Court Decision Making in Police Brutality Cases, 1990-2000

• DOI: 10.1177/1532673X04264986
• Author: Madhavi Mccall
• Date: 01 January 2005
• Published date: 01 January 2005
• Subject Matter: Articles

10.1177/1532673X04264986
ARTICLE
AMERICAN POLITICS RESEARCH / January 2005
McCall / POLICE BRUTALITY CASES, 1990-2000
COURT DECISION MAKING IN
POLICE BRUTALITY CASES, 1990-2000
MADHAVI MCCALL
San Diego State University
Prior research on state supreme courts tends to concentrate on evaluating the influence of gender
or elections on the behavior of justices but not both. By studying either gender or elections, the
research fails to address the joint influence of these variables on behavior, and consequently, I
posit that a composite model containing both gender and election variables provides a more thor-
ough understanding of judicial behavior. I examine police brutality cases decided in state
supreme courts between 1990 and 2000 and find that gender and elections are significant factors
in structuring judicial behavior. Consequently, I contend that state supreme court studies consid-
ering gender and institutions together provide greater understanding of the factors leading to
judicial behavior.
Keywords: state supreme court decisions; police brutality cases; state court elections;
effects of judge gender
Behavioral analysis within the judicial subfield is a long-standing
area of research (see Pritchett, 1948; more recently, see Epstein &
Knight, 1998). Although many theories abound, in recent years, as
applied to state supreme courts, institutionally based and gender-
based models appear more prevalent than others. Namely, one set of
analysts concentrates on discerning if there are voting differences
between male and female judges (Allen & Wall, 1993; Beiner, 1999;
Crowe, 2000; Davis, Haire, & Songer, 1993; Gryski, Main, & Dixon,
1986; Martin, 1993; Martin & Pyle, 2000; Maule, 2000; McCall,
2002, 2003a; McCall & McCall, 2002; Pruitt, 1994; Resnik, 1988;
Sherry, 1986; Songer & Crews-Meyer, 2000; Songer, Davis, & Haire,
Author’s Note: Please direct all correspondence to Madhavi McCall, Political Science Depart-
ment, College of Arts and Letters, San Diego State University, 5500 Campanile Drive, San
Diego, CA 92182; phone: 619-594-5050; fax: 619-594-7302; e-mail: mccall@mail.sdsu.edu. I
would like to thank Alex Bernstein, San Diego State University, for help in coding the data used
in this research.
AMERICAN POLITICS RESEARCH, Vol. 33 No. 1, January 2005 56-80
DOI: 10.1177/1532673X04264986
© 2005 Sage Publications
56

---

McCall / POLICE BRUTALITY CASES, 1990-2000
57
1994; Walker & Barrow, 1985)1, with most scholars positing that
female justices are more liberal than their male brethren (Allen &
Wall, 1993; Davis et al., 1993; McCall, 2002, 2003a; McCall &
McCall, 2002; Solimine & Wheatley, 1995; Songer & Crews-Meyer,
2000). Although the empirical results are mixed, judicial gender liter-
ature nonetheless contends that gender differences exist. A second set
of judicial literature holds that institutional features surrounding the
selection and retention of state supreme court justices structure their
decisional patterns (Dubois, 1986; Hall, 1987, 1992; Hall & Brace,
1989, 1996; McCall, 1999, 2001, 2003a, 2003b). These authors sug-
gest that elected justices exhibit more conservative voting patterns
than appointed justices because of the electoral necessity. Although
other institutional features may be used as control variables, research
concentrates on delineating the import of elections and provides
empirical evidence that elections matter (Hall, 1987; Hall & Brace,
1989; McCall, 1999, 2003a, 2003b; but see Traut & Emmert, 1998).
Although both sets of scholars support their contentions through
empirical analysis, neither set currently asks the next series of logical
questions: How do gender and institutional variables jointly affect
behavior? Are female justices constrained in their policy choices by
institutions? Are elected women more conservative than appointed
women? Do women suppress gender-motivated behavior to win elec-
tions, with the electoral process thus leading to greater homogeneity
on elected courts?2 Although the influences of gender may be miti-
gated by institutions, there is often a disconnect within the two sets of
literatures, with those who primarily study the influence of gender
failing to control for court-level institutional differences and those
who primarily study institutional traits failing to control for gender.
This point is further evident in two death penalty studies conducted
using the same data set. Songer and Crews-Meyer’s (2000) analysis of
the voting behavior of state supreme court justices indicates female
justices are more liberal than male justices in death penalty cases.
Moreover, Hall and Brace’s work (Hall, 1987, 1992; Hall & Brace,
1989, 1996) indicates elected justices are more conservative than
appointed justices in the same cases. So which factor, gender or elec-
tions, is the stronger element of judicial decisions, and are elected

