Mr. Chairman, Mrs. Maloney, and Members of the Committee:
It is a pleasure to be here today, 4 days after Census Day, to report on the status of Census 2000. When I last testified, the focus was on whether the Census Bureau could pull off the many complex and massive operations--questionnaire mailout and delivery, telephone questionnaire assistance, reminder card, data capture--that would occur between early March and now. Mr. Chairman, all of these operations were conducted successfully with no major problems that would put the census at risk.
We successfully delivered some 24 million questionnaires to housing units in the update/leave areas on schedule.
The U.S. Postal Service completed the delivery of some 98 million questionnaires to housing units in mailout/mailback areas.
Telephone questionnaire assistance (TQA) has worked well since we became fully staffed. We have about 6,000 agents working in 22 centers and they have answered nearly 5 million calls since TQA became active on March 3.
The U.S. Postal Service completed the delivery of 120 million cards reminding people to mail back their census forms.
Data capture operations are performing well, without any major problems.
Other operations, which I will discuss later, have also begun on schedule and are progressing.
In short, Mr. Chairman, the Census Bureau is keeping its part of the civic bargain. Operations are progressing on schedule, within budget, and without any disabling problems. Now the census is in the hands of the people and we hope they will answer the call to civic duty and respond. This is one of the few cases where the personal decisions of individuals could have a tremendous impact on governmental operations. The Congress has made response to the census mandatory. We will make repeat calls and visits to every housing unit that does not mail back a questionnaire. I urge everyone to help his or her community have the most complete count possible.
In your letter of invitation, you ask for The status of nationwide mail response rates and what those rates translate into for the Nonresponse Followup (NRFU) workload, hiring, other operations, and associated costs. As of this writing, the national mail response rate as posted on the Internet is 55 percent. The current estimates for nonresponse followup workload, hiring, and costs are based on achieving a 61 percent mail response rate; as you can see, we are not there yet. It is not too late for people to mail back their form, but time is running out. April 11 is the cutoff date for identifying housing units that have not mailed back a questionnaire so we can include them in the nonresponse followup workload. We will continue to receive and process mail returns after that date. On April 17, we will produce a late mail return file that we will transmit to the local census offices so they can delete those addresses from their nonresponse followup assignments.
On March 27, we successfully posted mail response rates on our Internet site--www.census.gov, have updated the numbers every day since, and will continue to do so daily until April 11. We will post final mail response rates on April 18. In January, Secretary Daley and I announced a new initiative to encourage grassroots participation in Census 2000 in every town, city, county, state, and tribal area in the nation. We are calling this initiative "How America Knows What America Needs." We are providing local elected officials with tools to use in encouraging their constituents to complete and mail back their census forms and to cooperate with enumerators. These tools include sample news releases, articles, talking points, and other written materials; a dedicated website to enable participants to obtain updated information and download promotional materials; and a toll-free number to allow elected officials to call for additional information. We are also challenging communities to increase their overall response rates in Census 2000 by at least five percentage points over their 1990 level. We are calling this component '90 Plus Five, which means we are encouraging a 70 percent national response rate--the 65 percent from 1990 plus 5. Obviously, we have not reached that goal nationally, but some jurisdictions have or will reach their goals. A second component of "How America Knows What America Needs" is called Because You Count. This component is aimed at increasing cooperation with census enumerators when they come knocking on doors. Our goal with the "How America Knows What America Needs" initiative is to convert the census into a civic event of the highest order.
Operational Update
In your letter of invitation, you asked for The status and a brief overview of the Census 2000 operational time line, and readiness for key activities and dates that lay ahead.
The next big operation is the nonresponse followup, which I will talk about at some length later. Let me just take a few minutes now, to provide additional details about operations we recently completed or started.
We began the update/leave operation on March 3, as planned. Census enumerators left questionnaires at approximately 24 million housing units in areas (including Puerto Rico) that have several different address types. These areas are mostly in small towns and rural areas where address systems have less geographic structure. Local census offices conducting update/leave operations had sufficient staff and some 70,000 enumerators were hired for this operation. We sent regional and headquarters staff, as needed, to troubleshoot in those few local census offices that experienced problems. This operation is complete and on schedule.
