Racial Bias in Juridic Judgment at Private and Public Levels

ROBERT K. BOTHWELL1
University of Louisiana at Lafayette
MELISSA A. PIGOTT
Magnus Research Consultants, Inc.
LINDA A. FOLEY
University of North Florida
ROBERT M. MCFATTER
University of Louisiana at Lafayette

Mock jurors (college students and prospective jurors) made individual decisions regarding
liability and damages (before and after deliberation) in response to a case of
sexual harassment. There were no significant differences in damage awards from
college students and prospective jurors. There was evidence of racial bias among
White mock jurors against plaintiffs who accused a Black supervisor of sexual harassment:
Lower damages were recommended for plaintiffs who accepted an offer to
meet for drinks in a Black supervisor’s room than for plaintiffs who accepted the
same offer from a White supervisor. There was also evidence of racial bias among
White mock jurors against Black plaintiffs: Lower damages were recommended for
Black plaintiffs than for White plaintiffs. These effects were present in the individual
judgments of college students and prospective jurors. However, these forms of racial
bias did not carry over into the decisions made by juries comprised of college students
or prospective jurors. Subtle racial biases operating primarily at a subconscious level
may get washed out in the complex task of coming to agreement on an appropriate
award. The effects of manipulated variables on damage awards probably are overestimated
in general in mock juror studies that do not examine group verdicts.
Public opinion surveys paint a rosier picture of the attitudes of Whites toward Blacks than
in the past (Schuman, Steeh, Bobo, & Krysan, 1997). Nevertheless, discrimination against
Blacks still prevails in many facets of life that have a negative impact on quality of life
(Sidanius & Pratto, 1999), and a substantial proportion of Whites still may be opposed to full
racial equality (Schuman et al., 1997).
The form of racism has changed (Dovidio, 2001; Dovidio & Gaertner,
2004; Gaertner & Dovidio, 1986; Kinder & Sears, 1981; McConahay, 1986; Sears & Henry, 2003).
Racism in contemporary American society involves a disinclination among Whites to engage in
obvious acts of discrimination against Blacks. When discrimination against Blacks occurs, Whites
need to be able to justify it on grounds other than just race.

Some research has suggested that racial discrimination against Black de- fendants continues to be
a problem for the American criminal justice system (Bedau, 1997; Nickerson, Mayo, & Smith, 1986;
Sidanius & Pratto, 1999). However, with a decrease in the old-fashioned form of racism in
American society, we expect that most discrimination against Blacks in the American criminal
justice system must exist beneath the surface. Because it has become unacceptable to make negative
statements about minority group members on the basis of group membership, even in homogeneous
majority group set- tings, it seems likely that a public buffer frequently would exist in
American juries against the direct influence of racial bias.
Most laboratory studies that have found evidence of racial bias have studied the judgments
of individual mock jurors without looking at what happens when they become part of an
interacting group of jurors (Field,
1979; Hymes, Leinhart, Rowe, & Rodgers, 1993; Johnson, Whitestone, Jackson, & Gatto,
1995; Knight, Giuliano, & Sanchez-Ross, 2001; Skolnick
& Shaw, 1997; Ugwuegbu, 1979; Wuensch, Campbell, Kesler, & Moore,
2002). The current study examines discrimination against Black plaintiffs in a simulated civil
trial. After participants indicated individual judgments re- garding defendant liability and
plaintiff compensation, they were placed in groups of interacting mock jurors (college
students or prospective jurors) and were required to return a group verdict. This design
permits the com- parison of private judgments with public decisions.
Civil trials, which involve the claims of an alleged victim against an individual or corporate
defendant, can evoke a bias against the plaintiff among jurors (Hans & Lofquist, 1992;
Lupfer, Cohen, Bernard, Smalley, & Schippmann, 1985). When huge sums of money are requested by
plaintiffs for their alleged injuries, jurors naturally consider the possibility that the
case might be at least partially motivated by greed. This places a heavy burden of proof
on the plaintiff: Jurors will be particularly sensitive to any evidence suggesting that it was
the plaintiff, rather than the defendant, who caused the event that resulted in injury.
The hypothesis that is tested in the current study is that bias against the plaintiff in
civil trials may be more pronounced when the plaintiff is Black. Under these circumstances,
White jurors may be more inclined to attribute the event in question to the plaintiff ’s ignorance
or carelessness, rather than to defendant misbehavior. However, we consider it very probable
that this type of bias will consist primarily of subconscious associations that would not
be expressed publicly during jury deliberations (Devine, 1989; Dovidio, Kawa- kami, Johnson,
Johnson, & Howard, 1997; Greenwald & Banaji, 1995; Kawakami & Dovidio, 2001;
Wittenbrink, Judd, & Park, 1997). Therefore, we might see evidence of racial bias at a private
level, but not at a public level in the verdicts decided upon by a group of individuals acting
collectively as a jury.

