The Insights Association on Tuesday said in a letter to Commerce Secretary Howard Lutnick that the market research and data analytics industries would be hobbled if the proposed rule is enacted. The Commerce Department oversees the U.S. Census Bureau, which conducts the once-a-decade head count of the United States.

“By hobbling the work of our $89.3 billion U.S. industry, the proposed elimination of race/ethnicity data from the decennial census would also create significant costs and huge risk across the many varied industries that rely on our insights for their decision-making,” the association said in the letter.

The private sector is no substitute for the depth of data that is collected during the decennial census, and the race and ethnicity data are a key component, the letter said. For business research, they serve as benchmarks for predicting retail store performance for certain groups, testing new products and determining how much advertising is driving sales.

“Race/ethnicity data underpins the industry’s work in many ways,” it said. “Eliminating race and ethnicity data collection would not simply create a gap, it would remove the anchor that the entire U.S. insights industry uses to gauge the accuracy of its own numbers and thus unmoor the decision-making insights necessary for the rest of the country.”

The letter asked the Commerce Department to either drop or revise the proposed rule, engage with data users about the changes and provide adequate time for public input on this decision.

Questions on race have been asked since the first census in 1790. The data can be used to make claims of racial voter dilution under the Voting Rights Act, and their elimination would make it impossible to pursue such claims.

The Commerce Department hasn’t yet made public the draft rule, but a copy was obtained by Talking Points Memo.

The Census Bureau may survive these assaults in the long run, but it’s getting smeared by the octopus’ ink spray.
The tentacles already have choked off, for now, the release of the most comprehensive survey of American life after the Trump administration’s Commerce Department surprised the statistical world, and the Census Bureau, this summer with an order restricting the types of privacy methods statistical agencies can use. The new order could limit the amount of data available for redistricting at the smallest geographies, and that could mean less data on the racial makeup of small communities.

The previously used method was too hard for people who aren’t statistical experts to grasp, according to George Cook, a Commerce Department official who is performing the duties of acting Census Bureau director in the second Trump administration without a Senate confirmation.

Per reporting from Wired and Talking Points Memo, a draft rule by the Trump administration also would prohibit the collection of race and ethnicity data in the 2030 census, even though those types of questions have been asked since the first census in 1790. This data can be used to make claims of racial voter dilution under the Voting Rights Act, and their elimination would make it impossible to pursue such claims.

Within the last week, the Census Bureau’s previous director condemned a decision by the bureau to release a report purporting to show that 24,000 noncitizens voted in the 2020 election. President Donald Trump falsely touted it last week as evidence that he had won the 2020 presidential election against Joe Biden. In a telling sign of its credibility, no Census Bureau researcher attached their name to the report. Rob Santos, who served as the Census Bureau’s chief for most of the Biden administration, said the report “besmirches” the bureau’s reputation. John Abowd, a former chief scientist at the bureau, called the report “nonsense” in a post and said its results were “at best irresponsible but more likely pure partisan manipulation of the Census Bureau’s reputation for reliable statistical products.”

NPR reported that the report’s analysis was conducted by a team affiliated with the America First Policy, which was co-founded by Stephen Miller, the current White House deputy chief of staff. If the think tank’s name sounds familiar, it’s because its legal arm has been leading courtroom fights in Florida, Louisiana and Missouri to keep unauthorized immigrants or noncitizens from being counted in the 2030 census and to throw out the results of the 2020 census. Those efforts also are tied to the Trump administration’s decision to use a questionnaire with a citizenship question for the 2026 census test currently being conducted in Alabama and South Carolina instead of a traditional census questionnaire that omits it.

The reason for all of this? A now-deceased Republican redistricting expert had argued that Republicans and non-Hispanic whites get an advantage if only U.S. citizens who are age 18 and older were counted when redrawing congressional and state legislative districts, instead of the total population. The Census Bureau’s own research has shown that adding a citizenship question to a census questionnaire produced lower self-response rates in households with noncitizens and immigrants from Latin America.

