U.S. Department of Justice
Office of Justice Programs
Bureau of Justice Statistics

January 2021

Statistical Brief

NCJ 255969

Race and Ethnicity of Violent Crime
Ofenders and Arrestees, 2018
Allen J. Beck, Ph.D., BJS Statistician

I

n 2018, based on data from the FBI’s Uniform
Crime Reporting (UCR) Program, black people
were overrepresented among persons arrested for
nonfatal violent crimes (33%) and for serious nonfatal
violent crimes (36%) relative to their representation
in the U.S. population (13%) (table 1).1 White people
were underrepresented. White people accounted for
60% of U.S. residents but 46% of all persons arrested
for rape, robbery, aggravated assault, and other assault,
and 39% of all arrestees for nonfatal violent crimes
excluding other assault. Hispanics, regardless of their
race, were overrepresented among arrestees for nonfatal
violent crimes excluding other assault (21%) relative to
their representation in the U.S. population (18%).

At the same time, white ofenders were underrepresented
among persons arrested for nonfatal violent crimes
(46%) relative to their representation among ofenders
identifed by victims in the NCVS (52%). When limited
to ofenders in incidents reported to police, white
people were found to be arrested proportionate to their
criminal involvement (48%). Hispanic ofenders were
overrepresented among persons arrested for nonfatal
violent crimes (18%) relative to their representation
among violent ofenders (14% of all violent ofenders
and 13% of violent ofenders in incidents reported to
police). However, victims were unable to determine
if the ofender was Hispanic in 9% of single-ofender

Tese UCR data on incidents of nonfatal violent crime
can be compared to data from the National Crime
Victimization Survey (NCVS) to determine how much
ofense and arrest diferences by race and ethnicity can
be attributed to diferences in criminal involvement.
Te NCVS collects information on victims’ perceptions
of ofenders’ race, ethnicity, and other characteristics in
incidents of violent crime. Tis survey is administered to
persons age 12 or older from a nationally representative
sample of U.S. households. Te 2018 NCVS data fle
includes interviews from 151,055 households.

Table 1
Race or ethnicity of the U.S. resident population and of
persons arrested for nonfatal violent crimes, 2018

An examination of ofenders’ characteristics, as
reported by victims in the NCVS, provides information
on racial and ethnic disparities beyond an arrestee
and population-based comparison. Based on the 2018
NCVS and UCR, black people accounted for 29% of
violent-crime ofenders and 35% of violent-crime
ofenders in incidents reported to police, compared to
33% of all persons arrested for violent crimes (table 2).
(See Methodology for diferences in how violent crimes
are measured in the NCVS and UCR.)
1Serious nonfatal violent crimes include rape, robbery, and

aggravated assault and exclude other assault.

Race/ethnicity
Whiteb
Blackb
Hispanic
American Indian/
Alaska Nativeb
Asianb
Native Hawaiian/
Other Pacific
Islanderb
Two or more racesb

U.S. resident
population,
July 1, 2018
60.4%
12.5
18.3

Persons arrested, UCR 2018
Nonfatal Nonfatal violent
violent crimes excluding
crimesa other assault
45.9%
38.7%
33.0
36.1
17.6
21.4

0.7
5.7

1.9
1.3

1.9
1.5

0.2
2.2

0.3
--

0.4
--

Note: See appendix table 1 for estimation of arrests by race and ethnicity.
--Not reported.
aIncludes rape, robbery, aggravated assault, and other assault. Excludes
murder and non-negligent manslaughter.
bExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Federal Bureau of Investigation, Crime in the United States,
2018, table 43; National Incident-Based Reporting System, 2016,
(National Criminal Justice Archive, Study Number 37066); and U.S.
Census Bureau, Annual Estimates of the Resident Population by Sex,
Age, Race Alone or in Combination, and Hispanic Origin for the United
States: April 1, 2010 to July 1, 2019.

incidents and 12% of multiple-ofender incidents,
which may have resulted in some underestimates of
Hispanic ofenders’ involvement in violent crime (not
shown in tables).
Among the most serious incidents of violent crime
(rape or sexual assault, robbery, and aggravated
assault), there were no statistically signifcant
diferences by race between ofenders identifed in the
NCVS and persons arrested per the UCR (table 3).
White and black people were arrested proportionate
to their involvement in serious nonfatal violent crime
overall and proportionate to their involvement in
serious nonfatal violent crime reported to police.
Hispanics accounted for 21% of persons arrested for
serious nonfatal violent crime but 12% of persons
involved in serious nonfatal violent crime reported to
police. However, some of this diference may be due to
victims not knowing the ethnicity of their assailants,
even if they knew their race.
TABLE 2
Race or ethnicity of ofenders in the NCVS and of
persons arrested for nonfatal violent crimes, 2018
Race/ethnicity
of ofender
Whitec
Blackc
Hispanic
American Indian/
Alaska Nativec
Asianc
Native Hawaiian/
Other Pacifc
Islanderc
Two or more
racesc

All nonfatal violent crimes
Ofenders in NCVS
Persons
Ofenders in incidents reported
arrested in
the NCVSa
to policea
the UCR*b
52.2% †
48.3%
45.9%
28.9
34.9
33.0
14.2 †
12.6 †
17.6
1.6
1.8

1.4
1.8

1.9
1.3

0.4 !

0.5 !

0.3

1.0

0.6 !