Method

Participants

Four hundred and six mock jurors participated in the study: 220 college students and 186
prospective jurors. College students participated as an option for extra credit toward their
grades in a psychology course. The age range for college students was 17 to 52 years with a median
age of 23 years.
Prospective jurors were selected randomly from a list of jurors by a
market research firm and were screened for jury eligibility during a tele- phone
interview. Prospective jurors were paid for their participation. The age range for prospective
jurors was 18 to 84 years, with a median age of 48 years. Most of the mock jurors were White
(71%), followed by Blacks (16%), and Hispanics (4%).
Stimulus Cases

Participants were exposed to one of eight versions of the case. Different versions of the case were
created by presenting computer-generated photo- graphs of either a Black or a White supervisor
and either a Black or a White plaintiff. In addition, the case involved either a male supervisor
and a female plaintiff, or a female supervisor and a male plaintiff.
Within each version of the case, the incident alleged by the plaintiff was
identical. According to the plaintiff, the incident occurred in the supervisor’s hotel room during
an out-of-town business trip. The supervisor invited the plaintiff to the room for drinks and
then coerced the plaintiff into sexual activity. The supervisor indicated that failure to
comply would jeopardize the plaintiff’s job. The plaintiff resisted strongly at first, but
eventually gave in out of fear that the supervisor would have him or her fired.
The plaintiff waged a complaint with the company upon return to work after the trip. The
company fired the plaintiff as a result of the complaint. There were two defendants in this case:
the supervisor and the company. Mock jurors apportioned responsibility to the plaintiff, the
supervisor, and the company.
Procedure

Mock jurors were tested in groups of 5 to 10. They attended to an audiotaped
presentation of the facts of the case and a standard set of jury instructions. The case was
presented by two experienced attorneys: One presented the plaintiff’s case, and the other
presented the defendant’s case.
The jury instructions (standard jury instructions in the state in which the research was
conducted) were presented by a psychologist with extensive experience in jury decision making. The dependent variablesFliability and damages were assessed
following the jury instructions. Liability was as- sessed by requiring mock jurors to distribute
responsibility to the plaintiff and the defendants. A percentage was assigned to each, with the
stipulation that the percentages had to total 100%. Damages were assessed by requiring mock jurors
to decide on an appropriate figure in response to the plaintiff’s request for $1 million.
After each mock juror had decided liability and damages, the group was required to
deliberate to reach a consensus regarding the apportionment of liability and the appropriate damage
award. Following deliberation, individual decisions regarding liability and damages were
assessed again.

Results

Repeated-measures ANOVA was conducted on two dependent variables separately: plaintiff fault
(percentage of responsibility attributed to the plaintiff) and plaintiff compensation
(amount of compensation recom- mended for the plaintiff). Each was assessed before and
after deliberation from individual mock jurors, and each jury was required to deliberate to a
consensus regarding both liability and damages.
The analyses presented first will be those conducted on the judgments of individual mock jurors
with race of victim, race of defendant, and type of mock juror (college student or prospective
juror) as independent variables. The analyses were conducted first on data only from White mock
jurors to assess racial bias at the private level. Then, those results will be compared to the
results of the same analysis on data from all participants, including those from other ethnic
groups.
In the analyses conducted on the amount of compensation mock jurors recommended for
the plaintiff, actual damages given were corrected first for extreme skewness with a log
transformation. The relevant means from both analyses are shown in Table 1. The means that are
shown were pooled across type of mock juror. There were no main effects for type of mock juror
on plaintiff fault, F(1, 280) 5 1.22, ns; or plaintiff compensation, F(1, 278) 5 1.96, ns.

