Research support for this article was provided by The Investigative Fund at The Nation Institute.

Houston

Even though Julia DeLeon has lived in Houston for more than half her life, built a business here, raised two daughters and is now helping to rear a toddler grandson, she doesn’t exist—not statistically at least. DeLeon first moved to Houston in the 1980s. When she got sick while pregnant and an emergency room attendant refused to admit her because she’s not a documented resident, she retreated to Guatemala. But DeLeon couldn’t stay away. Her sister had lived here and was certain the town held opportunity. So she returned with her newborn, settled in at a friend’s apartment in Southwest and started rebuilding her life.

If either the 1990 or 2000 Census crossed her mind amid all of this, she doesn’t remember it. She wouldn’t likely have participated, anyway. After all, she didn’t even have a valid lease, let alone a valid visa. Standing up to be counted by federal officials wasn’t high on her to-do list.

"I just make a little part in the corner where they grew up," DeLeon recalls of the apartment where she raised Evelyn DeLeon, now 21, and Sharon Garcia, 16. "I just try to find a way in the school, any program so they not just stuck in the house. Like, making the Girl Scouts, art classes, in the church—some kind of organization where they can go, and it’s not costing me much and they learn something."

Now, having painstakingly built a successful house-cleaning business, DeLeon sits in her own living room and listens proudly as her daughters talk about the importance of finally "being counted" in the 2010 census. "I do it," she shrugs. "Ten years back, I was maybe afraid if they deport us—what happen with these two girls? But now, if something happen with me, it’s not really a worry," she says, turning to Evelyn, who’s been volunteering in an ad hoc campaign to encourage Census participation in the neighborhood. "She can support now."

That seemingly blasé statement represents a huge victory for the Census Bureau. The DeLeon family—working-class, urban people of color—is just the sort of household the bureau has struggled to count for decades. Census statisticians discovered in 1940 that they were undercounting blacks by as much as double-digit percentages. Decade after decade, as the overall count grew more accurate, a racial gap in undercounting held firm. When the bureau began measuring across all races and ethnicities in 1990, it found blacks, Latinos and Native Americans were all undercounted at more than double the national rate.

As a result, the nation’s largest metro areas have also been consistently shorted in the decennial Census. The bureau keeps a list of characteristics that make it tough to enumerate a given tract: lots of renters, low household incomes, new immigrant communities, single-parent families and large populations of people of color, among others. The list informs a ranking of "hard-to-count areas" that’s essentially a big-city roll call—Los Angeles County, Brooklyn, Chicago’s Cook County and Houston’s Harris County fill the top four spots this year.

As of late April, a familiar pattern was emerging with the 2010 Census. Just over 70 percent of the 134 million mailed questionnaires had been returned. Small, largely white counties in the Plains and Midwest were registering response rates of 80 percent and up. Barely half of DeLeon’s neighborhood had responded, as was the case throughout Southwest Houston’s densely Latino neighborhoods; Brooklyn was at just about half, Chicago at 60 percent. In May, Census workers will start knocking on doors of homes that haven’t replied.

The consequences of undercounting stretch far past mapping Congressional districts. Decennial Census numbers are used to plan things ranging from city council seats to airports and to divvy up hundreds of billions of dollars in federal support for state and local initiatives every year. In 2000 Census monitors analyzed how the undercount that year would affect funding for just eight federal programs. New York City was predicted to lose a collective $847 million. Harris County looked to miss out on roughly $240 million. This loss is amplified by the fact that undercounted communities are often those most in need of government programs—Medicaid, Head Start and special education, for example—whose funds are determined by Census data.

Planners fear that this year the collapsed economy will complicate things further still. Unemployment rates among blacks and Latinos are as much as a third higher than the national average. Communities with high foreclosure and job-loss rates are more likely to have families in unstable, temporary housing. It’s a statistical negative feedback loop in the making.

In the past decade Washington has spent $14 billion, more than twice the cost of the 2000 Census, to avoid such bleak math. Over the past year, the Census Bureau has joined local governments and community groups to mobilize hundreds of thousands of neighborhood opinion-shapers to persuade families like the DeLeons to participate. It is by far the nation’s most ambitious attempt ever to enumerate itself.

