Hate crimes

Do We Really Need New Anti-Asian Hate Crime Laws?

A holistic look at the data shatters the narrative about bias-based violence.


In March, a man opened fire at Young's Asian Massage, just north of Atlanta. Later, he shot up two more Atlanta-area Asian spas. All told, eight people were killed. Six of them were Asian women.

Was this a hate crime?

Clearly, it targeted a certain sort of business. In the immediate aftermath, people chalked the killings up to anti-Asian sentiment. Many were quick to implicate the Trump administration's anti-Chinese rhetoric during the COVID-19 pandemic. After all, the shootings came amid a much-discussed uptick in alleged anti-Asian crimes. But neither the Atlanta massage parlor murders nor the broader narrative around anti-Asian incidents is so easily categorized.

The Atlanta shooter—Robert Aaron Long—told police he struggled with sex addiction. He was a devout Christian who felt guilty about visiting sex workers at Asian spas, friends said. Were Long's hateful acts really about race? Or were they more about misogyny—a man lashing out at women for inspiring lust in him? How significant is the fact that the victims were largely Asian women? Was his true bias against sex workers?

In one sense, none of this makes a difference. Eight lives were senselessly lost. Long's acts were morally heinous whether driven by anti-Asian racism, general misogyny, resentment of sex workers, or total randomness. And hate crime or not, murder is a serious criminal offense, punishable in Georgia by life in prison, with the possibility of life without parole or even execution.

Yet if Long was motivated by anti-Asian or anti-female bias, this would be considered, under Georgia and federal law, a hate crime. If he was motivated by hatred of sex workers, it would not. This ambiguity perfectly encapsulates the tangled logic behind U.S. hate crime laws.

Multiple high-profile bias crimes have gained attention in recent years: the 2020 shooting of Ahmaud Arbery, an unarmed black man killed by three white assailants; the 2019 massacre of nearly two dozen people at an El Paso Walmart by a 21-year-old white man who told police he was targeting Mexicans; the 2018 shooting at Pittsburgh's Tree of Life synagogue that killed 11 congregants.

Looking at these tragedies, which often splay out in wall-to-wall media coverage for days or even weeks, it's easy to understand the emotional appeal of hate crime laws. But there is no solid evidence that such laws have a deterrent effect on hate-driven attacks. And as hate crime statutes have expanded, they have increasingly ended up penalizing people for getting mouthy with cops or otherwise punishing pure speech in violation of the First Amendment. Meanwhile, media coverage of hate crimes and "awareness" campaigns around them tend to distort Americans' perceptions of the true frequency of bigotry and violence, exacerbating divisions and stirring up fear while giving strength to political agendas that threaten civil liberties and serve to uphold status-quo power. Hate crime laws might feel good. But it's far from clear that they do good.

Constructing a Hate Crime Epidemic

Crimes rooted in bias, prejudice, and identity-based hate have always existed, and too often they were greeted with insufficient concern. Attention to such crimes grew throughout the 20th century in tandem with the civil rights movement. But it wasn't until the 1980s that the concept of "hate crimes" took off as a special category of criminal offense with enhanced sentencing possibilities.

The first state hate crime laws were passed in Oregon and Washington, based on model legislation drafted by the Anti-Defamation League. By 1992, most states had them. At the federal level, the Hate Crimes Statistics Act of 1990 required the Department of Justice to tally up hate crime data from state and local law enforcement. More federal hate crime statutes were passed in the 1990s under President Bill Clinton.

The early laws focused on racial and religious motives, but as the '90s wore on, interest groups began lobbying to expand the definition of what could legally be called a hate crime. The Violence Against Women Act—authored by then–Sen. Joe Biden and eventually passed as part of the 1994 crime bill—said crimes of violence against women were a form of hate crime that robbed women of their civil rights. Soon, activists were pushing for protections based on sexual orientation.