Plaintiff Responsibility

The race of defendant main effect on plaintiff fault was significant, F(1, 280) 5 7.74,
p 5 .006. White mock jurors attributed more fault to the plaintiff when the defendant was
Black. This effect interacted with time of assessment, before or after deliberation, F(1,
280) 5 3.38, p 5 .067. The

Table 1 (See PDF)
effect was stronger after deliberation, F(1, 287) 5 21.32, p o .0001; than before
deliberation, F(1, 288) 5 2.02, p 5 .157. After deliberation, White mock jurors attributed
19.4% of the responsibility to the plaintiff when the defendant was Black, compared to 13.1% to
the plaintiff when the defen- dant was White. The effect of race of victim was
not significant, F(1, 280) 5 2.73, p 5 .099. There was an interaction of this weak effect with
time of assessment, before or after deliberation, F(1, 280) 5 3.34, p 5 .069. The effect was not
significant before, F(1, 287) 5 2.48, p 5 .117; or after deliberation, F(1, 286) 5 0.01, ns.
The tendency to attribute more fault to Black plaintiffs (20.8%) than to White plaintiffs
(17.4%) before deliberation vanished during deliberation.
There was a significant interaction between race of victim and type of mock juror, F(1, 280) 5
9.36, p 5 .002. This interaction reveals a tendency among White prospective jurors, but not
White college students, to place more blame on Black plaintiffs than on White plaintiffs.
Prospective jurors showed this bias before deliberation, F(1, 134) 5 7.19, p 5 .008; and after
deliberation, F(1, 134) 5 7.15, p 5 .008. Before deliberation, prospective ju- rors assigned 25.1%
of the responsibility to the Black plaintiff, compared to
16.6% to the White plaintiff. After deliberation, prospective jurors assigned
19.5% of the responsibility to the Black plaintiff, compared to 13.7% to the White plaintiff.
There was a general decrease in plaintiff responsibility dur- ing deliberation, F(1, 288) 5 14.59,
p 5 .0002.
Amidst this change, a bias against Black plaintiffs among White pro- spective jurors
remained. This type of bias was not found with White college students. There was no difference
before deliberation, F(1, 151) 5 0.20, ns; while a difference in the opposite direction was found
after deliberation, F(1,
150) 5 8.35, p 5 .004. Following deliberations, college students attributed
17.8% of the responsibility to the White plaintiff, compared to 12.9% to the Black plaintiff.
There was also a significant interaction between race of victim and race of defendant,
F(1, 280) 5 7.14, p 5 .008. The pattern of this interaction is shown in Table 1. The interaction
is clearest after deliberation, F(1, 285) 5 6.70, p 5 .010. White mock jurors showed a tendency
to place more blame on the Black plaintiff when the defendant was Black, but did not show this
distinction as strongly with the White plaintiff.
The interactions between race of victim and race of defendant, F(1,
394) 5 5.98, p 5 .015, and between race of victim and type of mock juror, F(1, 394) 5 7.09, p 5
.008, were still present when data from all participants, including those from other ethnic groups,
were analyzed. The pattern for the interaction between race of victim and race of defendant in
the data from all participants is similar to that shown in Table 1. There continued to be a
main effect for race of defendant, but this effect was most prominent with the Black plaintiff.
The bias against Black plaintiffs among White prospective jurors, but not among White
college students, was not present among prospective ju- rors from other ethnic groups.
Hence, the effect is watered down when the analysis is conducted on all participants, especially
after deliberations. Fol- lowing deliberations, there was not much remaining of the tendency to
show a bias against Black plaintiffs among all prospective jurors. They attributed
17.5% of the responsibility to the Black plaintiff, and 14.3% to the White plaintiff, F(1, 184) 5
2.85, p 5 .093. College students showed the reverse tendency, attributing more responsibility
to the White plaintiff (17.2%) than to the Black plaintiff (13.4%), F(1, 214) 5 7.74, p 5 .006.

