Policing by the Numbers


Mounting criticism of American policing following the violent death of George Floyd and others at the hands of law enforcement has prompted calls for enhanced police accountability and shifts in funding from police to social services and community-led safety efforts. At the same time, violent crime has spiked in many U.S. cities, with some experiencing record-setting numbers of homicides.

Efforts to develop responses that achieve the twin goals of crime control and justice must be grounded in hard data and research evidence, as well as personal and professional experience. This series of charts is intended to inform debates about the future of policing in America. It paints a statistical portrait of trends in key areas, ranging from the size and makeup of the nation’s police agencies to spending, reported crime and victimization rates, people killed by police and officers killed in the line of duty, and public perceptions of and trust in law enforcement.

The included metrics have been compiled to shed light on the bigger picture trends in policing in recent years as governments, agencies, and communities wrestle with priorities and reforms at a local level. The overall picture is complex and nuanced, reflecting the nature and extent of the challenges facing the country and individual communities.


Police Agency Size and Composition

Number of Officers

While the number of sworn police officers has increased by 26% since 1987, that expansion has not kept pace with the growth of the general population. As a result, there were 11.5% fewer officers per capita in 2016 than there were in 1987. However, because of the decline in serious (Part I) crime since the early 1990s, the number of officers per serious crime has steadily grown, nearly doubling from 41.1 to 76.5 officers per 1,000 Part 1 crimes.

Full-Time, Sworn Law Enforcement Personnel

Female Officers

Both the number and share of female officers have increased over time. The number of female officers more than doubled from 1987 to 2016, increasing by 112%, while their share grew from 7.6% to 12.3% of local officers during that same period. However, most of the growth in representation of female officers occurred in the 1990s, after which their share of the total number of sworn officers has remained relatively stagnant.

Share of Female Officers in Local Police Agencies

Non-White Officers

The number and share of Black officers have increased by about 22% from 1987 to 2016, at which time Black people made up 11.4% of police personnel and 13% of the U.S. population. By contrast, the share of Hispanic officers has quadrupled since 1987, rising to 12.5% of officers in 2016, but remains lower than the share of Hispanics/Latinos in the general population (18%). However, racial and ethnic representation of officers as compared to the local population varies by agency size and geography.

Share of Officers in Local Police Agencies by Race


Police Spending

State and Local Expenditures

Spending on police nearly tripled (increasing by 168%) between 1977 and 2018, when adjusted for inflation. Per capita spending on police also grew during this period, but at a lower level, close to doubling (83% higher in 2018 than in 1977). However, spending on law enforcement accounts for a small share of total state and local government expenditures, and that portion has remained virtually unchanged over the past four decades, hovering at around 3.7%. This indicates that total expenditures on other state and local functions, such as K-12 education, post-secondary education, public welfare and healthcare, and roadways, increased at roughly the same rate as police expenditures over the past four decades.


Police Contacts with the Public

Police-Initiated Contacts

National data on the nature and type of contacts between police officers and members of the public, and the reasons for those contacts, are not routinely collected. But one federal source gathers such data every few years through a nationally representative sample of households. The most recent dataset that reports contacts by race and ethnicity of respondent is from 2015. These data indicate that in 2015, a large majority of contacts that police initiated with the public were related to traffic stops. Black drivers were more likely to be pulled over by police than White drivers, and Hispanic drivers were less likely. The rate of Black pedestrians stopped by police was 1.7 times greater than the rate for White pedestrians.

Public-Initiated Contacts

Members of the public also initiate contacts with the police. The most common reason for public-initiated contact is to report a crime. White people are significantly more likely to call police to report a possible crime or a non-crime emergency.

Police Contacts Over Time

The rate of police-initiated contacts with the public fell by 19% from 2011 to 2015, from 26,071 per 100,000 residents to 21,085. In 2015, the most recent year for which nationally representative data are available, Black people were less likely to experience police-initiated contact than in 2011, but still slightly more likely than White people, whereas Hispanic people were less likely to be on the receiving end of a police-initiated contact. White people were much more likely to initiate contact with police than were people of other races and ethnicities, and this difference has grown over time. Because traffic accidents do not fall neatly into a “police-initiated” or “public-initiated” category, they are analyzed separately. There was little change in the rate at which members of the public and officers came into contact because of a traffic accident. Rates of traffic accidents were lower for Hispanic people, but similar among people who identify as White, Black, and of other races.

