Revised February 3, 2023
U.S. Department of Justice
Oce of Justice Programs
Bureau of Justice Statistics
Bureau of Justice Statistics · Statistical Tables
December 2020, NCJ 255730
Contacts Between Police and the
Public, 2018 – Statistical Tables
F
indings described in this report are
based on data from the Bureau of Justice
Statistics’ 2018 Police-Public Contact
Survey (PPCS), a supplement to the National
Crime Victimization Survey (NCVS). e
NCVS collects information from a nationally
representative sample of persons age 12 or
older in U.S. households. e PPCS collects
information on contact with police during the
prior 12 months from persons age 16 or older.
U.S. residents were asked about instances where
they sought help from police (resident-initiated
contacts) and when police approached or stopped
them (police-initiated contacts). Resident-
initiated contacts with police include reporting
a crime, disturbance, or suspicious activity;
reporting a non-crime emergency such as a
medical emergency; reporting a non-emergency
such as asking for directions; participating in
a block watch or other anti-crime program; or
approaching or seeking help from police for
another reason. Police-initiated contacts include
being stopped by police while in a public place or
a parked vehicle (i.e., a street stop), being stopped
by police while driving a motor vehicle (i.e., a
trac stop), riding as a passenger in a car that
was stopped by police, being arrested, or being
stopped or approached by police for some other
reason. e PPCS also collected data on contacts
resulting from a trac accident.
Erika Harrell, Ph.D., and Elizabeth Davis, BJS Statisticians
Highlights
In the prior 12 months, as of 2018, among
persons age 16 or older—
About 61.5 million residents had at least one
contact with police.
Twenty-four percent of residents experienced
contact with police, up from 21% in 2015.
Whites (26%) were more likely than blacks
(21%), Hispanics (19%), or persons of other
races (20%) to experience police contact.
ere was no statistically signicant dierence
in the percentage of whites (12%) and blacks
(11%) who experienced police-initiated contact.
Persons ages 18 to 24 were most likely to
have any contact with police (30%) and to
experience police-initiated contact (19%).
A higher percentage of blacks (4%) and
Hispanics (3%) than whites (2%) or other
races (2%) experienced threats or use of force.
Males (3%) were more likely than females
(1%) to experience threats or use of force.
Four percent of blacks and 4% of Hispanics
reported being handcued during their most
recent contact with police, compared to 2% of
whites and 2% of other races.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 2
List of tables
TABLE 1. Number and percent of U.S. residents age 16 or older with any police contact, by type of
contact and demographic characteristics, 2018
TABLE 2. Residents with police contact, by reason for contact, 2015 and 2018
TABLE 3. Residents who experienced nonfatal threats or use of force during contacts with police,
by demographic characteristics, 2015 and 2018
TABLE 4. Residents who experienced nonfatal threats or use of force during their most recent
police-initiated contact or trac-accident contact, by resident perception of force as necessary or
excessive and demographic characteristics, 2015 and 2018
TABLE 5. Percent of residents who experienced nonfatal threats or use of force during their most
recent police-initiated contact or trac-accident contact, by resident race or ethnicity and type
of force, 2018
List of appendix tables
APPENDIX TABLE 1. Standard errors for table 1: Number and percent of U.S. residents age 16 or older
with any police contact, by type of contact and demographic characteristics, 2018
APPENDIX TABLE 2. Standard errors for table 2: Residents with police contact, by reason for contact,
2015 and 2018
APPENDIX TABLE 3. Standard errors for table 3: Residents who experienced nonfatal threats or use of
force during contacts with police, by demographic characteristics, 2015 and 2018
APPENDIX TABLE 4. Standard errors for table 4: Residents who experienced nonfatal threats or
use of force during their most recent police-initiated contact or trac-accident contact, by resident
perception of force as necessary or excessive and demographic characteristics, 2015 and 2018
APPENDIX TABLE 5. Standard errors for table 5: Percent of residents who experienced nonfatal
threats or use of force during their most recent police-initiated contact or trac-accident contact,
by resident race or ethnicity and type of force, 2018
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 3
Police contact and resident demographics
In 2018, about 61.5 million persons age 16 or older
had at least one contact in the prior 12 months
with police: 28.9 million U.S. residents experienced
contacts initiated by police, 35.5 million initiated
contact with police, and 8.9 million had contact with
police as a result of a trac accident (table 1).
A higher percentage of males (13%) than females
(10%) experienced police-initiated contacts in 2018.
Whites (26%) were more likely than blacks (21%),
Hispanics (19%), or persons of other races (20%) to
experience police contact.
ere was no statistically signicant dierence in the
percentage of whites (12%) and blacks (11%) who
experienced police-initiated contact.
