Americans' Views on Supreme Court Tenure, Abortion, and Gay Marriage:
Why Recent Polling Results Are Open to Different Interpretations

What do mandatory retirement ages, abortion rights, and gay marriage have in common? These controversial legal issues were all subjects of a recent poll conducted for the Associated Press by Ipsos-Public Affairs. The poll asked 1,000 adults from all states except Alaska and Hawaii a series of questions about high-profile legal topics.

The poll found that a majority of respondents favored instituting an unspecified retirement age for Supreme Court justices. A majority of respondents wanted President Bush to nominate justices who would uphold the Supreme Court's key abortion decision, Roe v. Wade. And a majority of respondents opposed legalizing gay marriage.

Should we conclude, then, that the American people are radicals when it comes to judicial appointments, liberals when it comes to abortion, and conservatives when it comes to gay rights? Not necessarily.

As I will explain in this column, a closer look at the poll reveals that it's impossible to say why the respondents voted the way they did - or whether they would change their minds if confronted with arguments against their positions, or if asked questions with different emphases.

Thus, although it may be tempting to brandish poll results as evidence of what Americans really think, the real lesson of this poll might be that polls are of only limited value.

The Supreme Court Life Tenure Issue: Interpreting the Poll Results

Let's start with Americans' views on Supreme Court tenure. To begin, it turns out the respondents were not exactly legal eagles. Only 41% correctly identified William Rehnquist as the Supreme Court's Chief Justice - mirroring the finding of a 2003 FindLaw poll that 63% of American adults could not recall the names of any Supreme Court justices.

Respondents, then, were ill-informed even about the Court's composition. It's therefore reasonable to assume that they were also probably ill-informed about the arguments for, and against, life tenure. That raises an important question: If they were made familiar with these arguments, would they change their opinions?

Certainly, strong arguments have been made recently for instituting limited tenure (instead of life terms) for Supreme Court justices - a move that would require a Constitutional amendment. A June 2004 Note in the Virginia Law Review supported a single 18-year term. And recently, an op-ed by Norman Ornstein of the American Enterprise Institute advocated a 15-year term as a way to break Congressional standoffs over judicial confirmations.

Supporters of limited tenure have pointed out that it would decrease the incentives for strategic retirements - that is, retirements timed so that a successor will probably share the retiring justice's political views. Relatedly, the probability of justices' suffering diminished mental capacities while still on the Court - which has happened in the past - would be greatly reduced by limited-term appointments, for justices could no longer hang on through numerous Administrations to make sure they can retire at a strategic time.

Finally, imposing limited terms would reduce the temptation for Presidents to appoint young ideologues. If terms are only fifteen years long, actuarial statistics won't dramatically favor appointing, say, a forty-year-old over a fifty-five year old. And the political cost of appointing an ideologue may outweigh the benefit from having the ideologue on the Court for only fifteen years, not forty.

But were the respondents familiar with these arguments? And even if they were, which, exactly, persuaded them? The arguments, it's important to note, cut different ways. For every person who supports limited tenure to weed out young ideologues, another may support it out of a belief that older justices who have remained on the Court for decades tend to have stale or simply incorrect views. (Put more concretely, among those who actually know the names of the justices, some may be thinking "No more Clarence Thomases!" while others may be thinking "Justice Stevens: Retire, already!")

More to the point, were respondents familiar with the arguments against limiting judicial tenure? There's no reason to think so - and from the polls demonstrating ignorance even as to who is on the Court, good reason to think otherwise. And it turns out there are also strong arguments against the constitutional amendment that would be required.

First, consider that current justices usually keep their positions on the Court well into old age: They are usually the last jobs they will ever have. And that's a very good thing: Justices can completely forget about future job opportunities, including political offices or political appointments - such as, say, to an Attorney General position. So justices do not have an incentive to try to court favor with any person, or any segment of society; they are incentivized to be just what we want them to be: unbiased decisionmakers who look to the long-term - which, after all, is the time frame to which the Constitution itself looks.

Second, strategic thinking won't just go away by instituting a rule that supplies justices with retirement dates. Let's assume that strategic retirements are indeed reduced (and would they be? Justices could always retire early). Even then, having future vacancies become entirely predictable will lead to another and perhaps even uglier kind of strategizing by justices. They may not only be thinking about when to leave, but also about how to vote.

