REPWTP, PERWT04, EARNWT, WTSUPP which should I use with CPS?

I am using CPS data from 1990 to 2015 with the ASEC Supplement. I am unsure of which weights to use or if I should use all four weights I different times?

First I run the command

svyset [iw=wtsupp], sdrweight(repwtp1-repwtp160) vce(sdr)

do I then include both the EARNWT AND WTSUPP and PERWT04 weights in my analysis that follows?

Also, should I create a group variable with PRENUM, SERIAL and YEAR?

I am using the hourwage variable from the ASEC supplement for individual level analysis in Stata 13.

Thank you.


Since you are interested in the HOURWAGE variable, which is part of the Outgoing Rotation Group/Earner Study questionnaire, you will need to use the EARNWT weight variable in place of WTSUPP in your svyset command. You can then use the svy: commands to generate estimates. IPUMS-CPS also has more information about available weights and replicate weights that may be helpful.

I hope this helps.