Sociology

20 contexts.org

BLACK LIVES

AND POLICE TACTICS

MATTER

Do police

provide a

public good or

do they perpetuate

racial inequality?

Like most

institutions, they do

both. Nonetheless,

thanks to the tragic

deaths of Michael

Brown, Eric Garner,

Rekia Boyd, Tamir Rice,

and hundreds of others,

this blunt question

remains at the heart

of political debate.

Social movements

like Black Lives

Matter (BLM)

and Say Her

Name, focused on

civilian deaths at

the hands of cops,

have changed national

discourse around what cops

should be doing—and how they

should do it.

by rory kramer, brianna remster and camille z. charles

21S U M M E R 2 0 1 7 c o n t e x t sContexts, Vol. 16, No. 3, pp. 20-25. ISSN 1536-5042, electronic ISSN 1537-60521. © 2017 American Sociological Association. http://contexts.sagepub.com. DOI 10.1177/1536504217732048

Of course, social movements inspire responses. Two slo-

gans, in particular, are especially indicative of negative reactions

to BLM: “All Lives Matter” and “Blue Lives Matter.” The first

assumes that BLM suggests that Black lives matter more than

others, while the latter asserts that critiques of the police, such

as those lodged by BLM, have created a backlash—sometimes

called the “Ferguson effect” in reference to Mike Brown’s death

in Missouri—that feeds mistrust and hampers policing, incites

violence against police officers, and makes everyone less safe.

If the core divide between Black Lives Matter and All/Blue Lives

Matter is whether Black people are more likely to experience

violence at the hands of police than other Americans, then we

need to know the answer.

As sociologists of race and crime, we were surprised to dis-

cover that there have been few systematic analyses of disparities

in police use of force. Among those, none examines the specific

intersectional claims made by activists on both “sides” of the

issue. One big reason is a lack of good data—government data

and crowd-sourced data like that produced by the newspaper

The Guardian’s project, “The Counted”, consider only deaths

or shootings at the hands of the police. Without a comparison

group (such as non-fatal police encounters) and without adjust-

ing for completing explanations for variations in victimization, it

is hard to draw any meaningful conclusions. Fortunately, there is

an alternative data source: New York Police Department (NYPD)

records all investigatory civilian stops, regardless of whether or

not they involve the use of force. We harness these city-wide

data to investigate whether police use more force against Black

individuals (particularly Black youth) than Whites and whether

and how gender factors into this equation.

studying police encounters Despite sustained outrage around police use of force on

Black bodies, research has lagged. For instance, just because

“The Counted” finds that Black people were more likely than

Whites to be killed by police in 2016, it does not necessarily

mean that police are racially biased. Perhaps the most common

justification police offer for the overrepresentation of Black

victims in use of force incidents is that it follows from racial

disparities in crime rates. In this scenario, if police use of force

is random across civilian interactions, but Blacks commit more

crime, any racial disparities in shootings could very well be the

result of Blacks’ greater contact with police. It’s a simple, if

stereotypical and damaging, narrative, and it shows why it is

imperative that any estimates of disparities in police violence

adjust for crime rates by race.

Further, in each of the most infamous recent deaths in con-

nection with law enforcement, there have been other aspects

of the interaction that proponents of All Lives Matter or Blue

Lives Matter wield as their explanations for why the situation

ended in violence. In Missouri, officer Darren Wilson described

Michael Brown as being so physically large that his body alone

was construed as a deadly weapon; New York City’s Eric Garner

was accused of being non-compliant as officer Daniel Panteleo

attempted to handcuff Garner, thus Panteleo had to use what

would become a fatal chokehold; and Freddie Gray was stopped,

beaten, and arrested in a high-crime, high-poverty, and predomi-

nantly Black neighborhood—just the kind of neighborhood the

Blue Lives Matter folks insist are particularly dangerous to police.

An attendee at a January 2015 End the Madness/Sea of Blue march in Washington, D.C.

El ve

rt X

av ie

r B ar

n es

, el

ve rt

b ar

n es

.c o m

22 contexts.org

Importantly, any explanation is post-hoc; it may be true, it may

be false. After all, North Charleston police officer Michael Slager

insisted that Walter Scott had been threatening his life, but video

evidence shows Scott being shot while running away. If even

after adjusting for all of these potential explanations, including

any civilian behavior that inspired the stop, where and when

the stop occurred, local crime rates, and neighborhood poverty

and racial demographics, we still find evidence of disparities

in police violence, then BLM’s critiques of over-policing and

over-use of force against Black Americans would gain powerful

empirical support.

