Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1.1 Overview of the American Community Survey Census data have long played a central role in transportation planning and analyses. In partic- ular, transportation planners are heavy users of data products concerning population and house- hold characteristics that are derived from the decennial census Long Form, such as the following: â¢ Summary File data (SF3 and SF4), tables, and profiles; â¢ Public Use Microdata Sample (PUMS) data (1 percent and 5 percent samples); and â¢ Census Transportation Planning Package (CTPP) tables (Parts 1, 2, and 3). Beginning with this decade, the Census Bureauâs American Community Survey (ACS) will replace the decennial census Long Form as the preeminent source of U.S. population and household characteristics. In previous decennial censuses, residents of one out of every six addresses were asked to complete the Long Form of the census questionnaire, which gathered demographic, socioeconomic, and housing characteristics data, in addition to the base census questionnaire (or census âShort Formâ), which gathered constitutionally mandated population counts, along with age, race, and gender information. Beginning in 2010, the decennial census will focus solely on collecting the basic census counts by age, race, and gender. The Census Bureau is now collecting additional population and household characteristics data through the rolling sample ACS. Each year, about 3 million of the U.S. addresses, 36,000 Puerto Rico addresses, and residents of 2.5 percent of group quarters facilities will partici- pate in the ACS, providing data that are more up to date, timely, and probably more accurate than the decennial census Long Form data. The primary cost to the data user is a reduction in sample size and the corresponding need to accumulate data over time and/or across geography. The Census Bureauâs stated goals for the ACS are to â¢ Provide federal, state, and local governments with an information base for the administration and evaluation of government programs; â¢ Facilitate improvement of the 2010 Census by allowing the decennial census to focus on counting the population; and â¢ Provide data users with timely demographic, housing, social, and economic statistics updated every year that can be compared across states, communities, and population groups. The Census Bureau began developing the ACS in the mid 1990s after many years of research indicated the potential value of a âcontinuous measurementâ data collection program. During the initial ACS test period between 1996 and 1998, while the program was still somewhat exper- imental, preliminary ACS data were collected for a few test sites. In 1999, the number of test sites was increased to 31 locations, comprising 36 counties and representing a broad range of communities that were selected to provide different combinations 1 C H A P T E R 1 Introduction
of population sizes, population characteristics, population growth levels, and difficulty of enumeration. The data collection effort for the 31 test sites has been performed annually since 1999. In addition to the test site program, the Census Bureau performed a large-scale (1,203 counties) operational test of ACS methods in the year 2000, entitled the Census 2000 Supple- mentary Survey (C2SS). In 2002, 2003, and 2004, the data for the 31 test sites were supplemented with data collected for the 1,203 counties in the C2SS. The Census Bureau began full implementation of ACS for all housing units in 2005. Begin- ning in 2006, the Census Bureau will start collecting group quarters data, as well as housing unit data. Thus, from 2006 onward, ACS should provide comparable coverage to the decennial census Long Form data collection. 1.2 Some Important Implications of ACS for Data Users The discontinuation of the decennial census Long Form and the implementation of ACS will significantly affect how transportation planners access, use, and interpret data on population and household characteristics. Later sections of this guidebook will discuss the details of ACS implementation and how ACS will affect transportation data usersâ analyses. Some of the more important implications of ACS for transportation data users are summarized here. These, and other ACS issues, are developed further later in the guidebook. 1.2.1 Frequency of Data Releases The primary benefit of ACS is that the data are being collected and will be disseminated more frequently than the once-in-10-years decennial census Long Form data. Data users will no longer need to rely on aging âsnapshotâ estimates of population and housing characteristics. Instead, they will be able to use more recently collected data whose accuracy and relevance will not depend on how closely the analysis year conforms to the decennial census year. In addition, the increased frequency of data releases will enable data users to analyze trends over shorter time periods. 1.2.2 Differences between Census 2000 and ACS Unfortunately, the ACSâs differences with the previous census Long Form, in terms of data collection procedures and questions, will make it more difficult for users to compare ACS results with previous census estimates and to understand longer-term trends in demographic, socioe- conomic, and economic characteristics. Determining whether differences between ACS estimates and year 2000 Census estimates reflect actual differences in the populations of interest will require analysts to understand the survey differences and to be able to perform significance tests on sample data. 1.2.3 Reduced Sample Size and the Need for Data Accumulation For the next few years, the ACS sample size will be equivalent to slightly less than 1-in-40 addresses. The decennial census Long Form was sent to about 1-in-6 addresses, with some areas with slightly higher rates and some with slightly lower rates. Because of the reduced sample size of ACS, it is not possible to replicate the decennial census Long Form data on an annual basis. For most analyses, data users will need to rely on estimates derived from accumulating ACS data across years. ACS annual estimates will be released for areas 2 A Guidebook for Using American Community Survey Data for Transportation Planning
