My Two Census

Formerly the non-partisan watchdog of the 2010 US Census, and currently an opinion blog that covers all things political, media, foreign policy, globalization, and culture…but sometimes returning to its census/demographics roots.

Posts Tagged ‘questionnaires’

Must-Read: New York’s 2010 Census nightmare

Monday, June 7th, 2010

The following report comes from a Census Bureau official whose identity has been confirmed but will remain anonymous as she is a current Census Bureau employee:

The five boroughs of New York City and its diversified population of eight million have long eluded demographers and census employees in producing an accurate count. Having worked in three censuses now and living in New York for almost my entire adult life I notice that the socioeconomic spectrum of New Yorkers has widened, making the poor poorer and the rich richer. In the last ten years there is an influx of immigrants; some legal some illegal. It makes what was once a one family home in Queens, Brooklyn and The Bronx a two or even three family home. These people are living in converted basements or the second story of the houses some legal some illegal. On the other end of the spectrum, luxury rentals and condominiums have become even more exclusive with price tags in the millions of dollars. In both cases the immigrants and residents of these upscale housing units and their exclusive real estate management companies have ignored repeated attempts by phone or mail to allow enumeration.  Even in the face of a fine, the management companies are adamant about their policy and would willingly pay the fine rather than to allow enumerators to count their residents. The problem is the Census’ Bureau’s threat of a fine is merely used as a scare tactic. When a real estate mogul calls their bluff the actual fine like many other Census Bureau promises is empty.

As native New Yorkers we anticipated these problems. And sitting through four days of verbatim training where someone read through a book, we knew that it wasn’t as simple as the script made it to be to persuade these respondents about the importance of the census and their participation. As a group we brainstormed and created techniques through trial and error to get those who were non-responsive to fill out our questionnaires. Some of these tactics included: sending another enumerator of a different race or creed after several visits with no contact; leaving blank enumerator questionnaires under their door allowing them the privacy of completing it in their own home. One of us even went as far as sending well dressed suits or female fashion models to coerce participation. But all this takes time and money. All of which with 15 billion price tag the Census Bureau doesn’t have.

With inaccurate workload estimation models and front loading the Census Bureau overrecruited, overhired on many operations in preparation for the final major operation: non-response followup. One of the major costs was the paper based operational control system PBOCS which has been the subject of intense scrutiny by media, Congress and employees because of its inability to check out, check out and ship questionnaires and generate management reports. The managers who are monitoring productivity and costs are trained to believe if the reports don’t show it’s done then it isn’t done. With only erroneous reports to rely on, headquarters and regional offices are using a take no prisoners do whatever it takes attitude to pressure temporary employees to complete the task. PBOCS also moves assignment areas fooling LCO managers and field staff into thinking they have more or less work than they have. And ultimately this may have long term geography problems when the Census is completed and used for congressional redistricting.

Since PBOCS doesn’t work correctly and fails to handle the workload, The Census Bureau runs on a more is better attitude. The solution is hire more employees for manually counting and reviewing enumerator questionnaires when they should have slowed enumerator production. Local Census offices have gone from a simple 9am-5:30pm operation to running three shifts 24 hours a day seven days a week with triple to quadruple what their staffing authorizations originally allowed. This compounded the bottleneck, increased the backlog of questionnaires waiting to be checked in and slowed the re -interview and quality assurance phase. There is overwhelming suspicion of data falsification and false proxies but by the time this is figured out the operation will end and the enumerators already released for lack of work.

Now what was originally touted as the most accurate decennial count ever has quickly turned into a race to meet production goals and wrap up the operation as quickly as possible with procedural changes.  We have enumerators, telephone clerks in the LCO, and enumerators from other LCOs taking interviews ignoring the fact that PBOCS will only let you check it in under an enumerator and that if data falsification is happening it will be difficult to find the culprit. What were originally any six personal and telephone visits is now three visits go to a proxy. What used to be try to get the household member because he knows his own name, sex, age, DOB, Hispanic origin and race and whether he rents or owns has become going to a proxy on a first visit and sometimes writing don’t know on most if not all of those questions. Sadly this actually passes the office review portion and nothing in the enumerator procedures disallows that. If a respondent refuses and a proxy is able to give any of the information no matter how knowledgeable he/she is that doesn’t constitute marking it as a refusal, skewing the accuracy of the data.

