Donald Trump is, to put it mildly, a different kind of candidate. As such, there’s been a lot of speculation that his support may come from different parts of the United States than that for recent Republican nominees.
With Hillary Clinton replacing Barack Obama atop the Democratic ticket, we might also expect her appeal to be stronger than Obama’s in certain states, and weaker in others.
Though the polls so far suggest that this year’s candidates are not re-drawing the electoral map as much as that thinking might imply, there are important changes to the parties’ relative strengths in some states.
Clinton currently leads by 7 percentage points nationally. Obama won by 3.9 points in 2012. If the swing is uniform across all 50 states, we would expect Clinton to be running about 3 points better than Obama’s 2012 margin in each state.
According to FiveThirtyEight’s “now-cast” – Nate Silver and his team’s model for how the election would go if held today – about half the states roughly fit that uniform swing. In 24 states, the swing towards Clinton is within 4 points of her national swing.
The state that most strongly defies uniform swing is Utah. Helped by its 62 per cent Mormon population, Mitt Romney won Utah by a whopping 48 points in 2012, making it his strongest state in the country. In 2008, non-Mormon John McCain won it by 28 points, despite losing to Obama by 7.3 points nationally.
This year, the state looks a lot closer. The latest poll – by Dan Jones Associates – had the very non-Mormon Trump ahead by just 12 points. A previous iteration of the poll in March actually found Clinton ahead by two points, and another had the race tied – as did a Survey USA poll in early June.
Still, although Utah is much closer than in 2012, it doesn’t actually look all that close. FiveThirtyEight’s now-cast has Clinton losing the state by 8.6 points, even as she wins by 7 nationally. In its two forecasts for the actual election in November – “polls-only” and “polls-plus” respectively – FiveThirtyEight puts Trump’s chances of carrying the state at 77 per cent or 95 per cent.
On the other end of the spectrum are two states where Trump is doing significantly better than McCain or Romney. The first is his home state of New York. Obama won there by 27 points in 2008 and 28 in 2012, but Clinton currently leads by only 19 points according to FiveThirtyEight’s polling average.
The second is Vermont, home to Clinton’s erstwhile rival for the Democratic nomination, Bernie Sanders. Obama won the state by 37 and 36 points, but the two polls there so far have Clinton ahead by 15 and 22 points. Her lead may grow, however, as the salience of her primary battle with Sanders fades. And anyway, both of these states are very unlikely to go Republican or to affect the overall outcome of the election.
So what about the closer, more important states? Much of the speculation about a changing electoral map has focused on Trump’s appeal to white voters in the big Rust Belt swing-states – Ohio, Pennsylvania and Michigan – through his anti-immigrant, anti-free trade shtick.
Yet the polls do not suggest that Trump has any particular strength in those states. He’s down by 8.7 in Pennsylvania, which Obama won by 5.4 – almost exactly as you’d expect on a three-point national swing towards the Democrats. He’s doing a touch better than uniform swing would suggest in the other two, but he’s still down by 4.6 points in Ohio (which Obama won by 3.0 in 2012) and 11 in Michigan (which Obama won by 9.5).
Instead, Trump is polling well (relative to McCain and Romney) in two very different swing states: Iowa and Nevada. In both 2008 and 2012, Obama won both states by bigger margins than he won the national popular vote. He beat Romney by 5.8 points in Iowa and 6.7 in Nevada. The polls currently suggest that Clinton is ahead in both states, but not by as much.
According to FiveThirtyEight’s polling averages, Clinton leads by just 2.3 points in Iowa and 3.2 in Nevada – both well below her 7-point lead nationally. If the race tightens up as we head towards election day, it’s not hard to see Trump winning either state. Indeed, FiveThirtyEight’s polls-plus forecast gives him a 40 per cent chance of winning Iowa, and a 31 per cent chance in Nevada.
However, Clinton is offsetting her poor polls in Iowa and Nevada with stronger ones in Virginia (home of her running mate, Tim Kaine), New Hampshire and Colorado. She’s up by 9.6 points in Virginia, which Obama won by 3.9 in 2012. She’s up by 9.5 in New Hampshire, which Obama won by 5.6. And she’s up by 10 in Colorado, which Obama won by 5.4.
If Clinton wins those three states, as well as Pennsylvania and Michigan, it’d be very hard for Trump to make it to the 270 electoral votes he needs to become President – even if he were to win Ohio, Iowa, Nevada, and Florida.
And Clinton looks as if she might be able to win a number of states that Obama didn’t in 2012 – including some that Democrats haven’t won for a long time. FiveThirtyEight’s polling averages have her leading by 4.3 points in North Carolina, almost tied with Trump in Georgia, Arizona and South Carolina, and just 2.5 behind in Missouri. Both its forecast models make Clinton at least slight favourite to win North Carolina, and give her at least a 20 per cent chance in each of the other four states.
The last Democrat to win South Carolina was Jimmy Carter, 40 years ago.