In recent weeks, with the rise of Sen. Bernie Sanders (I-Vt.) in Democratic presidential primary polling, pundits have warned that the party faces a McGovern-like scenario in November: a moderately popular president gliding to reelection thanks to deep skepticism about his opponent. Republican consultant Rick Wilson, probably the most colorful member of the Never Trump pundit community, recently predicted that in a contest between Sanders and President Trump, Trump would win 44 states.
He would not. Setting aside questions of polling (in which Sanders, a well-known politician, consistently beats Trump), we can look at how the electorate and election results have shifted in the past 50 years to make clear that a similar margin of victory is unlikely.
Nixon won the presidency in 1968, beating Hubert Humphrey handily. Four years earlier, President Lyndon B. Johnson won reelection in a country still recovering from President John F. Kennedy’s assassination a year before. The swing from Johnson’s massive win to Nixon’s narrower one was wide; on average, states voted 21 points more Republican (in the two-party vote) in 1968 than they had four years before. Four years later, when Nixon beat McGovern, the average shift was even wider — 23 points. On average, states moved 45 points to the right in the two-party vote between 1964 and 1972.
Those are humongous swings. In 2016, the average shift was only 4 points. Only one state moved more than 20 points since 2012: Utah, which backed local(ish) boy Mitt Romney in 2012 and wasn’t terribly enamored of Trump. When electing candidates as different as Obama and Trump, though, 42 states shifted 10 points or fewer from 2012 to 2016.
Why? In part because states have become more consistent in their voting. In the three election cycles centered on the 1972 presidential contest (that is, in 1970, 1972 and 1974), there were 38 Senate contests in which states supported senators and a 1972 presidential candidate of the same party. In 60 Senate contests, states backed a senator from one party (generally a Democrat, given how 1972 turned out) and a presidential candidate from the other.
In the three elections centered around the 2016 contest, there were 84 same-party Senate elections and only 10 in which states backed a senator from one party and a president from the other. (Seven of those were in 2018.)
That’s in part a function of Nixon’s blowout win. But since the early 1990s, there’s been a consistent trend toward fewer Senate-president split-ticket votes.
Part of this is because of migration, the Big Sort that has, among other things, moved a lot of Democrats to large coastal cities. Part of it is due to the evaporation of the political middle.
The Pew Research Center has consistently measured the ideological views of partisans. Since 1994, its data shows there’s much less ideological overlap between the parties.
In 1994, for example, about 3 in 10 Democrats were more conservative than the median Republican. About a third of Republicans were more liberal than the median Democrat. By 2017, there were almost no outliers of that sort: Only a small percentage of each party was similarly more ideological than the other party’s median.
This is self-reinforcing. As Republicans become more conservative, for example, fewer are liberal, and the median shifts to the right, making it less likely that any Democrats will sit to the right of the median ideologically.
Pew’s data also suggests increased hostility from Democrats toward Republicans and vice versa. In 2016, nearly 6 in 10 members of either party viewed members of the opposite party very unfavorably — making it much less likely that they might vote for a member of the opposing party. The increase in self-identified independents hasn’t really meant more moderation. Instead, many independents are primarily motivated to vote against one party.
All of which is to say that, given a choice between Trump and Sanders, states that backed Hillary Clinton in 2016 are very likely to back Sanders this time. To win 14 more states, bringing his total to 44, Trump would need to win states that backed Clinton by as many as 16 points — Illinois, for example. That’s not going to happen barring some substantial, unimaginable collapse in the partisanship Americans have been nurturing for the past several decades.
The theory presented by those warning about a Sanders-Trump contest is that 2020 might be for millennials what Reagan-Mondale or Nixon-McGovern were for previous generations. It’s certainly possible. But in 2020, a landslide is much more likely to look like Barack Obama’s 192 electoral-vote win over John McCain in 2008 than like Reagan’s 512 electoral-vote win 36 years ago.