[House Hearing, 119 Congress]
[From the U.S. Government Publishing Office]
ENUMERATION OR ESTIMATION:
WHY INACCURATE CENSUS RESULTS HURT
AMERICAN CITIZENS
=======================================================================
HEARING
BEFORE THE
SUBCOMMITTEE ON THE CONSTITUTION AND
LIMITED GOVERNMENT
OF THE
COMMITTEE ON THE JUDICIARY
U.S. HOUSE OF REPRESENTATIVES
ONE HUNDRED NINETEENTH CONGRESS
FIRST SESSION
__________
WEDNESDAY, NOVEMBER 19, 2025
__________
Serial No. 119-38
__________
Printed for the use of the Committee on the Judiciary
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Available via: http://judiciary.house.gov
__________
U.S. GOVERNMENT PUBLISHING OFFICE
61-917 WASHINGTON : 2026
-----------------------------------------------------------------------------------
COMMITTEE ON THE JUDICIARY
JIM JORDAN, Ohio, Chair
DARRELL ISSA, California JAMIE RASKIN, Maryland, Ranking
ANDY BIGGS, Arizona Member
TOM McCLINTOCK, California JERROLD NADLER, New York
THOMAS P. TIFFANY, Wisconsin ZOE LOFGREN, California
THOMAS MASSIE, Kentucky STEVE COHEN, Tennessee
CHIP ROY, Texas HENRY C. ``HANK'' JOHNSON, Jr.,
SCOTT FITZGERALD, Wisconsin Georgia
BEN CLINE, Virginia ERIC SWALWELL, California
LANCE GOODEN, Texas TED LIEU, California
JEFFERSON VAN DREW, New Jersey PRAMILA JAYAPAL, Washington
TROY E. NEHLS, Texas J. LUIS CORREA, California
BARRY MOORE, Alabama MARY GAY SCANLON, Pennsylvania
KEVIN KILEY, California JOE NEGUSE, Colorado
HARRIET M. HAGEMAN, Wyoming LUCY McBATH, Georgia
LAUREL M. LEE, Florida DEBORAH K. ROSS, North Carolina
WESLEY HUNT, Texas BECCA BALINT, Vermont
RUSSELL FRY, South Carolina JESUS G. ``CHUY'' GARCIA, Illinois
GLENN GROTHMAN, Wisconsin SYDNEY KAMLAGER-DOVE, California
BRAD KNOTT, North Carolina JARED MOSKOWITZ, Florida
MARK HARRIS, North Carolina DANIEL S. GOLDMAN, New York
ROBERT F. ONDER, Jr., Missouri JASMINE CROCKETT, Texas
DEREK SCHMIDT, Kansas
BRANDON GILL, Texas
MICHAEL BAUMGARTNER, Washington
------
SUBCOMMITTEE ON THE CONSTITUTION AND LIMITED GOVERNMENT
CHIP ROY, Texas, Chair
TOM McCLINTOCK, California MARY GAY SCANLON, Pennsylvania,
THOMAS MASSIE, Kentucky Ranking Member
HARRIET HAGEMAN, Wyoming STEVE COHEN, Tennessee
WESLEY HUNT, Texas PRAMILA JAYAPAL, Washington
GLENN GROTHMAN, Wisconsin JOE NEGUSE, Colorado
MARK HARRIS, North Carolina BECCA BALINT, Vermont
ROBERT F. ONDER, Jr., Missouri SYDNEY KAMLAGER-DOVE, California
BRANDON GILL, Texas DANIEL S. GOLDMAN, New York
CHRISTOPHER HIXON, Majority Staff Director
ARTHUR EWENCZYK, Minority Staff Director
C O N T E N T S
----------
Wednesday, November 19, 2025
OPENING STATEMENTS
Page
The Honorable Chip Roy, Chair of the Subcommittee on the
Constitution and Limited Government from the State of Texas.... 1
The Honorable Mary Gay Scanlon, Ranking Member of the
Subcommittee on the Constitution and Limited Government from
the State of Pennsylvania...................................... 3
The Honorable Jamie Raskin, Ranking Member of the Committee on
the Judiciary from the State of Maryland....................... 6
WITNESSES
Trey Mayfield, Member, Chalmers, Adams, Becker, & Kaufman, LLC
Oral Testimony................................................. 8
Prepared Testimony............................................. 11
Jay Rodriguez, Assistant Attorney General, Office of the Kansas
Attorney General
Oral Testimony................................................. 13
Prepared Testimony............................................. 15
Joseph Wade Miller, Senior Advisor, Center for Renewing America
Oral Testimony................................................. 26
Prepared Testimony............................................. 28
John C. Yang, President, Executive Director, Asian Americans
Advancing Justice | AAAJC
Oral Testimony................................................. 30
Prepared Testimony............................................. 32
LETTERS, STATEMENTS, ETC. SUBMITTED FOR THE HEARING
All materials submitted by the Subcommittee on the Constitution
and Limited Government, for the record......................... 74
Materials submitted by the Honorable Becca Balint, a Member of
the Subcommittee on the Constitution and Limited Government
from the State of Vermont, for the record
An article entitled, ``Immigration Agents Have Held More Than
170 Americans Against Their Will,'' Oct. 16, 2025,
ProPublica
An article entitled, ``NPR fact checks Kristi Noem on ICE
detaining US citizens,'' Nov. 5, 2025, NPR
An article entitled, ``Citizenship question effects on
household survey response,'' Mar. 20, 2025, The Journal
of Policy Analysis and Management
A report entitled, ``Apportioning Seats in the U.S. House of
Representatives Using a 2013 Estimated Population," Oct.
2015, Congressional Research Service
An article entitled, ``What You Need to Know About the
Citizenship Question and the Census,'' Jul. 2, 2019, The New
York Times, submitted by the Honorable Harriet Hageman, a
Member of the Subcommittee on the Constitution and Limited
Government from the State of Wyoming, for the record
A report entitled, ``2020 Census: Coverage Errors and Challenges
Inform 2030 Plans,'' Nov. 21, 2024, U.S. Government
Accountability Office, submitted by the Honorable Mary Gay
Scanlon, Ranking Member of the Subcommittee on the Constitution
and Limited Government from the State of Pennsylvania, for the
record
APPENDIX
A written statement from Deborah Weinstein, Executive Director of
the Coalition on Human Needs, submitted by the Honorable Mary
Gay Scanlon, Ranking Member of the Subcommittee on the
Constitution and Limited Government from the State of
Pennsylvania, for the record
ENUMERATION OR ESTIMATION:
WHY INACCURATE CENSUS RESULTS HURT AMERICAN CITIZENS
----------
Wednesday, November 19, 2025
House of Representatives
Subcommittee on the Constitution and Limited Government
Committee on the Judiciary
Washington, DC
The Committee met, pursuant to notice, at 10 a.m., in Room
2141, Rayburn House Office Building, the Hon. Chip Roy [Chair
of the Subcommittee] presiding.
Members present: Representatives Roy, Jordan, McClintock,
Hageman, Hunt, Grothman, Harris, Onder, Gill, Scanlon, Raskin,
Jayapal, Balint, and Kamlager-Dove.
Mr. Roy. The Subcommittee will come to order. Without
objection, the Chair is authorized to declare a recess at any
time. We welcome everyone to today's hearing on the accuracy of
the census.
I will now recognize myself for an opening statement. There
will be some different conclusions and perspectives as there
always are, but I do think there are some areas here where we
can all agree that there are ways to improve the census process
and I do how we will be able to have conversations along those
lines where there can be some bipartisan agreement, that there
are things we might be able to do better for future census
activities.
Look, it is one of the Constitution's simplest commands.
Article 1, Section 2 states that an actual enumeration of the
American people shall be conducted every 10 years. The
decennial census of Americans, though it sounds mundane, is a
core pillar of our democracy. It determines how many people
represent each State in Congress, how many electoral votes each
State gets in our Presidential election, informs the drawing of
district lines for State and Federal offices, and affects how
much money each State receives from the Federal Government.
Simply put, it is important that the Census Bureau gets the
count right, and in 2020, it seems that it failed. Indeed, the
2020 Census should be called the Sanctuary Census. Just like
sanctuary jurisdictions, the 2020 Census unconstitutionally put
illegal aliens ahead of American citizens. Under Section 2 of
the 14th Amendment, known as the Apportionment Clause, the
census must count ``the whole number of persons in each
State.''
At the time of the amendment's ratification, the term
persons were understood to refer to only members of the
American political communities, citizens, and permanent
resident aliens who have been lawfully admitted to the country.
Just as the Supreme Court has held that jurisdiction in the
14th Amendment means, ``not merely subject in some respect or
degree to the jurisdiction of the United States, but completely
subject to their political jurisdiction and owing them direct
and immediate allegiance,'' when determining who is a U.S.
citizen who is born to illegal alien parents.
Indeed, the Supreme Court has recognized a similar
principle, noting that in 1992 the constitutional language
governing the census contemplates more than mere physical
presence and some element of allegiance or enduring tide to a
place. It is unthinkable that the 14th Amendment's Framers
sought to reward states that harbored illegal aliens with
additional Congressional representation. Yet, that is exactly
what happens now.
After the 2020 Census, States with more illegal aliens
often because of a conscious policy choice to hamper Federal
law enforcement and welcome immigration law breakers, get more
seats in Congress, more electoral votes, and more of your
Federal tax dollars. A lot of them, in fact. A leading study
found that if Congressional seats were properly apportioned
based on the number of citizens, Republican States would gain
between 10-12 seats in the House of Representatives and votes
in the Electoral College. That is right. Illegal aliens and
noncitizens have at last ten seats in the U.S. House of
Representatives and the Electoral College and that was before
the Biden Administration policies welcomed ten million illegal
aliens into our country, during his disastrous four years. How
many more seats will that give Democrats in the 2030 Census if
nothing changes?
The counting of illegal aliens in the census and as part of
the apportionment base also violates the core constitutional
principle of one person, one vote. The Supreme Court has
repeatedly held that State and Federal legislative districts
must be roughly equal in population so that each citizen's vote
is roughly equal in power. The 2020 Census undermined this
principle by counting illegal aliens in the apportionment base.
Today, a voter in a Congressional district with 730,000 citizen
residents and 30,000 illegal aliens has a stronger say in the
election of a representative than a voter in a Congressional
district with 760,000 citizens and no illegal immigrants.
Shockingly though, the problems of the 2020 Census did not stop
there.
Beyond the manifest unconstitutionality of including the
illegal aliens, the Census Bureau failed at its simplest task,
accurately counting the legal population of the United States.
The Census Bureau's own after-action report found that it
significantly undercounted the population of six States, five
of which including my home State of Texas, are predominantly
Republican. In fact, in addition to Texas losing its seat,
Florida lost out on two additional seats. Colorado gained a
seat that it did not deserve and Minnesota and Rhode each
maintained a Congressional seat that should have been lost.
Conversely, the Census Bureau found that it over-counted in
eight States, Seven of which are predominantly Democratic.
Estimates showed that these errors which were unprecedented in
the recent history of the census cost Republicans six seats in
the House, six seats, just from miscounting, and that is on top
of the 10-12 extra seats Democrats have from counting illegal
aliens.
What is more, because the Census Bureau used a technique
called differential privacy to scramble the data underlying the
2020 Census, these errors are impossible to correct without an
entirely new census. We have witnesses here today who can
explain the technicalities. As our witnesses will highlight,
the practice of differential privacy was derided by outside
groups across the political and ideological spectrum, calling
for the use of it to end. In fact, a 2021 Harvard study on
differential privacy's effects on redistricting noted that,
``The DAS has a tendency to transfer population across
geographies in ways that artificially reduce racial and
partisan heterogeneity. . . .,'' which makes it impossible to
follow the principle of one person, one vote, as it is
currently interpreted by courts and policymakers. Together
then, the 2020 Sanctuary Census denied Republicans 18 seats in
the U.S. House, 12 by counting illegal aliens, and another six
by counting inaccurately. More importantly, it cost law-abiding
citizens in States across the U.S. their equal say in Congress
and in the Presidential elections.
Congress must act to ensure that at the very least that
illegal aliens are not counted for the purposes of
apportionment. Let's follow the Constitution. Let's stop
enabling policies and encourage illegal immigration. Let's stop
weaponizing the census for political gain and let's find
bipartisan ways to improve the census and let's understand the
technicalities that we are going to hear about today that we
think have bipartisan agreement, create problems in terms of
the estimating procedures that are used in census, and let's
put the American people first. The people's house, the
Electoral College, and all the elections and policies that flow
from that should be representative of the American people.
I now recognize the Ranking Member, Ms. Scanlon, for her
opening statement.
Ms. Scanlon. Good morning and thank you to all our
witnesses for being here. I look forward to hearing from you.
As our Founders recognized, the census is fundamental to
our democracy playing a critical role not just in governance,
but in shaping decisions made by organizations and institutions
across American life. An accurate census is essential for a
fair distribution of political power here in Congress and for
the equitable allocation of billions of dollars of Federal
funds to States and communities including some programs like
Medicaid, free and reduced school lunches, Head Start, and
SNAP. Its data is used by businesses and nonprofits to
determine where and how to operate and by State and local
governments to decide where to invest in public infrastructure
like roads and hospitals and schools.
The census is not just some mathematical equation to
distribute Congressional seats and taxpayer dollars. It is
deeply rooted in our Founders' vision of what a government by
and for the people should be. Alexander Hamilton is on record
saying during the Constitutional Convention in Philadelphia,
There can be no truer principle than this--that every
individual of the community at large has an equal right to the
protection of the government.
Ultimately, the census is a national snapshot we take every
10 years that helps us understand and define who we are, how to
govern, and undermining that process in ways that distort that
picture will harm all us. That is why Democratic and Republican
administrations alike have historically tried to protect the
census from political interference and partisan pressures. This
President and his MAGA allies are deviating from that
precedent, once again trying to rig things in their favor. Now,
these efforts are not new. They began under the first
administration when they tried to add an unprecedented citizen
question to the census. This was a blatant attempt by political
appointees to exclude noncitizens from our count which they
expected to ``benefit Republicans and non-Hispanic Whites,''
and which was ultimately rejected by the Supreme Court. Those
efforts continued with the development of the Right-wing
Project 2025 manifesto which, as predicted, has become a
playbook for the Trump Administration. It devotes a whole
section to plans to appoint political appointees loyal to the
White House to lead the census ``to execute a conservative
agenda'' which includes trying to add a citizenship question
again, focusing census engagement on conservative groups and
limiting efforts to engage historically marginalized and under
counted groups.
