A series of national polls released on Tuesday showed Hillary Clinton leading Donald Trump by margins ranging from 5 to 11 percentage points — except for the USC Dornsife/Los Angeles Times tracking poll, which defiantly continues to show Trump up by 2 points. There isn’t yet enough data from after Sunday night’s debate to really gauge its impact, however. For that matter, the polls may not yet have fully caught up to the effects of the release on Friday of a 2005 videotape, which showed Trump making vulgar comments about women and condoning unwanted sexual conduct toward women. For the time being, Clinton’s lead is holding at about 6 percentage points in our polls-only model, which gives her an 84 percent chance of winning the White House. Clinton’s chances are 80 percent in our polls-plus forecast.
But while we’re in something of a wait-and-see mode, one demographic split caught my eye. That was from a Public Religion Research Institute poll conducted on behalf of The Atlantic. It showed a massive gender split, with Clinton trailing Trump by 11 percentage points among men but leading him by 33 points among women. To put those numbers in perspective, that’s saying Trump would defeat Clinton among men by a margin similar to Dwight D. Eisenhower’s landslide victory over Adlai Stevenson in 1952, while Clinton would defeat Trump among women by a margin similar to … actually, there’s no good comparison, since no candidate has won a presidential election by more than 26 percentage points since the popular vote became a widespread means of voting in 1824. To get to 33 points, you’d have to take the Eisenhower-Stevenson margin and add Lyndon B. Johnson’s 23-point win over Barry Goldwater in 1964 on top of it.
The PRRI poll is an outlier for showing quite so large a gender gap, however. In the table below, I’ve compiled the gender split from all national polls I could find so far in October.1
National polls in October show a wide gender gap
| POLL | WOMEN | MEN |
|---|---|---|
| CBS News | Clinton +18 | Trump +11 |
| CNN | Clinton +14 | Trump +4 |
| Fairleigh Dickinson | Clinton +24 | Trump +7 |
| Fox News | Clinton +10 | Trump +7 |
| Google Consumer Surveys | Clinton +13 | Trump +3 |
| Ipsos/Reuters | Clinton +9 | Clinton +5 |
| Morning Consult | Clinton +6 | Clinton +4 |
| PRRI/The Atlantic | Clinton +33 | Trump +11 |
| Quinnipiac University | Clinton +20 | Trump +12 |
| Rasmussen Reports | Clinton +11 | Clinton +2 |
| USC Dornsife/LA Times | Clinton +9 | Trump +14 |
| YouGov | Clinton +15 | Trump +2 |
| Average | Clinton +15 | Trump +5 |
There’s quite a wide range — with live polls showing a notably wider gender spread than online and automated polls. But on average, Clinton leads Trump by 15 percentage points among women while trailing him by 5 points among men. How would that look on the electoral map?
Here’s a quick way to estimate it. In the polls I cited above, Clinton is doing 10 points better among women than among the electorate overall.2 So we’ll add 10 points to her current polls-only margin in every state to forecast her performance if women were the only ones who could vote. In addition to the states where Clinton is already leading Trump, that would put her ahead in Alaska, Arizona, Georgia, Kansas, Indiana, Missouri, Montana, South Carolina, South Dakota, Texas and the 2nd congressional districts in Maine and Nebraska. Clinton would win 458 electoral votes to just 80 for Trump:
If men were the only voters, conversely, we’d have to subtract 10 points from Clinton’s current margin in every state — which would yield an awfully red map. Trump would win everything that could plausibly be called a swing state, with Clinton hanging on only to the West Coast, parts of the Northeast, Illinois and New Mexico. That would yield 350 electoral votes for Trump to 188 for Clinton:
Of course, if the electoral franchise were really restricted only to men, our politics would look a lot different. And if it were restricted only to women, they’d look really, really different, perhaps with entirely different coalitions than the ones that prevail now.
But it seems fair to say that, if Trump loses the election, it will be because women voted against him. I took a look at how men and women split their votes four years ago, according to polls conducted in November 2012. On average, Mitt Romney led President Obama by 7 percentage points among men, about the same as Trump’s 5-point lead among men now. But Romney held his own among women, losing them by 8 points, whereas they’re going against Trump by 15 points.
The gender gap in November 2012 national polls
| POLL | WOMEN | MEN |
|---|---|---|
| ABC News/Washington Post | Obama +9 | Romney +9 |
| American Research Group | Obama +11 | Romney +12 |
| CNN | Obama +8 | Romney +9 |
| Gravis Marketing | Obama +2 | Romney +2 |
| GWU/Battleground | Obama +15 | Romney +16 |
| Monmouth University | Obama +5 | Romney +7 |
| NBC News/Wall Street Journal | Obama +8 | Romney +7 |
| Pew Research | Obama +11 | Romney +10 |
| Public Policy Polling | Obama +5 | Obama +2 |
| Purple Strategies | Obama +5 | Romney +3 |
| YouGov | Obama +12 | Romney +9 |
| Average | Obama +8 | Romney +7 |
That’s the difference between a close election — as you’ll remember, those national polls in late 2012 showed the race neck-and-neck3 — and one that’s starting to look like a blowout.
Click here to view original web page at fivethirtyeight.com
