CASE STUDIES IN THE APPLICATION OF ADJUSTED CENSUS DATA FOR PLANNING PROJECTS

This paper details the reasons for differences between locally collected data and 1990 Census data as determined from a detailed analysis of model development efforts in two planning studies. Agencies around the country are beginning to use Census data that has been adjusted based on a newly released Federal Highway Administration publication. A number of issues persist on why it is so difficult to match locally collected data and Census data. A recently completed publication, Transportation Planner's Handbook on Conversion Factors for Use of Census Data, has been published to assist planners in using the 1990 Census to develop and calibrate local travel demand models. Collecting new data to complete the development of a local model is not always an option. The 1990 Census provides another source of information to assist in traffic model estimation. Potential users of the Census need to be aware that there would appear to be a variance between results obtained from the Census journey-to-work files and locally developed home interview surveys, even after the use of the Census adjustment factors. Recently completed projects in Hampton Roads, Virginia and Atlanta, Georgia involved detailed traffic model development and calibration, in conjunction with factor adjusted Census data. Because of the intimate understanding of the data for the study area, and the development of the model sets from the beginning, differences between the locally collected data and Census were explainable. This paper details possible problems that can arise when comparing the data as they relate to geography, data definition and accuracy of the data collection process.