This page has information on the multi-phase research project investigating the Medicaid "undercount". The research is designed to explain why discrepancies exist between survey estimates of enrollment in Medicaid and the number of enrollees reported in state and national administrative data.
This study, coined as the "SNACC" project, was conducted by:
- the University of Minnesota's State Health Access Data Assistance Center (SHADAC),
- the National Center for Health Statistics (NCHS),
- the Agency for Healthcare Research and Quality (AHRQ),
- the U.S. Department of Health and Human Services Assistant Secretary for Planning and Evaluation (ASPE),
- the Centers for Medicare and Medicaid Services (CMS), and the
- U.S. Census Bureau.
Phase I: Merging the National Level CMS Databases
This phase involved creating a national database of health-insurance enrollment and evaluate the quality of the information it contains. This was done by merging the CMS Medicaid Statistical Information System (MSIS) files with the CMS Medicare (EDB) files. The quality of the database of health-insurance enrollment was evaluated by assessing the ability to accurately merge the input files and by comparing the characteristics of the individuals in the database to expectations based on Medicaid eligibility rules and characteristics of the U.S. population.
Phase II: Matching the MSIS to the Current Population Survey (CPS)
In this phase data were merged from the Medicaid Analytic Extract (MAX), MSIS, the Current Population Survey (CPS), the Person Characteristic File (PCF), and the Master Address File Auxiliary Reference File (MAFARF). Once matches/non-matches between the MSIS and CPS were determined, the matched records were supplemented with information from the MAX and examine why there were discrepancies between MSIS records of enrollment and CPS reports of Medicaid coverage. The MAFARF and PCF are used to augment the analysis.
Phase III: Matching the State Frame, Household and Person MSIS data to the CPS
In this phase data from the state MSIS files, CPS, 2001 Supplemental Survey (SS01), as well as the MAX and Census Bureau Master Address File (MAF) were used. The Census Bureau worked with CMS to negotiate the acquisition of state MSIS files that enhance the capability of finding matches because they included person names and addresses (in contrast with the national file, which does not include this identifying information). The results of the Phase III analysis were compared to the results from the Phase II analysis to determine if the expected higher match rate (given the additional information for finding matches in Phase III) affects the substantive conclusions of the analysis. Frame coverage was also analyzed to determine if frame differences (i.e., incomplete identification of individuals) affect the substantive conclusions of the analysis.
Phase IV: Matching the MSIS to the National Health Interview Survey (NHIS)
This phase involved re-using the Phase II process, replacing the CPS data with NHIS data. In addition to providing explanations for discrepancies between the national NHIS and MSIS, comparisons to Phase II results allowed the examination of how survey design and implementation affect the quality of the resulting survey data and the estimates derived from it.
Phase V: Comparing MSIS and the CPS ASEC, 2003-2004
Phase V is a repeat of Phase II with later years of the CPS. The intent of this Phase is to see if the discrepancy measured and explained in Phase II changes noticeably in the size or the relationship to explanatory factors from 2000 and 2001 to 2002 and 2003.
Phase VI: Comparing MSIS and MEPS
Phase VI replicates the Phase II process but uses the calendar year 2003 Medical Expenditure Panel Survey (MEPS) data. The intent of Phase VI is to determine how the MEPS estimate of the number of Medicaid enrollees compare to the count from National MSIS as well as to determine factors associated with false-negative response about Medicaid in MEPS.