Medicaid Underreporting in the CPS: Results from a Record Check Study

In 2006 an inter-agency team of researchers was assembled to address the Medicaid undercount issue in the Current Population Survey (CPS). Records on enrollment in 2000-2001 were compiled from the Medicaid Statistical Information System (MSIS) and matched to the CPS survey data covering the same years. Assuming the MSIS data reflected “true” Medicaid status, the matched dataset was used to examine respondent and household characteristics associated with Medicaid underreporting in the survey. Hypotheses about possible causes of underreporting were generated from earlier cognitive testing of the CPS, which suggested that accurate reporting could be associated with the relationship between the household respondent and the other household members for whom he or she was reporting. Testing also indicated that the calendar year reference period could be problematic. Results of the record check study showed support for both hypotheses. Respondents were most accurate when reporting their own Medicaid coverage, and when respondents were asked about the Medicaid status of other household members, they were more likely to report that coverage accurately if they (the respondents themselves) were also covered by Medicaid. With regard to reference period, the more recent the coverage the more likely respondents were to report it accurately; that is, respondents tended to underreport coverage held in the more distant past. Other factors associated with accurate reporting were also explored, and duration of coverage, receipt of Medicaid services and various demographic characteristics proved to be significant.