182 – Estimation Issues

Study of Error in Survey Reports of Move Month Using the U.S. Postal Service Change of Address Records

at U.S. Census Bureau

Correctly recalling where someone lived as of a particular date is critical to the accuracy of the once-a-decade U.S. Decennial Census. In the 2010 Census, all persons living in the U.S. were counted at the place they were living or staying as of Census Day, April 1, 2010. The data collection period for that census occurred over the course of a few months: February to August, with some evaluation operations occurring up to 11 months after Census Day in April. The assumption was that respondents could accurately remember moves and move dates on and around April 1st up to 11 months afterwards. Our research uses statistical models to investigate the validity of this assumption by comparing reports of move months in a U.S. Census Bureau survey with an administrative records database from the U.S. Postal Service containing requests to forward mail filed in March and April of 2010. We found some evidence that the length of time since the move affects memory error in reports of a move and the month of a move. Respondents were less likely to report a move when responding to a survey 10 to 11 months later than when responding to an identical survey either 2 to 3 months or 5 to 6 months later. However, the error in reporting a move did not differ when responding to a survey 5 to 6 months later compared to responding 2 to 3 months later. For movers, the analysis of the discrepancy between the reported move month and the NCOA record showed the length of time since the move had a similar effect on the error in the reports.