Hispanics (10%) were less likely than whites (12%)
to experience police-initiated contact.
Whites (16%) were more likely than blacks (11%),
Hispanics (10%), or other persons (11%) to initiate
contact with police.
Among all age groups measured, persons ages 18
to 24 were most likely to have any contact with
police (30%) and to experience police-initiated
contact (19%).
TABLE 1
Number and percent of U.S. residents age 16 or older with any police contact, by type of contact and demographic
characteristics, 2018
Demographic
characteristic
U.S. population
age 16 or older
Any contact Police-initiated contact
Resident-initiated
contact Trac-accident contact
Number Percent Number Percent Number Percent Number Percent
Total 259,316,200 61,542,300 23.7% 28,880,900 11.1% 35,468,500 13.7% 8,882,000 3.4%
Sex
Male* 125,725,800 30,467,400 24.2% 15,707,300 12.5% 16,443,500 13.1% 4,141,300 3.3%
Female 133,590,400 31,074,900 23.3 † 13,173,600 9.9 † 19,025,000 14.2 † 4,740,700 3.5 ‡
Race/ethnicity
White
a
* 163,551,000 42,525,700 26.0% 19,216,500 11.7% 25,386,330 15.5% 5,885,200 3.6%
Black
a
30,973,900 6,545,700 21.1 † 3,393,800 11.0 3,280,700 10.6 † 1,086,900 3.5
Hispanic 43,135,000 8,238,400 19.1 † 4,221,800 9.8 † 4,449,900 10.3 † 1,222,600 2.8 †
Other
a,b
21,656,300 4,232,500 19.5 † 2,048,900 9.5 † 2,351,600 10.9 † 687,300 3.2
Age
16-17 8,046,500 1,
143,500 14.2% † 765,300 9.5% † 349,100 4.3% † 223,100 2.8% †
18-24* 29,941,900 8,859,700 29.6 5,730,700 19.1 3,984,200 13.3 1,769,400 5.9
25-44 86,126,600 23,518,700 27.3 † 11,791,500 13.7 † 13,314,800 15.5 † 3,255,500 3.8 †
45-64 83,540,200 19,160,700 22.9 † 7,744,300 9.3 † 12,075,100 14.5 ‡ 2,355,300 2.8 †
65 or older 51,661,000 8,859,600 17.1 † 2,849,000 5.5 † 5,745,200 11.1 † 1,278,700 2.5 †
Household income
$24,999 or less* 48,322,500 11,112,600 23.0% 5,487,200 11.4% 6,339,000 13.1% 1,624,500 3.4%
$25,000-$49,999 65,747,300 14,380,200 21.9 6,973,400 10.6 8,113,600 12.3 2,014,700 3.1
$50,000-$74,999 47,676,700 11,015,000 23.1 5,228,700 11.0 6,339,100 13.3 1,570,000 3.3
$75,000 or more 97,569,600 25,034
,500 25.7 † 11,191,700 11.5 14,676,700 15.0 † 3,672,800 3.8 ‡
Note: Details may not sum to totals because respondents could indicate yes to multiple reasons. See appendix table 1 for standard errors. Missing
data on annual household income was imputed. From July through December of 2018, 25% of persons age 16 or older in the NCVS sample had
missing data on annual household income. For more information on imputation procedures, see National Crime Victimization Survey, 2016: Technical
Documentation (NCJ 251442, BJS, December 2017).
*Comparison group.
†Dierence with comparison group is signicant at the 95% condence interval.
‡Dierence with comparison group is signicant at the 90% condence interval.
a
Excludes persons of Hispanic origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
b
Includes Asians, Native Hawaiians, other Pacic Islanders, American Indians, Alaska Natives, and persons of two or more races.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 4
Reasons for police contact in 2018 compared
to 2015
e percentage of U.S. residents age 16 or older
who experienced contact with police in the prior 12
months increased from 21% in 2015 to 24% in 2018
(table 2).
e percentage of persons who initiated contact with
police increased from 11% in 2015 to 14% in 2018.
e percentage of persons who experienced a
police-initiated contact did not change signicantly
from 2015 to 2018 (11% both years).
Compared to 2015, higher percentages of residents
in 2018 reported a possible crime or a non-crime
emergency to police (among the resident-initiated
contacts measured), while a lower percentage
participated in a block watch.