But the challenges the Census faces are both greater and more complex than the mechanics of a head count. Families like the DeLeons—young and brown-skinned migrants—are driving rapid demographic changes in the United States. Many of these new residents are uncertain about whether government is a source of support or a threat—the long arm behind immigration raids, detentions and record-high deportations. The answer becomes less clear as the right stokes an increasingly polarized debate over immigration. The tea party’s smears of the government as an intrusive, untrustworthy force are often vocalized simultaneously with the charge that government sold out "real Americans" in favor of "illegal" menaces. In October, Louisiana Senator David Vitter tried adding a question about immigration status on the stripped-down 2010 Census form. He hoped to spark a fight about whether undocumented residents should be enumerated at all. The Congressional Research Service countered that the Constitution clearly dictates that the Census count "persons" living in the United States, not citizens. But the question Vitter sought to force is one the modern Census—with its mandate of rendering a national portrait in hard, tangible numbers—cannot avoid: Who does and does not count?

Meanwhile, tea partyers aren’t the only ones voicing distrust; it runs deep in regions like Southwest. Recent Pew Research Center polls found new immigrants articulating greater trust in government than people born here. If so, it’s a complex faith in Southwest, where lived experiences like Julia DeLeon’s thwarted emergency room visit often overshadow the message of city and Census officials—that all residents matter. As a result, the nation may find itself unable to fulfill one of the Constitution’s most basic assignments—to count the population every ten years—at a time when policy-makers need the data it generates more than ever.

As goes Texas, so goes the nation. By far the country’s fastest-growing state, Texas is gaining on California for the title of most populous. As the housing crash reversed trends in such rapid-growth states as Georgia and North Carolina, the Texas population boom appeared to continue apace in 2008 and ’09. The decennial Census, which takes a closer look than the bureau’s yearly estimates, is expected to show this trend intensifying.

But it would be far more accurate to say that East Texas is the fastest-growing state in the nation. Rice University professor Steve Murdock was George W. Bush’s last Census director and has been Texas demographer for decades. He’s got a state map divided into counties, with red marking those that are shrinking and blue, those that are growing fastest. The largely rural, overwhelmingly white counties in the west form a sea of blood-red. The eastern half is dominated by three massive blue blobs, each a leg of the Houston-Dallas-San Antonio triangle and each a place where whites are in the minority. Actually, Texas stopped being a majority-white state in 2004. It’s one of four states, with California, Hawaii and New Mexico, the Census Bureau calls "majority-minority."

Texas’s new residents are not only overwhelmingly brown; they are overwhelmingly young. Nearly half of 18- to 29-year-olds in Harris County are Latino, according to Rice University’s Institute for Urban Research, while 70 percent of people over 60 are white. The bureau estimates that the county’s Latino population grew by 40 percent between 2000 and 2008, while the white population grew less than 1 percent.

Bureau estimates show a nearly identical growth pattern for US residents overall: a young, growing population of people of color, particularly Latinos, alongside an old, plateauing population of whites. In 2000 nearly half of the residents enumerated as "Hispanic—any race" were under 25, as were 42 percent of those counted as "Non-Hispanic Black." Just 30 percent of residents who identified as "Non-Hispanic White" were under 25. The Latino population likely grew by a third between 2000 and 2008, while the white population grew by 2.5 percent—a gap that will only grow as today’s youth have kids, and as their parents and grandparents die.

"The Texas of today is the United States of tomorrow," says Murdock, pointing to a national map in which states across the South and West—places with large nonwhite populations—are booming with new residents, while states in the country’s middle hold flat. "Where there are diverse populations in the United States, there is population growth. Where there is less diversity, there is less growth," Murdock concludes.

This blunt fact inspires one of two reactions in America’s political culture: hope or fear. To people like Angela Blanchard, who runs Houston’s Neighborhood Centers, a social service agency working largely in immigrant communities, Houston is a beacon on a hill. "There’s a huge entrepreneurial spirit that’s fed by a massive amount of immigration. Last year we crossed the line as the most diverse city in the United States," she says, shaking her head at the fact. "No one thinks of Houston that way. But it’s just amazing to live in a place where everybody’s here!"

Southwest Houston certainly feels more like Queens than middle America. A four-mile stretch along the main east-west artery is packed with retail shops targeting a broad Asian community. The area has been booming steadily since the mid-1980s, recessions notwithstanding; Houston boasts 16,000 Asian-owned firms, the sixth-largest concentration in the nation as of 2002. "When I was growing up, if we saw an Asian we’d stop the car and point," jokes Glenda Joe, a Chinese-American Houston native with whom city planners have contracted to promote the Census. "Now look left and right, and all you’ll see is Asians—Filipino, Korean, Chinese."