The overwhelming majority of Census Bureau demographers, statisticians and researchers who are still doing tremendous work inside the agency are upset by the political interference. But the octopus’ actions could alter how outsiders view the statistical agency as a source of the highest quality, nonpartisan data.
An octopus has mini brains in all eight of its tentacles, making it seem like they are operating independently, but they are connected to a central brain. That’s a good metaphor for how the Trump administration’s tentacles are reaching into the Census Bureau.
This octopus knows what it’s doing.

Rob Santos, who served as the Census Bureau’s chief for most of the Biden administration, called on the agency to take down from its website the report which President Donald Trump used as evidence this week to falsely claim that he had won the 2020 presidential election against Joe Biden.

“Its appearance there besmirches the Bureau’s reputation as an excellent nonpartisan scientific statistical agency serving the public good,” Santos said in a social media post on Thursday. “The document reflects neither the Bureau’s core values (transparency, scientific integrity, objectivity, independence) nor its mission to produce quality data on our people and economy.”

Most significantly, neither the bureau’s scientists nor experts were used in putting together the report, Santos said.

Indeed, no bureau researcher attached their name to the report which was posted to the agency’s website on Tuesday. Every other report on the bureau’s “Census Publications” site is credited to a statistician, economist, demographer or researcher.

Trump lost to President Joe Biden by 7 million votes in an election in which 160 million voters cast ballots.

The Census Bureau report, “Noncitizen Voting in the 2020 Election, A Beginning Analysis,” says that its “beginning” analysis determined with certainty that over 128 million citizens cast ballots and more than 24,000 noncitizens also voted. However, the report said the Census Bureau still has 32 million voter records to analyze.

Instances of noncitizens voting are rare, and the Census Bureau numbers, so far, back that up – 0.018% of votes cast in the 2020 election. Also, 24,000 votes weren’t enough for Trump to overcome Biden’s lead.

The noncitizens voting report seems to be annoying other former and current Census Bureau economists, statisticians and demographers.

“I’m proud of the work that I do at the Census Bureau (so proud, in fact, I put my name on it!),” labor economist Cody Orr said Wednesday in a social media post touting research on post-secondary employment outcomes.

Santos left the bureau in 2025, a short time after Trump returned to the White House for a second term. George Cook, a deputy undersecretary of the Commerce Department, which oversees the Census Bureau, is performing the duties of acting bureau director without a Senate confirmation.

The noncitizen voting report lacks methodological rigor and transparency, and its results should be discarded, Amy O’Hara, a research professor in the Massive Data Institute at Georgetown University, wrote in an analysis.

“Moreover, the Census Bureau should answer for the process concerns raised by the publication of this unsigned and shoddy report,” O’Hara said.

“When we’re providing a service, all taxpayers should be able to use it. There does need to be a certain accessibility and transparency,” George Cook, told outside demographers and statisticians Wednesday during an online discussion hosted by the American Statistical Association. “The public needs to have trust.”

Federal law prohibits people’s identities from being disclosed in the data the federal government releases. During the 2020s, the agencies have used statistical “noise,” which adds small errors to data sets so that people can’t be identified by their race, ethnicity or sex, particularly in small places like census blocks. The new Trump administration order prohibits that. Critics of the method had said it made the data at the smallest geographies less than accurate.

“It was difficult to explain to people, ‘Well, it’s pretty close to accurate,’” said Cook, who is performing the duties of acting Census Bureau director in the second Trump administration without a Senate confirmation.

His agency is bearing the brunt of the decision.

It has delayed the release the most comprehensive survey of American life, the American Community Survey, as the bureau looks for a way to comply with the order, and it will seemingly scramble the future releases of other data sets. The order also halted research on privacy protections for the 2030 census since that research was premised on the now-banned method, said Michael Lachanski, a deputy director of the Census Bureau.

“Some things are going to be delayed,” said Lachanski, who also serves as a senior advisor in the Commerce Department, which oversees the Census Bureau. “We’re trying to adhere as closely as we can to product release schedules.”

Instead of using the method which injected “noise,” the new order instructs the agencies to use an old-school method known as “coarsening,” which can reduce the level of details by rounding figures up or down, or grouping data sets together. If “coarsening” isn’t possible, the order says, the agencies shouldn’t release the numbers at all, and that could impact rural and small communities the most.

