ATUS variable CPS Weekly Earnings Detail

TIME USE
1

Hi, I have a question regarding the variable EARNWEEK_CPS8. I am doing research which uses this variable, but I have recently realized that in the CPS, non-responses for weekly earnings are imputed by the BLS. For the study of interest, the way that the BLS imputes these values is biased (the study includes details on the union earnings premium. The BLS imputes weekly earnings without consideration for union status).

I want to know if there is any way to distinguish between those earnings which are imputed, and those which are not. That, or if there is a way to determine if a person did not report earnings. Thank you so much!

2

The two data quality flag variables Q1EARNWEEK_CPS8 and Q2EARNWEEK_CPS8 flag observations with allocated or altered values of EARNWEEK_CPS8. Q2EARNWEEK_CPS8 identifies responses that were top-coded, and Q1EARNWEEK_CPS8 identifies responses that were imputed. You can use Q1EARNWEEK_CPS8 to identify individuals whose EARNWEEK_CPS8 values were imputed in the original data.

In general, you can find data quality flags for a variable by clicking the variable, and then clicking the “flags” tab. You can also add data quality flags to your extract when you “check out” your data cart:

unnamed
unnamed1070×814 72 KB