What is the universe of EMPSAME? (and some other details)

CPS
1

I have a sort of twofold question about the EMPSAME variable and the information provided in the “Description” and “Universe” sections in the codebook (here: https://cps.ipums.org/cps-action/vari…).

Under “Description,” the codebook says, “EMPSAME indicates whether or not the respondent was employed by the same employer and the same job he/she reported working as his/her main job in the previous month’s survey.”

Under “Universe,” the codebook says, “Employed persons not in month-in-sample 1 or 5, and were employed last month.”

However, I am able to find in the data respondents who answered “yes”/“no”, etc, to this question in their month-in-samples 1 and 5. Is this a mistake in the data? Otherwise, should the universe actually be all individuals who were employed in the last month?

And if the latter is true, should we interpret the question in month in sample 5 as the same employer, relative to the previous month, or the same employer, relative to month-in-sample 4?

Thank you! This is important to me because I am trying to do a study where I take individuals who participated in the ASEC twice, so that I have two points of income data for them. I would then like to find whether they have changed employers between the two points, so any information on this detail would be helpful.

2

I am able to replicate your observation, however I only find about 5 such cases over the past decade of basic monthly samples. Additionally, this seems to be an input data quality issue, rather than an IPUMS coding issue. So, there is nothing much we can do to “fix” these cases. If these cases are concerning to you, feel free to simply drop them from your analysis. The universe for EMPSAME remains “Employed persons not in month-in-sample 1 or 5, and were employed last month.” The reason for the restriction based on MISH is due to the fact that for EMPSAME to hold any meaning we need to have information from the “last month”. Additionally, since it sounds like you are aiming to analyze ASEC samples, note that none of the 5 cases that I observe occur within ASEC samples.