A reality check about the presidential election, as we approach the political conventions: #Hillary Clinton, by any objective standard, is winning. You may have heard that the race is “even” or a “toss up” or that it’s “up for grabs” or “basically tied” or “could go either way.” And if you go by the daily noise of the campaign, or selective claims by those supportive of Trump, it may appear that way, but it’s not.
Which polls to believe.
If the 2008 and 2012 elections taught us anything, it’s that you really shouldn’t take too much stock in one individual poll on one individual day. There are many, many national presidential polls each week and, as the calendar moves closer to November, each day. Individual polls on individual days may show Trump ahead, just as outlier polls showed McCain ahead and Romney ahead in the last two elections. But that’s not what you should be looking at.
Real clear politics.
There’s a relatively well-known political website called Real Clear Politics. Each day the site publishes the most recent of each of the eight national presidential polls, and then it publishes an average of those polls. Then, it publishes a chart of where that average has been throughout the campaign.
Take a look at RCP as of Sunday night. It has Clinton leading the eight polls by an average of 3.2 points. Of the eight individual polls, six list Clinton leading. One lists a tie, and one (the notoriously Republican-leaning Rasmussen) has Trump leading. The polls list Clinton’s lead as everywhere from 11 points (Reuters/Ipsos) to 2 (Economist/YouGov.) But more important is this: Take a look at the RCP average, from July 2 to today. Aside from one day last January when it was tied and one day in May when Trump pulled ahead by two-tenths of a percentage point, Hillary Clinton has led the average every day of the last year.
Google the phrase “shock poll” and see what comes up. Most of the results are individual polls on individual days in the past few months that have shown Trump ahead. This is instructive for a couple of reasons: Because it’s truly shocking, even to pro-Trump websites, for their candidate to be leading even for one day.
But it’s also a familiar trick from the last two campaigns. In the waning days of the 2008 and 2012 campaigns, when the majority of polls showed Barack Obama ahead, Matt Drudge made a daily habit of picking the poll each day most favorable to McCain or Romney, in order to give his Republican readers hope that maybe their guy was winning after all. No, it never mattered if the most pro-McCain/Romney poll today was the same one as yesterday. Yes, this contributed to Republican wish fulfillment, but that's not all. This tactic also gave mainstream news- which is heavily invested in the appearance of a close, competitive race- excuses to create that impression.
So are the polls “biased”? In a word, no. For a simple reason: Pollsters’ greatest professional incentive isn’t to help candidates they like- it’s to be accurate and right. If pollsters biased against Trump stated Clinton was winning, and Trump won, why would that be good for the pollsters?
Could #Donald Trump still win? He could. He could have a better convention than the Democrats, or the race could turn some other way; it's already been the craziest election in history so nothing would be a surprise. But as of now, it’s clear: Hillary Clinton is winning, and has been all along. #Election 2016