Inaccurate Age and Sex Data in the Census PUMS Files

In April 2009, the U.S. Census Bureau acknowledged problems in the techniques it uses to prevent the identification of specific individuals in public-use microdata sample (PUMS) files. These techniques produced inconsistent sex ratios for people ages 65 and older in the PUMSs produced from the 2000 Census, the 2003-2006 ACS, and the 2005-2006 PRCS. These errors carried through to the 2005 and 2006 cases in the 2005-2007 3-year ACS/PRCS.

In December 2009, the Census Bureau released corrected age data (see Erratum #50 at this link) for 2006. The IPUMS versions of the 2006 ACS/PRCS now contain the correct data in the AGE variable, and the original (incorrect) data are contained in the AGEORIG06 variable.  Almost all AGE values 65 and over have changed, suggesting that the revised data were imputed or calculated with synthetic data methods; QAGE has been set to 4 (allocated) for all persons age 65 and up. 

The Census Bureau has not announced plans to release any more corrected data files. Researchers interested in Census plans to re-release these data are encouraged to contact the Census Bureau's ACS Office, at acso.users.support@census.gov.

For a full discussion of the problem and its implications for researchers, see:

Alexander, J. Trent, Michael Davern, and Betsey Stevenson. 2010. "Inaccurate Age and Sex Data in the Census PUMS Files: Evidence and Implications." National Bureau of Economic Research Working Paper No. 15703.