The Commerce Department’s order could limit the amount of data available for redistricting at the smallest geographies. That could mean less data on the racial makeup of small communities and no data on the voting-age population. The lack of data could make it more difficult to pursue claims of racial voter dilution under the Voting Rights Act.

Cook said the decision on privacy protections was made at the level of the Commerce Secretary’s office, and the Census Bureau just provided technical advice.

“The Census Bureau basically interacted on a kind of technical and advisory basis for the department,” Cook said.

Then again, few Census Bureau publications are ever used as evidence by the president of the United States to claim an election was stolen. But that’s exactly what President Donald Trump did yesterday, proclaiming on “Truth Social” that he had won the 2020 presidential election. Just to be clear, Trump lost to President Joe Biden by 7 million votes in an election in which 160 million voters cast ballots.

The Census Bureau report, “Noncitizen Voting in the 2020 Election, A Beginning Analysis,” says that its “beginning” analysis determined with certainty that over 128 million citizens cast ballots and more than 24,000 noncitizens also voted. However, the report said the Census Bureau still has 32 million voter records to analyze.

“The Census Bureau is going to analyze the next 32,000,000 Voters, and this number will explode,” Trump wrote. “I WON THIS ELECTION!”

Trump was sending a message to Republicans in Congress to pass the stalled SAVE America Act, which would require voters to provide proof of citizenship to register to vote and a photo ID at the time of voting.

Instances of noncitizens voting are rare, and the Census Bureau numbers, so far, back that up – 0.018% of votes cast in the 2020 election. Also, 24,000 votes weren’t enough for Trump to overcome Biden’s lead.

There should be some concern with how the Census Bureau used administrative records to conduct this analysis. The voting data came from a compilation of state registration rolls and vote-history extracts, according to the report. The bureau also used its Person Identification Validation System that links anonymously a single person to a wide range of data sets from the federal government and surveys. It may sound Big Brotherish, but it can be used, among other things, to help fill in gaps in data collection in an age when people aren’t responding to surveys or answering their doors when census takers come knocking.

However, incomplete and mismatched records can inject bias into the statistics using the linked data. A 2014 Census Bureau study suggested that young children, minorities, immigrants and recent movers were less likely to have information that could be tagged into the system.

For the election study, the Census Bureau noted, “the team conducting this beginning analysis did not have access to the names of the population analyzed or any other identifying information.”

As it turns out, no names were attached to the Census Bureau report either.

The Trump administration’s order already has delayed the planned release in September of data from the most comprehensive survey of American life, the American Community Survey, and it will seemingly scramble the future releases of other data sets.

That’s my quick take from today’s blog post about the new order from George Cook, who is performing the duties of acting Census Bureau director in the second Trump administration without having had a Senate confirmation.

It goes without saying there’s a method to the madness. Per reporting from Wired and Talking Points Memo, a draft rule by the Trump administration would prohibit the collection of race and ethnicity data in the 2030 census, even though those types of questions have been asked since the first census in 1790. This data can be used to make claims of racial voter dilution under the Voting Rights Act, and their elimination would make it impossible to pursue such claims. Along those same lines, the Commerce Department order could limit the amount of data available for redistricting at the smallest geographies. That could mean less data on the racial makeup of small communities and no data on the voting-age population.

According to Cook:

_ The Census Bureau is currently researching new privacy methods for the American Community Survey to comply with the order. That’s like giving a lesson about landing to a pilot while the plane is still in the air. You would have thought they would have come up with a way to land the plane before the change was made in midair.

_ Flying with that metaphor, how long will the plane be forced to stay in the air? Cook gives no indication when the latest 1-year American Community Survey data will be released. He says the transition to privacy methods that comply with the order will “occur gradually,” and that a full transition will be done by 2029. Does that mean the plane stays in the air for another three years?

_ The order is causing the Census Bureau to shift its research agenda for the 2030 census, according to Cook. 2030 may seem like a long time away, but the Census Bureau begins planning for each once-a-decade census 10 years before the head count begins. It takes years of evaluations and planning, and big changes midway through are extremely disruptive.

