 |
| PHOTOGRAPH BY ANDREW HARNIK / AP |
NEW YORKER, JOHN CASSIDY [Condensed, with add'l photos]
With nine weeks left until the general election, most signs point to a victory for Hillary Clinton. In head-to-head national polling she has been ahead of Donald Trump for much of the past year, and in most recent surveys she has retained the lead. At the state level, too, Clinton holds the advantage: over the summer she moved ahead of Trump in many key battleground states, greatly complicating his path to accumulate two hundred and seventy votes in the Electoral College.
Of course, polls aren’t infallible—we relearned that lesson in the recent Brexit referendum. But in the past eight U.S. elections, the candidate who was leading on Labor Day went on to become President. At the online bookmakers, where real money is wagered, Clinton remains the strong favorite. On Tuesday thepolls-based forecasting model maintained by the Times Upshot team estimated the probability of her winning at eighty-four per cent. FiveThirtyEight’s “polls-only” model put the probability of a Clinton victory at 68.5 per cent.
Why, then, is there so much anxiety among Clinton supporters? One reason is a new CNN poll, released on Monday, which generated headlines saying Trump had taken the lead. Among respondents the pollsters deemed “likely voters” that was true: the poll showed Trump at forty-four per cent, and Clinton at forty-two per cent. But among the broader pool of registered voters, Clinton was still ahead, forty-four per cent to forty-one per cent.
Essentially, the pollsters screened out some of Clinton’s supporters because they didn’t adjudge them likely to turn out on November 8th. There’s nothing unusual or untoward in that. As elections approach, many polling organizations switch their focus from registered voters to likely voters. But all such screens are somewhat arbitrary, because each polling organization has its own criteria for screening out unlikely voters.
When inspecting the trend in a given poll, it is also better to compare like with like. In this case, we can still look at registered voters: among those voters, Clinton’s lead in the CNN poll has shrunk from eight points a month ago to three points now. That finding is in line with other recent polls. On August 10th the Real Clear Politics poll average, which combines the results of numerous surveys, showed Clinton leading Trump 47.8 per cent to 39.9 per cent, a gap of almost eight percentage points. By Tuesday morning the gap had narrowed to 3.3 percentage points. (Clinton: 46.2 per cent; Trump: 42.9 per cent.)...[But] The Huffington Post’s poll average, which is somewhat different than the Real Clear Politics average, indicates that Clinton is still leading by more than five percentage points—48.1 per cent to 42.5 per cent. But it, too, indicates that the race has narrowed over the past month.
Why has it narrowed? Part of the explanation may be that Clinton’s recent slippage reflects a predictable correction to the polling gains she enjoyed after a successful Democratic Convention and Trump’s self-defeating attacks on the Khan family. As memories of Conventions fade, the bounce that candidates get from them often subsides, partially or wholly.
In this case, it could be argued that the polls have largely reverted to where they were before the Conventions. On July 11th, a week before the gavel came down in Cleveland, the R.C.P. poll average showed Clinton leading Trump by 4.5 percentage points—not far from where the gap is today. Looking at the numbers this way, the key fact about modern American politics is that the country is divided pretty evenly, with this year being no exception. We should always have expected a close race.
A more pointed explanation for the recent polling trends acknowledges the possible effect of the recent barrage of negative press about the Clinton Foundation and Clinton’s private e-mail server. In an ABC News/WashingtonPost poll, which was carried out on August 24th through 28th, fifty-six per cent of respondents said they had an unfavorable opinion of Clinton. That was a jump of six points compared with a poll taken at the start of August. The Post’s Aaron Blake noted that Clinton’s favorable/unfavorable numbers were the worst she “has had in her quarter-century of national public life.”
But it needs putting in perspective. To come out on top in November, Clinton doesn’t need to transform herself into a beloved leader. She just needs to defeat Trump, who, by most measures, is even more unpopular than she is. Here again, the poll averages provide more reliable information than individual surveys. According to the Huffington Post’s poll average, Trump’s net favorability rating—that is, his favorable rating minus his unfavorable rating—is minus nineteen. Clinton’s figure is minus 14.6.
Ultimately, of course, the race will come down not to polls but to the Electoral College and the outcome of voting in the battleground states. For these purposes, I’ll put eleven states in the battleground category: Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, Virginia, and Wisconsin. According to the Real Clear Politics polling database, Trump is narrowly ahead in only two of these states: Iowa and Missouri. Clinton is up by five points or more in six: Colorado, Michigan, New Hampshire, Pennsylvania, Virginia, and Wisconsin. In the three remaining states—Florida, Ohio, and North Carolina—Clinton holds narrow leads.
Mimicking the national trends, polls published last week showed the race tightening in Pennsylvania, Virginia, and Wisconsin. But Clinton was still ahead of Trump in these surveys, and her unfavorability ratings were better than his. In Pennsylvania, for instance, which is shaping up as a must-win state for the Republicans, a poll from Franklin & Marshall College showed Clinton leading Trump by seven percentage points: forty-seven per cent to forty per cent. Clinton’s net favorability rating was minus sixteen, which is pretty bad. But Trump’s net favorability rating was minus twenty-one.
t should be noted that numbers like these have seldom, if ever, been seen before in a U.S. Presidential election. To many Americans, the election has come down to a choice between the unpalatable and the unthinkable. But in this strange and dystopian contest, Clinton retains a distinct advantage.