with population greater than 65,000, starting with 2005 data that will be released in the fall of 2006. So, for these larger population areas, data users will get new independent ACS estimates for the previous year. Three-year accumulated average estimates will be released for areas with population greater than 20,000 starting in year 2008. In that year, the Census Bureau will release estimates for those areas that are based on averaging the ACS data from 2005, 2006, and 2007. Five-year accumulated averages will be released for all areas starting in year 2010, using accu- mulated averages for the years 2005, 2006, 2007, 2008, and 2009. The three- and five-year accumulated averages will be developed every year using the preceding three or five years of ACS data, but analysts will need to be aware that multiyear estimates reported for a specific year are not independent of previous multiyear estimates that have overlapping years. In addition, data users will need to understand the potential impacts of using characteristics data accumulated over time when those characteristics are changing year to year. 1.2.4 Understanding and Reporting Sample Data Both the decennial census Long Form and ACS are samples of the overall population. There- fore, estimates from both data sources contain uncertainty. Despite the fact that the Census Bureau provides variance estimates for Long Form data, in most cases when using census Long Form data, analysts take the reported estimates as simple points and ignore the level of variance present. However, for the ACS, with its lower sample sizes, the Census Bureau is instructing and enabling users to account for the inherent sampling error in their analyses. The Census Bureau ACS data tables include 90 percent confidence intervals for all estimates so that users can readily see the relative level of uncertainty in the estimates. Data users will need to determine how the higher uncertainty levels affect their analyses, and they will need to develop effective ways of presenting information with uncertainty. 1.2.5 Data Disclosure Avoidance Before releasing any ACS data, the Census Bureau first determines whether the informa- tion could be used to identify specific households, individuals, or establishments. When the information is deemed to potentially result in wrongful disclosure, Census Bureau staff are required by law to take actions to prevent such identifications. Three types of data disclosure avoidance procedures will be applied to the ACS data: imputation, rounding, and data suppression. Disclosure avoidance is an important issue for transportation planning uses of Census data, because transportation data users rely on small-area data to a greater extent than almost all other data users. Analyses of journey-to-work flow data will be particularly affected, because home- to-work matrices, even for mid-sized geographic units, will generally consist of small numbers within individual cells, and thus will be subject to data disclosure avoidance. 1.3 Purpose and Organization of this Guidebook In this guidebook, attempts have been made to identify the key issues that will face transporta- tion planners as they use ACS data to complete analyses that have been heretofore performed with decennial census Long Form data. Potential new transportation planning analyses that ACS may enable also are outlined. Section 2 describes the implementation of ACS, including the operational steps that the Cen- sus Bureau follows to collect and disseminate ACS estimates. This section tries to highlight the Introduction 3
differences between ACS and the traditional census Long Form data collection, as these differ- ences will have direct impacts on data users, such as transportation planners. Section 3 describes the ACS data products and how users can access these products and other information related to ACS through the Census Bureauâs American FactFinder website. This sec- tion provides lists of available tables and examples of the different types of ACS tables that are available to users. Section 4 includes several sections that summarize the special challenges of using ACS data. This section begins by summarizing research on ACS data quality. The first subsection attempts to inform data users about the accuracy and potential biases in the ACS estimates. Next addressed are two aspects of ACS that may be new to traditional Census data usersâthe need for the accumulation of data across geographic areas and over time and the need to con- sider the effects of data disclosure limitations in designing analyses. The ways in which these issues should affect data usersâ choices about data analysis strategies are demonstrated. Section 4 then turns to less strategic, and more hands-on, ACS data use issues. Described is the importance of measuring, understanding, and reporting the precision levels in ACS esti- mates. There is discussion of the comparison of ACS estimates to Census 2000 results, and finally, discussion of effective ways that analysts can take advantage of having more frequent estimates of census variables. Sections 5 through 9 provide case study examples of potential transportation planning analyses using ACS data. Section 5 discusses the most common uses of census dataâdescriptive analyses and policy planning analyses. Section 6 describes the application of ACS to trend analyses. With ACSâs annual data releases, the opportunities for performing more interesting trend analyses increase. Section 7 provides case studies on transportation market analyses, including environmental justice analyses. Section 8 discusses the use of ACS in the design and analysis of transportation surveys, particularly household travel surveys. Finally, Section 9 describes how ACS data can be used in place of decennial census data in travel demand modeling analyses. 1.4 Additional Information Sources for an Introduction to ACS The Census Bureau and others have developed several documents that provide an introduc- tion to ACS. The Census Bureau American Community Survey website (www.census.gov/acs/www/index. html) includes several introductory documents in Portable Document Format (pdf) in the Survey Basics section, including â¢ American Community Survey: A Handbook for State and Local Officials (issued December 2004); â¢ Congressional Toolkit (issued Spring 2004), which includes several documents that âexplain how and why the survey is conducted, its benefits, and how to obtain additional information.â The documents, which are provided on CD-ROM to congressional staff, include â ACS Tool Kit, an introductory summary of ACS; â ACS Housing Fact Sheets, a summary of housing information collected in the ACS and justification for why this information is collected; â ACS Population Fact Sheets, a summary of population information collected in the ACS and justification for why this information is collected; 4 A Guidebook for Using American Community Survey Data for Transportation Planning
â ACS Questionnaire, a copy of the mailed questionnaire; â ACS Instruction Guide, a copy of the instructions that accompany the questionnaire; and â Questions and Answers, an introduction to ACS presented in a question-and-answer format. â¢ ACS News Media Toolkit (updated October 2005), which includes a number of documents that summarize ACS for media data users. In addition to the Census Bureau overviews, other researchers and data users have assembled many good descriptions of ACS that are suitable for interested new potential users. We recom- mend the Population Reference Bureauâs September 2005 issue of its Population Bulletin, which can be found at www.prb.org/pdf05/60.3The_American_Community.pdf. Transportation users of ACS can find summaries of many relevant ACS issues in FHWAâs Census Transportation Planning Package (CTPP) status reports, which can be found at www.fhwa.dot.gov/ctpp/status.htm. Finally, motivated prospective ACS users could benefit greatly from the on-line course on ACS offered by Statistics.com ($399 as of January 2006) and can be found at www.statistics.com/ content/courses/census/index.html. Introduction 5