The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Harvard / Harris Poll on Abortion

Interesting results about public attitudes, including the complicated gender gap.


From a survey taken June 28-29 by the Center for American Political Studies at Harvard University and The Harris Poll:

[1.] Abortion rights are listed as the greatest concern by 14% of the respondents and as the second greatest by 11%, far behind inflation (total 62%) but at roughly the same level as energy prices, as crime, and as immigration.

[2.] Dobbs appears to have no visible net effect on the midterms, with 36% saying it makes them more likely to vote Democrat and 36% more likely to vote Republican.

[3.] Dobbs is opposed by a slim majority, 55%-45%; the question wording is, "Do you support or oppose the Supreme Court's decision to overturn Roe vs Wade, which allows each state to decide its own standards for abortion instead of a set right?" At the same time, when asked, "Do you think it is better for abortion standards to be set by judges of the supreme court, by a vote of congress or by the legislatures of each individual state?," only 25% go say "the supreme court," 31% say "Congress," and 44% say "state legislatures."

[4.] The results as to "Do you think your state should allow abortion…?" were

  • Up to 9 months [10%]
  • Up to 23 weeks [18%]
  • Up to 15 weeks [23%]
  • Up to six weeks [12%]
  • Only in cases of rape and incest [37%]

The strong national majority thus appears to be both pretty far removed from Roe and Casey (which would have allowed abortions until about 23 weeks or later), which only get 28% support (including the up-to-9-months respondents), and from near-total bans, which only get 37% support. The 15-week mark people seem to be the swing voters.

[5.] There appears to be something of a gender gap on abortion, but it's complicated. Women oppose Dobbs by 61%-39% and men support it by 52%-48%, so that's a significant though not vast difference, a swing of about 13% (I expect the margin of error for each sex category is about 4%, based on a total respondent pool of 1308). But on the other hand, when asked,

Do you think your state should allow abortion

  • Up to 9 months
  • Up to 23 weeks
  • Up to 15 weeks
  • Up to six weeks
  • Only in cases of rape and incest

women respondents preferred the three most restrictive options by 75%-25%, while men preferred them only by 69%-31%. (The published data doesn't break down each individual answer by sex, perhaps because at that level the margins of error would be too large.)

Naturally, this has all the limitations of polls generally, and in particular it doesn't tell us much about state-level data, or about how people would respond to particular abortion laws, such as ones that only have an exception for the life of the mother but not for maternal health or for rape or incest. And of course this doesn't tell us what the right answer is, either as to whether abortion should be banned as a substantive matter, or whether Casey should have been overturned as a constitutional matter (including as a matter of constitutional stare decisis). Still, I thought it was worth noting, especially since one of the main questions going forward is about what can be done through the political process, and how it can be done.

As a personal matter (and noting that these questions aren't within my core area of expertise), I generally think that Roe and Casey are hard to justify as a  matter of constitutional law first principles (including the precedents as of the time Roe was decided); I'm not sure about the question whether they should still have been preserved as a matter of stare decisis; and I generally support abortion rights as a policy matter, which is one reason I'm particularly interested in seeing what paths there are towards state statutes and voter initiatives that protect abortion rights.

But I also think that there's a vast gulf (from the perspective of both the policy arguments for allowing abortion and against) between "abortions allowed up to 15 weeks" and "abortions generally banned, except as to life or physical health of the mother, and perhaps as to rape or incest"—far smaller than the gulf between "abortions allowed up to 15 weeks" and "abortions allowed up to 23 weeks." If this data suggests that 15 weeks would be a politically workable compromise in many (though likely not all) states, that strikes me as important information.