Plaintiff Compensation

The effects of race of victim and race of defendant on White mock jurors’ judgments regarding
plaintiff compensation are similar in some respects to those found with the analysis of
attributions of plaintiff fault (see Table 1). The main effect for defendant race was present,
with less compensation recommended when the defendant was Black, F(1, 278) 5 8.90, p 5
.003. This effect did not interact with type of mock juror, F(1, 278) 5 1.33, ns. The main effect
for race of victim was stronger in this analysis than in the analysis on plaintiff
compensation, with less compensation recommended for the Black plaintiff, F(1, 278) 5 8.38, p 5
.004. This effect did not interact with type of mock juror, F(1, 278) 5 0.92, ns.
Neither of the effects shows significant change during deliberation, de- pending on whether
they were assessed before or after deliberation. The effect of race of defendant after
deliberation, F(1, 288) 5 14.58, p 5 .0002, was much stronger than the race of victim
effect after deliberation, F(1, 287) 5 5.30, p 5 .022. While there was a significant effect
for race of victim on log awards after deliberation (M 5 11.51 for the Black plaintiff
compared to 12.25 for the White plaintiff), this effect was not strong enough to come through all
the variance in actual awards, F 5 0.00, ns (M 5 $376,367 for the Black plaintiff, compared to
$376,324 for the White plaintiff).
The much stronger race-of-defendant effect was evident after delibera- tion with both
compensation variables. The mean log award when the de- fendant was Black was 11.20, compared
to 12.42 when the defendant was White, F(1, 288) 5 14.58, p 5 .0002. The corresponding mean for
the actual awards recommended by individual White mock jurors after deliberation was
$297,124 when the defendant was Black, compared to $438,416 when the defendant was White, F(1,
288) 5 6.56, p 5 .011.
There was an interaction between race of victim and race of defendant on log awards, with
White mock jurors recommending substantively lower damages for the Black plaintiff who joined a Black defendant for drinks in the defendant’s hotel
room, F(1, 278) 5 4.32, p 5 .039 (see Table 1). How- ever, this interaction was not present
when data from all participants were analyzed, F(1, 396) 5 1.35, ns. There was also no effect for
race of victim in that analysis, F(1, 396) 5 2.51, p 5 .114. The main effect for race of defen-
dant was the only one still present, F(1, 396) 5 8.51, p 5 .004. This effect was weak
before deliberation, F(1, 399) 5 3.31, p 5 .074 (M 5 10.92 when the defendant was
Black, compared to 11.61 when the defendant was White), but became more substantial during
the course of deliberation, F(1, 402) 5 14.68, p 5 .0001 after deliberation (M 5 11.33 when the
defen- dant was Black, compared to 12.39 when the defendant was White).
While it is not apparent in the analysis of log awards, it becomes evident when actual damages
recommended after deliberation are analyzed that the effect is weakened by the inclusion of mock
jurors from other ethnic groups. Although the difference was still present ($481,391 was
recommended when the defendant was White, compared to $348,034 when the defendant was Black),
it was no longer significant, F(1, 402) 5 2.48, p 5 .116.

Type of Case

By now, it should be clear that the most consistent finding in the analysis of the private
judgments of individual mock jurors is the effect for race of defendant. Mock jurors
attributed more fault to the plaintiff when the de- fendant was Black and recommended lower
damages. The analysis that fol- lows examines the generalizability of the race-of-defendant
effect across two cases: either a case in which the plaintiff was female and the defendant was
male, or a case in which the plaintiff was male and the defendant was female.
This variable is included along with race of victim, race of defendant, and
mock juror gender as an independent variable in ANOVA on judgments regarding plaintiff
responsibility and plaintiff compensation (log damage awards) after deliberation. The focus in
these analyses is on the decisions that mock jurors made after deliberation in anticipation of
upcoming anal- yses that would examine the verdicts of deliberating groups of mock jurors. Data
from all participants were included in these analyses.
In these analyses, the race-of-defendant effect was found for plaintiff responsibility, F(1, 386) 5
31.33, p o .0001; and plaintiff compensation, F(1,
387) 5 18.93, p o .0001. The race-of-victim effect was found for plaintiff compensation, F(1,
387) 5 3.99, p 5 .047; but not with plaintiff responsibil- ity, F(1, 386) 5 0.03, ns.
The mean log award for the Black plaintiff (M 5 11.52) was less than the mean award for the White
plaintiff (M 5 12.08). There was no main effect for type of case with either plaintiff responsibility, F(1, 386) 5 0.71, ns; or plaintiff
compensation, F(1, 387) 5 0.32, ns. Differences between male and female mock jurors were weak but
present with both plaintiff responsibility, F(1, 386) 5 3.22, p 5 .073; and plaintiff
compensation, F(1, 387) 5 3.75, p 5 .054. As would be expected, females attributed less fault to
the plaintiff than did males (M 5 15.30 and 17.28, respectively), and females recom- mended
more damages than did males (M 5 12.07 and 11.53, respectively).
There was also a three-way interaction between race of victim, race of defendant, and type of case
on plaintiff responsibility, F(1, 386) 5 21.03, p o
.0001; and on plaintiff compensation, F(1, 387) 5 4.78, p 5 .03. There was a significant
interaction between race of victim and race of defendant on judg- ments of plaintiff
responsibility, F(1, 191) 5 29.51, p o .0001; and on recom- mended damages, F(1, 191) 5 4.17, p 5
.042, when the plaintiff was a female.
Mock jurors attributed more fault and awarded lower damages to Black female plaintiffs when the
defendant was Black. There was a weaker race-of- defendant effect for White female plaintiffs,
resulting in significant main ef- fects for race of defendant with both plaintiff responsibility,
F(1, 191) 5 41.66, p o .0001; and plaintiff compensation, F(1, 191) 5 11.72, p 5 .001.
The pattern with female plaintiffs is similar to that shown in Table 1. The pattern for male
plaintiffs was in some respects similar to that found for females: Male plaintiffs received
lower damage awards when the defendant was Black, but the difference was very weak. When
the plaintiff was a male, there were no significant main effects for race of defendant on
either plaintiff responsibility, F(1, 203) 5 2.70, p 5 .10; or plaintiff compensa- tion,
F(1, 204) 5 3.36, p 5 .07; and the two-way interaction between race of victim and race of
defendant was not found for either plaintiff responsibility, F(1, 203) 5 2.03, p 5 .16; or
plaintiff compensation, F(1, 204) 5 1.47, ns.
The two-way interaction was not strong enough, even with female plain- tiffs, to come through
all of the variance in actual damages. In fact, ANOVA on actual damages failed to reveal a
significant main effect for race of de- fendant with either the female plaintiff, F(1, 93) 5
1.44, ns; or the male plain- tiff, F(1, 207) 5 1.13, ns. For the female plaintiff, mock jurors
recommended an average damage award of $342,099 when the defendant was Black, com- pared to
$536,228 when the defendant was White. For the male plaintiff, mock jurors recommended a
mean damage award of $352,842 when the de- fendant was Black, compared to $424,039 when the
defendant was White.