Then, earlier this year, President Trump reinstated a memo
to the Commerce Department ordering that the census not count
undocumented immigrants when determining each States' share of
Congressional seats. We just heard a little bit of that
rationale which really conflates the idea of who is in this
country legally, not legally. I mean we constantly hear from
our colleagues here about illegal aliens not understanding that
fully 44 percent of immigrants are citizens, that another 26
percent are legal permanent residents, that another five
percent are here on visas, so the vast majority of the
immigrants in our country are, in fact, here legally as opposed
to the constant refrain we hear from the other side of the
aisle. At any rate, the Constitution's command is that the
``whole number of persons be counted.''
The President has also posted on his social media that he
wants his administration to start work on a new census and in a
common theme of this administration, it blatantly ignores the
fact that the Constitution gives Congress, not the President,
the authority to conduct the census. Our Constitution's Framers
understood the importance of having an accurate and
comprehension population count that is reflected in the fact
that Article 1, Section 2, Clause 3, the Enumeration Clause,
mandates that representatives be apportioned among the States
according to the whole number of persons in each State using an
actual enumeration.
Later, the drafters of the 14th Amendment kept this
phrasing while amending the clause to account for the end of
slavery, reiterating that a State's representation in Congress
must be determined by counting the whole number of persons in
each State. It really could not be clearer, the census must
count the whole number of persons, not just a subset of
persons, living in the U.S. The Constitution's drafters could
have used a narrower term like citizens or voters or something
else, which they did, for example, in other parts of the 14th
Amendment, but they purposely used the broader term persons in
the Apportionment Provision, making it clear that all people
count, including women, who could not yet vote at the time of
that drafting, children, indentured servants, who were often
immigrants, and former slaves, and they rejected proposals that
would restrict the census to voters.
Additionally, the requirement of actual enumeration has
never meant that census takers are only required to count only
those who they can reach directly or in person. We know that
the effort to count every person every decade has never been
perfect and that is why there have been continual improvements
over time, and that is for a variety of reasons, chief among
them being that certain populations are difficult for census
takers to reach. Renters are harder to reach than homeowners.
Children are particularly tough to reach, especially if they
are not reading and writing yet. The Constitution does account
for this. The Supreme Court made clear in Utah v. Evans that
the Constitution does not specify exactly how the census should
be counted. As the Court noted in that case, census takers
implementing our very first count in the late 1700s were only
required to report the names of household heads which are very
different from our system now.
Our Framers knew better than to try and anticipate the
technology or methodology that America would need to conduct
the count in the future, one that must capture the hundreds of
millions of people living in the country today. Instead, the
constitutional North Star they gave us for the census is simply
that it be accurate. These on-going efforts to cast doubts
about the results of the 2020 Census, a census that was
conducted, I might add, while Donald Trump was President, a
part of a larger scheme, one MAGA extremists have designed to
redraw maps mid-decade in a blatant power grab, so that
Republicans can win more seats in Congress and keep the
Majority despite the increasing unpopularity of their Project
2025 agenda. In effect, they are seeking a pregerrymander even
as they attempt other gerrymanders.
The ongoing lawsuit in Florida which seeks to overturn the
2020 Census by challenging some of its counting methodologies
fits right into this scheme. The suit filed against the Trump
Administration that is presumably sympathetic to the plaintiffs
may be just the pretext this administration is looking for to
push for a census redo. If their efforts succeed, Americans
will suffer.
Under counts and other counting errors, particularly hurt
vulnerable and hard-to-reach populations including young
children, racial and ethnic minorities, the transient and
homeless, those in nontraditional housing arrangements, and
those living in very rural or very dense urban communities. In
other words, people living in America who patronize its
businesses, drive on its roads, attend its schools and use its
essential public services, and people who we need to have
included in the count. If these lawsuits by plaintiffs get
their way, and the methods used to account for these
populations are prohibited or limited, essentially putting the
Federal Government's head in the sand, will only move farther
away from, not closer to an accurate and complete count.
Federal funds will be misdirected. States, local governments,
businesses, nonprofits, and others will make less-informed
decisions, inadvertently depriving people of what they need to
live and thrive in our communities and that is devastating for
millions.
Of course, we all want, as the Chair suggests, our census
picture to be fuller and more accurate. It is what our
democracy demands, and we should all strive to improve our data
collection and prevent under counting as we look forward to the
scheduled, regular 2030 Census. To make that a reality, we
should be talking about reforms that limit the Executive
Branch's ability to interfere in the count for political
purposes or reforms to ensure that the Census Bureau has
adequate funding and flexibility to do the job right. We should
not try to curtail or ignore scientific methods which are
available to complete such a huge undertaking or consider
changes that make people trust this process less, rather than
participating in it more.
The census and its results are too consequential for that,
so I look forward to hearing our discussion today and I yield
back.
Mr. Roy. I thank the Ranking Member, and I will now
recognize the Ranking Member of the Full Committee, Mr. Raskin.
Mr. Raskin. Thank you, Mr. Roy. I appreciate it. The call
for a census in Article 1 of the Constitution plays an
essential role in our democracy, guaranteeing for centuries
that there will be an appropriate allocation of House seats and
districts of equal population within each State.
President Trump's extraordinary mid-decade partisan
gerrymander offensive has hit major roadblocks recently in
California, in Utah, and in Federal Court in Texas, but his
forces have continued on their path of destruction and are now
targeting the 2020 Census itself as a way to tear up the
political map of the country.
This week, President Trump released this telltale statement
on social media. This all began with the rigged census he said.
We must keep the Majority at all costs. Republicans must fight
back.
What a syllogism we are offered because Trump must keep the
Majority at all costs, therefore, the 2020 Census was rigged.
This is lunacy. It is like saying because Trump had to win the
2020 election, he really wanted to win it, therefore the
election was rigged and he won it. That is just derangement.
The 2020 Census was not rigged. A 2024 report by the
Government Accountability Office reviewing the accuracy of the
census found that the accuracy of the 2020 national population
count was consistent with the previous census. While the
Bureau's Post-Enumeration Survey estimated significant net
coverage errors in certain regions and States, the accuracy of
the count was overall consistent and over counts and under
counts of the margins are nothing new. The census was not
rigged.
The actual institutional weakness of the census, as the GAO
reported, is the chronic under counts of communities of color
which persisted through the 2020 Census. In 2020 and 2010,
``Black or African American and Hispanic persons, young
children, and renters, were systematically under counted while
non-Hispanic White persons, adults over 50, and homeowners were
over counted.'' Are we going to have a hearing planned on that?
Maybe on the same day we are going to have a hearing with the
Epstein victims. I do not know.
The Census Bureau faced unprecedented challenges in
conducting the 2020 Census because of the raging global
pandemic and the Government's inadequate response to it. COVID-
19 significantly complicated census activities, many of which
were suspended or delayed. In many States and localities,
lockdowns, and travel restrictions stopped the Census Bureau
from accessing communities entirely. The Bureau also had to
contend with national disasters in a number of States, as well
as a planning budget that had been dramatically cut in the
first Trump Administration. These two are the real problems
that we should be discussing today. Instead, we seem to be
wasting our time on a long list of fanciful, self-pleading
theories that have no basis in law or history, and they fly in
the plain text of the Constitution.
Before we assign blame for any actual or imagined
inaccuracies in the results of the 2020 Census, let it be clear
that it was the first Trump Administration that was responsible
for preparing the 2020 Census and the greatest obstacle to an
accurate 2020 Census count was President Trump himself. He
repeatedly sought to undermine the accuracy of the count by
adding a citizenship question which experts warned would
dramatically depress participation including by American
citizens and skew the results for Democratic and Republican
States alike. Thank goodness that the Roberts court thwarted
that when it found that Trump's process is completely contrary
to law and the President's pretext for adding the blatantly
political citizenship question was ``contrived.''
The Trump Administration told the court that they needed
citizen questions to help enforce the Voting Rights Act and
then after oral argument, the evidence came to light indicating
that the request of the administration, a GOP redistricting
expert, had provided an analysis showing that the addition of
the citizenship question to the census and the use of citizen-
only data for redistricting would politically benefit
Republican and the White community. The Supreme Court found the
administration's formal excuses completely unbelievable,
calling them contrived.
Now, in his second term, Trump and his allies are once
again seeking to weaponize the census in an unlawful way and
undermining its accuracy. Trump instructed the Commerce
Department to exclude undocumented immigrants from the census
count, a policy that one three-judge panel said, ``has already
been rejected by the Constitution, the applicable statutes in
230 years of our history.''
The Constitution makes clear that the census includes,
``the whole number of persons in each State to ensure a
complete picture of who resides in the country.'' As Republican
Senator Jacob Howard of Michigan stated during his introduction
of the 14th Amendment in 1866, ``the basis of representation is
numbers. That is the whole population, numbers, not voters;
numbers, not property. That is the theory of our
Constitution.''
In the face of this overwhelming textural and historical
evidence telling us that every person must be counted, some
State officials have filed meritless lawsuits against the
Commerce Department challenging the constitutionality of the
Census Bureau's longstanding, constitutionally mandated
practice to count all persons who reside within the U.S.
regardless of their citizenship status. Right-wing, MAGA
activists in Florida have also filed suit against the Commerce
Department claiming that the census' use of group quarters
count imputations and differential privacy violated the
Constitution in the Census Act and called the accuracy of the
census in this question.
Do not let the jargon or these tortured arguments fool you.
The lawsuits are part of a coordinated strategy to give
coverage to this mid-decade attack on the 2020 Census'
apportionment count ahead of the midterms. Our democracy and
our society cannot afford our constitutional census being
turned into a constant instrument of partisan conflict,
division, and advantage. Let's reject this dangerous direction.
Mr. Chair, I yield back to you.
Mr. Roy. Well, I thank the Ranking Member. Without
recognizing--Mr. Jordan is not here, without objection all
other opening statements will be included in the record. We
will now introduce today's witnesses.
We will start with Mr. Mayfield. Mr. Mayfield is a member
at Chalmers, Adams, Backer & Kaufman, where his practice
includes complex commercial litigation, constitutional law, and
administrative law. Prior to that, Mr. Mayfield was a partner
in a boutique litigation firm, an Assistant United States
Attorney, and served as Counsel to the Director of the United
States Census Bureau.
Next, Mr. Jay Rodriguez. Mr. Rodriguez is an Assistant
Attorney General in the Office of the Kansas Attorney General.
He primarily works on cases involving constitutional issues and
was the attorney of record on behalf of the Kansas Attorney
General in Louisiana versus Department of Commerce.
Mr. Wade Miller. Mr. Miller currently serves as a Senior
Advisor at the Center for Renewing America. Wade is a combat
veteran who served in the United States Marine Corps as an
infantryman deploying in combat theaters three times in Iraq
and the Horn of Africa. He has significant experience and
involvement in campaign activities, conservative grass roots
movement and the public-policy development and for the record,
he also served in my office in my first term in Congress as my
Chief of Staff.
Mr. Yang. Mr. Yang is the President and Executive Director
at Asian Americans Advancing Justice. AAJC works to eliminate
discrimination against and to advance the civil rights of Asian
Americans.
We thank our witnesses for appearing today and we will
begin by swearing you in. Would you please rise and raise your
right hand? Do you swear or affirm under penalty of perjury
that the testimony you are about to give is true and correct to
the best of your knowledge, information, and belief so help you
God?
The witnesses answered in the affirmative. Let the record
reflect that they have. Thank you. Please be seated. Please
know that your written testimony will be entered into the
record in its entirety. Accordingly, we ask that you summarize
your testimony in five minutes.
I am going to start on the right here with Mr. Mayfield.
You have five minutes and you may begin.
STATEMENT OF TREY MAYFIELD
Mr. Mayfield. Thank you, Chair Roy, Ranking Member Scanlon,
and the Members of the Committee for inviting me here today to
discuss issues related to the 2020 Census. My name is Trey
Mayfield and I served as Counsel to the Director of the Census
Bureau at the end of the first Trump Administration during
which time many legal issues concerning the methodology,
accuracy, and use of the 2020 Census were raised.
Our Constitution is largely concerned with structuring our
form of government and placing textural limits on power. The
decennial census is one of the very few specific exercises of
power mandated by the Constitution. The census clause is firmly
lodged in Article 1 of the Constitution. It is, for
constitutional purposes, a Congressional responsibility. To
apportion Members of the House of Representatives among the
several States so that House membership proportionally reflects
each State's population, Article 1, Section 2, Clause 3
mandates that there be ``an actual enumeration'' every 10
years. Section 2 of the 14th Amendment reiterates that
requirement.
When interpreting the Constitution, we must look to its
text, using the commonly understood meaning of the words at the
time it was ratified in 1788. The Constitution requires the
actual enumeration. Not surprisingly dictionaries from the time
demonstrate that enumeration requires an actual counting and
not just an estimation of numbers. The Founders were well aware
of the opportunities for fraud to inflate or deflate a given
State's population to alter representation in the House. For
that reason, the Federal Government was charged with conducting
the census and doing so by actually counting, not estimating,
how many people were living in the United States.
Since 1790, Congress has consistently reaffirmed in census-
authorizing legislation that statistical adjustments are
prohibited in determining our national population for
apportionment purposes. To be sure, since this data is used for
many purposes other than apportionment including dispersing
Federal funds to States and localities, obtaining demographic
data and ascertaining the State of the Nation in such ways as
the Congress and President deem appropriate. Some kinds of
statistical adjustment are permissible in performing those
functions, but not for enumeration.
Congress has mandated the census ``be as accurate as
possible'' and banned the use of statistical adjustments for
enumeration because it ``poses the risk of an inaccurate,
invalid, and unconstitutional census.'' Within those
limitations, Congress has delegated responsibility for
implementing the census to the Secretary of Commerce and the
Census Bureau. Congress also requires the President to report
to Congress the total population of the Nation, each State, and
the number of representatives to which each State is entitled.
There is no law, however, prohibiting the President from
reporting additional information to Congress in that report
such as the number of illegal aliens in the count and what the
results of the count would be in their absence.
