# Time-Invariant variables? **URL:** \ **Category:** USA\ **Created:** [July 5, 2022, 9:20am UTC](https://forum.ipums.org/t/time-invariant-variables/4762 "2022-07-05T09:20:36Z")\ **Posts on this page:** 1\ **Showing post:** 5
**Author:** ![Alex](https://avatars.discourse-cdn.com/v4/letter/a/db5fbb/32.png) [@Alex](https://forum.ipums.org/u/Alex)\ **Post date:** [July 13, 2022, 3:39pm UTC](https://forum.ipums.org/t/time-invariant-variables/4762/5 "2022-07-13T15:39:28Z")
Thank you, I misleadingly thought that unique identifiers created from _serial_, _person_ and _sample_ could represent representative individuals across years. I add this related link for anyone falling in the same trap. > [@Could I construct panel data using multi-year ACS data](https://forum.ipums.org/t/could-i-construct-panel-data-using-multi-year-acs-data/4687): > > Hi, there, As far as I know, ACS surveyed different people in different years, meaning that it is not a panel design. However, in the ACS data I downloaded from IPUMS, it has variables like household serial number and personal serial number, which can be combined to uniquely identify individuals (I suppose). Can I use this approach to link individuals in multi-year ACS data to construct panel data? Or, simply put, is it legit or possible to contract individual-level panel data using multi-year … --- _[View the full topic](https://forum.ipums.org/t/time-invariant-variables/4762)._