When Minneapolis Police Officer Derek Chauvin suffocated George Floyd, it already had become prevalent for bystanders to use smartphones to record police officers using force during arrests. This was not the case when, on March 3, 1991, George Holliday was awakened by the sound of sirens and a police helicopter outside his second-floor apartment in Los Angeles. Instead of watching events unfold across the street, Holliday did something that then was novel. Using his recently purchased SONY Handycam, he videotaped Los Angeles Police Department (LAPD) officers beating Rodney King with batons. Within days, Holliday’s video had, in modern parlance, “gone viral,” stoking worldwide outrage. The following spring, the recording of the King beating was the centerpiece of a state court prosecution in Simi Valley, California, of four of the police officers involved. When all four were acquitted (with the jury unable to reach a verdict as to one count against one officer), the worst rioting in U.S. history engulfed much of Los Angeles, leaving almost 60 dead, hundreds injured, and a billion dollars in property damage from arson, looting, and vandalism. A generation later, history seemed to repeat itself when a bystander’s video of Chauvin killing George Floyd triggered more widespread (but less destructive) civil unrest.
The George Floyd killing brings the Rodney King beating to mind—not only because both were captured on video by a bystander and both sparked turmoil but also because both involved victims who were Black and police officers who were not. The interplay between policing and race in the United States has a troubled history. As public attention has focused on police use of force against members of marginalized communities, both prosecutors’ decisions whether to bring charges and the outcome of cases in which charges are brought increasingly have become perceived as barometers of racial justice. Although there is reason to question whether successful prosecutions of accused officers have a meaningful deterrent effect, it is undeniable that they drive public perceptions about the criminal justice system in ways that other, perhaps more effective, mechanisms for addressing police misconduct—such as administrative sanctions, civil “pattern and practice” lawsuits, and consent decrees—do not. This fact, coupled with the threat that unrest or rioting may follow an unwelcome verdict, makes it imperative for the criminal justice system to respond fairly, transparently, and effectively to allegations of police criminality.
My views on these matters were shaped by the events in Los Angeles in the 1990s, when I was an assistant U.S. attorney in the Central District of California. While rioting raged in late April and early May 1992, I was assigned to investigate whether the LAPD officers who were acquitted in state court had violated Rodney King’s constitutional rights and, if so, whether a significant federal interest merited a federal prosecution. My co-counsel was the late Barry Kowalski of the Civil Rights Division of the Department of Justice. Barry and I were assisted by Assistant U.S. Attorney Lawrence Middleton and Department of Justice Civil Rights Division attorney Alan Tieger, as well as other attorneys and paralegals from our offices, and agents from the Federal Bureau of Investigation. When the Department of Justice decided that federal prosecution was warranted and a federal grand jury agreed by returning an indictment, our team prosecuted the officers in what became known as “the federal Rodney King case.” In April 1993, after a month-long trial, the jury convicted two of the defendants, LAPD Sergeant Stacey Koon and Officer Laurence Powell, and acquitted two other officers, Theodore Briseno and Timothy Wind. Thereafter, I continued to represent the United States when the two convicted officers appealed to the Court of Appeals for the Ninth Circuit.
In the years since, I have followed media accounts of other controversial police uses of force—both those that have resulted in prosecution and those that have not—through the lens of my experience in the Rodney King case. Some common themes have emerged that may aid an understanding of where justice lies when there are allegations that a use of force by police has crossed the line into criminality.
One of the most prominent features of modern media coverage of video-recorded police uses of force, especially on cable news networks and in the opinion pages of newspapers, is the instant certainty about the guilt or innocence of involved police officers, along with calls for swift initiation or declination of criminal charges. This generates considerable pressure on prosecutors—who usually are publicly elected—to make swift decisions whether to seek or reject indictment. This pressure to move quickly only increases when outraged community members threaten or resort to civil unrest, which in turn can provoke rioting and looting, or when opponents of prosecution claim that the specter of possible criminal charges will cause other police officers to forgo effective enforcement of the law.