Thousands of them don’t exist in the official population count, however. The 2000 Census tallied around 5,500 Koreans in Houston; Dongwook Yang, managing editor of the Korean Journal, says the city’s two largest Korean churches alone have at least 3,000 members between them, and his paper’s community directory includes nearly fifty more churches. "If I buy the pastors’ estimation," says Yang, "it can be more than 40,000." But that sort of napkin calculation is meaningless when businesses are struggling to convince investors and would-be migrants of Houston’s burgeoning Asian community. "They think about Dallas instead," Yang complains.

The Korean undercount is infamous among city planners, too. "Houston Independent School District’s statistics indicate there are at least 30,000 Koreans in HISD," scoffs Margaret Wallace, assistant director of Houston’s planning department. Wallace has a map of neighborhoods where the city has struggled most to get an accurate demographic picture and where the department is targeting its 2010 effort. "We are treating this essentially as an election campaign, and we’re gonna do that kind of ground campaign."

DeLeon’s neighborhood sits inside a stretch of districts in Southwest Houston that feature prominently on Wallace’s map. The area’s hub is Gulfton, a small, tightly packed neighborhood dominated by a handful of massive apartment complexes. The area was developed in the 1960s as a convenient suburb for young, upwardly mobile white families. A deep recession in the 1980s pushed those families out, leaving affordable housing stock for new Central American immigrants. By the 2000 Census, the neighborhood had grown by 40 percent while white residents in the broader area had fallen by half. The city estimates that three-quarters of Gulfton’s residents are Latino.

All these characteristics—the dense housing complexes, the new migrants, the mixed languages—give Gulfton and the surrounding area a sky-high score on the hard-to-count matrix. Wallace’s office believes it’s one of the city’s most undercounted neighborhoods.

Enrique Sapon, a church-group leader in Gulfton’s Guatemalan community, is sitting in Benito Juarez’s office in Gulfton’s new city-funded community center. Juarez, a community liaison in the mayor’s immigrant affairs office, spends his days finding neighborhood surrogates like Sapon to pull in residents for all manner of services, from prenatal health to English classes. But all conversations lead back to one place.

"We are planning for immigration reform, and to inform and raise the consciousness of the people," Sapon says when asked about the Census. For many in the neighborhood, the word "government" only triggers fears about immigration policy. "Because people don’t want any more raids and families being separated because they are deported." The only question to be answered about the Census is how participating will impact the path to citizenship.

Residents’ answers vary widely but are most often rooted in suspicion. Sapon is not terribly moved by the promises that Census information is confidential. "It’s the same thing they were saying about the Social Security and the hospitals," he says, before citing neighborhood lore about people being deported out of hospital beds. "There have been cases." Another rumor in Southwest is that the effort to get everyone counted is a ploy to set up a mass deportation sweep. Others fear that showing how many immigrants are truly here will fan public animosity toward them. "Anyway, if they don’t fill out the form, somebody is going to come knock on the door," Sapon says of his neighbors, "so they are just praying that nothing happens."

"People really have a lot of fear," Juarez explains. "They go to work, and they do normal everything, but when it comes to getting in contact with government agencies? That’s a no-no." He says when the center opened, it took six months to persuade people to walk through the door. Today, an estimated 6,000 people pass through every week.

By the time the center opened in 2007, it had been on the city’s wish list for more than a decade. Juarez says that’s in part because Census data have never adequately reflected the population that locals know to be wedged into the tiny space. Houston’s city-run community centers are funded in part by federal Community Development Block Grants, and those are allocated based on a formula that includes Census data. So, for that matter, are many of the social services offered at the center.

The Brookings Institution estimates that Census data guided nearly $3 billion in federal grants made to Harris County in the 2008 fiscal year and $420 billion in federal grants nationally. Medicaid and the State Children’s Health Insurance Program, highway construction, special education, Head Start, home-heating assistance, senior housing—funding for all these things is based on Census data.

Wallace and others stress that nobody really knows how much the city’s undercount affects specific programs, because the formulas for allocating money are complex. But from service providers to elected officials, people trying to tackle community needs day to day say they consistently bump against bad data.

"A hospital wants to relocate, they look at numbers," says Harris County commissioner Sylvia Garcia, who also heads the National Association of Latino Elected Officials. "A bank wants to open up a branch, they look at numbers. A foundation wants to make their donations, they look at numbers. And one of the very basic tools of any of that kind of research is Census."