_ Cook says the Census Bureau will release a demonstration data set in mid-2027 showing how the changes will impact the 2030 census.

Data advocates are genuinely worried that the new order will limit the type and amount of data that the Census Bureau releases each year.

Federal law prohibits people’s identities from being disclosed in the data the federal government releases. In recent years, the agencies have used statistical “noise,” which adds small errors to data sets so that people can’t be identified by their race, ethnicity or sex, particularly in small places like census blocks. The new order prohibits that. Critics of the method had said it made the data at the smallest geographies less than accurate.

Instead, the new order instructs the agencies to use an old-school method known as “coarsening,” which can reduce the level of details by rounding figures up or down, or grouping data sets together. If “coarsening” isn’t possible, the order says, the agencies shouldn’t release the numbers at all, and that could impact rural and small communities the most.

Cook said that the goal of the new order is to provide the public with accurate data.

“The new order marks a positive shift in how the Census Bureau balances confidentiality, accuracy and data availability,” he wrote. “Implementing the policy will involve tradeoffs and resource implications but will allow us to meet our most important goal: ensuring accurate and objective statistics.”

_ The U.S. Census Bureau said it was delaying the release of its most comprehensive annual survey of American life as it figures out how to deal with the administration’s order. The 1-year estimates from the American Community Survey were supposed to be released in September.

_ Planning for the 2030 census has been disrupted because of the order, according to attendees of a presentation earlier this month by U.S. Census Bureau leaders at a statistical conference.

_ The order could limit the amount of data available for redistricting at the smallest geographies, according to attendees at the Joint Statistical Meetings. That could mean less data on the racial makeup of small communities and no data on the voting-age population.

“We interpret that to mean that data users should expect less detail in 2030 Census redistricting data products,” data advocates Beth Jarosz, Jae June Lee, and Chris Dick wrote in a post about the presentation.

_ The order may limit how the Census Bureau measures its accuracy in the 2030 census.

Federal law prohibits people’s identities from being disclosed in the data the federal government releases. In recent years, the agencies have used statistical “noise,” which adds small errors to data sets so that people can’t be identified by their race, ethnicity or sex, particularly in small places like census blocks. The new order prohibits that.

Instead, the order instructs the agencies to use an old-school method known as “coarsening,” which can reduce the level of details by rounding figures up or down, or grouping data sets together. If “coarsening” isn’t possible, the order says, the agencies shouldn’t release the numbers at all, and that could impact rural and small communities the most.

In a related move, a draft rule by the Trump administration, viewed by WIRED and Talking Points Memo, would prohibit the collection of race and ethnicity data in the 2030 census, even though those type of questions have been asked since the first census in 1790. This data can be used to make claims of racial voter dilution under the Voting Rights Act, and their elimination would make it impossible to pursue such claims. The draft rule also would prohibit any questions about sexual orientation, although that has never been asked previously on a census questionnaire.

The draft rule also would prohibit unauthorized immigrants from being counted in the numbers gathered in the 2030 census and used to determine how many congressional seats and Electoral College votes each state gets. This is no surprise. During his first term, President Donald Trump unsuccessfully tried several times to keep people in the U.S. illegally from being counted in the apportionment process following the 2020 census.

The idea has its roots in the writings of a now-deceased Republican redistricting expert who argued that Republicans and non-Hispanic whites would get an advantage if only U.S. citizens who are age 18 and older were counted when redrawing congressional and state legislative districts, instead of the total population.

Finally, the draft rule would change the time period for collecting data for the 2030 census to Jan. 3 until April 1. During the 2020 census, residents started getting invitations to respond to the census questionnaire in March.

The U.S. Census Bureau said Tuesday that it is delaying the release of results from the largest annual survey of American life as it figures out how to deal with the administration’s order. Instead of being released next month, the bureau said it was looking for solutions that would allow it to comply with the Trump administration’s order and release the 1-year estimates from the American Community Survey later this year.

The survey of 3.5 million households asks about more than 40 topics, including income, housing costs, veterans status, computer use, commuting, and education.

