American Community Survey 5-Year Data Product Plans
The Bureau of the Census (Census Bureau) currently releases American Community Survey (ACS) data products in the form of 1-year estimates and 3-year estimates. Most recently, the 2008 ACS 1-year estimates were released in September 2009, and the 2006-2008 ACS 3-year estimates were released in October 2009. By this notice, the Census Bureau announces plans for the release of ACS 5-year data products covering the period of 2005-2009. The release of the ACS 5-year estimates will achieve a goal of the ACS to provide small-area data similar to the data published after Census 2000, based on the long-form sample. This notice provides general information on the Census Bureau's modifications to its current line of ACS data products to accommodate the 5-year estimates.
Table of Contents Back to Top
- FOR FURTHER INFORMATION CONTACT:
- SUPPLEMENTARY INFORMATION:
- I. Purpose of the ACS 5-year Data Products
- II. Summary of Comments Received and the Response of the Census Bureau
- 1. Block Group Level Geography
- 2. Types of Data Products
- 3. Restrictions Required for Disclosure Avoidance or Statistical Reliability
- 4. Frequency of Data Release
- III. ACS 5-year Data Products Plans
- Paperwork Reduction Act
DATES: Back to Top
The Census Bureau plans to release 2005-2009 ACS data in December 2010. The plan for the 2005-2009 ACS data products will be implemented on September 20, 2010.
ADDRESSES: Back to Top
Please send any correspondence about the Census Bureau's American Community Survey 5-year estimates or data product plans to Sharon M. Stern, Assistant Division Chief, American Community Survey Office, Room 3H463, Mail Stop 7500, Washington, DC 20233-7500.
FOR FURTHER INFORMATION CONTACT: Back to Top
For information about the Census Bureau's American Community Survey, contact Sharon M. Stern, Assistant Division Chief, American Community Survey Office, on (301) 763-5638, by e-mail at firstname.lastname@example.org, or by mail at Room 3H463, Mail Stop 7500, Washington, DC 20233-7500.
SUPPLEMENTARY INFORMATION: Back to Top
I. Purpose of the ACS 5-year Data Products Back to Top
The ACS is part of the 2010 Decennial Census Program and provides annually updated, detailed demographic, socioeconomic, and housing information for communities across the United States and Puerto Rico. One goal of the ACS is to provide small-area data similar to the data published after Census 2000, based on the long-form sample data. This goal will be met with the release of the 2005-2009 ACS 5-year estimates.
On March 6, 2009, the Census Bureau published a Federal Register notice (74 FR 9785) that proposed releasing the 5-year estimates using the same set of ACS data products that were produced for the ACS 3-year data estimates, and included proposed geographic summary levels for the 5-year data products.
Descriptions of the suite of ACS data products follow:
Detailed tables include the most detailed ACS data and cross-tabulations of ACS variables.
Download files provide the detailed table estimates in comma-delimited, ASCII-formatted files that are in the standard Census “Summary File” format.
Data profiles provide separate fact sheets on social, economic, housing, and demographic characteristics.
Narrative profiles provide clear, concise, textual descriptions of the data included in the data profiles.
Subject tables include detailed ACS data, organized by subject such as employment, education, and income.
Selected population profiles provide social, economic, and housing characteristics for a large number of groups based on race, Hispanic origin, country of birth, and ancestry.
Geographic comparison tables allow the comparison of ACS data for a given time period across a variety of geographic areas.
Thematic maps provide graphic displays of the data available from the geographic comparison tables, which compare ACS data for different areas in a given time period.
Public Use Microdata Sample (PUMS) Files provide access to ACS microdata for data users to create summaries that are not available as ACS summary products.
In the March 6, 2009 Federal Register notice, the Census Bureau sought input and feedback on the suite of data products for the ACS 5-year estimates and in particular, expressed interest in data users' specific feedback on the following four dimensions:
1. Block Group Level Geography—The Census Bureau proposed releasing block group data only as downloadable Summary Files through the American FactFinder Download Center. The American Factfinder is the electronic system for access and dissemination of Census Bureau data on the Internet. Tables can be accessed through the American Community Survey Data Sets page on American FactFinder or downloaded in file format from the American FactFinder Download Center.
2. Types of Data Products—The Census Bureau proposed releasing 5-year estimates in detailed tables, summary files, subject tables, data profiles, narrative profiles, selected population profiles, thematic maps, geographic comparison tables, and PUMS files. Narrative profiles and selected population profiles were not proposed for particular geographic summary levels, such as block groups.
3. Restrictions Required for Disclosure Avoidance or Statistical Reliability—As done with all data released by the Census Bureau, the proposal included restrictions on the release of 5-year estimates that were based on disclosure avoidance requirements.
