For about two weeks now, Donald Trump has been making a slow, steady climb back into competition. In that time, whenever a national poll showed the race getting a little too close for comfort, liberals would seek solace in the same way as all weary travelers who suddenly feel lost in a land they thought they knew — by looking at a map.
One gaze at that blue-and-red battlefield, and the mogul’s chances would once again feel safely remote: There were just so many different things that needed to happen for Trump to become president! After all, the authoritarian insult comic could hold on to his slim advantage in Ohio and Iowa, win a toss-up in Florida, mount an improbable comeback in North Carolina, fend off outraged Latino voters in Nevada, and still lose to Clinton — as long as she held on to Colorado and New Hampshire, where she’d led in damn near every poll since July.
This comforting reasoning was true enough, on its face. But it always contained a hint of fallacy — that each state election was its own independent event, and that what shifted the winds in Florida would never be felt up in New Hampshire. Which is to say, there was an assumption that Clinton had a structural advantage in the Electoral College, which she’d retain even in a tied national race. But according to FiveThirtyEight’s model, the opposite is true: The abundance of white voters without college degrees in several key swing states provides Trump with the best chance of losing the popular vote, while winning the presidency.
On Thursday morning, a CBS News/New York Times national poll showed Clinton leading Trump by just 3 points, down from 9 in mid-October — and new surveys showed the Democratic nominee and Donald Trump deadlocked in New Hampshire and Colorado.
A WBUR/MassINC poll of the Granite State shows Trump leading by a skinny digit, 40 to 39 percent. Two weeks ago, that survey had Clinton up 41 to 38. Both of those findings were well within the poll’s 4.4 percent margin of error.
A new Suffolk University/Boston Globe poll shows a similar result, with Clinton and Trump tied at 42 percent. In October, Clinton enjoyed a two-point lead in the survey.
Other recent surveys of New Hampshire have shown a significant tightening of the race, albeit while keeping Clinton solidly ahead. WMUR/UNH had Clinton up 7 earlier this week, after finding her leading by 15 in mid-October; Emerson had Clinton up by 8 two weeks ago, and now by just 3.
And the campaigns’ internal polling seems to show a close race in New England’s only swing state. Trump is scheduled to campaign in New Hampshire on both Friday and Monday, when President Obama will also be in the area, trying to seal the deal for Clinton. With only four electoral votes on the line, it seems doubtful that the Democratic nominee would send her most valuable surrogate to the Granite State on the eve of Election Day if her campaign wasn’t worried that it might slip out of her column.
Meanwhile, a University of Denver poll shows Trump and Clinton tied at 39 in Colorado, where the third-party candidates are enjoying aberrant strength, drawing 15 percent of likely voters. The last four polls of the Rocky Mountain State have all had Clinton ahead — but by the slim margins of 1, 2, and 3 percent.
That said, the University of Denver poll comes with a big caveat.
Pollsters have a famously hard time assembling accurate samples in states with heavy Latino populations. And so there is some reason to think Clinton’s Election Day performance will outpace her polling. At present, Democrats have cast roughly 23,000 more early ballots than Republicans have in Colorado.
Regardless, Clinton remains the overwhelming favorite. Individual polls are chock-full of noise.
But the race is getting closer. And if it continues in this direction, progressives will find little comfort in consulting their maps.