The recent order from Trump’s Department of Commerce, which oversees the Census Bureau, limited the types of privacy protections that can be used by the Census Bureau and the Bureau of Economic Analysis, a move that could restrict the amount of data available for research and redistricting.

Federal law prohibits people’s identities from being disclosed in the data the federal government releases. To prevent that, the agencies have used statistical “noise,” which adds small errors to data sets so that people can’t be identified by their race, ethnicity or sex, particularly in small places like census tracts. The method had been controversial among some who worry it causes the data to veer too much from reality.

The Department of Commerce order instead instructed the agencies to use an old-school method known as “coarsening,” which can reduce the level of details by rounding figures up or down, or grouping data sets together. If “coarsening” isn’t possible, then the agencies shouldn’t release the numbers at all, the Department of Commerce said in the order.

The order was issued without public comment, and researchers say it could diminish the level of detailed data drastically.

“It undermines statistical integrity. It reduces the tools the Bureau has to produce necessary public data while protecting individual privacy,” children’s advocate Deborah Stein wrote recently. “As a result, it puts the public in a lose-lose situation: it will result in less data, weaker privacy protections, or (most likely) both.”

In the meantime, statisticians and data advocates are trying to get the order reversed.

That thought crossed my mind this week as I read Realtor.com’s recent report estimating that a third of young adults lived in their parents’ home last year. Because an overwhelming majority of these young adults between ages 25 and 34 had jobs, the culprit was the skyrocketing cost of housing, not unemployment, according to Realtor.com.

This coincides with a drop in babies being born in the United States over the past two decades. Since the Great Recession, when the fertility rate was 2.08, around the rate needed for children to numerically replace their parents, it has been on a steady decline, hitting 1.6 in 2025. Besides unaffordable housing, poor access to childcare

and the growing expenses of child-rearing also add up to fewer kids.

The fertility rate drop has caused a lot of handwringing among conservative policymakers like Vice President JD Vance. Leading by example, Vance and his wife, Usha, welcomed their fourth child on Sunday. Keep it up, Veep!

Vance has suggested giving parents more voting power, according to their numbers of children, or following the example of Hungary’s former leader, Viktor Orbán, in giving low-interest loans to married parents and tax exemptions to women who have four children or more.

The conservative Institute of Family Studies this month issued a report with the ominous headline, “Without Action, American Population Will Soon Decline.” Among the actions proposed to solve the problem are five-or-six-figure baby bonuses paid by the federal government, credits toward Social Security benefits for stay-at-home parents and incentives for the construction of more housing.

Paul Quirk, who lives at his grandma’s house in Atlanta, recently told me that there’s a reason so many members of his generation haven’t flown the coop like previous generations.

“They were able to buy a lot of things, a lot cheaper,” Quirk said.

Among “a lot of things,” nothing is more consequential than housing. Home construction in the United States hasn’t kept pace with population growth, and the U.S. had a housing deficit of 4.7 million homes at mid-decade, which was the primary driver of the affordability crisis, according to a report released this month by Zillow. The median home listing price of $430,000 is a third higher than what it was in 2019.

“That’s forced millions to share housing with people outside their family -- doubling up when they would likely prefer to have their own place,” the Zillow report said.

The report, though, highlighted a silver lining around the doomsday clouds: the housing deficit has stopped growing in a meaningful way after years of deepening significantly.

But University of Maryland demographer Philip Cohen noted in a recent post that the “fertility crisis” may not be the crisis it has been made out to be, particularly since the greater threat to population growth in the United States is the Trump administration’s immigration crackdown.

“With a mass deportation target of a million people per year — and negative net migration already achieved, for the first time in many decades — we may end up with a population crisis in the U.S., just not one caused by falling birth rates,” Cohen wrote.

Want an alternate solution to the housing crisis, and perhaps the “fertility crisis” too?

Move to my hometown, St. Louis. A Neighbors Bank report just ranked it as the metro where it takes the least amount of time to save for a down payment on an entry-level home on a median family income. It was less than four months in St. Louis compared to Miami’s more than 40 years.

Yeah, I know. I wrote about this same topic last week involving a case in Missouri.

