Kermit Roosevelt III is an associate professor of law at the University of Pennsylvania Law School.

 

Print

E-mail

Synopsis

In this debate I will review the argument from my just-published book, The Myth of Judicial Activism. The book attempts to achieve two things: first, to counteract what I felt was excessive and unhelpful political use of rhetoric about “judicial activism,” and second, to offer a more useful perspective from which to judge Supreme Court decisions, for lawyers and non-lawyers alike.

The negative argument of the book is that “judicial activism,” as the phrase is typically used, is essentially empty of content; it is simply an inflammatory way of registering disapproval of a decision. It is supposed to indicate that a judge has decided a case based on personal policy preferences rather than law. But this is not a fair description of controversial constitutional cases for two reasons. First, I’m pretty confident that almost no judge has ever believed that he or she was deciding a case this way, and I’m certain that they don’t do so with anything like the frequency critics charge.

Second, and more important, most of the difficult and controversial constitutional cases don’t have clear right answers. The charge of activism relies on the idea that there is a clear answer that judges are ignoring (and therefore that they must be motivated by something other than a good faith desire to apply the Constitution). But this so-called “plain meaning” turns out to be indeterminate in many cases: the First Amendment doesn’t tell us what counts as speech (art? dance? nude dance?) or what’s an abridgment (firing from a government job?). Likewise, the Equal Protection Clause doesn’t tell us which kinds of discrimination are forbidden and which are permitted. Because there is no plain meaning that decides the difficult cases, I argue, courts must create doctrine that takes the judges from the generalities of the Constitution to decisions in specific cases. Doctrine is what we should be evaluating, and the key issue in evaluating doctrine tends to be whether the Court is showing an appropriate level of deference to the government actor whose conduct it is reviewing.

Before turning to a discussion about appropriate deference, a quick digression on doctrine and meaning is appropriate. Originalism, in one variant—which I call “application originalism”—suggests that doctrine is unnecessary. Instead, we can use the expectations of the drafters or ratifiers as to how general constitutional language would apply in specific cases. So, for instance, rather than crafting a doctrinal rule to reflect how deferential or suspicious courts should be with respect to racial discrimination by states, we can ask whether the ratifiers would have believed that a particular act of discrimination violated the Equal Protection Clause. This approach to constitutional interpretation is frequently mistaken because we would better serve both the purpose of many general constitutional provisions and the understanding of their ratifiers if we read them to announce general standards that are to be applied in light of contemporary circumstances. So if, for example, we take the basic meaning of the Equal Protection Clause to be a ban on unjustified discrimination, the relevant constitutional question is not whether racially segregated state public schooling appeared unjustified in 1868, but rather whether it appears unjustified now. Because history can’t answer that question, I argue, we do in fact need doctrine.

So this brings up the question of how to decide whether the doctrine that the Court has crafted is sensible. The best way to do this, I suggest, is to think in terms of deference. Deference strikes me as the central concept for two reasons. First, a surprising amount of doctrine boils down to the choice of a particular level of deference—for instance, that is how I read the tiers of scrutiny, which of course crop up all over the place. Second, if we are concerned about courts abusing their power and usurping the policymaking function of other governmental actors, a lack of appropriate deference is the basic form that this abuse and usurpation would take. Also, the deference inquiry is manageable, whereas trying to decide whether a particular decision is correct or not is very hard and very technical and hence not likely to be a useful endeavor for non-lawyers.

How do we decide whether the level of deference is appropriate? I think there is a fairly small set of factors that suggest when courts should defer or should not. In the book, I identify what I think are the four most important. Institutional competence goes to the respective abilities of courts versus legislatures or executive officers to decide a particular question. Because I read many important constitutional provisions to have a cost-benefit balancing component (e.g., the Equal Protection Clause in part tells states not to discriminate against some group unless the discrimination produces net benefits to society, though also in part not to do so out of hostility), which institution is better at assessing costs and benefits in a particular context will frequently be relevant. And, to the extent that the assessment relies on weighing competing policies, the decision should generally be left to the representative branches for reasons of democratic accountability. This is a Legal Process sort of consideration. Defects in democracy attempts to identify cases in which the legislative assessment will be suspect for structural reasons—as when benefits of a law go to in-staters and costs fall on out-of-staters. This is a John Hart Ely/process theory consideration. Lessons of history considers whether particular actors have proven themselves trustworthy or untrustworthy in particular circumstances. This is where a history of discrimination against a particular group would come into the equal protection analysis, for instance. Costs of error attempts to decide whether it is worse in a certain set of cases for the courts to err by upholding an unconstitutional act or striking down a constitutional one. This consideration will lead courts to adopt stances that produce different error distributions. If, for instance, the only constitutional question is whether a law is in the public interest, the costs of error factor will ordinarily suggest deference. The harm of a mistake in a close case is small, and a legislature that makes a mistake can realize that and correct it fairly easily, whereas a mistaken judicial decision is harder to fix. But if the error is one that would impose very high costs on a small group of individuals and is unlikely to be remedied through the democratic process, the costs of error factor might suggest less deference (the “beyond a reasonable doubt” standard, evidently reflecting a belief that it is worse to convict the innocent than to acquit the guilty, is an example).

All of this sounds reasonable, I hope. Disagreement will arise in applying the factors to particular cases. But here I try to stay reasonable by judging the Court’s decisions using a very weak criterion. In most difficult cases, there will be factors pointing in different directions, and any reasonable reconciliation of the competing factors is legitimate. That is, different judges will, appropriately, have different views as to the relative significance of these factors, and they will reach different conclusions about how much deference is appropriate in particular cases. There really aren’t right answers here, so all we can ask is that the judges be sincere and, to some extent, consistent. Most decisions will turn out to be legitimate. But, I like to say: this is not a bug, it’s a feature. Most of what the Court has been doing is legitimate.

A couple of examples will illustrate how these factors may be applied to Court decisions. Heightened scrutiny (a lack of deference) for discrimination against women is supported by the lessons of history factor (see, e.g., coverture). How gender discrimination doctrine fares under a defects in democracy factor is unclear: women are underrepresented in legislatures but are a majority of the electorate. The institutional competence factor almost always suggests deference. The costs of error factor will probably depend on the particular issue. Taking all of these factors into account, a court could legitimately come out either way on the question of whether heightened scrutiny is appropriate. So what’s the use? There is some value in focusing disagreement on the relevant factors and their respective significance. But there might also be a value in structuring analysis in a way that highlights consistency, or the lack thereof. If the Court says that heightened scrutiny is not appropriate, or not appropriate only because of women’s voting power, then it will be hard for that Court to explain why heightened scrutiny is appropriate for affirmative action.

In other cases, it’s very hard to justify the level of deference the Court uses. Too much deference to the purported innocent justification for racial segregation was the problem in Plessy, and it’s hard to see why the Louisiana legislature of 1890 should have been trusted. Too little deference to the legislature was the problem in Lochner, and it’s not obvious why the legislature wasn’t simply the better decisionmaker (though I think that Lochner is more defensible than commonly supposed). And, to be provocative, I find the lack of deference hard to justify in the Section 5 cases like Garrett and Kimmel, in Commerce Clause cases like Morrison (Lopez is a little easier), in the adoption of strict scrutiny for affirmative action, and in Bush v. Gore (though I also think there’s more to say about that case). I look forward to your thoughts.

Richard W. Garnett responds to Kermit Roosevelt III.