Firstly, the notion that unanimity might be important depends upon some conception of dis-
sent. That is, in a situation where opinions of judges are ordinarily delivered individually and
seriatim, as in the British House of Lords Appellate Committee, unanimity as to some issue may
only be visible to the skilled observer and is unlikely to be a basic feature of public perception of a
case. Thus, the first important component of the Supreme Court’s unanimity institutions has been
the institution at a very early stage in the Court’s history of majority opinions — indeed, initially,
unanimous opinions — to which other Justices joined
2
. In a period when the Court needed to
exercise its powers quite defensively, the capacity to speak with a unified voice may have both
increased the impact of decisions, by making it seem that any court would have reached the
same judgement in all details, rather than clarifying that different court memberships would have
reached different decisions, and emboldened individual Justices to write more far-reaching deci-
sions by rendering each decision institutional rather than personal. Where the Justices perceive
their powers as being under threat, the same calculation may still apply
3
.
The second major point is the much-analysed one that the institution of majority-opinion pro-
duction was, almost from the beginning, substantially dominated by unanimity. Dissent increased
somewhat in the latter half of the 19th Century, then increased suddenly and spectacularly and
rapidly from around 1940. The latter change is usually ascribed to the normative impact of per-
sonnel change — in particular, to the consquences of Harlan Fiske Stone’s tenure as Chief Justice,
although there may also be an impact from other institutional changes
4
. While these changes are
the subject of important debate, they do not bear directly on the desegregation cases from Brown
onwards. By the late 1940s and early 1950s, it must have been clear that a general norm of una-
nimity across all cases was no longer sustainable.
Thirdly, while there remains substantial unanimity in some cases — approximately 37% both
for the Warren era and for the whole period 1953–1990 — many important cases are, of course,
now not unanimously decided.
5
Unanimity must therefore be explained where it occurs for the
post-1940 court: we need to explain why dissensus on the Court is not being made public through
dissents and separate opinions in a particular case.
2.2
Modern unanimity
We may best understand these issues in broadly strategic terms
6
. The justices have limited re-
sources: each chamber can be responsible for only a certain amount of legal product in a given
term, given the number of clerks available, the nature of bureaucratic resources available, and so
on. These limits may have changed over time, with the creation and expansion of the clerking
system,
7
and with increases in bureaucratic productivity through (for example) the introduction
of photocopiers in the early Burger Court and of a word-processing system in the later Burger
Court. But in a given period, the limits are effectively fixed. Within that, success in meeting the
goals of a Justice (whatever they are) is likely to be influenced by appropriate use of these fixed
resources.
2
See Klarman (2001, 1157–8).
3
An obvious example is US v. Nixon, and a similar perception meant that many external commentators were led to
predict a unanimous decision in Bush v. Gore.
4
On the impact of Chief Justice Stone, see Walker et al. (1988); Haynie (1992); Caldeira and Zorn (1998); on the
possible impact of institutional changes such as the move to a new building and the increasing numbers of law clerks,
see O’Brien (1999); Best (2002).
5
Hensley and Johnson (1998), 388.
6
Epstein and Knight (1998); Maltzman et al. (2000).
7
Garrow (1999).
2
Convention