expecting
from
just
what.
What
I'm
seeing
and
my
putting
myself
out
on
a
there's,
there's
a
disclaimer
on
the
opening
slide
about
these
being
the
comments
of
the
the
speaker
and
not
of
the
agency.
So
this
falls
directly
under
that
that
clause.
But
I
I
I'm
thinking
that
we
will
have
all
the
geography
files
to
all
of
the
states,
the
district
of
columbia
and
puerto
rico.
By
the
end
of
the
first
week
of
march.
A
Okay-
and
we
have
a
question
from
hanzi
lawang,
who
you
probably
know,
is
a
reporter
for
national
public
radio.
He
asks.
Has
the
census
bureau
received
any
requests
from
any
states
or
other
stakeholders
for
citizenship
data
on
the
pl
94
171
file?
Any
requests
come
in
since
september
and,
if
so
from.
B
A
A
My
two
questions
about
the
census
relate
to
whether
I
know
you're,
always
in
a
continuous
improvement
model
of
learning,
from
what
you've
done
before
so
will
there
be
a
opportunity
for
enumerators
to
either
do
focus
groups
or
do
a
survey,
and
some
of
you
know
that
my
husband
was
an
enumerator,
so
I
know
he'd
love
to
give
you
his
thoughts
if
there's
ever
a
way
to
funnel
in
some
of
his
experiences.
So
do
you
take?
Do
you
take
information
from
illuminators
and,
if
so,
how.
B
That's
a
good
question
and
I
would
have
to
get
back
to
you
on
that
as
well.
I
know
that
my
sister
was
also
a
numerator
in
the
state
of
virginia
and
she
would
very
much
like
to
be
able
to
give
her
thoughts
because
she
had
a
lot
of
thoughts
in
in
both
interesting
and
good,
and
also
suggestions
for
improvement.
So
I
will
ask
about
that
and
make
sure
I
get
back
to
you.
A
Can
I
just
say
don't
put
that
at
the
top
of
your
list
that
could
be
like
a
year
from
now.
You
can
look
into
that.
We
fully
understand.
You've
got
more
important
things
to
be
doing
right
now
than
that.
My
husband's
experience
was
also
a
very
interesting
one
and
kind
of
a
mix
of
a
few
things
that
were
a
little,
not
not
how
they
could
have
been,
but
most
of
it
be
very
good
clark.
Benson
asks.
Will
the
cvap
data
also
include
cpop
citizen
population.
B
A
One
thing
that
was
different
in
2020
than
in
2010
was
that
a
great
amount
of
money
came
from
the
states
and
localities
and
from
outside
philanthropic
interests
on
complete
count
committees,
and
it
ranged
from
a
million
dollars
in
alabama
to
many
millions
of
dollars
in
california.
What
I'm
thinking
is
that
next
time
for
2030
I'm
going
to
guess
that
states
are
going
to
put
that
same
kind
of
money
in
possibly
a
little
earlier,
because
it
appears
from
stephen's
comments
that
that
money
paid
off
in
getting
the
message
out
steven.
A
C
Well,
the
state
support
of
outreach
efforts
was
definitely
uneven.
A
number
of
states
really
spent
a
huge
amount
of
money.
Some
states
spent
almost
none
and
the
philanthropic
community
really
stepped
up
to
the
plate
nationwide
and
one
really
good
aspect
of
what
their
contribution
was.
Is
that
effort
started
pretty
early?
We
were
involved,
like
I
said,
in
2010,
leading
up
to
the
2010
census,
with
interactive
maps
and
all
of
the
hard
to
count
areas
and
that
wasn't
until
like
2009.
C
this
time
around
the
philanthropic
community
and
civil
rights
groups
and
stakeholders
started
the
planning
process
in
2015
2016..
Thank
goodness,
they
did
not
only
because
of
all
of
the
difficulties
from
the
white
house,
but
also
because,
of
course,
we
had
a
pandemic,
which
no
one,
of
course
could
plan
for.
Hopefully
that
will
continue.
