Responding to Aaron’s post about changes to the Census’ survey methodology that will make it difficult to measure the impact of the ACA, Genevieve Kenney wrote me to remind us that there are many other good survey sources:
I know you are thinking about what might be lost/gained as the CPS [Current Population Survey] adopts a new approach to measuring health insurance coverage. In considering what we’ll know from the federal surveys and how we’ll be able to assess the role that the ACA is playing in affecting coverage, I think it’s important to step back and consider the full array of federal surveys that will be available to provide information on how insurance coverage and related outcomes is changing over time. While the CPS was one source of information on how the expansion of dependent coverage to young adults under the ACA affected coverage, the NHIS [National Health Interview Survey] and the SIPP [Survey of Income and Program Participation] were also used to addressed that issue. Likewise, the ACS [American Community Survey], not the CPS was used to assess the coverage effects of the early Medicaid expansions under the ACA and the BRFSS [Behavioral Risk Factor Surveillance System] has been used to examine the impacts of a number of prior Medicaid expansions. While I would have preferred that Census introduce the change to the survey instrument using a split sample approach so that there would be a bridge from the past to the new regime, it will be a relief to move away from the messy measurement issues that were inherent in the CPS.
We will have a lot more confidence that we have a handle on how coverage is changing and what is driving those changes if we observe the same patterns across different surveys.
As I put it some email or other, the lights may have dimmed a little, but they have not gone out.