What is the difference between 'pwgtp' and 'perwt'? Did the name of the person weight change '12 data?

I am just curious about the change to the person weight in the 2012 PUMS data (3 year). I can’t find the old ‘pwgtp’ weight in the extract I just downloaded, but ‘perwt’ seems to be the correct person weight to use. (However, the ACS codebook still lists the population weight variable as ‘pwgtp’, as seen here: https://www.census.gov/acs/www/Downloads/data_documentation/pums/DataDict/PUMS_Data_Dictionary_2010-2012.pdf.

Can anyone confirm that this is the same weight and the correct one to use for population estimates?



PERWT is actually the IPUMS-USA version of the Census Bureau variable ‘pwgtp’. IPUMS-USA copies the values directly from ‘pwgtp’ to create PERWT so the two variables are identical.

I hope this helps.

Hi Joe,

Thanks so much, this helps alot. I just have one confusing thing that I’d like cleared up before I feel 100% relieved…

Can you tell me why my 2011-5year ACS dataset has ‘pwgtp’, but my 2012 3-year ACS dataset has ‘perwt’? I downloaded both datasets from iPUMS, but the 2011 data was pulled about a year ago (maybe a little more) and I just downloaded the 2012 data today.

Thanks again!!


As far as I am aware, IPUMS-USA has always used PERWT as the Person Weight variable. I am not sure how pwgtp would have ended up in your extract from IPUMS-USA. Perhaps you renamed the variable to make the dataset similar to the original PUMS files? If you are unsure about the accuracy of your 2011 5-year data file, you can use the “resubmit” link that corresponds to that particular extract on your Download or Revise Extracts page. This will request the same variables and samples from our system that you requested last year and the resulting dataset will represent our current data holdings. Because some variables do change over time you may want to check the Errata and Revisions page to see if any revisions have effected your data since your initial download.

I hope this helps.