Most people attribute the origin of the census to the Constitution, but credit should really go to the Bible. As Moses was instructed in the Book of Numbers: “Take ye the sum of all the congregation of the children of Israel, after their families, by the house of their fathers, with the number of their names, every male by their polls.”
But, really, counting the children of Israel pales in comparison to the task of tallying each and every man, woman, and child living on American soil each decade, as the Constitution instructs, also in a single (if slightly less eloquent) sentence: “An actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.”
This year, despite pouring an estimated $14 billion into the count—more than twice as much as the last time around, and almost $14 billion more than the framers of the Constitution probably had in mind—the Census Bureau is facing one of its most daunting challenges yet. Given rancorous antigovernment sentiment and threats of boycotts from angry groups on both the right and the left, the most expensive and labor-intensive census in U.S. history could also turn out to be one of the most troubled ever. With issues of racially charged language on the census form, cost overruns, and the recently resolved debate over sampling hanging over the count, there’s a feeling that the 2010 census may be in trouble before it even starts.
It’s easy to forget how much is at stake in the decennial count—especially after an idle glance at the short, ten-question form that will land in every resident’s mailbox on April 1. But the data the census provides play a pivotal role in nearly every democratic process—a role that underscores the justification offered for the massive bureaucracy by the census’ staunch defenders, aside from the constitutional mandate that makes it necessary.
“It’s the very heart of our system of fundamental democracy,” says Kenneth Prewitt, who served as director of the U.S. Census Bureau from 1998 through 2001. Prewitt is now a political-science professor at Columbia University’s School of International and Public Affairs, and he speaks like a man who has had to defend the whole idea of the census more than once.
Prewitt is right, in one sense. Census data are used to determine the distribution of more than $400 billion in federal funds for schools, hospitals, bridges, tunnels, and other services. By some estimates, local governments stand to lose as much as $30,000 for every resident the census fails to count. While it is obviously impossible to gauge precisely how many people the census skips, New York City alone has forgone as much as $850 million over the past ten years thanks to undercounting, according to an estimate offered by New York’s governor, David Paterson.
The census also controls the allocation of seats in the House of Representatives and the redrawing of district lines in every state. Districts are carved city block by city block, and seats can swing on just a handful of votes. In 2000, for instance, Utah was forced to cede the 435th House seat to North Carolina because of just 857 extra heads. This makes an accurate count crucial to the political process, even though that isn’t mentioned in the constitutional mandate—nor is the exorbitant cost covered, or even mentioned, in the Constitution. Yet, last time around, the government spent approximately $6.5 billion on the decennial count: It hired nearly one million workers—the nation’s largest-ever peacetime mobilization—and flooded the airwaves with a $100-million ad campaign urging participation.
By most measures, the government’s efforts in 2000 were a rousing success. Sixty-seven percent of households mailed back their completed forms, reversing a three-decades-long trend of diminishing returns even though shifting demographic trends should have made the population more elusive. The 2000 census also made great strides in counting more of the people it systematically skips, including African Americans, Hispanics, and the poor. But this year, the bar has been lowered. Despite a record-smashing $300-million-plus outreach and marketing blitz, the Census Bureau predicts that only 64 percent of households will mail back completed forms—a 3 percent drop from 2000. This means that census workers will have an estimated four million more doors on which to knock over the coming months, a task that will add billions to the already enormous cost. Every percentage point of nonresponse drives the price of the census up an estimated $85 million, the bureau reports.
The bureau maintains that its benchmarks are actually higher than they were a decade ago, when the bureau grossly underestimated the number of people who would mail back their forms and set a return-rate goal of just 61 percent. The current projection, explains a census spokesperson, is based on the percentage of households that returned their forms before the bureau began sending workers door-to-door to collect missing data last time around.
But this time, some critics suggest that reaching even that number will be difficult.
“I’m pretty pessimistic about trying to get a mail-back rate as high as we did in 2000,” says Prewitt. Instead, he says, most experts are expecting a mail-back rate of just 61 to 62 percent. That’s not quite what the framers ordered.
That so much taxpayer money should be spent on a less successful count baffles many—especially those unconvinced of the count’s purpose (antigovernment and privacy advocates) or those who don’t appreciate unwelcome knocks at the door (almost everyone).
