Polls

Don't Believe Any Headline Showing Hillary Clinton with a 12-Point Lead over Donald Trump

Journalists are hyping poll results that look quite different when Gary Johnson is properly included

|

"A new poll shows Hillary Clinton with a 12 point lead over Donald Trump," trumpets ABC News Radio, hyping a broadcast of This Week with George Stephanopoulos. Well, yes, but mostly no.

There is indeed a 51 percent to 39 percent advantage for Clinton over Trump in newly released Washington Post/ABC News poll, conducted from June 20-23. But that same survey also asked the same pool of voters to react to a far more representative ballot, i.e., one that includes Libertarian Party nominee Gary Johnson (who is widely expected to end up on the ballots of all 50 states), and the Green Party's Jill Stein, who is projected to wind up on as many as 47. Leaving Johnson and Stein off polls is bad science; electing to emphasize the numbers that exclude them is journalistic malpractice.

How do the numbers differ with or without the third-party candidates included? Here's the WashPost/ABC poll, with the first line the misleading Clinton-vs.-Trump matchup, and the second line better reflecting reality. OT = "other," NV = "not voting."

HC 51% DT 39% OT 2% NV 6%

HC 47% DT 37% OT 1% NV 0% GJ 7% JS 3%

So it's a 10-point lead, not a 12-point lead, and Clinton is not supported by a majority of voters. (In fact, this marks her high-water mark in any poll this season that has included Johnson and Stein; in all others, she has been between 39 percent and 44 percent.)

The split is even more dramatic with the other poll making news today, from NBC News and the Wall Street Journal:

HC 46% DT 41%

HC 39% DT 38 GJ% 10 JS 6%

And yet here is how numbers wizard Nate Silver reacted to this poll:

That quip would read a lot differently if it were "The bad news for Trump is that a poll showing him 1 point down is considered good news for Trump," and yet the edited version is the more accurate one. (Silver, in other venues, has argued for Johnson to be included in polling, saying "it's not a pollster's job, in my view, to take that choice away from the voter when they'll have it on the ballot. They can always ask the question both ways, too — with Johnson and without.") 

Such selective, inaccurate reading is not just the stuff of Twitter jokes. This is how the Wall Street Journal headlined its own damned poll: "Hillary Clinton Holds 5-Point Lead Over Donald Trump, Latest Poll Finds." Right, except when you include the names of the candidates who will actually be on the ballots. ABC News did the same with its survey: "Clinton Opens 12-Point Lead on Trump."

Well, just because some news organizations are elevating clickbait over accuracy doesn't mean the rest of us need play along. There have been, to my count, seven national polls over the last month that asked about all four candidates, and also provided results for just Hillary Clinton and Donald Trump. Before you read or share any headline about presidential polls, make sure you check to see whether they've included Johnson and/or Stein, and then keep this month-long average in mind:

HC 46% DT 38% OT 7% NV 5% 

HC 42% DT 36% OT 2% NV 2% GJ 7% JS 5%

Notice the interesting symmetry in those numbers? Without the third-party candidates, "other" and "not voting" have 7 percent and 5 percent, respectively. Include them, and its Johnson and Stein who get 7 percent and 5 percent, while the number of voters feeling left out of the process plummets. And yes, the third-party challengers on net are drawing more from the Democrat.

The other thing that jumps out of the four-way polling is this: Clinton has yet to reach 50 percent when her proper competition is included, and Trump hasn't even cracked 40. It's early yet, and third-party support historically dwindles toward Election Day, but preliminarily the numbers illustrate what our gut already tells us: America is not enthusiastic about its major-party presidential choices.