TABLE 2
Residents with police contact, by reason for contact, 2015 and 2018
Reason for contact
2015 2018*
Number Percent Number Percent
Any 53,469,300 21.1% † 61,542,300 23.7%
Resident-initiated contact 27,060,200 10.7% † 35,468,500 13.7%
Reported possible crime 16,928,100 6.7 † 19,109,200 7.4
Reported non-crime emergency
a
8,841,900 3.5 † 9,971,500 3.8
Reported non-emergency
b
... ... 10,068,700 3.9
Block watch 2,366,200 0.9
‡
2,160,900 0.8
Other 2,478,400 1.0 † 641,200 0.2
Trac-accident contact 7,950,500 3.1% † 8,882,000 3.4%
Police-initiated contact 27,415,900 10.8% 28,880,900 11.1%
Driver during a trac stop
c
19,204,500 8.6 † 18,666,000 8.1
Passenger during a trac stop 5,964,100 2.4 5,702,600 2.2
Street stop
d
2,503,700 1.0 † 3,528,100 1.4
Arrested only
e
814,800 0.3 † 386,000 0.1
Other 1,946,700 0.8 † 3,638,100 1.4
Note: Details may not sum to totals because respondents could experience multiple types of contact. See
appendix table 2 for standard errors.
*Comparison year.
†Dierence with comparison year is signicant at the 95% condence interval.
‡Dierence with comparison year is signicant at the 90% condence interval.
...Not available.
a
Includes medical emergencies and trac accidents that were witnessed by the respondent in which the
respondent was not involved.
b
Not available as a separate category in 2015. Includes asking for directions, custody enforcement, court orders,
or other non-emergency situations.
c
Based on the driving population of 223,315,375 persons in 2015 and 231,290,951 persons in 2018.
d
While in a public place or parked vehicle.
e
Includes residents who reported, as the sole type of contact with police, an arrest that occurred outside of the
context of a trac stop, street stop, or trac accident.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 5
Nonfatal threats or use of force during contacts
with police by resident demographics
Among the 61.5 million U.S. residents age 16 or
older in 2018 who had contact with police during
the prior 12 months, 1.3 million (2%) experienced
threats or use of force from police (table 3).
Males (3%) were more likely than females (1%) to
experience threats or use of force.
Whites (2%) were less likely than blacks (4%) or
Hispanics (3%) to experience threats or use of force.
Persons ages 18 to 24 (3%) were more likely than
persons ages 45 to 64 (1%) or 65 or older (less
than 0.5%) to experience threats or use of force
from police.
TABLE 3
Residents who experienced nonfatal threats or use of force during contacts with police, by demographic
characteristics, 2015 and 2018
Demographic characteristic
2015 2018
Persons age 16
or older with any
police contact
Experienced threats or force
at any time during the year
a
Persons age 16
or older with any
police contact
Experienced threats or force
at any time during the year
a
Number Percent Number Percent
Total 53,469,300 985,300 1.8% 61,542,300 1,254,300 2.0%
Sex
Male* 27,038,300 735,100 2.7% 30,467,400 917,900 3.0%
Female 26,431,000 250,200 0.9 † 31,074,900 336,400 1.1 †
Race/ethnicity
White
b
* 37,334,200 485,700 1.3% 42,525,700 647,100 1.5%
Black
b
6,146,400 201,100 3.3 † 6,545,700 250,700 3.8 †
Hispanic 6,680,700 203,100 3.0 † 8,238,400 280,100 3.4 †
Other
b,c
3,307,900 95,500 2.9 † 4,232,500 76,300 1.8
Age
16-17 1,188,300 24,300 ! 2.0% ! 1,143,500 39,200 3.4%
18-24* 8,248,000 261,100 3.2 8,859,700 280,000 3.2
25-44 19,998,800 473,100 2.4 23,518,700 625,500 2.7
45-64 17,290,700 222,900 1.3 † 19,160,700 274,400 1.4 †
65 or older 6,743,400 4,000 !
0.1 † ! 8,859,600 35,200 0.4 †
Note: Details may not sum to totals due to rounding. See appendix table 3 for standard errors.
*Comparison group.
†Dierence with comparison group is signicant at the 95% condence interval.
! Interpret with caution. Estimate is based on 10 or fewer sample cases, or coecient of variation is greater than 50%.
a
Includes threatening use of force, pushing, grabbing, handcung, hitting, kicking, using chemical or pepper spray, using an electroshock weapon,
or pointing a gun. Includes persons reporting threats or use of force during the most recent contact or any earlier contacts with police in the last
12 months. In 2018, but not in 2015, force also included shooting a gun.
b
Excludes persons of Hispanic origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
c
Includes Asians, Native Hawaiians, other Pacic Islanders, American Indians, Alaska Natives, and persons of two or more races.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
Revised February 3, 2023
Most recent contact
Tables 4 and 5 present information on the most recent
police-initiated contact or trac-accident contact with
police reported by residents.