The latter effort was deeply intertwined with the 1998 murder of Matthew Shepard, a gay college student in Wyoming. The 21-year-old's brutal killing at the hands of two young men—Russell A. Henderson, 21, and Aaron J. McKinney, 22—would be a driving force in federal hate crime expansion efforts for more than a decade. "Matthew's death, I hope, will bring about a better and deeper understanding of hate-crime laws," Elizabeth Birch, executive director of the Human Rights Campaign, told The New York Times in the days following his death.

Police said from the outset that Henderson and McKinney's primary motive was robbery. "But investigators also said Mr. Shepard's sexual orientation was a factor," the Times reported in October 1998. "They said the suspects lured Mr. Shepard from the bar by saying they, too, were gay." Some accounts indicated one of the killers had been embarrassed when Shepard made a pass at him.

In addition, Wyoming had recently failed to pass proposed hate crime legislation. One of the first people to talk to the national press about the murder was a college instructor and friend of Shepard's named Walt Boulden, who drew a link between the failed legislation and the murder.

That was enough to make Shepard a symbol for anti-LGBT violence and a poster child for expanding hate crime statutes. Celebrities including Ellen DeGeneres, Elton John, Madonna, and Barbra Streisand got involved. "While new laws are unlikely to cure the hate crime problem," a Southern Poverty Law Center statement said, "one thing seems clear: They will draw the nation's attention to it and make an important statement."

Shepard's murder was a textbook example of what Robert Blanchard, writing in Reason in 1999, called the "hate crime news formula," in which a coalition of media, activists, academics, and politicians—"all of whom have vested interests in fomenting a sense of continuous social crisis"—descends on high-profile tragedies "to suggest that the United States is a seething cauldron of hate directed at members of unpopular groups."

Throughout the '90s, the idea that America was experiencing an "epidemic" of hate crimes had been building. The San Francisco Chronicle reported in 1993 that "hate-motivated violence is spreading across the United States in 'epidemic' proportions." Democratic Lieutenant Gov. Leo McCarthy announced "an epidemic of hate crimes and hate violence rising in California."

"Hate crime is so often referred to as an 'epidemic' that one might well believe that there is a solid foundation of facts documenting that this social problem is out of control and getting worse," wrote James B. Jacobs and Jessica S. Henry in "The Social Construction of a Hate Crime Epidemic," published in the winter 1996 volume of the Journal of Criminal Law & Criminology. But no one, they noted, had actually established facts to back up the narrative, and there was ample evidence to the contrary.

"It is impossible to prove the null hypothesis: there is no hate crime epidemic," Jacobs and Henry wrote. But the data used as evidence for such an epidemic "can be used to tell a very different story than that which prevails in the media, government, and the legal academy." In other words, it wasn't clear there was a hate crime epidemic even then. The same metrics show a drop in such crimes since the publication of Jacobs and Henry's paper. Yet the same dire warnings about a rising tide of hate have persisted.

Stopping Anti-Asian Hate

The first half of 2021 was awash in stories about an alleged spike in bias-based actions against Asians in the United States. From TV ads to newspaper articles to the halls of Congress, stopping "Asian hate" became a major rallying cry. A New York Times headline from April 3 conjured "swelling anti-Asian violence" in America. "Covid 'hate crimes' against Asian Americans on rise," warned the BBC, while Voice of America reported that "Hate Crimes Targeting Asian Americans Spiked by 150% in Major US Cities."

The narrative was based on a grain of truth: In the early days of the COVID-19 pandemic, Asians do seem to have faced an increase in verbal harassment—and occasionally worse—in some U.S. cities. But increases were far from consistent, and overall incident numbers remained quite small.

For instance, New York City saw an 833 percent rise in anti-Asian incidents between 2019 and 2020, according to the Center for the Study of Hate and Extremism at California State University, San Bernardino (CSUSB). That certainly sounds dire. Yet the leap represents a rise from three incidents in 2019 to 28 incidents in 2020—in a city with an Asian population of 1.2 million overall.

