Senator Gregg, Senator Hollings and members of the Subcommittee:

Good morning, Mr. Chairman and thank you for the opportunity to appear before you today to discuss the Bureau's proposed FY 2000 budget amendment for Census 2000. I am pleased to be here to share with you the reasons why this additional funding is necessary, and to let you know how much I appreciate the attention you've given Census 2000. I would like also to acknowledge the nearly full funding for Census 2000 provided by the House Appropriations subcommittee in their mark-up of our FY 2000 request. And, I would like also to express my gratitude to this Committee for recognizing that the rest of the Census Bureau must continue its critical work in the economic and demographic areas while Census 2000 takes place. We appreciate the funding for the non-decennial census programs provided in the Senate Appropriations bill, and strongly urge the Committee to maintain that level of funding when the bill is conferenced with the House bill because of the disparity between your level of support and that provided in the House Appropriations subcommittee mark-up.

The Census Bureau is requesting an amendment to the fiscal year 2000 budget for the decennial census totaling 1 billion 723 million dollars, bringing the total request for the decennial census to 4 billion 512 million dollars. The additional funding requested is a direct result of the January 25, 1999 Supreme Court ruling that statistical sampling could not be used in Census 2000 to determine the population count used to reapportion seats in the House of Representatives. There are also a few low-cost improvements based on recent dress rehearsal evaluations and lessons learned from developing the address list. I am sure that you are well aware of the impact the Court ruling had on our previous plan for conducting Census 2000, and that you also understand that Census 2000 was already in process when this ruling was handed down. The Bureau responded quickly to this ruling, preparing a revised operational plan by mid-March. Completion of the revised plan permitted the Bureau to reassess its funding needs and prepare the amendment you have before you today.

There are two primary reasons for the amended budget request: to improve the accuracy and completeness of the count without sampling we've added new operations designed to enumerate households not otherwise included in the census, and we've significantly increased the level of quality assurance on existing operations that affect the accuracy and completeness of the count. But before I outline for you the major components of the revised plan for which additional funding is required, I should note the importance of Congress providing adequate funding for Census 2000 in a timely manner. As I will explain in my testimony, Census 2000 is well underway. The train has left the station. We cannot park this train on October 1, even for a few days, without severe consequences to the schedule and accuracy of the census. For example, October 1 is a critical date for buying air time for our campaign scheduled to launch Census 2000 media advertising in November. Even a day or two delay will cause us to miss out on bidding for prime advertising time, seriously diluting the effectiveness of the campaign. Other field operations also cannot be interrupted without causing severe disruptions in the census.

By far the most important aspect of this new, revised plan and for which we're now requesting the largest increase is in our field data collection activities and support systems. In this component alone we're requesting an additional 1 billion 451 million dollars, or 84 percent of the total.

A 100 percent nonresponse followup operation increases the total expected door-to-door workload from thirty million housing units in our earlier design to forty-five million housing units - that is, about a fifty percent increase in the number of housing units that must be visited, many on more than one occasion, because they will not have returned a census questionnaire through the mail. This workload increase over the previous census design requires an additional four weeks of labor-intensive field work, for a total of ten weeks, to complete this nation-wide effort. Because of the significant increase in the housing units to be visited, we must seek out many more people willing to take a temporary job knocking on doors. We will review and screen about 1,000,000 more applicants than expected under the previous plan, in order to reach our ultimate yield of about 800,000 qualified candidates. And, because of the increase in the number of housing units to be enumerated and the need to hire hundreds of thousands more applicants, resulting in lower overall skill level, we expect the average productivity of the temporary workforce to decrease from previous estimates. The combination of the increased workload and lower productivity estimates requires 938 million dollars to put more people on the streets for a longer period of time to ensure we count the entire population where they live.

We require an additional 229 million dollars to put into place a thorough effort to improve operational coverage. By that I mean we must include extraordinary measures to improve the quality of the census data by including both a program for checking on an estimated seven and a half million housing units initially classified as either vacant or non-existent and for visiting areas of new construction where the address list completed earlier would not have picked up new housing units.

There are also other critical activities we must conduct to improve the accuracy and completeness of census data for apportionment without sampling. For example, because the 1998 Dress Rehearsal coverage evaluation for Columbia, South Carolina indicated a high probability for seriously undercounting rural areas, we've enhanced quality assurance procedures in the final update of the address list for rural areas. There are also more people in segments of the population in special circumstances that require increased effort if we are to have a complete and accurate count without sampling. These segments are often the most difficult to count. Included are those in urban and rural areas where standard procedures are not effective and thus require special procedures. Then there are the linguistically isolated, people in shelters, and that segment of the population now living in group homes or quarters. These operations require about 214 million dollars.

We're also increasing the infrastructure needed to support these many expanded operations. This requires about 279 million dollars, providing for forty-four additional offices in order to effectively distribute the increased nonresponse followup workload, increases in office space and staff for the originally planned four hundred seventy-six offices, as well as outfitting the new and expanded offices with all the equipment and supplies necessary to conduct the field operations and administratively support a workforce in excess of eight hundred thousand enumerators and supervisors. We will also keep all five hundred twenty field offices open one month longer than originally planned, and we must expand the telecommunications capabilities of all field offices to accommodate the significant increase in the data transmission requirements associated with one hundred percent nonresponse followup and to handle the administrative data such as payroll for the much expanded workforce. Finally, we expect that, because we're expanding the media and promotional aspects of the Census 2000 effort, the number of telephone inquiries will increase significantly, necessitating expansion of the contractor-operated Telephone Questionnaire Assistance operation.

