Attention, Paul Krugman: Not everyone votes

Am I crazy or is Paul Krugman making a really elemental mistake here?

As Greg Sargent points out, the choice of Tom Price for HHS probably means the death of Obamacare. Never mind the supposed replacement; it will be a bust. So here’s the question: how many people just shot themselves in the face?

My first pass answer is, between 3.5 and 4 million. But someone who’s better at trawling through Census data can no doubt do better.

Here’s my calculation: we start with the Census-measured decline in uninsurance among non-Hispanic whites, which was 6 million between 2013 and 2015. Essentially all of those gains will be lost if Price gets his way.

How many of those white insurance-losers voted for Trump? Whites in general gave him 57 percent of their votes. Whites without a college degree — much more likely to have been uninsured pre-Obama — gave him 66 percent. Apportioning the insurance-losers using these numbers gives us 3.42 million if we use the overall vote share, or 3.96 million if we use the non-college vote share.

The assumption appears to be that everyone votes. Turnout this year is currently reported at 58.6%. It's probably lower for most of those who gained insurance via the ACA, as their incomes are below median: according to the Current Population Survey, almost three quarters of those who gained insurance in both 2014 and 2015 have incomes below twice the Federal Poverty Level, and nearly all have incomes under 300% FPL. Lower income people have lower turnout rates.  It would appear that Krugman's total needs to be sliced almost in half.