Even without this pressure, there is much at stake in criminal prosecutions. On the one hand, there is a need to preserve order, deter future crime, achieve just punishment, and vindicate victim interests. On the other, it is imperative to ensure fair process and avoid a wrongful charging decision or conviction. When a controversial police use of force is captured on video, these competing demands can become particularly acute, especially when there is considerable media attention, as was true in both the Rodney King beating and the George Floyd murder.
Prosecutors are duty-bound to ignore outside pressure to make speedy charging decisions. And they must. Even when prosecution is warranted, a hasty charging decision can make a successful outcome at trial less likely. Criminal charges initiate adversarial litigation, which can make it more difficult to gather reliable evidence. Further, an indictment terminates the grand jury’s investigative role, denying the prosecutor a powerful tool for gathering evidence. But unnecessarily limiting investigative options is not the worst possible outcome from a hurried charging decision. A prosecutor who brings charges quickly in response to public pressure risks getting it wrong by prosecuting when there is insufficient evidence to establish guilt or, even worse, when there is actual innocence. On the flip side, a rushed decision to forgo prosecution allows police criminality to go unaddressed.
On April 30, 1992, the day after the state court acquittals of the LAPD officers and while the riot was raging in Los Angeles, then President George H. W. Bush announced on national television that a federal grand jury had issued subpoenas in a civil rights investigation. President Bush’s confirmation of a federal investigation was his last official word on the matter, either in public or to our team. Similarly, no Department of Justice official ever communicated or suggested to us that there was a deadline for a charging decision or an expected or preferred outcome of our investigation. Our team was tasked with methodically investigating whether the beating violated Rodney King’s constitutional rights; we never were instructed to rush to judgment or to seek charges.
We followed our training as federal prosecutors, which means that we dispassionately gathered and analyzed facts before we made decisions. Department of Justice policy requires that its prosecutors believe “that the person’s conduct constitutes a federal offense and that the admissible evidence will probably be sufficient to obtain and sustain a conviction” before seeking charges. It is not uncommon for federal prosecutors to be faced with facts as troubling as those depicted on the Rodney King beating video, but if there is insufficient evidence to convict, they nonetheless must abide by this policy requirement and forgo seeking unprovable charges. Although the videotape of the Rodney King beating was a compelling piece of evidence—one that clearly would be vital to any prosecution—we could not rely on its visceral impact as sufficient to establish guilt.
Even when there is what appears to be indisputably incriminating video evidence, viewers can perceive such proof differently. This point was brought home during voir dire when we asked a potential juror who had seen the video on television to describe what it showed. She described the confrontation as a stand-up fight between King and police and not, as the video unambiguously showed, police officers hitting King with batons and kicking him while he was on the ground. This potential juror’s description not only made clear that we did not want her to serve on the jury—it also showed that perceptions of video evidence can vary dramatically.
Further, the video of the King beating showed much that was not immediately apparent. We watched it literally hundreds of times, always noticing things that previously had escaped our attention. Among other things, we tracked what each of the many bystander police officers did during the critical events and how their actions reflected on claims that King posed a threat; we tried to identify civilian onlookers depicted in the video; we watched what each involved officer did after the use of force had ended; we strained to hear if any commands were audible despite the sound of the LAPD helicopter hovering over the scene; and we tried to determine how police had treated the other two occupants of King’s vehicle. We made efforts to enhance the video, hiring an expert and having a laboratory employ what then were sophisticated techniques to clarify an out-of-focus portion of the video and correct for camera movement.
What we were able to glean from the video required follow-up investigation. Did the events on the video align with or contradict accounts that the involved and bystander police officers had given after the arrest? Were the actions of the police officers consistent or inconsistent with their training and LAPD policy? Could we locate the civilian bystanders? How did civilians’ accounts compare with the video? Could the civilians whom we interviewed help us identify and locate other eyewitnesses who were not shown on the video?
