The “Myth” is a Myth: the Bad Data Behind Deniers of Systemic Racism in Policing
On June 2, 2020, The Wall Street Journal published an op-ed by Heather MacDonald entitled, “The Myth of Systemic Police Racism.” I wouldn’t normally lend any credence to something with that title, but it was forwarded to me by relatives and I felt the duty to debunk it. I’ve expanded my reply email here to provide the research, evidence, and links in one place for use by all thinking people.
An outspoken pro-police commentator, Ms. MacDonald is transparent about her pre-formed opinion. She’s written books and columns decrying criminal justice reform, and has spoken out against the Black Lives Matter movement specifically. I have not read nor listened to Ms. MacDonald outside this latest WSJ column; however, if this writing is indicative of her approach to evidence-based debate, she has no credibility in my eyes. Ms. MacDonald’s thesis for her column is “there’s no evidence of widespread racial bias” in policing. I will not only show that that absolutist statement is patently false, but I will also demonstrate the bias that exists in Ms. MacDonald’s own writing and, by extension, the outlets that publish her.
Who is Armed and Who is Dangerous?
The introduction to the column contains ostensible background, referencing the “Obama era,” as if the 2017 change in the Administration made the lived realities of millions of Americans vanish. This type of introduction is unsurprising in an op-ed piece, but it’s clear that MacDonald is raising the specters of Obama and Biden as dog whistles to conservatives, priming them to view her argument favorably.
But Ms. MacDonald and the WSJ purport to ground their opinions in data and evidence, so I will try to stick to analyzing the statements MacDonald presents as fact.
MacDonald says, “In 2019 police officers fatally shot 1,004 people, most of whom were armed or otherwise dangerous.”
MacDonald takes her data from The Washington Post’s exhaustive work on the Fatal Force Database, which amasses more comprehensive data than the FBI, and forms a partially true statement. 774 of the suspects shot were armed with guns or knives — indeed “most.” But the data could just as easily be couched this way: “In 2019, police officers fatally shot 230 people who were unarmed or holding toys.” Her focus on armed encounters erases over 20% of these homicides by police.
MacDonald also uses the phrase “otherwise dangerous,” an unsupported, subjective term that means nothing in this context. A body of evidence has demonstrated that people view black men as “dangerous” simply because of their race and gender; for example, this 2017 study published in the Journal of the American Psychological Association. Lead author John Paul Wilson, PhD, of Montclair State University said:
“Unarmed black men are disproportionately more likely to be shot and killed by police, and often these killings are accompanied by explanations that cite the physical size of the person shot. Our research suggests that these descriptions may reflect stereotypes of black males that do not seem to comport with reality.”
It is important to note that of all people shot and killed by police in 2019 who were categorized specifically as “unarmed,” 100% were black.
MacDonald says, “African-Americans were about a quarter of those killed by cops last year (235), a ratio that has remained stable since 2015.”
Black people make up 13% of the U.S. population, so the fact that they make up 25% of those killed by police is itself an indicator of racial bias. This statistic does the opposite of what she intends, undermining her argument by demonstrating that in the past 5 years, police departments have made no headway in diminishing their racial bias.
MacDonald attempts to mitigate the toothlessness of her evidence by saying:
“That share of black victims is less than what the black crime rate* would predict, since police shootings are a function of how often officers encounter armed and violent suspects.”
This is an obviously specious assertion. Firstly, to make this a valid statement, you would first have to assume that police intervene in crime at a rate consistent with demographic ratios. You also have to assume that police intervene in crime at a rate consistent with the occurrence of such crimes. As it is documented, “crime rate” is at best incomplete, and at worst a fiction.
“Crime Rate” Isn’t Real
Racially Biased Policing Creates a Disproportionate Number of Interactions Between Police and Black People
Blacks are 2.7 times more likely to be pulled over in an investigatory stop. Courts shut down “stop and frisk” because police were targeting minorities. In 2012, 53% of people stopped by NYPD cops were black, and only 10% were white. NYC is 25% black and 33% white. Even when you controlled for crime rate*, precincts that have higher populations of people of color had disproportionately higher rates of racial bias. The racial bias found in the “stop and frisk” policy is well documented; even though the policy has been stopped, the shadow of its influence remains.