The incentives of career census employees at RCC and headquarters are in contradiction with each enumerator who wants our city to be accurately counted. The career census employees’ evaluation of performance is purely based on numbers how many cases are completed with little regard to the demography or difficulty of enumerating the population. Their expectation is that the enumeration of traditionally undercounted minorities of Bedford Stuyvesant be just as quick as the white, upper middle class of Upper West Side of Manhattan. The very same agency whose motto has always been the leading source of data about the nation’s people and economy has become a competition between area managers and local census offices.

The leadership in the local census offices isn’t the strongest either. Those who made hiring decisions in New York RCC had every chance to hire the best managers but instead resorted to nepotism to make decisions. When it was clear these decisions were poor the career census employees terminated LCO managers’ employment to cover it up. But then found another disappointing replacement. In an attempt to bring operations up to speed the Census Bureau flew in managers from Denver into Manhattan and headquarters to Staten Island.

The goal is for enumerators to get as many cases in and clerks process work as quickly as possible doing whatever it takes to get the job done, otherwise there will be a formal written reprimand and termination of their employment. It is the chest beating, gorilla apelike attitude of the managers that will ultimately be the demise of New York City.

Lester Farthing, the Regional Director and his managers of the New York Regional Census Center have no intention of an accurate count in the five boroughs. Instead their goal is to appease headquarters, finish as quickly as possible so that the career census employees will be viewed as productive team players who are not questioning the possible inaccuracy of this count. As one of our area managers will say “it’s a hot mess.” I only hope the mayor of our great city Michael Bloomberg, city census coordinator Stacey Cumberbatch, politicians and congressmen are reading this letter and will intervene because ultimately it is the city that will suffer for the next ten years. They were quick to make public announcements touting the importance of participating in the census by returning the forms. But have yet to do anything to persuade non cooperative households and real estate management companies to allow enumerators in to complete their job. The sad reality is that it may be too little too late.

With the way the census works can any of us ever trust census data again?

Ed O’Keefe reports on Census Bureau’s final stats on mailback “participation” rates

Thursday, April 29th, 2010

H/t to Ed O’Keefe of The Washington Post:

72% of households responded to 2010 Census

Take a gander at the documents Groves shared with reporters at his announcement earlier today:

The 2010 Census response rate matched returns for the 2000 Census, the U.S. Census Bureau said Wednesday.

Seventy-two percent of American households returned questionnaires by last week and 28 states had higher response rates than 10 years ago. Seven of the 10 most populous counties matched their 2000 response rates as did eight of the 10 most populous cities, the Census Bureau said.

Census Director Robert Groves estimated that between 46 million and 49 million households did not return questionnaires. Temporary census takers hired by the agency will hit the streets starting this week and will visit those addresses up to six times to get answers. The agency will further outline those plans at a news conference Monday.

Groves said he anticipates critics will question why this year’s results only matched the 2000 response rates despite a multimillion-dollar advertising and outreach campaign, but he called this year’s results “unbelievable” because survey response rates have dropped significantly in the past decade.

Socioeconomic concerns rather than race or ethnicity appeared to drive lower response rates, Groves said. Less-educated, lower-income households appeared to respond less. The nation’s foreclosure crisis also contributed to the lower rates, he said.

The total cost of 2010 Census operations — budgeted for about $14 billion — will be known once officials get a complete tally of households that did not respond, Groves said.

Belated Earth Day Special: The Census Bureau Waste Continues (with hard evidence attached)

Wednesday, April 28th, 2010

Last week, I planned to publish this piece, but the data from a New York area census office didn’t come in until yesterday…Check it out:

Here’s the hard evidence:

10-02 DISPOSITION OF 2010 GQAV MATERIALS (1)

10-10 DISPOSITION OF 2010 QAC MATERIALS

It seems like the Census didn’t know April 22nd was Earth Day. In honor of it the printers ran non stop from morning to midnight in 494 offices across the nation printing out all the address listing pages and assignment preparation for Non Response Followup.