Garcia is still frustrated, for example, by a recent flap with the city schools. When officials closed a school in a historic building in her district, they at first agreed to open an early-childhood education center in its place. But they consulted Census data that said there weren’t enough kids under 6, so they shelved the project. "I just said, ‘OK, the crying room of the church is always full. How can you tell us there’s not enough kids?’ And we didn’t believe it, but they showed us." Garcia bought the building herself and is raising funds to open a center like the one she envisioned.

Money aside, the politics matter, too. Based on the 2010 Census, Houston must draw two new city council districts to account for the city’s population growth. "We need an accurate representation of where we live in order to do a fair redistricting plan that furthers representation in our neighborhoods," Wallace says. The implications aren’t geographic alone. Just two of the council’s fourteen members are Latino in a city where Latinos are estimated to be nearly 40 percent of the population.

Cynthia (not her real name) is being polite, but she’s not terribly interested in talking, certainly not about the government’s desire to know who she is and where she lives. She listens patiently as Evelyn DeLeon explains in Spanish why the ten-year Census is so important. But DeLeon might as well be telling her to drink poison.

"She feels like it’s some sort of trap," DeLeon says, translating Cynthia’s response. "She knows that she’s seen it on TV," DeLeon explains, "and that this will benefit our children so that they can be counted and get help." But that sounds like BS to Cynthia. "Whatever they say it is, she thinks it’s the opposite."

Cynthia is, however, thrilled to talk about her daughter in Honduras. The proud mom holds out a cellphone picture of the 18-year-old decked out in a graduation robe. She’s off to college, and Cynthia, 47, is in Houston breaking her back for $230 a week because it’s contributing to her daughter’s future. She figures whatever kind of help this Census thing is offering, it’s not likely to advance that singular goal. So it’s not worth the risk of getting deported.

Her suspicions about whether and how she counts to the government would be confirmed later that night. Her closest friends, who, like her, are undocumented, were called to the airport to pick up their young son, a US citizen by birth. They had sent him home to Honduras with friends who are also citizens, but border cops wouldn’t let the group back into Houston without the boy’s legal guardians. When the parents arrived to claim him, they were arrested for visa violations. Separated from their son, they will be detained indefinitely and likely deported.

As the Census Bureau has flooded neighborhoods like Gulfton with its "be counted" message, the Obama administration has ratcheted up deportations to a record high. Immigration and Customs Enforcement deported nearly 400,000 people in Obama’s first year. And the president has hardened his tone on immigration; he moved from using the term "undocumented workers" on the campaign trail to "illegal immigrants" during the healthcare debate, when anti-immigrant rhetoric on the national stage soared.

"What’s really different that wasn’t there even ten years ago is this evolving anti-immigrant sentiment," says Commissioner Garcia. "We really had not felt it here as strongly as we see it on TV and around this country," she adds, "but in this past year, we’re starting to see more raids."

The nightmarish idea of undocumented parents being separated from their children looms over any family’s daily risk calculations. Cynthia, for instance, avoids food stamps, despite her challenges buying groceries for her youngest daughter, a grade schooler in Houston and a US citizen. "She stopped going because they were asking for way too much information, and she felt uncomfortable," DeLeon explains. That fear also means Cynthia’s daughter won’t be counted toward her neighborhood’s elementary school or toward education grants based on Census numbers. She’s among nine out of ten Houston public school students who are either Latino or black.

Garcia says this is the sort of thing that makes the anti-immigrant furor so counterproductive. It’s clear that new immigrants infuse a city with the entrepreneurial ethos Neighborhood Centers’ Blanchard gushes about. But it’s also true they are often starting from scratch in an exploitative labor market. Nearly a quarter of both Latinos and blacks lived in poverty in 2008. A Southern Education Foundation study noted earlier this year that the South’s public schools are now majority-minority; three years ago the region became the first to have a majority of its students be so poor they qualify for free lunch. Those students are the future workforce.

"We’re already here, and it’s quite doubtful we’re going to be going anywhere," Garcia lectures. "So what we need to do is work together on solving these challenges to ensure that there’s not some sort of meltdown in ten years. We gotta get ready."

