An error in data processing led to approximately 10 percent of cases being incorrectly set out of universe for Tax Return content. The correct universe for EFILING – the highest level Tax Return variable – should be all cases who were aged 15 or over at some point during the reference period. This universe inconsistency also influences downstream variables: EWILLFILE, EFSTATUS, EDEPCLM, and EEITC.

In all cases, the incorrect records would have been imputed with a hotdeck (or transferred from a spouse who had their value imputed from a hotdeck) if the error had not existed. Data users can impute missing data for those who are in universe using their preferred imputation method but should analyze these data with caution.