Cost to print NRFU Address Listing Pages of every housing unit in the United States single sided and then ship it to the National Processing Center Fed Ex Priority Overnight

Cost to print out hundreds upon thousands of maps single sided only to not even be looked at

Cost to print all the training materials on high quality printer quality paper

Cost to print all the glossy recruiting brochures, partnership posters only for them to be unopened and thrown out by the palette like this everyday (see pictures below)

–  Some food for thought. These boxes are filled with 500 brochures a piece and has been happening everyday for months and in all 494 offices everyday –

Cost to print all the Be Counted Questionnaires which were all taken back from the Be Counted and Questionnaire Assistance Centers to be thrown away even though New York City wanted to extend the program by 30 days and some to count the estimated 500,000 illegal immigrants.(see attached disposal list)

Cost to print all the GQV Questionnaires which we still have two palettes left. (see attached disposal list) And that is just one of the forms on the attached list to throw out…Here we go:

10-02 DISPOSITION OF 2010 GQAV MATERIALS (1)

10-10 DISPOSITION OF 2010 QAC MATERIALS

Photos of materials on their way to be destroyed/recycled:



Security Alert: Haz-Mat teams at Baltimore Census processing facility…again.

Wednesday, April 7th, 2010

From the Baltimore Sun…too bad this plot line never made it into The Wire:

Suspicious powder was found this morning at a processing center of the U.S. Census Bureau in Essex, though it was found to be harmless.

The Baltimore County police and fire departments were called around 10 a.m. to the U.S. Census Bureau’s Data Capture Center, in the 8400 block of Kelso Drive, to examine a package containing a small amount of powder, said Elise Armacost, the fire department spokeswoman.

“We did not evacuate the building, no one was showing symptoms, there were no transports,” Armacost said. “People were removed from the immediate vicinity while we were testing.”

The substance was deemed non-threatening, Armacost said.

The plant — which processes 2010 Census questionnaires as they are returned — resumed operations by early afternoon, said Markia McLeod, a spokeswoman for the U.S. Census Bureau.

McLeod said that Baltimore County Hazmat teams have responded to the Essex facility for six incidents since it opened last year. The last time was in March, when powder found inside a census envelope turned out to be coffee creamer.

– Erica L. Green

Census Bureau Press Release: Second Round of Census Forms Mailed to 40 Million Households…Targeted Mailing Reminds Residents There is Still Time to Return Questionnaires

Tuesday, April 6th, 2010

The following is a Census Bureau press release that just came into the inbox:

To reduce the estimated $2.7 billion cost of following up with
households that fail to mail back their 2010 Census questionnaires, the
U.S. Census Bureau has begun mailing second forms to approximately 40
million housing units in areas that had below-average response rates in the
2000 Census.

“Census Bureau and a multitude of private sector research shows that
sending a replacement questionnaire to households can significantly
increase response rates in the end,” Census Bureau Director Robert Groves
said. “We estimate that the second mailing could increase America’s mail
participation rate in the 2010 Census by 7 to 10 percentage points, and
doing so would save taxpayers more than $500 million.”

According to the Census Bureau, every percentage point increase in the
national participation rate by mail saves about $85 million. It costs the
government just 42 cents in a postage paid envelope to get a questionnaire
back in the mail, but it costs taxpayers an average of $57 to count a
household that fails to mail it back.

Second questionnaires were mailed last week to every housing unit in
areas that had a mail response rate of 59 percent or less in 2000, or about
24.7 million households. The questionnaires were sent to all households,
regardless of whether they had already returned their 2010 Census form.

In areas that had response rates between 59 and 67 percent — below the
national average of 67 percent — replacement forms will be sent only to
households that have not yet mailed back their completed 2010 Census form.
These 15 million households will receive a second form April 6-10.

Households have until mid-April to mail back their forms before census
takers begin going door to door to residences that failed to respond.

“We understand that people lead busy lives and may not have gotten
around to sending back their forms yet,” Groves said. “The replacement form
gives them a second chance to get counted and help ensure that their
community gets its fair share of political representation and federal funds
over the next 10 years.”

Currently, the national mail participation rate is 60 percent, with some
of the lowest rates in Alaska, California, Texas, New Mexico and Oklahoma.
The latest national and local participation rates can be viewed at
http://2010.census.gov/2010census/take10map.

State-by-state breakdown: How was the response rate in your state?