Overpolicing, especially “Broken Windows” style quality-of-life interventions, is an increasing issue affecting communities of color, that coincides with and correlates to gentrification and the displacement of people of color. Anecdotal examples have gone viral:
- Alison Ettel (aka #PermitPatty), a white woman, called police because a black girl was “illegally selling water.”
- In Rialto, California, an unidentified white woman called the police on three black black people exiting an AirBnB because they did not smile or wave at her; at least three police cars and a helicopter responded.
- At Lake Merritt in San Francisco, an unidentified white woman called the police on a group of black men because they were using a charcoal barbecue in an area designated for non-charcoal barbecues.
- Most recently, Amy Cooper called the police on a black man birdwatching in Central Park, telling them “there’s an African American man threatening my life.” Another black birder, Walter Kitundu, has had the police called on him so many times that he posted signs about it in his neighborhood.
Beyond the anecdotes, though, the trend has been studied and the evidence of racial bias in quality-of-life interventions is abundant. In 2015, Dr. Joscha Legewie, a sociologist at New York University, and Dr. Merlin Schaeffer of the University of Cologne, conducted a broad analysis of 311 quality-of-life complaint calls in New York City to study the impact of racial demographics on neighborhood strife. In 2018, the Community Service Society of New York conducted a study of over 108,000 311 calls for “quality-of-life complaints” from neighbors against neighbors that resulted in police involvement. Buzzfeed News conducted a similar study the same year, analyzing calls back to 2015 for a specific geographic area.
All these studies, and others, concluded that racial demographics — especially changing demographics — are a significant factor in discord among neighbors. White people call the police on black people who look “out of place” in white neighborhoods, and white people call the police on black people who live in the “fuzzy” borders between white and black neighborhoods. The racial segregation of neighborhoods in American cities that contributes to this strife is itself an outcome of system racism, going back to redlining and earlier. The net result is a disproportionate number of encounters between blacks and police.
Related to overpolicing is proactive policing. In Proactive Policing, a Consensus Study Report out of the National Academies of Sciences, Engineering, and Medicine, the authors state, “proactive policing strategies by design increase the volume of interactions between police and the public, such strategies may increase the overall opportunity for problematic interactions that have disparate impacts.” (In the same chapter, the authors also cite bountiful evidence that demonstrate racial bias in the perception of suspicion, use of force, and the decision to shoot.)
Quota Culture Exacerbates Negative Outcomes From Interactions With Police
A byproduct, or perhaps simply product, of the increased police interaction with communities of color is an increase in citations and arrests, which is due in large part to the culture and demands of quota systems that have been, or still are, employed in police precincts across America. While they’re deployed in predominantly black neighborhoods, police are encouraged, and even mandated, to arrest and ticket as many people as possible to make quotas. This video shows NYPD police officers discussing the practice of “hunting” the “vulnerable” to rack up their numbers.
All of this is to say that the idea of “crime rate,” often cited to be higher among black people, is an inherently unreliable data point, based on law enforcement reports which reflect only crimes that are reported, taken seriously, recorded, and not subsequently canceled, and victim surveys that rely on individual memory and honesty. Further, the “crime rate” discussion centers almost entirely around so-called “street crimes,” often ignoring white-collar crimes that are near exclusively the purview of rich, powerful whites.
Crime rate is a direct result of police involvement; this circular logic taints it as a useful measure when analyzing police involvement itself.
Secondly, MacDonald’s assertion that “police shootings are a function of how often officers encounter armed and violent suspects” implies that every person shot by police is an armed and violent suspect. This is not true, as discussed above. Over 20% of police shooting victims in 2019 were unarmed. Breonna Taylor was asleep in her bed when police shot her eight times in March 2020; no suspect was even present at the scene.
Error 404: Data Doesn’t Exist
MacDonald says, “In 2018, the latest year for which such data have been published, African-Americans made up 53% of known homicide offenders in the U.S. and commit about 60% of robberies, though they are 13% of the population.”
