PREPARED STATEMENT OF
CHARLES LOUIS KINCANNON
DIRECTOR
US CENSUS BUREAU
Coverage Improvement Plans for the 2010 Census
Before the Subcommittee on Information Policy, Census,
and National Archives,
U.S. House of Representatives
Field Hearing, San Antonio, Texas
9 July 2007
Chairman Clay, on behalf of the U.S. Census Bureau, thank you for inviting me to testify. Congressman Gonzalez, Congressman Cuellar, Congressman Rodriguez, I’m very pleased to be in San Antonio and back in my home state to discuss our plans to improve coverage and accuracy for the 2010 Census. In the past decade a number of important changes have taken place both here in Texas and throughout the country. Our nation is growing and becoming increasingly more diverse. Among the noteworthy changes over the last decade is the continued growth of the Hispanic community—our nation’s largest minority community. The importance of this growth cannot be underestimated. In the last census, we noted the presence of new Latino communities throughout the nation, as well as in traditional states such as Texas. And the growth of the Latino community has continued throughout this decade. It is therefore critically important for the Census Bureau to ensure that our plans for 2010 will provide accurate data that will be used to create opportunity for thousands of Latino communities throughout America.
Addressing this challenge is the primary goal of the 2010 Census—that is our goal is to improve census coverage and accuracy. Improved coverage and accuracy encompasses reductions in census undercounts, overcounts, and other errors, such as geocoding (locational) errors. Today’s hearing comes at a key moment in the census lifecycle. Preparatory field activities for the 2008 Dress Rehearsal and the 2010 Local Update of Census Addresses (or LUCA) program are underway. These activities will help us to improve the accuracy of the short-form census count—the count used to apportion the U.S. House of Representatives.
We are optimistic based on the experience from Census 2000 that we can further improve the accuracy and coverage in 2010. One of the most significant achievements of the 2000 Census was reducing the undercount of the nation’s historically undercounted communities, including Hispanics, African Americans, and American Indians living on reservations. In fact, the final estimated undercount in Census 2000 for Hispanics was less than one percent (.71), whereas 1990 the undercount was nearly five percent (4.99). The undercount of African Americans was reduced from approximately four and half percent (4.57) to less than two percent (1.84). And the undercount of American Indians on reservations was reduced from twelve percent (12.22) to less than one percent (.62). This success was in large part the result of the extraordinary outreach effort and the efforts of the Census Bureau to assess and meet the needs of local communities in conducting the enumeration whether we were in the inner city neighborhoods of St. Louis or the colonias of South Texas. Working with partners in the nation’s hardest-to-count communities was key to achieving trust and encouraging participation.
We worked with national organizations, including NALEO, MALDEF, LULAC, and the National Council of La Raza, as well as the faith-based community and locally elected officials to reach out to the Hispanic community. And to meet the needs of the hardest-to-count communities, such as colonias, in addition to outreach, we also used different strategies in the field. As all of you know, gaining trust from colonias residents is particularly challenging. To meet this challenge in 2000, we worked with the state of Texas to identify the areas with colonias. In addition, we partnered with the University of Texas-Pan American in Edinburg in developing our enumeration strategies, which included the use of “update/enumerate” methodology. With update/enumerate, we not only update the census address list, but we also conduct enumeration at the same time. (Meaning we do not leave the questionnaires to be mailed.) We also developed a new strategy, the use of cultural facilitators—or “promotoras.” The promotoras were our partners at the street level. They accompanied the enumerator, or census taker, assisting the census taker in gaining access to the community and encouraging participation. Their assistance was invaluable. For the 2010 Census, we are trying to build on this success and we anticipate using these or similar strategies.
In fact, based on our experience with past censuses and the plans and testing developed this decade, we know improvements can be gained throughout the census process—from address list development to better response strategies, as well as outreach among the nation’s hard-to-count communities. And it is especially meaningful that we discuss these activities at a field hearing, because many of our efforts are focused on operational improvements and better strategies to encourage participation, ranging from the comprehensive communication and outreach program to the content of the census questionnaire itself.
In preparation for the 2010 Census, we have conducted extensive planning and
testing, including two national mailout tests and two large-scale field tests,
and we are now integrating planned improvements in the 2008 Dress Rehearsal.
Thanks to congressional support, this will be the best-planned and most tested
census in our nation’s history. These efforts, which began shortly after
Census 2000, are based not only on the coverage evaluations, but also evaluations
of other census programs. These evaluations have allowed us to generally characterize
both those population groups missed in the census and those counted in error.
Among those populations groups missed or undercounted are children, especially
young children; adult African American males; recent immigrants; and persons
living in large households, complex households, or in group quarters. Those
who are counted in error, or overcounted, include college students and persons
with multiple residences, as well as persons living in group quarters.
The Census Bureau’s primary objective for the 2010 Census is to improve coverage to reduce these errors. These strategies encompass operational improvements and strategies to encourage participation. They are targeted on eight specific areas.
Throughout the decade, the Census Bureau has placed a high priority on improving the accuracy of the census address list (the Master Address File or MAF). We are working to develop a complete, up-to-date address list and to identify the best enumeration method to use in every census block throughout the country. The Census Bureau uses a “mailout/mailback” enumeration method to obtain responses for the majority of the country. For these areas, which have city-style addresses (e.g., 101 Main Street), we rely on the Postal Service to deliver the census questionnaire through the mail and the respondent to return the questionnaire. In other areas, we may use “update/leave” in areas that have non-city style addresses (e.g., Rural Route 202, Box 34, or a locational reference, such as “white house with green shutters, end of the road”). With update/leave, a census-taker delivers the questionnaire to the household and the household can then respond by mail.
