2021 ASEC Weights

CPS
1

So I’m looking to update my data for the 2022 CPS and noticing some odd things with weights in prior years I’m hoping someone can clarify.

The person weights (asecwt) in 2021 are 100x larger than in all other years. Is this just an error in accounting for the implied decimal places in that year? I also see similar differences in the household weight (asecwth). It’s off for both 2020 and 2021, estimates of the total number of households in those years are 100x larger than all other years.

I already know about using the asecwtcvd in 2020, and adjusting the weights for the 2014 split sample, but this is a real mystery especially as I hadn’t noticed this issue when working with these files previously. Is it possible that prior year estimates have been modified with the addition of the 2022 ASEC file?

1 Like
2

Thank you for reporting this. It is indeed an error and the IPUMS CPS team is aware of this. They are looking for the source and will correct the public data as soon as possible, most likely by tomorrow. In the meantime you can recover the proper weights by dividing by 100. Sorry for the inconvenience.