Perception of nonfatal use of force during the
most recent police-initiated contact or trac-
accident contact in the last 12 months, 2018
About 3% of U.S. residents experienced a threat or
use of force during their most recent contact with
police (table 4).
Males (4%) were more likely than females (1%) to
experience force during their most recent contact
with police.
Of residents who experienced a threat or use of force
during their most recent contact with police, about
28% felt that the action was necessary, while 51% felt
it was excessive.
Whites (32%) were more likely than Hispanics (17%)
to view the threat or use of force during their most
recent contact as necessary.
Whites who experienced a threat or use of force
were less likely than blacks to perceive it as excessive
(44% to 63%).
Persons ages 18 to 24 who experienced a threat or
use of force were less likely (16%) than persons ages
25 to 44 (31%) and 45 to 64 (34%) to view the force
as necessary.
TABLE 4
Residents who experienced nonfatal threats or use of force during their most recent police-initiated contact or
trac-accident contact, by resident perception of force as necessary or excessive and demographic characteristics,
2015 and 2018
Demographic
characteristic
2015 2018
Most recent police-initiated or trac-accident contact
a,b
Most recent police-initiated or trac-accident contact
a,b
Of those who
experienced force—
Of those who
experienced force—
Number
Percent
experienced
force
c
Perceived
force as
necessary
d
Perceived
force as
excessive
d
Number
Percent
experienced
force
c
Perceived
force as
necessary
d
Perceived
force as
excessive
d
Total 30,195,900 3.3% 30.2% 48.4% 31,089,300 2.8% 27.5% 50.9%
Sex
Male* 16,522,500 4.4% 28.2% 50.1% 16,569,300 4.2% 26.0% 52.9%
Female 13,673,400 1.8 † 36.2 43.4 14,520,000 1.1 † 34.0 42.5
Race/ethnicity
White
e
* 20,282,400 2.4% 32.4% 42.7% 20,560,500 2.0% 31.7% 44.3%
Black
e
3,886,200 5.2 † 32.0 59.9 ‡ 3,713,800 5.3 † 25.8 62.9 †
Hispanic 3,986,400 5.1 † 20.8 52.5 4,534,400 4.8 † 16.9 † 53.7
Other
e,f
2,040,900 4.7 † 35.5 ! 43.8 2,280,600 1.9 50.1 ‡ ! 43.4 !
Age
16-17 988,900 2.5% ! 35.3% ! 22.1% ! 871,200 3.9% ! 30.4% ! 62.0% !
18-24* 6,107,600 4.3 28.2 49.2 5,890,100 3.5 16.3 59.4
25-44 11,757,400 4.0 34.1 44.0 12,292,200 3.6 30.9 † 46.8
45-64 8,422,100 2.6 † 24.4 58.5 8,435,500 2.0 † 33.9 † 47.7
65 or older 2,919,900 0.1 † ! <0.1 † ! <0.1 † ! 3,600,300 0.4 † ! 4.7 ‡ ! 63.1 !
Note: Details may not sum to totals due to rounding. See appendix table 4 for standard errors.
*Comparison group.
†Dierence with comparison group is signicant at the 95% condence interval.
‡Dierence with comparison group is signicant at the 90% condence interval.
! Interpret with caution. Estimate is based on 10 or fewer sample cases, or coecient of variation is greater than 50%.
a
Includes persons stopped by police during the last 12 months for whom the most recent contact involved being stopped by police in a street stop,
stopped as a driver or passenger in a trac stop, arrested, in a trac accident reported to police, or approached by police for another reason.
b
Denominator includes approximately 0.3% of respondents who had missing data on the most recent contact due to recoding of contacts originally
classied as “other-specify.” See Methodology.
c
Includes threatening use of force, pushing, grabbing, handcung, hitting, kicking, using chemical or pepper spray, using an electroshock weapon, or
pointing a gun during the most recent police-initiated contact or trac-accident contact. In 2018, but not in 2015, force also included shooting a gun.
In 2018, 0.3% of persons had a gun pointed at them or were shot by a gun during their most recent police-initiated contact or trac-accident contact.
d
Respondents who reported experiencing force during their most recent police-initiated contact or trac-accident contact were asked two
questions: if they felt that the actions were necessary and if they felt that any of the actions used against them were excessive. Respondents could respond
“yes” to both questions, “no” to both questions, or “yes” to one question and “no” to the other.
e
Excludes persons of Hispanic origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
f
Includes Asians, Native Hawaiians, other Pacic Islanders, American Indians, Alaska Natives, and persons of two or more races.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 7
Revised February 3, 2023
Type of nonfatal force and race or ethnicity
during the most recent contact
Among the types of force measured in 2018, being
handcued (2%) was the most prevalent type of
force that U.S. residents experienced during their
most recent contact with police in the last 12 months
(table 5).
