Status:Closed    Asked:Nov 13, 2016 - 08:23 AM

How to use Stata svyset on IPUMS CPS data correctly?

I am using yearly data from IPUMS CPS for an empirical analysis at the individual level. From the FAQs on the IPUMS CPS website I understood that I need to use the sample weights (WTSUPP) in order to obtain correct estimates. My analysis is in Stata, so I intend to use the command "svyset". I am having difficulties in finding the information and understanding from the website of IPUMS CPS what exactly are the following parameters that I need to specify in Stata:

- the primary sampling units

- the strata

- the finite population correction

Is this information available?

Thank you for your help!

Do you have the same question? Follow this Question

Staff Answer


Jeff Bloem


First, a couple notes about the various person-level weight variables available for use in the CPS data: WTFINL is a person-level weight that should be used in analyses of CPS basic monthly data. When analyzing ASEC data, researchers should use the person-level weight WTSUPP. Finally, EARNWT should be used for any analysis including a small number of person-level variables (EARNWEEK, HOURWAGE, PAIDHOUR, and UNION).

Second, STATA has several commands that can be used to weight sample data. Here is a bit of the documentation from STATA about the weight options that can be added onto the end of most STATA commands. Typically, the "pweight" option is what should be used with microdata. However, there is not one single "correct" way to weight data, so be sure to take a look at the STATA documentation.

I hope this helps.


Nov 14, 2016 - 10:18 AM

Report it


negative values of wtsupp 1975 and before
Should you use WTSUPP and the replicate weights for state level analysis usin...
I need to calculate the " STATAXAC " variable (state taxes after deducting ta...
Why are not all the "not in the labour force" categories filled for the EMPST...
When was contingent worker supplement data (1995 and 2005) added to IPUMS-CPS...
Login   |   Register

Share |