August 2007

New Medicaid Undercount Research! Sponsored by the Robert Wood Johnson Foundation and the Offices of the Assistant Secretary for Planning and Evaluation, a group of inter-agency researchers, collectively referred to as "SNACC," have proposed an "imputational model" that can be used for adjusting for the underreporting of Medicaid enrollment.

The document Medicaid Under-reporting in the Current Population Survey and One Approach for a Partial Correction describes the project, the logistic regression models, and how to use the coefficients to derive imputed values for CPS respondents and thus a new estimate of Medicaid coverage.

To expedite this process for researchers, next year we plan to release variables that provide IPUMS-CPS users instant access to the imputed Medicaid data.  At the time of this release, there will be an announcement on our home page, as well as a detailed note on our Revision History page.

SNACC Researchers are from: the State Health Access Data Assistance Center, the National Center for Health Statistics, the Office of the Assistant Secretary for Planning and Evaluation, the Center for Medicare and Medicaid Services, and the US Census Bureau.

As always, your feedback is welcome. Please send any comments or questions to ipums@pop.umn.edu.