We are working extensively with the U.S. Postal Service to update the address list twice a year using the Delivery Sequence File. This is a file maintained by the Postal Service that includes every address for which they deliver mail. It is a rich resource and we are using this file to a greater degree than in the previous census. In addition, we are working with local communities. Many communities have undergone what is known as Emergency-911 (E-911) address conversions where local governments have developed a number-street address for emergency services. While the Postal Service may not recognize these addresses for many of these communities, we can use these E- 911 addresses to locate households more effectively.
Similarly, we are also working to ensure we have an accurate, up-to-date list of group quarters, including college dormitories, nursing homes, military barracks, jails, prisons, and other facilities. We are working to improve our methods for identifying, locating, and enumerating these facilities. As part of the address canvassing operation, which takes place in the spring of 2009, we are including group quarters. We did not do this in the past, and our goal is to ensure that we can locate and properly identify or determine whether the address is, in fact, a group quarters location. We have also included the group quarters in the address lists that are being sent later this year to governments as part of the LUCA program. By including the group quarters, we hope local governments will verify or update the group quarters locations. This will help to ensure we have an accurate and complete list of group quarters, and that none are missed.
Many of the planned coverage improvements are focused at key field and processing operations. For instance, working for the first time with handheld computers we are using for data collection, our census takers will be able to take advantage of GPS technology to locate addresses much more precisely. We are implementing improved data processing and questionnaire tracking—we will follow the outcome for each questionnaire we send out much more closely. The handheld computers will also improve coverage in other ways. We intend to update the assignments after the field work begins—something that a paper-based census does not allow. If we receive a late mail return from a household before sending the census taker to the respondent’s address, we will remove that household from their assignment lists. This will greatly increase our efficiency in the field and also help to reduce duplicates or overcounting.
It is important to remember, however, complete coverage in the census ultimately depends on participation. To encourage participation requires a comprehensive approach, encompassing a wide-range of strategies, including questionnaire content, response options, and outreach. Beginning with the census questionnaire, we have worked throughout the decade to minimize respondent burden and encourage every household in America to respond to the census.
It is also worth noting our efforts to make the short-form census questionnaire more user friendly. In an effort to obtain more accurate information, we conducted extensive testing of the wording for the race and Hispanic origin questions and have made some modifications that should improve our reporting for detailed race and Hispanic Origin. We are also providing clearer instructions for the census residence rules. We have tested a number of options to ensure we get an accurate count of each and every person, once and only once, at the right place. The Census Bureau’s task is to count every person at their “usual residence”—where they live most of the time. We are reminding respondents to include everyone who should be counted. We are also including additional queries to determine whether a person may also have been counted somewhere else, such as a college dormitory.
Another important area of outreach is with regard to language. Our goals for language are to provide effective ways for our nation’s diverse communities to participate and respond to the census. For the first time, we will mail bilingual Spanish-English census forms to neighborhoods with higher proportions of people who need assistance in Spanish. We are also identifying areas throughout the United States where informational materials in other languages, including language assistance guides, may help respondents in completing the census form—which is also available in five languages. We will also encourage local hiring efforts to ensure we can hire enumerators with relevant language skills to work in our nation’s diverse neighborhoods.
The success of the census will also be enhanced through partnership and outreach. It is our commitment to work with communities throughout the nation so each and every person living in the United States is counted on Census Day. To do so, the Census Bureau is planning a multi-faceted and integrated communications program to reach every community. As part of the overall communications strategy, we intend to incorporate the lessons and successes of Census 2000, which included both paid advertising and reliance on partnerships. The goals of the 2010 Census communications program are to increase the mail response rate, encourage cooperation during the non-response follow-up operations, and to improve accuracy and reduce the differential undercount. We will announce the award of the 2010 Census communications contract in early September, and the first deliverable of this contract will be an integrated communications plan in May 2008.
One of the most important lessons of Census 2000 was the need to engage state, tribal, and local governments. The LUCA program is an important partnership opportunity for every tribal, state, and local government to review the Census Bureau’s address list and to submit either corrections or additions. LUCA contributes directly to the accuracy of the census. We have made significant improvements to the program for the 2010 Census, including a longer review period and three participation options. We sent advance letters to each of the nearly 39,000 governments throughout the country early this year, and we will mail the official invitations later this summer. Governments will be allowed, starting this fall, 120 days to review and update the Census Bureau’s address materials. The Census Bureau is also conducting 14 tribal consultation meetings with federally recognized tribes across the country. Our mutual goal is to identify ways for the Census Bureau and the tribes to work better together to ensure the most accurate count in Indian country for the 2010 Census.
Partnerships, whether they are formed through tribal, state, or local governments or through community organizations, are important to the success of the census. These partnerships encourage participation and help demonstrate the importance of the census to every community. Partners are often our best ambassadors in hard-to-count communities.
As I mentioned at the beginning, this census is the most planned and tested census in our nation’s history. Our primary goal for the 2010 Census is improved coverage and accuracy. We believe all of these coverage improvement efforts contribute to this goal, because the census affects every community throughout America from the historic African American neighborhood of North Webster in St. Louis to the Frio Heights colonia in Frio County, Texas. Representative Clay, I can assure you that the Census Bureau is fully committed to the goal of reaching the residents of America’s communities, especially the hard-to-count.
Thank you for supporting the goals of the census. I would be happy to answer
any questions.
Source: U.S. Census Bureau
Public Information Office
(301) 763-3030
[an error occurred while processing this directive]