--

Note: Based on incidents in which the victim reported the race/ethnicity
of the ofender(s). See appendix table 2 for standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
--Not reported.
aIncludes rape/sexual assault, robbery, aggravated assault, and simple
assault. Excludes murder and non-negligent manslaughter.
bIncludes rape, robbery, aggravated assault, and other assault. Excludes
murder and non-negligent manslaughter.
cExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization Survey,
2018; Federal Bureau of Investigation, Crime in the United States, 2018,
table 43; and National Incident-Based Reporting System, 2016, (National
Criminal Justice Archive, Study Number 37066).

The NCVS provides the basis for comparing the
race and ethnicity of violent ofenders with persons
in the U.S. population and persons arrested
Previous reports from the Bureau of Justice Statistics
(BJS) on the race and ethnicity of violent ofenders
have been based either on violent victimizations or on
violent incidents.2 Estimates of violent victimization
represent the number of persons victimized and
count the characteristics of ofenders multiple times if
more than one victim was present during an incident.
Estimates of violent crime incidents represent the
number of incidents but do not account for the number
of ofenders or their characteristics when incidents
involve two or more ofenders. Estimates in this report
are based on violent ofenders to facilitate comparisons
with the race and ethnicity of U.S. residents and of
persons arrested. (See Methodology.)
2See Race and Hispanic Origin of Victims and Ofenders, 2012-15

(NCJ 250747, BJS, October 2017) for an analysis based on violent
victimizations. See Criminal Victimization, 2019 (NCJ 255113, BJS,
September 2020), for an analysis based on incidents of violent crime.

TABLE 3
Race or ethnicity of ofenders in the NCVS and of persons
arrested for serious nonfatal violent crimes, 2018
Race/ethnicity
of ofender
Whitec
Blackc
Hispanic
American Indian/
Alaska Nativec
Asianc
Native Hawaiian/
Other Pacifc
Islanderc
Two or more
racesc

Serious nonfatal violent crimes
Ofenders in NCVS
Persons
Ofenders in incidents reported
arrested in
the NCVSa
to policea
the UCR*b
43.8%
40.9%
38.7%
35.9
42.8
36.1
15.5 †
12.0 †
21.4
1.1 !
2.6

1.4 !
2.5

1.9
1.5

0.3 !

0.2 !

0.4

0.7 !

0.3 !

--

Note: Based on incidents in which the victim reported the race/ethnicity
of the ofender(s). See appendix table 2 for standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
--Not reported.
aIncludes rape/sexual assault, robbery, and aggravated assault. Excludes
murder, non-negligent manslaughter, and simple assault.
bIncludes rape, robbery, and aggravated assault. Excludes murder,
non-negligent manslaughter, and other assault.
cExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization Survey,
2018; Federal Bureau of Investigation, Crime in the United States, 2018,
table 43; and National Incident-Based Reporting System, 2016, (National
Criminal Justice Archive, Study Number 37066).

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

2

Victims were able to report the race or ethnicity of
the ofender in 86% of violent crime incidents
Victims in two-thirds (67%) of nonfatal violent crime
incidents said they knew all or some of the ofenders
or had seen them before. Among incidents involving a
single ofender, 69% of victims said they knew or had
seen the ofender before. Among incidents involving
multiple ofenders, 34% of victims said all of the
ofenders were known, 18% said some were known, and
48% said all were strangers they had never seen before
or said they did not know (not shown in tables).
Victims reported the race or ethnicity of ofenders
in 75% of nonfatal violent crime incidents involving
Asian victims, 82% of incidents involving Hispanic
or black victims, and 88% of incidents involving
white victims (table 4). Overall, victims in 3.9 million
nonfatal violent crime incidents reported information
on the race or ethnicity of nearly 5.0 million ofenders
linked to these incidents (table 5).
Te number of ofenders per incident varied by the
ofenders’ race and ethnicity. Victims reported an
average of 1.5 black ofenders in incidents involving
at least one black ofender, compared to 1.2 white
ofenders in incidents involving at least one white
ofender and 1.3 Hispanic ofenders in incidents
involving at least one Hispanic ofender. Te ratios
were similar in incidents involving at least one
TABLE 4
Reported race or ethnicity of violent crime ofenders by
race or ethnicity of victim, NCVS 2018
Race/ethnicity
of victim
Total
White*a
Blacka
Hispanic
American Indian/Alaska Nativea
Asiana
Native Hawaiian/Other
Pacifc Islandera
Two or more racesa

Race/ethnicity of ofender
Reported
Not reported
85.8%
14.2%
87.8
12.2
82.0
18.0
81.8
18.2
91.9
8.1
75.4
24.6
100.0 !
90.2

0.0 !
9.8

Note: Estimates exclude incidents in which the victim did not know the
race or ethnicity of the ofender(s). Percentages based on nonfatal violent
crime incidents, including rape or sexual assault, robbery, aggravated
assault and simple assault. See appendix table 3 for standard errors.
*Comparison group.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