Many reports of a supposed surge in anti-Asian animosity relied on data from CSUSB, which culled police records to assess bias-based incidents in 16 big U.S. cities. It found only one (Washington, D.C.) with a decline in anti-Asian incidents and one (Chicago) with no change. Data from the other cities looked grim: Anti-Asian incidents were up 150 percent in San Jose, 133 percent in Boston, and 114 percent in Los Angeles.

Yet expressing the data in terms of the percentage increase can be misleading. The raw numbers went from four to 10 in San Jose, six to 14 in Boston, and seven to 15 in Los Angeles. Cleveland, Dallas, and Philadelphia each saw six incidents in 2020, up from zero to two in 2019. Cincinnati and San Diego went from zero to one; Phoenix from two to three; San Francisco from six to nine; and Seattle from nine to 12.

Another much-cited figure came from a group called Stop AAPI Hate, which reported a staggering 3,795 hate incidents against Asians and Pacific Islanders between March 19, 2020, and February 28, 2021. But unlike the CSUSB study, this figure came from self-reports to the group's hotline, not police records. And its reporting went far beyond potentially criminal incidents.

The Stop AAPI Hate tally lumps together physical attacks and serious crimes with verbal insults, discrimination, and "shunning." If someone crossed the street or moved aside when an Asian person walked by, and the Asian person perceived it as deliberate avoidance based on race, that would be counted among the group's statistics. (Notably, the period in question was during a pandemic, when many were going out of their way to avoid crossing paths with others, regardless of race.) Overall, 68.1 percent of reported incidents were verbal harassment, an additional 6.8 percent were online harassment, and 20.5 percent were shunning. Only around 11 percent of incidents reported to AAPI—or 421—alleged physical contact.

None of this is to diminish the emotional pain or fear that taunts or avoidance can cause. But it does add important context. Talk of hate crimes and bias incidents tends to conjure images of vandalism and violence. This makes the idea of even a small increase appear extremely dangerous to the targeted group and drives up trepidation among members of the community. As an example, several Asian teen girls told NPR in April that they were afraid to leave home or partake in ordinary activities.

Unfortunately, conflation of serious crimes with verbal insults and perceived avoidance has become standard in press coverage of crimes against Asians. Media accounts have also taken to framing every crime against an Asian person as a likely hate crime, even in the absence of any evidence that the incident was racially motivated.

Reporting last March on the stabbing of a Chinese man in Manhattan, The New York Times noted that "the perpetrator, a 23-year-old man from Yemen, had not said a word to the victim before the attack" and that prosecutors were not prosecuting it as a hate crime. Nonetheless, the Times reported that "for many Asian Americans, the stabbing was horrifying, but not surprising. It was widely seen as just the latest example of racially targeted violence against Asians during the pandemic." Hunches, feelings, and blind assertions were all that was needed to make the case.

Then there was the killing of 84-year-old Vicha Ratanapakdee, who died after being knocked down on the street by 19-year-old Antoine Watson. "The fatal assault in San Francisco on a defenseless older man was the latest terrifying episode for Asian-Americans, many of whom have endured racist taunts, rants and worse during the pandemic," said The New York Times. But according to San Francisco's district attorney—who charged Watson with murder and elder abuse—there was no evidence of an anti-Asian motive. "This unfortunate assault has to do with a break in the mental health of a teenager," said Sliman Nawabi, the public defender assigned to Watson's case. "Any other narrative is false, misleading, and divisive."

A number of crimes against elderly Asian people in California were similarly pegged by the press and social media onlookers as hate crimes even though they were not charged as such. While it's possible that race or ethnicity was a motivating factor, the perpetrators didn't say or do anything leading police to believe this.

One such assault involved a suspect—Yahya Muslim—with a history of random and unprovoked attacks on people of various races and whom a judge described as having significant mental health issues. The fatal assault on Pak Ho, a 75-year-old Asian man in Oakland, took place during the course of a robbery. Sakhone Lasaphangthong of Family Bridges in Oakland's Chinatown told The Oaklandside, "I don't see people just targeting Asians. They're targeting vulnerable people. I don't see it as Blacks against Asians. I see it as crimes of opportunity."