This process took time but was necessary to determine whether to seek charges. It paid dividends. A small example: Well after the use of force depicted on the video ends, defendant LAPD Officer Powell can be seen taking an object from his belt and raising it to his mouth. It is his police radio. On another piece of evidence—the audio recording of the police radio communications with the LAPD dispatcher—Powell can be heard calling for an ambulance for King. Identification of this event on both the video of the beating and audio recording of the radio calls enabled us to synchronize the two recordings. Because the dispatcher recording also captured police radio transmissions that preceded the video, such as Powell’s radio call about the initial stop of the car King was driving, we were able to determine exactly how long LAPD officers had been on scene with King before the video started. Surprisingly, the events had unfolded quickly; the videorecorded police use of force began only about two minutes after the car stop. This understanding proved useful in showing that another defendant, LAPD Sergeant Koon, had exaggerated when, on direct examination, he testified that a prolonged encounter with a combative King preceded the video and justified the amount of force that the police officers had used.
There was much to investigate beyond the video. We needed to ascertain what happened before the video started and what occurred afterward. In doing so, we uncovered critical new evidence, including LAPD internal affairs interviews showing that when two of the officers, Powell and defendant LAPD Officer Timothy Wind, transported King from a medical clinic, where his face was stitched, to a hospital for treatment of his more serious injuries (including multiple facial fractures), they first went to the police station. Powell left Wind, his rookie partner, in the car with the injured King and went inside the station, where Powell described the police encounter with King to other police officers, one of whom went outside to look at King. Powell and Wind omitted mention of this detour when completing their patrol log, which was supposed to fully document their whereabouts and activities during their shift. This evidence showed that the officers had disregarded LAPD policy and demonstrated Powell’s callous attitude toward a badly injured suspect in his custody, shedding light on Powell’s mental state during the earlier encounter with King.
The timing of our presentation of a proposed indictment to the grand jury—which did not occur until more than three months after we began our investigation—was an indication of our commitment to a thorough independent exploration of the facts, one uninfluenced by calls that we move quickly. Our refusal to rush to judgment, despite significant public sentiment that the LAPD officers had violated King’s constitutional rights, enabled us to conduct an exhaustive investigation, leaving us better situated when trial began, and made us confident that our ultimate decision to seek charges was supported by evidence sufficient to earn a conviction at trial.
One of the most controversial aspects of the state court prosecution was the successful defense motion for a change of venue, from downtown Los Angeles, where there was a racially and ethnically diverse jury pool, to Simi Valley, a predominately Caucasian community in Ventura County, about 40 miles from downtown. The jury drawn from Ventura County that acquitted the four officers had no Black members.
With this as a backdrop, it was remarkable that attorneys representing the defendant police officers in the subsequent federal prosecution tried to prevent one of the few Black members of the venire from serving on the jury. From a racially neutral perspective, this man seemed to be an ideal juror for the defense. He was in his early 60s, had been employed his entire adult life, and was law-abiding. There appeared to be no legitimate reason for the defense to view him as a problematic juror or to treat him differently from others in the jury pool. But from the outset, counsel for some of the defendant police officers did so. Whereas they had conducted gentle and often cursory questioning of similarly situated potential jurors, they aggressively “cross-examined” this man, apparently searching for some reason to have him excused for cause. When, despite prolonged inquiry, they were unable to ferret out any answers suggesting that he could not be fair, they exercised a peremptory challenge to remove him from the jury.
As experienced criminal practitioners well know, in its landmark decision in Batson v. Kentucky, 476 U.S. 79 (1986), the Supreme Court held that a prosecutor violates the Equal Protection Clause by exercising even a single racially motivated peremptory challenge. The Batson Court also established a process by which a criminal defendant can persuade a district court to prevent the discriminatory removal of a prospective juror. The right that the Supreme Court described in Batson is unusual. Unlike other rights in criminal cases, which belong to the defendant, this right protects jurors, guaranteeing that they will not be excluded from participation in the criminal justice system for impermissibly discriminatory reasons. If the right belongs to the juror, not the criminal defendant, then it stands to reason that neither the prosecution nor the defense can excuse potential jurors based on race. The Supreme Court reached this conclusion six years after Batson in Georgia v. McCollum, 505 U.S. 42 (1992), in which it held that “a defendant’s discriminatory exercise of a peremptory challenge is a violation of equal protection” and that the prosecutor may employ the process set out in Batson to deny a defendant’s effort to impermissibly remove a potential juror.