American Indian or Alaska Native (AIAN) ofender
(1.2), at least one ofender who was a Native Hawaiian
or Other Pacifc Islander (1.2), or at least one Asian
ofender (1.1).
Te average number of ofenders per incident did not
vary signifcantly by victims’ race or ethnicity (table 6).
Overall, victims reported an average of 1.3 ofenders
per incident. Te average ranged from 1.2 ofenders
per incident among Asian and AIAN victims to 1.4
among black victims, though the diferences were
not statistically signifcant. Te percentage of singleofender incidents was lower among black victims
(78%) than among white (90%) or AIAN (91%) victims.
Black ofenders accounted for 22% of single-ofender
incidents and 43% of multiple-ofender incidents
(table 7). White ofenders accounted for 59% of
single-ofender incidents and 38% of multiple-ofender
incidents. Hispanic ofenders accounted for similar
portions of single-ofender (14%) and multiple-ofender
incidents (16%).
TABLE 5
Number of violent crime incidents and number of
ofenders, by race or ethnicity of ofender, NCVS 2018
Race/ethnicity of
ofender
Total
Whitec
Black*c
Hispanic
American Indian/
Alaska Nativec
Asianc
Native Hawaiian/
Other Pacifc
Islanderc
Two or more racesc

Number of
incidentsa
3,857,100
2,250,300
980,800
560,900

Number of
ofendersb
4,957,600
2,588,800
1,432,600
702,300

Ratio of
ofenders
to incidents
1.29
1.15 †
1.46
1.25 †

64,700
78,700

78,600
86,800

1.21 †
1.10 †

17,800 !
47,300

21,100 !
47,300

1.19 !
1.00 †

Note: Estimates are based on nonfatal violent crime incidents, including
rape or sexual assault, robbery, aggravated assault and simple assault.
Incident weights were not adjusted for series victimization. Estimates
exclude incidents in which the victim did not know the race or ethnicity
of the ofender(s). See appendix table 4 for standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
aDetails may not sum to totals because multiple-ofender incidents with
ofenders of more than one race or ethnic group are counted more than
once (e.g., incidents with both white and black ofenders are counted in
each group).
bDetails may not sum to totals due to rounding.
cExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

3

Violent crime involvement by race and ethnicity
relative to the U.S. population
Relative to their share of the U.S. population (60%),
white people were underrepresented among ofenders
in nonfatal violent crimes overall (52%) (table 8). Tey
accounted for 45% of ofenders involved in aggravated
assaults and 31% of ofenders involved in robbery. Tey
were not underrepresented to a statistically signifcant
degree among ofenders involved in rape or sexual
assault (56%) or simple assault (59%).
Black people were overrepresented among ofenders
in nonfatal violent crimes overall (29%) relative to
their share of the U.S. population (13%). Half of all
ofenders involved in robbery (51%), a third involved
in aggravated assault (34%), and more than a ffh
involved in simple assault (23%) and rape or sexual
assault (22%) were black.
Hispanic ofenders were involved in serious nonfatal
violent crimes (16%) nearly proportionate to
their representation in the U.S. population (18%).
Hispanics were underrepresented to a statistically
signifcant degree among ofenders involved in simple
assault (13%).
Among other racial groups, Asians (6% of the U.S.
population) were consistently underrepresented
among violent ofenders, except for their involvement
in rape or sexual assault (5%). Between 1% and 2% of
ofenders involved in robbery, aggravated assault, or
simple assault were Asian.3
3Te NCVS provides rich detail on the relationship between

race and ethnicity and violent crime victimization. For example,
a previous BJS report found that at each level of poverty, there
were no statistically signifcant diferences between whites and
blacks in the rate of violent victimization. (See Household Poverty
and Nonfatal Violent Victimization, 2008-2012, NCJ 248384, BJS,
November 2014.) However, the NCVS has little information on
violent ofenders beyond their demographic characteristics. More
detailed information on violent ofenders is available from BJS
based on personal interviews of prison and jail inmates. (For more
information, see the Survey of Inmates in Local Jails (https://www.
bjs.gov/index.cfm?ty=dcdetail&iid=274) and Survey of Prison
Inmates (https://www.bjs.gov/index.cfm?ty=dcdetail&iid=488) data
collections on the BJS website.)

TABLE 6
Percentage of single- and multiple-ofender violent
crime incidents, by race or ethnicity of victim, NCVS 2018

Race/ethnicity
All
of victim
incidents
Total
100%
Whitea
100
Black*a
100
Hispanic
100
American Indian/
Alaska Nativea
100
Asiana
100
Native Hawaiian/
Other Pacifc
Islandera
100
Two or more racesa 100

Percent of incidents
Single- Multipleofender ofender
incidents incidents
87.6%
12.4%
90.0 †
10.0 †
77.6
22.4
86.2
13.8

Average
number of
ofenders
per
incident
1.29
1.29
1.35
1.32

90.7 †
89.9

9.3 †
10.1

1.16
1.16

72.3 !
83.7

27.7 !
16.3

1.28 !
1.22

Note: Based on nonfatal violent-crime incidents in which the victim
reported the race/ethnicity of the ofender(s). See appendix table 5 for
standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

TABLE 7
Percentage of ofenders by race or ethnicity in singleand multiple-ofender violent crime incidents, NCVS
2018
Race/ethnicity
of ofender
Total
Whitea
Blacka
Hispanic
American Indian/Alaska Nativea
Asiana
Native Hawaiian/Other Pacifc
Islandera
Two or more racesa
Number of ofenders

Percent of ofenders
Single-ofender Multiple-ofender
incidents*
incidents
100%
100%
58.8
38.0 †
22.2
43.4 †
13.6
15.5
1.4
2.0
2.1
0.9 !
0.5 !
1.4
3,399,000