Congress has also prohibited the release of any census data
whereby the person who provided the data could be identified.
Purportedly to meet this requirement, civil servants at the
Bureau apply the methodology to the 2020 Census called
differential privacy. It violates the Constitution's actual
enumeration requirement and the Congressional prohibition on
the use of such methods. They refused to allow their method to
be peer reviewed or to disclose the data on which they
concluded the differential privacy was needed instead of the
method the Bureau had previously used known as data swapping.
Using differential privacy altered both the count and the
characteristics of individuals and households reported at the
region, district, town, and census block level. In essence, the
Bureau's reported population moved people around within States.
For instance, when testing differential privacy using the data
from the 2010 Census, the population of Port Royal, Virginia
was artificially increased 87 percent while the population of
Stony Creek, Virginia, was decreased by 43 percent. In other
words, differential privacy results in people from larger
subgroups and geographic units being moved to smaller ones.
Opposition to differential privacy was unanimous among
census stakeholders. Every letter the Bureau received from
outside scientists and statisticians came to the same
conclusion: Differential privacy is illegal, it is data
corrupting, and it addresses a nonexistent problem. It was also
opposed unanimously by governmental actors and interest groups
for reasons as straight forward that using fake population data
makes it difficult for governments to carry out core
responsibilities such as drawing district lines and determining
where to locate bus stops and schools and how to allocate
welfare.
Studies since the 2020 Census have confirmed that this is
exactly what differential privacy does. It adds fake people
where they do not live and subtracts real people from where
they do live. Thus, all the reported enumeration census data
from 2020 below the State level is false.
I welcome the Committee's questions.
[The prepared statement of Mr. Mayfield follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
STATEMENT OF JAY RODRIGUEZ
Mr. Rodriguez. Thank you, Chair Roy, Ranking Member
Scanlon, and the Members of the Committee.
I'm an Assistant Attorney General in the Special Litigation
and Constitutional Issues Division of the Kansas Attorney
General's Office. In that role, I'm leading a multistate
Federal lawsuit against the Department of Commerce challenging
its practice of counting illegal aliens in the census.
Mr. Roy. Mr. Rodriguez, can you move a little closer to
your microphone or ensure that it's on?
Mr. Rodriguez. Sorry. Is that better?
Mr. Roy. That's better, yes.
Mr. Rodriguez. OK. When illegal aliens are counted, it
inflates the count for a few States with high numbers of
illegal aliens and gives those, those States more political
power and representation at the expense of States with smaller
numbers of illegal aliens. That violates the Constitution,
which requires an actual enumeration of inhabitants to be used
for apportioning House seats, and it violates the principle of
equal representation by making votes in some States count less
than votes in others.
The census comes out of Article I, Section 2, of the
Constitution. It requires Congress to conduct an actual
enumeration to be used in the apportionment of Representatives
and Electoral College votes, according to the respective
numbers in each State.
In the Founding Era, numbers referred to inhabitants, which
meant someone with a fixed and permanent connection to the
place and some legitimate tie to the political community. Under
that definition, illegal aliens are not inhabitants. They're
present in the U.S. illegally, subject to removal, and their
residence is, therefore, temporary and contingent. They're not
bona fide members of the State. They should not be included in
any enumeration for purposes of apportionment.
Historically, Congress has delegated its constitutional
authority over the census to the Secretary of Commerce and to
the Census Act, and the census is conducted according to the
rules promulgated by the Secretary.
Every census and apportionment, including the most recent
one in 2020, has counted illegal aliens at their usual
residence, although it's only in the last few decades that the
scale of illegal immigration has significantly affected
apportionment.
In the Supreme Court, there are two broad categories of
cases that are relevant to these issues. The first concerns
equal representation or the principle of one person, one vote.
Starting in the 1960s, the Supreme Court found that debasement
of votes, or vote dilution, is a justiciable injury, and when
State voting districts are unequal in population, it means that
votes in less populous districts count more than votes in more
populous districts, which violates the Equal Protection Clause
and the Apportionment Clause. The Court has held that districts
must be equal in terms of total population, not in terms of the
number of eligible voters.
In these cases, the Court set the lower boundary of who
should be counted. It hasn't set the upper boundary. Even
though total population we know is more than just eligible
voters, we don't know that it must include illegal aliens.
The second category of relevant cases concerns the conduct
of the census. For example, in Department of Commerce v. New
York, the Court held that asking a question about citizenship
was permitted by the Enumeration Clause, given Congress',
quote, ``unlimited discretion'' to conduct the census, but
including the question violated the Administrative Procedure
Act. Without an APA violation, a citizenship question could
likely be included in the next census if it was produced by a
reasoned decisionmaking process.
In Trump v. New York in 2020, the Court heard that
challenge to a Trump Administration memo which planned to
exclude illegal aliens from the already completed census to the
extent feasible. The Court dismissed the case for lack of
standing, since there was no certainty that anyone would
actually be removed from the census pursuant to the memo, but
it did not rule on the legality of excluding illegal aliens
from an actual enumeration.
The census cases show that Congress has broad authority
over what information is collected in the census and that
counting illegal aliens remains an open question. The Court has
not said whether the Constitution requires counting illegal
aliens or prohibits it, or whether Congress has discretion to
decide one way or the other. The future of the census will be
decided in the courts and in Congress.
Thank you.
[The prepared statement of Mr. Rodriguez follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Roy. Thank you, Mr. Rodriguez. I'll now turn to Mr.
Miller. You have five minutes.
STATEMENT OF JOSEPH MILLER
Mr. Miller. Mr. Chair and the Members of the Subcommittee,
thank you for inviting me to testify about the serious flaws in
the 2020 Census caused by the Census Bureau's first-ever use of
differential privacy.
Differential privacy is an algorithm that deliberately
injects noise--in plain English, false data--into census
results below the State-level data, all the way down to
individual blocks. It alters racial and household composition,
changes population counts, and even shifts those fabricated
numbers to entirely different neighborhoods, cities, or
counties.
Census employees will say they are statutorily required to
protect privacy as justification for choosing differential
privacy to achieve that goal. Under differential privacy, the
Bureau insists that State-level totals remain accurate, but
everything below that level is intentionally distorted. The
amount of distortion is Top Secret. Only a handful of Bureau
officials know the true, unaltered numbers. The consequences
are severe and criticism is bipartisan.
Harvard data scientists found that differential privacy
artificially reduces racial and partisan diversity, making it
impossible to achieve genuine one-person, one-vote equality.
The Mexican-American Legal Defense and Education Fund and Asian
Americans Advancing Justice concluded that the data altered by
this method are too inaccurate to satisfy the Voting Rights Act
or the equal population requirements for redistricting.
Rural and low-population areas suffer the greatest
proportional harm, destroying the partisan makeup of Congress
and potentially affecting billions in Federal funding.
Differential privacy also has three fatal structural flaws:
(1) It renders the count question resolution process
meaningless. Localities can no longer tell whether a number is
wrong because of an actual counting error or because the Bureau
intentionally falsified it.
(2) It injects distorted data into every State's
redistricting process, casting doubt on the legitimacy of
district lines.
(3) It effectively gives bureaucrats a permanent tool to
obscure noncitizen and illegal immigrant populations,
nullifying any future citizenship question and shifting
political power away from American citizens.
Some limited analysis suggests an estimated two percent
movement of numerical population data, but some reporting has
suggested that upwards of 20 percent movement is allowable.
Either way, this poses serious constitutional concerns.
Compounding these problems, the 2020 Census was the most
inaccurate in modern history. The Bureau's own Post-Enumeration
Survey admitted statistically significant undercounts in six
States, mostly Red States--Arkansas, Florida, Illinois,
Mississippi, Tennessee, and Texas--and overcounts in eight
States, mostly Blue States--Delaware, Hawaii, Massachusetts,
Minnesota, New York, Ohio, Rhode Island, and Utah.
Those errors alone shifted at least six Congressional seats
and Electoral votes to the wrong States. Those errors all
benefited Democrats. When you combine known political
imbalances in the raw count with an opaque algorithm that
deliberately moves and alters population data, and all those
problems, coincidentally, mathematically tend to benefit the
Left, it is reasonable for Americans to conclude that the 2020
Census was weaponized, if not intentionally, through
malpractice.
Public trust in the Census Bureau has been badly damaged as
a result. Restoring that trust is simple. Congress and the
current and next administration must correct the errors
committed by the Census Bureaucracy leading up to the 2020
Census and permanently prohibit the use of differential
privacy.
Differential privacy policy for the purposes of satisfying
statutory privacy requirements is insufficient as a
justification. Statutory privacy requirements cannot serve as
grounds for ignoring the constitutional requirements that flow
from Census data.
The 2030 Census returns need to be accurate, transparent,
and politically neutral, the way the Constitution demands.
Thank you. I look forward to your questions.
[The prepared statement of Mr. Miller follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Roy. Thank you, Mr. Miller. I'll now turn to Mr. Yang,
and you have five minutes.
STATEMENT OF JOHN C. YANG
Mr. Yang. Thank you. Good morning. Chair Roy, Ranking
Member Scanlon, Committee Chair Jordan, and Ranking Member
Raskin, thank you for inviting me to testify here today.
My name is John C. Yang. I am the President and Executive
Director of Asian Americans Advancing Justice, AAJC.
AAJC is part of a national network that includes
independent affiliates in Atlanta, Chicago, and Los Angeles, as
well as 37 community partners in 37 different States, as well
as the District of Columbia.
We have served on numerous advisory Committees for the
Census Bureau, including the most recent 2020 Census Advisory
Committee. We maintain a permanent Census Program that monitors
Census policy, educates policymakers, and conducts community
outreach and education.
I appreciate this opportunity to testify before all you on
the importance of the accuracy of the census. An inaccurate
census hurts everyone in this country. Over $2.2 trillion in
Federal funding is allocated based on Census data. State and
local governments decide where to build hospitals, where to
build schools, how to widen highways, based on the Census data.
Businesses decide where to put new stores, what inventory to
include in those new stores, based on demographic data provided
by the census.
That's why every administration going back to 1790, whether
Republican, Democrat, Whig, Federalist, and Democratic-
Republican, has supported policies and supported a budget to
fund an accurate census.
Traditionally, ensuring an accurate census has been a
bipartisan endeavor, and it has been about science and has
about statistics. Under the Constitution, the census requires
the counting of the whole number of people. The Supreme Court,
as well as the text and history of the 14th Amendment, makes
clear that apportionment is based on total population and not
subsets.
Arguments about counts based solely on immigration status
or eligible voters have been repeatedly rejected by Congress
and the courts. Any attempt to add a question concerning
citizenship or immigration status will only make the census
inaccurate.
Studies have repeatedly shown that questions such as a
citizenship question only cause people to decline to
participate in the census. To be clear, those people are not
limited to noncitizens, but also include households that
include both citizens and noncitizens.
Such inaccuracies will mean that resources would not be
distributed appropriately throughout the country. It also
contradicts the Constitution's edict to equal representation
for all.
The Census Bureau is committed to producing and
disseminating objective, credible, and useful statistics to
support the decisionmaking of governments, businesses,
philanthropy, and social service organizations.
Although the total population count of the 2020 Census did
not see a significant overcount or undercount, we did see that
specific populations did have undercounts. The 2020 Census
showed an undercount for children under four of over five
percent; of rural areas were undercounted by around four
percent; the Hispanic community was undercounted by over five
percent; the African American community by around three
percent; and the American Indian Tribal population by over five
percent.
Now, two common reasons exist for undercounts. First, anti-
government sentiment and privacy concerns are causing people to
resist sharing information with the government. Less than a
third of the U.S. population have trust in the Federal
Government.
Accordingly, the Census Bureau takes its responsibility to
maintain the confidentiality of the information very, very
seriously by ensuring that personally identifiable information
is not shared or distributed to any other government or any
other agency at any level.
Now, differential privacy is simply one means for
accomplishing that confidentiality. Basically, it creates
statistical noise to mask personally identifiable
characteristics that could be revealed when census data is
overlaid with--against other publicly available data. Now, to
be clear, and everyone has agreed, differential privacy is not
used at the State level, and therefore, has no effect on
apportionment.
The second common reason for undercount is a lack of a
robust get-out-the-count program, especially for communities
that are traditionally undercounted or difficult to count.
States that invest in their census outreach and invest in their
operations are more likely to have accurate counts when
compared to States that did not invest in their populations. A
State's lack of investment in their census, therefore, not only
results in a potential loss of representation in Congress, but
also in a loss of resources.
For example, many experts expected that Alabama would lose
a Congressional seat in the 2020 Census. The State, however,
invested in get-out-the-count operations at an early stage, and
Governor Ivey was very vocal in pushing Alabamians to
participate. As a result, the State did not lose a seat.
In conclusion, it is clear that calling it ``all persons,''
not just citizens or legal residents, is a constitutional
requirement that is also critical for determining the
allocation of Federal funds and for making data based business
decisions. Rather than seeking to exclude certain classes of
people from the count, policymakers and the Census Bureau must
do everything they can to ensure that everyone is counted.
Thank you again for the opportunity to testify. I look
forward to questions.
[The prepared statement of Mr. Yang follows:]
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
Mr. Roy. Thank you for your opening statement, Mr. Yang.
We will now proceed under the five-minute rule with
questions. The Chair recognizes the gentleman from North
Carolina for five minutes.
Mr. Harris. Thank you very much, Mr. Chair, and thank you
for holding this hearing on this very important matter.
Thank you all on the panel for your presence today and for
your expertise that you bring to the table.
I want to take just a moment to talk about the importance
of including a citizenship question in our census. I want to
come to you, Mr. Miller. Should the census include a
citizenship question?
Mr. Miller. Yes, I would argue that it is constitutionally
mandated in terms of what the census is seeking to produce from
a constitutional perspective. Knowing citizenship or residency
status, lawful permanent residence status or otherwise, is
important in understanding how all the total population fits
into what you must do with that data.
Mr. Harris. Thank you. Has the census ever included a
citizenship question?
Mr. Miller. That's a good question historically. Anyone--I
think it has historically but has not recently. They tried to
get that going last time and ran into APA violation issues. It
was not included.
Mr. Harris. A followup to that, because I think it has--and
why was the question removed? That would be the real question
here.