Telephone questionnaire assistance (TQA) also began on March 3 and will run through June 8 and outbound calling from the TQA sites as part of our coverage edit program will continue into mid-June. There are 7 toll-free telephone numbers (in English, Spanish, Chinese, Korean, Vietnamese, Tagalog, and TDD) where people can call to get assistance in filling out their questionnaire, get replacement questionnaires, get language assistance guides, or provide their census questionnaire information. As I mentioned earlier, we have answered nearly 5 million calls. Just over 4 percent of calls were unable to get through; almost all of those were on the first two days after questionnaires were delivered in mailout/mailback areas. There was a higher volume of calls earlier than expected but we moved quickly to increase staff so that we could successfully handle the calls. There were also some early problems in validating the questionnaire data that was taken over the telephone; these problems have now been resolved.
As you will recall, in our advance letter, we also provided an opportunity for those who want a form in Spanish, Chinese, Korean, Vietnamese, or Tagalog to request one. We have received about 2.5 million such requests, the vast majority for the Spanish form, and this is an indication that the advance letter worked.
From March 13 through March 15, the U.S. Postal Service successfully delivered questionnaires to some 98 million addresses in what we call mailout/mailback areas of the country. These addresses are generally in urban areas but can occur in small and mid-sized towns and even some rural areas. There were some instances where households received duplicate questionnaires. This occurred because during all the overlapping processes used to build the Master Address File, we wanted to minimize the chance that we would eliminate an address that should be retained. We have procedures in place to eventually remove these duplicate addresses from our files before the final census data are tabulated.
Also beginning March 13, and continuing throughout March and April, census enumerators are visiting about half a million housing units in list/enumerate areas, in an operation similar to that initiated in Alaska on January 19. These are remote, sparsely populated areas where it is not efficient to compile a precensus address list. At the time enumerators visit each housing unit, they are listing the unit and completing a questionnaire. Also at this time, we began the update/enumerate operation, which is conducted in communities with special enumeration needs and where most housing units may not have house number and street name addresses. These areas include selected American Indian Reservations, unincorporated Spanish-speaking communities along the border of Texas and Mexico (colonias), and resort areas with high concentrations of seasonably vacant housing units. Both of these operations progressed more slowly than we anticipated at the start, but we are confident that we will be able to complete these operations on time.
Beginning on March 20, we mailed out a reminder card to those housing units we are asking to return a form by mail. Many had already mailed back their forms, but this reminder hopefully spurred others to do so as soon as possible. That mailing was also completed successfully.
Last week, we completed the service based enumeration, with the goal of including in the census those people without conventional housing who might be missed in the regular enumeration procedures. On the evening of March 27, census enumerators interviewed people in shelters, on March 28 at soup kitchens and mobile food van stops, and on the early morning of March 29 at targeted outdoor locations. We enumerated about 22,000 such places over the course of the three days.
On the evening of March 31, we conducted the transient night operation, which is designed to enumerate people at locations where residents are highly transient in nature, such as campgrounds and parks, commercial or public fairs, carnivals, marinas, racetracks, and recreational vehicle parks. This operation will extend until April 14 for a few very large and relatively stable Transient Night locations.
Beginning April 1 and continuing until May 6, we expect to count about 7 million people in about 125,000 special places during group quarters enumeration. These are people living in such places as college and university dormitories, hospital and prison wards, migrant farm camps, and nursing homes. From April 4 to May 4, we will also conduct the enumeration of land-based and shipboard military personnel and people aboard U.S. flag-bearing merchant vessels--about 1,000 ships and over 500 military reservations in all. This group includes all U.S. land-based military personnel, their dependents living on base, and U.S. Navy and U.S. Coast Guard ships not deployed or assigned to foreign ports.
In your letter of invitation, you asked about The status of data capture systems for all four sites. Data capture operations began shortly after the first questionnaires were delivered in update/leave areas and the volume of returned questionnaires picked up significantly with the delivery of forms in mailout/mailback areas. Data capture is working very well and we are aware of no serious problems at this point. We have checked in about 67 million questionnaires and we have scanned about 24 million of them. Scanning accuracy, which is over 99%, has exceeded our expectations and we are also maintaining a very high level of keying accuracy.