We are not the first to use police records to study BLM

critiques, but data limitations prevent most from accounting for

more than one or two competing explanations. As an excep-

tion, economist Roland Fryer made quite a media splash when

he found no evidence of racial bias in police shootings (though

he did find evidence of bias in lesser forms of police violence).

But in his analyses, Fryer compared police shootings to police

encounters in which the officer could have been justified in

shooting but did not (such as when a civilian is resisting arrest).

Sociologists know, however, that such arrests are themselves

racialized: officers are more likely to perceive people of color as

aggressive or noncompliant compared with Whites and, as such,

are more likely to arrest them on such a charge. In short, Fryer

obscured racial bias in shootings by using a biased comparison

group in that part of his project. The question remains: are there

racial disparities in police use of force, even after we consider

justifications?

data strikes back As part of a 1999 settlement agreement in Daniels et al v.

City of New York, New York City agreed to make public detailed

data on all civilian stops executed by NYPD officers under its

highly-contested stop-and-frisk policy. These reports, based on

officers’ own reports, are the best available data for studying

racial disparities in policing. In fact, their release contributed to

a federal judge’s 2013 ruling that New York City’s stop-and-frisk

policy was racially discriminatory. Thereafter, the NYPD dramati-

cally reduced the number of its civilian stops and revised training

and policies toward more civil, less confrontational community

interactions. The changes mean we can now see whether police

tactics and training matter in reducing disparities in police vio-

lence. Or, to be a bit more wonky, can organizational change

reduce victimization inequalities in police officers’ use of force?

With data from 3.3 million NYPD investigative pedestrian

stops conducted between 2007 and 2014, supplemented by U.S.

Census and crime data provided by the New York City Division

of Criminal Justice Services, we directly tested whether policy

and training changes affected associations between race, stops,

and violence. In addition, as BLM makes intersectional claims

cr e

C o u rt

es y

A va

D u ve

rn ay

/N et

fl ix

F ilm

s b ey

o n d 1 3 th

.c o m

Training reforms include efforts to counter cops’ implicit biases.

C o u rt

es y

A va

D u ve

rn ay

/N et

fl ix

F ilm

s b ey

o n d 1 3 th

.c o m

23S U M M E R 2 0 1 7 c o n t e x t s

that Black children (like 12-year-old Tamir Rice who was shot to

death by Cleveland police) are more likely to experience police

force than are White children and that Black women (like Korryn

Gaines, a Maryland mother shot by police after earlier violent

interactions) are more likely than White women to endure the

experience. Since deaths at the hands of police are rare (in 2013,

NYPD shot and killed a total of 8 people), we examined incidents

in which police pointed a firearm at a civilian, which occurred

in less than one percent of all stops (police are trained to only

draw their guns when prepared to use lethal force). To broaden

our lens as much as possible, we then considered any use of

force by police, ranging from “hands on suspect” to “pointing

firearm at suspect.” We then adjusted our estimates for every

justification for police use of force put forward to date.

targeting young black bodies We find gaping inequalities in police violence. Black people

are more likely than Whites to be subject to police violence—

regardless of whether they are doing anything illegal, the time

of day, their height, their age, their gender, their behavior dur-

ing the stop, the police justification for the

stop, neighborhood socioeconomic and

racial characteristics, the precinct’s stop

rate, or local arrest rates broken down by

race. In fact, the majority of stops in which

police used force against Black individuals

did not lead to arrest or the discovery of

any contraband.

Even as scholars well versed in racial disparities in criminal

justice contacts, we were surprised by the sheer number of

young Black children in our data. On average, more than seven

Black children aged 10-13 were stopped by police in New York

City each day between 2007 and 2014 (approximately 20,000

stops). More Black teens experienced police violence during a

stop than did Whites of all ages combined.

When we add controls, including teens’ behavior during

police stops, when and where they were stopped, and whether

or not they were arrested, the racial disparity in police violence

persisted. We have long known that young Blacks are more likely

to be stopped, but our results show that this disparity extends

to police violence. The first figure above plots the association

between race, age, and the likelihood of experiencing force

once stopped across our entire sample. Overall, and at any age,

Black individuals stopped by police are more likely to experience

force (including an officer drawing and/or pointing a gun). The

racial gap is largest among youth: stops of Black youth are

more likely than those of White youth to include force, though

the effect shrinks, albeit only slightly, with age. Although we

don’t know for sure why the likelihood of experiencing police

violence declines with age, it may be that Black men have simply

had more time to learn how to navigate police encounters and

decrease the likelihood of enduring police violence. Importantly,

these results hold even after adjusting for police justifications

and a host of variables in stops.