Mr. Miller. Well, are you asking me to speculate? I suspect
that it was an attempt to try to increase population levels in
ways that would help certain States, because they're now
counting illegal immigrants for the purposes of apportionment
and illegal immigrants and legal immigrants for the purposes of
redistricting.
Mr. Harris. OK. What are some of the effects of leaving a
citizenship question out of the census?
Mr. Miller. If you don't know accurate citizen status, then
if your, if your argument is that, constitutionally, you should
only be creating political districts with those that are here,
lawful permanent residents and citizens, or just citizens, then
you need to know that data to create political maps that
reflect that dataset, if that's your objective.
Mr. Harris. Sure. The result of our census not only
controls how Congressional and State districts are drawn, but
it also directs the flow of Federal dollars. We talk about
Medicaid, highway funds, education, and SNAP benefits, all use
population and demographic data, demographic data from the
census, and to award funds. When we don't distinguish between
citizens and noncitizens when determining our population, it
incentivizes States to mass import illegals through sanctuary
policies.
We just went through the Big Beautiful Bill or Working
Families Tax Act, and we made an effort to ensure illegals are
not accessing welfare programs. I'm worried that, when we allow
them to be considered for purposes of determining benefit
formulas, it undermines that purpose.
Mr. Miller, do you agree that including illegal aliens in
population totals for Federal programs undermine the recent
reforms to prohibit illegal aliens from taking benefits away
from Americans?
Mr. Miller. Certainly, there's two questions here: Should
you count them, and then, how should that data be used, and for
what purposes?
Certainly, if you are counting illegal aliens and you don't
know their location, and the location of those illegal aliens,
because of differential privacy, is being moved around, that
would have significant impacts on every single Federal funding
formula, assuming that memorandums of understanding are not
being properly submitted to the census to ascertain the
accurate information on ways to disperse those Federal funds.
Mr. Harris. Agreed. I would submit that, when we allow
illegal aliens to be counted toward the total population for
the purpose of determining welfare award numbers, it distorts
the proper amount that should be given to that community, and
that, ultimately, puts a further strain on, on the Federal
budget.
Mr. Mayfield, let me ask you really quickly in the last
moment I'm very grateful to be serving in the House of
Representatives, which is a democratic body, but I worry that
the errors in the 2020 Census undermine this institution and
its purposes, which is to represent the American people. Can
you, Mr. Mayfield, explain how the way in which the 2020 Census
was conducted undermined the representation of the American
people?
Mr. Mayfield. Thank you for that question, Congressman.
It's important in asking that question to make a
distinction between two things: The gathering of the data, the
collection, and the processing of that data.
The gathering of the data by the census in 2020 was
exemplary. As your colleague Mr. Raskin pointed out, the COVID
pandemic made things extremely difficult. There were also
hurricanes that affected, in particular, Alabama and Louisiana.
The collection was delayed.
The collection was tied to the most successful census in
history. We had a 99.98 household contact rate. I say, ``we'';
I had nothing to do with that; the Bureau did.
There were three primary reasons for that success:
(1) President Trump insisted that the census follow regular
order and not make shortcuts in the census.
(2) The Census had done a good job over the previous decade
implementing technology and cutting out many, many layers of
needless followup, including using administrative records, such
as the Postal Service.
(3) As you may recall, the Census lockdown day was March
18th. As chance would happen it, Census Day is April 1st.
Virtually, everyone was in their home. Once they got done
watching Tiger King, they decided to fill out their census
forms.
The problem that the people here have spoken here today
comes in the Census's data processing. The largest problem was
differential privacy, which, as several speakers have alluded,
moves people around.
In the modeling prior to the census that was done in the
University of Virginia under a Census contract, it, basically--
I'm summing this up--moved people from urban areas to rural
areas, where they did not exist. You can all discern from that,
if that's true, what the political modeling would look like.
Another major problem in--
Mr. Roy. Mr. Mayfield, will you wrap this up?
Mr. Mayfield. Sure.
Mr. Roy. We're a little over time.
Mr. Mayfield. Grouped quarters were a problem because
colleges, in particular, and other large quarters were double-
counted. The Census went back and changed the data to assume
that people--where they thought they were instead of just where
they were when they answered the census.
Mr. Harris. Thank you. I'm out of time, Mr. Chair. I yield
back.
Mr. Roy. Thank you, Mr. Mayfield. Thank you to the
gentleman from North Carolina.
I'm now going to recognize the gentleman from Maryland, the
Ranking Member of the Committee, Mr. Raskin.
Mr. Raskin. Thank you kindly, Mr. Chair.
Mr. Yang, first, Section 1 of the 14th Amendment states
that, ``All persons born or naturalized in the United States
and subject to the jurisdiction thereof, are citizens of the
United States. . . .'' Then, in Section 2, it says that,
``Representatives shall be apportioned among the States
according to their respective numbers, counting the whole
number of persons in each State. . . .''
Obviously, the framers of the 14th Amendment understood the
difference between citizens and persons, and they used
``citizens'' when they meant citizens; they used ``persons''
when they meant persons.
Doesn't that pretty much settle this bizarrely lingering
question of whether or not we should be counting all persons or
just citizens, when we perform the census?
Mr. Yang. That's absolutely correct. Thank you for that
question. If you look at the text of the Constitution and the
14th Amendment, it counts all persons in the State. Frankly,
some of the history, both of the 14th Amendment and going back
to the founding of the Constitution, illustrate that they
wanted to count the entire population and not simply voters,
not simply landowners. Obviously, at that point, women could
not vote; obviously, children could not vote. They want to
include the entire population.
Frankly, this debate also came up in 1929 during what was
called the Apportionment Act. That was the only time in
Congress at the 1920 Census where we did not have
apportionment. In that debate, part of the debate was whether
certain people should be counted. During that debate, at the
end of the debate, everyone decided that it is still based on
all population, but during that debate, some argued that people
should not be counted. Specifically, one Representative argued
that the--a counting, it should count only voters or citizens
because, otherwise, it would include the thousands of
unnaturalized aliens.
Another Representative said that they didn't want to count
thousands of Mexicans and Oriental aliens, and that those
people would end up somehow giving extra seats to California in
that particular case.
When they talk about Oriental aliens, I will take that
personally because at that time there was a Chinese Exclusion
Act, which meant that Chinese were not citizens. Nevertheless,
during that time, it was settled that it should still be all
population for purposes of the census.
Mr. Raskin. Right. Of course, at various points in American
history, the determination to count just voters would have
excluded women, Native Americans, today children still, right?
Will you explain this controversy around differential
privacy which seems to be a new one? Will you explain what that
means, what the practice is, and then, what your response is to
your fellow panelist there, and specifically, to address Mr.
Mayfield's last point that: I think this is inflating the
population of rural communities by falsely attributing people
from urban communities to rural communities, which would seem
to me to be inflating the numbers in Red districts that--
Mr. Yang. No. Thank you for asking me for that question.
Again, the starting point is that differential privacy was not
used at the State level.
Mr. Raskin. It is for--
Mr. Yang. Sure. Differential privacy is, basically, the
notion that you will inject, some call it, ``statistical
noise,'' but masking data in the, the census data. Typically,
you would take a certain piece of data and transfer it to
another individual, so that you can't identify individual A.
There's concern with Big Tech now, with big data. Individual A,
if you have enough information, both from the census and what
is available on the web you might be able to identify that
person. ``Oh, that must be so-and-so who lives on this block.''
The concern was we needed to devise a system that would
take some of the data, switch it around, so that you can't
identify that individual. Again, it was not used at the State
level. Any argument that somehow a State got less seats or a
State got more seats because of it is just not--
Mr. Raskin. Where was it used?
Mr. Yang. It was used certainly at the lower levels, most--
so--
Mr. Raskin. Do you mean at the county or municipal levels?
Mr. Yang. Right. At what's called a Census Block level,
Census Tract level, it was used there. I agree that there could
be concerns about how you use that, but at the end of the day,
everyone that's been doing redistricting still uses that data
that's available. What's it called? PL 90.
Mr. Raskin. When the Trump Administration administered that
or implemented that in the 2020 Census, was that the first time
it was ever used?
Mr. Yang. Yes, that was the first time that it was ever
used.
Mr. Raskin. Uh-hum, OK. Mr. Mayfield, was your point that
we shouldn't be using that anymore? Even if it doesn't affect
Congressional reapportionment, it affects local apportionment
in city council races, county council, things like that?
Mr. Mayfield. To clarify, Congressman, it affects all
redistricting within a State. My friend's point is that, if
Indiana has a population, say, of three million, that's
probably correct within the, the normal parameters of accuracy.
How Indiana's nine Congressmen--I apologize if I've got that
wrong--are allocated is going to be incorrect and down to the
legislative, district, and town levels. That error will
permeate within each State below the State level. When he
refers to ``apportionment,'' he means the number of
Representatives that are allocated to each State after the
census.
Mr. Raskin. With your indulgence, Mr. Chair, if I could
just ask him to finish this point.
Why was this differential privacy thing adopted? What
problem was it addressing, and what would the alternative be to
the problem of not being able to get in particular people's
houses if they don't open the door, or whatever?
Mr. Mayfield. Sure. Well, differential privacy has nothing
to do with collection. It has to do with ensuring that Section
9 of Title XIII is adhered to. In that, Congress has prohibited
the release of personal information based on using the data.
In other words, we don't release the raw data, the short
form that everyone fills out. That is very basic. There are
basically nine questions: Name, phone number, address, racial
characteristic, your relationship to the household head, and
your gender. The only real personal information on that is
name, address, phone number.
The census has always kept the granular data out of the
public eye. You can't get census data that's actually
collected.
In the past three censuses, 2010, 2020--sorry--2010, 2000,
and 1990, the Bureau used what's called ``data swapping.''
Imagine you have a block. There are five houses on each side of
the block. Because the smallest unit of data that is publicly
released is at the block level, the Bureau scrambles that data.
You don't know if there's a 20-year-old living on this side and
a young daughter living on this side.
That was always used. There was never a problem with it. No
one outside of the people driving differential privacy thought
there was a problem, and they advocated for continuing to use
that
method.
Mr. Roy. I thank the Ranking Member for his question, and
I'm going to use a little prerogative, because there's some
technical stuff here, particularly when it's a nonpartisan
point. I don't mind it going over on either side, because this
differential privacy is highly complex, and I do think this
Committee, on a nonpartisan basis, we need to understand it and
understand it fully. I'll try to allow some of that
explanation, as necessary, on both sides, where it's
appropriate.
With that, I would be recognizing the gentleman from
California, my friend, Mr. McClintock.
Mr. McClintock. Thank you. Mr. Rodriguez, if the 14th
Amendment actually confers automatic citizenship to anyone
who's born here, as the Democrats contend, it seems to me it
would have said, ``All persons born or naturalized in the
United States are citizens of the United States.'' Period.
That's pretty clear.
As you know, that's not what it says. It says, ``All
persons born or naturalized in the United States, and subject
to the jurisdiction thereof, are citizens of the United States.
. . .'' What does that mean?
Mr. Rodriguez. Well, the courts have interpreted ``subject
to the jurisdiction thereof '' in other contexts for purposes
of due process or equal protection, and it's certainly been
broader than naturalized or legal residents. It's still--
Mr. McClintock. Has the Supreme Court ever rendered a
decision on whether it applies to those who are illegally in
this country who have never sworn allegiance to it, and have
defied its laws to be here in the first place, and under its
laws, are subject to deportation as soon as encountered?
Yes, I realize the Ark Decision was rendered 127 years ago,
but that applied to legal immigrants who accepted the
jurisdiction in the United States by obeying our immigration
laws and who had taken up legal permanent residence, subject to
a treaty ratified by the Senate.
Mr. Rodriguez. Right. There seems to be no directly on-
point Supreme Court decision that says that the children of
illegal immigrants are or are not citizens.
Mr. McClintock. Perhaps we're due to have that at some
point decided by the Supreme Court or decided by the people
through constitutional amendment or statute. Am I correct?
Mr. Rodriguez. It does seem like it will be heard in the
near future.
Mr. McClintock. Mr. Mayfield, it seems to me that a census
is neither science nor statistics. A census is counting real
people. People have names, dates of birth, addresses, and other
identifying information. How can it possibly be that we include
in our census any count without names and such identifying
characteristics? It seems to me, if you don't have a person,
they shouldn't be counted in a census of people. How did we
deviate from this simple principle?
Mr. Mayfield. If I understand your question, Congressman,
the first aspect of the census is voluntary compliance with it.
Every American is required to answer it by law, but the deal
that Congress has made is that this information will not be
disclosed. In fact, they cannot even be subpoenaed by a court.
That's just how tight the information is kept.
Approximately 67 percent of Americans voluntarily filled
out the short form, which is: What is your household data as of
April 1, 2020? For the remaining third of the country, the
census endeavors to learn that information. It does so--
Mr. McClintock. You knock on the door. You get a count of
the people who actually live there. You enter them on the
Census Register. That's been done since time in memoriam. I'm
reading now of statistical errors that shifted the--a portion
of Congressional seats away from Republican States toward
Democratic States. Now, how can that be if we're actually
counting real people?
Mr. Mayfield. It's important to remember, Congressman, that
several Members have alluded to so-called errors after the
fact. The Census does what's called the Post-Enumeration
Survey. That is modeling. That is statistical guess, is what
it--
Mr. McClintock. Well, how did we reach a point where we're
guessing in the census? Either you're a real person or you're
not. If you're not, you don't belong in the census. If you are,
we will count you.
Mr. Mayfield. As Justice Scalia pointed out in Department
of Commerce v. The House of Representatives, it's imperfect
science enumeration, but it's better than anything else we've
got.
Mr. McClintock. It's not. It's not science; it's a simple
matter of counting real people.
Mr. Mayfield. Well, Congressman, there are many people who
are very hard to reach. For instance, homeless people, which
the Census works very hard to count. There are people who won't
allow--
Mr. McClintock. You work very hard to find them, but if you
can't find them, you don't make them up. That's not a real
census. That's guesswork. Mr. Miller, any thoughts?
Mr. Miller. Yes. It is extremely important that the data
that is collected, at least the sensitive, Top-Secret version
of this data, be highly accurate and that we don't just invent
people. Because former processes have just swapped people. The
new process literally deconstructs a person and every
characteristic, and then, just in the aggregate, creates them
elsewhere in a whole bunch of different profiles all over the
map. It's a complete mess.