You also asked about Preparation and issues concerning Internet response to Census 2000 questionnaires and tabulation processes. Internet data collection and questionnaire assistance began on March 3. Internet data collection will operate until April 15 and Internet questionnaire assistance will end the first week of June. We have received nearly 50,000 responses through the Internet. Some respondents have experienced problems completing the questionnaire through the Internet. This is primarily due to a bug in their browser and not caused by the website itself. If they are unable to get through after trying a different browser, we have recommended that they fill out and mail back their paper form.
Your letter also asked about Status and issues confronting Questionnaire Assistance Centers and Be Counted sites. Questionnaire Assistance Centers opened on March 8 and will be open through April 14. There are no major problems related to QAC's at this time. We have provided assistance to over 100,000 individuals in the QAC operation. Approximately 15,000 volunteers and 11,000 paid clerks have been assigned to staff the QAC's. To maximize use of staff, we have eliminated redundant sites and currently have 24,000 in operation.
Be Counted forms became available on March 31 at approximately 19,000 sites in addition to the QAC's, where they are also available. The Be Counted sites are not staffed, but clerks are assigned to set them up, restock them, and close them after April 11.
Your letter also asked about Any difficulties confronting Local Census Offices. None of the 520 LCO's is experiencing problems that have prevented normal operations. The tornado last week in downtown Ft. Worth, Texas, did not harm the office, but staff were prevented from entering the downtown section of the city for the next few days. Some LCO's are reporting minor problems with their telephone systems and headquarters staff are working closely with the General Services Administration and telecommunications service providers to resolve the problems. At present, all systems are up and running.
At the last hearing, I reported problems with a New York City LCO. GSA sought another location but no suitable site could be found. GSA has now assumed many services to assure the office is able to perform its work. Basic housekeeping services are being provided by a GSA contractor. An extermination service has solved the rodent problem and the water leak has been fixed. We are now working on getting an elevator into working order for freight and employees who need assistance and in the meantime are using an elevator in an adjacent facility.
Nonresponse Followup
Mr. Chairman, the nonresponse followup (NRFU) operation is the largest, most complex, and most costly operation in Census 2000. NRFU is scheduled to begin April 27 (enumerator training begins April 24) and will continue for 10 weeks until the first week of July. Extending NRFU beyond that date would not only increase census costs, it could lead to a reduction in data quality. Experience teaches us that the longer we are in the field and the farther we get from Census Day, the more the quality of respondents' answers deteriorates. Therefore, our goal is to complete NRFU in 10 weeks. This is an ambitious schedule, but as I describe below, we have productivity goals and multiple quality checks built into the process. And we will stay in the field as long as it takes to complete this process.
You asked about The status of the hiring process for NRFU. We believe we have a sufficiently large recruiting pool to fill all positions. While we have met our national goal of having 2.4 million qualified applicants well in advance of our April 19th target date, we are continuing to accept applications and to actively recruit in local areas where we have not yet met our recruiting goals. This targeted approach will ensure that we have sufficient enumerators in all areas of the country. We have selected supervisory staff for NRFU and about one-third of the enumerators. We will complete selection of enumerators over the next three weeks.
I would now like to describe in some detail the enumerator's job and our procedures for assuring the quality and completeness of their work.
Each NRFU enumerator is assigned a specific area in which to work, called an assignment area, and is given a binder of addresses in that area that includes all those addresses for which we have not received a completed questionnaire. Because houses without numbers and street name addresses can be difficult to find, enumerators in rural areas also receive maps that have the housing unit locations spotted on them. The NRFU enumerator must go to each address in the assignment area to complete the appropriate questionnaire (either short form or long form) for the housing unit and its occupants.
If the current household lived at the address on Census Day, the enumerator interviews a household member at least 15 years of age and completes the assigned questionnaire. If the unit was occupied by a different household on Census Day, the enumerator completes a questionnaire for the occupants who lived there on Census Day by interviewing a knowledgeable person, such as a neighbor. If the current occupants were not enumerated elsewhere, the enumerator will also complete a census questionnaire for them for their Census Day address.
