Transportation planners and modelers need journey-to-work (JTW) flow data for a variety of purposes. Historically, this data was available through a special tabulation of the Census Long Form data produced at a Traffic Analysis Zone (TAZ) level. However, since 2005 with the advent of the ACS, the quality and level of detail of the JTW flow data is diminished. Because the ACS is a continuous survey with a small sample, the margins of error are higher than the decennial long form. Therefore, alternative sources and methods to combine ACS with alternative data sources are needed to enhance its usability. To assist in this challenge, the Census Bureau has a new dataset called the Longitudinal Employment Household Dynamics (LEHD). The LEHD merges the Bureau of Labor Statistics Quarterly Census of Employment and Wages (QCEW) with other federal data sets to produce a synthetic JTW file at a block level.
The purpose of this research is to examine the data and methods for merging the ACS and LEHD data. To accomplish any merger a first step is to examine the home and work combinations in both the ACS and LEHD at an aggregate level to see how they compare. The next step is to identify the individual records from ACS that are in LEHD and see what, if any, differences there are in the individual residential and workplace addresses. It is expected that not all records will match and the reasons for any mismatch will need to be documented. For example, why do differences in the home and work locations occur? Can the source of the difference be identified? Once these differences are understood and documented methods can be developed for successfully integrating the two data sets.
This research serves as the foundation for making the ACS a more robust, accurate and useable data set for transportation planning and analysis. The work will need to be conducted at the U.S. Census Bureau Suitland Office, with researchers who have “special sworn status” to access the confidential micro data. Report documenting what was found in the micro level analysis of the home and work locations. Recommendations on what, if any, “fixes” need to occur followed by recommendations on how best to merge the two data sets