The Trump Effect

The suspects in many of these cases were young black men. That doesn't rule out xenophobia as a motive, but it also isn't quite what most people picture when they imagine MAGA-loving, "Wuhan flu"–fearing hate crime perpetrators. Yet President Donald Trump was often pegged as complicit in alleged anti-Asian hate crimes.

"After Donald Trump repeatedly used his platform to try to racialize this disease, we continue to see a spike in rhetoric and actions against the Asian American community because of misguided fears surrounding the COVID-19 outbreak," Sen. Tammy Duckworth (D–Ill.) said, sans evidence, in a May 2020 statement.

"Inflammatory and racist rhetoric from officials at the highest level of our government has contributed to a disturbing rise in hate crimes targeting Asian Americans during the coronavirus pandemic," Sen. Mazie Hirono (D–Hawaii) announced—also without evidence—alongside Duckworth's statement.

In covering the Atlanta massage parlor shootings, the Australian Broadcasting Corporation reported that "no-one did more to legitimise violence than former president Donald Trump." Late-night host Stephen Colbert commented after the shootings that Trump "bears a particular responsibility for amplifying this form of hatred."

Again and again while reporting on a brutal attack on an Asian person that has no apparent connection to race, media will adorn the story with statistics based on self-reported surveys of perceived discrimination conducted by activist groups and supplement it with mentions of white supremacy, MAGA-crowd racism, or Trump's anti-China rhetoric. By tying together these disparate threads, stories often gave the overall impression that Trump was seeding and his followers carrying out a widespread uptick in violent assaults on Asians.

The underlying idea had, in fact, been percolating since before the pandemic. After Trump was elected president, Democratic politicians and liberal commentators took to insisting that he had provoked a spate of violent bigotry.

Many of the most notable incidents used to back this assertion turned out to be hoaxes, fabrications, or misinterpretations. Some went viral under the assumption that they were carried out by Trump supporters—a Nazi flag showing up over a home in San Francisco, a message left on an Elon University whiteboard reading "Bye Bye Latinos Hasta La Vista"—but were later revealed to be anti-Trump antics.

Reporting on this supposed hate crime spike often relied on numbers from the Southern Poverty Law Center (SPLC). "Over 400 Reports of Hateful Harassment and Intimidation Post-Election," said an ABC News segment based on the group's data. "Report: Hate crimes surge post-election," read a headline in The Hill.

But the incidents compiled by the SPLC were not only unverified; they included many things that—while objectionable—are far from crimes. The data included students in a cafeteria chanting "build a wall," a white man telling a Hispanic man to "go back to Mexico," and elementary school children telling their parents that classmates had said they would be deported.

After a gunman opened fire at the Tree of Life synagogue in Pittsburgh in 2018, media members and activists turned to the same playbook, warning of a supposed surge in anti-Semitic violence and pointing their fingers at Trump. Many cited an Anti-Defamation League study showing a nearly 60 percent jump in overall anti-Semitic incidents between 2016 and 2017. But the same study showed a 47 percent decrease in physical assaults driven by anti-Semitism. It also included in its data some questionable incidents, such as synagogue bomb threats that had since been determined not to be based on anti-Semitism.

Claims of a subsequent spike in anti-Jewish hate crime—spurred by news of a December 2019 machete attack on Hasidic Jews in Monsey, New York—were based on similarly murky evidence.

"We're facing an anti-Semitism crisis, and not just in this city," asserted New York City Mayor Bill de Blasio in January 2020. "It's happening across our country and planet." Reports of anti-Jewish crimes were up significantly in New York City in 2019 (242 complaints, compared to 186 in 2018 and 151 in 2017). But they dropped precipitously in 2020, down to 116 complaints.

National data also paint a more complicated picture. After steadily falling since 2008, reports of anti-Jewish bias incidents began rising again in 2015, dropped in 2018, then rose again in 2019, per FBI data. But the latest number—953 incidents in 2019—was still lower than in many years throughout the '90s and '00s, despite a rise in reporting agencies.

