NRL News

Gallup poll a gold mine of useful information and misleading inferences

by | May 18, 2016

By Dave Andrusko

gallup779As is always the case with Gallup and abortion, the latest post from veteran Gallup writer Lydia Saad is informative. In places it also makes you scratch your head and occasionally it’s deeply misleading. (The figures come from Gallup’s 2016 Values and Beliefs poll, conducted May 4-8.)

The headline is understandable, given that Donald Trump’s views on abortion are a work in progress: “Most ‘Pro-Life’ Americans Unsure About Trump’s Abortion Views.”

But what is far more interesting is that after decades of unabashed, straight-forward, unambiguous promotion of abortion, Hillary Clinton’s views are unknown, or at least unclear, to many pro-choice adults.

Let’s compare the responses to the presumptive nominees of the Republican and Democratic parties, respectively. Saad writes

Sixty-three percent of Americans who describe themselves as “pro-life” are unable to say whether they agree or disagree with Trump on the abortion issue. The rest are about equally divided between agreeing (19%) and disagreeing (18%) with him.


Hillary Clinton is unambiguously pro-abortion rights and has talked about her strong support for women’s reproductive rights for decades. Nevertheless, the public is only a bit less vague about her stance on the issue compared with Trump’s. Overall, 22% of Americans say they agree with her views and 32% disagree, while 46% are unsure.

Notice she doesn’t identify specifically how “pro-choice” Americans feel about Clinton (as she did with pro-lifers and Trump), which is far more intriguing in light of her vocal support for abortion for decades.

Later we learn that 15% of pro-choice adults disagree with Clinton on abortion as compared to 38% who agree.

Also 47% of pro-choicers have “no opinion,” presumably how Gallup categorizes those who answer they don’t “know enough to say.”

Let’s think about all this, and more. Stay with me, it’s complicated but worth your time.

Once again Saad underestimates the importance of the single-issue voter. She writes

One in five say they will only vote for a candidate who shares their views on abortion. Another 49% say it is one of many important factors they consider when voting, while 28% say it is not a major issue for them.

As in the past, slightly more in the pro-life than in the pro-choice camp — 23% vs. 17%, respectively — say that a candidate must share their views on abortion to win their support.

First, what is the “past” Saad refers to? It’s a 2015 Gallup poll about which Rebecca Riffkin wrote

The percentage of Americans who say they would only vote for a candidate who shares their views on abortion has been edging up over the past seven years. The 21% who currently say this is, by one percentage point, the highest Gallup has found in its 19-year history of asking the question.

In other words, this year’s 20% is just one point under the highest figure ever recorded by Gallup.

Second, in 2015, “23% of pro-life Americans sa[id] they will only vote for a candidate who shares their views, compared with 19% of pro-choice American,” according to Riffkin–a four point pro-life advantage.

How about 2016? The pro-life advantage is 6 points (23% to 17%). Good–but unmentioned–news.

This is even trickier.

In 2014, addressing the same issue, Saad went into a long discussion of how there were more “pro-choice registered voters,” a dimension I don’t see here.

But as we explained back then, when you factored in all the variables, it netted out to be a +3% pro-life advantage among the population taken as a whole. A 3% advantage can be crucial in close elections and is nothing to be dismissed.

Finally, there was a lot of huffing and puffing back in 2015 when 50% self-identified as “pro-choice” to 44% pro-life. In 2016 the difference is not six points but one: 47% pro-choice to 46% pro-life.

Saad describes this as “a more even division.”

The Gallup poll is fascinating, almost as fascinating as what Saad does with it

Categories: Polling