In 2018, blacks and Hispanics (5% each) were more
likely than whites (2%) or persons of other races
(2%) to experience at least one type of force during
their most recent contact with police.
Fewer than 1% of members of any race or ethnicity
had a gun pointed at them during their most recent
police-initiated contact or trac-accident contact.
TABLE 5
Percent of residents who experienced nonfatal threats or use of force during their most recent police-initiated
contact or trac-accident contact, by resident race or ethnicity and type of force, 2018
Type of force
a
Race/ethnicity
Total White
b
* Black
b
Hispanic Other
b,c
Any 2.8% 2.0% 5.3% † 4.8% † 1.9%
Threat of force 0.7 0.5 2.0 † 1.2 ‡ 0.2 † !
Handcung 2.3 1.6 4.4 † 3.5 † 1.9
Pushing/grabbing/hitting/
kicking 0.7 0.4 1.6 † 1.4 † 0.8 !
Spraying/shocking <0.1 ! 0.1 ! <0.1 ! <0.1 ! <0.1 !
Pointing/shooting gun 0.3 0.1 0.8 † 0.7 ‡ ! 0.4 !
Note: Details may not sum to totals because respondents could experience multiple types of force. See appendix table 5 for standard errors.
*Comparison group.
†Dierence with comparison group is signicant at the 95% condence interval.
‡Dierence with comparison group is signicant at the 90% condence interval.
! Interpret with caution. Estimate is based on 10 or fewer sample cases, or coecient of variation is greater than 50%.
a
Denominator includes approximately 0.3% of respondents who had missing data on the most recent contact due to recoding of contacts originally
classied as “other-specify.” See Methodology. Denominator also includes less than 0.1% of respondents who indicated that police used some other
type of force.
b
Excludes persons of Hispanic origin (e.g., “white” refers to non-Hispanic whites and “black” refers to non-Hispanic blacks).
c
Includes Asians, Native Hawaiians, other Pacic Islanders, American Indians, Alaska Natives, and persons of two or more races.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 8
Methodology
e Police-Public Contact Survey (PPCS) is
a supplemental survey to the National Crime
Victimization Survey (NCVS). e NCVS collects
data on crime against persons age 12 or older from a
nationally representative sample of U.S. households.
e NCVS sample includes persons living in group
quarters (such as dormitories, rooming houses, and
religious-group dwellings) and excludes persons living
in military barracks and institutional settings (such
as correctional or hospital facilities) and homeless
persons.
1
e NCVS and PPCS are administered by the
U.S. Census Bureau on behalf of the Bureau of Justice
Statistics (BJS).
Since 1999, the PPCS has typically been administered
every 3 years.
2
It is administered at the end of the
NCVS interview to persons age 16 or older within
households sampled for the NCVS. Proxy responders
are ineligible to receive the PPCS. In 2018, persons
who completed their NCVS interview in a language
other than English could complete the PPCS in the
same language. All NCVS and PPCS interviews
were conducted using computer-assisted personal
interviewing by telephone or personal visit.
e 2018 PPCS was administered between July 1, 2018
and December 31, 2018. e survey asked respondents
if they had experienced dierent types of specic
police contacts during the prior 12 months. Persons
who reported contact were asked to describe the nature
of the contact and their most recent contact during that
period if more than one contact had occurred.
e PPCS sample increased 47% from 2015 (70,959
respondents) to 2018 (104,324 respondents), due to an
increase in the NCVS sample designed to facilitate the
ability to produce state- and local-level victimization
estimates for the largest 22 states. For more
information on the NCVS sample sizes in 2017 and
2018, see Criminal Victimization, 2017 (NCJ 252472,
BJS, December 2018) and Criminal Victimization, 2018
(NCJ 253043, BJS, September 2019).
1
For more information, see Methodology in Criminal Victimization,
2019 (NCJ 255113, BJS, September 2020).
2
e 2015 PPCS was delayed one year to further improve the
instrument aer the 2011 redesign. For more information, see
Police-Public Contact Survey: Assessment and Recommendations
for Producing Trend Estimates aer 2011 Questionnaire Redesign
(NCJ 250485, BJS, April 2017). e 2018 PPCS marked the survey’s
return to its 3-year cycle.
PPCS non-respondents consisted of persons within an
interviewed NCVS household who did not respond to
the NCVS, NCVS proxy interviews, those who refused
to participate in the PPCS, those who were unavailable
to complete the PPCS, and other non-respondents.
ere were 35,368 non-respondents in the 2018 PPCS.