There's all sorts of constitutional problems with the way
we've done it that have impacted redistricting, and there are
some arguments that, if you don't want to count illegal aliens
for the purposes of apportionment, because your argument is
that they are not under the political jurisdiction of the
United States of America at the time of birth, that this would
impact apportionment because, if you're using differential
privacy, you may not know--we don't know what they're doing
with that characteristic data. Is it actually adding up all the
way to the top accurately or not?
The biggest problem with all this is we assume it is. It
may not be impacting apportionment, but it's all Top Secret. We
don't have eyes on it. We can't verify it. No one can go right
now and say that, 100 percent, yes, all this data, even the
characteristic data, let alone the population data, is
accurate. That's a huge problem.
Mr. Roy. I thank the gentleman from California. I will now
recognize the gentlelady from Washington, Ms. Jayapal.
Ms. Jayapal. Thank you so much, Mr. Chair.
I think we all agree on the importance of accuracy in our
census results. I am concerned about some of the actions that
are being advocated for are actually undermining some of that
accuracy.
I want to turn to the citizenship question first and put
aside the Constitution, which is a strange thing to say in the
Constitution Subcommittee, but to not necessarily look at what
the Constitution says, specifically, persons rather than just
citizens, and actually talk about how adding a citizenship
question would make the census less accurate.
I'm going to come to you Mr. Yang because there is a lot of
data that shows that communities of color, including American
citizens, have been historically undercounted. This
undercounting question is directly related to what might happen
if a citizenship question was added to the census.
Can you speak to how adding a citizenship question actually
hurts us in terms of accuracy and increasing undercounting?
Mr. Yang. Thank you for that question, Representative.
The bottom line is we want an accurate census. All the
studies that we have seen, included by the Census Bureau, show
that the addition of a citizenship question will undermine the
accuracy of the census, not only for immigrants, but for all
Americans, because there's a hesitancy about how that data is
going to be used.
That was documented, especially and even going up to the
2020 Census. Even after the Supreme Court decision that said
that we cannot have a citizenship question, there are still
people that were concerned about whether that question
appeared. That is part of the reason why many believe that
Hispanic Americans were undercounted by over five percent.
The other thing that we should remember about this,
actually, there is a citizenship question on what's called the
long form, that American Community Survey of the census. Now,
that is a form that contains 50 questions. It has been tested
in terms of how we can do that in a way that captures the data
accurately. Again, the bottom line is accuracy.
Ms. Jayapal. Let me go to the question of undercounting. If
there is undercounting, substantial undercounting, how does
that undercount harm American citizens in terms of how the
census is utilized?
Mr. Yang. Well, the undercount affects American citizens,
then, because we will have areas where there's more population
than was expected, even though the roads are not wide enough,
even though the schools are not big enough, even though the
public services are not big enough.
Let's be clear, right, highways, public services aren't
dependent on immigrant status. It's dependent on the actual
residents that live in that community.
Ms. Jayapal. Give me some of the other reasons, for anyone
that's listening to this debate and trying to understand it,
for why noncitizens should be counted. Again, we're clear about
what the Constitution says, but go ahead.
Mr. Yang. Right. Again, why noncitizens should be counted
is because they are part of the entire community that we're
talking about. First, we break it down in terms of noncitizens.
Obviously, that would include people that are legal permanent
residents. That would include people that are refugees. It
includes a large body that participated in that community, that
deserves to have a voice in that community, and nevertheless,
has not yet attained this notion of citizenship.
Historically, that's been part of the debates that we've
had, whether it was for the 14th Amendment, the original
founding of the Constitution, or as I pointed out, in the 1929
Apportion Act. We've always agreed with the general principle
that we should represent the entire population and not just a
small subset of that population.
Ms. Jayapal. Yes, that it's very real that we have many,
many mixed-status families in this country. If you insert fear
into one group of people, that certainly affects the entire
family's trust in government.
Mr. Yang. I appreciate that you bring up the notion of
mixed-status families. What that means is you have a family
that has some citizens, some noncitizens, maybe even someone
that is undocumented. In that situation, that citizen in that
household would not get the services that they would be
entitled to because of the fact they are not being counted
correctly.
Ms. Jayapal. I wanted to go back to differential privacy,
because, as the Chair knows, this is an issue where we all have
a lot of concerns about surveillance. I guess I want to
understand why we would want to eliminate that ability to
really mask the characteristics, so we can get accurate
assessments. I wonder if you want to speak to that.
Mr. Yang. Yes, the bottom line here is, how do we guarantee
confidentiality for the community, so that the community can
trust that, when they are responding to the census, their
identity will not be made public?
The reason this differential privacy policy was developed
was because of the advent of big data.
Ms. Jayapal. Uh-hum.
Mr. Yang. Now, you are able to get so much more information
about individuals off the internet.
Ms. Jayapal. Uh-hum.
Mr. Yang. That what might have been unidentifiable before
by simply swapping data at that block level is not going to be
enough. That is why this was developed. I agree that we all
have concerns, we've asked questions about how this is
implemented, and perhaps some fine tuning is appropriate, but
that is a proper source of inquiry. The notion of disregarding
it altogether I think would--
Ms. Jayapal. What would be the fine tuning?
Mr. Yang. This is where we all need to continue to work
with the Census Bureau to see what worked and what didn't. The
Census Bureau, when they said that they were announcing this
differential privacy policy, among other things, they said
that, for 95 percent of the jurisdictions, they expect an error
rate of under five percent. In that context, then you start to
understand, OK, this data at these levels still should be
accurate.
Ms. Jayapal. Uh-hum.
Mr. Yang. Now, I would agree we need to test that and to
see where those five percent did not work; what that actually
looks like. I think that's an appropriate area for inquiry.
Ms. Jayapal. Great. Thank you. Thank you, Mr. Chair.
Mr. Roy. I thank the gentlelady for her questions. I will
now represent--Mr. Onder?
Mr. Onder. Thank you, Mr. Chair. I couldn't agree more with
President Trump that we need to restore the integrity of our
census, and that this begins with refusing to count illegal
aliens in our census. For too long, sanctuary States have
openly defied Federal laws and been rewarded with more
political power, more Congressional seats, and more influence.
This is unacceptable and it undermines the very foundation of
equal representation.
Meanwhile, States like Missouri that I represent, that
follow and respect the law, have a diluted voice in Washington,
and not because we have done anything wrong, but because other
States have chosen to disregard Federal immigration law, and
they benefit politically by doing so.
Mr. Rodriguez, doesn't this mean that sanctuary policies,
if we count illegal aliens in the census, have the effect of
giving Democrats an unfair advantage in Congress, in the
Electoral College, and in Federal funding?
Mr. Rodriguez. To some extent. I don't think it's entirely
partisan, though. There are Red States like Texas which also
have disproportionate levels of illegal aliens in their State,
and they also benefit--
Mr. Onder. Right.
Mr. Rodriguez. Sure, California has been the State that has
benefited most from this policy.
Mr. Onder. Yes, but Texas does not have a sanctuary State
policy, of course.
Mr. Rodriguez. Yes, of course.
Mr. Onder. If illegal immigrants are excluded from
apportionment counts, wouldn't we see some shift in political
power back to States that follow the law?
Mr. Rodriguez. Yes.
Mr. Onder. Yes, because census results also determine how,
literally, trillions of dollars of Federal dollars are
distributed, are my constituents in Missouri subsidizing
sanctuary States with large illegal alien populations?
Mr. Rodriguez. They likely are.
Mr. Onder. OK.
Mr. Rodriguez. It's important to note that, under Federal
law, the Personal Responsibility and Work Organization
Reconciliation Act of 1996 prohibit illegal aliens from
receiving Federal benefits. Calculations for Federal benefit
purposes that are based on illegal alien populations are wrong
under the law.
Mr. Onder. Right. Again, we know that, looking back on the
2020 Census, it is clear the Census Bureau has admitted that
there were errors in States, in the counting of States, and
five of the six States that were undercounted were Red States
and seven of the eight States that were overcounted were Blue
or swing States.
Mr. Miller, can you explain how these inaccuracies in the
2020 Census occurred and what can be done differently in the
2030 Census?
Mr. Miller. The Census Bureau would argue that these were
difficulties that they encountered due to COVID, and that this
was kind of a historical unprecedented event that occurred
during a census, at least in modern times, that I'm aware of.
Mr. Onder. Excuse me if I interrupt. It seems like
everything is blamed on COVID.
Mr. Miller. Right.
Mr. Onder. Supply chains are blamed on COVID; educational/
lack of educational is blamed on COVID; workforce, COVID.
What specifically--how did fighting this virus lead to
these inaccuracies?
Mr. Miller. Well, that's why it just doesn't make much
sense to me, because the States that were undercounted were,
generally, more open than the States--
Mr. Onder. Where people wearing gloves and masks and not
able to be out properly?
Mr. Miller. Right.
Mr. Onder. I don't know.
Mr. Miller. Yes, the States that were overcounted were the
ones that were locked-down and closed, more so, I just don't
understand the whole COVID logic. I'll take them at their word,
but it's certainly a strange outcome.
Mr. Onder. Yes. Maybe everyone at the Census Bureau working
from home and not actually out counting persons.
Do we really know the full extent, Mr. Miller, of the 2020
Census, the census inaccuracies? If they admit to this, is--
Mr. Miller. No, they have admitted to major miscounts.
Mr. Onder. Yes.
Mr. Miller. We do not know the extent to which differential
privacy has had an impact, because the TIGER file that all this
exists at the Census Bureau is Top Secret. Only a few people
have access to it. This is an oversight problem that we really
need to get some data statisticians in there to really analyze,
with clearance, to analyze the impacts. Because whether it's
urban to rural or rural to urban, there has been a lot of
population movement, and I've seen studies arguing both sides
of this. There are big problems with this on a variety of
fronts.
Mr. Onder. Again, remind us what's the alternative to the
differential privacy scheme?
Mr. Miller. Well, there can be a lot of these. IFMS has a
really good analysis on how to address this. Past practices
have used swapping, limited swapping.
Mr. Onder. Right.
Mr. Miller. One solution is just turn off characteristic
data until you get to a statistically significant level--
Mr. Onder. Level where you can--
Mr. Miller. --of numerical data available, but turn off all
the characteristic data until you're up at the census tract or
county level. That way, it's really hard to reverse-engineer
the algorithm with population numbers that are strong.
Mr. Onder. It makes perfect sense. Thank you so much. I
yield back.
Mr. Roy. I thank the gentleman.
I will now represent the gentlelady from Vermont.
Ms. Balint. Thank you, Mr. Chair. Thank you to the
witnesses. I appreciate your time.
This is a hearing about the census, so, essentially, a
hearing about counting people, or put another way, determining
who counts and who doesn't.
President Trump has made clear that he feels that
birthright citizens is not constitutional. He feels that
noncitizens should not be counted in the census. He has made
clear that he views the 2020 Census as illegitimate.
The truth is the President can feel any way he wants; the
courts have spoken. Birthright citizenship is the law of the
land. The census counts for everyone, and the 2020 Census was
as accurate as the 2010 Census.
Here we are again in a hearing to make these claims, these
nonsensical legal theories more legitimate by holding hearings
about this. More broadly, this hearing props up baseless legal
theories that undermine the census and threaten our
representative democracy.
I want to associate myself with the remarks of my
Democratic colleagues. I would like to take my time to go in a
little bit different direction here. I want to spend my time
getting to the heart of this question of who counts and who
doesn't count in the eyes of this administration. Who is seen
as a real American and who is not?
Investigative report by ProPublica has found that more than
170 American citizens have been detained by immigration agents.
They've been dragged; they've been kicked; and they have been
detained for days at a time. These are U.S. citizens. They are
our fellow Americans. Top Trump Administration officials keep
denying that U.S. citizens have been arrested or detained at
all in the immigration raids that are taking place across this
country.
Department of Homeland Security Secretary Kristi Noem has
repeatedly falsely claimed that no American citizens have been
arrested or detained. There are two reasons that we have that
she could be doing this:
(1) That she is blatantly lying to the press. That is one
option.
(2) That she doesn't know what is happening under her own
leadership.
Either option is deeply disturbing to me.
We have so much evidence that what Noem has been saying
about this is untrue. Many news outlets have confirmed these
arrests and have spoken directly to U.S. citizens who have been
arrested or detained by immigration agents. They have gathered
these disturbing stories about how our own citizens are being
treated by this administration. We have the evidence.
There's the case of American citizen George Retes, an Army
veteran who was pulled out of his car by Federal agents who
were raiding a farm North of Los Angeles. In his words,
They sat me down on the dirt with my hands zip-tied behind my
back. I told them I was a citizen. I told them I had an ID.
They didn't care. They never asked me for it.
This U.S. citizen was held for three days before finally
being released. He was never charged with a crime. He's just
one of many U.S. citizens that we can document who have been
detained or arrested. Again, we have the proof. Secretary Noem
continues to spread false information.
Then, there's the case of Jason Brian Gavidia. Jason is a
U.S. citizen who worked at a tow truck yard when Border Patrol
agents stormed in. Jason was handcuffed, as he pleaded with the
agents. He said, ``Let me show you my ID. Let me prove my
status.'' He said, ``They didn't care. All they cared about was
the fact that I was Brown and I was in the wrong place at the
wrong time.'' That's not right. We're in America. That's not
right.
The point is that the only people who seem to count as
Americans to this administration are the people who look
American. Any of us sitting here in this room, regardless of
party, would be shocked, would be outraged, would express
righteous indignation if we were detained based on our physical
appearance or on our profile alone. Regardless of what we
believe about this President, I would hope that we would agree
we would howl at our due process rights being denied.
Mr. Raskin. Well, would the gentlelady yield for a quick
question?
Ms. Balint. I will. I will.
Mr. Raskin. What does it mean to ``look like an American''
in 2025?
Ms. Balint. Yes, thank you very much. It seems clear to me
that looking like an American is somebody who does not have
Brown or Black skin, and sounding like an American is someone
who does not have an accent. That seems like that's been
established.
I just want to end with this: There is nothing more
important than a person's liberty, nothing. I would hope that
we can all agree on that.
