Secretary's Speech

Statement of U.S. Secretary of Commerce Carlos M. Gutierrez
Before the United States House Appropriations Subcommittee on Commerce, Justice and Science

Thank you for the opportunity to come before you today to discuss the Census Bureau and the status of the 2010 Census. The goals of the Department of Commerce are far reaching, from increasing American competitiveness to growing American exports and protecting America’s environment. Measuring American life is also a significant ongoing goal, which comes into sharper focus every 10 years with the decennial census.

As you know the Census Bureau is part of the department’s Economics and Statistics Administration. The 2010 Census is one of the highest priorities and most important responsibilities of the department.

Today, I am here to update you on the progress we have made toward addressing some of the challenges currently facing the 2010 Census and to seek the committee’s support in ensuring that the upcoming Census is our best ever. The American people expect and deserve a timely and accurate decennial census, and we at the department are working tirelessly to ensure they get it.

The 2010 Census
First, I thought it might be helpful to review briefly how the Census field operations work.

Since last October, the Local Update of Census Addresses, or LUCA, has been underway. This program provides state, local and tribal governments with an opportunity to review their address lists. The Census Bureau will then verify and update their address list in the address canvassing operation, which will begin in mid-April 2009.

In mid-March 2010, the Census Bureau will mail a form to every address on the list, with the expectation that a large number (60 to 70 percent) will respond. A second, follow-up mailing will be sent out on April 10, 2010, to those who have not responded. As you know, it is vital that people respond by mail, as this will help save the government significant costs.

The intense field work starts in the Non-Response Follow-Up, or NRFU, which begins in April 2010 and goes through the end of June, where enumerators go door-to-door to personally interview those who have not mailed in their forms. The Census Bureau also conducts special field operations for group quarters and other residences not effectively covered by mail at that time.

From July through November 2010, the Census Coverage Measurement process, an intense effort with specially trained people to do a detailed analysis of the count, takes place. This effort helps evaluate if there is an under- or over-count by providing a detailed measure of the quality of coverage in the Census.

The final count by state will be sent to the President on December 31, 2010, and detailed block-level counts for redistricting at the state level will be delivered by April 1, 2011. The full data tables will be made available in late 2011 to early 2012.

While taking the Census sounds simple and straightforward, it is anything but. The scope and complexity of the Census has steadily increased. It is the largest peacetime mobilization in the United States.

The sheer volume of participants and infrastructure necessary to carry out the Census can be overwhelming. To staff 12 regional Census Centers (RCCs) and over 450 local Census offices with the best possible workforce, the Census Bureau conducts over one million applicant interviews and hires 140,000 address canvassers and more than 580,000 enumerators.

This undertaking is increasingly supported by state-of-the-art technology, including GPS address mapping, wireless handheld computers and sophisticated electronic and paper-based data integration systems.

The challenges with the Decennial Census are increasing. Population growth means there are more people to count and more housing units to visit. Increasing amounts of “junk mail” may also reduce household responses to mailed questionnaires.

Two-income families, increasing transience, and the changing shape and diversity of the American family all combine to make an accurate count increasingly difficult. Enumerators often have to return to homes numerous times in order to find someone at home.

We also know that it is more difficult to reach people with our advertising. In Census 2000 we could rely on television, radio, and print media. The migration away from passive, traditional forms of media means we have to find new ways to get our message across to the people we need to reach.

And, survey responses are down, in part because there is in general less of a willingness to complete a government survey than may have existed in the past. Recent data suggest that a general distrust of government has increased, as have concerns about privacy.

As a result of these factors, the number of enumerators needed has steadily risen and the productivity in NRFU, in terms of households answering per hour, has declined.

Planning for the 2010 Census
To deal with some of these challenges, at the beginning of the decade, the Census Bureau reengineered the 2010 Census to build a better, faster, and simpler Census.

The plan was to:

1) Leverage technology more effectively, including GPS mapping and handheld computers;

2) Replace the Decennial Census long form with the short form only Census;

3) Introduce the American Community Survey to capture long form annual data for economic and social characteristics;

4) Take further steps toward a multi-language Census; and

5) Coordinate and target messaging with integrated communications and partnerships.

I’m pleased to report that the Master Address File/Topologically Integrated Geographic Encoding and Referencing (MAF/TIGER) Accuracy Improvement Project, the GPS portion of the reengineering, is on schedule for completion this month (other than island areas). Utilizing GPS will reduce the geographic coding problems and help get more accurate counts in rural areas.

And, the American Community Survey is fully implemented. It has a response rate of 97 percent and is providing far more timely data than the once-a-decade long form.

With the 2010 Census we are aggressively trying to knock down anything that keeps people from answering the Census. Language is increasingly one of those barriers.

In 2010, for the first time a bilingual form will be sent to neighborhoods with large Spanish speaking populations, as identified in the American Community Survey. And, as before, the form will be available in other major languages such as Chinese, and a “help guide” will be printed in 40 additional languages.

This multi-language approach is helping remove obstacles and encourage participation in the Census by households in which English is not the primary language.

In addition, we are working closely to ensure that our messages and outreach are coordinated and targeted more effectively with our integrated communications and partnerships.

However, while we are pleased that these components are on track, problems have developed with the Field Data Collection Automation Program, or FDCA. The FDCA is a key component of the reengineered short-form 2010 Census. Non-FDCA related planning challenges have also surfaced that require immediate attention.

I am here today because the FDCA project has experienced significant schedule, performance, and cost issues. A lack of effective communication with one of our key contractors has significantly contributed to the challenges.

As I have said before, the situation today is unacceptable, and we have been taking steps to address the issues. But first, let’s answer the question, how did we get here?

