69
Endnotes
1
In a previous study (Lee 2003a), for example, I found that the approximately 80% of
subsequent citations to the sampled circuit precedents were positive, and that only about 5% of
subsequent citations to circuit precedents involved kinds of negative treatment other than
distinguishing treatment (about 15% of subsequent citations). These findings suggest that circuit
judges generally follow circuit precedent rather try to avoid following precedents by
distinguishing them or calling them into question.
2
There are situations, of course, where judges disagree about whether certain legal materials are
binding on them and what those materials mean (see Grutter v. Bollinger 2002). In some cases,
policy-oriented judges may be able to manufacture legal ambiguity to enable them to pursue
policy goals. But there are other matters that are beyond reasonable dispute. This is not to say
that judges do not, at times, defy clear precedent, but rather that such clear defiance is
indefensible under the rules of the game. Removal from office is not likely as a result; the
impeachment of federal judges is exceedingly rare, and then involves breaches much more
serious than failing to follow precedent. But the violation of judicial norms will have an impact
on a judge’s reputation as a judge (see, e.g., Coffin 1980, 196-97). Defiance of precedent may
also invite reversal by the Supreme Court. To the extent that a judge’s professional reputation
and the threat of reversal matter to that judge, either for purely personal reasons or because of
ambition for higher office, the rules of the game will act as a constraint on the pursuit of policy
goals.
3
This includes state courts, typically state courts of last resort, which are not addressed in this
paper.
4
The Court has adopted a new approach to stare decisis in constitutional cases in recent years in
which it sometimes—but not always (see Lawrence v. Texas 2003)—requires that a “special
justification” exist to overrule the precedent. For a discussion of this approach, and how it
differs from the traditional approach, see Lee (2002).
5
See Kelman (1985, 232-33) for a lengthy “list of dubious constitutional propositions that are
realistically irreversible.”
6
The issues of drug quantity and drug type are inseparable—under federal law, it is not quantity
per se that matters, but quantity of a particular controlled substance. Repeating “drug quantity
and drug type” is thus unnecessary. But readers should keep in mind that drug type may also be
an issue in some cases.
7
See, e.g., United States v. Webb (2001, 894) (D.C. Cir.) (discussing pre-Apprendi law); United
States v. Eirby (2001, 36) (1st Cir.) (same).
8
Section 841(b)(1)(C) specifies the default statutory maximum for most drug types; the default
maximum for marijuana, however, is found in § 841(b)(1)(D).
9
McMillan (1986, 92) actually refused to “constitutionaliz[e]” the burden of proof at sentencing.
But as “preponderance of the evidence” means simply “more likely than not,” it would seem to
be the baseline for any factual finding.
10
The majority opinion emphasized the distinction as follows: “The Pennsylvania Legislature
did not change the definition of any existing offense. It simply took one factor that has always
been considered by sentencing courts to bear on punishment—the instrumentality used in
committing a violent felony—and dictated the precise weight to be given that factor if the
instrumentality is a firearm. Pennsylvania’s decision to do so has not transformed against its will
a sentencing factor into an “element” of some hypothetical “offense” (89-90).
Convention