But the latest apparent warning reinforces, if there were any lingering doubts, that Trump and his immigration hardliners are plowing ahead with years-long efforts to exclude people in the United States illegally from the once-a-decade head count. The first Trump administration failed at these efforts, including trying to add a citizenship question to the 2020 census questionnaire, but it wasn’t for lack of trying.

The latest tip-off comes from a court filing made late last week, this time in a Louisiana case. In it, DOJ lawyers reiterate that the Trump administration is going through the process of changing a rule dealing with “residence criteria” for the 2030 census. The criteria set out who is counted in the census, and the Census Bureau has said historically that immigrants are counted regardless of their legal status.

The numbers derived from the census determine how many congressional seats, and Electoral College votes, each state gets, as well as the distribution of $2.8 trillion in federal government spending.

In both the Louisiana and Missouri cases, Republican state attorneys general sued the U.S. government, trying to stop unauthorized immigrants from being tallied in the population count following a census. The Trump administration hasn’t yet publicly said how the criteria will change, but it’s latest litigation messages to the attorney generals in both the Louisiana and Missouri cases are, more or less, “Wait and see. I think you’re going to like this!”

“Defendants have taken concrete steps towards finalizing their policy position on issues that bear directly on the ones presented here,” Trump administration lawyers said Friday in a court filing in the Missouri case.

The U.S. Constitution’s Fourteenth Amendment says “the whole number of persons in each state” should be counted for the numbers used for apportionment, and the Census Bureau for decades has interpreted that to include people living in the U.S. illegally.

To find out if someone is an unauthorized immigrant, though, the 2030 census questionnaire would have to ask about a respondents’ immigration status, which is unchartered territory. It likely would create confusion and reduce participation, noted Thomas Saenz, president of the civil rights group, MALDEF, during an online forum this week.

“This goes beyond the citizenship question,” Saenz said. “It would have some of the same effects as a citizenship question, which is to deter participation, create all kinds of confusion, and require the Bureau to engage in an unprecedented educational effort to try to get good data in response to such a question. All of that would be multiplied if there were a status question.”

Meanwhile, we wait.

The Trump administration hasn’t given any indication when it will publish the rule change.

The state of Missouri in January sued the U.S. government, trying to stop unauthorized immigrants from being tallied in the once-a-decade U.S. head count. The “Show-Me” state wants to prevent them from being included in the population numbers used to divide up congressional seats and Electoral College votes among the 50 states based on their population size. Missouri claims their inclusion unfairly gives states with large numbers of unauthorized immigrants more political representation at its expense.

The Trump administration said late last month that it has submitted a proposed rule that may soon eliminate Missouri’s concerns.

The Trump administration on June 26 asked a federal judge to put Missouri’s lawsuit on hold since the yet-to-be-published rule “may ultimately obviate or significantly narrow many of the issues at the heart of the litigation.”

The rule change was submitted late last month for review by the Office of Management and Budget.

“This rulemaking may narrow or eliminate issues presented here,” Trump administration lawyers wrote in a proposed order for the judge.

The residence rule

The rule in question is the “residence criteria” that the Census Bureau sets out for who is counted during the census. It has said that immigrants regardless of their legal status are counted at the place where they live and sleep most of them time.

The U.S. Constitution’s Fourteenth Amendment says “the whole number of persons in each state” should be counted for the numbers used for apportionment, and the Census Bureau for decades has interpreted that to include people living in the U.S. illegally.

The Missouri lawsuit claims a person can’t be considered an inhabitant of a state if they don’t have “a lawful intent” to be there. It also asks the court to order a redo of the 2021 apportionment process so that people in the U.S. illegally are excluded.

Civil rights groups have asked to join the Missouri lawsuit as intervenors, claiming the Trump administration isn’t going to protect the interests of immigrant groups and others.

The population figures derived from the census not only determine political power among the states, but they also guide the distribution of $2.8 trillion in annual federal spending on health care, education, school lunch programs and many other things.

Other lawsuits

Two other lawsuits challenge the rules and methods of the 2020 and 2030 censuses.