The Supreme Court decided McCollum less than a year before our federal civil rights prosecution began. It then was novel for a prosecutor to bring a so-called Batson challenge, but we believed that we were on firm ground in contesting the defense effort to exclude this Black venireman. We argued that the stark disparity between the way in which the defense attorneys had questioned this potential juror and their questioning of jurors of other races demonstrated that the real reason for the defense peremptory challenge of this juror was because he, like Rodney King, was Black. The differential treatment was so blatant that the judge, although generally sympathetic to the police defendants, could not ignore what everyone in the courtroom knew had occurred. He sustained our Batson challenge, and the man remained on the jury.
Although evidence of a police officer’s racial animus can be relevant to establish motive, there is a danger in thinking that evidence of racial motivation is either necessary or sufficient in an excessive force case. As often is true, in our prosecution of the LAPD officers, it was neither.
The federal criminal statute under which we proceeded, 18 U.S.C. § 242, which was enacted in its present form in 1874, traces back to criminal provisions of the Civil Rights Acts of 1866 and 1870. In somewhat archaic language, it provides that “[w]hoever, under color of any law, statute, ordinance, regulation, or custom, willfully subjects any person in any State, Territory, or District to the deprivation of any rights, privileges, or immunities secured or protected by the Constitution or laws of the United States, or to different punishments, pains, or penalties, on account of such person being an alien, or by reason of his color, or race, than are prescribed for the punishment of citizens,” shall be subject to specified criminal penalties. The district court judge in our case initially opined that section 242 required proof that the accused LAPD officers had used force against Rodney King “by reason of his color, or race.” In fact, the statute can be violated by a willful deprivation of a constitutional right without regard to the race of a victim. Ultimately, after briefing, the judge agreed that section 242 did not require that we prove that the King beating was racially motivated.
We nonetheless took the position that evidence of the defendant LAPD officers’ attitudes about race was relevant and admissible to show motive. Most significantly, we sought to show that one of the officers, Powell, harbored negative attitudes about Black people, ones that might cause him to treat King harshly because of his race. Shortly before his encounter with King that night, Powell and his partner had answered a radio call about a noise complaint concerning a party attended by Black party-goers. Powell later sent to a fellow LAPD officer through the patrol car computer system a message describing the party as “right out of ‘Gorillas in the Mist,’” referring to the title of a then recent movie about naturalist Dian Fossey’s work with gorillas. Remarkably, the judge deemed this evidence irrelevant and, thus, inadmissible. Similarly, when cross-examining another defendant, LAPD Sergeant Koon, we asked him about a draft of a book that he coauthored about the Rodney King incident, one that was published between the state court acquittals and our trial. When our questions focused on Koon’s use of a racial epithet for Black people in the book draft, the district judge sustained an objection and ruled generally that this area of inquiry was off limits.
Ultimately, the court did not permit us to present any evidence that King’s race or the officers’ views about Black civilians played a role in the use of force. Although such evidence was relevant to show bad motive, it was not necessary to satisfy the requirements in the statute. As the conviction demonstrated, it also was unnecessary to persuade the jury that Powell and Koon had willfully violated Rodney King’s constitutional rights.
