Is Fixed Effects possible for ACS data?

Hi, I am currently using the data from 2012 to 2022 to estimate the wage returns of an individual employed in a green job.

I would like to do a fixed time, industry, and PUMA (location) fixed effect; however, due to the nature of ACS data being cross-sectional data, I am unsure if I could do fixed effect estimations on the dataset.

Can anyone please advise? Thank you.

IPUMS USA data are microdata. Each observation (row) is a person or household. Each observation has information on year (the year their survey was taken), PUMA (where they live), industry (if applicable), and other characteristics. These variables on each observation make up the columns of the dataset. You can apply fixed effects to a regression using microdata by adding the fixed effect variables as binary independent variables. For example, adding year fixed effects means controlling for a binary variable for each year in the dataset. Adding industry fixed effects means controlling for a binary variable for each industry. For industry fixed effects, you should use a harmonized industry variable (IND1950 or IND1990) so that the codes are consistent between samples. For PUMA fixed effects, be sure to pair PUMA with state, as PUMAs are state dependent. Also note that the PUMA definitions used for the 2022 ACS differ from the definitions used for the other samples you are interested in, so you will need to include an additional interaction for 2022 PUMAs.

There is no reason why fixed effects would not be feasible using IPUMS USA data from the U.S. census or American Community Survey. Fixed effects can be used with panel data and cross-sectional data. Determining whether they are appropriate for you to use in your specific analysis is beyond the scope of IPUMS User Support–we can answer questions about IPUMS data and documentation, but do not provide analytical advice. I recommend reviewing the literature in your field and consulting with mentors, coauthors, and colleagues to make this determination.

Thanks Isabel for your insights, that is very helpful.