Jury Verdicts

To examine the effects of racial bias on jury verdicts, ANOVA was performed on the amount of
responsibility that mock juries assigned to the plaintiff and on the amount of compensation they awarded. Three inde- pendent variables
were considered in these analyses: race of victim, race of defendant, and type of case (either a
case in which the plaintiff was female and the defendant was male; or a case in which the
plaintiff was male and the defendant was female).
Another analysis conducted with type of mock jury (i.e., juries comprised of college students,
compared to juries comprised of prospective jurors) included as an independent variable
shows that there were no main effects for that variable, and it was not involved in significant
interactions with race of victim or race of defendant. Because these analyses were conducted at
the group level, the power to identify interaction effects is weak.
There was a total of 56 mock juries, resulting in a range of 5 to 8 juries in each of the eight
conditions. Nevertheless, there should be enough power to detect any substantial main effects.
This expectation was confirmed in the analysis of verdicts regarding plaintiff responsibility,
where a main effect for race of defendant was found, F(1, 48) 5 7.14, p 5 .01. More fault was
at- tributed to the plaintiff when the defendant was Black (20.2%) than when the defendant
was White (13.2%). The race-of-defendant effect was not found in the analysis of jury
awards, F(1, 48) 5 0.005, ns. When the de- fendant was Black, mock juries returned a mean
damage award of $370,000. The comparable award when the defendant was White was $372,000.
There was no main effect for race of victim or type of case in either analysis. The mock
juries assigned 15.3% of the responsibility to the Black plaintiff, compared to 17.6% to the
White plaintiff, F(1, 48) 5 0.36, ns. Black plaintiffs received a mean award of $373,000, and
White plaintiffs received a mean award of $368,000, F(1, 48) 5 0.004, ns.
In the analysis of the percentage of responsibility that mock juries as- signed to the
plaintiff, there was a near significant three-way interaction between race of victim, race of
defendant, and type of case, F(1, 48) 5 4.02, p 5 .05. The pattern of this interaction is
similar to the interaction found at the private level, but not as pronounced. When the
plaintiff was a female, the most fault was attributed to her and the lowest damages were awarded
when she was Black and the defendant was a Black male. This pattern was not found for male
plaintiffs. When the plaintiff was a male, the most fault was attributed to him and the lowest
damages were awarded when he was White and the defendant was a Black female.
The three-way interaction was not significant in the analysis of damage
awards, F(1, 48) 5 2.06, p 5 .16. Overall, only meager success was found in predicting the
amount of responsibility that mock juries attributed to the plaintiff in this case with the
three variables that were manipulated in this study, F(7, 48) 5 1.88, p 5 .09. These variables
did not contribute at all to a prediction of the variance in mock jury damage awards, F(7, 48) 5
0.32, ns.

