Attitudinal Model is of lesser importance. In general, however, little progress has
been made in determining the relative importance of the various institutional
attributes characterizing decision-making on the United States Supreme Court.
The Attitudinal Model relies heavily on the ability to measure the attitudes of
judges apart from their decisions. In a highly original approach to this problem,
Segal and Cover ( 1989 ) used newspaper editorials to derive a measure of the
justices’ values at the time of their appointment (see also Segal and Spaeth 2002 ,
322 , for an updated set of attitude scores for the justices). Since 1953 , the most
conservative justice to serve on the United States Supreme Court is Justice Scalia
(followed closely by Justice Rehnquist); the most liberal are Justices Brennan,
Fortas, and Marshall (all tied).
The Attitudinal Model has been used to predict the decisions of the justices of the
US Supreme Court. In an interesting investigation of the model, two political
scientists pitted their predictive skills against legal experts. The experts were relying
upon their knowledge of law, cases, and justices, whereas the political scientists used
a simple predictive algorithm grounded in the Attitudinal Model. The objective was
to predict each individual vote cast in the 2002 – 3 term of the United States Supreme
Court. As it turns out, the political scientists were able to predict 75 percent of the
case outcomes; the legal experts did somewhat worse at 59. 1 percent (Martin, Quinn,
Ruger, and Kim 2004 ). In terms of predicting the individual votes of the justices,
both approaches did equally well ( 67. 9 percent vs. 66. 7 percent, for the experts and
political scientists, respectively). What is perhaps most interesting is that the statis-
tical model is so simple, and that by relying on such non-legal factors as the circuit of
origin, the type of the petitioner and respondent, etc., the model produced such a
high level of predictive success. This is likely a function of unexplored correlations
between the factual elements of the cases and the attitudinal predispositions of
the justices (just as the expert predictions undoubtedly built in implicit judgments
about the ideologies of the justices). That three-fourths of the decisions of
the Court can be predicted without knowing anything at all about legal
doctrine, precedents, and constitutional law is impressive indeed.
Thus, it seems highly probable that judges rely upon their own ideological
predilections in making their decisions. This is undoubtedly true of the rariWed
atmosphere of decision-making by the Supreme Court (e.g. no accountability, no
review of decisions), but may also be true of lower court judges who often work
with laws delegating enormous discretion (e.g. sentencing laws) and without fear
of review by a higher court (because appeals are so rare). 2 And few judges
would disavow the intention to do justice by their decisions. What is less well
2 For an example of research based at least in part on the Attitudinal Model in Courts of Appeal see
Klein 2002. At the level of the federal district courts, the Attitudinal Model plays an extremely prominent
role (e.g. Rowland and Carp 1996 ), even if attitudes are typically inferred, rather than directly measured,
from attributes such as the party of the president who appointed the judge. On the Attitudinal Model
and state supreme courts see Langer 2002 ; Brace and Hall 1997 ; and Brace, Langer, and Hall 2000.
judicial institutions 517