To highlight the experience of Black youth, the second and

third figures show the predicted probability that police will use

force or draw a firearm during a stop with underage (p. 24, left)

Black and White civilians (ages 10-17), split by gender (again,

this data holds after adjusting for the circumstances of the stop).

Consistent with work by social psychologists, such as Philip Goff

who has shown that Black youth are perceived as older and

more threatening than White youth of the same physical size

and age, we find bias. The figure above (at right) shows that

Black youth are 41% more likely to be victims of any police use

of force than comparable White youth. Sadly, the patterns are

similar when we predict stops in which officers pull their guns

and are prepared to shoot—Black youth are more than 50%

more likely than White youth in similar stops to find themselves

staring down the barrel of a gun. So Tamir Rice and Tyre King,

aged 12 and 13, are not tragic outliers, but important examples

Overall and at any age, Black individuals stopped by police are more likely to experience force.

5

10

15

20

25

50+45-4940-4435-3930-3425-2920-2418-1916-1714-1510-13

Predicted probability of police using force for all individuals stopped, by race and age

black

white

Note: Sample includes 1,766,743 Black individuals and 311,308 White individuals stopped by NYPD between 2007-2014.

percent

0

5

10

15

20

25

Predicted probability of police violence for children stopped by police, by race and gender

black white

Note: Sample includes 277,992 Black males, 21,844 Black females, 38,898 White males, and 4,718 White females aged 10-17 stopped by NYPD from 2007-2014.

percent

male female

24 contexts.org

of a systematic pattern of racial inequality in police violence.

We also tested Say Her Name activists’ critique that Black

women are more likely than White women to experience police

violence and that such violence is especially unlikely to be cov-

ered in debates about police violence. We have shown predicted

probabilities for girls ages 10-17, but the findings hold across all

age groups in our sample. Although women overall experience

fewer stops than men, Black women are more likely to experi-

ence police use of force—including having a gun drawn on

them—during a stop than White women, even after accounting

for alternative explanations. And Black girls’ predicted probability

is about the same as White boys’ (15% to 14% respectively).

Within race, gender is not associated with a difference in police

drawing guns during a stop (see above); in other words, it’s

about as likely that a Black woman or a Black man will have

a gun drawn on them, and the same holds for Whites. Taken

together, we find that gender influences whether force is used

with adults, but does not affect whether police draw and/or aim

their guns at children.

police tactics matter The history of policing is intimately linked to the history of

racial subjugation. Even though today’s police officers and those

in leadership roles are more diverse than even a few decades ago,

police—and citizens—have not escaped the institution’s legacy

of racially disparate violence. Not only do police stop people at

different rates according to their race, age, and gender, but they

also behave differently during otherwise similar interactions.

There is scant evidence to support stop-and-frisk as a

crime-fighting strategy in New York City. As Mayor Bill De

Blasio tweeted, “We’ve reduced stop-and-frisk by 97% in New

York City and crime is down significantly. Not everyone knows

that.” Moreover, we find that when the NYPD reduced the use

of stop-and-frisk and instituted changes in officer training in

late 2013—due in part to the policy being ruled unconstitu-

tional—police used force less frequently in all stops. Still, the

racial disparity in who is subjected to force did not shrink. The

figure above (at right) shows the Black-White disparity in use

of any force and in gun use in the periods before and after

stop-and-frisk reform. The racial disparity in police violence of

any type grew after the NYPD’s 2013 reform. This is because

use of force declined more dramatically for

Whites than Blacks, widening the violence

gap even as violence, overall, decreased in

police interactions. In better news, there

are hundreds of thousands fewer stops

made by NYC cops, the rate of force has

declined for both Black and White civilians,

and, perhaps most importantly, the 2013 reforms shrank racial

inequities in police gun violence. But that race is no longer a sig-

nificant predictor of whether a NYPD officer will draw a firearm

in a citizen interaction owes to a change in how and how often

NYPD officers stop civilians, rather than to any “disappearance”

of racial inequality in policing and use of force.