Democratic politicians and their activist/media allies were similarly quick to blame Trump's rhetoric for the 2019 Walmart shooting in El Paso. "People say to me, 'Did Donald Trump cause those folks to be killed?' Well, no, of course he didn't pull the trigger, but he's certainly been tweeting out the ammunition," said then–Sen. Kamala Harris (D–Calif.).

To back up such assertions, politicians from Sen. Bernie Sanders (I–Vt.) to Rep. Ilhan Omar (D–Minn.) and media outlets from CNN to the Associated Press all cited a working paper purporting to show that hate crimes spiked after Trump rallies.

To reach this conclusion, Regina Branton and Valerie Martinez-Ebers of the University of North Texas and Ayal Feinberg of Texas A&M University Commerce had compared crimes in counties that held rallies with crimes in those that did not. When other researchers attempted to replicate the study, they too found evidence of a Trump rally effect. But they also found that hate crimes spiked even more following Hillary Clinton campaign rallies. The researchers hypothesized that this was not because the rallies were causing crime but because the rallies were held in bigger cities, where crime already tends to be higher. "Adding a simple statistical control for county population to the original analysis causes the estimated effect of Trump rallies on reported hate incidents to become statistically indistinguishable from zero," wrote Harvard economics Ph.D. students Matthew Lilley and Brian Wheaton in a 2019 article for Reason.

Hate Crimes Mirror General Crime Trends

Decades of data on hate crimes tell a much different story than the one that gets attention. It's nowhere near as simple—nor as compelling a cause for urgent action—as the conventional narrative suggests.

Since the '90s, the FBI has been collecting hate crime data from police departments across the country as part of its annual Uniform Crime Reporting (UCR) initiative. The stats on hate crimes have loosely mirrored larger crime trends, with steady falls from the 1990s through the middle of the last decade and smaller upticks beginning around 2015.

As with rising homicide numbers more broadly, the recent uptick in hate crimes may be cause for concern. But any such analysis needs to come with a few key caveats.

The first is that these numbers are not comprehensive: Not all law enforcement agencies in the United States report UCR data, and the total number of participating agencies has increased over time. It's also worth keeping in mind that UCR data does not measure hate crime arrests, charges, prosecutions, or convictions. Instead, it measures potential hate crime incidents known to police. That means it may still count incidents later found to be hoaxes—a situation surprisingly common among high-profile stories of hate crimes—or simply crimes not based on bias.

From the data we do have, at least three encouraging observations can be made. First, hate crimes make up a very small percentage of overall crime. In 2019 (the most recent year for which we have data), law enforcement agencies reported to the FBI some 7,314 bias-based incidents, including 4,526 crimes against people (violent and otherwise) and 2,811 crimes against property. That compares to 1,203,808 overall violent crime incidents and 6,925,677 incidents of property crime. While the FBI counted 16,425 total murders in 2019, it classified just 18 murder or nonnegligent manslaughter incidents as hate crimes. Thirty rapes (out of 139,815 total) and 866 aggravated assaults (out of 821,182 total) were classified as hate crimes.

Second, most incidents classified as hate crimes do not involve serious violence. Of the 4,526 incidents categorized as crimes against persons, the vast majority were either intimidation (1,849)—that is, words, circumstances, or actions intended to induce fear—or simple assault (1,730), a label that includes things like spitting, slapping, pushing, or threatening assault while behaving in a menacing way. The vast majority of crimes against property were categorized as vandalism (2,152 incidents).

Third, hate crimes appear to be less common now than in the '90s and early '00s, even with the uptick in the last few years. Compare the 7,314 incidents in 2019 or the 7,120 incidents in 2018 to 9,730 incidents in 2001, 7,876 incidents in 1999, or 8,759 incidents in 1996.

By far the most common category of bias over time has been race, ethnicity, or ancestry bias, with anti-black incidents dramatically outnumbering all others. But there has at least been progress on this front, with the number of anti-black incidents dropping from 3,674 in 1996 to 2,640 in 2006 to 1,739 in 2016. (Anti-black bias incidents ticked up slightly in 2017—to 2,013—but by 2019 had once again fallen, to 1,930.)

