Four days before the presidential election, the newest polls continue to suggest that Hillary Clinton is still narrowly ahead — though it’s close enough that the election could well turn out to be a nail-biter.
Clinton’s decline in national polling averages, which are heavily influenced by daily tracking polls at this point, has reversed itself a bit. As of Friday afternoon, Clinton was leading Trump by 2.4 points in the RealClearPolitics average, while the FiveThirtyEight model estimates she’s up by 2.9 points.
The various forecasting models all currently give Clinton an 84 percent chance of winning or greater — except for FiveThirtyEight, which gives her a significantly lower 65.1 percent shot of winning. (I wrote about why FiveThirtyEight’s forecast differs from the others earlier this week.)
As for state-level results, we’re now being flooded with them, and many seem to show somewhat contradictory results — varying between reassuring for Clinton and worrying for her.
How’s Clinton’s firewall doing?
Clinton’s easiest path to victory appears to be by holding onto her six “firewall” states of Wisconsin, Michigan, Virginia, Colorado, Pennsylvania, and New Hampshire, along with the other solidly Democratic states.
Right now, the weakest firewall state appears to be New Hampshire. Though she has long led the vast majority of polls in the state, four new ones were released on Thursday and zero of them showed her ahead. (Two showed a tied race, one a 1-point Trump lead, and one a 5-point Trump lead).
Then today, we got a poll from the Democratic firm PPP showing Clinton up 5 and one from Breitbart/Gravis showing Trump up 1, so the state of the race in New Hampshire is uncertain. Luckily for Clinton, though, New Hampshire is the smallest of her firewall states, with only four electoral votes up for grabs. She could make up for a loss there by winning pretty much any other swing state (Nevada, Florida, North Carolina, Arizona, Ohio, or Iowa).
Pennsylvania is much more important to the firewall, and while Clinton has continued to lead most polls there, her margin does appear to have tightened to the low single digits, with one new poll even showing a tied race. (No polls of Pennsylvania have shown Trump ahead since July.) Since there is no early voting in Pennsylvania, election day turnout will be crucial here, and Clinton will spend much of the campaign’s last few days in the state.
There’s also been new attention on Michigan, a firewall state that had long appeared safe for Clinton. A poll from the Detroit Free Press released Friday shows Clinton up 4 points, which is close to a poll from earlier this week showing a 3-point race. Clinton is campaigning in Detroit today amid chatter that Democrats could be concerned about low black voter turnout in the state.
But Clinton got better news in Wisconsin, where a poll from Loras College shows her up 6, corroborating a bunch of recent results showing Clinton in a reasonably strong position. She also got a good poll in Virginia from Roanoke College, which shows her up 7, her best result all week.
What about the other swing states?
Of course, Clinton is competitive beyond those firewall states. Her next best shots for important victories appear to be in Nevada, Florida, and North Carolina.
These states have been sparsely polled over the past few days. Perhaps more importantly, in all three, early voting has been ongoing for weeks. And in each, votes equivalent to more than half the 2012 vote total have already been cast. So, essentially, it’s not all that clear how much new polls would really tell us about these states with so many votes locked up.
Naturally, then, people have been trying to interpret early vote tea leaves in each of them — check out Jon Ralston in Nevada, Daniel Smith and Steve Schale in Florida, and Nate Cohn analyzing North Carolina numbers. Right now, the consensus is that the Nevada numbers appear to look good for Democrats, but there’s mixed signs in North Carolina and Florida.
Then there are three other states Clinton is still competing in that seem like more of a reach for her: Ohio, Iowa, and Arizona. Most recent polls have shown Clinton down in all of them, and nothing has come out today to change that.
Overall, then, the polls we have do show Clinton definitely ahead nationally and in a critical mass of swing states — but exactly how strongly she should be favored depends heavily on a few key assumptions about how likely it is the polls might be wrong, and what might be happening in states we haven’t gotten much polling from lately.