4. Frequency of Data Release—The Census Bureau proposed that ACS 5-year estimates be released annually.
II. Summary of Comments Received and the Response of the Census Bureau Back to Top
The Census Bureau received comments from 26 organizations and individuals regarding the four above-mentioned categories, in response to the March 6, 2009 Federal Register notice. Some commenters addressed more than one category in their comments. All comments have been summarized and organized according to subject matter. The subject matter categories are: (1) The option of alternative dissemination methods for data at the block group geography level, (2) the types of data products to be included in the 5-year data products, (3) the limitations on the availability of the 5-year estimates due to restrictions required for disclosure avoidance and statistical reliability, and (4) the proposed annual release for the ACS 5-year data products. Comments were provided by a variety of Federal and State agencies and organizations, non-profit policy research and analysis organizations, non-governmental organizations, and a private sector company. Federal, State, local, and private sector organizations from agencies representing the transportation community provided 17 of the total 26 comments received. All comments received are posted on the Census Bureau's ACS Web site, http://www.census.gov/acs. A summary of the comments and the Census Bureau's response is below.
1. Block Group Level Geography
The Census Bureau received six comments in response to the question of using downloadable Summary Files rather than releasing tables on American FactFinder for block group data. All six comments were in favor of ACS producing block group level data and releasing the block group data tables separately from the standard ACS tables currently found on American Factfinder. Two comments strongly recommended not releasing block group data tables on American Factfinder. Specifically they wanted the block group data released with cautions and instructions for combining data for block group areas into larger geographic areas “to achieve greater reliability.”
In considering this proposal, the Census Bureau reviewed the complexity of using the block group data with the sheer volume of the estimates to be produced for approximately 210,000 block group geographies and agreed that releasing tables on American Factfinder was not the preferred approach. As a result of public comments and staff review, the Census Bureau will release to the public through the American Factfinder Download Center the block group estimates only as files that can be accessed by more sophisticated users.
2. Types of Data Products
The Census Bureau received thirteen comments in response to the proposed data products for the ACS 5-year estimates. The comments were all in support of the data products proposed; one group interested in data on American Indian and Alaska Native (AIAN) populations requested specific AIAN data products. The Census Bureau agrees with the comment and plans to produce data specifically for American Indians and Alaska Natives with the release of the 2006-2010 ACS American Indian and Alaska Native Summary File. One comment from the transportation community requested expanded transportation data, and six comments were concerned about how confidentiality protections might limit the availability of detailed transportation data at very low levels of geography. These confidentiality protections focus on tables that have many cells but few sample cases. To address the concerns expressed in the comments, the Census Bureau has for several years been working with the transportation community to develop ACS data products that provide a balance between the low geographic levels required by the work of the transportation community and confidentiality protections required by the Census Bureau's collection authority, Title 13. For example, the ACS 1-, 3- and 5-year standard data products now contain 59 new transportation tables that were not a part of the Census 2000 standard data products; some of these ACS transportation tables had formerly only been provided through custom tabulation requests paid for by the transportation community (in Census 2000 and earlier censuses). The Census Bureau is also providing some tables with low geographic levels of transportation data sooner and more frequently in the ACS data products than they have from past Censuses. Finally, the Census Bureau continues to work in collaboration with the transportation community to determine the best set of products to provide more data for very small geographic areas without violating confidentiality protections. One comment from a private sector firm offered suggestions for organizing ACS data for download to enhance analysis. The Census Bureau agrees with this comment and is developing improvements for data products available for download.
To arrive at a final plan for the data products to be released for the 2005-2009 ACS, the Census Bureau considered all comments and also undertook a comprehensive staff review of the many ACS data products released for the 3-year estimates. The objective of this review was to determine if those products were appropriate for very small counties, towns, and incorporated places, as well as for specific government data uses and public use. The final plan for the ACS 5-year data products will provide a very large percentage of the data that were previously found in Census Summary Files 1 and 3. Some tables previously provided only upon request for a custom tabulation will be made available routinely to the public in the standard ACS data products. Additionally, tables not present in Census 2000 data products have been added to this set of available ACS tables. When comparing the plan for the 2005-2009 ACS to what was released in Census 2000, most of the new tables reflect new content, but some tables were added because they were determined by subject matter experts to be desirable by data users.
Census Bureau staff also reviewed the practical matter of providing public access to the large volume of data being produced by releasing 5-year estimates for such a large number of geographic areas. Staff reviewed the available data products and tables and determined that a reduced set of tables will be released on American FactFinder with the remainder to be available to the public through downloadable Summary Files from the American FactFinder Download Center.