If the housing unit was vacant on Census Day, the enumerator completes appropriate housing questions on the questionnaire by interviewing a knowledgeable person, such as a neighbor or apartment house manager. If the housing unit was demolished or otherwise nonexistent under census definitions, the enumerator completes a questionnaire that provides the reason why the unit should be deleted from the census address list, by interviewing a knowledgeable respondent such as a neighbor or apartment house manager.
The enumerator must make up to six attempts to complete a questionnaire. If no one is home at an occupied housing unit, the enumerator obtains as much information as possible about how to contact the occupants from a neighbor, building manager, or another source. The enumerator also leaves a notice at the address that they have visited and provides a telephone number so the occupant can call back. The enumerator then makes up to two additional personal visits (3 in all) and three telephone attempts at contacting the household before obtaining as much information as possible to complete the questionnaire from a knowledgeable source. Enumerators are instructed to make their callbacks on different days of the week and at different times of day. The enumerator must maintain a record of callbacks that lists each type of callback made (telephone or personal visit) and the exact date and time it occurred. Enumerators are expected to obtain complete interviews but must obtain at least the status (occupied or vacant) and the number of people living in the unit. If the enumerator submits a questionnaire that contains this minimal level of data, the crew leader must check the enumerator's record of callbacks for the housing unit to determine that procedures were properly followed. The crew leader also holds these cases for possible further followup to obtain more complete data.
Supervisors, called crew leaders, meet daily with each enumerator to pick up and check completed work. These meetings enable crew leaders to monitor progress and quality. Crew leaders are expected to make sure that the enumerators produce quality work at a rate of 1 to 1.5 completed questionnaires per hour depending on the type of area the LCO covers. Crew leaders check each completed questionnaire for completeness and accuracy.
In order to prevent falsification of the data by enumerators, a percentage of each enumerator's work is verified for accuracy by a reinterview staff. This staff verifies a sample of each enumerator's work and may also verify additional questionnaires from enumerators whose work differs significantly from that of other enumerators working for the same crew leader. An enumerator who is discovered falsifying data is dismissed immediately and all the work must be redone by another enumerator.
Daily production levels begin to decrease toward the end of NRFU. Sometimes enumerators completed the easier cases first, finished the work closest to their homes first, or believed that the quicker they finished their assignment, the sooner they would be out of work. In order to bring the NRFU to closure within the scheduled 10 weeks, we implement a procedure known as "final attempt." When the area covered by a crew leader has completed 95 percent of its workload, the crew leader consolidates the remaining work and gives it to the most productive and dependable enumerators. These enumerators then make one final visit to each outstanding NRFU address and to some of the housing units for which only minimal data was earlier collected to complete as much of the questionnaire as possible. Final attempt must resolve all outstanding cases within a few days, but NRFU is not over until a questionnaire is completed and checked into the local census office for every unit.
Long Form Questionnaire
Now, I want to address concerns you are hearing from some Members and constituents about the Census 2000 long form. We take these questions and concerns seriously. We are aware of the time pressures confronting people in our overworked and over stressed society and of the concerns we all have about privacy and confidentiality. I am confident that once the public understands that their answers are protected by law, that every question asked in the census serves an important purpose, and that every question has a specific federal legislative or judicial mandate or requirement, they will perform their civic duty to respond to the census. Failure to collect the data on the long form would prevent agencies from carrying out existing laws and could result in loss of benefits to local communities.
I mentioned earlier the importance of answering the census, whether one has the short form or the long form. I also want to stress that Title 13 of the United States Code not only requires respondents to answer the census, but also requires the U.S. Census Bureau to maintain the strictest confidentiality of the data collected. The law forbids the Census Bureau from giving an individual's information to anyone else--not the Federal Bureau of Investigation, the Internal Revenue Service, or the Immigration and Naturalization Service; not local governments, or police or landlords. All census employees are subject to severe fines and imprisonment if convicted of violating this law. I understand that many people have deep distrust about providing information to the government. At a time of increasing concerns about the security of our personal data, people will understandably want to know why the data are needed, what they will be used for, and how they will be protected. Answers to census questions are confidential because the Congress has provided iron-tight protections to the data and serious penalties for those who are not vigilant in protecting the data.
