This article is from the Legal Research FAQ, by Mark Eckenwiler with numerous contributions by others.
It's not enough simply to be able to point to words in a book
that say what you want to prove. Different sources have varying
degrees of persuasiveness, and certain kinds of beside-the-point
remarks may have little weight, even in Supreme Court opinions.
A spectrum of significant authority, from weightiest to least
important, is as follows: court opinions; statutes and codified rules;
regulations; model codes; scholarly treatises; survey works (e.g., CJS
& Am. Jur.); and dictionaries.
Within each category above, there are (naturally) still finer
gradations. Appellate court opinions carry more weight than lower
court decisions; on questions of federal law, federal court cases
carry more authority than state cases, and vice versa on questions of
state law. Similarly, when a federal law and a state law conflict,
the federal law controls (under the Constitution's Supremacy Clause).
It is a common error to assume that everything in a judicial
opinion carries the full weight of that court's authority. An opinion
is nothing more than a court's decision resolving a *particular*
dispute between 2 (or more) parties on the basis of the specific facts
presented. Accordingly, the most significant part (or parts) of an
opinion are those where the court answers directly the questions
placed before it. (In Supreme Court opinions, the Court frequently
identifies these questions early on with a phrase like "The question
placed before us is whether or not . . . .") This answer (or answers)
on the main legal issues in a case is called the "holding"; in other
words, the holding is the proposition for which the case should be
Opinions often contain what are called "dicta," remarks or
conclusions not necessary to deciding the case before the court.
(Judges include such comments, often in footnotes, either to note the
existence of an interesting legal issue not fully presented in the
case, or to suggest how the court might analyze such an issue in the
future.) Dicta are not binding on lower courts (as is the holding in
a case), and should not be relied on as conclusive statements of the
The distinction between holding and dictum is important because
of a bedrock principle in American law known as "stare decisis". That
phrase -- Latin for "to stand on matters decided" -- simply signifies
that a court should, whenever possible, adhere to its own precedents
(and to those of higher courts). (This rule promotes consistency in
decisionmaking, which in turn allows individuals and businesses to
structure their affairs in reliance on the belief that the law will be
the same tomorrow as it is today.) In this context, a prior "holding"
is precedent to be respected; dictum, by contrast, is not binding.
It is perhaps obvious that "stare decisis" is not an ironclad
rule. Many famous Supreme Court decisions -- notably _Brown v. Board
of Education_ -- even overrule prior cases. It should be understood,
however, that such deviations are the exception and not the rule. For
a fuller discussion from the Supreme Court itself on the value of
stare decisis, see _Planned Parenthood v. Casey_, 112 S. Ct. 2791
(1992) (declining to overrule _Roe v. Wade_).
Note also that a single appellate case in a reporter may include
several different opinions. These can be divided into three basic
categories: majority opinions, concurrences, and dissents. Only the
majority opinion constitutes binding law; concurrences and dissents,
while they may contain discussions useful to later courts (and
researchers) interested in the issues involved, have no legal effect.
Sometimes a judge may concur and dissent in a single opinion.
Such opinions are usually labelled "concurring in part and dissenting
in part" at the beginning. This happens when the judge agrees with
some or all of the majority's conclusions, but disagrees strongly with
at least one portion of the analysis.
On courts with more than 3 members -- such as the Supreme Courts
of the United States and the various states, certain intermediate
state courts (such as New York's Appellate Division), and the federal
Circuit Courts when sitting **en banc -- there is also the phenomenon
of the "plurality" opinion. A plurality exists when the largest
voting bloc does not constitute a majority (e.g., 4 Justices on the
U.S. Supreme Court). Plurality opinions are generally strongly
persuasive as statements of the law, but are not binding authority;
at the U.S. Supreme Court, for example, it is common for plurality
opinions to be adopted as law by a majority of the court in a similar
case a few years later.
One last point closely tied to the phenomenon of plurality
opinions: It is essential to bear in mind that an appellate court
has one overriding function -- to affirm or reverse the decision of
the lower court. Courts exist to resolve specific, concrete disputes
(see the discussion re "dicta" above); thus, it is the "judgment" of
the court (AFFIRMED or REVERSED, or sometimes VACATED) that is most
It is possible for a court -- say, the Supreme Court -- to
affirm or reverse without reaching majority agreement on why a
particular outcome is appropriate. When a Justice (or judge)
"concurs in the judgment," s/he is agreeing with the outcome reached
by the majority (or plurality), but with little or none of their
For example, consider Prisoner X's appeal (on Fourth and Fifth
Amendment grounds) from a Circuit Court decision affirming his
conviction. Four Justices might conclude that Prisoner X should be
freed because his Fifth Amendment rights were violated at trial; if
another Justice disagrees with that reasoning, but finds that X's
Fourth Amendment claim justifies voiding the conviction, then five
Justices have agreed on the "judgment," and the conviction is
reversed. Because there was no majority agreement, however, the case
does not furnish binding precedent on the constitutional issues