0.3 !
-1,558,600

Note: Based on nonfatal violent-crime incidents in which the victim
reported the race/ethnicity of the ofender(s). See appendix table 6 for
standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer
sample cases or coefcient of variation is greater than 50%.
--Not reported.
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

4

TABLE 8
Percentage of ofenders in violent crime incidents by type of crime and race or ethnicity compared to the U.S. resident
population, 2018
Race/ethnicity
Total
Whiteb
Blackb
Hispanic
American Indian/Alaska
Nativeb
Asianb
Native Hawaiian/Other
Pacifc Islanderb
Two or more racesb

U.S. resident
population,
July 1, 2018*
100%
60.4
12.5
18.3

Nonfatal violent Serious nonfatal
crimes
violent crimes
100%
100%
52.2 †
43.8 †
28.9 †
35.9 †
14.2 †
15.5

Percent of ofendersa
Rape/sexual
assault
Robbery
100%
100%
55.7
31.0 †
22.1 †
51.1 †
15.0
15.6

Aggravated
assault
100%
44.8 †
33.9 †
16.6

Simple assault
100%
59.2
23.3 †
12.7 †

0.7
5.7

1.6 †
1.8 †

1.1 !
2.6

0.3 !
4.7 !

0.4 !
1.9 ! †

2.0 !
1.9 ! †

1.9 †
1.1 †

0.2
2.2

0.4 !
1.0 †

0.3 !
0.7 !

0.5 !
1.7 !

0.0 !
0.0 !

0.3 !
0.6 ! †

0.6 !
1.2 †

Note: See appendix table 7 for standard errors.
*Comparison group.
†Signifcant diference at the 95% confdence level.
! Interpret estimate with caution. Estimate is based on 10 or fewer sample cases or coefcient of variation is greater than 50%.
aExcludes incidents in which the victim did not know the race or ethnicity of the ofender(s).
bExcludes persons of Hispanic/Latino origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization Survey, 2018; and U.S. Census Bureau, Annual Estimates of the Resident Population
by Sex, Age, Race Alone or in Combination, and Hispanic Origin for the United States: April 1, 2010 to July 1, 2019.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

5

Methodology
National Crime Victimization Survey data on
violent ofenders’ race and ethnicity
Ofender characteristics
Te National Crime Victimization Survey (NCVS)
collects ofender information from victims in the
Crime Incident Report (CIR). Ofender demographic
characteristics are based on the victims’ perceptions.
Te section in the CIR on ofenders begins with a
question about the number of ofenders. (For a more
detailed description of the NCVS methodology,
see Criminal Victimization, 2018, NCJ 253043, BJS,
September 2019).

times, up to a maximum of 10 times.6 Additional
information on series victimization is detailed in the
report Methods for Counting High-Frequency Repeat
Victimizations in the National Crime Victimization
Survey (NCJ 237308, BJS, April 2012).
Incident weights were adjusted by the number of
violent ofenders and their characteristics as reported
by victims. Incidents in which the race or ethnicity
of the ofender was not reported were excluded. Te
adjustments varied by the type of incident and number
of ofenders:
Single-ofender incidents—Te incident weight was not
adjusted (adjustment factor = 1.0). Persons of Hispanic
or Latino origin were classifed as Hispanic, regardless
of their reported race.

For violent crime incidents involving a single ofender,
respondents are asked about the ofender’s relationship
to the victim, demographic characteristics (including
sex, race, ethnicity, and age), membership in a street
gang, use of alcohol or drugs at the time of the
incident, and previous crimes against the respondent
or respondent’s household.

Multiple-ofender incidents, all ofenders of the same
race—Te incident weight was adjusted based on
the number of ofenders (adjustment factor = 2.0 for
incidents with two ofenders; 3.0 for incidents with
three ofenders; and up to 15.0 for incidents involving
15 ofenders). In six incidents, the number of multiple
ofenders was unknown and was imputed to equal two.

For violent incidents involving multiple ofenders,
respondents are asked similar questions, such as
whether the ofender demographic characteristics
apply to all or most of the ofenders. Respondents are
asked if any of the ofenders were Hispanic or Latino,
followed by whether they were mostly Hispanic, mostly
non-Hispanic, or an equal number of Hispanic and
non-Hispanic. Respondents are then asked what the
race or races were of their ofenders.4

Multiple-ofender incidents, at least one ofender was
Hispanic—Te adjustment factor for “mostly Hispanic
or Latino” ofenders (“all Hispanic” ofenders was not
provided as a response category) was 2.0 for incidents
involving two ofenders; 2.5 for incidents involving
three ofenders; 3.5 for four ofenders; and up to 7.0
for nine ofenders. Te adjustment factor for “mostly
non-Hispanic” was assigned the middle value of
the minority range (e.g., adjustment factor = 1.0 for
incidents involving two or three ofenders; and 1.5 for
fve or six ofenders).

Following the Ofce of Management and Budget
(OMB) standards for measuring race and ethnicity, the
ofender race categories in the 2018 NCVS were white;
black; American Indian or Alaska Native; Asian; Native
Hawaiian or Other Pacifc Islander; and persons of two
or more races.5 Ofenders reported as Hispanic were
classifed as Hispanic, regardless of their reported race.
Estimating the race or ethnicity of ofenders
NCVS estimates in this report use incident weights
unadjusted for series incidents to produce estimates
of the number ofenders. Inclusion of series weights
would result in counting the same ofender(s) multiple
4For all questions included on the NCVS CIR, see https://www.bjs.

gov/content/pub/pdf/ncvs18_cir.pdf.