Discussion

White mock jurors in the current study showed an inclination at the private level to give
lower damage awards to plaintiffs who accepted an offer to have drinks in a Black defendant’s
hotel room. We consider it very probable that this type of bias consists primarily of
subconscious associ- ations that typically are not expressed publicly (Devine, 1989; Dovidio et
al.,
1997; Greenwald & Banaji, 1995; Kawakami & Dovidio, 2001; Wittenbrink et al., 1997).
From the perspective of hindsight, many White mock jurors apparently thought that the plaintiff
should have known better than to go to the Black defendant’s hotel room, and they were inclined
to assign a higher level of fault to the plaintiff and to award lower damages when these were the
circumstances of the case. Even though less racial bias would be expected from college students
(Citrin, Sears, Muste, & Wong, 2001; Firebaugh & Davis, 1988; Schuman et al., 1997), the bias
against the plaintiff who went to the Black defendant’s hotel room was found among college
students and among prospective jurors. This is not so surprising when we recognize that racial
stereotypes can operate subconsciously in individuals low and high in modern racism (Devine,
1989).
White mock jurors in the current study also showed an inclination at the
private level to give lower damage awards to Black plaintiffs than to White plaintiffs. We
consider it very probable that this type of bias also consists primarily of subconscious
associations, with a bias against plaintiffs in gen- eral perhaps playing a nuclear role in the
network of relevant associations (Hans & Lofquist, 1992; Lupfer et al., 1985). In civil
trials, jurors are likely to be particularly sensitive to any evidence suggesting that it was the
plaintiff rather than the defendant who caused the event that resulted in injury. With
certain cases, jurors may be more inclined to attribute the event in question to the
plaintiff’s ignorance or carelessness, rather than to defendant misbehavior.
When a bias against Black plaintiffs operates among White jurors in civil trials, it may take the
form of what Pettigrew (1979) called the ultimate attribution error. This results in more fault
attributed to out-group members involved in a particular case than would be attributed to in-group
members. In the current study, this kind of bias was evident in the private judgments of White mock
jurors, but appeared to be filtered out at the public level. It is not something that White mock
jurors would want to admit in the company of others, if they are aware of their bias at all.
The race-of-plaintiff effect was not strong enough to influence jury decisions in any
way, while the race of defendant was present in judgments of fault, but not in recommended
damage awards. These results are interpretable from a modern racism perspective in which it is anticipated that direct forms of
racial discrimination (e.g., bias against Black plaintiffs) would be less probable than indirect
forms of racial bias (e.g., bias against women who agreed to go to a Black defendant’s hotel room
to have a drink). Racism may still occupy the 13th chair in the deliberation room (Frank,
1949), but today that chair may not get pulled up to the table as often as it did in the past.
The more subtle form of racial bias that is revealed with the race-of- defendant effect may
operate consistently in trials when alleged female vic- tims claim they were sexually assaulted
or sexually harassed by Black male defendants. This implies that stereotypes regarding the sexual
and aggressive proclivities of Black men may underlie this effect. Devine (1989) showed that such
stereotypes may unconsciously affect the behavior of those individuals who have no particularly
overt hostility toward Black men.
The race-of-defendant effect was reported in a study conducted in another Southern
venue with a different kind of case (Landwehr, 1995). In that study, White college students who
viewed the videotaped testimony of a White or a Black female who claimed she was raped in a White
or a Black defendant’s car after leaving a bar with him attributed more fault to alleged victims
who left the bar with a Black man than to those who left the bar with a White man. There was no
race-of-victim effect and no interaction between race of victim and race of defendant. However,
there was strong evidence of the influence of extra-evidentiary impressions of the alleged victim’s
per- sonality on the amount of fault attributed to her (r 5 -.62, p o .0001).
The finding that mock jurors made inferences about an alleged rape vic- tim’s personality based
on what she said about her interaction with a White or Black defendant and that these inferences
influenced the amount of fault they attributed to her (Landwehr, 1995) is consistent with an
accumulating body of research findings that indicates that jurors consider both the alleged
victim’s character and the character of the defendant in their determination of dam- ages in
civil trials (Greene & Bornstein, 2003). Furthermore, impressions of the character of the
victim appear to have a strong influence on the legal decisions of both male and female
jurors, and sometimes can be even more pronounced among female jurors because of a stronger
need for them to distance themselves psychologically from the victim.
While gender differences in the expected direction are sometimes found in studies of mock jurors’
evaluations of cases of sexual assault and sexual harassment, there is a great deal of
inconsistency in the size of this difference across studies, and there is usually a considerable
amount of within-group variation (Bridges & McGrail, 1989; Pollard, 1992; Schutte & Hosch,
1997; Thornton, Robbins, & Johnson, 1981). We found only weak differences between male and female
mock jurors in the current study. Furthermore, because most juries are comprised of a collection of both males and females, it
is obvious that whatever differences exist between males and females at the beginning of
deliberation, these usually are re- solved by the end of deliberation, at least at the public
level. For example, in a recent study conducted on mock jurors’ evaluations of the Kobe Bryant
rape case (Bothwell, 2004), all 10 of the mixed-sex juries that were tested returned not-guilty
verdicts. A high level of fault (more than 50%) was attributed to the alleged victim by
both male and female mock jurors, and authoritarian female mock jurors were found to be
particularly critical of the alleged victim.
This brings us to the last issue of why the race-of-defendant effect, still evident in the private
judgments of mock jurors after deliberation, failed to carry over into the awards recommended by
mock juries. The damage award recommended by White mock jurors when the defendant was
Black ($297,124) was substantially lower than the award recommended when the defendant was
White ($438,416). An attenuated difference in this direction remained when the data from mock
jurors from other ethnic groups were included in the analysis: $348,034 was recommended when
the defendant was Black, compared to $481,391 when the defendant was White.
This effect vanished completely in the awards that the mock juries returned: When the
defendant was Black, mock juries returned a mean dam- age award of $370,000; the comparable award
when the defendant was White was $372,000. It could be that the presence of others in the jury
who were not similarly biased engaged in arguments that provided an above-average anchor
when the defendant was Black and a below-average anchor when the defendant was White. However,
if this is what is happening, why don’t we see the same kind of anchoring in decisions with
plaintiff responsibility?
Mock juries assigned more responsibility to the plaintiff when the de- fendant was Black
(20.2%) than when the defendant was White (13.2%). After all, judgments of plaintiff
responsibility are correlated strongly with recommended awards. The problem may center on an
inherent unreliability of damage awards when 12 jurors with a variety of opinions are forced to
come to some kind of consensus: ‘‘The assessment of damages is not a straightforward process, and
jurors who are assigned this task are not in- fallible thinkers’’ (Greene & Bornstein, 2003, p.
148).
What is troublesome about jury damage awards is that two juries who are exposed to the same
case can come up with awards that are drastically different: You never know what you are going to
get. This is why attorneys usually are not motivated strongly to take any given case to trial.
Often the prospect of a jury trial begins to receive serious consideration only after any chance
for settlement has evaporated. The current data reflect the kind of enormous variability
in awards that can occur across different juries considering the same case (from $10,000 to $1 million). There was not any clear relation of this
variance to either race of plaintiff or race of defendant, and there was also no difference in
awards from juries comprised of college students compared to those comprised of prospective
jurors. The variance across juries in damage awards appeared to be largely random.
Damage awards appear to be largely overdetermined, meaning that the award from each jury is
based on a complex idiosyncratic combination of those facts that the individual jurors find
important and the way that these facts become integrated into a gestalt as particular jurors
present to others during deliberation the schemes they have used to tie certain facts together.
Awards are the outcome of a complex process that involves a sequence of decisions about what
happened, who is at fault, and how much it is worth. The effect of these decisions is cumulative
so that the tacitly agreed-upon narrative influences the amount of responsibility that is
attributed to the parties involved, which influences in turn how much money, if any, the jury
decides that the plaintiff should receive. Analyses limited to the opinions of individual mock
jurors that do not take into account the group dynamics on which the decisions of actual juries
are based often may implicate factors that really do not make much of a difference.

References

Bedau, H. A. (Ed.). (1997). The death penalty in America: Current contro- versies. New York:
Oxford University Press.
Bothwell, R. K. (2004). Bias against the victim in the Kobe Bryant rape case.
Manuscript submitted for publication.
Bridges, J. S., & McGrail, C. A. (1989). Attributions of responsibility for date and stranger
rape. Sex Roles, 21, 273-286.
Citrin, J., Sears, D. O., Muste, C., & Wong, C. (2001). Multiculturalism in American
public opinion. British Journal of Political Science, 31,
247-275.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled
components. Journal of Personality and Social Psychology,
56, 5-18.
Dovidio, J. F. (2001). On the nature of contemporary prejudice: The third wave. Journal of
Social Issues, 57, 829-849.
Dovidio, J. F., & Gaertner, S. L. (2004). Aversive racism. In M. P. Zanna (Ed.), Advances in
experimental social psychology (Vol. 36, pp. 1-52). San Diego, CA: Elsevier Academic Press.
Dovidio, J. F., Kawakami, K., Johnson, C., Johnson, B., & Howard, A. (1997). On the nature
of prejudice: Automatic and controlled processes. Journal of Experimental Social Psychology, 33,
510-540.
Field, H. S. (1979). Rape trials and jurors’ decisions:: A psychological analysis of the
effects of victim, defendant, and case characteristics. Law and Human Behavior, 3, 261-284.
Firebaugh, G., & Davis, K. E. (1988). Trends in anti-Black prejudice,
1972-1984: Region and cohort effects. American Journal of Sociology, 94,
251-272.
Frank, J. (1949). Courts on trial. Princeton, NJ: Princeton University Press. Gaertner, S. L.,
& Dovidio, J. F. (1986). The aversive form of racism.
In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 61-89).
Orlando, FL: Academic Press.
Greene, E., & Bornstein, B. H. (2003). Determining damages: The psychology of jury awards.
Washington, DC: American Psychological Association.
Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: At-
titudes, self-esteem, and stereotypes. Psychological Review, 102, 4-27. Hans, V. P., & Lofquist,
W. S. (1992). Jurors’ judgments of business liability
in tort cases: Implications for the litigation explosion. Law and Society
Review, 26, 85-115.
Hymes, R. W., Leinhart, M., Rowe, S., & Rogers, W. (1993). Acquaintance rape: The effect of race
of defendant and race of victim on White jurors’ decisions. Journal of Social Psychology, 133,
627-634.
Johnson, J. D., Whitestone, E., Jackson, L. A., & Gatto, L. (1995). Justice is
still not color blind: Differential racial effects of exposure to inadmissible evidence.
Personality and Social Psychology Bulletin, 21, 893-898.
Kawakami, K., & Dovidio, J. F. (2001). The reliability of implicit stereo- types. Personality
and Social Psychology Bulletin, 27, 212-225.
Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics: Symbolic racism versus
racial threats to the good life. Journal of Personality and Social Psychology, 40, 414-431.
Knight, J. L., Giuliano, T. A., & Sanchez-Ross, M. G. (2001). Famous
or infamous? The influence of celebrity status and race on perceptions of responsibility for rape.
Basic and Applied Social Psychology, 23, 183-190. Landwehr, P. H. (1995). The impact of
extra-evidentiary factors in rape cases.
Unpublished master’s thesis, University of Louisiana, Lafayette.
Lupfer, M., Cohen, R., Bernard, J. L., Smalley, D., & Schippmann, J. (1985). An
attributional analysis of jurors’ judgments in civil cases. Journal of Social Psychology, 125,
743-751.
McConahay, J. B. (1986). Modern racism, ambivalence, and the Modern
Racism Scale. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and
racism (pp. 91-125). Orlando, FL: Academic Press.
Nickerson, S., Mayo, C., & Smith, A. (1986). Racism in the courtroom. In J. F. Dovidio & S. L.
Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 255-278). Orlando, FL: Academic
Press.
Pettigrew, T. F. (1979). The ultimate attribution error: Extending Allport’s
cognitive analysis of prejudice. Personality and Social Psychology
Bulletin, 5, 461-476.
Pollard, P. (1992). Judgments about victims and attackers in depicted rapes: A review. British
Journal of Social Psychology, 31, 307-326.
Schuman, H., Steeh, C., Bobo, L., & Krysan, M. (1997). Racial attitudes in America: Trends and
interpretations (rev. ed.). Cambridge, MA: Harvard University Press.
Schutte, J. W., & Hosch, H. M. (1997). Gender differences in sexual assault verdicts: A
meta-analysis. Journal of Social Behavior and Personality, 12,
759-772.
Sears, D. O., & Henry, P. J. (2003). The origins of symbolic racism. Journal of Personality and
Social Psychology, 85, 259-275.
Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and
oppression. New York: Cambridge University Press. Skolnick, P., & Shaw, J. I. (1997). The O. J.
Simpson trial: Racism or status
shield? Journal of Social Issues, 53, 503-516.
Thornton, B., Robbins, M. A., & Johnson, J. A. (1981). Social perception of the rape victim’s
culpability: The influence of respondents’ personal– environmental causal attribution
tendencies. Human Relations, 34,
225-237.
Ugwuegbu, D. C. E. (1979). Racial and evidential factors in juror attribu- tion of legal
responsibility. Journal of Experimental Social Psychology,
15, 133-146.
Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racial prejudice at the
implicit level and its relationship with questionnaire measures. Journal of Personality and
Social Psychology, 72, 262-274.
Wuensch, K. L., Campbell, M. W., Kesler, F. C., & Moore, C. H. (2002).
Racial bias in decisions made by mock jurors evaluating a case of sexual harassment. Journal of
Social Psychology, 142, 587-600.