It is not unusual for pundits to criticize prosecutors when they refrain from seeking charges in controversial use-of-force cases, especially when police shoot and kill an unarmed Black person. Although each charging decision is fact-dependent, as a general matter there is good reason for prosecutors to tread carefully before pursuing charges in these cases. Most civilians respect police. They view police officers in general as dedicated civil servants who protect law-abiding members of the community, sometimes under difficult and dangerous circumstances. That some police officers fail to live up to this ideal does not alter the general perception that most officers strive to do so and that their failures often are aberrational. Prosecutors, who routinely work with police, are aware of this perception and how it affects the probability of persuading a jury to convict a police officer who is charged with a crime.
Further, unlike civilians, police can lawfully use force, sometimes including deadly force, as part of their job responsibilities, and sometimes must do so. This makes it more difficult to prove that any particular police use of force, no matter how harmful the outcome, merits a conviction. Prosecutors also are aware that, unlike other kinds of cases, which routinely are resolved by guilty plea, obviating the need to prove guilt beyond a reasonable doubt, police excessive force cases often go to trial, requiring prosecutors to satisfy that daunting standard of proof.
Because the law permits police to use force under certain circumstances, use-of-force cases often turn on whether the accused officer acted with a criminal state of mind, which, depending on the nature of the charges, can be as “high” as “willfulness” or “specific intent” or occasionally as “low” as recklessness or even gross negligence. It is rare to have direct evidence of state of mind. Accordingly, in prosecutions of police officers, proof of a guilty state of mind almost invariably will depend on circumstantial evidence.
These obstacles to successful prosecution of police officers help explain why prosecutors often do not seek charges when police shoot and kill a civilian. Such cases typically involve a split-second decision by a police officer, a circumstance that often will not support an inference that the officer acted with bad purpose. Even if the suspect is later determined to have been unarmed, unless the police officer can be shown to have known this at the time, it may be difficult, if not impossible, to disprove that the officer reasonably feared that the suspect had a weapon.
This was another way in which the uses of force against George Floyd and Rodney King were similar. In both cases, the police use of force could be measured in minutes, not split seconds. Chauvin kept his knee on Floyd’s neck long after there was any legitimate reason to do so and when it was clear that Floyd could not breathe. The LAPD officers continued hitting King with batons and kicking him when he was on the ground, when he obviously posed no threat. The length of these events helped to show that the force used was impermissible and, perhaps more critically, also was compelling circumstantial evidence of the mental state of the officers involved. From the passage of time, a jury could infer that the officers were aware that there was no reason to continue using force and that they acted with a guilty mental state when they nonetheless did so. In the federal Rodney King case, when a defense “use-of-force expert”—an LAPD sergeant who trained officers in the use of force—was pressed on cross-examination about the extent and length of the King beating, he attempted to justify its continuation by arguing that LAPD policy permitted officers confronted with a combative suspect to “beat him into submission.” During closing argument, we reminded the jury of this testimony and asked whether police officers should be held to use only reasonable force or instead be permitted to beat uncooperative arrestees into submission. The compelling circumstantial evidence of bad purpose, shown in the length of time the police used force, made both the George Floyd and Rodney King cases less daunting than many other potential police prosecutions.
The Rodney King beating was not the first use of force by the police that resulted in widespread rioting. In 1965, after police used force to arrest a Black man stopped for driving under the influence of alcohol, a riot broke out in the Watts neighborhood of Los Angeles. In the aftermath of the Watts riot, the governor of California at the time, Pat Brown, tasked a commission to investigate the root causes of the unrest. The commission’s report cautioned that the “situation” was “[so] serious and so explosive . . . that . . . the August riots may seem by comparison to be only a curtain-raiser for what could blow up one day in the future.” Sadly, this passage has proven prophetic at least twice, first in 1992, following the Simi Valley acquittals in the state court Rodney King case, and again in 2020, following the George Floyd killing. The ubiquity of smartphones makes it certain that controversial police use-of-force incidents will continue to be recorded and disseminated. Without better regulation of the use of force by police and efforts to foster mutual respect among police and those who live in the communities that officers are sworn to serve and protect, it tragically may take less than another generation before we again see the kind of turmoil that resulted from the Watts arrest, the King beating, and the Floyd killing.
