Can you tell me which IPUMS variables and values I can use to identify persons who were coded with PEMLR = 3 in the CPS? My apologies if this has already been asked and answered. I found some discussions about variables that are similar, but was unable to figure out how I could exactly reproduce the CPS assignment. Right now, I’ve selected the following variables: EMPSTAT, ABSENT, WHYUNEMP, WHYABSNT, UHLAY6M_B1, UHLAYAVL_B1, and UHLAYOFF_B1. WHYUNEMP = 1 looks like it should be the same as PEMLR = 3, but I haven’t seen any confirmation of that.
The unharmonized variables UH_MLR_B2 and UH_MLR_B3 correspond to the variable PEMLR for 1989-1993, and 1994-forward, respectively. I would suggest using these variables to identify persons with a specific PEMLR code.
That being said, a crosstab between these unharmonized variables and the harmonized IPUMS CPS variable WHYUNEMP show a 1:1 relationship between values of “on layoff” in the unharmonized variables and a value of 1 for WHYUNEMP (crosstab below), confirming your proposed identification strategy.