Current Population Survey (CPS) Matching

The programs below are written to match consecutive March CPS surveys. They can be easily modified to match other CPS surveys where matching is possible. A list of variables for matching CPS basic and CPS March files is also available. Each recent CPS March Annual Demographic File includes a section on matching across years. The Stata .do files accompany

Madrian, B. and L. J. Lefgren (1999). "A Note on Longitudinally Matching Current Population Survey Respondents." NBER Working Paper T0247, which is available at https://www.nber.org/papers/t0247 and as "An Approach to Longitudinally Matching Current Population Survey (CPS) Respondents." Journal of Economic and Social Measurement, 2000, 26: 31-62.

NBER internal users can obtain the programs at /home/data/cps/match.

Programs 

• readme.txt - Describes the files below
• matchcps.do - The master program. The other files are all called within this program. This program is written to match consecutive March CPS surveys but can be easily modified to match other CPS surveys where matching is possible.
• educate.do and educate2.do - redefine the education variables in the CPS which are not consistently coded over time for the time t and time t+1 data to be used in the CPS merge.
• married.do and married2.do - Redefine the marital status variables in the CPS for the time t and time t+1 data to be used in the CPS merge. 
• race.do - Recodes the race variable for the CPS merge.
• multobs.do - Deals with the fact that some individuals do not actually have a unique identifier at a point in time in the CPS. The working paper does not have much detail on this problem. This is discussed further in readme.txt.
• year.do - Creates a variable with the year