David Wasserman on FiveThirtyEight claims that even if Trump achieves historically high 1992 white voter turnout (i.e., the year Ross Perot ran), he will still lose the election because, surprisingly, it only swings Florida, Nevada, and Ohio Republican, all other things equal.
Of course, all other things are not equal. Black turnout may fall from
66% back to 58% (e.g., 2004) when President Obama is no longer running. Hispanic turn out may rise in opposition to Trump. And some share of college-educated white voters may vote against Trump. As Wasserman points out: “In 2012, Romney carried [college educated white voters] 56 percent to 42 percent. But in the most recent Washington Post/ABC News poll, whites with a college degree favored Trump over Clinton by just 46 percent to 45 percent.”
FiveThirtyEight’s Swing-O-Matic is a very cool tool for running election scenarios. I tried one: Non-college educated white voter turnout rises from 57% to 66%, with the whole 9-point increase voting for Trump (i.e., [[66%-57%]+[62%x57%]/66%]), and the same turnout of Blacks for Pres. Obama. Surely, that gives Trump support the benefit of the doubt, but Trump still loses, although it’s a close election. If Black participation falls back to 58%, Trump then pulls out a victory. But if only 2 points of college-educated whites vote against Trump (54% vs. 56%), Trump loses narrowly.
Clearly Trump needs to do more than just energize non-college educated white voters. As Wasserman points out: “Whites without college degrees — a group that made up 65 percent of voters in 1980…is on pace to make up just 33 percent in 2016. Meanwhile, nonwhite voters, with whom Trump is extremely unpopular, rose from 12 percent of the electorate in 1980 to 28 percent in 2012,” and, according to Wasserman, is on pace to reach 31% in 2016.
No surprise, Betfair currently pegs the chances of a Trump victory at only 30%.