Rate of Police-Public Contacts by Type and Race

Overall Crime Trends

The rates of serious crimes reported to the police have declined dramatically since the early 1990s. Reported property crime was 58% lower in 2019 than it was at its peak in 1991. Violent crime was 52% lower during that same period, but the decline has been uneven in the last 15 years. A recent analysis by the Council on Criminal Justice indicates that murder and aggravated assault have increased during 2020, while property crime fell.

Violent and Property Crime Rates by Offense Type

Homicide Trends

Homicide victimization rates have fallen over time, especially for Black people, whose rates of victimization plummeted from a peak of 37 per 100,000 in 1993 to a low of 15 per 100,000 in 2014. However, the rates for both Black and White people increased between 2014 and 2016. Over the past 20 years, racial disparity in homicide victimization rates has been relatively stable, at six times greater for Black people than White people.

Homicide Victimization Rates by Race


Victimization Reporting Rates

Crime Reported to Police

One metric of the degree of trust members of the public have in police is the rate at which they report crimes they have experienced. According to victimization surveys, the average percentage of crimes the public reports to police is similar in recent years to what it was in the early 1990s when the surveys began, with reporting rates of 41% for violent crimes and 33% for property crimes. However, after trending slightly upward during the late 1990s and 2000s, over the past ten years, the share of victims who reported violent crimes such as robbery, assault, and rape to the police declined 10 percentage points, from a high of 51.1% in 2010 to 40.9% in 2019. In the same time period, the share of households that reported property crimes such as burglary and auto theft declined by 6.7 percentage points, from 39.2% to 32.5%.

Share of Victimizations Reported to the Police

Crime Reported by Race

Victimization reporting rates by race and ethnicity are remarkably similar for both violent and property crimes. However, Black people are slightly more likely to report experiencing a crime than members of other racial groups, particularly with regard to violent offenses. From 1995 to 2019, 51% of victims who were Black reported their violent person crimes to the police compared to 44% of White victims and 47% of Hispanic victims.

Victimizations Reported to the Police by Race


Arrest and Clearance Rates

Arrest Rates

Arrest rates have declined in recent years, particularly for public order offenses. However, those declines are largely explained by dramatic reductions in arrest rates for Black people. Racial disparities in arrest rates have narrowed across crime types over time, although Black people remain far more likely to be arrested for violent and drug offenses than White people. Of particular note is the reduction in disparity between Black and White arrest rates for drug offenses, which dropped from a high of 4.9-to-1 in 1991 to a low of 2-to-1 in 2019. The arrest rate disparity for violent crime also dropped from a peak of 6.2-to-1 in 1987 to 3.5-to-1 in 2019.

Arrest Rates by Crime Type and Race

Clearance Rates

The crime clearance rate – the proportion of crimes that result in an arrest or are otherwise “cleared” by police – is one metric of police effectiveness. Clearance rates have remained consistent for the past three decades and across crime categories. However, homicide clearance rates—a subset of violent crimes—dropped by 11 percentage points (or by 16.5%) between 1987 and 2018.

Violent Crime, Property Crime, and Homicide Clearance Rates


Officer-Involved Fatalities

Confirmed cases of people killed by law enforcement are not systematically documented by any one entity. Moreover, no single data source has fully verified comprehensive details on contextual factors surrounding these fatalities, such as whether the victim was armed or unarmed, the number of officers present, the race and ethnicities of both victim and officer, and whether the case was criminally charged. Each available data source tells a somewhat different story, in part reflecting the scope and limitations of the data.

The Fatal Encounters data depicts the highest volume of deaths, as these include cases for which law enforcement personnel were present but may have had no direct role in the fatality. The Mapping Police Violence data draws from multiple sources and extracts relevant details from the three largest and most comprehensive crowdsourced databases, FatalEncounters.org, the U.S. Police Shootings Database, and KilledbyPolice.net. These original data sources are further researched for verification purposes and to extract additional information with which to improve the quality and comprehensiveness of the data. The Washington Post tracks people who were fatally shot by police with a firearm but not those killed by other means. Both Mapping Police Violence and Washington Post data show similar trends, but both projects are new and the data are only available for the past several years.

The two data sources that most closely align with each other, the CDC's mortality statistics and the FBI’s justifiable homicides by law enforcement data, are both generated by events reported by law enforcement agencies themselves. One would expect the volume of these counts to be lower, as the chart illustrates, because not all fatalities of members of the public by police are justified. These data sources suffer from the lack of third party validation about what is considered “justifiable homicide.”