5See Revisions to the Standards for the Classifcation of Federal

Data on Race and Ethnicity, Ofce of Management and Budget,
Statistical Policy Directive No. 15, October 1997.

In incidents that involved an “equal” number of
Hispanics and non-Hispanics, the adjustment factor
for Hispanics was coded as the middle value of the
number of ofenders (e.g., adjustment factor = 1.0 for
incidents involving two ofenders). Odd numbers of
ofenders were divided by 2 and assigned the integer
value plus 0.5 (e.g., Hispanic adjustment factor = 1.5
for incidents involving three ofenders).
6Series victimizations involve multiple crimes that are ofen

indistinguishable to victims, making it difcult for them to separate
the details of each event. To handle these repeated victimizations,
the NCVS records a series victimization when the respondent
reported experiencing six or more similar crimes during the
6-month reference period and was unable to recall or describe each
event in detail. Te NCVS records the victim’s report of the number
of times this type of victimization occurred and collects detailed
information only for the most recent victimization.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

6

In incidents that involved offenders of more than one
racial group among non-Hispanics, the adjustment
factors for “mostly” were assigned the average of
the majority values (e.g., black adjustment factor
= 4.5 and other-races adjustment factor = 1.5 in
incidents of “mostly black” offenders with six nonHispanic offenders).
Multiple-offender incidents, no offender was Hispanic—
Adjustment factors were assigned to incidents for
each racial group. In incidents with four or fewer
offenders, “mostly” was the number of offenders minus
1 (e.g., adjustment factor = 3.0 in incidents with four
“mostly white” offenders; and 2.0 in incidents with
three “mostly white” offenders). Adjustment factors
for offenders not among the “mostly” group were
assigned the middle value of the range of minority
values (e.g., adjustment factor = 1.0 in incidents with
three offenders).
For incidents with five or more offenders, “mostly” was
the middle value of the range of majority values (e.g.,
adjustment factor = 3.5 in incidents with five offenders
of “mostly” some race; and 7.5 in incidents with 10
offenders of “mostly” some race). Adjustment factors
for offenders not among the “mostly” group were
assigned the residual value (e.g., adjustment factor =
1.5 in incidents with five offenders; and 2.5 in incidents
with 10 offenders).
For incidents with an “equal” number of offenders in
each racial group, the adjustment factor was the middle
value of the number of offenders. Odd numbers of
offenders were divided by 2 and assigned the integer
value plus 0.5 (e.g., adjustment factor = 1.5 in incidents
with three offenders).
The adjustment factors when applied to the unadjusted
incident weights provide an estimate of the total
number of offenders. (See table 5.) Overall, victims
reported the race or ethnicity of more than 4.9 million
offenders of violent crimes.
An incident is a specific criminal act involving one or
more victims and one or more offenders. In 2018, there
were an estimated 6.0 million violent crime incidents.
To avoid double counting of the same offenders by
victims in series crime, the incident series weights
were not adjusted (that is, they were treated as a
single incident rather than weighted by the number
of incidents, up to a maximum of 10 incidents). The
number of unadjusted violent crime incidents totaled
4.6 million in the 2018 NCVS.

Standard error computations
When national estimates are derived from a sample, as
with the NCVS, caution must be used when comparing
one estimate to another or when comparing estimates
over time. Although one estimate may be larger than
another, estimates based on a sample have some degree
of sampling error. The sampling error of an estimate
depends on several factors, including the amount of
variation in the responses and the size of the sample.
One measure of the sampling error associated with
an estimate is the standard error. The standard error
may vary from one estimate to the next. Generally,
an estimate with a smaller standard error provides a
more reliable approximation of the true value than an
estimate with a larger standard error. Estimates with
relatively large standard errors have less precision and
reliability and should be interpreted with caution.
For complex sample designs, there are several methods
that can be used to generate standard errors around
a point estimate (e.g., numbers, percentages, and
rates). In this report, Taylor Series Linearization (TSL)
methods were utilized to estimate the standard errors
for numbers and percentages by the race and ethnicity
of offenders. The TSL method directly estimates
variances through a linearized function by combining
variance estimates from the stratum and primary
sampling units (PSUs) used to sample households and
persons. In the NCVS, the design parameters used for
computing TSL variances are PSEUDOSTRATUM
(stratum) and HALFSAMPLE (PSU).
Because some multiple-offender incidents involved
offenders of different racial or ethnic groups, standard
errors were estimated iteratively for each group. The
standard errors for the number and percentage of
white offenders were estimated first, based on the
white adjustment factor times the unadjusted incident
weight. To account for multiple-offender incidents
with a combination of white offenders and offenders of
other racial or ethnic groups, the numbers of offenders
were allocated proportionately to ensure that the sum
of offenders equaled the total number of offenders.
The same procedures were applied to each subsequent
racial or ethnic group. As a result, the estimated
standard errors for each group account for the overlap
of groups among multiple-offender incidents.
Statistical tests were conducted to determine whether
differences in percentages of offenders by race or
ethnicity were statistically significant from percentages

