Nagged for weeks by a census count that was lagging behind most of the country, Mayor Richard Daley and regional census officials in June made a rousing pledge: Chicagoans would dramatically step up their cooperation with the census, and federal headcounters would help them do it.
As night follows day, things improved–fast.
But seven former and current U.S. Census Bureau employees now say in interviews that the sudden surge in recorded cooperation in at least part of Chicago may have been more on paper than real.
In short, the employees allege that bureaucrats in one key Chicago census office, in a rush to finish the main counting phase on time, lowered rigorous standards established to ensure the statistical validity of the count and allowed workers to accept questionnaires with less-reliable information.
“The message was: Get the numbers up, get the numbers up, get the numbers up–any way you can,” said Bob Rucker, 61, a former supervisor who oversaw the processing of questionnaires in the bureau’s office in charge of counting Lincoln Park and Lakeview residents until he was fired June 2.
A retired postal supervisor, Rucker and the bureau agree he was fired because he failed to make production goals.
“But I didn’t meet those goals because I wasn’t willing to do what they wanted me to do–I didn’t agree with it and I told them so,” Rucker said.
All those interviewed stress they were instructed not to lie or falsify any information. The instructions they received, as they understood them, were aimed at finishing the count as soon as possible.
What they were told, they say, was to turn to neighbors for information without trying as hard to reach members of households themselves. That method, census officials agree, typically results in less-reliable, less-complete information.
The lower standards are alleged to have been adopted just after Memorial Day, days after census officials in Washington announced management shifts and other measures aimed at revitalizing the slumping count in Chicago.
Regional census officials vigorously deny allegations about the Near North office, asserting that credit for the rapid gains in questionnaires gathered goes to hard work and a last-minute media blitz by the city.
Conducting a reliable and thorough census is critical in determining the redrawing of congressional and other political district boundaries as well as the dispersal of government aid.
The results of the census won’t be released until the end of the year, but the willingness of the public to cooperate with the census is an important early indicator of how sound and statistically valid those results will be.
In Chicago, ensuring cooperation with the count took on added urgency because a low turnout would reflect poorly on the mayor and his brother, outgoing Secretary of Commerce William Daley, who has been overseeing the national count.
Chicago is not the only place where questions about accuracy and speed have emerged in recent week, as census officials in Washington applaud themselves for finishing the bulk of the count well ahead of a publicly stated schedule.
In at least one case, concerns spurred the bureau to embark on an internal investigation into whether workers in a Hialeah, Fla., office cut corners in racking up 82,000 completed questionnaires in 22 days, according to U.S. Rep. Dan Miller (R-Fla.), chairman of the House subcommittee on the census.
Miller also said he has received “numerous” complaints from census employees across the country offering similar allegations that they were pressured to find shortcuts.
The Census Bureau mailed questionnaires to 98 million American households earlier this year and began following up in April by sending census takers, officially dubbed enumerators, to those households that had failed to return completed forms.
Census director Kenneth Prewitt has explained that door-knocking work finished earlier than expected nationwide because more Americans than expected returned their questionnaires by mail.
Yet Miller’s aides say they want particular scrutiny paid to places such as Chicago where census progress accelerated suddenly.
The census relies on a rigid set of procedures that spell out in precise detail the steps workers at every level of the process are to follow in gathering and processing questionnaires. For all but the final 5 percent of the count, headcounters must make six attempts to reach any household–three in person and three by phone–before turning to neighbors for information. Supervisors must reject forms filled out otherwise, and the clerks responsible for reviewing forms before they are scanned offer another check.
By May 24, roughly a month after headcounters had gone out nationwide, the Near North office had become a puzzling and frustrating case for the Census Bureau. Though much of its district was white and affluent–the demographics that typically cooperate most with the census–it was running second to last among the city’s nine offices in total forms gathered.
The bureau complained publicly that doormen were turning away headcounters. And gradually, officials say, some doors opened.
But several of the employees interviewed, most of whom are still on the census payroll and asked not to be named, said they were instructed after Memorial Day and during the week of June 5 to accept information they had earlier been told to use as a last resort.
Two employees said that beginning the week of June 5, they were told by higher-ups to get forms in faster and were permitted to try households only three times before turning to neighbors or building managers.
Census officials say turning to neighbors after so few attempts is usually allowed only when a counting district has finished 95 percent of its door-knocking workload.
Progress reports released by the bureau indicate that by June 5 the Near North office had completed 62 percent of that workload.
Moreover, two former supervisors, who resigned recently because they say they were criticized for not turning in forms fast enough, independently described seeing another supervisor marking forms “population unknown” and submitting them.
“Something like that might happen during closeout [after reaching 95 percent], but this was two weeks before that,” said one of the former supervisors, who asked not to be identified.
The Near North office closed out the main door-knocking phase on June 15, officials said.
Two clerks responsible for reviewing questionnaires said in separate interviews that they were told during the week of June 5 to accept forms that showed only three attempts.
One clerk said she first felt pressure to “do what you need to do to get these forms through” during the week of June 5.
“They would bring in these questionnaires with very little information or none,” said the clerk, who is still employed by the bureau and asked not to be named. “It might say there was one person living at this place–no name, no age, nothing, but they supposedly talked to a neighbor … and they told us to pass these through however you can pass these through.”
Regional and local census officials say they have heard no allegations of any Near North employees relying improperly on information from neighbors or other sources or filling out forms inappropriately.
The census is now conducting a process designed to audit the work of headcounters, supervisors and others.
Stanley Moore, who oversees the three-state region that includes Chicago, said he has confidence in the Near North office, though he added he could not rule out the chance that some employees cut corners.
“I know there are people out there who will take shortcuts, but the process is designed to catch that,” Moore said, alluding to smaller door-to-door campaigns going on now to fill in gaps. “The census isn’t over, and we are still checking.”