By this point in her column, MacDonald is starting to give up even the pretense of objective data analysis. Using the weasel word “known” to describe homicide offenders excludes cases for whom police failed to complete their job. With this in mind, it’s just as (il)logical to ask why aren’t police solving more homicide cases with white suspects?
Saying blacks “commit about 60% of robberies” is just lazy. The statement implies that all robberies are reported and all robbery suspects are caught and determined guilty. That’s obviously impossible. The statement is just as absurd as saying “Law enforcement was only able to catch and convict 40% white robbery suspects, though they are 60% of the population.”
Fruit of the Poison Tree
MacDonald says: “The police fatally shot nine unarmed blacks and 19 unarmed whites in 2019, according to a Washington Post database, down from 38 and 32, respectively, in 2015. The Post defines “unarmed” broadly to include such cases as a suspect in Newark , N.J. , who had a loaded handgun in his car during a police chase.”
You can look at the database for yourself and see that MacDonald is making false statements again. The Post does not “define” unarmed. The Post reports on fatal shootings as reported by other news outlets, typically citing two different sources. The original stories define “armed” vs. “unarmed;” if MacDonald wants to argue this, she has to argue with the journalistic standards of news outlets from all 50 states. Also by looking at the database for yourself, you can see all the ways you can manipulate the data, including the way MacDonald chose to. MacDonald talks about “unarmed” as a distinct and complete category, but she is excluding the categories of “vehicle,” “toy,” “other,” or “unknown.” Politifact researched this exact statistic as it was quoted by Larry Elder on Twitter, and found it Mostly False because of how misleading it is. MacDonald goes on to extrapolate from this bad data, and as the saying goes, it’s the fruit of the poison tree.
Incidentally, in the story of the NJ shooting she cites as anecdotal evidence, whether the shooting officer knew that a gun was present in the vehicle is under hot debate in a courtroom setting. The officer, who has been indicted for manslaughter but not tried, maintains that the suspect pointed a gun at him, but the prosecutor argues it could not have been seen through the dark tinted windows of the vehicle. Body and dash cam footage do not corroborate the officer’s story. The officer shot both driver and passenger in the vehicle. Both were black men.
Just as MacDonald’s purported evidence is fruit from the poison tree of her poor analysis of incomplete data, crime statistics for black Americans — from rates of school suspensions, to reports, to interventions, to arrests, to convictions, to incarcerations — are the fruit of a poison tree rooted in 400 years of systemic racial injustice and the blood of black bodies.
Black Lives Are People; Blue Lives Are a Job
At this point in the column, MacDonald turns to the tired trope of “black on black crime,” using her bad data to support an argument she is failing to make. To paraphrase Trevor Noah, if you think blacks don’t care about black-on-black crime, you’ve never been to the hood.
MacDonald says, “A police officer is 18½ times more likely to be killed by a black male than an unarmed black male is to be killed by a police officer.”
I’m going to step out of my narrative voice for a moment. THIS IS A BONKERS THING TO SAY.
First of all, police officers are more likely to be killed by another person in the course of their job than anyone in any other job; that’s the job they signed up for. But their job isn’t that much more dangerous than working at Wal-Mart, and it certainly is not the most dangerous job there is. But, and this shouldn’t even have to be said, being a black person is not a job. By even making this comparison, MacDonald is saying that just existing in the world as a black male puts you at a risk of homicide comparable with the risk police officers take when entering the force. And she’s right: homicide by police is the number one cause of death of black men and boys. But black people don’t take off a uniform at the end of a shift, and comparing a black life to a job is unacceptable.
Second of all, MacDonald doesn’t provide a source for this wild statistic, which is obviously, provably, and enormously wrong. I tried to find a source, but all I could come up with were quotes from MacDonald herself.
She seems to be saying that for every unarmed black man killed, 18+ police officers are killed by black men. In 2018 (latest year available), 33 police officers were killed by another person in the line of duty. Even if every single one of those homicide suspects was black (they aren’t), her math would indicate that police killed just one unarmed black man in 2018. At least 16 unarmed black men were killed by police in 2018. (That number could be as high as 65, depending how one defines “armed” when deadly force is used. One man had a lawnmower blade. Another had a baseball bat. Another had a “pole.”)
