The U.S. presidential campaign seems to have been going on for an eternity. The Labor Day holiday on Sept. 5 marks the start of the final stretch.
Here are some leading indicators to determine whether Donald Trump will make this a competitive contest or whether Hillary Clinton can build on her current advantage to open a more commanding lead.
• Debates: The initial forum is slated for Sept. 26. The event is being anticipated as a a smack-down brawl between two candidates who show contempt for each other. Each will try to bait the other. Will either take the high road and rise above the insults?
There are two more scheduled presidential debates, on Oct. 9 and 19, and a vice-presidential session on Oct. 4. Several of these events coincide with National Football League games, which could cut into the audience, a possibility that has drawn Trump's ire.If history is a guide, these showdowns are too hyped and won't alter the basic contours of the race. There are exceptions. In 1976, President Gerald Ford took a hit after he mistakenly suggested that the Soviet Union-dominated Eastern Europe was free. And in 1980, when the forum took place one week before the election, the challenger Ronald Reagan was able to reassure voters about his competence. On two other occasions, the incumbent president -- Reagan and Barack Obama -- performed poorly in the initial debate, though the contests soon reverted to what they had been.
• College-educated white voters: Mitt Romney carried this bloc, about a third of the electorate, by 14 points in 2012. Polls have shown Trump trailing with these voters by double digits.
He may do better than Romney with less-educated white voters, even as he gets trounced among nonwhites who likely will account for a slightly larger slice of the electorate than in 2012. Thus the Republican nominee can ill afford to lose college-educated voters, many of whom seem turned off by his invective and negativity.
• Candidates: If on Oct. 20 the focus of the race is on Trump, he's a goner; the same is probably true of Clinton. Rarely have there been two presidential candidates as unpopular. So the trick for both is to keep the attention on the opponent. The candidates' schedule also may be revealing. If either is campaigning in states they're expected to win -- say Clinton in Michigan or Trump in Missouri -- it means their base isn't secure.
What's equally crucial, though harder to assess, is how energized this core support is. Will evangelicals deliver for a Republican candidate who before this election showed no cultural or political affinity with them? Will blacks and Hispanics turn out for Clinton as they did for Obama? She's counting on Trump to deliver them to her column. Young people will go decisively Democratic -- Trump is anathema to them -- but will they turn out to vote?
"With such a large pool of ambivalent or unenthusiastic voters, gauging turnout will be particularly problematic," says Charles Cook, a leading expert on U.S. elections. He projects a high turnout because "the intensity of the hatred" for one or the other candidate "is so high among so many that the option of refraining from voting at all seems abhorrent."
• October surprise: Every presidential election is rife with suggestions of a last-minute development that will shake up the race. It almost never occurs. Yet both sides are preparing for a possible terrorist attack, a health problem or a really big, new scandal. If one could anticipate the likelihood, it wouldn't be a surprise.
Finally, be a discerning poll-watcher. Some of the better brand-name surveys are pretty reliable; be more leery of others.
If any candidate declares, "The only poll that counts is the poll on Election Day," that politician is facing defeat.