How Not to Read the Abortion Polls
earlier today that a
new post in a section of the Washington Post he oversees explained why “Republicans need to stop talking about abortion. Immediately.” What it really shows is that polling on abortion needs to be read carefully rather than excitably.
In that post Aaron Blake writes, “The trend line is clear, and Americans are becoming more accepting of abortion rights.” His thesis depends entirely on an overreaction to a few bits of poll data. A fuller look at the evidence does not bear it out.
First, let’s look at the polls on Roe v. Wade. Blake’s major piece of evidence here is a
new NBC/Wall Street Journal poll that shows that 70 percent of Americans believe that Roe should be upheld rather than overturned. He also cites a new Pew poll that he claims “largely confirms those findings, showing 63 percent of Americans oppose overturning Roe.”
Actually, Pew did not find that support for Roe has been increasing. It found less support for Roe than it did in 2005, which appears to be the last time it asked the question. The ABC/Washington Post poll also found declining support for Roe between 2005 and 2010.
Public support for Roe is not new*–and it is increasing only in one pollster’s trendline. Even in that trendline, fluctuations have occurred in the past without signaling much of anything. So, yes, the don’t-overturn-it view has gone up 4 points in the NBC/WSJ poll between December 2005 and now; it fell the same amount between 1989 and 1991.
Blake’s next piece of evidence is that the NBC/WSJ poll “shows for the first time that a majority of Americans (54 percent) support abortion rights.” That’s true of the NBC/Wall Street Journal poll. But plenty of other polls over the years have found that most Americans believe abortion should be legal in “all” or “most cases.”
Other polling does not find any leftward shift. The University of Michigan’s polling finds no clear change from 1990 through 2008. The CBS/New York Times poll shows no movement between 2003 and 2012.
Gallup shows no clear change in either direction from 2002 to 2012. (It also finds no pro-choice majority: In May of 2012, 59 percent of respondents told Gallup abortion should be legal in a few circumstances or illegal in all circumstances, while 38 percent said it should be legal in “all” or “most” circumstances.) Harris’s numbers show a movement in the pro-life direction from 1993 to 2009 on the question of under what circumstances abortion should be legal.
Maybe it will turn out that the public is becoming more supportive of abortion. I’d wait to see more evidence before calling that trend, which may not exist at all, “clear.”
Pro-lifers have been able to succeed politically even while Roe has been popular and majorities have given pro-choice answers in many abortion polls (while giving answers more congenial to pro-lifers in other abortion polls). Blake’s evidence doesn’t offer much reason to think that they will no longer be able to do so.
He leaves the data altogether to opine about the political dynamics of abortion—which is perfectly fine; I did the same thing in a recent issue of NR. The problem is that what Blake writes makes no sense. He writes that “talking about the issue opens the door for the party’s more conservative elements to do unhelpful things.” Does anyone think that Todd Akin needed Mitt Romney to “open the door” for him to talk about abortion? For that matter, how much did Romney talk about abortion?
In 2012, some Republicans said ill-considered things about abortion and others ran away from the issue. The combination was indeed harmful to the party (as I explain in that NR article). More running away is probably not the solution. Maybe Republicans should instead emulate reasonably successful pro-life politicians, who have managed to both talk about the issue on occasion and not say anything disastrous.
- Especially when the pollsters include misinformation in their questions about Roe, as both the Pew and NBC/WSJ polls do. They suggest falsely that Roe limits the abortion license to the first three months of pregnancy. (The combined effect of Roe and its companion case Doe v. Bolton is to make abortion legal at any stage of pregnancy.) The latter poll even uses the phrase “completely overturn” in its question, a qualifier that can be expected to lower support for the option. We have no reason to think that a strong majority of the public has a clear sense of what Roe held or what overturning it would do. In my own book on these issues, I argue that what the Roe polls are probably picking up is that a strong majority of the public does not favor a ban on all first-trimester abortions.
nationalreview.com/corner/338358/how-not-read-abortion-polls-ramesh-ponnuru
So I vow not to use the line just for the rest of today. 
Thanks for the laugh.