Friday, October 2nd, 2009

The following is a list of the percentages of U.S. households that returned census questionnaires in 2000 by mail or submitted information on another form, over the phone or by Internet (The nation averaged a 67 percent response rate):

- Alabama: 61 percent

- Alaska: 56 percent

- Arizona: 63 percent

- Arkansas: 64 percent

- California: 70 percent

- Colorado: 70 percent

- Connecticut: 70 percent

- Delaware: 63 percent

- District of Columbia: 60 percent

- Florida: 63 percent

- Georgia: 65 percent 

- Hawaii: 60 percent

- Idaho: 67 percent

- Illinois: 69 percent

- Indiana: 69 percent

- Iowa: 76 percent

- Kansas: 71 percent

- Kentucky: 66 percent

- Louisiana: 60 percent

- Maine: 61 percent

- Maryland: 69 percent

- Massachusetts: 69 percent

- Michigan: 71 percent

- Minnesota: 75 percent

- Mississippi: 63 percent

- Missouri: 69 percent

- Montana: 68 percent

- Nebraska: 75 percent

- Nevada: 66 percent

- New Hampshire: 67 percent

- New Jersey: 68 percent

- New Mexico: 62 percent

- New York: 63 percent

- North Carolina: 64 percent

- North Dakota: 72 percent

- Ohio: 72 percent

- Oklahoma: 64 percent

- Oregon: 68 percent

- Pennsylvania: 70 percent

- Rhode Island: 67 percent

- South Carolina: 58 percent

- South Dakota: 74 percent

- Tennessee: 65 percent

- Texas: 64 percent

- Utah: 68 percent

- Vermont 60 percent

- Virginia: 72 percent

- Washington: 66 percent

- West Virginia: 64 percent

- Wisconsin: 75 percent

- Wyoming: 66 percent

Source: U.S. Census Bureau

It’s Too Late To Stop The Presses…

Thursday, July 23rd, 2009

The following is a press release from the U.S. Census Bureau:

FOR IMMEDIATE RELEASE
THURSDAY, JULY 23, 2009

Jack Martin/Shelly Lowe                               CB09-CN.17
Public Information Office
301-763-3691                                                2010 Census Web
site
e-mail: <pio.2010@census.gov>                         2010 Census sample
form

Printing of 2010 Census Questionnaires Under Way
New 10-Question Survey Among Shortest Since First Census in 1790

The U.S. Census Bureau has begun printing 2010 Census questionnaires as
the agency continues preparations for next year’s count of the U.S.
population. The new questionnaire, which every residential address will
receive, is designed to be one of the shortest since the first census in
1790, asking just 10 questions and taking about 10 minutes to complete.

“Our goal is to count everyone living in the United States once, only
once, and in the right place,” said Census Bureau Director Robert M.
Groves. “Making that happen begins with the 2010 Census questionnaire, a
powerful tool that provides critical data that will guide representation in
Congress and the distribution of more than $400 billion in federal funds to
state, local and tribal governments every year.”

Beginning in mid-March 2010, more than 120 million questionnaires will
be delivered to U.S. residential addresses. To meet the goal, the Census
Bureau will print more than 1.5 million documents every day.

For the first time, more than 13 million questionnaires will be
bilingual (English – Spanish). The move is based on tests showing that
targeting the bilingual questionnaires toward areas with high
concentrations of Spanish-only speakers will improve response rates.
Questionnaires are also available on request in Spanish, Chinese
(simplified), Korean, Vietnamese and Russian. Language guides, which
provide instructions on how to complete the questionnaire, are available in
nearly 60 languages.

“The Census Bureau has gone to great lengths to make the printing
process as efficient and eco-friendly as possible,” Groves said. “The
printing of 2010 Census questionnaires uses
30 percent less ink than 10 years ago and will be printed on 30 percent
recycled paper.”

Another critical factor in the success of the census is the quality of
the address list used for delivering the questionnaires next March. This
spring, Census Bureau workers walked every street in the nation to match
actual residential addresses on the ground with those provided in lists
from the U.S. Postal Service and local governments.

The 140,000 workers who verified addresses operated out of 151 local
census offices in the U.S. and Puerto Rico.  In the fall, an additional 344
local census offices will open.

The Census Bureau will hire approximately 1.4 million people to conduct
the 2010 Census, including following up with households that do not return
their questionnaire.

“The 2010 Census is easy, important and safe,” Groves said. “The Census
Bureau is ready to undertake this massive domestic operation and looks
forward to everyone’s participation in the national count.”