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021			

7

of U.S. residents and of persons arrested for violent
crimes. Te primary test procedure was the Student’s
t-statistic, which was used to test the diference
between the sample estimates and the percentages
among U.S. residents and among persons arrested for
violent crimes in 2018. Estimated percentages and
standard errors provided in this report may be used to
generate a confdence interval around the estimate as a
measure of the margin of error.
Te following example illustrates how standard errors
may be used to generate confdence intervals: Based
on the 2018 NCVS, the percentage of black persons
among violent ofenders was 28.9%, with a standard
error of 2.6%. (See table 2 and appendix table 2.) A
confdence interval around the estimate is generated
by multiplying the standard error by ± 1.96 (the t-score
of a normal, two-tailed distribution that excludes 2.5%
at either end of the distribution). Terefore, the 95%
confdence interval around the estimated percentage of
black violent ofenders, for example, would be 28.9%
± (2.6% × 1.96) or (23.8% to 34.0%). In other words,
if the same methods to select samples and compute
a confdence interval estimate for each sample were
repeatedly used, then the true population parameter
(percentage of black persons among violent ofenders)
would be expected to fall within the interval estimates
95% of the time.
Comparisons of race and ethnicity percentages
between NCVS-reported ofenders and U.S. residents
or persons arrested for violent crimes were determined
to be statistically signifcant when the percentages were
outside of the confdence interval for ofenders in the
NCVS. For example, the percentage of black persons
among U.S. residents was 12.5% in 2018, which was
outside of the 95% confdence interval of 23.8% to
34.0% of violent ofenders identifed as black in the
NCVS. (See appendix table 7 for standard errors.) As a
result, the percentage of black persons among violent
ofenders was determined to be statistically signifcant
and not due to sampling error.
A coefcient of variation (CV) is another measure
of reliability and a means for determining precision
across estimates with difering levels or metrics.
CVs (not shown in tables) represent the ratio of the
standard error to the estimate. Estimates with a CV of
greater than 50% are fagged and should be interpreted
with caution.

Data on race or ethnicity of persons arrested for
violent crime
Te FBI collects data on arrested persons through the
Uniform Crime Reporting (UCR) Program, which
includes the Summary System (SRS) and the National
Incident-Based Reporting System (NIBRS). Unlike
data reported through the UCR’s traditional SRS—an
aggregate monthly tally of crimes known to police
and of arrests—NIBRS captures details on each crime
incident and characteristics of victims and persons
arrested. In 2018, 7,283 law enforcement agencies
submitted NIBRS data to the UCR, representing 43.7%
of the 16,659 law enforcement agencies that submitted
data to the UCR program during the year.
Te UCR counts one arrest for each instance in
which a person is arrested, cited, or summoned for
an ofense. Te SRS collects arrest data separately by
race and ethnicity, while NIBRS provides the ability to
combine race and ethnicity into a single classifcation,
consistent with OMB reporting standards. For
comparison purposes in this report, information
from the SRS and NIBRS were combined to provide
estimates of race and ethnicity.
Te estimation process was based on the arrests by
race for each crime type and arrests of Hispanics
by crime type from the combined SRS and NIBRS.
(See appendix table 1.) NIBRS data from 2016, the
most recent comparable data available, were used to
estimate the proportion of Hispanic persons by race
for each crime type in the SRS. For each crime type,
the number of Hispanics arrested was multiplied by
the proportion of Hispanics within each racial category
to obtain an estimated number of Hispanics by race
by crime type. Tese estimates were then subtracted
from the total counts for each racial category to obtain
counts that separated Hispanics from non-Hispanics of
diferent races.
Te following example illustrates how these estimates
were calculated: In 2018, the UCR reported a total
of 12,794 whites (regardless of ethnicity) and 4,090
Hispanics (regardless of race) who were arrested for
rape. To determine the number of Hispanic whites
among arrestees for rape, the proportion of white
persons among Hispanic arrestees for rape in NIBRS
(0.9675) was applied to the Hispanic total (4,090). Tis
estimate (3,957) was subtracted from the total arrests of
whites (12,794) to obtain the number of non-Hispanic
whites arrested for rape (8,837).

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

8

Te estimation process was repeated for each crime
type and racial group. Te estimates by race and
ethnicity were then summed across crime types and
converted to percentages. Overall, an estimated 17.6%
of persons arrested for nonfatal violent crime were
Hispanic; 45.9% were non-Hispanic white; 33.0%
were non-Hispanic black; 1.9% were non-Hispanic
American Indian or Alaska Native; 1.3% were nonHispanic Asian; and 0.3% were non-Hispanic Native
Hawaiian or Other Pacifc Islander. Percentages were
also calculated for nonfatal violent crime excluding
other assaults.

Diferences in how violent crimes are measured in
the NCVS and UCR
Te NCVS measures crime reported to and not
reported to police. Te UCR measures only crime
recorded by police. Te NCVS and UCR measure an
overlapping, but not identical, set of ofenses:
„

Te NCVS measures rape and sexual assault.
Although the UCR measures rape similarly, it
includes a broader category in its arrest statistics:
“sex ofenses, excluding prostitution and
commercialized vice.” Tis includes “ofenses against
chastity, common decency, morals, and the like.
Incest, indecent exposure and statutory rape are
included.” Because this broader defnition does not
match up with the NCVS, arrest data for these sex
ofenses were excluded from this report.

