Legitimacy in the Public Eye: Justice Sandra Day O’Connor and Public Opinion

By Thomas R. Marshall

Supreme Court Justice Sandra Day O’Connor, who has announced her retirement upon confirmation of a successor, was one of the few justices who explicitly discussed the role American public opinion should play in influencing Supreme Court decisions. In the landmark 1992 abortion case, Planned Parenthood of Southeastern Pennsylvania v. Casey, O’Connor wrote that the Court should reverse its best-known decisions only if those decisions proved unworkable and were widely rejected by public opinion. If the Court too often abandoned its own landmark decisions, she wrote, it might lose public confidence and be unable ever to recoup its legitimacy in the public eye. O’Connor also set a relatively lenient standard for introducing polls and surveys as evidence in court cases.

Several justices rejected O’Connor’s argument of the proper role of public opinion. In the Casey decision, Chief Justice Rehnquist argued that “the Court’s duty is to ignore public opinion and criticism on issues that come before it,” and further complained that Supreme Court judges “are in perhaps the worst position to judge whether a decision divides the Nation deeply.” Justice Scalia called O’Connor’s argument about public opinion “outrageous” and accused her of “almost czarist arrogance.”

As Justice O’Connor ends her tenure on the Supreme Court, two questions arise. First, for a justice who was so important in American public opinion, how closely did O’Connor’s on-the-bench votes actually represent it? Second, as the first female justice in the Court’s history, did O’Connor’s votes represent women’s attitudes more often than men’s?

We can offer answers to both these questions because, since the origins of modern polling during the 1930s, American pollsters have frequently attempted to tap attitudes on Supreme Court decisions. This is not surprising—as Alexis de Tocqueville remarked during the early 1800s, most American political disputes become lawsuits, and many of these reach the U.S. Supreme Court.

Consider the 2003 case Lawrence and Garner v. Texas, in which the Court ruled, 6-3, that Texas’s ban on consensual sexual relations between adults of the same sex was unconstitutional. A May 2003 Gallup poll reported that a 60 percent to 35 percent majority of Americans thought “homosexual relations between consenting adults should … be legal.” Accordingly, we can count the Lawrence decision as “consistent” with American public opinion.

By contrast, in FDA v. Brown & Williamson Tobacco Company in 2000, the Supreme Court ruled, 5-4, that the Food and Drug Administration could not regulate tobacco products as a drug. In May 1998, Gallup had reported that by a 51 percent to 47 percent margin, Americans favored “classifying nicotine as a drug so that it can be regulated by the same government agency that regulates prescription drugs.” The Brown & Williamson decision is counted as “inconsistent” with public opinion. In short, Supreme Court decisions that agreed with poll majorities (or occasionally, pluralities) can be counted as consistent with public opinion; those that did not are counted as inconsistent.

By this method of matching Supreme Court decisions with nationwide public opinion polls, we can analyze some ninety-nine decisions during Justice O’Connor’s tenure on the Court. Because the justices routinely record their votes individually, each one’s votes can also be compared to nationwide public opinion. For example, in the Lawrence decision, O’Connor’s vote with the Court’s 6-3 majority is counted as consistent with American public opinion. In the Brown & Williamson decision her vote with the 5-4 majority is counted as inconsistent with public opinion.

Of necessity, these tallies exclude all decisions in which the Supreme Court declined to hear a dispute. Although poll questions are available for many such denials of certiorari, the justices rarely record individual votes in these cases. Nor do the tallies below include a handful of full, written decisions in which poll results were evenly divided, or available polls yielded conflicting results, such as the 2000 Florida presidential vote recount dispute in Bush v. Gore and the school vouchers dispute in Zelman v. Simmons-Harris in 2002.