We should all be standing up to our government when it is
arresting and detaining citizens illegally and then lying about
it. Citizenship in this country is not based on the color of
your skin; it is not based on the accent that you have. Based
on the actions of this administration, I'm not confident that
it believes this, and I find this deeply chilling.
Before I yield back, I have several unanimous consents that
I would like to enter into the record.
First, ProPublica, ``Immigration Agents Have Held More Than
170 Americans Against Their Will.''
Second, ``NPR Fact Checks Kristi Noem on ICE Detaining U.S.
Citizens.''
Third, an article in The Journal of Policy Analysis and
Management, ``Citizenship Question Effects on Household Survey
Responses.''
Fourth, an October 2015 report by The Congressional
Research Service entitled, ``Apportioning Seats in the U.S.
House of Representatives Using a 2013 Estimated Population,''
concluding that excluding noncitizens for apportionment
purposes likely requires a Constitutional Amendment.
I yield back.
Mr. Roy. Without objection, those will be entered into the
record. I will now recognize the Chair of the Judiciary
Committee, Mr. Jordan.
Chair Jordan. Thank you, Mr. Chair. Mr. Mayfield, we've
asked about it before, haven't we? We've asked the citizenship
question on this census before, haven't we?
Mr. Mayfield. That's correct, Congressman, many--
Chair Jordan. From 1820-1950, we asked that question.
That's 130 years. From 1970-2000, another 30 years. One hundred
fifty years, we asked the question.
Mr. Mayfield. That's correct.
Chair Jordan. The previous speaker just said, if we ask the
citizenship question, it's going to threaten democracy. Well,
it didn't hurt democracy for 150 years. In fact, I would bet if
you--I'll use my--I'll just our district. You can go to Urbana,
Ohio, our hometown. I bet if I went to Urbana, Ohio; I walked
to the post office--or walked anywhere, any store. I walk in. I
see someone I know:
Chair Jordan. ``Mr. Smith, do you think on the census--you
know what the census is, of course?''
Mr. Smith. ``Oh, sure. Every 10--every so often, we find out
how many--we count the number of people in the country.''
Chair Jordan. ``Mr. Smith, do you think, when we're counting
the number of people in this country, it's appropriate to find
out how many are citizens?''
What do you think he would say? What do you think the average
person would say?
Mr. Mayfield. I suspect the answer would be, ``Yes, sir.''
Chair Jordan. It would not only be yes; they would say,
``Well, aren't you already doing that? That's pretty stupid if
you're not.''
That's where the American people are. Everyone's for
counting it. We did it for 150 years of our 250 years as a
country--or since we declared independence, 249 years, almost
250 years now. We've done it for most of our history. Everyone
thinks we're doing it. Everyone thinks we should do it, except
Democrats in Congress.
It was interesting, too, in Mr. Yang's testimony, he said,
``Oh, if we ask the question, we'll get inaccurate counts.''
Then, he went on for the rest of his five minutes and talked
about all the inaccuracies in the recent censuses where we're
not asking it--undercounts, overcounts, rural America, Hispanic
Americans, pointing out all the problems, and we're not asking
the question.
Somehow, if we ask it, oh, it might get worse because we
have some study. He even said, people thought it might be on
the census questionnaire, and it wasn't, and that hurt the
count--which makes no sense to me. It wasn't a question.
Think this is so darn simple. Count the persons. That's
what the Constitution says. Count the number of people, and
while you're doing that, find out how many are citizens. What's
wrong with that? It might actually help policymakers when we're
determining policy if we know those two numbers. That's all we
need to focus on.
I'll go to Mr. Miller, if he's got any comments about what
I commented on.
Mr. Miller. On the last part, I absolutely agree. There is
some question on who to count. I think that this, if you get
into it technically, answer is you can count everyone. You just
need to know certain population numerical data, then, residency
status, because that could bleed into policy questions later.
That that's an important distinction to know, that you can use
different datasets within the totality of that data for
different purposes.
Chair Jordan. Of course, that's useful information for
policymakers. That's why we do it that way. I get that.
Let me ask you this question, Mr. Miller: Do you think
asking a citizenship question on the census is any way a threat
to democracy?
Mr. Miller. I do not. I think it's constitutionally
mandated if you really--
Chair Jordan. Mr. Rodriguez, do you think, as a
constitutional scholar here, do you think that's somehow a
threat to democracy if we ask how many people are here and how
many citizens are here?
Mr. Rodriguez. No, it's entirely consistent with
Congress'--
Chair Jordan. Mr. Mayfield, do you think that's a threat to
democracy?
Mr. Mayfield. The census has no effect on democracy,
Congressman.
Chair Jordan. Exactly. Mr. Yang, do you think asking a
citizenship question, while you say it might impact the
accuracy--I don't know that I'd agree with that. I, frankly,
disagree. Do you think that might threaten democracy in some
manner?
Mr. Yang. It's a threat to accuracy, which means that it
would be a threat to how we allocate funds in the United
States. That's how--
Chair Jordan. That's not what I asked you. I said, ``Do you
think it's a threat to democracy?'' because that's what my
Democrat colleague said just four minutes ago.
Mr. Yang. How people are being represented, and the funding
that they get, would be a threat, yes.
Chair Jordan. You think asking a citizenship--the Democrat
witness thinks asking a citizenship question on the census is a
threat to democracy. Do you think that's true?
Mr. Yang. As currently constituted, it could be, yes.
Chair Jordan. Well, that's amazing. That's amazing. Finding
out the number of people and the number of citizens in our
country is somehow a threat to democracy, that's what the
Democrat witness said in the House Judiciary Subcommittee
hearing on the Constitution. That is amazing.
Mr. Chair, I yield back.
Mr. Roy. I thank the Chair of the Judiciary Committee. I
will now represent--the gentlelady from California?
Ms. Kamlager-Dove. Thank you, Mr. Chair.
Apologies if this question has already been asked, but I
like to hear things with my own ears. Mr. Yang, what is the
census sort of used for? Why do we conduct this onerous
process?
Mr. Yang. The principal use for the census is twofold.
First, is for the apportionment of, Congress, seats in
Congress, and second, again, accurate count of the population.
When we're distributing funds, whether it's $2.2 trillion in
Federal funds or State and local funds, that we're distributing
the right way. What that means, literally, is where roads are
built; where schools are built; what services are needed in
different communities. For businesses, what that means is what
stores get built and what you stock on those shelves of those
stores.
I live in Fairfax County. The fact that our Costco has
approximately seven different, seven different types of rice is
a reflection of the fact that Fairfax County is about 20
percent Asian American.
Ms. Kamlager-Dove. Uh-hum. Thank you for that.
I would agree with that, right? Someone who lives in a
city, a county, and a State, we're always fighting to get the
resources that we need. It's important to know how many people,
regardless of where they come from, regardless of who they are,
what they look like, right, what language they speak, what rice
they like, it's important to know how many people are in a
place, so how many hospitals to have; how many schools to have;
and how many parks to have.
Am I missing something here?
Mr. Yang. No. It is as a very basic as that. Again, the
point is to get an accurate count.
One of the things that we've been talking about here is, so
what affects accuracy? We've talked specifically about Texas
and Florida as being inaccurate counts. Well, I would point out
that, in the case of Florida, the Federal--the State government
only allocates $300,000 to the count for the entire State
compared to the State of Alabama, a much smaller State, which
allocated $1.2 million to the count of their residents.
How you do your get-out-the-count operations will translate
into how accurate the results will be. There is a correlation
there.
Ms. Kamlager-Dove. I do understand that it seems like my
Republican colleagues want to get rid of every single foreign-
born person in this country. It boggles the mind because--
Ms. Hageman. I would object. I think that this is
attacking--
Ms. Kamlager-Dove. It's--
Ms. Hageman. I think that is attacking individuals--
Ms. Kamlager-Dove. Well, you've been taking my time. I'd
like to have my time--
Ms. Hageman. I think that is attacking individuals--
Ms. Kamlager-Dove. Point of order.
Mr. Raskin. Mr. Chair, could you restore the order, if you
will?
Mr. Roy. Well, does the gentlelady have--
Ms. Kamlager-Dove. I made an opinion. I can have an
opinion.
Mr. Roy. Is the gentlelady asking to strike--
Ms. Hageman. I am. I'm asking to strike those words,
accusing us of that level of racism.
Ms. Kamlager-Dove. It's not racist. If you think it--it's
my opinion. I have a right to have an opinion.
Mr. Roy. We'll suspend for one second. Will the gentlelady
express the specific words you're asking to be--
Ms. Hageman. Specifically, what she stated was that the
people on the other side wanted all foreign people to be
removed from the country, and that's absolutely untrue.
Ms. Kamlager-Dove. As I mentioned, Mr. Chair it is my
opinion.
Ms. Hageman. That is attacking every person on this side of
the aisle.
Ms. Kamlager-Dove. I have a right to my opinion. I said,
``I believe''--
Ms. Hageman. I am tired of all the allegations of racism.
Mr. Roy. Order. Order.
Mr. Raskin. Mr. Chair, that's not--
Mr. Roy. Order. Let me suspend for a second.
Ms. Kamlager-Dove. I didn't use race at all.
Mr. Roy. Let me suspend for one moment.
Mr. Raskin. Mr. Chair, if I could, that's not engaging in
personalities. It's not an ad hominem attack.
Ms. Hageman. Yes, it is.
Mr. Raskin. If she disagrees, she can use her own time to
explain why she disagrees.
Mr. Roy. Everyone suspend for one moment.
[Pause.]
Mr. Roy. The gentlelady from Wyoming, are you withdrawing
the request to strike the words from the gentlelady from
California?
Ms. Hageman. Yes. Yes. I just wanted to point out the
allegations are absolutely unsubstantiated and it is ridiculous
to--
Mr. Roy. OK. I'd yield--
Mr. Raskin. Mr. Chair, could you instruct her to point it
out on her own time?
Mr. Roy. Reclaiming the time, but, the gentlelady from the
California, please proceed. We will restore the clock.
Ms. Kamlager-Dove. Thank you, Mr. Chair. I guess I don't
want to use my own time saying, but I guess now perspectives
and opinions are no longer important here in this Committee.
That seems like it would be a shame.
Mr. Yang, we are talking about the census, and the question
is--it kind of doesn't matter where you come from. If you are
someplace, if you are using resources, it would be important
for that jurisdiction, for that municipality to know that you
are there. Where I come from, if you get sick, you should go to
the hospital. The hospital is not supposed to say, well, who
are you? Let me see all your paperwork. That doctors take a
Hippocratic Oath to help anyone who is coming in bleeding who
would be wanting to need or use the resources that are
available to them in that hospital. Do you understand sort of
question that I am asking about why it is important just to
know who is in a particular place?
Mr. Yang. That's absolutely right. With respect to public
services and with respect to roads, that immigration status
does not matter when it comes to the need for good roads and
the need for good public services. Making sure we have an
accurate count of our entire population--and again, if you even
go to Supreme Court precedent in Evenwel, they recognize that
you should be doing total population counts, whole population
counts for purposes of apportionment. The history of this has
been very clear that the census is designed to count the entire
population.
Ms. Kamlager-Dove. Thank you for that. As I look at all my
questions, not one word in any of these questions is the word
race. Mr. Yang, Assistant Attorney General Rodriguez says in
his written testimony that apportionment should not account for
those in a State unlawfully. What would be your response to
that?
Mr. Yang. Well, I don't see anything in the Constitution
that makes any sort of distinction. Rather, what the 14th
Amendment provides in the census--in terms of apportionment in
Article 1, Section 2, with respect to census is the counting of
whole persons, and it makes no distinction.
Now, in the written testimony it makes reference to the
fact that somehow they should be part of a political community,
but the notion of political community does not exist anywhere
in the text of the Constitution, the 14th Amendment, or the
Apportionment Act. Certainly, even the contemporary
dictionaries that he cited, except for a reference of parish
law, did not make any reference of being part of a political
community. Rather, simply, a resident is someone that has a
normal place of domicile.
Now, to be clear, there are rules that the census with
respect to people that are transient, with respect to people
that have tourist visas, for example, should not be counted in
the census. There are residency rules with respect to census as
to what is considered a resident.
Ms. Kamlager-Dove. You mentioned whole person in the--I see
the constitutional framers use the phrase whole number of
persons rather than citizen or voter. Why do you think they did
that? Because of course when you say whole person, of course,
this time I am going to use race, but I am going to talk about
myself. There used to be a time when a Black person was not
even considered a whole person. It was like three-fifths of a
person. I am getting a little triggered by that phrase. Can you
explain sort of why the framers used that phrase?
Mr. Yang. You're correct, representative. Unfortunately,
when the Constitution was first adopted slaves were not
considered citizens--for purposes of a census was considered
three-fifths of a person, as a distinguish from a so-called
whole person. It wasn't until the 14th Amendment that
citizenship was given to slaves and anyone born of slaves, and
that making clear that when we are counting, we are still
counting all persons in the State.
Ms. Kamlager-Dove. Tell me again, can you reiterate for the
folks that can't seem to maybe hear what you are saying about
why it is important to have an accurate count when you were
doing the census?
Mr. Yang. Well, having an accurate count for persons of
data allows us to make sure that we have the services that we
need in the United States for everyone. We need a society that
has infrastructure for everyone. It is true for immigrants,
nonimmigrants alike. Certainly, there can be policy debates
around different aspects of that but at least having the set
accurate data that all us can work from is critical to how this
Nation functions.
Ms. Kamlager-Dove. Thank you, Mr. Yang. Mr. Chair, I yield
back.
Mr. Roy. I thank the gentlelady from California. I will now
recognize the gentlelady from Wyoming.
Ms. Hageman. Well, first, I disagree with the idea that
there is a constitutional requirement for conducting a census,
so that we can have different varieties of rice at Costco. I
almost want to say that almost all countries conduct a census.
Ms. Kamlager-Dove. Mr. Chair, that is racist.
Mr. Roy. Does the gentlelady have a--
Ms. Kamlager-Dove. Yes, that should be stricken.
Mr. Roy. Does the gentlelady from Wyoming have a response?
Ms. Hageman. Rice, or hamburger, or any other foodstuff.
That isn't why we conduct a consensus.
Mr. Raskin. Mr. Chair, those words, as ill-advised and
foolish as they were, were probably not engaging in
personalities, we can let that go and call it even after--
Ms. Kamlager-Dove. Yes, Mr. Chair, I will withdraw.