“Why do they want to know so much about me?” asks Margarita Rivera, 51, bundled in a purple parka as she rummages through bins of plastic clothes hangers outside a Spanish Harlem discount store, just as the census media blitz begins to ramp up to full speed. Rivera says she already is concerned about the level of government intrusion into her life and her home. “Don’t bother me,” she warns census workers, her gold hoop earrings shining as she shakes her head in opposition. A recent poll by the Pew Research Center for the People & the Press suggests that nearly one in five residents still is not sure if he or she will participate in this year’s census, or says that he or she is unlikely to or definitely will not.
“That would be an absolute disaster,” warns Roderick J. Harrison, a senior fellow at the Joint Center for Political and Economic Studies and former chief of the U.S. Census Bureau’s Racial Statistics Branch. For most of the survey’s respondents, the decision not to participate is practical, not political: More than half of those polled by the Pew Center said they are too busy, not interested, or don’t know what the census is. For a small but vocal minority, however, the motive lies deeper—in concerns over personal privacy, safety, and government intrusion. And that small minority has census experts deeply concerned.
Harrison says that this year’s political climate is the most toxic he has ever witnessed. The biggest threat to this year’s count, he says, is the growing wave of antigovernment sentiment that has swept the nation and that has begun to fuel a violent backlash against the census count. “It’s overwhelming,” he says. “Some of the poison from the polarization of the country is really seeping in.” After months of crushing recession compounded by bank bailouts, foreclosures, and no clear end in sight, people are deeply frustrated. Harrison fears that boycotting the census may be an easy way for people to take out their anger. “There is a tide saying we need to send a message to Washington, to get heard,” he says. “Here’s a very convenient way to be heard.”
Harrison anticipates significantly lower mail-back rates as well as abuse directed toward enumerators—the folks who travel door-to-door collecting information from residents who did not mail back their forms.
Minnesota Representative Michele Bachmann, a Republican, has been the leading voice in the growing anti-census movement. In July, she told the Washington Times that her family will protest the count, leaving most of its form blank except for the question that asks directly how many people reside in her home. “We won’t be answering any information beyond that,” she says. She complains that the remaining nine questions are “very intricate, very personal” and argues that “the Constitution doesn’t require any information beyond that.”
More moderate members of the Republican party, who have repeated their support for the count, have readily criticized Bachmann’s views. Harrison says that having a legitimate conservative leader publicly question the value and constitutionality of the census is “unprecedented.”
“I can’t remember when there has ever been a time when you had people in mainstream political parties arguing not to take the census,” Harrison says. “That used to be one thing everyone could agree on. . . . The polarization has gotten to a point where there are elements that are willing to exploit the census to reinforce antigovernment sentiment.” He fears this change represents “a horrifying erosion of a common sense of civic duty.”
If someone chooses to boycott the census, there is little the bureau can do. Those who refuse to complete the form technically risk a $5000 fine, but the bureau is reluctant to publicize or enforce the rule for fear of further alienating a public that already is on the fence.
In general, those on the left have traditionally favored as inclusive a count as possible, both for ideological reasons and because the victims of the undercount are generally assumed to be part of the Democratic base. As of today, as the Pew data show, attitudes toward the count differ markedly by party, with 71 percent of Democrats saying the census is “very important,” and just 56 percent of Republicans.
But this year, threats of boycott are being raised from all sides, suggesting an added possibility of trouble.
The National Coalition of Latino Clergy and Christian Leaders, which represents 22,000 pastors in 34 states, is calling on millions of undocumented immigrants and sympathizers to boycott the 2010 census in protest of government inaction on passing comprehensive immigration reform. “We can assure you that over 3 million participants and members of our church will not participate,” says Reverend Miguel Rivera, the coalition’s head.
Undocumented immigrants have long feared that census data will be used against them or will spell disaster if the data land in the wrong hands. That is part of the reason why, as the undocumented population swells, so too will the tally of the undercounted. The census does not ask participants about their immigration status, and the bureau insists that all information is confidential and strictly protected by law. Rivera maintains, however, that participation puts the undocumented at risk.
“They are liars. They have been lying for years,” Rivera says of census officials’ assurances. “People in our churches are being hunted like animals,” he says. “Undocumented immigrants should not take more chances.”
Rivera’s fears aren’t entirely without merit. The Census Bureau has, in the past, turned over privileged information—including, most recently, the location of prominent Arab populations to the Department of Homeland Security. The department maintains that it requested the information only to help it determine which languages to use on signs posted in major international airports. But critics remained unconvinced.
