Is Per weight a sample weight, ie. on a 5 yr ACS est the weights should sum to a fraction of total pop?


I’m working with 2014 ACS-5yr extracts and at a county level (both by PUMS and county FIPS) and the sum of HH and Per weights are between 1/6 and 1/7 of actual population. While I’m not using the dataset for total population estimates I would like to make sure that I’m using this dataset properly. I understand ACS 5-yr sample is only a fraction of HHs, does the weight roll up to sample totals thus explaining the discrepency?

Thank you!



Nervermind, I used select cases, which reduced the headcount.

Apologies for the inconvenience.