Ron Brownstein has an excellent and important new article out on the Atlantic website. And I say that not just because he uses my States of Change data and quotes me in a few places! No, it's because his analysis is spot on and he's really done his homework, digging into the data in a thorough way most political journalists never bother with.
"Signs are growing that voter turnout in 2020 could reach the highest levels in decades—if not the highest in the past century—with a surge of new voters potentially producing the most diverse electorate in American history.
But paradoxically, that surge may not dislodge the central role of the predominantly white and heavily working-class voters who tipped the three Rust Belt states that decided 2016: Pennsylvania, Michigan, and Wisconsin. Even amid a tide of new participation, those same voters could remain the tipping point of the 2020 election....
Data from States of Change show that over the past quarter-century, white voters without a college education have typically declined as a share of actual voters little by little over each four-year presidential cycle: They fell from 61 percent of voters in 1992 to 44 percent in 2016. Minority voters, meanwhile, have increased over those seven election cycles from 15 percent in 1992 to 26 percent in 2016. And college-educated whites have drifted up, from 24 percent in 1992 to 30 percent in 2016.
The 2016 election, however, was something of an anomaly for blue-collar white people. Partly because Trump inspired so many non-college-educated white voters to turn out—and partly because African American turnout skidded so badly—white working-class voters declined less than usual in 2016 as a share of the electorate, States of Change concluded. But in 2018, as turnout surged among minority and younger voters, non-college-educated whites suffered a much sharper decline: Compared with 2014, they fell by 4 to 5 percentage points as a share of the total vote, according to both Catalist’s estimates and McDonald’s analysis of census figures. That’s about double their average decline from one presidential election to the next over the past quarter-century.
Ruy Teixeira, a veteran liberal analyst and a co-founder of States of Change, believes it’s likely that in 2020 the decline in blue-collar white people’s share of the total vote could again push toward the high end of recent experience, shrinking by as much as 3 percentage points, to just over 40 percent. “I think if we do have a high-turnout election that builds on the trends we saw in 2018, you might see the white non-college share decline significantly more than it did in 2016,” Teixeira says.
Those changes pose obvious problems for Trump in winning the national popular vote. But they also present a challenge for Democrats, because these shifts are not evenly distributed among the states. The electorate is not diversifying nearly as fast in the three Rust Belt states that Trump dislodged from the Blue Wall—Michigan, Pennsylvania, and Wisconsin. Those states, for years to come, will remain older and whiter than the nation overall, meaning that to win them, Democrats have to run better with older, whiter voters than they do in most places.....
Unless and until Democrats can tip some of the potential Sun Belt battlegrounds, particularly Arizona, Florida, and North Carolina, the party can’t reach 270 Electoral College votes without recapturing some of the Rust Belt states least affected by demographic change.
Teixeira is one of many Democratic strategists who say the party’s top priority must remain regaining those Rust Belt states, because it cannot yet rely enough on the Sun Belt. “How can you possibly count on these states?” he says. “Democrats haven’t won Florida for a while. Arizona, they haven’t won in a million years. Georgia, Texas—are you kidding me? These are hard states. You cannot build a strategy around having to win those states.”
Chipping into Trump’s base of non-college-educated and rural white voters isn’t the only way for Democrats to win back the Rust Belt states he took in 2016: They could also theoretically recapture them by increasing turnout among young people and minorities, and converting more suburban white people. But in a 2020 election likely to be defined by a historic surge of new voters, many Democrats are resigned to facing the same old challenge of scratching out a few more votes in mostly white union halls and country diners across the Upper Midwest."
So in 2020 we'll likely have a spectacularly large, diverse electorate, with high turnout across the board. But the white working class will still matter. A lot.