The decision to automate data collection was made at the beginning of the decade, just after the 2000 Census. Initially, the Census Bureau worked to develop an automation effort using in-house resources. The Bureau realized after the 2004 test that this was a massive technology development effort that required outside expertise.

In 2006, a contract for FDCA was signed with Harris Corporation providing for a comprehensive approach to the use of handheld computers and information technology for managing the decennial operations. In 2007, the Address Canvassing dress rehearsal was conducted, at which time development and scoping problems emerged. Reports from the Census Bureau’s field staff, consultants from the non-profit MITRE Corporation working for the Bureau, and the Government Accountability Office confirmed these problems. The department’s Inspector General also raised concerns.

In late 2007 and early 2008, more than 400 new or clarified technical requirements were identified by the Census Bureau. Upon the realization of the large scope of requirement changes, Census Director Murdock established the 2010 Census FDCA Risk Reduction Task Force, to begin to propose and evaluate options to keep the FDCA program on track. These efforts served to clarify the issues and confirm the urgent need for action.

To address the issues, we sought advice from a wide range of respected sources, including the Office of the Inspector General, the Government Accountability Office, the MITRE Corporation, the 2010 Census FDCA Risk Reduction Task Force, and an independent panel of experts I formed last month, which included Dennis Hastert, Former Speaker of the House; Kenneth Prewitt, former Census Director now of Columbia University; Vincent Barabba, former Census Director now of Market Insight Corporation; George Ligler, an independent software developer and consultant; Ron Ponder of Hudson Heights Partners; and John Thompson, the Associate Director for the 2000 Census who is currently at the National Opinion Research Center.

The expert panel evaluated the “go forward” options identified by the 2010 Census FDCA Risk Reduction Task Force. A majority of the expert panel members recommended that we proceed with the alternative recommended by the Task Force, which calls for the Census Bureau to utilize a paper-based NRFU methodology, as opposed to using the electronic handheld units for NRFU.

Assessments by the MITRE Corporation, our Inspector General and the leadership of the Census Bureau echoed this recommendation to utilize a paper-based NRFU. Based on these inputs and serious consideration of the benefits and pitfalls of each option, today I am reporting to this committee that we will move forward with the recommendation to use a paper-based NRFU in the 2010 decennial census.

Pursuing this recommendation allows both the Census Bureau and the Harris Corporation to focus on what they can do best. This path still represents a significant step forward for the Census Bureau in the use of automation. Address Canvassing, the operation that occurs the year before the decennial to validate and update the location of every household location in the country will be conducted using GPS-enabled Harris handheld computers allowing for the most accurate and comprehensive address list in the Census’ history.

The paper-based NRFU we propose will be supported by state-of-the-art hardware and software that will help ensure the most accurate Census. The Census Bureau has extensive experience conducting paper NRFU and they will build on that experience for the 2010 decennial census.

Our flexibility to respond to unforeseen difficulties is greater under the paper-based option because our ability to deploy people is greater and faster than the ability to manufacture additional handhelds on short notice. Additionally, the majority of expert panel members recommended this alternative, believing that it offers greater flexibility to ensure a successful decennial at this late time in the planning and development cycle.

Cost Estimates
The effect of moving forward with this alternative, as well as the non-FDCA related planning challenges we have faced will require an increase of $2.2 to $3 billion dollars through Fiscal Year (FY) 2013. This will bring the total lifecycle cost of the 2010 Census to between $13.7 to $14.5 billion.

These costs are driven in large part by increases in the numbers of people who will be needed to carry out the 2010 Census; these include enumerators and personnel to service the help desks, data centers, and the control system for the paper-based NRFU. There are also additional costs that result from more recent increases in gas prices, postage, and printing.

We estimate that to successfully pursue this new course, an increase of approximately $160 to $230 million is needed for FY 2008. The exact amount needed is still contingent on final procurement decisions to be made soon. This funding will allow Harris to continue its development of an automated Address Canvassing operation and the operational control system, and for the Census Bureau to immediately begin developing a paper-based NRFU system for 2010. While we continue to assess the cost impact for FY 2009 and beyond, our current rough estimates place the additional 2009 funding need at approximately $600 to $700 million. We expect to have more concrete numbers for 2009 when we submit to the Congress an amendment to the President’s 2009 Budget that will be fully offset. The majority of the lifecycle cost increase occurs in 2010, and we and the Census Bureau are evaluating all options to keep the cost increase towards the low end of the range, while ensuring an accurate count.

To address the funding need for 2008, we have developed a transfer proposal from existing departmental resources that will fully cover the resources required for the 2010 Census. Our transfer proposal would require legislative authority to remove certain limits on the department’s ability to make intra-departmental funds transfers, and we are submitting to you proposed legislative language that would provide this authority.

I realize that a transfer proposal such as this is extraordinary; however, I believe it is the responsible thing to do in light of the importance of the Census and our shared need to engage in fiscal restraint and avoid additional burdens on the American taxpayer.

Executing the Replan
Both at the Census Bureau and at the Department of Commerce we have made substantial management changes to address the challenges facing the 2010 Census. We are working to ensure that there is clear accountability and that we have set specific leadership expectations. This includes better integration between Census and Harris personnel; rapid decision making; real-time problem solving; and improved transparency, oversight, and communication.

We are taking this very seriously and hope these changes and others reflect our concern and ultimately our resolve to better serve the American people.

I am personally engaged in this matter and will continue to devote my time to this issue until we can be assured that we have established a sustainable and achievable path forward to a successful 2010 Census. In addition, I have devoted significant resources from the Department of Commerce to assist our colleagues at the Census Bureau. The 2010 Decennial Census is a responsibility we all share at the Department of Commerce and we are pulling together to ensure the problems we face are addressed. I look forward to your questions and working together on a path forward.