Mr. Roy. Proceed.
Ms. Hageman. I was using the terminology that Mr. Yang
used. If you weren't listening.
Ms. Kamlager-Dove. I was listening but I was also getting
your vibe.
Ms. Hageman. OK. Mr. Roy. The gentlelady from Wyoming will
proceed.
Ms. Hageman. Almost all countries that conduct a consensus
actually have a citizenship question, and the U.N. says that it
is best practice to include a citizenship question in these
censuses. Counties that ask about citizenship are Canada,
Australia, Ireland, Germany, and Mexico. The source of this is
The New York Times.
Mr. Mayfield, I would like to focus on the Census Bureau's
decision to utilize the differential privacy methodology for
the first time in the 2020 Census.
I am incredibly concerned that intentionally creating
structural inaccuracy census data collection at the lowest unit
has impacted the reliability of nationwide census counts,
Congressional reapportionment, assignment of Presidential
electors, redistricting, and Federal funding allocations.
In light of these concerns, could you briefly explain the
steps that the Census Bureau took when adopting and
implementing this differential privacy within the 2020 Census?
Mr. Mayfield. To put it in shorthand, Congresswoman, this
was done at the civil servant level without any legal opinion
being obtained. The decision was made by John Abowd, who is the
Director of Privacy and Statistics, and a small group of people
working with him. There was actually opposition from other
parts of the bureau because the test repeatedly demonstrated
that it would falsify data, including the block level that you
mentioned, but anything below the State level.
They proceeded with that because they wanted to do
differential privacy, not because there was any need for
differential privacy. Data swapping, which was the method used
in the previous three censuses, had always been adequate, and
to this day there is still no problem with it. There is a 2023
study by researchers from Harvard, NYU, and Columbia that finds
that data swapping is just as good as differential privacy, but
it doesn't alter the data. There was no reason to do this.
Now, to give you an example of just how bad it is, this is
post-use of differential privacy--that study found the use of
differential privacy, for instance, if you had three Hispanics
on one block, you create a variance anywhere in the State of
zero Hispanics or six Hispanics. In other words, 100 percent
difference either way.
All the data, as Mr. Miller alluded, below the State level
has been subjected to differential privacy. Because of the way
the algorithm works--it's a top-down algorithm--none of the
data can be released in unaltered form. You cannot actually get
the results of the survey below the State level. You may know
there are nine million people in Indiana, but you don't know
how many people there are in Muncie, you don't know how many
people there are on Mayfield Street in Muncie.
Ms. Hageman. They cooked the books, and they did so
intentionally. Is that a good way of describing it?
Mr. Mayfield. That is accurate.
Ms. Hageman. All right. They didn't follow the
Administrative Procedures Act, they didn't put this out for
notice and comment, they didn't solicit information from
people, statisticians, economists, States, or local governments
about how best to do the census of counting their people to do
as Mr. Yang says, which is to know how many people live in our
country. They didn't do any of that?
Mr. Mayfield. It's even worse than that, Congressman and
Congresswoman. Excuse me. They didn't do notice and comment.
There's a debate about whether or not that's required. There
are numerous Committees to allude to something that
Congresswoman Scanlon mentioned in her opening statement the
Census interacts with consumers of census data and partners,
including State and local governments, and including interest
groups. Every single statistical group objected to this by
letter. Every single State.
The National Conference of Legislators objected. Maine
objected. It's a Blue State. Utah objected, a Red State. The
city of Alexandria, just across the water here, very Blue,
objected. Maricopa County, very Red, objected. The House
Progressive Caucus. They were right. They objected. The Mexican
American Legal Fund. They objected. The Native American Tribes
objected. Everyone knew--it was well-known that this would not
work. The Census civil servants, a small group of them, went
ahead with it because they wanted to do it, not because there
was any need, and disregarding the fact that it made the census
inaccurate the enumeration stage.
Ms. Hageman. They did it because they wanted to cook the
books and they wanted to manipulate the data to have the
outcome that we actually had, which is overcounting people in
six Blue States, or liberal States, and giving them additional
Members in the House of Representatives. Is that fair?
Mr. Mayfield. If I might respectfully disagree with that
part, I believe it was done out of professional pride because
there was an obsession with using differential privacy. Several
other speakers have alluded to the use of internet data and
supercomputers, which is what John Abowd mentioned.
Congress does not exclude the use of outside data. If
anyone here wants to grab census data and buy something from
the internet, or buy a credit card company's data and combine
them, that creates no violation. There is no problem with doing
that. Section 9 of Title XIII bars only the disclosure of
personal data; in other words, your census form that you filled
out for your house.
The errors that have been attributed with respect to the
number of people are based on post-enumeration surveys. In
other words, the Census does quality checks after the census.
Those are models. They are not themselves enumerations. They
are no good--are no better than the models and the data that's
put in them. There are errors after every census that are
believed.
What you do is use that data to try to focus on your next
census and all your other data collections to make sure that
you don't miss other people, that you don't overcount them. It
may be an educated guess that some States received more people
and some States received fewer, but it is a guess. It is
certainly no better than the modeling that the people in the
bureau did.
Ms. Hageman. OK. I know that I am out of time. What I hope
comes out of this hearing, however, is that we have solutions
of what Congress can do to make sure that these inaccuracies
are addressed and that we never have this kind of mess again.
I also would ask for unanimous consent to submit into the
record, ``What You Need to Know About the Citizenship question
in the census,'' July 2, 2019, by The New York Times.
Mr. Roy. Without objection, that will be entered into the
record. I thank the gentlelady from Wyoming. I now recognize
the Ranking Member, Ms. Scanlon.
Ms. Scanlon. Thank you, Mr. Chair. Thank you to all our
witnesses because there is a lot to dig into here. It was a
really interesting discussion.
I guess there are a couple points of clarification. Mr.
Yang, do you think there's general agreement that the purpose
of the census clause is to get an accurate count of all
persons?
Mr. Yang. Certainly, that has been the history. It has been
nonpartisan, that is the purpose of the census, as defined in
the Constitution, and defined in the 14th Amendment for
purposes of apportionment, is to count all persons. That is the
literal text.
Ms. Scanlon. We have considerable constitutional history
and court opinion saying they did not choose to say
inhabitants. They did not choose to say taxpayers. They did not
choose to say citizens or voters. It is persons.
Mr. Yang. That is correct. This came up in the context of a
Supreme Court case called Evenwel where an argument was made
that apportionment should be based solely on legal voters. The
court, the Supreme Court in that case rejected that approach
and said that we should be using total population. Total
population or whole population for purposes of apportionment.
Ms. Scanlon. Sure.
Mr. Yang. That case was in 2016, so it was only 10 years
ago, 5-9 years ago.
Ms. Scanlon. Of course, there has been a lot of discussion
about what questions should be on the census or not be on the
census. I heard my colleague across the aisle talk about it
being a good thing to put a citizenship question on the census,
although context is important when we are talking about Canada,
Ireland, or some of the other Nation States that were
mentioned. They certainly do not have administrations that are
actively engaged in hostility toward immigrants. That is an
important distinction.
When you were testifying you talked about the two biggest
things that undermine accuracy in the census, and first, was
privacy concerns and antigovernment sentiment. The second, was
a lack of robust--get-out-the-census, get-out-the-count
efforts. I certainly participated in count efforts in my home
State.
With respect to the privacy and antigovernment concerns, I
wish Mr. Massie were here, because I know it is a big interest
of his. In an era where we have got DOGE trolling through
everyone's Social Security, and SNAP records, and voting
records, and everything, can you talk a little bit about how
those privacy concerns might be heightened right now?
Mr. Yang. Well, this goes to--thank you for that question.
This goes to the decrease in trust that Americans have in U.S.
Government, which is unfortunate, but that is the case. That is
because of--whether we want to talk about DOGE or what
Representative Balint talked about, immigration enforcement.
Certainly, that also breeds distrust within certain pockets of
our community, whether it's the Hispanic community or the Asian
community.
When we are trying to determine an accurate census, all
that context matters. Simply citing the history of how the
census was used in the past when there might have been more
trust in the government, is different than what we are looking
at today.
Ms. Scanlon. Yes, I certainly hear from my constituents
that they are very, very concerned about these efforts to troll
all the Federal data, and certainly the census would provide
another opportunity to do that.
We also have in your testimony on page 5, information about
which groups were overcounted or undercounted in the 2020
Census. I would note that non-Hispanic Whites were overcounted,
but there were some pretty serious undercounts, as you've
mentioned before, among Hispanic or Latino, Blacks, and
children, et cetera. It does seem important that we try to have
more robust counting in those areas.
Mr. Yang. That's correct. To be clear, representative,
overall the census was accurate. When compared to the 2010 and
2000 Census, they were within--they were all considered
statistically insignificant with respect to overcount or
undercount. Certainly, when you compared the 1990, 1980 Census
we did much better.
That said, we recognize that there are overcounts and
undercounts. We all agree we should continue to work on
figuring that out and how best to do that. Certainly, for my
community, the Asian community, part of this is outreach. Part
of it is also language barriers that people face. Part of it is
distrust of the government. Making sure we address all these
barriers. For rural communities it's internet access,
especially in this age where the census can be conducted by the
internet. We should be thinking about those barriers to make
sure how we get an accurate count.
The last thing I might add quickly is also I would ask the
Members of Congress to fully fund the Census Bureau. The budget
right now--the request is around two billion, by our community
1.8 billion. Right now, the funding is at 1.35 billion. If we
want to address a lot of these technical issues as well as
figuring out how to do it right to get an accurate count,
robust funding is necessary.
Ms. Scanlon. Yes, and the last question I had was with
respect to some attempt to conflate the undercounting, which
was admitted, of certain Republican-led States conflating that
with some political animus as opposed to being undercounting
relating to these already marginalized populations. Is there
anything that you have seen to indicate that it was based on
political motivation as opposed to those States just having
marginalized populations which have been undercounted across
the board?
Mr. Yang. Yes, thank you for that question. We have the
honor of working with the Census Bureau and they're dedicated
civil servants. They're trying to get this right. They
recognize that they did make mistakes and there were certain
undercounts and overcounts. The reason is complex, like we've
talked about. Part of it is distrust. Part of it is
underinvestment by certain communities in making that count
happen.
When you look at it in the aggregate, especially when you
look at where the overcounts and undercounts occurred, it's not
fair to say that it only occurred in Red States or only
occurred in Blue States. It occurred in a lot of different
States, but it also was overall pretty accurate. We should
think about all these in that context before making any snap
judgments.
Ms. Scanlon. I appreciate the Chair calling this so we can
work on continual improvement of the process. Thank you.
Mr. Roy. I thank the Ranking Member. I will now represent
the gentleman from Wisconsin.
Mr. Grothman. Thank you for being easygoing on the five-
minutes. That is what I do, too.
OK. First, Mr. Yang, I just got to interject myself in here
on something that you said that was I thought offensive. There
is a three-fifths clause in our original Constitution. That was
the result of a compromise. I suppose it could have been 100
percent, but the reason it wasn't 100 percent is the freedom-
loving people did not want to give the slave States too much
power in Congress. OK?
You imply that the reason it wasn't 100 percent was because
people were trying to say Black people were less than White
people. That wasn't the point at all, you understand? The point
is that the free States: Pennsylvania, New York, and New
Jersey, those states did not want to give too much power to the
slave States. It was not done to punish slaves. It was done to
look down the road--eventually, we had a civil war anyway, but
it was looking down the road to make sure that the slave States
did not have too much power. OK? You understand that? I hope
you understand that, but you just explain the three-fifth thing
in a way to make America look worse in the eyes of young
people. Do not appreciate that.
Now a question for Mr. Miller. Has the census ever--
Mr. Raskin. Mr. Chair, just a quick point of order. Would
it be possible for me, after the gentleman speaks, to respond
to that, because of the profundity and the depth of the
statement he just made? I think it--
Mr. Roy. It is the gentleman from Wisconsin's time.
Mr. Raskin. I don't mean to--exactly. I don't want to
detract from his time. I am wondering if maybe we could go for
a second round just to explicate that.
Mr. Roy. You and I can talk offline.
Mr. Raskin. OK.
Mr. Roy. We will let the gentleman from Wisconsin ask his
questions.
Mr. Raskin. OK. Very good. Thank you.
Mr. Grothman. OK. Mr. Miller, has the census ever included
a citizenship question?
Mr. Miller. Yes, historically it has included the
citizenship question.
Mr. Grothman. OK. Why was it removed?
Mr. Miller. That's speculative. That there were motives for
that. States would be motivated by it if they get more money
from the Federal Government. There're partisan reasons why you
would want to do it. There's a lot of reasons and it's probably
fairly complicated.
Mr. Grothman. Who was the President at the time it was
removed, do you know?
Mr. Miller. The last one was at '50s, is what we said--
1950s.
Mr. Groth. Truman? OK. Assuming you are right. Has the
Supreme Court ever held that it is illegal to include this
question?
Mr. Miller. No, I don't believe the Supreme Court has ever
weighed in to say that you cannot ask a citizenship question.
Mr. Grothman. Could President Trump and Congress Secretary
Lutnick add a citizenship question to the 2030 Census?
Mr. Miller. Yes.
Mr. Grothman. OK. Could you give me some benefits, I am
shocked that it's not there already, but do you want to
speculate on some benefits we would get if we added such a
question?
Mr. Miller. Well, it would impact especially if you got rid
of differential privacy, you could know how to allocate Federal
funding formulas more accurately to make sure that it's going
to citizen or lawful permanent resident populations that
legally qualify for them. If you want to have this impact
redistricting and you want to argue on a constitutional basis
that political districts should only represent citizens, or
citizens and lawful permanent residents, that you would need to
know that data. That would have a massive impact on how
political maps are made and who is in Congress.
Mr. Grothman. Couldn't we also assume that people who are
not here legally probably aren't going to be hanging around
that much longer?
Mr. Miller. I don't know. A lot of them have been here for
a long time. Some of them are seasonal workers and leave after
a couple of years. It's hard for me to answer that precisely
with such a large population.
Mr. Grothman. OK. Not across the board, but wouldn't one
expect that if some people were here illegally or on a
temporary visa that they weren't going to be here for the whole
10 years, or maybe even--
Mr. Miller. That's correct. There's a lot of movement of
that population both into and out of the United States of
America.
