Dave Wasserman, of the nonpartisan Cook Political Report, recently tweeted:
A very plausible 2020 scenario Democrats have to take seriously:
Trump loses popular vote by about 5 million as Dems narrow deficit in Texas, expand margin in California, and flip Michigan and Pennsylvania blue.
But Trump narrowly holds onto Arizona, Florida, North Carolina, and Wisconsin and wins reelection by a single electoral vote.
First observation would be that, as you'd expect, Wasserman's math checks out: if 48 states vote as they did in 2016, with only Pennsylvania and Michigan flipping to the Democrats, then Trump wins the Electoral College, 270-268. How likely is this? In the thread, someone asks how likely it is that the Democratic candidate does much better in Texas, thereby adding to the popular vote margin attained by Clinton, but can't make up two-thirds of one percent in Wisconsin. Wasserman's answer is that an influx of new voters--Latinos and young professionals--accounts for the Dem's gains in Texas, and these conditions do not pertain in Wisconsin. Another reasonable question: why would Michigan and Pennsylvania fall into the Democratic column but not Wisconsin? Wasserman's answer concerns the different demographics of these states, specifically the percentage of non-college whites compared to the percentage of African Americans:
Pennsylvania 50%/10%
Michigan 53%/13%
Wisconsin 57%/6%
I would add that in the 2018 midterms Republicans fared significantly better in Wisconsin than in either Pennsylvania or Michigan. All three had statewide races for governor and US Senate, and the Democrats went 6-for-6, but their narrowest win by far was the Wisconsin governor's race. The Democrats did not pick up any US House seats in Wisconsin, and, in the state legislature, Republicans held their large majorities in both chambers. (State legislative districts in Wisconsin are so gerrymandered, however, that there should almost be a different word invented to describe the situation: in 2018, for example, Democratic candidates for State Assembly won 53% of the statewide vote, but Republicans won 63 of the 99 seats.)
So Wasserman's scenario seems plausible, especially if Trump's position improves modestly between now and November of 2020.
Second observation would be that, once you have contemplated the prospect of a 270-268 Trump victory in the Electoral College, it's natural to consider the possibility of a 269-269 tie. As it happens, there are two states--Maine and Nebraska--that award their electoral votes by congressional district. In 2016, Trump pulled one of Maine's four electoral votes, because, while he lost the statewide vote, he did outpoll Clinton in one of the two districts. Meanwhile, Nebraska, though a deep red state, has one district, centered on the city of Omaha, that is competitive: Trump won it in 2016 by 47% to 45%. If the Democratic candidate were to win one of these districts, but not the other, and all else shook out as described in Wasserman's scenario, then the Electoral College would not render a verdict, despite the Democratic candidate's popular vote margin of 5 million. I described here what happens in that event. Briefly, the president is then selected by the House of Representatives, but, crucially, not by the vote of the full House--instead, each state delegation gets one vote, and the winner is the presidential candidate who is preferred by the delegations of at least 26 states. Notwithstanding the Democratic majority in the current House, there are exactly 26 state delegations with more Republicans than Democrats. In the scenario we are now considering, therefore, the Democratic candidate wins the national popular vote by about 5 million ballots, which, in percentage terms, would be something like 51-48. Trump, however, ekes out a tie in the Electoral College, and is then "elected" president when, despite the Democratic majority in the House, he is the choice of a majority of state delegations--Wyoming and Idaho having between them twice the pull of California or Illinois or New York.
Third observation relates to the prospect of the presidential election triggering a national crisis of legitimacy. Trump has "joked" persistently about not leaving office after his second term. What if he loses in 2020 and declines to step down? Since we're considering scenarios where the outcome is very close, we should remember that "very close" can equal "ambiguous," as in 2000 when Bush's Electoral College victory was secured after he was declared the winner in Florida by fewer than 600 ballots. Gore, who won the national popular vote by more than half a million, conceded defeat after the Supreme Court ruled against him. Who thinks Trump would act similarly in the event that, say, the election result hinged on the outcome in Wisconsin, where state election officials deemed the Democrat to have won by 500 votes? What if some alleged "irregularity" was the basis of a lawsuit that then went to the Supreme Court, where Trump appointees supplied the votes that set in motion events that reversed the outcome? What then would Democrats do? Lay down like lambs? Might people stop going to work and take to the streets in mass protests? It's mainly on account of our democratic traditions that such a scenario seems outlandish, but, as is often said, Trump has no regard for these traditions--he's proud of smashing them.
Fourth and last observation would be that the Electoral College is the root cause of these various scenarios, all regrettable to varying degrees and the last outright dangerous, for in every scenario there is no question about which candidate got more votes. It wasn't even particularly close--5 million.
Comments