Incidents targeting Asians or  Pacific Islanders also fell, from 355 in 1996 to 138 in 2011. After 2012, the data categorization changed slightly, with anti-Asian bias as one category and anti–Pacific Islander and anti–Native Hawaiian bias as another. In 2019, there were 21 incidents in the latter category and 158 incidents in the former—up somewhat from the start of the decade but still down from the '90s and early '00s.

The data also don't support the narrative that rising white supremacy is driving a hate crime spike. In cases with a known offender where that offender's race was available, 65.9 percent of offenders in 1996 and 65.5 percent of offenders in 2001 were white. Last decade, the percentage of white offenders dropped below 60 percent—59 percent in 2011, 52.4 percent in 2013, 48.4 percent in 2015, 50.7 percent in 2017, and 52.5 percent in 2019.

Of course, there's another way to read these data: As more attention is paid to hate crimes, more minority communities might be getting targeted for hate crime prosecution in a discriminatory way. And that could interact with another trend: hate crime statutes being used to protect police.

'Blue Lives Matter' and Other Hate Crime Dangers

This year, the bipartisan COVID-19 Hate Crimes Act funneled more money to state and local police departments under the mantle of fighting hate crimes. While progressives celebrated the law, it will encourage more intensive policing at a time when criminal justice reformers have started successfully campaigning for the opposite. And history shows that intensified policing too often trickles down to disproportionate policing against minorities.

Since their earliest days, hate crime laws have been used to attempt to impose extra punishment on people who insult police officers. In 1991, a Florida man was charged with violating the state's hate crime law for telling a police officer responding to a domestic disturbance, "I'll shoot you, white cracker." The charge—which was eventually dropped—would have raised his potential punishment for simple assault from one to three years.

Flash forward a few decades, and we've seen hate crime enhancements added to charges for spray painting negative messages about police, calling cops "Nazis" or "Gestapo," and stomping on a "Back the Blue" sign.

With the rise of the Black Lives Matter movement, this tendency to use existing hate crime laws to punish people for hurting cops' feelings has solidified into an effort to make police a protected category under hate crime statutes. Never mind that assaults on police officers are already punished more harshly than assaults on ordinary citizens. In 2015, the National Fraternal Order of Police started pushing for cops to be included in a proposed federal hate crime law. "Right now, it's a hate crime if you attack someone solely because of the color of their skin, but it ought to be a hate crime if you attack someone solely because of the color of their uniform as well," Jim Pasco, executive director of the union, wrote in a letter to then–President Barack Obama.

In 2016, Louisiana's "Blue Lives Matter" law officially extended hate crime protections to cops and first responders. It was subsequently used to charge a man with a felony hate crime for using sexist and racist slurs against officers arresting him (the charge was later dropped) and against a man who threatened to shoot police officers. One Louisiana police chief said it could even apply to simply resisting arrest.

A few years later, Kentucky passed a Blue Lives Matter law of its own, and other states have considered doing the same. In 2018, the idea made it all the way through the U.S. House of Representatives, where just 35 members voted against a federal Blue Lives Matter law.

This year, New York City has charged people with hate crimes for using anti-Asian slurs against plainclothes New York Police Department (NYPD) officers. Charges like these implicate two hate crime dangers: a tendency to use them to go after First Amendment–protected activity, and the way changes in enforcement can produce misleading statistics.

After reporting just 28 anti-Asian hate crimes in 2020, the NYPD launched a new Asian Hate Crime Task Force, which included the use of undercover Asian officers. The number of anti-Asian hate crimes in New York City shot up to 80 between January 1 and April 4, 2021, the department says. That jump has been cited as evidence of an anti-Asian hate wave. But could it simply be evidence of more policing, including more arrests for insulting undercover cops?

As vile as they may be, such insults are protected by the First Amendment. Even though the charges won't stand legal scrutiny, it's not uncommon to see authorities use hate crime statutes to target mere hateful speech. In 2017, a Maryland teenager was charged with a hate crime for burning a Trump sign. Earlier this year, a teen was arrested for a racist social media post.

