I got a little excited when I saw the headline to this article at Real Clear Politics—especially the part about the possibility of a "tie" in this year's presidential election. The most likely scenario would be a fairly static outcome in which only Pennsylvania and Michigan, Trump's two narrowest wins, change their 2016 color on the ubiquitous Election Night map. That would leave Trump ahead in the Electoral College, 270-268. Nebraska, however, is one of two states—Maine is the other—that awards its electoral votes by congressional district, and the state's 2nd district, which is basically Omaha and its suburbs, went for Trump by just 47-45 percent. Omaha's county voted for Clinton but Trump made it up in the surrounding suburban counties—the kinds of places where he and the GOP have been losing support during his time in office. If the Democratic nominee flipped Pennsylvania, Michigan, and Nebraska's 2nd congressional district, and no other electoral votes switched columns, the result would be a 269-269 tie.
The article, I'm sorry to say, is a disappointment—so badly written that it's not always even clear what the author, Bill Whalen, of the conservative Hoover Institution, is trying to say. But I take it he doesn't think there's much chance of a tie. Well, of course not! When you're considering the possible permutations of 50-some separately weighted but binary and distinct outcomes, there aren't going to be many that come out exactly even. I'd bet against it, too. But the above scenario is highly plausible, and others aren't that far behind.
Whalen's basic error, I think, is assuming that a presidential election result is what I'll call "interlocking by state." Since Pennsylvania, Michigan, and Wisconsin all went for Trump by less than 1 percent of the statewide vote, he thinks it's very unlikely that in 2020 the Democratic candidate will capture one or two of them but not all three. He also puts forward another 269-269 scenario that involves the Democrat winning Arizona but losing New Hampshire. He puts it forward in order to argue that it won't happen, since to win Arizona a Democrat would have to do about 4% better than Clinton did—and, at the same time, do worse than she did in New Hampshire. Very unlikely, Whalen thinks.
But Wisconsin isn't Pennsylvania, Arizona is even farther from being New Hampshire, and, to fall back again on my awkward phrase, a presidential election result is not even remotely "interlocking by state." For example, Trump, compared to Romney in 2012, made substantial gains in Pennsylvania, Michigan, and Wisconsin, but he fell back in California and Texas. The rough numbers are interesting. Trump ran ahead of Romney in Pennsylvania by about 330,000, in Michigan by about 460,000, and in Wisconsin by about 220,000. Meanwhile, he lost California by 1.2 million more votes than Romney did, and won Texas by 450,000 fewer. I say the numbers are interesting mainly because, if you sum it all up, in these five states—the country's two largest considered together with three "Rust Bucket battlegrounds"—Romney ran ahead of Trump by around a half million votes but won 46 fewer electoral votes. (The figures are derived not from raw vote totals but by the difference between the opposing candidates' totals: for example, I say Trump "ran ahead" of Romney in Wisconsin by about 220,000 because Romney had lost the state by 200,000, then Trump won by 20,000, a net gain for the Republican candidate of 220,000.)
Contrary to what Whalen seems to think, a Democrat could, from one election to the next, gain considerable ground in a state like Arizona (growing, browning) while actually going backward in a state like New Hampshire (static population that's aging, with few nonwhites). It's exactly what happened in the last election. It's more like the rule than an exception to the rule. The states' demographic profiles are key, which also accounts for why it's no sure thing that this year the same candidate wins Pennsylvania, Michigan, and Wisconsin. Remember how, in our 269-269 scenario, the Democrat flips Pennsylvania and Michigan but Trump holds on to Wisconsin? Most election gurus think that is not at all a remote possibility. The demographic differences make Wisconsin more Trump friendly than the other two. The main consideration is that in both Michigan and Pennsylvania there are about 4 or 5 whites without a college degree for every African-American. In Wisconsin, this ratio is more like 9 to 1.
It's also sort of weird that an article about an Electoral College tie does not take up the question of who then wins and becomes president. The 12th Amendment describes the process; I summarized here. Short answer is that it's another gob of spit in the face of democracy.
Comments