Officer-Involved Fatalities

An examination of all five data sources on police-involved homicides indicates no clear aggregate trend over the past five years. However, the estimated number of civilians killed in incidents that directly involve the police averages between 1,000 and 1,100 per year.

Officer-Involved Civilian Fatalities

By Race and Ethnicity

While the previous chart depicts the numbers of annual police fatalities (or those in which officers were involved), only Mapping Police Violence allows for an analysis of the rate of fatalities by race and ethnicity. Roughly three out of every four people killed by police are White, but the rate at which people of color are killed by police is considerably higher. Black people are four times more likely and Hispanic people are two times more likely than White people to be killed by police.

People Killed by Police by Race and Ethnicity

By Type of Jurisdiction

The estimated number of people killed annually by police has declined in urban areas in recent years, dropping by 27% from 2013 to 2019, from 363 to 264, according to the Mapping Police Violence data. Over the same period, the number of people killed by police in rural areas increased by 34%, from 200 to 268. Suburban areas experienced a marginal (8%) uptick in police killings during this same period, with the number rising from 516 to 560.

People Killed by Police in Urban, Suburban, and Rural Areas

By Race and Jurisdiction

From 2013 to 2019, trends in the number of people killed by police by type of location and race depict a clear pattern of decline for all races and ethnicities in urban areas, with a sizeable (24%) drop for Black people in urban areas (from 143 to 109). During this same period, the number of Hispanic people killed by police in suburban areas increased by over a third (34%), from 75 to 114. In rural areas, killings of White people rose 34%, marking the largest increase among racial groups.

People Killed by Police, by Geography and Race

By Cause of Death

Fatal Encounters is the best source of data for discerning the cause of civilian death, although it includes cases in which law enforcement may have had no direct role in the fatality. By far the most common means of death is by firearm, followed by motor vehicle (e.g, vehicle crashes and pedestrian fatalities by motor vehicle). Death by asphyxiation, which would include police use of chokeholds and other methods to subdue people by neck restraint or compression, represents a very small share of all fatalities and one that has decreased over time.

Fatalities in Which Police Were Involved, by Cause of Death


Officers Killed or Assaulted

Officers Killed

Over the past three decades, the number of law enforcement officers feloniously killed declined from 74 in 1987 to 48 in 2018 and the rate of officers who were feloniously killed followed a similar downward trend, roughly halving over the same time period.

Officers Feloniously Killed

Officers Assaulted

For 25 years, both the number and rate of officers assaulted declined, but the rate of officers assaulted experienced a notable and steady uptick from 2014 to 2019.

Officers Feloniously Assaulted


Public Opinion of Police

Confidence in the Police

While White and Black people report crimes at similar rates, suggesting similar levels of confidence in the police, they provide starkly different responses when questioned directly about their trust levels. In polls, White people consistently report higher rates of confidence in the police and that high confidence has been stable over time. Overall, confidence in 2020 among White people is about the same as it was in 1994. Black people report having much lower levels of confidence in the police, averaging about 40% lower and showing considerably more volatility over time. Most notably, confidence in police among Black people plummeted to a decades-long low in 2020, 16 percentage points below the 1994 level.

Confidence in Police

Respondents who said they have a "great deal" or "quite a lot" of confidence in the police.

Police Interactions

In a survey taken in the weeks after the police killing of George Floyd, a majority of Black respondents (62%) were either very or somewhat confident that police would treat them with courtesy and respect. That is 29 percentage points lower than White respondents and 16 points lower than Hispanic respondents. Black people are much less likely to express that they are “very confident” that they will be treated respectfully.

Public Confidence in Police Interactions

Question: If you had an interaction with police in your area, how confident are you that they would treat you with courtesy and respect?

Defunding the Police

In a survey taken shortly after the death of George Floyd and the outbreak of protests across the nation, relatively few Americans favored cutting spending on police. Three-quarters of survey respondents said police funding in their area should stay the same or increase while one quarter said it should be cut either by a little or a lot.

Americans' Views on Police Funding

Percent of respondents who say spending on policing in their area should...

Policing Preferences

Just 14% of Americans want police to spend less time in their neighborhoods, according to a survey taken about a month after the death of George Floyd. Asian Americans (28%) were the most likely to say they want less police presence. Large majorities of people, regardless of race or ethnicity, prefer that police spend the same or more time in their areas.

Preference for Amount of Time Police Spend in Your Area

Question: Would you rather the police spend more time, the same amount of time, or less time as they currently spend in your area?