Prior to 2009, hate crimes were only a federal criminal concern if they involved interference in a federally protected activity, such as voting, or in certain other cases. But in 2009, the Matthew Shepard and James Byrd Jr. Hate Crimes Prevention Act expanded federal jurisdiction to include all crimes of violence based on actual or perceived race, color, religion, national origin, gender, sexual orientation, gender identity, or disability status.

Unnecessarily federalizing crime can lead to prosecutorial overreach. For instance, an early 2010s case involving Amish-on-Amish beard and hair cutting—a serious transgression in that community—was tried by federal prosecutors, who successfully convicted Samuel Mullet Sr. and 15 others of federal hate crimes. The convictions were overturned a few years later by a federal appeals court, with judges writing that the prosecution "treads uncomfortably close to the line separating constitutional regulation of conduct and unconstitutional regulation of beliefs" and also brings up "whether the federal hate-crime statute exceeds Congress's Commerce Clause powers as applied to the facts of this case."

Entrenching Power and Dominance

American discourse tends to treat hate crime laws either as condemnations of bigotry with no legal meaning or as a way to punish actions that wouldn't otherwise be punished. But neither representation is accurate.

Hate crime statutes generally do one specific thing: enhance criminal punishments for actions that are already against the law. They say that for whatever the underlying offense is—vandalism, harassment, theft, assault, murder—the sentence will be harsher if the offense was committed out of identity-based bias or prejudice instead of, say, pure greed or lust or non-specific anger.

Hate crime laws work similarly to the drug sentencing enhancements that were popularized in the 1980s, '90s, and '00s. The latter have since fallen out of fashion, with reformers pointing out that these enhancements drive up prison sentences, incarcerated populations, and parole logs without any evidence of a payoff in public safety. If we're serious about reforming the U.S. criminal justice system, hate crimes must receive the same scrutiny.

Hate crime statutes may make people feel like they're doing something about a serious problem. But judged by their results, they're likely to be harmless but ineffective at very best. At a 2018 U.S. Commission on Civil Rights briefing on hate crimes, none of the panelists could point to data, studies, or other evidence showing that designating something a hate crime deters, prevents, or reduces that crime or helps authorities catch perpetrators.

At worst, hate crime laws and their emphasis on individual bad motives can distract from more systemic issues.

In his 1996 book Virtually Normal: An Argument About Homosexuality, Andrew Sullivan argued that hate crime and anti-discrimination laws can function to undermine the classical liberal project as well as distract from fights against a more potent driver of abuse and inequality. "The truth is, leaving private discrimination aside, liberalism has not yet succeeded in achieving even the most basic public equality to homosexuals," Sullivan pointed out. "Homosexuals are still systematically discriminated against by the state in the military, in the law, and in marriage rights. By first emphasizing discrimination by private citizens, or even by emphasizing it at all, liberals actually undermine the strength of their argument. By inference, they have ceased to focus on the most pressing discrimination of all—that of the government."

In the case of Matthew Shepard, the crime was portrayed as two homophobes so angry over a stranger at a bar hitting on one of them that they lured him outside, beat him savagely, set him on fire, and then left him for dead, tied to a fencepost. But the act-of-hate story never quite added up. Why did police insist that the perpetrators' primary motive was robbery? Why did the allegedly violently homophobic perpetrators supposedly pretend to be gay? And if hate crime protections were needed to stop horrific acts like this from happening, why did Wyoming have no trouble convicting the men of first-degree murder and sentencing them to life in prison?

For many years, such questions were largely ignored. Activists, politicians, and media benefited from turning Shepard's story into a simple, lurid tale of bigotry turned violent. Eventually, in 2009, this narrative led to the passage of a  federal law bearing Shepard's name that further expanded the definition of hate crimes, the role of federal law enforcement in policing and prosecuting such crimes, and the punishments possible.