Police have shot to death 5–20+ unarmed black men in the first half of 2020. 21 police officers have been shot in the same time period; 9 of the shooting suspects were black men.
The Black-on-Black Deflection
MacDonald then goes on in her column to recount the homicides of 31 residents of Chicago, another racist dog whistle for her followers. “What about Chicago” is a common, and tired, deflection in the discussion about police violence. Closer examination of the point actually demonstrates the contradictory problem of underpolicing in black communities. Why aren’t we asking whether high crime rates are a result of poor police performance? From the Wall Street Journal itself:
[P]olice are better at stopping African-Americans at random than at halting an epidemic of murder.
Concluding her analysis of two weekends in Chicago,
MacDonald says: “Police shootings are not the reason that blacks die of homicide at eight times the rate of whites and Hispanics combined; criminal violence is.”
First of all, no one is making an argument that more blacks die by homicide because of police. This is a classic example of the special pleading fallacy, commonly known as “moving the goalposts.” The argument at hand is about the disproportionate and unnecessary killing of people of color during interactions with the police. I don’t think MacDonald is trying to make the point that black people deserve to be killed by the police because they’re already killing each other. At least I hope not.
Second of all, this is another statistic for which she does not provide a source, and the only instances I can find of it are, again, quotes from MacDonald herself. Whether or not this is an accurate number, MacDonald’s statement has racist implications: that something about being black makes you more likely to be involved in criminal violence. The truth is much more complicated, and actually rooted clearly structural causes, particularly economic disparity.
It’s also interesting to note that she uses the term “criminal violence” in reference to civilian homicide. When police kill innocent people, no one refers to it as “criminal violence,” even when the officers are, rarely, charged with and convicted of a crime. Police are given cover by law to act violently without repercussion. This is what we mean by “systemic.”
Cherry-Picking Writ Large
Next, MacDonald delves into what she calls “a series of studies undercutting the claim of systemic police bias.” She cites three studies; none of them undercut the claim of systemic police bias, and in fact reach the opposite conclusion.
The main findings of the first study she mentions are as follows (emphasis added):
1) As the proportion of Black or Hispanic officers in a FOIS increases, a person shot is more likely to be Black or Hispanic than White, a disparity explained by county demographics; 2) race-specific county-level violent crime strongly predicts the race of the civilian shot; and 3) although we find no overall evidence of anti-Black or anti-Hispanic disparities in fatal shootings, when focusing on different subtypes of shootings (e.g., unarmed shootings or “suicide by cop”), data are too uncertain to draw firm conclusions. We highlight the need to enforce federal policies that record both officer and civilian information in FOIS.
This isn’t exactly the clear cut myth-bust that MacDonald claims. MacDonald quotes from the study’s synopsis to say “no significant evidence of antiblack disparity in the likelihood of being fatally shot by police” but provides no context or analysis. The authors also state (emphasis added):
“Examination of National Violent Death Reporting System data shows racial differences across types of fatal shootings. Black civilians fatally shot by police (relative to White civilians) are more likely to be unarmed and less likely to pose an immediate threat to officers. In contrast, White civilians (relative to Black civilians) are nearly three times more likely to be fatally shot by police when the incident is related to mental-health concerns and are seven times more likely to commit ‘suicide by cop.’ ”
Did MacDonald read the full report, or did she just stop at the line she agreed with?
The next study she cites concludes:
“Our assessment uncovered policy, training, and operational deficiencies in addition to an undercurrent of significant strife between the community and department. It yielded 48 findings and 91 recommendations for the department to reform its deadly force practices.”
MacDonald states that “white police officers were less likely than black or Hispanic officers to shoot unarmed black suspects.” From the report, this is factually correct. The study states:
“Our analysis shows that the suspects in officer-involved shootings were overwhelmingly Black. That same pattern was apparent in unarmed persons shot by the PPD. Our analysis also shows that threat perception failures (TPF) occur with suspects of all races. Black suspects have had the highest TPF rate (8.8 percent), more than twice the rate of White suspects (3.1 percent). It is clear that the Black community is disproportionately impacted by extreme violence involving the police.”