Consider, hypothetically, a politically-minded justice who, in 2000, is deciding how to vote in Bush v. Gore and who knows he (or another like-minded justice) will be forced to retire in 2002. Voting to stop the recount - and effectively make Bush President - will also mean, to that justice, paving the way for a conservative replacement. If the justice is liberal, he won't like that; if he's conservative, he will. Will the justice's decision depend wholly on the law and not on Court politics?

I doubt it. It's one thing to speculate that a given president will be able to nominate several justices, and another to know exactly how many positions he'll fill, and when they will open up. Could a justice really separate from his decisionmaking process the knowledge that, if he votes one way, he'd be replaced by a successor whose views are anathema to his own, and who will try to reverse all the precedents he has labored over?

In sum, there are strong arguments both for and against limited Supreme Court tenure. And there is no reason to think the public is familiar with any of these arguments. But their representatives will be - or ought to be - before they vote on the Constitutional amendment, if one is proposed. And by then, the public, too, ought to be far more informed.

But at this stage, all the poll's support for limited tenure really suggests is that the debate is worth having - not which arguments will ultimately persuade the public.

The Other Polling Results: Equally Ambiguous on Gay Rights and on Abortion

Meanwhile, the poll numbers on abortion and gay rights might not be much more informative than those on tenure limits.

Perhaps, as the AP-Ipsos poll indicates, Americans really are liberal when it comes to abortion rights but conservative when it comes to gay rights. Indeed, polls consistently show support for upholding Roe, and citizens in several states registered their disapproval of gay marriage through ballot initiatives in November.

But how much can the AP-Ipsos poll really tell us? To see how little they actually prove, it's important to note, initially, that the poll's abortion and gay rights questions each asked for different kinds of information.

The abortion question sought views about a particular Supreme Court precedent. Thus, the respondents who favored upholding Roe might in fact disfavor abortions in general - but might support the stability of judicial precedent. Or, such respondents might support Roe's core holding but also oppose, for instance, so-called "partial birth" abortions.

In contrast, the gay marriage question sought personal views not tied to a particular case, court, or law. And that raises different ambiguities: Respondents might, if asked, have other views on the various legal provisions and decisions pertinent to that issue.

For example, the Massachusetts Supreme Court has ruled, in Goodridge v. Dep't of Public Health, that the Equal Protection clause of the Massachusetts Constitution forbids denying marriage rights to gay and lesbian couples. The Bush administration, however, supports a Constitutional amendment outlawing gay marriage regardless of public opinion or constitutional provisions in particular states.

These subtleties mean every answer only spawns additional questions: Does an anti-gay marriage respondent also oppose the Massachusetts decision - or as a supporter of federalism, believe the question is up to each state to decide for itself? Would that respondent oppose civil unions? Would he or she support Bush's constitutional amendment? We simply don't know.

Poll Interpretation: Why It's Difficult Even Given Ostensibly Neutral Questions

My point here isn't that these are "push polls" designed to evoke specific responses. (Here's an example of a "push poll" question: "If you knew that candidate X was a murderer, would you vote for him?"). Rather, my point is that polls need not be "push polls" to be ambiguous in crucial ways - and to be open to interpretation and misinterpretation.

Indeed, the AP-Ipsos poll shows that even ostensibly neutral questions can influence or distort public opinion. The Roe question must have pleased Roe's defenders because it told the respondents that Roe was decided in 1973, thus emphasizing that overruling it would erase over 30 years of Supreme Court precedents and a right that has become somewhat entrenched. And the gay marriage question must have pleased supporters of the Bush administration's proposed amendment - for it asked respondents whether they supported allowing gays to marry, without asking whether they believed it necessary to amend the Constitution to impose such a rule even on states that oppose it.

Unfortunately, reports of poll results rarely are accompanied by sufficient disclaimers. Beyond the question of what a poll's margin of error may be (given its sample size), there is always a deeper question: What do the poll results really mean, given the exact questions that were asked?

As long as we fail to ask these deeper questions, questions like those in the AP-Ipsos poll - once publicized nationwide - can do as much to influence opinion as to discern it. Politicians and pundits may shift or explain their views based on what they think the public believes, when in fact the public believes something completely different. And some Americans may be told incorrectly that their views are scorned by their fellow citizens, while others are incorrectly led to believe that their views are widely shared.

In the end, polls work well as tests: We can be fairly sure that the respondents weren't kidding when they said they couldn't recognize William Rehnquist. Yet they are rather incomplete - and therefore, potentially dangerous - barometers of public opinion.

For news organizations, polls are easy to report. And for news junkies, they are interesting to discuss. But for people serious about important legal issues, they can only mark the beginning of the debate - not any kind of summary of its outcome.