In the 2008 (waves 1-5), 2004 (waves 1-12), and 2001 (waves 1-12) SIPP panels, the mover recode variable (TMOVRFLG) is incorrectly coded for some respondents. This incorrect coding manifests itself in one of two ways. First, for some respondents, the mover recode incorrectly indicates that the respondent is new to the sample, even though the person existed in at least one previous survey wave. We estimate that this problem affects approximately 2.5 percent of respondents. For other respondents, the value of the address ID (SHHADID) changes, indicating that the respondent has moved. However, even though the address ID has changed, the mover recode incorrectly indicates that the respondent did not move. We estimate that this problem affects somewhere between 2 percent and 2.5 percent of respondents. Thus, between the two problems, approximately five percent of respondents have incorrect values for their mover recode.

The mover recode should be corrected in future releases of the 2008 panel data (waves 6+) with the exception of situations in which the address ID changes between waves and the mover recode incorrectly indicates the respondent did not move. These instances will not be corrected because the Bureau does not currently longitudinally edit SIPP. The mover recode, minus the longitudinal problem, will also be fixed in the previously released panels, starting with the 2008 panel and moving backwards to the 2004 and 2001 panels. We do not currently have an estimated release date for the fix. In the meantime, you can use the address ID (SHHADID) to identify whether a respondent moved. However, you will not be able to determine any further information about this move, such as whether it is intercounty, intracounty, etc. Should you want more technical details concerning this error, or if you have additional questions, please contact the Journey-to-Work and Migration Statistics Branch at (301) 763-2454.