Concern with overburdening respondents with too many questions led the Census Bureau to introduce a long form on a sample basis into the 1940 census, and we have used this
approach in each decennial census since then. The selection of a sample, based on established
scientific methods, means that not everyone is asked every question; the large majority receive
only a short form. Thus, the need for information is met with much less burden on the public than if everyone had to answer the questions on the long form.
In particular, the long form for Census 2000 includes the questions asked on the short form and additional questions, for a total of 52. In contrast, the long form included 57 questions in 1990, and 81 questions in 1940. This is the shortest long form the Bureau has ever used.
The questions on the long form provide essential data so that decision makers can help move our rapidly growing and economically complex country in the 21st century. Every question on the form serves an important purpose. Let me just give a few examples.
The answers to the question on a telephone in the home are used to help plan local 911 emergency services and, in response to the Older Americans Act, to provide emergency and healthcare services to areas with relatively more homebound seniors who do not have telephone service.
The answers on plumbing facilities are used by the Department of Housing and Urban Development to administer housing programs and by local communities to apply to the federal government for community development funds.
The answers to the question on income are used to allocate Title 1 education funds to school districts with high rates of student age poverty.
The information on how long it takes to commute to work is used by federal, state, local, and private transportation planners who design new roads, bus routes, mass transit systems, and manage traffic congestion, as well as for the distribution of federal transportation funds.
The information on disability is required to help design and deliver public and private services to the elderly and disabled.
The data on veterans' status are used to plan the location of veterans' hospitals and cemeteries and for the delivery of veterans' healthcare and nursing services.
The law requires that, three years prior to Census Day, the Census Bureau report to Congress the subjects proposed for inclusion in the census. The Census Bureau reported this information to Congress in a letter and accompanying materials dated March 28, 1997. The law also requires that we report to Congress the specific questions we intend to ask two years prior to Census Day, and we did that on March 30, 1998. The materials we submitted to Congress describe each question we included on the long form and, more importantly, provide detailed legal citations that indicate that each item is mandated or required by congressional legislation or federal judicial decisions.
Accurate census data also provide the underpinnings for other federal surveys and data collections. The decennial census forms the sampling base for other national surveys and is used to compute rates of various indicators. Therefore, it is directly linked to the statistical system's
ability to provide current unemployment data; to provide data for making cost of living adjustments; to calculate numerous vital statistics and rates for health service utilization; to calculate crime, imprisonment, and victimization rates; and the like.
Only one new subject is included in the long form for Census 2000: grandparents as caregivers. This addition complies with legislation passed in the 104th Congress requiring that the decennial census obtain information about grandparents who have primary responsibility for care of grandchildren.
We intend for Census 2000 to be the last decennial census to include a long form. The American Community Survey, scheduled for nationwide implementation in 2003, is one of the most important improvements in federal statistics, and it is the cornerstone of our effort to keep pace with ever increasing demands for timely and relevant data. The American Community Survey will revolutionize the way we take the decennial census by making the next one simpler and making it possible to provide detailed socioeconomic and housing data throughout the decade.
There were no viable alternatives to having a long form for Census 2000. No other data source could provide all the information the Nation needs in a cost-effective manner. But, in the long term, the American Community Survey will replace the long form by Census 2010.
Without the Census 2000 long form and the American Community Survey during the next decade, the federal government would not have the data currently required to allocate billions of dollars in federal funds for education programs, veterans' services, programs to reduce unemployment and stimulate economic growth, public healthcare, and services for the elderly. The states and local governments would not have the data they need to help them plan new school construction, occupational and vocational education programs, programs to assist the elderly and the disabled, highway safety and public transportation systems, the location of police and fire department personnel, and rural development.
If there were no long form in Census 2000, each federal agency would have to undertake its own data collection efforts to fill the gap. The burden on respondents would actually increase; moreover, this would be an inefficient use of federal dollars. Alternatively, if the agencies did not conduct their own data collections, they would be forced to use ever more outdated information from the 1990 census.
Mr. Chairman, that concludes my testimony and I will be happy to answer any questions.