,
9 tweets,
2 min read
Read on Twitter
A quick thread on yesterday’s Gallup data: I would caution against concluding that there has already been a large increase in the uninsured rate under the Trump Administration on the basis of these data.
Notably, Gallup’s estimate of the uninsured rate has been drifting upward relative to estimates from the ACS and NHIS, the two most comparable federal surveys, for about 10 years now. 
In 2008, Gallup’s estimate of the uninsured rate was about 2 p.p. lower than the corresponding estimates in the ACS and the NHIS. By 2017, Gallup’s estimate was approaching 2 p.p. higher.
Who’s right? My money is on the ACS and NHIS. Notably, the federal surveys have much higher response rates, which makes them much less vulnerable to changes in what types of people opt to participate.
Additionally, Gallup made a major change to its survey methodology from 2017 to 2018. They report having made adjustments for that change, but this can be hard to do perfectly.
Together, these factors make me discount Gallup’s estimate of the increase in uninsured rate since 2016. Indeed, the ACS and NHIS showed much smaller upticks for the same age group in 2017. I would bet on that happening again in 2018.
(Sidenote: The upward drift in Gallup’s estimates has implications for other time periods too. In particular, Gallup’s data appear to *understate* the 2013-2016 decline in the uninsured rate.)
To be clear, I do expect recent policy changes will put significant upward pressure on the uninsured rate over time. But I think those effects will appear only gradually and are largely still to come.
And upward pressure from federal policy changes may be offset in part by other factors, including a strengthening labor market and various state efforts to expand insurance coverage.