You
know
when
we
lead
up
to
2030.
B
All
I
can
say
is
from
from
what
I
know
it
is
an
invaluable
part
of
the
work
of
getting
people
to
respond
and
making
sure
people
are
aware
about
the
census,
making
sure
that
they
understand
how
to
get
in
touch
with
the
census
and
how
to
how
to
respond
to
make
sure
that
you
know
bad
actors.
Don't
don't
don't
interfere
so
it
the.
B
I
can't.
I,
the
the
census
bureau
is
extremely
grateful
for
all
the
support,
the
unprecedented
support
we
had
from
from
the
philanthropic
community
from
the
public
from
community
service
organizations
just
across
the
board.
It
was,
it
was
bigger
and
better
than
it's
ever
been,
and
we're
very
thankful
for
that.
A
Thank
you
now.
I
know
it's
102.,
so
for
those
of
you
who
need
to
leave
feel
free,
we've
got
a
few
more
questions
coming
in
and
if
stephen
and
james
can
hang
on
for
those
that
would
be
great.
But
I
do
want
to
thank
everyone
who
has
been
with
us
for
this
length
of
time,
and
I
know
that
james
and
stephen
are
both
open
to
receiving
questions
from
you
all.
A
If
you,
if
james
and
steven,
if
you
would
be
willing
to
put
your
email
addresses
in
the
chat
box,
then
you
would
prove
that
I
meant
what
I
said
that
you're
willing
to
be
responsive
to
anybody
that
going
along
here.
So
mickey
responded
saying
that
this
is
a
very
excellent
time
to
be
thinking
about
the
2030
census
and
how
the
partnerships
can
be
even
stronger,
and
can
we
get
that
moving
earlier?
A
I
think
the
philanthropic
community
would
be
interested
in
that
and
I
think
the
u.s
census
bureau
would
be
interested
in
that
as
well,
and
I
I
hope
that
the
philanthropic
community-
and
I
certainly
know
with
christie
zamarripa
here
at
ncsl-
will
be
bundling
up
what
did
happen
amongst
the
states
this
year.
Let's
see,
I
did
want
to
ask
on
behalf
of
clark
what's
the
status
of
the
reassignment
of
the
group
quarters
persons
and
the
d
duplicate,
I
don't
know
if
I
get
de-duplication
aspect.
B
B
A
B
This
is
again
me
speaking
from
from
myself
and
not
necessarily
the
agency,
but
I
because
I
haven't
actually
read
the
data
processing
final
operational
plan,
but
from
the
discussions
that
I've
had
and
the
things
that
I've
heard,
the
the
spirit
of
that
plan
is
still
operational.
There
may
be
some
some
additional.
B
I
know
that
there
was
a
mention
yesterday
that
there
was
a
math
quality,
a
master
address,
file,
quality
check
that
we
really
didn't
feel
needed
to
be
done
because
we'd
already
done
multiple
quality
checks
through
the
entire
process
of
the
census,
so
that
was
removed.
But
from
my
it's,
my
understanding
that
the
remainder
of
the
plan
is
still
in
place,
it's
just
being
done
more
efficiently.
B
Alfontnow
mentioned
during
his
press
briefing.
The
other
day
that
you
know
technology
is
playing
a
big
part
in
that
we've
got
more
horsepower
for
some
of
these
operations.
That's
allowing
it
to
run
faster,
we're
having
people
work,
nights
and
weekends,
but
I
believe
the
the
the
spirit
and
the
content
of
that
plan
is
still
intact.
A
James
that
it
was
such
a
wonderful
bridge
to
my
closing,
which
is
to
say
I
know,
you've
been
working
nights
and
weekends,
stephen,
I'm
fairly
sure.
In
the
final
days,
you
were
also
working
nights
and
weekends,
and
I
know
that
many
of
the
people
who
are
on
the
line
here
are
working
nights
and
weekends,
if
not
only
about
the
census
but
about
the
upcoming
election
and
getting
prepared
for
redistricting
next
year.