Throughout the primaries, Donald Trump regularly boasted that the polls showed him with a substantial lead over his rivals. And the polls that didn’t? Well, those he brushed off as inaccurate. Meanwhile, Hillary Clinton, who has already been through a couple presidential elections, says it’s way too early to read anything into the polls.
So as we edge closer to the general election, Trump will most like face Clinton in the general election. Maybe. That is, unless delegates at the Democratic Convention heed the latest polls, showing that Hillary Clinton would lose to Donald Trump by an average of .2 percent while Sanders could beat Trump (by an average of 10.8 percent). It just may be the Democratic Convention that becomes contested.
On the other hand, Moody’s Analytics released their election model just a couple of days earlier, suggesting a very different outcome. In their report, based on two year’s worth of economic statistics, the Dems have it. Moody’s, which has successfully predicted the next president since the 1980 election, said that current stats coupled with the incumbent president’s favorability rating, basically predicts whether the same Party will cruise to victory or be overtaken. No matter who that Party’s nominee is.
So what can we really take from the latest polls? Well, give or take the margin of error, the biggest take-away is this – both of the presumptive candidates are considered to be unfavorable choices. In fact, 54 percent of those surveyed in an NBC News/Wall Street Journal poll viewed Clinton unfavorably and 58 percent were put off by Trump.
But who is the least unfavorable? Well, although the momentum may be with the Trump campaign, the pair is now statistically in a dead heat. It’s just too close to call.
What does that really mean? Can we trust these numbers? Or are they just headline-catchers, designed to draw us in? It depends on the poll.
If you’re like a lot of Americans you have your own opinion of the candidates and whom you’d most want to win. (Note, we say most want – not necessarily actually want!) No matter how vocally you share your opinions with friends and family would you answer the phone and talk to the pollsters? Probably not – and the number seems to declining by the day! Back in the 1990s (still in the heyday of landlines), political telemarketers could expect a 40 percent response rate. Today? The number has dropped to just 9 percent. Maybe it’s due to landlines going the way of the dinosaur. Or maybe the public just doesn’t have a strong enough preference between the candidates – so why bother to talk about it?
The question that arises, then, is this: with so many people declining to respond, do the opinions of those that do answer, represent a valid reflection of the general public? According to the Washington Post, they do. Recent research has shown that the response rate doesn’t seem to affect accuracy. At least in the general election.
But in a polling analysis by political pundit Nate Silver, presidential primaries seem to have a much worse record of accuracy – with an average error rate of 7.7 percentage points – than polls associated with the general election. One reason for the huge gap may be that primaries tend to have a lower turnout and they attract more swing voters, swaying the results. Or conversely, they may attract people who will vote the Party line no matter what. You might like Bernie Sanders way more than Hillary Clinton, for instance, but will still vote for any Democrat rather than any Republican. Or vice versa. And those primary polls, of course, are the ones we’re still hearing about.
So with new polls coming out practically every day, what can we believe? Nothing is certain. But if you feel strongly about who you want for our next president, don’t take any candidate’s leading poll numbers for granted. Get out and vote.
