I see the NY Times, and other sources shifting to a sense of blue inevitability. I remember 2016 too well. Beware.

Election coverage can make a race feel settled before voters have finished deciding it. Looking back at the New York Times’ 2016 coverage, then-executive editor Dean Baquet said he would have edited early stories to give less of a sense Hillary Clinton’s ‘win’ as inevitable. He also acknowledged relying too heavily on party insiders and spending too little time reporting across the country. Jay Rosen’s account of Baquet’s interview records those admissions. They basically had noone on the ground in the Midwest; just opined from their tall skyscrapers in Manhattan.

The election result cannot be reduced to a story about supporters staying home. Census Bureau figures show that overall turnout was statistically similar to 2012. Within that total, Black turnout fell from 66.6% to 59.6%, while white turnout rose. Those shifts mattered, but turnout was only part of what changed. Voters changed their choices, too. A study of six states found that switching between parties explained more of the change in results than turnout did in four of them, including Michigan and Pennsylvania.

Beware hubris. On the part of the parties, on the part of the media.

The lesson for anyone reading an election forecast is to leave room for what it cannot see. Confidence about a result is a description of the evidence available at the time. It is not a substitute for reporting on voters, and it is not a result.