For the 2018 PPCS, the NCVS household response
rate was 72%, and the person response rate, which
is based on persons in responding households, was
75%. By comparison, for the overall NCVS in 2018,
the household response rate was 73%, and the person
response rate was 82%. PPCS interviews were obtained
from 104,324 of the 139,692 U.S. residents age 16 or
older in the 2018 NCVS sample (75%), which includes
all persons in responding households.
To produce national estimates on police-public
contacts, sample weights designed for the PPCS
were applied to PPCS data so that respondents
represented the entire population, including
non-respondents. PPCS weights were generated
starting with the nal NCVS person weights for each
sample case. A non-response adjustment was applied
to PPCS respondents to reduce the eects of PPCS
non-response. e non-response adjustment was
based on characteristics believed to be correlated with
police contacts, such as respondents’ place of residence,
race or ethnicity, sex, age, and response propensities.
e sample cases in 2018 were weighted by the PPCS
weights to produce a national population estimate of
259,316,200 persons age 16 or older.
Despite non-response adjustments, low overall
response rates may still result in biased estimates if
non-respondents have characteristics associated with
the outcomes of interest that dier from respondents.
e Oce of Management and Budget guidelines
require a non-response bias analysis to be conducted
when the overall response rate for a survey is below
80%. e overall 2018 PPCS response rate was
53.7%. Accordingly, BJS and the U.S. Census Bureau
conducted a non-response bias analysis, comparing
distributions of respondents and non-respondents
and non-response estimates across various household
and demographic characteristics. e analysis also
examined the impact of any dierences on key
PPCS estimates.
Findings from the analysis suggested that any
non-response bias could be corrected through
weighting adjustments. e non-response bias analysis
found signicant dierences in response rates and
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 9
in respondent and non-respondent distributions
among dierent demographic subgroups. However,
non-response weighting adjustments were expected
to minimize these dierences as the demographic
subgroups were used in creating the adjustment. Based
on model-based predicted estimates, there was no
evidence of non-response bias in any of the
police-contact key estimates before or aer
non-response weighting adjustments.
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. e sampling error of an estimate
depends on several factors, including the amount
of variation in responses and sample size. When
the sampling error around an estimate is taken into
account, estimates that appear dierent may not be
statistically signicant.
One measure of the sampling error associated with
an estimate is the standard error. e standard error
may vary from one estimate to the next. Generally,
an estimate with a small 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 are associated with less
precision and reliability and should be interpreted
with caution.
In this report, to generate standard errors from the
NCVS data, BJS used the Taylor Series Linearization
(TSL) method. e 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.
3
Using statistical analysis programs developed
specically for the NCVS, all comparisons in the text
were tested for signicance. e primary test procedure
was the Student’s t-statistic, which tests the dierence
between two sample estimates. Findings described in
this report as higher, lower, or dierent passed a test at
either the 0.05 level (95% condence level) or
0.10 level (90% condence level) of signicance.
3
Woodru, R.S. (2012). A simple method for approximating
the variance of a complicated estimate. Journal of the American
Statistical Association, 66(334), 411-414. https://doi.org/10.1080/01
621459.1971.10482279
Tables in this report should be referenced for testing
on specic ndings.
Estimates and standard errors of the estimates
provided in this report may be used to generate a
condence interval around the estimate as a measure
of the margin of error. e following example
illustrates how standard errors may be used to generate
condence intervals:
Based on the 2018 survey, an estimated 12.5% of
male U.S. residents age 16 or older experienced
some type of police-initiated contact during
the year (see table 1); a standard error of 0.25%
was determined for the estimate (see appendix
table 1). A condence interval around the estimate
was 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). erefore, the 95% condence
interval around the 12.5% estimate from 2018 is
12.5 ± (0.25 * 1.96) or (12.00 to 12.98). In other
words, if BJS used the same sampling method to
select dierent samples and computed an interval
estimate for each sample, it would expect the true
population parameter (the percentage of males
who experienced some type of police-initiated
contact) to fall within the interval estimates 95%
of the time.
For this report, BJS also calculated a coecient of
variation (CV) for all estimates, representing the ratio
of the standard error to the estimate. CVs (not shown
in tables) provide another measure of reliability and a
means for comparing the precision of estimates across
measures with diering levels or metrics.