Rice’s Institute for Urban Studies, which has been tracking opinions and demographics in Houston for three decades, has found an increasing number of people who disagree. Just 64 percent of people surveyed last year agreed that kids of undocumented parents should be able to go to public schools, down from 71 percent three years ago. Similarly, half of those surveyed last year wanted to deny "health and welfare services" to undocumented residents. On a question measuring opinions about a range of populations—from whites to Muslims to gays and lesbians—"undocumented immigrants" scored the worst.

Immigrants, of course, aren’t the only ones who distrust the Census. During the 2000 count, candidate Bush and Senate majority leader Trent Lott suggested that the process is too intrusive; Bush said on the campaign trail he was "not sure" he’d fill out his form. The fallout is one reason Congress and the bureau stripped the 2010 questionnaire down to ten questions. This time, the right’s anti-Census chorus is being led by tea party cheerleaders like Congresswoman Michele Bachmann, who has said she won’t fully complete the form and invoked the specter of Japanese internment camps.

Meanwhile, the demographic realities that Rice’s Murdock lays out have savvier Republicans, worried about their shrinking base of nonurban white voters, offering a much different take. Karl Rove starred in a Census Bureau ad in April, calling the Census an "instrument of democracy."

In a number of small, largely white counties around the country, however, the Census has become an instrument of something quite different. Census methodology turns on the bureau’s "usual residence" rule, meaning that you are counted in the place you live most of the time, regardless of whether you consider that home. So college students are counted on campus, and prisoners are counted where they are incarcerated.

Problem is, state prisons nationwide are located overwhelmingly in small, aging white towns and are filled with young black and brown people. That generates all manner of statistical absurdities. Prisoners of the Census, an advocacy group urging a change in methodology, found twenty-one counties nationwide in which more than a fifth of the population enumerated in 2000 was behind bars. The group counted 256 counties in which more than half the enumerated black population was incarcerated. In Brown County, Illinois, 18 percent of the county’s 6,950 residents are supposedly black, but just five of them were free.

The enumeration of these prisoner-residents conceals broader trends. Ten overwhelmingly white Texas counties that appeared to be growing after the 2000 Census were actually shrinking. Meanwhile, a fifth of the state’s prisoners came from Harris County but were tallied elsewhere—in rural counties in which they participated in none of the programs or services the Census informs.

This distortion has real consequences. Seven New York State Legislature districts would not exist but for their nonvoting prison population. Peter Wagner, who runs Prisoners of the Census, says a handful of states have begun subtracting prisoners from the Census count before redistricting and at least three, including New York, have pending legislation to make similar adjustments.

But it’s not just about politics. Waupun, Wisconsin—which is among the counties with more than a fifth of their populations locked up—recently commissioned a study to reconfigure its profile, absent the three state institutions it hosts. The town, it seems, has become a prisoner of its jail-based economy. As Mayor Jodi Steger told the local paper, "Developers look at a site and they don’t know the income includes the prisons, and their maximum wage is something like $2 an hour."

The Census taps a conflict deep in the American psyche. Our multiple identities have long been at odds. We’re a nativist country, built on immigration. The Statue of Liberty welcomes newcomers in New York Harbor, as a fence turns them away at the Texas border. Little wonder a process that quantifies this debate’s futility is fraught.

The Constitution nonetheless demands it be done, and myriad government functions have come to depend on it. After decades of bitter legal and political wrangling over whether and how the bureau could use a sampling formula to fill in the undercount, bureau scientists voluntarily shelved the idea after the 2000 count. Disparate experts agreed that the formula wasn’t working, though it could be figured out if politics allowed science to run its course.

The unintended result is that Congress has been backed into budgeting previously unthinkable sums for an old-fashioned tally. In 2000 that investment appeared to pay off; the Census got its first-ever paid-advertising budget, inaugurated the community partnerships that have been a definitive part of the 2010 count and beefed up staff for in-person follow-ups. All of this drove up overall response rates and helped close the racial gap.

The bureau hopes to roll that success over into this year’s count. But few in Gulfton expect it to succeed in the current climate. "People have been saying there’s no need to count everybody because in the end, even if everybody’s counted, those resources are not going to trickle down to the undocumented people," says Juarez. "I’m optimistic they’re going to get a better count, because of the resources," he concedes, before adding, "but it’s not going to be enough."

Ironically, it may be that the politics of immigration have made places like Gulfton too American for government to do its job, even among Census believers like Evelyn DeLeon. "It’s important and I hope there is change, but…" She stops to gather her thoughts, before settling on a starkly American idea about it all: "Whether the Census is supposed to provide this help or not, I’m going to have to do this on my own."