Given these findings, there is no good reason for any police

department to pursue stop-and-frisk. There is growing evidence

that it sours public perceptions of police, especially in commu-

nities of color, which have been treated violently by powerful

institutions since the very birth of the country.

what now? Conversations need to shift from whether there are racial

disparities in policing to what can be done to reduce or eliminate

racial disparities in policing. Our findings suggest that policy

changes such as New York City’s discontinuation of stop-and-

frisk practices and the revision of police training courses are

promising avenues, at least with regard to disparities in gun use

and in reducing the sheer quantity of force used. Because police

force is not restricted to investigatory pedestrian stops, police

0.00

0.05

0.10

0.15

0.20

0.25

0.30

Predicted probability of police drawing a gun for children stopped by police, by race and gender

black white

Note: Sample includes 277,992 black males, 21,844 black females, 38,898 white males and 4,718 white females aged 10-17 stopped by NYPD from 2007-2014.

percent

male female

5

10

15

20

25

30

Black-White disparity in police force during a stop before and after reform

any force gun use

Note: Figure based on sample of 3,097,893 stops of people of all racial backgrounds before 2013 and 205,283 stops after 2013. Of those, 1,653,559 Blacks and 289,754 Whites were stopped between 2007 and 2012 and 113,184 Blacks and 21,554 Whites were stopped in 2013 or 2014.

percent

pre 2013 post 2013

Tamir Rice and Tyre King, aged 12 and 13, are not tragic outliers, but important examples of a systematic pattern.

25S U M M E R 2 0 1 7 c o n t e x t s

training and policy also deserves a much broader reconceptu-

alization; currently, officer training modules prioritize weapons

training over de-escalation techniques. However, law enforce-

ment agencies like those in Sanford, FL (where Trayvon Martin

was killed) are beginning to incorporate implicit bias training pro-

grams (to help counter cops’ sense of Black civilians, particularly

young, Black males, as threatening), while others have adapted

their training and policies to emphasize de-escalation tactics.

These steps are vital toward reducing force as a primary tool

of policing. Holding police departments and individual officers

accountable for disparities in civilian interactions is also critical in

improving police-civilian relations. After all, when people don’t

trust the police, they don’t call the police; communities and crime

victims suffer when they are unable to use law enforcement to

enforce the laws meant to protect them.

Although many good proposals have been put forward

with an eye toward reducing inequalities in police violence, at

a basic level, we need better data to understand the problem

we want to solve. The FBI, the agency tasked with overseeing

systematic data collection, has begun its work with federal

agencies in 2017, but we do not yet know when state and

local departments will be phased in, which local agencies will

agree to share their data (the FBI has said all data collection will

be voluntary), nor even the level of detail the FBI will require on

each incident. Further, the FBI has announced that information

will only be gathered on officer-involved deaths or instances

involving “serious bodily injury”. So rather than understanding

inequalities, this data collection effort is geared toward obtain-

ing accurate incident counts. As our analysis demonstrates, for

public oversight to be effective and to reduce racial inequality,

data must consider all police contacts. Unfortunately, as Attor-

ney General Jeff Sessions has already disbanded the National

Commission on Forensic Science and questioned the value of

consent decrees focused on fixing systemic racial biases in local

police forces, we are not optimistic about the role of the federal

government in obtaining, releasing, and responding to data on

racial disparities in police use of force in the near future.

recommended resources The Center for Policing Equity. July 2016. “The Science of Jus- tice: Race, Arrest, and Police Use of Force.” Compiles data from a nationally representative sample of police departments to identify trends in police use of force and crime rates and finds that the latter cannot explain the former. policingequity.org.

“The Counted,” The Guardian. An interactive news organization website tracking all U.S. police killings, including victim and inci- dent details along with stories about their deaths. theguardian. com/us-news/series/counted-us-police-killings. Alternate formats available at thesocietypages.org.

Ava Duvernay. 2017. “13th.” Described by The New York Times as “powerful” and “infuriating,” this documentary explores policing as a form of social control, making a provocative argu- ment connecting modern policing to historical slavery and includ- ing a section on the morality and use of videos of police brutality to inspire social movements.

Charles R. Epp, Steven Maynard-Moody, Donald P. Haider-Markel. 2014. Pulled Over: How Police Stops Define Race and Citizenship. Chicago, IL: University of Chicago Press. Enhances our understand- ing of police investigatory stops, racial disparities in the likelihood and experience of such stops, and how police encounters affect perceptions about citizenship and police legitimacy.

President’s Task Force on 21st Century Policing. 2015. “Final Report of the President’s Task Force on 21st Century Policing.” Washington, DC: Office of Community Oriented Policing Services. A report commissioned by President Barack Obama after the death of Michael Brown and protests in Ferguson, MO, this document includes academics’ and police chiefs’ recommendations of specific policy proposals to promote police legitimacy and safety.

Rory Kramer and Brianna Remster are in the department of sociology and

criminology Villanova University. Kramer studies the physical and social boundaries

between races in the U.S., and Remster studies inequality, crime, and formal social

control and Camille Z. Charles is in the departments of sociology and Africana

studies at the University of Pennsylvania. She studies U.S. racial attitudes.

An attendee at a November 2015 rally.

Jo h n n y

Si lv

er cl

o u d , Fl

ic kr

C C