National and state polls have certainly tightened in the last week between Trump and Clinton. CBS and YouGov have Clinton +3, ABC/WaPo has Clinton +2, and IBD/TIPP is tied. GOP outlier Rasmussen has Trump +3, and the wacky LA Times/USC tracker has Trump +5. What’s important isn’t so much what the absolute values of these numbers are, so much as the fact that individually they are mostly leaning farther in Trump’s direction than their previous tallies. We are also seeing the same effects in state polling–and the Clinton campaign is responding in kind by sending surrogates and resources to states like New Hampshire and Michigan which were previously considered mostly safe.
But there are three big reasons for cautious optimism among Democrats.
First, Democrats and liberal constituencies are trouncing Republicans in the early vote. Latino turnout in Georgia, North Carolina and Florida is well beyond 2012 and 2008 benchmarks. In Nevada Democrats lead the statewide early vote by over 45,000, with a whopping 72,000 vote lead in Clark County. Over 30 million Americans have already voted early–and in most of the states (Ohio excepted) the numbers seem to favor Democrats.
Second, while the Trump campaign has been seeking to turn out “missing white voters”–i.e., those who demographically and culturally should be voting for Republicans but who don’t usually turn out to vote–evidence suggests we are seeing a surge of nonwhite “missing” voters this year. As Greg Sargent notes, this is likely in part because of the very same racist tactics that Trump is using to turn out non-political white voters is energizing nonwhite voters against him.
Third, if it seems unlikely that voters are changing their minds this much between Trump and Clinton in the waning days of the election, that’s because it is. When polls seem to “swing” away from the mean–whether because of a convention bounce or a debate–it’s not usually because voters changed their minds so much as because supporters of candidates who recently did well are likely to respond to polls. This is known as response bias, and the closer we get to election day the likelier swings in the polling are to reflect that rather than genuine shifts in the electorate.
At this point most voters have already long decided which side of the aisle they are on. The only question will be how many turn out on each side–and possibly how many disaffected voters will cast protest votes for Johnson or Stein. Which leads us to the final cause for cautious optimism: the Clinton campaign has a strong ground game, and the Trump campaign has very little of it.
If Democrats are nervous about Clinton’s chances and the fortunes of their candidates downballot, it won’t help to refresh the polling aggregators. It’s all about getting out the the vote–the old-fashioned democracy of knocking on doors and making phone calls.