Missing data for most recent contact
e PPCS screener was divided into two sections, with
the rst series of questions asking about dierent types
of resident-initiated contact and the second asking
about dierent types of police-initiated contact. Aer
each series of questions, respondents were given the
opportunity to report any other contacts that were
not already asked about. U.S. Census Bureau eld
representatives recorded all of these responses as
“other-specify” text responses even though a large
portion of them t into pre-existing categories. For the
current analysis, other-specify responses were recoded
into the correct screener categories when possible.
However, this aected how the respondent answered
questions about the most recent contact. Also, in some
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 10
instances, respondents should have been administered
the questions about the nature of their trac or street
stop but were skipped out of that series of questions.
Missing data accounted for 10.8% of the most recent
contacts in 2018.
2016 NCVS sample redesign
Every 10 years, the NCVS sample is redesigned to
select PSUs based on population data from the most
recent decennial Census. e most recent redesign of
the NCVS sample involved adjusting U.S. population
counts to reect counts from the 2010 decennial
Census rather than those from the 2000 decennial
Census. For more information about the most recent
redesign of the NCVS sample, see National Crime
Victimization Survey, 2016: Technical Documentation
(NCJ 251442, BJS, December 2017). In 2015, the 2010
NCVS sample design was phased into production with
the 2015 data collection. e 2015 NCVS marked the
rst year of the overlap between the 2000 design and
the 2010 design NCVS samples. In the 2015 PPCS, 29%
of all cases were drawn from the 2010 sample, with the
remainder being drawn from the 2000 sample design.
In the 2018 PPCS, all cases and interviews were drawn
from the 2010 sample design.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 11
APPENDIX TABLE 1
Standard errors for table 1: Number and percent of U.S. residents age 16 or older with any police contact, by type
of contact and demographic characteristics, 2018
Demographic
characteristic
U.S. population
age 16 or older
Any contact Police-initiated contact
Resident-initiated
contact Trac-accident contact
Number Percent Number Percent Number Percent Number Percent
Total 2,582,440 1,073,740 0.30% 593,005 0.18% 694,308 0.21% 232,308 0.08%
Sex
Male 1,325,854 594,950 0.37% 367,070 0.25% 382,506 0.25% 136,550 0.10%
Female 1,400,949 575,107 0.33 299,561 0.19 396,261 0.24 149,542 0.10
Race/ethnicity
White 2,196,168 882,891 0.37% 490,125 0.24% 575,845 0.26% 184,786 0.10%
Black 937,498 242,562 0.64 155,244 0.42 153,035 0.44 76,132 0.24
Hispanic 931,575 276,436 0.47 189,382 0.37 178,345 0.34 79,283 0.17
Other 471,120 159,775 0.63 98,356 0.42 110,431 0.47 56,476 0.25
Age
16-17 255,343 69,978 0.92% 58,153 0.74% 42,511 0.53% 32,475 0.41%
18-24 638,394 305,767 0.75 223,771 0.63 183,184 0.52 105,187 0.31
25-44 1,077,657 510,340 0.43 310,744 0.29 328,786 0.30 122,431 0.13
45-64 945,394 397,820 0.37 214,735 0.22 296,149 0.29 87,875 0.10
65 or older 690,952 219,460 0.34 100,948 0.17 165,663 0.28 62,392 0.12
Household income
$24,999 or less 873,420 336,256 0.57% 213,991 0.38% 203,734 0.37% 84,797 0.16%
$25,000-$49,999 990,852 369,258 0.48 217,426 0.30 242,909 0.33 92,802 0.14
$50,000-$74,999 839,491 276,084 0.44 175,061 0.33 193,497 0.33 79,603 0.15
$75,000 or more 1,430,414 497,188 0.36 302,036 0.27 343,858 0.26 137,297 0.13
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2018.