Mr. Grothman. OK. Could you explain the differential
privacy controversy, I will put it?
Mr. Miller. Yes. Basically, you have data, and there's been
previous methods of keeping basically recognizing data privacy
laws. There were different methods of protecting privacy. This
is a new method, but it's highly flawed because it scrambles
all the data. It literally deconstructs an entire person,
reconstructs them in the aggregate all over the place.
There are arguments that it will disproportionately help
urban areas. There are arguments it will disproportionately
hurt rural areas. There are racial issues here. Basically, it's
fake data below the State level and it's really hard to create
political districts that are accurate and representative of the
voting population especially when you don't have accurate
numerical population data at a granular and semigranular level.
Mr. Grothman. Would you support returning to what we will
call swapping for the 2030 Census?
Mr. Miller. I think there's several other proposals--that
is combined with the ability to turn off characteristic data
until you're at statistically significant levels. There are
other methods that strategies are multitiered that are
mentioned by entities like IMPS that would work. Essentially
yes.
Mr. Grothman. OK. Thank you.
Mr. Raskin. Would the gentleman yield? FBI--
Mr. Roy. I think the gentleman from Wisconsin would--
regarding?
Mr. Raskin. I was just wondering if you--I just wanted the
opportunity to respond to Mr. Grothman's intervention.
Mr. Roy. It is the gentleman's time, but out of courtesy to
the Ranking Member, I'll give the gentleman 30 seconds.
Mr. Raskin. OK. I just want to say that the analysis he
advanced is historically obtuse and morally disoriented. The
slave States took the position that the enslaved African
Americans should count 100 percent for the purposes of
reapportionment, but for those purposes only. They didn't have
any civil rights. They didn't have any voting rights. They
couldn't run for office. They had no civil liberties. They were
property. The Northern States said they should not count at
all.
Mr. Roy. Reclaiming our time.
Mr. Raskin. They should count zero. They arrived at the
three-fifths--
Mr. Roy. Reclaiming the time. We are not going to have a
history lesson here. The gentleman from Wisconsin was making a
point with respect to the often-used three-fifths reference.
With respect--
Mr. Raskin. I was trying to explain what the historical
genesis of it is.
Mr. Roy. The history of it, the gentleman from Wisconsin
characterized it from his perspective. It was important for him
to put that on the record. It has been asked and answered and
we don't need to have a history lesson it. I am going to go
ahead and proceed with my questions.
Mr. Roy. This hearing has been illuminating on a number of
different levels. I would like to ask just a simple question
for the panel generally, which is very specifically was the
2020 Census accurate? Mr. Mayfield?
Mr. Mayfield. At the State and national level it was within
a margin of error for modern censuses. The State level it was
not.
Mr. Roy. Mr. Rodriguez?
Mr. Rodriguez. Considering that it counted illegal and
nonimmigrant aliens, it was inaccurate according to the
dictates of the Constitution.
Mr. Roy. Mr. Miller?
Mr. Miller. It was the most inaccurate census in modern
times in my opinion.
Mr. Roy. Mr. Yang?
Mr. Yang. At the State and national levels it was accurate.
At the sublevels below that it was inconsistent at times, but
overall, still did very well.
Mr. Roy. Was it true that in 2022 that the Census Bureau
itself acknowledged that there were significant errors and
significant discrepancies? Is that correct, Mr. Mayfield?
Mr. Mayfield. That is correct.
Mr. Roy. All right. Did it have a significant impact on
States in that there was undercounting and overcounting of at
least 14 States? Mr. Mayfield?
Mr. Mayfield. That is the Census' position as of 2022.
Mr. Roy. Now, was it also accurate that, in fact, the
undercounting was often directed toward more conservative
States with more Republican representation and the overcounting
was directed toward what we might call Blue States with more
Democrat representation?
Mr. Mayfield. I can't address your use of the word
direction. That is certainly what the modeling suggests was the
result.
Mr. Roy. Mr. Miller, would you add to that?
Mr. Miller. Yes, the Census Bureau would acknowledge that
this would have been the outcome.
Mr. Roy. To the tune of as many as 500,000 in the State of
Texas, for example?
Mr. Miller. That's correct, costing Texas an additional
Congressional seat.
Mr. Roy. Let's get to the heart then of why. Why is that
the case? We have posited here in this hearing--I have heard a
number of folks represent that one of the reasons is because of
illegal immigrants. True or false, is the issue of how we count
involving illegal immigrants having a direct impact on
representation in Congress? Mr. Mayfield?
Mr. Mayfield. That is undeniably true.
Mr. Roy. Mr. Rodriguez?
Mr. Rodriguez. I agree. It's been true for at least the
last three censuses.
Mr. Roy. Mr. Miller?
Mr. Miller. It's a mathematical certainty.
Mr. Roy. Mr. Yang?
Mr. Yang. Undocumented immigrants are required to be
counted under the Constitution.
Mr. Roy. Do they have a direct impact on how we apportion
and how districts are getting set up?
Mr. Yang. That is less clear to me.
Mr. Roy. Let me ask you a question, Mr. Yang. You posited
earlier some theories about how we count people and that we
should count everybody for apportionment purposes. If 75
million people were to come into the State of California
between now and 2030, should we have 100 new Congressional
seats in California, if all 75 million of those people were
illegally present in the State of California?
Mr. Yang. That's the way the Constitution is currently
constructed as I understand it.
Mr. Roy. That is taking issue with that, by the way, that
the Constitution requires that we apportion based on people who
are here illegally, which a number of the other people on the
panel have questioned, and at no point has the Supreme Court
said that is the case, then we are saying here--the Democrat
witness is saying that if we had 75 million people illegally
flood into the State of California, we must have 100
representatives--again, rounding the math there--added to the
State of California? That is extraordinary to me.
I do want to followup and finish on this question of the
differential privacy. The other question that has been raised
here, which is highly problematic on a less-partisan basis, is
the extent to which we have the Census Bureau adopting, by Mr.
Mayfield's testimony, by edict of a civil servant bureaucrat at
the Census Bureau, an approach to estimating and guessing the
population based on something that was attacked by numerous
groups, questioned by numerous groups, and is now proven to be
statistically problematic, that it is actually causing massive
problems with respect to our counting. Is that fair, Mr.
Mayfield?
Mr. Mayfield. That is accurate.
Mr. Roy. Mr. Rodriguez, is that fair?
Mr. Rodriguez. It's fair.
Mr. Roy. Mr. Miller?
Mr. Miller. We may disagree on a lot of different things,
but there's a bipartisan consensus that this is a problem, a
big problem.
Mr. Roy. Mr. Yang, you said something that needed fine
tuning, but you also seem to acknowledge that there are
significant problems with respect to the way this is being
handled now.
Mr. Yang. As a new policy I agree that this is something
that needs to be evaluated, absolutely.
Mr. Roy. Mr. Miller, you have thrown out a couple ideas. I
want to conclude on that point because I think it is really
important that there are other ways to go about doing this.
Swapping, you get into that. The issue here for everybody to
understand is you have laws that say you got to protect
privacy. You have got a massive file that is protected at
unbelievable levels of security and inability to look at the
data. Therefore, we have to figure out how to know that we have
an accurate census. This differential privacy is raising
massive concerns on a nonpartisan basis, right? A Harvard
study: Independent groups, groups on the Left, and groups on
the Right. It would seem to me that we should come to a
consensus on how we should approach this.
Could the gentleman finish filling in that question and
that point?
Mr. Miller. Absolutely. At the end of the day there is a
constitutional purpose for all this. A privacy statute cannot
be a reason by which to ignore the constitutional outgrowths
and outflows of what the purpose of this data is.
If we have a privacy standard that is causing--I think most
people would agree with this at this point, it's significant
data distortion that is impacting all sorts of Federal funding
formulas, all sorts of potentially definitely redistricting,
potentially depending on how it's not everything that we
believe it is, and not everything that they're purporting to
say it is, in theory it's not supposed to affect apportionment,
I don't have access to the data. I can't verify that. That's
how it's supposed to work.
At the end of the day, we should defer to a privacy method
that is also as good and does not create as many constitutional
issues as we have seen. If that is achieved by swapping and
turning off data, we need to get our smart people together and
figure out another option.
Mr. Roy. You were suggesting--in closing, my time is over--
the--turning off the data. That was an approach you are
suggesting to try to have an approach.
Mr. Yang, can I just ask you on a nonpartisan basis with
just a little indulgence, do you have a perspective on that?
Because I saw you nodding--maybe all of a sudden that
intellectual chewing on it as to whether that would be an
approach that might help solve the problem of managing privacy
but not having data that is problematic.
Mr. Yang. Like I said, this is worth evaluating. This is a
part of what happens after ever census is the Census Bureau
looks at what worked, what didn't. It acknowledges what
mistakes are made. It acknowledges in this case that there were
certain overcounts and undercounts. That's no different than
any other census. Now, the job of the Census Bureau is--with
guidance from all of you, is to move forward and make that
system even better for the next census.
Mr. Roy. All right. I thank the panel. I now know that my
colleague from Texas--
Ms. Scanlon. I would just seek unanimous consent.
Mr. Roy. Sure.
Ms. Scanlon. We have discussed the census after review
quite a bit, so I would like to seek unanimous consent to
introduce the GAO report on the 2020 Census: ``Coverage,
Errors, and Challenges to Inform the 2030 Plans.'' There have
been a lot of misstatements about what it shows, so it is good
for everybody to have access to it.
Mr. Roy. Without objection, we will insert that into the
record.
Ms. Scanlon. Thank you.
Mr. Roy. I will now recognize the gentleman from Texas, Mr.
Hunt.
Mr. Hunt. Thank you, Mr. Chair. The time has come to
elucidate the American people in the census manipulation that
Democrats have been using to influence American elections for
years.
In 2018, Republican leadership, including the President
himself, called for a citizenship question to be included in
the 2020 Census Reports. To no one's surprise, when the
Democrats took control, these requests went unanswered and the
question of citizenship was waived. The fact remains that the
concept of sanctuary cities and radical Blue States ignoring
immigration law and using the illegal population to manipulate
the lines of Congressional districts is in fact
unconstitutional.
Not only are there severe constitutional concerns from the
inclusion of illegal immigrations, but the Census Bureau
admitted to having made several, quote, ``errors'' in their
2020 Census Report. Convenient, seeing that as in 2020 Donald
Trump was on top of the ticket.
In 2020, the bureau undercounted in primarily deep Red
States such as Arkansas, Florida, Mississippi, Tennessee, and
Texas, all Red, while overcounting in radical Left Blue States
like Delaware, Hawaii, Massachusetts, Minnesota, New York,
Ohio, and Rhode Island. This egregious error led to many States
being denied proper representation in Congress and the
Electoral College. So much so that these errors cost
Republicans not one, not two, not three, not four, not five,
but six seats in the House.
In addition to the 2020 miscounting, including illegal
immigrants in the census has improperly granted radical Left
Blue States 12, 12 additional seats in the U.S. House of
Representatives. That is a total of 18 seats gained. That is a
huge problem. Those are the facts.
Mr. Mayfield, thank you so much for being here. Sir, I have
a question for you if you don't mind. You, sir, are a seasoned
litigator who has practiced in matters ranging from complex
commercial litigation to constitutional and administrative law.
In your expert opinion,--I want you to really speak about this
in long form because I want this recorded--should illegal
aliens be included in the apportionment process under the 14th
Amendment?
Mr. Mayfield. The Supreme Court has never directly
addressed that, Congressman. For apportionment purposes as
Congress has interpreted it the Supreme Court has held that it
currently applies to all people regardless of citizenship.
There is certainly an open question as to whether or not--if
Congress were to direct the Census Bureau to do it differently
as to whether or not that would be legal.
Mr. Hunt. OK. In regards to thinking more of a high-level,
in terms of being a citizen and what it means to be a citizen
in this country--and I do believe that being a citizen in this
country means that you have the right to choose the leadership,
if you are a tax-paying legal citizen in this country. If you
are an illegal person in this country, should you under any
circumstances be allowed to vote and choose leadership in this
country?
Mr. Mayfield. That wouldn't be an expert opinion that--as a
citizen I would object to that.
Mr. Hunt. What about you, Mr. Rodriguez?
Mr. Rodriguez. I also object to that.
Mr. Hunt. Mr. Miller?
Mr. Miller. Correct. Citizens should not be--or illegal
aliens should not be voting.
Mr. Hunt. That applies to you also, Mr. Yang. I would love
to hear your response to this question.
Mr. Yang. Noncitizens are not allowed to participate in
Federal elections.
Mr. Hunt. OK. Are they participating in Federal elections?
Mr. Yang. The studies have shown that there are very few
instances in which they have.
Mr. Hunt. That number needs to be zero. As somebody who is
a proud Texan, we, and in Florida, have done our best to get
that number to zero. In spite of the fact that I just voted a
couple of weeks ago, and we voted with paper ballots, that we
had to show an ID for, they then verified my address and where
I lived. Then, I had to put said piece of paper through a
machine, copied it. They gave me another copy. They gave me my
very proud ``I Voted'' sticker. Even with those stringent
rules, with my daughter who is seven years old, was actually
with me. She too got an ``I Voted'' sticker. Guess what, with
those stringent laws we still have illegals voting in Federal
elections.
The question is not even about what is happening in Texas.
If it is happening in Texas, you have to imagine that it--while
it may seem like a minuscule amount to you, sir, that number
needs to be zero. If it is happening in Texas with those
stringent rules, imagine what is happening in other Blue States
that don't have anything that closely resembles that level of
redundancy to ensure that the person that voted in that
election is in fact a citizen that has an address and has an
ID.
With that, I yield back the remainder of my time. Thank you
very much.
Mr. Roy. I thank the gentleman from Texas.
That concludes today's hearing. We thank the witnesses for
appearing before the Subcommittee today. Without objection, all
Members will have five legislative days to submit additional
written questions for the witnesses or additional materials for
the record.
Without objection, the hearing is adjourned.
[Whereupon, at 12:14 p.m., the Subcommittee was adjourned.]
All materials submitted for the record by Members of the
Subcommittee on the Constitution and Limited Government can
be found at: https://docs.house.gov/Committee/Calendar/ByEvent
.aspx?EventID=118665.
[all]