Abortion Sentiment Affected by Rating Scale

by David W. Moore, Senior Gallup Poll Editor

American public opinion on abortion is decidedly mixed. For almost three decades, Gallup has measured public opinion on this issue, asking whether abortions should be "legal under any circumstances, legal only under certain circumstances, or illegal in all circumstances." Throughout most of those years, a majority of Americans have declared themselves in the middle of the issue, with the rest leaning more for legality in all, rather than in no, cases. A recent Gallup survey, conducted May 2-4, finds 24% of Americans opting for unrestricted abortions, 19% for a complete ban, and 56% for legality only under certain circumstances.

Can opinion be more finely tuned than what this three-part question provides? In 1994, Gallup first asked a follow-up question to determine whether the public leaned more in favor or in opposition to abortion. Respondents who chose the middle alternative (legal under certain circumstances) were then asked whether they thought abortions should be legal "in most circumstances or only in a few circumstances." Typically, the middle group has disproportionately chosen the more restrictive response. 

Do you think abortions should be legal under any circumstances, legal only under certain circumstances, or illegal in all circumstances? (If certain circumstances) Do you think abortion should be legal in most circumstances or only in a few circumstances?



under any circum-

under most circum-

Legal only
in a few circum-

in all circum-








2004 May 2-4






The 55% who chose the middle option in the first question divided 42% for legality in only a "few" circumstances, with just 13% choosing "most" circumstances. This division suggests that a majority of Americans -- 61%  -- give at best only reluctant support for abortion, while only a minority (37%) give relatively strong support. 

In 1995, the ABC News poll began asking a single question on abortion that was modeled closely after the two Gallup questions. The original and follow-up options were combined in this format:

Do you think abortion should be legal in all cases, legal in most cases, illegal in most cases, or illegal in all cases?

The results of ABC's latest poll show virtually the same division that Gallup found -- 23% for legality in all cases, 20% for illegality in all cases. But the two middle categories (legal in most cases, illegal in most cases) show different percentages from Gallup's two middle categories. Instead of a 3-to-1 margin for the more restrictive category (illegal in most cases), ABC found most of the middle category respondents opting for abortion in "most" cases. 

The ABC results suggest that the public leans more in favor of abortion than opposition, by a 54% to 43% margin, rather than opposed by the 61% to 37% margin found by Gallup. The difference in results between the two polls have been consistent over the past decade.

While Gallup shows on average about 40% of Americans favor abortions in most or all circumstances, ABC finds on average about 55% in favor.

The consistency of the differences over the years suggests a root cause related to question wording. 

  • Gallup has a two-tiered structure, the first question a three-part response, followed by a question asking respondents in the middle group to make a distinction between legality in "most" or "some" circumstances. ABC has just the one question, giving the respondent a look at all the alternatives before making a choice.
  • Gallup's scale is different from ABC's scale. Gallup goes from legal under "any" circumstances, to legal under "most" circumstances, to legal "only in a few" circumstances, to illegal in "all circumstances." By contrast, the first two responses of the ABC question are mirror opposites of the second two responses -- legality in all or some cases versus illegality in all or some cases.

It seems unlikely that differences between the two polls is related to Gallup's two-tiered structure versus ABC's one question. The "all" or "none" responses in both polls -- despite slight question wording differences -- produce similar percentages. It's the middle categories of the two polls that are different.

When Gallup initiated the follow-up question, the intent was to preserve the trend of the original question that goes back to 1975. Rather than change that question to a four-part response, it was felt the follow-up would suffice to further divide respondents by their views on the issue. Instead of asking whether abortion "should be legal in most cases or only in a few cases," we could have asked whether abortion "should be legal in most cases or illegal in most cases." That would have made the final four-part scale virtually identical to the ABC scale and perhaps would have produced similar results. 

Which scale produces the more "valid" results?  In fact, both scales measure a slightly different aspect of public opinion, and neither scale can be demonstrated to be better or more valid than the other. Both provide some insight into what Americans are thinking on the issue.

What do these differences in poll results reveal about public opinion on abortion? People are more willing to say abortion should be legal in only a few cases than they are to say abortion should be made illegal in most cases. The latter option is a bit harsher in tone than the Gallup option, though in practice the two options appear to be similar. This type of finding -- where tone appears to have a major effect on which option respondents select -- appears in other experiments that show Americans more willing to say the law should "not allow" some action than the law should "forbid" those actions.

In a broader context, these results suggest that not only are most Americans conflicted about the issue, but that a relatively minor wording difference can push many people into a more favorable, or a more negative, expression about abortion.  


Gallup http://www.gallup.com/poll/12037/Abortion-Sentiment-Affected-Rating-Scale.aspx
Gallup World Headquarters, 901 F Street, Washington, D.C., 20001, U.S.A
+1 202.715.3030