In Florida, Republican U.S. Rep. Byron Donalds, who is running for Florida governor, as well as Young Republicans and College Republicans, sued the federal government, claiming that the statistical methods used by the Census Bureau during the 2020 census had robbed Florida of political representation and federal funding. A majority of a three-judge panel on Tuesday dismissed the lawsuit for a second time. The majority said the plaintiffs lacked standing and that there was no adequate remedy to address their claims given that they were made almost half-a-decade after the apportionment process.

In Louisiana, four GOP state attorneys general sued the federal government in an effort to keep people in the U.S. illegally from being counted in the 2030 census. That lawsuit currently is on hold.

No surprise

This shouldn’t come as a surprise. During his first term, President Donald Trump

unsuccessfully tried to keep people in the U.S. illegally from being counted in the apportionment process following the 2020 census. He also ordered citizenship data to be collected through administrative records. The first Trump administration tried to add a citizenship questionnaire to the 2020 census questionnaire but was blocked by the U.S. Supreme Court on procedural grounds.

In his second term, Trump in August 2025 instructed the Commerce Department, which oversees the U.S. Census Bureau, to exclude people in the U.S. illegally from the census count. On the first day of his second term, Trump rolled back a Biden executive order reaffirming that all people in a state are counted in census results.

The desire to eliminate illegal immigrants from the head count is tied to a now-deceased Republican redistricting expert who argued that Republicans and non-Hispanic whites would get an advantage if only U.S. citizens who are age 18 and older (voting age) were counted when redrawing congressional and state legislative districts, instead of the total population.

But that could shrink the head count in both red and blue states because the states with the most people in the U.S. illegally include California, Texas, Florida and New York.

It’s about the U.S. statistical system, which may put some of you in snooze mode. The truth is, the system helps us understand who we are as a nation as we celebrate our 250^th year of independence, and it provides information that keeps the economy running. It is vitally important, but researchers are concerned that it’s being disrupted by the Trump administration.

What is the federal statistical system? It is comprised of more than a dozen federal agencies that gather data on the economy, employment, housing, trade, our health, our population and much, much more. For decades, it has been considered the gold-standard of statistical systems – robust, unbiased and reliable.

People at these agencies are still doing good work. But in the past year and a half, figures on LGBTQ people and other data sets have disappeared. Some agencies have lost more than a third of their staff, resulting in delays and concerns about data quality. Political interference is creeping in. Federal data websites have been down.

“They’re really reflecting an instability across the entire data ecosystem,” Darby Steiger, a vice president of the SSRS survey research firm, said Tuesday during a discussion on this topic. “We used to be able to rely on federal data websites, as stable go-to resources … Those sites are frequently down, they’re not being updated or they’re simply not functioning. And that’s making it difficult to answer even the most straightforward questions that people have.”

Limits on privacy protections

Are you still with me? Your eyelids aren’t drooping, I hope? Because there’s more.

Last month, the Department of Commerce limited the types of privacy protections that can be used by the Census Bureau and the Bureau of Economic Analysis, a move that could restrict the amount of data available for research and redistricting.

Federal law prohibits people’s identities from being disclosed in the data the federal government releases. To prevent that, the agencies have used statistical “noise,” which adds small errors to data sets so that people can’t be identified by their race, ethnicity or sex, particularly in small places like census tracts. The method has been controversial among some who worry it causes the data to veer too much from reality.

The Department of Commerce instead instructed the agencies last month to use an old-school method known as “coarsening,” which can reduce the level of details by rounding figures up or down, or grouping data sets together. If “coarsening” isn’t possible, then the agencies shouldn’t release the numbers at all, the Department of Commerce said in the order.

Researchers say it could diminish the level of detailed data drastically.

“What we do know is that the result of this ill-conceived order will be either less privacy for our personal information, or less usable data, or both,” a coalition of statisticians and demographers said in a statement last month.

Concerns about the future

SSRS conducted a survey of more than 520 public data users earlier this year, and a majority said the changes to the federal statistical system have had a major impact on their ability to do their job. More than three-quarters of those surveyed said they were very concerned about whether federal statistical data will continue to be collected.

