Weights in Regressions

Dear All,

I would like to learn about what predicts the duration of unemployment (time it takes people who got laid off to find a new job), using monthly CPS data.

To that end, in a nutshell, I am planning to regress the duration of unemployment (DURUNEMP) on variables like AGE, SEX, OCC, IND, and especially EARNWEEK (and other variables from the Earner Study/Outgoing Rotation Groups).

Could someone please point me toward the correct weight I should use in such a regression exercise?

Thank you very much!

Best wishes,
Chantal

Cross-sectional analyses using any outgoing rotation group (ORG) variables should use EARNWT even when other non-ORG variables are included. Researchers should use the weights that correspond to the strictest sampling universe in their analysis. Since ORG variables are drawn from a universe of respondents in month-in-sample 4 and 8, they are more restrictive than the universe of all basic monthly respondents or that of any monthly supplement. Since DURUNEMP is available in all basic monthly surveys, you can run this analysis without needing to link respondents across their time in the panel. If you are not using the data cross-sectionally and are leveraging the panel aspect of the CPS, you will instead need to use longitudinal weights since this represents a further restriction on your sampling universe.

1 Like