As a direct result of the Supreme Court ruling prohibiting the use of sampling for determining the population count used for apportioning Congressional seats, we redesigned the Accuracy and Coverage Evaluation to now restrict the focus to all other uses of census data. This allowed a reduction in the number of housing units to be included in the survey from seven hundred fifty thousand to about three hundred thousand. In terms of field costs we expect this operation to cost almost 209 million dollars less in 2000 than the Integrated Coverage Measurement survey.

Continuing updates for the Master Address File are fundamental to the preparations for mailing to and following up on housing units from which we do not receive a questionnaire. The Bureau must conduct a comprehensive, just-in-time address list update to include all newly-constructed housing units. This includes preparing and shipping additional address lists and maps, and then updating the geographic database with information provided by government units and other community entities participating in the new construction program. We will also launch a new operation to locate the correct address for people with multiple residences and people who have moved during the now extended data collection period. These address listing operations account for about 10 million dollars of the increased funding requirement.

Because the redesigned census operations will increase the data capture workload by an estimated twenty three million forms and questionnaires, the Bureau must extend operations by two months at all four data capture centers to ensure adequate opportunity to acquire the data through the data capture operation and conduct quality assurance activities. The equipment, software licenses, and telecommunications infrastructure for Census Bureau headquarters and in the National Processing Center in Jeffersonville, Indiana also must be expanded to handle increased data processing and data transmission requirements associated with the expanded volume and complexity of census data and the need to monitor more closely a much larger overall census operation.

The increased processing volume and extended and expanded operations requires 136 million dollars, which has been offset by a 5 million dollar reduction in the processing requirements of the Accuracy and Coverage Evaluation relative to the earlier design.

The funding required for Puerto Rico, the Virgin Islands and the Pacific Territories has increased by just over 34 million dollars. The largest contributing factor is that the one hundred percent nonresponse followup requirement increases the workload in Puerto Rico by an estimated 140,000 units. Comparable to enumeration activities of the fifty states and the District of Columbia, we'll conduct added coverage improvement operations, expand field infrastructure, increase telecommunications capability, add staff, and expand other activities such as recruiting. Operations in Puerto Rico account for 25 million dollars of this increased funding requirement. In addition, recently revised estimates we have received for the work to be done in the Island Areas will cost 9 million dollars more.

For the Bureau's program to reach out to the American people through advertising and other mediums, we're requesting an additional 88 million dollars. The bulk of this increase is associated with additional media messages, expanded promotion activities designed to achieve a higher level of responsiveness from those segments of the population traditionally most difficult to enumerate, and increases for the Census in the Schools program. These changes account for roughly 73 million dollars. One of the new media messages is designed for late Fall and early Winter to educate the American people about the census and the benefits to their communities. The second new media message is designed to encourage cooperation with field interviewers when they are following up on housing units from which no questionnaire was returned. The remaining additional 15 million dollars is to provide expanded promotion and outreach to State, local and tribal governments and community-based organizations. This funding primarily will be directed to in-kind support such as posters, defrayment of printing costs, and other activities which encourage local participation in all aspects of Census 2000.

Earlier in my statement I indicated that I would briefly outline some significant accomplishments of ongoing Census 2000 operations.

1. Since Fiscal Year 1999 began, we've completed planned address list development activities in both rural and urban areas, including a visual spotting of about 99 percent of housing units in the country. Work is continuing with local officials to resolve potential address list conflicts.
2. We've awarded 35 printing contracts and over 180 million forms are already printed, and we're currently preparing the computer file of addresses to be used for printing addresses on the questionnaires that we will mail or deliver next March.
3. We've opened and staffed 130 Local Census Offices and have leased space for 380 of the remaining 390 offices. Leases for the remaining 10 offices will be executed before the end of the Summer. These offices will begin opening this Summer and all offices will be occupied by October 1, 1999, with management and support staff in place.
4. We've opened and staffed the Regional Census Centers, which are the twelve temporary offices that will monitor operations at the Local Census Offices, collecting not only the operational data but maintaining the administrative support such as the daily payroll updates for all offices. These offices must remain open through completion of all Census 2000 field work.
5. We've opened the Baltimore, Maryland data capture center and are in the process of testing and proving in the equipment and software there. And today we're opening the data capture center at our National Processing Center in Jeffersonville, Indiana. We're also working on final preparations for opening the two remaining data capture centers, in Pomona, California in October and Phoenix, Arizona in November in order to conduct necessary testing in preparation for Census 2000.
6. We've overseen the design of the television advertising, and our contractors are filming and producing the ads. The schedule for purchasing air-time is complete, and we're making purchases.
7. Finally, the partnership activities are continuing at an increasing pace, with approximately 22,000 agreements already in place. Over 400 partnership specialists have been hired to promote the census at a grassroots level.

Census 2000 is well underway, with many critical activities accomplished, and I'm pleased to report, on schedule. However, the success of Census 2000 depends on continuing scheduled activities at an ever accelerating pace through the balance of Fiscal Year 1999 and on into Fiscal Year 2000. Without the additional funding we are requesting, it will not be possible to carry out a fair and accurate census. With your continued attention to the critical nature of census activities and the timely manner in which funds are provided, we can maintain the pace necessary to accomplish what is required to fulfill our Constitutional duty to provide census data.

Again, thank you for this opportunity to express our additional funding needs. I will be happy to answer any questions you may have about this request or any other aspect of Census 2000 operations.