“If they [undocumented residents] are needed so much to be counted, they need to be legalized first,” Rivera maintains. The coalition is now calling on the White House to sign an executive order halting all deportations and releasing those imprisoned for immigration violations from jail. “It’s in the hands of the president and Ms. Pelosi and Mr. Reid,” Rivera says.
All this would have worked better had the Census Bureau been allowed to use sampling methods to deliver an accurate count—or would it? Statisticians have long argued that using sampling instead of manually counting each and every individual in the nation would save billions of dollars, hours, and headaches. Sampling, in which a representative subset of the population would be questioned, would also provide the bureau with a more accurate count, says former census director Prewitt, because statistical techniques could be used to correct for errors such as the repeated undercounting of certain groups—making sampling especially popular among Democrats in the House. This is precisely why Republicans have argued the opposite.
After a bitter partisan dispute that threatened to derail the 2000 count, the issue finally came to a head before the Supreme Court in 1999 in the case of Department of Commerce v. U.S. House of Representatives. In a 5-to-4 decision, the court sided with House Republicans, ruling that sampling violates the Constitution’s mandate of “an actual enumeration” and so cannot be used in the reapportionment of Congressional seats. This literal interpretation, which many Democrats still oppose, has ruled out sampling, at least for now.
In the end, even those who want to be counted can pose a challenge. As with the growing number of undocumented residents, other demographic trends already make the U.S. population harder to count with each passing year. We are more mobile; we move more often and live in multiple or broken homes; we rent more and own less; more of us are foreign born. Our phones no longer are fixed to our addresses; we are easier to reach via our email inboxes than via the mailboxes on our front lawns. Response rates for surveys also have dropped across the board. “People are over-surveyed and inundated with market research,” says Harrison. “They don’t want to be bothered.”
Before the growing hostilities, the Census Bureau’s most pressing concern was reversing the systematic undercounting of groups such as racial and ethnic minorities and the poor. For African Americans, for instance, the undercount has hovered between 4.5 and 5.6 percent since the 1970 count. As a 1967 report by the Conference on Social Statistics and the City, which was spearheaded by the late Senator Daniel Patrick Moynihan, once argued, “Where a group defined by racial or ethnic terms and concentrated in special political jurisdictions is significantly undercounted in relation to other groups, individual members of that group are thereby deprived of the constitutional right to equal representation. . . . In other words, miscounting the population could unconstitutionally deny minorities political representation.”
Leaving aside the boycott threats, the Census Bureau can—and has—improved these omissions by administering a better count. This year, however, the bureau also has made a serious gaffe that threatens to depress participation. After more than 50,000 individuals spontaneously wrote in the word Negro in response to the question that asked about race in 2000, the bureau decided to include the term on this year’s form, listing “black, African-Am. or Negro” as one of several potential answers to Question 9. News of the wording spread quickly and sparked outrage among some black leaders who consider the term racist and outdated.
It isn’t just outside factors such as boycotts, anger, and political beliefs that dog the census, however. Community organizations that partnered with the Census Bureau in 2000 are also complaining that the bureau is moving too slowly to reach key populations and has been less eager to reach out for assistance this time around.
In addition, if the first phase of the census is any indication, ballooning costs could become a serious threat. Last fall, the Census Bureau hired nearly 150,000 temporary workers to canvass neighborhoods to update the address files that determine where census forms will be sent. The project went over budget by $88 million—a whopping 25 percent. While bureau director Robert Groves has promised to rein in costs and recover the funds through other savings, some fear this could foreshadow events to come.
Other concerns about how cash is being spent are also starting to emerge. According to an audit by Commerce Department inspector general Todd Zinser, obtained by the Associated Press, the Census Bureau already has wasted $3 million paying temporary workers for work they never did and reimbursing them for millions of miles they never drove.
“There’s a lot of frustration out there among people who did a great deal to make the 2000 census successful,” Harrison says. “There’s a lot to worry about.” Inadvertently, Harrison has summed up the problem this time around: a nagging, anticipatory fear among those who follow the census that this year’s count may miss its mark. But by how much? Sadly, in the greatest irony of all, the census is always least capable of counting the precise number of Americans it couldn’t find.
Jill Colvin is a New York–based journalist whose writing has appeared in the New York Times and the Washington Post.