„

Te NCVS distinguishes aggravated assault from
simple assault. Te NCVS defnes aggravated assault
as an attack or attempted attack with a weapon,
regardless of whether an injury occurred, and an
attack without a weapon when serious injury results.
It defnes simple assault as an attack or attempted
attack without a weapon, resulting either in no
injury, in a minor injury (e.g., a bruise, black eye,
cut, scratch, or swelling), or in an undetermined
injury requiring less than 2 days of hospitalization.

„

Te UCR arrest data, used in this report, utilizes
similar distinctions (unlike the UCR summary data
which does not include simple assault) but with the
addition of an “other assault” category that includes
stalking, intimidation, coercion, and hazing. Tese
crimes categorized under “other assault” are not
measured in the NCVS.

Methodology for U.S. population of persons age 12
or older
During 2018, more than half of violent victimizations
were committed by someone known to the victim
(not shown in tables). Because most violent crime
victims know their ofender, most can provide reliable
information about their race and Hispanic origin. Even
in incidents of stranger-perpetrated violent crime,
victims may be able to accurately identify the ofender’s
race and ethnicity using cues, such as the location of
the crime, appearance, language used, mannerisms
of the ofender, and information from other sources,
including police reports or arrest records. NCVS
questions aim to establish general characteristics
of ofenders instead of unique identities. Research
on eyewitness testimony indicates that victims can
identify general characteristics of ofenders, such as
sex, race, ethnicity, and age, and are more ofen correct
than incorrect when identifying these characteristics.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

9

APPENDIX TABLE 1
Estimating the number of persons arrested for violent ofenses by race/ethnicity, UCR 2018

All racesa
Total
White
Black
American Indian/Alaska Native
Asian
Native Hawaiian/Other Pacifc
Islander
Hispanics onlyb
Total
White
Black
American Indian/Alaska Native
Asian
Native Hawaiian/Other Pacifc
Islander
Race/ethnicity combinedc
Total
Whited
Blackd
Hispanic
American Indian/Alaska Natived
Asiand
Native Hawaiian/Other Pacifc
Islanderd

Rape

Number of persons arrested by ofense
Aggravated
Robbery
assault
Other assault

All violent ofenses
Number
Percent

18,776
12,794
5,376
267
289

66,789
29,025
36,187
676
641

298,040
184,527
100,393
6,736
5,078

794,787
512,025
254,360
15,711
10,348

1,178,392
738,371
396,316
23,390
16,356

50

260

1,306

2,343

3,959

4,090
3,957
121
4
4

12,823
11,972
607
102
142

65,056
61,991
2,575
299
158

125,007
119,341
4,549
735
308

206,976
197,261
7,852
1,140
612

4

0

33

74

111

18,776
8,837
5,255
4,090
263
285

66,789
17,053
35,580
12,823
574
499

298,040
122,536
97,818
65,056
6,437
4,920

794,787
392,684
249,811
125,007
14,976
10,040

1,178,392
541,110
388,464
206,976
22,250
15,744

46

260

1,273

2,269

3,848

100%
62.7
33.6
2.0
1.4
0.3
100%
95.3
3.8
0.6
0.3
0.1
100%
45.9
33.0
17.6
1.9
1.3
0.3

Note: Race and ethnicity totals are collected separately in the UCR summary system. The National Incident-Based Reporting System (NIBRS) provides
incident-level detail needed to classify arrestees by race and ethnicity in combination. In 2016, 6,849 law enforcement agencies (approximately 37% of
all agencies that were eligible to participate in the UCR Program) reported crime data via NIBRS. These agencies covered approximately 31% percent
of the United States’ population.
aBased on total arrests by race for each crime type as reported by state and local law enforcement agencies. Excludes arrests in agencies that did not
provide race information.
bFor each crime type and racial group, the total number of Hispanics arrested was estimated by multiplying the proportion Hispanic for race (as
reported in NIBRS) by the total number of Hispanics arrested (as reported in UCR summary data, table 43a). For example, 96.75% of Hispanics arrested
for rape were Hispanic/Latino. Consequently, of the 4090 Hispanics arrested for rape, an estimated 3,957 were white.
cBased on total arrests by race minus estimated arrests of Hispanics by race for each crime type.
dExcludes persons of Hispanic/Latino origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Federal Bureau of Investigation, Crime in the United States, 2018, table 43a; and National Incident-Based Reporting System, 2016, (National
Archive of Criminal Justice Data, Study number 37066).

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

10

APPENDIX TABLE 2
Standard errors for tables 2 and 3: Race or ethnicity of
violent crime ofenders in the NCVS, 2018

Race/ethnicity
of ofender
Whiteb
Blackb
Hispanic
American Indian/
Alaska Nativeb
Asianb
Native Hawaiian/
Other Pacifc
Islanderb
Two or more racesb

All violent crime
offenders in the
NCVSa
Reported
All
to police
2.7%
3.6%
2.6
4.0
1.5
1.8

Violent crime
offenders in the
NCVS, excluding
simple assault
Reported
All
to police
3.5%
5.1%
3.8
5.9
2.5
2.0

0.5
0.4

0.5
0.6

0.5
0.9

0.9
1.2

0.1
0.2

0.3
0.3

0.2
0.3

0.2
0.1

aIncludes rape/sexual assault, robbery, aggravated assault, and simple

assault. Excludes murder and non-negligent manslaughter.

bExcludes persons of Hispanic/Latino origin (e.g., “white” refers to

APPENDIX TABLE 3
Standard errors for table 4: Reported race or ethnicity
of violent crime ofenders by race or ethnicity of
victim, NCVS 2018
Race/ethnicity
of victim
Total
Whitea
Blacka
Hispanic
American Indian/Alaska Nativea
Asiana
Native Hawaiian/Other Pacifc
Islandera
Two or more racesa

