Status:Closed    Asked:Jun 05, 2015 - 03:19 PM

Thanks for the help

Dear Joe,

THanks a lot for the article. This is exactly what i was needing. I have actually decided to do the match itself based on the household characteristics and demographics, that should otherwise not change, to identify the same person in the basic march CPS and ASEC.

Under the principle that they should be the same samples, I thought was the most likely aproach.

For the 1994, 2003 and 2012 March CPS data, I managed to match about 130K observations with their "twins" in the ASEC. I can share you the code, which might help regarding the creation of the link id for the IPUMS. my email friosa@gmail.

Bellow is the table of matched (right) and un match (left) observations

| match and Survey year

Flag for | ----------- 0 ----------- ----------- 1 -----------

ASEC | 1994 2003 2012 1994 2003 2012

---------- --+-------------------------------------------------------

asec | 11,325 82,437 70,667 139,618 133,987 130,731

March Basic | 1,007 7,301 641 139,618 133,987 130,731


Thanks again for the article. Fernando

Do you have the same question? Follow this Question

Staff Answer




I am glad you found the article helpful. You may want to pay special attention to the tables at the end of the file, particularly for 1994, as they highlight special cases identified during the MARBASICID generation process.


Jun 05, 2015 - 04:11 PM

Report it


weights for partial sample in March CPS 2014
How to link original ASEC family relationship variable with RELATE variable o...
Before the last change people born in Mexico in 2014 were 11,458,133 and 23,1...
Change in sample size in 94 and 95 when linking individuals with non-missing ...
Is there a reason that CPSIDP equals 0 for various months in 1978 through 198...
Login   |   Register

Recently Active Members

View More »

Share |