Accounting for multiple responses with over 5 years of PUMS files

From the ACS design and methodology report (page 45 of the PDF):

"One of the ACS design requirements is that no HU address be in sample more than once in any five-year period."

www2.census.gov/.../acs_design_methodology_report_2014.pdf

I just reread your question and missed the "over 5 years of data" part.

The way households are sampled, a HU could be sampled again 5 years later (but not 6 or 7 because of rotating sub-frames).  But there is no way of identifying if two observations 5 years apart came from the same HU.

I just reread your question and missed the "over 5 years of data" part.

The way households are sampled, a HU could be sampled again 5 years later (but not 6 or 7 because of rotating sub-frames).  But there is no way of identifying if two observations 5 years apart came from the same HU.