Race/Hispanic origin of ofender
Reported
Not reported
1.2%
1.2%
1.2
1.2
3.7
3.7
3.1
3.1
2.9
2.9
8.7
8.7
-3.9

-3.9

--Not calculated.
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

APPENDIX TABLE 4
Standard errors for table 5: Number of violent crime
incidents and number of ofenders, by race or ethnicity
of ofender, NCVS 2018
Race/ethnicity of
ofender
Total
Whiteb
Blackb
Hispanic
American Indian/
Alaska Nativeb
Asianb
Native Hawaiian/
Other Pacifc
Islanderb
Two or more racesb

Ratio of
ofenders to
incidents
0.03
0.03
0.07
0.07

Number of
incidentsa
147,813
124,146
67,160
46,013

Number of
ofenders
224,538
157,363
158,116
80,426

17,438
22,397

20,634
21,382

<.01
0.07

675
11,242

6,335
11,242

0.12
0

aMultiple-ofender Incidents with ofenders of more than one race or

ethnic group are counted more than once (e.g., incidents with both
white and black ofenders are counted in each group).
bExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

11

APPENDIX TABLE 5
Standard errors for table 6: Percentage of single- and
multiple-ofender violent crime incidents, by race or
ethnicity of victim, NCVS 2018
Race/ethnicity
of victim
Total
Whitea
Blacka
Hispanic
American Indian/
Alaska Nativea
Asiana
Native Hawaiian/
Other Pacifc
Islandera
Two or more racesa

Percent of incidents
SingleMultipleofender
ofender
incidents
incidents
1.2%
1.2%
1.2
1.2
4.7
4.7
2.9
2.9

Average
number of
ofenders per
incident
0.04
0.05
0.07
0.10

4.5
5.1

4.5
5.1

0.11
0.09

23.6
8.1

23.6
8.1

0.24
0.09

Note: Based on violent crime incidents in which the victim reported the
race/ethnicity of the ofender(s).
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

APPENDIX TABLE 6
Standard errors for table 7: Percentage of ofenders
by race or ethnicity in single- and multiple-ofender
violent crime incidents, NCVS 2018
Race/ethnicity of ofender
Whitea
Blacka
Hispanic
American Indian/Alaska Nativea
Asiana
Native Hawaiian/Other Pacifc
Islandera
Two or more racesa
Number of ofenders

Percent of ofenders
Single-ofender Multiple-ofender
incidents*
incidents
2.0%
6.3%
1.4
6.7
1.1
3.0
0.5
0.7
0.6
0.5
0.2
0.3
138,443

0.3
-94,174

--Not calculated.
aExcludes persons of Hispanic/Latino origin (e.g., “white” refers to
non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization
Survey, 2018.

APPENDIX TABLE 7
Standard errors for table 8: Percentage of ofenders in violent crime incidents by type of crime and race or ethnicity
compared to the U.S. resident population, 2018
Race/ethnicity
Whiteb
Blackb
Hispanic
American Indian/Alaska
Nativeb
Asianb
Native Hawaiian/Other
Pacifc Islanderb
Two or more racesb

Nonfatal violent
crimesa
2.7%
2.6
1.5

Serious nonfatal
violent crimes
3.5%
3.8
2.5

Percent of ofenders
Rape/sexual
assault
Robbery
5.3%
3.2%
3.9
4.0
3.4
2.5

Aggravated
assault
4.9%
6.6
2.6

Simple assault
2.8%
2.4
1.5

0.4
0.4

0.5
0.9

0.3
1.8

-1.2

1.0
0.6

0.6
0.5

0.1
0.2

0.2
0.3

0.5
0.9

0
0

0.3
0.1

0.2
0.3

--Less than 0.05%.
aExcludes incidents in which the victims did not know the race or ethnicity of the ofender(s).
bExcludes persons of Hispanic/Latino origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
Source: Bureau of Justice Statistics, National Crime Victimization Survey, 2018; and U.S. Census Bureau, Annual Estimates of the Resident Population
by Sex, Age, Race Alone or in Combination, and Hispanic Origin for the United States: April 1, 2010 to July 1, 2019.

R ace and Ethnicit y of Violent Crime O ffenders and Arrestees, 2018 | Januar y 2021

12

Te Bureau of Justice Statistics of the U.S. Department of Justice is the
principal federal agency responsible for measuring crime, criminal
victimization, criminal ofenders, victims of crime, correlates of crime,
and the operation of criminal and civil justice systems at the federal, state,
tribal, and local levels. BJS collects, analyzes, and disseminates reliable
statistics on crime and justice systems in the United States, supports
improvements to state and local criminal justice information systems, and
participates with national and international organizations to develop and
recommend national standards for justice statistics. Jefrey H. Anderson is
the director.
Tis report was written by Allen J. Beck. Barbara Oudekerk, Heather
Brotsos, Jenna Truman, and Alexia Cooper verifed the report.
Edrienne Su edited the report. Teodore Robinson produced the report.
January 2021, NCJ 255969

11111111 1111111111111111111IIIIIII
NCJ 255969

Ofce of Justice Programs
Building Solutions • Supporting Communities • Advancing Justice
www.ojp.gov