APPENDIX TABLE 2
Standard errors for table 2: Residents with police contact, by reason for contact,
2015 and 2018
Reason for contact
2015 2018
Number Percent Number Percent
Any 998,696 0.30% 1,073,740 0.30%
Resident-initiated contact 560,931 0.19% 694,308 0.21%
Reported possible crime 383,674 0.13 412,694 0.13
Reported non-crime emergency 259,485 0.09 249,595 0.09
Reported non-emergency ~ ~ 286,859 0.10
Block watch 117,969 0.05 98,499 0.04
Other 105,856 0.04 49,544 0.02
Trac-accident contact 216,355 0.08% 232,308 0.08%
Police-initiated contact 617,902 0.20% 593,005 0.18%
Driver during a trac stop 447,052 0.16 443,859 0.16
Passenger during a trac stop 223,034 0.08 186,564 0.07
Street stop 131,084 0.05 148,721 0.06
Arrested only 67,043 0.03 43,045 0.02
Other 106,102 0.04 151,032 0.05
~Not applicable.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 12
Revised February 3, 2023
APPENDIX TABLE 3
Standard errors for table 3: Residents who experienced nonfatal threats or use of force during contacts with
police, by demographic characteristics, 2015 and 2018
Demographic characteristic
2015 2018
Persons age 16
or older with any
police contact
Experienced threats or force
at any time during the year
Persons age 16
or older with any
police contact
Experienced threats or force
at any time during the year
Number Percent Number Percent
Total 998,696 74,277 0.13% 1,073,740 81,519 0.13%
Sex
Male 552,138 63,004 0.22% 594,950 68,306 0.22%
Female 537,568 34,301 0.13 575,107 34,660 0.11
Race/ethnicity
White 742,753 51,452 0.14% 882,891 50,964 0.12%
Black 252,583 33,752 0.54 242,562 34,872 0.54
Hispanic 285,587 36,167 0.51 276,436 34,110 0.41
Other 154,572 23,497 0.69 159,775 18,740 0.43
Age
16-17 94,639 12,478 1.03% 69,978 13,215 1.15%
18-24 300,725 45,098 0.55 305,767 47,055 0.53
25-44 431,294 48,030 0.23 510,340 54,278 0.23
45-64 373,907 27,091 0.15 397,820 28,088 0.15
65 or older 195,756 2,796 0.04 219,460 8,182 0.09
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
APPENDIX TABLE 4
Standard errors for table 4: Residents who experienced nonfatal threats or use of force during their most recent
police-initiated contact or trac-accident contact, by resident perception of force as necessary or excessive and
demographic characteristics, 2015 and 2018
Demographic
characteristic
2015 2018
Most recent police-initiated or trac-accident contact Most recent police-initiated or trac-accident contact
Of those who
experienced force—
Of those who
experienced force—
Number
Percent
experienced
force
Perceived
force as
necessary
Perceived
force as
excessive Number
Percent
experienced
force
Perceived
force as
necessary
Perceived
force as
excessive
Total 649,600 0.23% 3.20% 3.51% 585,206 0.21% 2.40% 2.99%
Sex
Male 388,272 0.37% 3.66% 4.06% 380,333 0.35% 2.61% 3.45%
Female 348,227 0.24 6.55 6.60 290,205 0.17 5.33 5.50
Race/ethnicity
White 461,083 0.25% 4.70% 4.50% 485,481 0.20% 3.08% 3.35%
Black 194,060 0.84 7.50 7.66 160,283 0.86 4.91 5.44
Hispanic 209,932 0.85 5.48 8.36 202,465 0.72 5.84 8.60
Other 115,175
1.12 11.47 12.66 99,888 0.69 8.87 8.64
Age
16-17 85,289 1.24% 26.39% 20.06% 60,941 1.43% 15.45% 15.61%
18-24 248,610 0.73 7.02 8.33 227,371 0.67 4.16 7.76
25-44 300,496 0.39 4.19 4.52 315,257 0.36 3.44 3.46
45-64 228,594 0.31 5.93 7.38 196,433 0.27 4.73 4.69
65 or older 115,645 0.10 ~ ~ 112,565 0.16 4.65 14.58
~Not applicable.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2015 and 2018.
Contacts Between Police and the Public, 2018 – Statistical Tables | December 2020 13
Revised February 3, 2023
APPENDIX TABLE 5
Standard errors for table 5: Percent of residents who experienced nonfatal threats or use of force during
their most recent police-initiated contact or trac-accident contact, by resident race or ethnicity and type
of force, 2018
Type of force
Race/ethnicity
Total White Black Hispanic Other
Any 0.21% 0.20% 0.86% 0.72% 0.69%
Threat of force 0.12 0.10 0.62 0.39 0.12
Handcung 0.19 0.17 0.79 0.65 0.69
Pushing/grabbing/
hitting/kicking 0.11 0.09 0.56 0.39 0.47
Spraying/shocking ~ 0.05 ~ ~ ~
Pointing/shooting gun 0.08 0.05 0.30 0.29 0.42
~Not applicable.
Source: Bureau of Justice Statistics, Police-Public Contact Survey, 2018.
e Bureau of Justice Statistics of the U.S. Department of Justice is the
principal federal agency responsible for measuring crime, criminal
victimization, criminal oenders, 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. Jerey H. Anderson is the director.
is report was written by Erika Harrell and Elizabeth Davis. Grace
Kena and Stephanie Mueller veried the report.
eodore Robinson and Edrienne Su edited this report. eodore
Robinson produced this report.
December 2020, NCJ 255730
Oce of Justice Programs
Building Solutions • Supporting Communities • Advancing Justice
www.ojp.gov
