Background. In 2017, the Census Bureau fielded a Split-Panel test of the Current Population Survey Annual Social and Economic Supplement (CPS ASEC). One panel included the traditional health insurance questions (asked prior to the 2014 CPS ASEC), and the other panel included the redesigned questions (asked in the 2014 CPS ASEC and beyond).
Objective. To estimate differences in the uninsured rates due to previous changes in the health insurance questions.
Data and Methods. There were 100,000 addresses in the 2017 CPS ASEC sample; a subsample of 5,000 addresses were randomly assigned to be eligible to receive the traditional health insurance questions (i.e., pre-2014 CPS ASEC questions). The remaining sample (95,000 addresses) were eligible to receive the set of the redesigned health insurance questions.
Findings. Overall, the uninsured rate of 11.0 percent from the traditional health insurance questions was statistically different from the uninsured rate of 8.8 percent from the redesigned questions.