Picking Apart the Polls

Is opinion really shifting on abortion?

If you’re an avid news reader, there’s a good chance that sometime in the last day or so you’ve come across a headline trumpeting an apparent decline in public support for abortion rights. But depending on the source you were reading—and how close you were paying attention—you may not have gotten the full picture.

The source of the claim that opinion is shifting is a poll conducted by the Pew Research Center for the People & the Press, which does, indeed, find a substantial drop in people who say abortion should be legal in most or all cases. As with all poll results, though, there are a few questions to ask about this one, including: was the poll methodologically sound? And how does it compare to findings from other organizations? In other words, is this result part of a trend, or is it an outlier?

The first of those questions requires technical expertise that it’s not reasonable to expect most reporters (including this one) to possess, or most readers to sift through. The second, though, can be answered with a quick Web search, and it’s an important part of how any poll is presented. In this case, as people who follow poll results know, public opinion on abortion has historically been remarkably stable. That very stability is what makes an apparent shift newsworthy, but it raises the burden of proof considerably, and it is itself an essential part of the story.

So how has the media handled this story? It depends on which media you’re reading. A brief Reuters story available through a number of outlets begins with the declaration that “support for abortion rights has slipped in America this year,” and doesn’t examine whether other polls have reached different conclusions. A blog post at USA Today’s site takes the same approach: the Pew poll stands on its own, with no connection to other findings.

The same can’t be said of this post at The Christian Science Monitor’s “The Vote Blog.” After summarizing the findings, that post declares that “the Pew data tracks with trends found by other pollsters.” That sounds, at first, like important context. But the post cites only one other pollster with similar data—Gallup, which in May found for the first time that more Americans identified as “pro-life” than “pro-choice.” That survey appeared about the same time as earlier Pew survey which also found a conservative shift—and both encountered skepticism, partly on methodological grounds and partly because they were at odds with the bulk of recent research. So citing that Gallup study as confirmation of this Pew result doesn’t tell us much that we didn’t already know—especially since a subsequent Gallup poll in August found the trend moving back in the other direction.

So did anyone do better? Yes, actually. The lede and headline of the New York Times’s story don’t differ much from Reuters’, but the piece quickly starts throwing in the caveats, noting that “the apparent shift… contradicts some other recent polls.” Later, the article notes that “polls conducted by some other organizations within the last few months have found opinion on abortion to be more stable,” and adds that the inconsistency “makes it hard to draw a firm conclusion about whether attitudes are shifting.” Another interesting nugget: Pew noted that the movement it found coincided with Democratic control of the White House; one counterintuitive reading is that abortion opinion may shift against the prevailing political winds. But, the Times notes, no such shift happened when Bill Clinton moved into the Oval Office.

With so many caveats needed, it’s reasonable to ask whether the Pew poll warrants a story at all. Indeed, a search of The Washington Post’s Web site doesn’t turn up an article on the topic. The Post’s “Behind the Numbers” blog, though, flagged the findings while injecting the appropriate caution: “other polls, including the Washington Post-ABC News poll, have not picked up such a basic reorientation on this divisive issue.” That’s not just boosterism for the in-house product: “National polls by the AP, CBS News and the New York Times and Quinnipiac University also show no big increase in opposition to legal abortion,” Post blogger/pollster John Cohen continues.

That’s not to say the Pew results should be dismissed—as Cohen acknowledges, they may be catching early signs that “opinions thought settled for a decade are now in flux.” At the very least, it’s worth keeping a closer eye on the next round of data from other sources. But until we see more, what we have is suggestive, not conclusive—and stories that don’t make that distinction are letting their readers down.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Greg Marx is an associate editor at CJR. Follow him on Twitter @gregamarx.