---

58
AMERICAN POLITICS RESEARCH / January 2005
women more greatly influenced by gender or the electoral necessity? I
posit both individual approaches are flawed, contending instead that
gender differences among the justices can only be accurately dis-
cerned if institutional variables are also considered, and consequently,
I expand upon past works by presenting a more inclusive behavioral
model.
Toward this end, I examine the outcome of 170 votes cast by state
supreme court justices in police brutality cases decided between 1990
and 2000. The dependent variable is the direction of the justice’s vote
and is coded 1 if the justice renders a vote for the brutality victim or the
police department attempting to suspend an allegedly abusive officer
and 0 if the justice votes for the police officer. I consider votes for bru-
tality victims or departments as liberal and votes for police officers as
conservative.3 Using logit, I regress the dependent variable against
institutional, policy, electoral, and legal variables to discern the
factors contributing to case outcome.
Before turning to a description of the current literature, it is impor-
tant to note that this work merges institutional and gender-based anal-
yses and thus is a major departure from past research. To justify not
only the use of both institutional and gender variables but also to sug-
gest that the integrated model provides more information than either
gender-based or institution-based models can provide alone, I con-
duct this analysis in three stages. First, consistent with most gender
studies (Smith, 1994), I consider only the importance of gender, elec-
toral, and legal variables on the dependent variable, leaving out all
institutional considerations (for an example of this type of study, see
Songer & Crews-Meyer, 2000). For simplicity, I call Model 1 the gen-
der model. Second, consistent with most institutional studies, I con-
sider all the variables except the gender variable (for examples of this
type of work, see Hall & Brace, 1989; McCall, 1999). Again for sim-
plicity, I call Model 2 the institutional model. Although both models
perform well with the variables of interest significant and signed
appropriately, the results point to the need for a more integrated
approach to judicial decisions. Consequently, in the final model,
labeled composite model, I include al variables, anticipating a signifi-
cant gender variable controlling for, namely, elections.

---

McCall / POLICE BRUTALITY CASES, 1990-2000
59
LITERATURE REVIEW OF GENDER
AND INSTITUTIONAL STUDIES
Although several authors question the influence of a judge’s gender
on voting behavior, studies have produced mixed empirical results,
and consequently the impact of gender remains debatable. Some
authors find gender significant in cases dealing with women’s rights
issues4 (Allen & Wall, 1993; Davis et al., 1993; McCall, 2003a;
McCall & McCall, 2002) whereas others see differences in some
women’s rights issues and not others (Martin & Pyle, 2000). For
instance, although some studies document the import of gender in
sexual harassment (McCall, 2003a) and sex discrimination (Davis
et al., 1993) cases, Martin and Pyle (2000) find gender is insignificant
in both issue areas on the Michigan Supreme Court but find gender
significant in family law cases. The difficulty in part appears to be a
lack of agreement on what constitutes a women’s rights issue and thus,
the analyses fail to provide consistent results.
The mixed results are evident in cases that do not contain an obvi-
ous gender component as well (see Davis, 1993; McCall, 2002;
McCall & McCall, 2002; Songer & Crews-Meyer, 2000; Songer et al.,
1994; Walker & Barrow, 1985). For example, although Songer and
Crews-Meyer (2000) find women more liberal than men in obscenity
and death penalty cases, earlier work by Walker and Barrow (1985)
reveals women to be less liberal than men in cases dealing with eco-
nomic regulation, criminal justice, personal liberty, and women’s pol-
icy issues. Moreover, although recent work by McCall and McCall
(2002) finds significant gender-related differences in search-and-
seizure cases from state supreme courts, Davis et al. (1993) find gen-
der insignificant in U.S. courts of appeals’ search-and-seizure cases.
The works, taken as a whole, seem to leave questions regarding the
import of gender unresolved.
Part of the confusion might stem from the perception that female
justices’ behavior differs from males at all times regardless of the
institutional differences among the courts. Female justices may be
unable to advance policy preferences...