he success or failure of Harriet Miers' nomination depends on how people use her past to predict her future. This is a very iffy process, and in many ways is no more than dressed-up fortune-telling. The most obvious reason we try so hard to use Miers' past to predict her future is that other ways of doing this are even less reliable.

But there is something else about predicting judicial behavior that grabs the attention of those involved and leads them to focus on — even exaggerate — the ability to predict. It is the momentum that comes from short-term political agendas and long-term aspirations about the Supreme Court.

Even on abortion, an issue where her views are relatively clear, it is impossible to predict how Miers will rule as a justice. There is more than enough evidence that shows she is against abortion, as a general proposition. She has been for a long time, and she is not going to go through a dramatic pro-choice conversion. She has previously supported a constitutional amendment outlawing abortion and has long been active in a church that opposes the practice. President Bush has sent out indirect, but hard to miss, signals to reassure his anti-abortion supporters. Still, this bit of information on Miers says almost nothing about how she would rule in future Supreme Court cases involving right-to-life matters.

Defending Miers, White House press secretary Scott McClellan said that "being a judge is different from the role of a candidate or political office-holder."

Under the circumstances, you would not expect him to be the most objective observer of the legal process. But he's right.

People behave differently when they become justices. Of course, new justices don't jettison their former beliefs, but they have to apply them in different contexts and with different rules. Also, times change and new issues emerge. Being a justice on abortion cases differs from deciding about abortion as a private citizen or as a legislator because the court typically deals with issues in ways that are less starkly framed than the most ardent anti-abortion partisans want.

Since Roe v. Wade, the Supreme Court has never directly confronted the question of whether that decision should continue to be precedent. Typically, post-Roe cases involve government attempts to limit abortion, not to get rid of it entirely. Anti-abortion groups have not been able to get the majority of the justices to use one of these cases to overrule Roe. Even conservative justices who have not shown much sympathy toward abortion have been unwilling to go that far, mainly because they felt that Roe had become established law and because the facts of the case allowed the justices to make abortion-limiting decisions without confronting the more basic and highly polarizing up-or-down alternatives.

There is no way to predict whether Miers will follow the path of these cautious conservatives or the more radical path of overruling Roe entirely. And it would be politically infeasible, ethically unacceptable, and professionally impossible for her to promise publicly that she would always vote to overturn Roe v. Wade, no matter what the circumstances are.

Despite this clamor for more information, since the defeat of Reagan nominee Robert Bork, the most successful strategy has been to give as little information about the nominee as possible. The Bork nomination has become an object lesson on how fatal it is for a nominee to have a large body of widely publicized, controversial views and be eager to discuss them. Lesson learned. Since that time, no nominee has lost because he or she was too evasive, and at least one, Justice David Souter, sailed through partly because he was so unknown.

Being unknown has not been working for Miers.

Nevertheless, her nomination has been following this post-Bork strategy. Miers' responses to the recent Senate Judiciary Committee's questionnaire were terse and bland. They appeared calculated to send reassuring messages to both the left and the right. For example, the right could read her statement "I grew to appreciate even more the importance of predictability and stability in the law" as a signal that she accepts conservative tenets about activist judges. But the statement can also be read as a signal that she will not overrule Roe v. Wade.

Judging from the Judiciary Committee's response to her answers, her strategy is not working. In a tone reminiscent of a high school English teacher returning a carelessly prepared draft, both Democrats and Republicans on the committee have demanded a do-over within a very short time frame.

It is clear that the Judiciary Committee's views about prediction differ from that of President Bush. Consequently, the administration feels pressure to furnish more information than it ever wanted to offer, yet at the same time it has been following a path that has done quite the opposite.

The president already has a strategy for dealing with this difference. During the first stages of the nomination, Bush responded to hostile conservatives by emphasizing the predictive powers of the information that already exists, the crux of which comes from his own impressions. He asked then and continues to ask that people rely on the predictive power of his own personal assessment.

The Judiciary Committee is not buying this, and among the president's conservative supporters, this strategy has met with hugely varying responses. Whether Bush's own conservative supporters buy into it depends on their aspirations about the next justice and about the Supreme Court itself.

Conservative pundits like George Will, David Brooks and some writers for the National Review want the next Supreme Court justice not simply to rule correctly on individual hot-button cases. They want the next justice to be a powerhouse who will fundamentally change American constitutional law. These conservatives mention Oliver Wendell Holmes, Felix Frankfurter, and even their old nemesis Earl Warren as models of intellectual vigor or organizational skill that this next justice should follow. By these criteria, it is far from enough to make calculated guesses on how Miers might rule on, say, abortion. At the very least, they want much more information about her overall legal philosophy and her ability to write. This sets the bar of both candidate quality and information quality very — and perhaps impossibly — high and is much different from anything the Bush administration has been willing or even able to put forward.

On the other hand, other conservatives, including evangelical Bush supporters with less-encompassing visions of what the new justice should accomplish, are more willing to see Miers' religious beliefs and Bush's endorsement as a good enough indicators that she will be a proper justice.

Miers' success as a nominee depends on which predictability standards in play become dominant. Regardless of which ones ultimately win out, in the short run, the politics of prediction will be the center of attention because talking prediction is an outwardly neutral and thus a safe way to discuss the nomination while masking its politics.

At the same time, we should remember how limited the ability to predict really is. These limits go far beyond the politics of the moment. Predicting Harriet Miers is so dodgy because law constantly changes to deal with new situations, and because the Supreme Court is different from other political institutions. That will influence the successful nominee, whoever he or she is.