In The Book of Matt: Hidden Truths About the Murder of Matthew Shepard, published in 2013, author Stephen Jimenez—a gay man and respected investigative journalist who had no ideological reason to oppose the original narrative—suggests it was drugs, not homophobia, that fueled the murder. In the course of research for a screenplay about Shepard's death, Jimenez discovered that the young man had had a crystal meth habit and had been dealing the drug. He also already knew one of his assailants, McKinney, with whom he had allegedly had previous sexual encounters. And both men had reportedly dabbled in gay sex work.

"It was fairly well known in the Laramie community that McKinney wouldn't be one that was striking out of a sense of homophobia," Flint Waters, one of the officers on the Shepard case, told The Guardian in 2014. "Some of the officers I worked with had caught him in a sexual act with another man, so it didn't fit—none of that made any sense."

At the time of the murder, McKinney reportedly knew that Shepard had or was about to receive a shipment of meth worth $10,000.

From the beginning, Shepard's parents asked that their son's death not be made into a political cause. "We should not use Matt to further an agenda," Dennis Shepard reportedly told Wyoming's governor at the time. "Don't rush into just passing all kinds of new hate crimes laws. Be very careful of any changes and be sure you're not taking away rights of others in the process to race to this."

But Shepard's murder took place in a rural Western state without a hate crime enhancement on the books. Even though Laramie, home to the University of Wyoming, was a relatively liberal area, the idea of hillbilly homophobes was an easy sell nationally. The original story was more useful to mainstream activists than a gay sex worker and drug dealer murdered by a greedy associate.

Perhaps Shepard's death wasn't a tragic illustration of the need for new hate crime laws. Perhaps it was a tragic case of yet another victim of drug prohibition. What was more culpable here—what the government didn't do, or what it did?

This retelling casts a new light on the Atlanta spa shooter case. As with Shepard's murderers, no hate crime enhancement was necessary to serve a severe punishment to the killer. In July, Long pleaded guilty to four of the murders. (The others are being tried in a different county, where prosecutors are trying it as a hate crime.) He received four life sentences without parole.

"This was not any kind of hate crime," reiterated Cherokee County District Attorney Shannon Wallace at the time of Long's plea. Yet slotting the killings into a narrative about anti-Asian hate and COVID-19 gave both parties an easy way to condemn the violence and to purport to do something about it (while giving Democrats a bonus way to lash out at Trump and Republicans). The findings section of the 2021 COVID-19 Hate Crimes Act mentions victims of the Atlanta spa shooter by name, then uses this as partial justification for giving more money to police, to federal programs to raise "awareness of hate crimes during the COVID-19 pandemic," and to anti–hate crime hotlines, as well as for allowing courts to impose more supervised release requirements on people convicted of such crimes.

Unexamined by most mainstream coverage of the killings was the way that government scrutiny of Asian massage parlors puts workers at these places at more risk and fuels narratives in which staff members are assumed to be sex workers or even sex slaves. Activists have actually pushed local ordinances to disallow precautions such as locked doors during business hours, buzzing customers in, or using multiple entrances and exits—things that could help protect staff and clients alike—under the theory that these measures facilitate sex trafficking. Meanwhile, overzealous vice cops and Homeland Security agents mean immigrant massage and sex workers who do face harassment or violence from customers may be afraid to report them. (In one recent case, after a massage parlor owner helped law enforcement bring down a stalker and suspected serial killer, the FBI went after her for sex work allegedly happening at her business.) Yet nothing in the COVID-19 Hate Crimes Act will help decrease stigma and violence against sex workers. To the extent that it encourages more policing of these places, it could actually increase danger for people like Long's victims.

Progressives and others hail hate crime laws as a way to protect vulnerable groups, yet they often end up being used to entrench power and dominance. And the statistics used to justify hate crime laws tell a much more nuanced—and even optimistic—story than "hate crime epidemic" narratives suppose. There may be reasons for worry, but there are also reasons for hope. America hasn't eradicated bias-based crime and probably never will, but a holistic look at the data suggests a nation growing more tolerant over time.