The Bounce Continues
Two weeks ago, I noted that Hillary Clinton’s convincing defeat of Donald Trump in the first presidential debate had improved her probability of victory from 52.1 percent as of the day of the debate to 71.9 percent three days later. I used the “now-cast” index developed by Nate Silver of FiveThirtyEight.com as the most accurate reflection of the short-term impact of election-related events.
As of this writing, the “now-cast” gives Clinton a 90.8 percent chance of victory, meaning that her chances have improved by as much after the initial post-debate bounce as they did in the bounce itself.
The “now-cast” also shows an important shift in the states seriously in play. In addition to its national projection, the “now-cast” assesses the probability of each candidate winning in each individual state. Where neither candidate had a winning probability of 70 percent or greater, I considered the state to be a battleground state. Two weeks ago, the battleground states were Arizona, Florida, Iowa, Nevada, North Carolina, and Ohio. As things stood then, even if Trump won every one of them, he would lose the electoral vote by a narrow margin.
Trump’s electoral map has gotten quite a bit worse in the last two weeks. As of today, the “now-cast” indicates that there are five battleground states: Alaska, Arizona, Iowa, Georgia, and Ohio. Florida and North Carolina have moved from the battleground to Clinton’s column in the “now-cast,” and Alaska and Georgia have moved from Trump’s column to the battleground.
When Alaska and Georgia are in play, Republicans are in serious trouble. But I guess we already knew that.
The “now-cast” has Clinton slightly ahead in Arizona, Iowa and Ohio. But even if Trump were to win all five of the current battleground states, he would be 54 electoral votes short of victory. In other words, Trump would have to sweep the battleground PLUS win 54 electoral votes in states where the “now-cast” gives him less than 30 percent chance of victory.
Oh, and just for yuks, a Utah state poll released today shows Clinton and Trump tied at 26 percent each in that state, with third-party candidate Evan McMullin taking 22 percent. Utah, like Alaska, hasn’t voted for a Democratic presidential candidate since 1964. (Due disclaimer: the poll in question is by Y2 Analytics, which FiveThirtyEight rates as a C+ polling firm.)
The “now-cast” is a portrait of where things stand now; it is not a prediction of the actual election vote. But as the election comes closer, the “now-cast” becomes more of an election forecast. In fact, to the extent that voters have cast or are about to cast early ballots, the “now-cast” is an election forecast.