Donald Trump has eaten away at most of Hillary Clinton’s lead in the national head-to-head poll averages after trailing badly in August, leading to a surge of enthusiasm/panic (depending on how you view this race). It’s the third time Trump has surged in the poll average, and unlike late May (after his primary opponents dropped out and before Hillary’s did) and late July (after his convention “bounce”), he has yet to pull ahead of her:
The RealClearPolitics map of state-by-state polling still has Trump losing the Electoral College 293-245, even with Ohio, Florida and Iowa (278-260 if North Carolina shifts his way, and the pickups get tougher from there):
The state-by-state polls are more important than the national averages right now, but they also tend to lag the trends in more regularly-conducted national polls, so you really have to keep an eye on both. Also the state-by-state poll averages now have the advantage (if you think Gary Johnson and Jill Stein will draw a non-trivial number of votes) that most of them are drawn from 3- or 4-way polls. Consider how the 4-way polls have Trump a little closer to Hillary, due mainly to Johnson sapping her advantage with younger voters (who mostly loathe Trump), and a less volatile race (partly because a 4-way average excludes some of the “swingier” polls):
But Trump still trails. Let’s add a little historical perspective, then. Four years ago, Mitt Romney was much further back than Trump at this point, with President Obama still riding high on his convention bounce (the conventions were a month earlier this year than in 2012 and 2008) – but would go on to surge into the lead in the national averages in mid-October after a strong win in the first debate:
The overall pattern of that race was not so dramatically different, a seesaw between a Democratic lead and a few tightenings that pushed the GOP ahead for a moment. Romney was always weaker in the state-by-state polling, although in mid-October he was leading or even in several swing states that he ended up losing, some of them not especially narrowly: in Colorado, he led from October 9-29; in New Hampshire, he led from October 19-21; in Virginia, from October 19 to November 2; in Florida, from October 8 to the end of the race.
What about 2008, the last time we had an election without an incumbent on the ballot? John McCain was in better shape in mid-September in the RCP national average than Trump, but with three big differences: he was coming down from a lead, his lead was the immediate post-convention bounce, and September 15, 2008 was the day Lehman Brothers filed for bankruptcy, the moment when the 2008 financial crisis hit “jailbreak” levels of panic, driving the race decisively to the party out of power:
Finally, we have 2004, and a trend that mirrored 2012: Bush was up big in September, but the race tightened after the first debate:
You’ll notice that in absolute terms, Trump even in a 2-way race is at 44.2%, compared to Bush at 49.2%, McCain at 45.7%, and Romney at 45.5% – he’s still got the weakest support at this stage. Hillary at 45.7% is running even with Obama ‘08 and ahead of Kerry’s 43.5%, but well below Obama ’12’s 48.5%. Again, the two incumbents were much closer to 50% at this stage than any of the other candidates (Obama had a lot more locked-in support by this stage in 2012, but ended up with fewer votes than in 2008 because he didn’t get the big break at the end from the financial crisis).
What we can take away is that there’s still probably at least one, maybe two more shifts to come in this race, and while a game-changer of 2008 proportions is unlikely, it would be foolhardy to just project a straight-line trend where Trump’s current momentum holds. He still has more people to convince than any Republican nominee at this point since Bob Dole. But we also know that with Gary Johnson in the mix and two outrageously unpopular candidates, one of them completely unconventional, history may not have all the answers.