The Census Bureau plans to deliver to the public the tables for the 5-year estimates on American FactFinder in a single release. The release of the Summary Files (including all data at the block group level) and the PUMS files will follow soon after the initial release of tables on American FactFinder. The plan for future releases of the ACS 5-year data products may be subject to change as Census Bureau staff improves the data products and receives input and feedback from data users.
3. Restrictions Required for Disclosure Avoidance or Statistical Reliability
The March 6, 2009 Federal Register notice directed readers to a file containing supplementary information located on the Census Bureau's Web site (http://www.census.gov/acs). The table describing disclosure avoidance protections was in the file that provided this supplementary information, and these protections were listed by number. The Census Bureau received 20 comments in response to the proposed disclosure avoidance. Three comments supported the Census Bureau's plans for disclosure avoidance. Some commenters, mostly from the transportation data community, had comments or suggestions concerning disclosure avoidance. Disclosure avoidance number seven from the March 6, 2009 Federal Register notice stated
For the residence and workplace tables where means of transportation (mode) is crossed with one or more other variables, there must be at least three unweighted workers in sample for each transportation mode in a given place for the table to be released. Otherwise the data must be collapsed or suppressed and complementary suppression must be applied. There is no threshold on univariate tables.
Commenters expressed objections to disclosure avoidance number seven, stating that its implementation would negatively impact data needed for planning requirements, particularly for very small geographic levels including traffic analysis zones and block-groups.
In response to these concerns, Census Bureau staff reviewed the published disclosure avoidance and determined that number seven, which impacts residence and workplace tables where means of transportation (mode) is crossed with one or more other variables, does not apply to the standard ACS data products, but rather it applies only to some custom tabulations that the Census Bureau produces upon request. However, because of general Census Bureau Disclosure Review Board restrictions pertaining to non-residential geographies, some of the tables of interest to transportation data users will still be limited or suppressed. The Census Bureau is aware of the needs that the transportation community has for these data and has expanded the number of transportation tables in the ACS standard annual data products to include ones that were not provided in Census 2000 standard tabulations. In addition, the Census Bureau continues to work closely with staff from the Department of Transportation to produce custom tabulations that will fit the needs of transportation data users and that also uphold the Census Bureau's duty to protect the confidentiality of ACS respondents.
Additionally, another commenter was concerned that disclosure avoidance number six would suppress data for small reservations and many Alaska Native Village Statistical Areas. Disclosure avoidance number six from the March 6, 2009 Federal Register notice stated:
For workplace tables, there must be at least 50 unweighted or 300 weighted workers in sample over the 5-year period in a given workplace for the table to be released.
Census Bureau staff recognizes the difficult balance in producing tables for small populations and ensuring that confidentiality is protected. Disclosure avoidance number six, originally developed for Census 2000 data, had a restriction of workplace tables to areas with 50 unweighted or 300 weighted workers. The 300 weighted workers restriction was based on the 50 unweighted workers and the Census 2000 average long form weight of six. The Disclosure Review Board, upon closer review for the ACS 5-year data products, decided that the key restriction to protect confidentiality for the ACS was the 50 unweighted workers, so the reference to a weighted number of workers has been dropped. Census Bureau staff also expanded the language on disclosure avoidance number six to clarify that in addition to workplace tables, the requirement for at least 50 unweighted workers in sample over the 5-year period applies to all non-residential geographies including residence 1 year ago and place-of-birth tables.
4. Frequency of Data Release
The Census Bureau received five comments on the proposed annual release of the ACS 5-year estimates. All five comments were in favor of the annual release.
III. ACS 5-year Data Products Plans Back to Top
The Census Bureau is releasing its plans for the ACS 5-year data products via the Web. The plan provides a list of the tables and geographies expected to be included in the ACS 5-year products and will be updated periodically with new and expanded information. This information can be accessed at: http://www.census.gov/acs.
Paperwork Reduction Act
Notwithstanding any other provision of law, no person is required to respond to, nor shall a person be subject to a penalty for failure to comply with, a collection of information subject to requirements of the Paperwork Reduction Act (PRA), unless that collection of information displays a current, valid Office of Management and Budget (OMB) control number. In accordance with the PRA, 44 United States Code, Chapter 35, the OMB approved the ACS under OMB Control Number 0607-0810. We will furnish report forms to organizations included in the survey, and additional copies will be available upon written request to the Director, U.S. Census Bureau, Washington, DC 20233-0001.
Dated: September 13, 2010.
Robert M. Groves,
Director, U.S. Census Bureau.
[FR Doc. 2010-23373 Filed 9-17-10; 8:45 am]
BILLING CODE 3510-07-P