Proposed NCHRP Study: Second Stage Problem Statement

I. PROBLEM NUMBER

2002-B-06/B-15

II. PROBLEM TITLE

Using American Community Survey Data for Transportation Planning

III. RESEARCH PROBLEM STATEMENT

Transportation planners have relied heavily on the decennial Census "long form" data, because it provides detailed demographic characteristics along with journey-to-work data for small units of geography such as census tracts or Traffic Analysis Zones.  The 2000 Census "long form" was probably the last time the "long form" will be included in the decennial Census, because of Congressional concerns about privacy and burden to the American public.  It is the "long form" that provides the data for the Census Transportation Planning Package (CTPP), the mostly widely used database for transportation planning.

The U.S. Census Bureau plans to replace the "long form" with a continuous data collection program called the American Community Survey (ACS). The ACS differs from the decennial census in many ways, especially as it represents a change from data collected at a single point-in-time (April 1, 2000) to data collected continuously throughout the year, and summarized annually for large geographic units.  Data for Traffic Analysis Zones or tracts would become available based upon a floating average of data accumulated over 5 to 7 years.   The transportation planning community needs to know how to use this new source of data in applications such as long range planning and forecasting, environmental justice analysis, specific project analysis and descriptive interpretation.

IV. RESEARCH OBJECTIVE

This research will compare results from the American Community Survey test sites (1999-2001) and explain how these results differ from decennial Census long form data.  The research will provide methods of incorporating these differences into existing transportation planning applications, such as travel demand forecasting, sketch planning and microsimulation, as well as methods for presenting this data for decision makers, the public, and the media.

1. Compare data from the Census 2000 long form to the data from the ACS test (31 test sites).   All 31 test sites are not required to be evaluated, however the research will include sites with different population characteristics, e.g. seasonal population shifts, degree of urban development and transit accessibility, racial/ethnic diversity.   The comparisons will include journey to work characteristics, geographic flow between home and work, as well as household characteristics.  Geographic comparisons will be made for counties, places, census tracts, and block groups/Traffic Analysis Zones.  Seasonality, "moving averages" from accumulations over time, differences in response rates, and sample weighting, are some of the issues to be addressed.  Standard errors based on the different surveys (decennial census long form and ACS) will be calculated for related variables and the geographic units listed above.  The research will include the calculation of point-in-time estimates versus moving averages.  A report on the results will be included in this task.
2. Develop recommendations for a Census Transportation Planning Package (CTPP) and a schedule of data release, based on the ACS data.   Develop recommendations for integrating the CTPP into the standard ACS data dissemination system.
3. Prepare guidance manuals for statistical analysis, training course for MPO and State DOT staffs, detailed case studies of ACS comparison site data, and guidance materials for presenting continuous census data to decision-makers, the public and the media.

V. ESTIMATE OF PROBLEM FUNDING AND RESEARCH PERIOD

The estimated funding for this project is $300,000.  The research will require approximately 24 months to complete.

VI. URGENCY, PAYOFF POTENTIAL AND IMPLEMENTATION

This is a high priority issue submitted by three states and FHWA. It is a time-sensitive issue because the ACS is currently in a 3-year testing period, with full implementation scheduled to begin in 2003.

Anticipated products from this research will be used directly by planners at Metropolitan Planning Organizations and State Departments of Transportation.

VII. PERSONS DEVELOPING THE PROBLEM

VIII. PROBLEM MONITOR

Chuck Purvis  (Chair A1D08)
Metropolitan Transportation Commission
101 Eigth Street
Oakland, CA  94607
510-464-7731

IX. DATE AND SUBMITTED BY

November 29, 2000

Ed Christopher
Bureau of Transportation Statistics
400 7th St SW
Washington, DC 20590
202-366-0412

In conjunction with support from the following Transportation Research Board Committees: