Introduction

The words “gang-related” give rise to strong emotions in jurors.  Many of us fear and hate criminal street gangs.  “It is fair to say that when the word ‘gang’ is used in Los Angeles County, one does not have visions of the characters from the ‘Our Little Gang’ series.  The word gang . . . connotes opprobrious implications.”¹

The California legislature captured the prevailing antigang sentiment in its preamble to the Street Terrorism Enforcement and Protection (STEP) Act:

In a criminal trial, gang evidence can arise in a host of different contexts.  “Evidence of the defendant’s gang affiliation—including evidence of the gang’s territory, membership, signs, symbols, beliefs and practices, criminal enterprises, rivalries, and the like—can help prove identity, motive, modus operandi, specific intent, means of applying force or fear, or other issues pertinent to guilt of the charged crime.”³

One way in which the prosecution can almost guarantee that the jury will hear gang evidence is to charge the STEP Act gang enhancement.⁴  If the jury finds that the defendant committed the crime for the sake of a gang, the gang enhancement can increase the sentence anywhere from two years to life, depending on the underlying crime.⁵  In order to prove the gang enhancement, the prosecution must present evidence of the gang’s size, its primary purposes, and its signs or symbols.  The gang enhancement requires proof that the gang has previously committed certain crimes off a list that features many of the most heinous offenses.⁶  The gang enhancement can thus serve as a conduit for gang evidence: “In general, where a gang enhancement is alleged, expert testimony concerning the culture, habits, and psychology of gangs is permissible because these subjects are ‘sufficiently beyond common experience that the opinion of an expert would assist the trier of fact.’”⁷

This study seeks to answer the question: Does the introduction of gang evidence cause jurors to convict defendants whom they might otherwise acquit?  Are jurors more likely to convict just because the defendant has a connection to a criminal street gang?  While the answers may seem obvious, there has been a dearth of scientific research as to the effect of gang evidence on jury decisionmaking.

I. Gang Status as a Relevant Extralegal Factor

Jurors are often influenced by extralegal factors in determining guilt or innocence,⁸ and the effects of these extralegal factors tend to be most potent when the evidence is equivocal.⁹  In other words, the weaker the case, the stronger the effect of extralegal factors.  There is good reason to believe that the label “gang member” invokes an extremely potent stereotype that can have a biasing effect on jurors, particularly when the evidence linking the defendant to the crime is weak.  Once a negative stereotype like gang member is activated, people often seek information that further supports the instilled perspective.  Frey referred to this as “confirmation bias.”¹⁰  Once this bias is instilled, the jurors may tend to filter the evidence presented through the negative stereotype that has been activated.

Recently, Eisen et al. directly examined the effect of gang evidence on mock juror decisionmaking.¹¹  These researchers created three versions of a simulated trial, all of which included opening statements and closing arguments by the defense and prosecution, together with direct and cross-examination of the investigating officer and the victim/eyewitness.  The three versions differed only with regard to the prosecutor’s reference to the defendant’s gang association in opening statements and closing arguments.  Gang association was manipulated by describing the defendant as either having been seen hanging out with gang members on the night of the incident (the “gang affiliate” condition), or as being a documented gang member with a gang tattoo (the “hardcore gang member” condition).  In the control condition, no mention of gangs was made at trial (the “no-gang” control condition).

When the prosecutor introduced testimony on gang affiliation in argument, guilty verdicts rose significantly from 48 percent in the no-gang control condition to 60 percent in the gang affiliate condition, and 64 percent in the hardcore gang condition.  Although data from this study supported the hypothesis that introducing evidence of gang affiliation increases guilty verdicts, it was not exactly clear what caused the effect, or how powerful it was.  It is possible that mentioning the gang association worked very much like 1101(b) evidence,¹² in that it implied the defendant had a history of criminal behavior, leading the jurors to think it more likely that the defendant committed the act in question.  Alternatively, the gang evidence may have colored the jurors’ perception of the defendant’s character just enough to push the mock jurors over the edge in a close case.  Or, it may be that introducing the defendant’s gang association prejudiced the jurors to an extent that they were willing to ignore reasonable doubt in order to incarcerate a defendant whom they perceived to be a danger to society by virtue of his gang association.  Unfortunately, the data from this study could not be used to test this latter possibility, because the underlying facts of the case were not skewed enough toward acquittal to gauge whether jurors were willing to disregard reasonable doubt just to convict a gang member.

A. The Current Study

The current study was designed to build on the initial work by Eisen et al., and to further examine the extent to which gang evidence can bias jury verdicts.  The authors hoped to examine whether the presentation of gang evidence can prejudice jurors to such a degree that they would ignore obvious reasonable doubt and vote to convict even in the absence of sufficient evidence.  In the previous study, guilty verdicts were split fairly evenly in the control condition, indicating that the evidence was equivocal, but reasonable doubt was not clearly established.  In this study, we created an eyewitness identification case in which the evidence was so weak that few jurors, if any, would vote guilty in the absence of gang evidence.  There was no evidence of the defendant’s involvement in the crime whatsoever; he became the target of the investigation solely by virtue of his association with one of the actual robbers who confessed to the crime and his documented association with a street gang.  Based on the gang connection alone, and despite the fact that the defendant did not fit the description provided by the only eyewitness, police put the defendant’s picture in a six-pack photo array.  According to the testimony, the witness viewed the six-pack for some time and then told the officer only that the defendant’s picture looked similar to the gunman.  Apart from the gang affiliation, that tentative identification was the only evidence against the defendant.  Given this scenario, it would be reasonable to conclude that any jurors who voted guilty were simply ignoring reasonable doubt.

This experiment was designed to improve on Eisen et al.’s previous methodology by addressing threats to that study’s generalizability.  First, jurors in the previous study decided on a verdict without deliberating in groups, unlike the deliberation process at trial.  In the current study, jurors deliberated in groups in addition to providing predeliberation verdicts.  Second, in the previous study, the information regarding the defendant’s gang affiliation was only presented through the prosecutor’s opening statement and closing argument.  In the current study, a gang expert testified, just as in a typical trial with gang enhancement charges.  Finally, in the previous study, jury instructions were not used.  In this study, standard instructions from the CALCRIM pattern instructions, including CALCRIM no. 1401 on the gang enhancement, were read to participants before testimony and just before deliberation.¹³  Taken together, these changes to the previous method increased the similarity between experiment conditions and typical criminal trial procedure.

The trial was designed for acquittal.  We predicted that few if any jurors would vote guilty in the no-gang control condition.  But when the defendant was charged with a gang related crime and gang testimony was added, we predicted that guilty verdicts would increase significantly despite the clear lack of evidence of guilt.

II. Method

A. Participants

Participants included 212 undergraduate psychology students (139 females, 73 males) from California State University, Los Angeles.  The students all participated in the study voluntarily and were recruited based on their enrollment in various undergraduate psychology classes.  The participants ranged in age from eighteen to thirty-six (M=21.80, SD=5.28).  The participants varied in ethnicity, with 56 percent identifying themselves as Latino, 22 percent as Asian/Pacific Islander, 4 percent as White, 7 percent as African American, 8 percent describing themselves as something other than the aforementioned ethnicities, and 2 percent not identifying their ethnicity.  This distribution generally reflects the ethnic representation in the University and the surrounding geographic area.

B. Procedures

Participants were told that they would were going to see a video of a condensed trial, and that they were being asked to act as a juror in this trial.  Participants were informed that they would see opening statements from the prosecution and defense followed by the presentation of evidence, which included testimony from the investigating officer and the eyewitness, who was also the victim in this case.  After the presentation of evidence, both sides would present closing arguments.

1. Jury Instructions

Following the informed consent procedures, participants listened to a set of selected CALCRIM pretrial jury instructions.  The instructions were selected by a California Superior Court judge who had worked for many years as a gang prosecutor before moving to the bench.  These instructions were read aloud to the participants on a prerecorded CD while they read along with their own printed copy.  The following CALCRIM pretrial instructions were used: no. 110 (trial process), no. 102 (note taking), no. 103 (reasonable doubt), and no. 104 (evidence).

Following the trial, jurors were read a set of CALCRIM posttrial instructions.  As with the pretrial procedure, a CD played the instructions, while participants read along on printed copies.  The following CALCRIM instructions were used: no. 300 (all available evidence), no. 301 (single witness’s testimony), no. 315 (eyewitness identification), no. 1600 (robbery defined), no. 1603 (intent of aider and abettor in a robbery case), no. 3500 (unanimity), and no. 3550 (predeliberation instructions).  In the gang condition, CALCRIM no. 1401 (crimes committed for the benefit of a criminal street gang) was also read.

2. Trial Video

Participants viewed the forty-five minute staged trial on a twenty-inch computer screen, in groups of four to seven.  Two current Superior Court judges, both experienced former gang prosecutors and both currently adjunct professors at a local law school, played the prosecutor and defense attorney.  A Los Angeles County deputy sheriff, who is frequently called to provide expert gang testimony in Superior Court, played the investigating officer.  The deputy also testified as an expert witness in the gang condition.  A professional actor played the victim.

The simulated trial contained the following stages: (1) Opening statements by the prosecution and defense counsel, (2) direct and cross-examination of the investigating officer/gang expert, (3) direct and cross-examination of the victim/witness, and (4) closing arguments by the prosecution and defense.

The victim testified that he had been robbed at gunpoint by two men after leaving a bar late at night.  The witness acknowledged having had several drinks just before the event.  Evidence was introduced that days after the robbery, a deputy sheriff approached a car similar to the one used in the robbery.¹⁴  The deputy testified that several young Hispanic men were surrounding the car listening to music, drinking, smoking pot, and working on the car.  The car belonged to a man named Alvarez.  In the car, the police found a gun that matched the one used in the robbery, as well as property stolen from the victim.  The jurors heard testimony that Alvarez confessed to his involvement in the robbery and settled his case before this trial.

The defendant was one of several men seen hanging around the car when the police arrived to arrest Alvarez.  The deputy testified that the defendant only became a suspect because he was a young Hispanic male, wearing a sleeveless white undershirt similar to the one described by the victim days earlier.  Beyond the defendant’s association with Alvarez, the police found no other evidence tying him to the crime.  Further, the defendant did not match the victim’s description in significant ways.  Most notably, he was covered in tattoos on his arms, chest, and neck, which were never mentioned by the victim/witness.  Nevertheless, the deputy put the defendant’s picture in a six-pack photographic lineup.  The witness reportedly studied the six-pack for some time before stating that the defendant’s face looked similar to the second robber.

Experimenters created two versions of the trial tape: In the no-gang control condition, the trial contained no mention of gang involvement.  In the gang condition, the investigating officer was qualified as a gang expert and testified that the defendant was a high-ranking member of a well-known criminal street gang.  The prosecutor reiterated this point in his closing argument.

3. Posttrial Procedures

Following the presentation of the trial video and the posttrial jury instructions, participants were given a form titled “Verdict Form 1.”  Participants were reminded to vote as if they were a juror for a real trial, and the defendant was a member of their community.  The form told participants to assume the defendant would return to their community if found not guilty, as he would after an actual trial.  The form then asked participants to indicate how they would vote if they were asked to render a verdict “right now.”  In addition, participants were asked to indicate how confident they were in their verdict on a scale of 0–100, and to provide the reasons for their verdict.¹⁵

Participants were then given a second sheet of paper, titled “Trial Questionnaire.”  This form asked participants to mark their age, gender, and race.  The form also asked several questions about the case to ensure participants had paid attention.  The form asked for the defendant’s name, Alvarez’s identity and role (the robber who plead), the defendant’s charge, and the primary evidence against the defendant.  Ten participants confused Alvarez with the defendant and/or could not identify one or either individual.  Their data were dropped from the sample.

Participants were then placed in groups of four to seven and listened to an audio recording of additional jury instructions.  The following CALCRIM posttrial concluding instructions were used: no. 3500 (unanimity) and no. 3550 (standard predeliberation instructions).  The panel was then lead to a room for deliberation.

After deliberation, participants were each given a form titled “Post Deliberation Verdict Form.”  This form instructed participants to vote on the verdict a second time, “Now that you have had the time to deliberate with your fellow jurors we would like you to vote one more time.  It is assumed that you kept an open mind during the deliberations and considered your fellow jurors[’] perspectives.”  They were then asked for a verdict and to rate their confidence in the verdict.  Participants were also asked if the verdict was unanimous, how long it took the group to reach the verdict, and the reasons supporting the verdict.  Finally, participants were asked if they had any prior knowledge of this study from discussions with other students.  If they reported having prior knowledge of the study, their data were dropped from the sample.¹⁶

C. Jurors’ Attitudes

After deliberations, follow-up questions were asked to examine the participants’ experience with and attitudes toward the police and gang members.  In summary, 42 percent of participants reported that they had friends or family that had been part of a gang; 35 percent stated that they know someone who was falsely accused of a crime; 77 percent reported being concerned about the police lying to prosecute an innocent person; and 56 percent reported being concerned about putting a gang member back on the street.

III. Results and Discussion

This study was designed to examine the potential prejudicial effect of gang evidence on juror decisionmaking.  We were interested in investigating whether including a gang enhancement charge in a criminal trial would affect juror verdicts when the evidence was weak and reasonable doubt was clear.

As predicted, participants were more likely to render a guilty verdict when evidence was introduced that the defendant was a member of a gang.  When participants were polled prior to deliberations, significantly more participants indicated that they would vote guilty in the gang condition compared to the no-gang control condition.¹⁷  Guilty verdicts in the gang condition exceeded not-guilties by nearly a three-to-one margin (33 percent vs. 12 percent).  The 21 percent increase in guilty verdicts from the no-gang controls to the gang condition is in line with data reported by Eisen et al.,¹⁸ who found a 19 percent increase in guilty verdicts when evidence of gang membership was introduced in a simulated trial (44 percent vs. 63 percent).¹⁹  In that study, when testimony was introduced by the investigating officer that the defendant had merely been seen hanging out with local gang members on the night of the incident (the gang-affiliate condition), guilty verdicts rose 15 percent from 44 percent in the no-gang control condition to 60 percent.  Taken together, these studies show that the introduction of testimony indicating any sort of association with a gang, even a weak one, can have a significant prejudicial effect on jury verdicts.

When examining verdicts after deliberations, none of the participants in the no-gang control condition voted guilty, whereas 10 percent of the participants in the gang condition continued to vote guilty after deliberation.  This included two panels that voted unanimously for a guilty verdict and one hung jury.  The fact that none of the mock jurors continued to vote guilty after deliberating shows that the evidence was clearly insufficient to support a guilty verdict.  So why did 10 percent of the mock jurors continue to vote guilty despite the clear lack of evidence in this case?  The qualitative data give us a look into the participants’ thought process.

A. Qualitative Analyses

Among participants in the gang condition who voted guilty prior to deliberation, the most common reason provided for finding the defendant guilty was related to the defendant’s gang membership and/or his previous criminal history (35 percent).  The second most common reason provided was that incriminating evidence was found in his friend Alvarez’s car (26 percent).  Following deliberations, thirty-one participants in the gang condition changed their verdicts from guilty to not guilty.  Participants stated that they changed their verdicts due to lack of evidence (48 percent); difference in the defendant’s appearance from the suspect in the victim’s description (29 percent); not understanding the facts well before deliberations (10 percent); and lack of witness credibility (10 percent).  Deliberations of the groups that continued to vote guilty in the end showed that they all focused on the defendant’s gang membership and/or criminal history.  This latter point is significant, as no mention of criminal history was introduced, so jurors must have inferred this by virtue of his gang affiliation.

In the no-gang control condition, only 13 percent of the mock jurors voted guilty before deliberations, and all of them either reported voting guilty because of his affiliation with one of the actual robbers or the eyewitness testimony.  All of these mock jurors changed their verdicts to not guilty after deliberations.

B. How Did the Gang Evidence Affect Verdicts?

The prejudicial gang effect observed in this study could be attributed to confirmation bias.  Accordingly, charging the crime as gang-related may have convinced the mock jurors that the defendant was a bad actor, independent of any evidence presented by the prosecutor.  This may have further led the jurors to infer that the defendant had probably committed this crime, based on his presumed history of criminal activity.  From this perspective, the gang evidence simply bolstered the prosecution’s case, resulting in a greater likelihood of guilt, and thus tipping the scales enough to affect verdicts.

Alternatively, the mock jurors who continued to vote guilty may not have simply viewed the gang evidence as an additive factor that helped tip the scales toward a guilty verdict.  Rather, those mock jurors may have ignored reasonable doubt altogether because of the gang evidence.  We argue that this alternative explanation is clearly plausible, and is best supported by the data.

C. Ignoring Reasonable Doubt

As noted above, after deliberation, none of the jurors in the no-gang control condition voted guilty.  These data provide strong support for the notion that reasonable doubt was well established.  Thus, it appears that the mock jurors in the gang condition who continued to vote guilty after deliberating simply ignored reasonable doubt in spite of clearly insufficient evidence.  These jurors voted to convict the defendant based at least in part on the fact that he was a member of a criminal street gang.  The qualitative data supports this explanation.  The most common reason provided by participants in the gang condition who voted guilty predeliberation was the defendant’s gang affiliation and/or his criminal background.  As noted above, the only bit of evidence suggesting that the defendant even had a criminal background was his gang membership.  Moreover, the defendant’s gang affiliation and criminal history were discussed during the deliberations of each panel when the mock jurors ultimately voted guilty.

D. Jury Nullification in the Reverse?

Taking a dangerous gang member off the street sounds like commonsense justice to many, and the merits of such an outcome are hard to argue with.  If, however, the decision to convict and incarcerate the defendant is made based in part or whole on the argument that he is a dangerous gang member and a threat to the community, and not based on the evidence of the charged crime, then this type of decision could be characterized as a type of jury nullification in the reverse.

Jury nullification occurs when sympathetic jurors disregard the specific letter of the law and instead vote to acquit morally acceptable (but legally guilty) defendants.²⁰  These jurors take the law into their own hands in order to administer their own version of commonsense justice.  Most research on jury nullification has focused on the instances in which jurors acquit legally guilty but perhaps morally innocent defendants.²¹  This study presents an opportunity to examine what happens when jurors face the opposite scenario: in which the evidence against the defendant is weak, but the defendant is understood to be a morally corrupt member of a dangerous street a gang whose prime objective is to terrorize the community at large.  In such an instance, if reasonable doubt has been clearly established by the defense, the jurors would be placed in an arguably difficult position: They could elect to put a dangerous criminal back on the street, or they could decide to act with their conscience, ignore the reasonable doubt, and lock up the dangerous criminal, and by extension complete their civic duty as a juror by reducing the larger gang threat to the community.  In this instance, many ordinary folks would agree that locking him up is a notion that has genuine merit and may be the morally correct choice whereby the ends ultimately justify the means.

In the current study, the gang expert described in detail to the jury his personal knowledge that the defendant was a high-ranking member—a “shot-caller”—in a violent criminal street gang.  The officer explained that the gang was in the business of intimidation, extortion, drug dealing, and murder.

Finkel describes jury nullification as commonsense justice, noting that nullification decisions by jurors result in judgments of “what ordinary people think the law ought to be.”²²  Finkel notes that jury nullification requires four criterion to be met.  First, the proper verdict must be apparent, and jurors must knowingly rule differently.  Second, the correct verdict must be an actual option for the jury.  Third, jurors must fully understand the concept of reasonable doubt, to a degree that their verdict does not merely represent a misunderstanding of that concept.  Finally, the jurors must understand the legal instructions, and knowingly act against them.²³

Our reverse nullification argument meets Finkel’s four criterion for jury nullification.  First, reasonable doubt was apparent, as demonstrated by the fact that no participants in the no-gang control condition voted guilty in the no-gang condition after deliberation.  Second, the correct verdict—acquittal—was an actual option for the jury.  Third, an understanding of reasonable doubt was established the same way it is done in any actual trial, through reading legal instructions to jurors.  Finally, the fact that no participants voted guilty in the no-gang control condition indicates that reasonable doubt was in fact established, and the standard was understood by the participants in this study.  Thus, those jurors who voted guilty postdeliberation knowingly acted against those instructions.

Horowitz noted that in nullification cases, jurors spend time discussing the defendant’s characteristics and often acquit those defendants whom they find morally acceptable.²⁴  In this study, an examination of the deliberations among those jurors who voted guilty in the gang condition consistently revealed that discussions of the defendant’s gang membership was central to the ultimate verdict.  In this instance, when discussions turn to the moral acceptability of the violent gang member, it appears likely that jury verdicts were influenced in the reverse direction.

E. Bifurcation: A Rarely Used Remedy

In order to prevent gang evidence from biasing the trier of fact, a defendant may seek to bifurcate the trial, separating the gang enhancement charges from the underlying charges, just as allegations regarding prior convictions are commonly bifurcated.  The party seeking bifurcation has the burden of clearly establishing that there is a substantial danger of prejudice, which requires that the charges be separately tried.  Bifurcation, however, is generally denied when “gang affiliation and activity [are] woven into the factual fabric” throughout the case.²⁵

When bifurcation is granted, the gang evidence is considered in a separate phase of the trial and only if the defendant is first convicted.  Certainly, in some cases, as in retaliation shootings, the gang allegation is central to the prosecution’s case in chief.  In these cases, gang evidence must be presented to the jurors in order to prove the charges.  A similar situation frequently arises when the perpetrator announces his or her gang affiliation as part of the offense.  Not presenting gang evidence when it is central to the prosecution’s case would be problematic, as it would deprive the jurors of getting a full picture of the totality of the evidence.

In many trials that involve gang enhancement charges, however, the defendant’s gang affiliation is not central to proving the facts of the case.  Yet prosecutors still argue that just about any crime committed by a gang member is done for the benefit of the gang.  Gang experts frequently assert that all criminal activity perpetrated by individuals associated with a gang is part of “the work” each gang member needs to put in to rise in the ranks of the gang.²⁶  Greater utilization of bifurcation when the gang evidence is not central to proving the prosecution’s case may provide an effective remedy to eliminate the prejudicial impact of the gang evidence on jurors’ decisionmaking.

IV. Summary

The findings in this study replicate and extend those of Eisen et al. and show that introducing gang evidence at trial can have a significant prejudicial effect on juror decisions as to the defendant’s guilt or innocence.  Taken together, these data show that informing a jury that the defendant is a gang member significantly increases the likelihood of a guilty verdict.  Further, the current data indicate that when a gang expert is called to inform the jury that the defendant is a member of a dangerous criminal street gang, as is frequently done at trial, a significant minority of jurors will vote to convict even when reasonable doubt has been clearly established.

A. Limitations

There were several limits to this study that are worth noting.  First, the participants knew that they were taking part in a simulation.  No matter how seriously the participants took their role in the research, the decision they were being asked to make was qualitatively different than what jurors would face in an actual trial.  While the decision to convict or acquit would not have a direct effect on any actual person’s life, it also would not in fact result in a dangerous gang member being released back into the neighborhood.  It is also worth noting that the sample of young college students from in or around East Los Angeles used in this study is not typical of jurors in actual trials.  Most of the participants grew up in neighborhoods where many of their peers were affiliated with gangs at some level.  As noted earlier, 42 percent of participants in this study stated that they had friends or family who were gang members.  This sample of urban college students was also highly suspicious of the authorities in general.  When asked about their backgrounds and attitudes after participation, 77 percent reported being concerned about the police lying to prosecute innocent people, and 35 percent personally knew someone who was had been falsely accused of a crime.  It is likely that if this study were conducted in a different setting, the gang effect would have been much larger.  Future studies might sample older participants and individuals from other areas where the participants would have had less contact with and understanding of people associated with gangs.

Table 1.  Participant’s attitudes towards law enforcement, gang members,
and contact with gangs.