Foot-dragging on race and ethnicity categories

On top of all this, the administration is dragging its feet implementing the first changes in more than a quarter century to how people are officially grouped based on race and ethnicity. Advocates believe the changes better reflect

who we are as a nation.

They add a new category, “Middle Eastern and North African,” for people descended from places such as Lebanon, Iran, Egypt and Syria who previously had been encouraged to identify as white.

The changes also combine the queries about race and ethnicity into a single question on forms, giving respondents the option to pick multiple categories at the same time, such as “Black,” “American Indian” and “Hispanic. Research has shown that large numbers of Hispanic people aren’t sure how to answer the race question when that question is asked separately because they understand race and ethnicity to be similar, and they often pick “some other race” or do not answer the question.

But the administration recently pushed back the deadline for agencies to come up with a plan for complying with the changes, and the changes won’t be used for the 2027 iteration of the Census Bureau’s largest survey of American life.

I know that’s a lot to chew on, but thanks for hanging in there. You’re still awake, right?

Subscribe now

METRO DALLAS HAS become the leading magnet of diversity in the United States this decade. Only metro Houston comes close when looking at the combined growth of its Asian, Black, Hispanic and white populations, according to U.S. Census Bureau estimates released Thursday.

If the 1980s TV soap opera “Dallas” was filmed in 2026, a more accurate setting than Southfork Ranch would be the suburban Frisco mansion of a South Asian family or an African American family living in the new subdivisions of suburban Grand Prairie.

The influx of African Americans, Asian Americans and Hispanic Americans comes from the growth of the Dallas-Fort Worth metroplex’s overall population. The metro has attracted people from almost every background. No other metro area has gained more residents this decade.

The Dallas-Fort Worth metro area grew by almost 840,000 new residents between 2020 and 2025, with more than two-thirds of the growth coming from people moving there from other places. The rest came from natural increase, or births outpacing deaths. About 55% of the migration was from immigrants, and the rest came from people moving from other U.S. states.

The Rev. Frederick Haynes, a Dallas pastor who is running for Congress, told me in February that he has seen an explosion of new parishioners who moved seeking Dallas’ big-city amenities, mild weather and relative affordability compared to the coasts. He says it reminds him of growing up in San Francisco decades ago when no one actually was from San Francisco.

The increasing diversity of the Dallas-Fort Worth metroplex this decade hasn’t been devoid of racial tensions or prejudices.

The affluent Southlake suburb faced a reckoning on racial harassment, and then a conservative backlash to the reckoning, that originally was triggered by a social media video showing white high school students chanting a racist slur. And Texas politicians have tried to block the East Plano Islamic Center from constructing a community development project, fanning the flames of Islamophobia in the process.

The count:

_ Metro Dallas was No. 1 in growth in pure numbers for the Black and Asian populations this decade among U.S. metros. It was No. 2 for growth in the Hispanic and white populations, both behind Houston.

_ The New York and Seattle metro areas had the next highest Asian population growth this decade. For African Americans, it was Houston and Atlanta, and for the Hispanic population, it was New York and Miami. Phoenix and Miami were the next top spots for white population growth this decade.

_ The metro with the largest number of Hispanic residents, Los Angeles, also lost the most Hispanic residents this decade. Los Angeles, San Jose and Santa Cruz were the only U.S. metros that experienced declines in their Hispanic populations in the 2020s so far.

_ The Pacific Northwest is attracting Hawaiians and Pacific Islanders. The Seattle and Portland metro areas saw the largest growth in this population this decade, followed by the Dallas, New York and Las Vegas metro areas. Fayetteville, Arkansas was the No. 5 metro, but that is not as surprising as it sounds since Marshall Islands residents have been moving to northwest Arkansas for decades to work in the poultry industry.

_ Los Angeles, New York and San Francisco have lost the most white residents this decade of any U.S. metro areas. But the Los Angeles and San Francisco metro areas also have lost the most residents overall so far this decade. New York, the most populous metro area in the U.S., has grown by only 29,000 people this decade.

_ The biggest gain in the American Indian and Alaska Native population was the New York metro area.