Aside from the fact that race of police officers is not party to the overall argument of systemic racism, this statement actually supports the conclusion of widespread systemic racism. The fact that police officers of every race — including police officers of color— disproportionately target black citizens is not evidence against systemic racism, but rather evidence for it. Furthermore, racism against blacks is not and has never been the sole domain of whites, especially among police.
Finally, MacDonald alludes from research from Ronald G. Freyer, but does not quote it or link to it. I assume she is referring to his 2017 article “An Empirical Analysis of Racial Differences in Police Use of Force.” MacDonald only paraphrases Freyer’s conclusions on officer-related deaths, and leaves out Freyer’s conclusion that (emphasis added):
“On non-lethal uses of force, blacks and Hispanics are more than fifty percent more likely to experience some form of force in interactions with police. Adding controls that account for important context and civilian behavior reduces, but cannot fully explain, these disparities…. [E]ven when officers report civilians have been compliant and no arrest was made, blacks are 21.2 percent more likely to endure some form of force in an interaction.”
MacDonald concludes her paragraph by saying “Any evidence to the contrary fails to take into account crime rates* and civilian behavior before and during interactions with police.” Freyer explicitly controlled for these variables in his research, as emphasized above. Plainly, MacDonald has selectively chosen evidence that supports her argument, and attempts to cast doubt on — or outright deny the existence of — data that doesn’t, even when they’re from the same exact study.
“In Conclusion, I’m Wrong About Everything”
Concluding her column with a return to the dog whistle of “Obama,”
MacDonald says: “The false narrative of systemic police bias resulted in targeted killings of officers during the Obama presidency. The pattern may be repeating itself.”
Instances of police killings, including by ambush (“targeted killings”) fell continuously from 2011 to 2015, rose temporarily in 2016, then began to fall again in 2017. But to put “targeted killings” into context, criminal justice reporter Mark Berman said in The Washington Post (emphasis added),
“These figures are subject to significant year-over-year fluctuation in part because these incidents are so rare. When only 10 or 20 incidents happen in a given year, as with ambush killings, it’s easy for one or two outliers to drastically affect the numbers.”
Overall, policing during the Obama administration was safer than it had ever been. That is a pattern worth repeating, isn’t it?
On November 20, 2016, three police officers were targeted and shot in three apparently unrelated ambush attacks within one 12-hour period. To be fair, this high-profile day in Obama’s last month as President may have skewed MacDonald’s perception of the previous 8 years.
MacDonald says: “If the Ferguson effect of officers backing off law enforcement in minority neighborhoods is reborn as the Minneapolis effect, the thousands of law-abiding African-Americans who depend on the police for basic safety will once again be the victims.”
Well, Minneapolis is abolishing their police by choice; they don’t have to wait for the threat of a racist “Ferguson effect” to come to fruition. The “Ferguson effect,” by the way, is a concept that was popularized by MacDonald herself, but never proven to exist. The real “effect” was a $65 million investment in Ferguson and the surrounding community, the curtailing of predatory and discriminatory fees and fines, the hiring of a diverse police department, and the election of reform-minded politicians.
In concluding her column,
MacDonald says, “But Floyd’s death should not undermine the legitimacy of American law enforcement, without which we will continue on a path toward chaos.”
It’s 2020, and we are very much on a path toward chaos. But that chaos is not because of Black Lives Matter; in fact, the Black Lives Matter movement against systemic police racism is one of the things bringing the world together. Anti-BLM stances are increasingly becoming fringe.
The “myth” that MacDonald is hawking is itself a myth. She has made a career of speaking and selling books to audiences that already agree with her, and are too biased or too lazy to question her bad research, bad data, and bad conclusions.
WSJ says of their opinion columns, “While we aim to persuade, every word we publish is the product of rigorous reporting, research and debate.” Ms. MacDonald’s work did not meet these standards and should not have been published. It’s interesting to note that the WSJ Opinion page is meta-tagged “Guided by Principle; Grounded in Fact,” but the page itself says “Guided by Principle; Grounded in Purpose.” By publishing MacDonald’s propaganda, they have demonstrated that they’ve left fact behind in favor of purpose. And their purpose here seems to be to make their readers believe that black lives don’t matter.
They do. Black Lives Matter.
