Explore crime, fear, and police contact

Explore data on the intersection of reported, unreported, and perceived crime with police contact in the United States.

Data Overview

Learn about how perceptions of crime and crime victimization differ within the U.S., and how different groups feel about and experience crime and police contact.

Explore a Subgroup

Investigate how crime victimization, perceived crime, and police contact relate for a specific subgroup within the population and how they compare to the population overall.

Compare Groups

Break down rates of crime victimization and police contact and beliefs about crime by age, sex, race/ethnicity, or geographic region.

Synthetic Data Explorer

Explore relationships of reported, unreported, and perceived crime in a synthetic data set combining measures from three surveys.



This application presents public use data from three different sources:

NCVS

The National Crime Victimization Survey (NCVS) is a product of the Bureau of Justice Statistics and collects information on both reported and unreported nonfatal personal crimes and household property crimes. Read more about the NCVS on the About page.

GSS

The General Social Survey (GSS) is administered by NORC at the University of Chicago and surveys adults in the United States to understand a variety of opinions, attitudes and behaviors in American society. Read more about the GSS on the About page.

PPCS

The Police-Public Contact Survey (PPCS) is a product of the Bureau of Justice Statistics and collects information on the characteristics of persons who had some type of contact with police during the past year. Read more about the PPCS on the About page.

Data Overview
Explore a Subgroup
Compare Groups
Synthetic Data Explorer

Explore crime, fear, and police contact in the U.S.

Discussion of crime among family, friends, and in the media is often centered around the idea of fear, where people are afraid of being the next crime victim. Often, those who are most afraid of crime are the ones who are, statistically, least likely to experience it or encounter it on a regular basis. There are others, however, for whom fear of crime is more warranted but might not manifest in the same way. This intersection between perceptions of and experience with crime and policing is the motivation behind this collation of nationally representative survey data. Through this data visualization explorer, we hope to spark interest in researcher and lay audiences in studying how perceptions and experiences of crime are uniquely intertwined.

As a starting point, consider the percentage of U.S. adults who report being afraid to walk alone at night in their own neighborhood. Perhaps unsurprisingly, women, overall, report greater rates of fear than men, but this is especially true for young, black women.




This gender, age, and race gap in fear of crime extends to attitudes related to public policy and spending. For example, an even more exaggerated gap exists in beliefs about whether the U.S. is spending enough on halting the rising crime rate. Specifically, over 85% of young, black women said that the U.S. was spending too little on crime reduction, which towers over their young, white male counterparts where less than 7% believe the U.S. is spending too little on crime.

Patterns of actual crime reported to the police show a different story than the patterns observed in perception of crime. Although people aged 65 and older thought that the U.S. spent too little on crime compared to people younger than them, their personal crime victimization rate was lowest among all age groups. The comparative personal reported victimization rate for young, black females, however, is consistent with the differences observed in perceptions of crime, as they had the highest reported personal victimization rate compared to all other groups.




Of course, one can have negative perceptions of crime without having experienced it personally, as we know that crime affects more people in communities than just those directly involved in any one incident. But another potential contributing factor in this seeming discrepancy between perception and experience is that not all crime that is committed is reported to the police. In fact, a key component of the NCVS is that it not only measures experiences with crime that was officially report to local authorities, but also measures experience with crime that went unreported. As one might imagine, patterns of crime victimization for unreported crime can look very different from demographic patterns of reported crime victimization. In contrast to reported crime victimization, young, black males reported higher rates of unreported crime victimization compared to all other groups, suggesting that demographic patterns of crime victimization, overall, are different than patterns reflected in official crime statistics, especially for certain subgroups.

Finally, do patterns of crime perception versus experience relate to how people interact with police in general, especially in terms of who is willing to initiate contact with police? One could make an argument that crime victims might be more likely to initiate contact with police, potentially to stay vigilant against the next threat of crime. Alternatively, those who have more fear related to crime, might be more likely to initiate contact with police to assuage fears related to potential crime activity. Observed patterns of resident-initiated contact with police from the PPCS show that police contacting patterns do not necessarily line up with demographic patterns related to perceptions of or experience with crime in general. Instead, there is an overall pattern where white residents seem to have an edge on contacting police compared to other racial groups, a pattern that appears to be consistent across age groups.




The demographic patterns in crime perception and experience highlighted here provide a brief example of the complexities inherent in how U.S. citizens think about crime. We built this data visualization explorer to give others a chance to explore these complexities in a single, coherent system that considers the intricacies of each survey from which these data come.

Please explore for yourself!



Data Overview
Explore a Subgroup
Compare Groups
Synthetic Data Explorer

Explore a subgroup.

Use the filters to select a population subgroup of interest.





NCVS measures
GSS measures
PPCS measures


Data Overview
Explore a Subgroup
Compare Groups
Synthetic Data Explorer

Compare across groups.

Select a variable and a demographic grouping to see how rates compare.

Click on a group in the graph to learn more.

Selected Group



Overall





Data Overview
Explore a Subgroup
Compare Groups
Synthetic Data Explorer

Explore relationships using synthetic data

One challenge of examining variables distributed across multiple complex surveys is the limitation of not being able to directly compare outcomes at the microdata level. For example, no single survey asks respondents about their personal crime victimization, fear of crime, and how often they have had contact with the police all together. However, one can create a synthetic dataset that combines variables of interest at the microdata level if there are enough shared variables measured across datasets. Using statistical matching techniques, we combined all our crime-related outcomes of interest into one micro dataset where one can examine novel relationships between variables that, otherwise, could not be examined in the source data.

Specifically, we used weighted random hot deck matching procedures to combine estimates across surveys that matches on the demographic variables (i.e., age, sex, race, region) while accounting for the survey weights of each donor survey in the process. Respondents in NCVS were matched to respondents in PPCS based on the combination of the demographic variables. A case was then randomly selected from the NCVS to be imputed onto the PPCS dataset, weighted by the NCVS survey weight (e.g., the larger the weight, the higher the chance of being selected). The same statistical matching process was then carried out for imputing estimates from the new NCVS/PPCS dataset onto the GSS dataset.

The final synthetic dataset contains microdata on all outcomes of interest as well as demographic information. This dataset should be used for exploratory purposes only. Estimates of relationships between synthesized variables do not represent real-world data but only approximate the relationships based on broad patterns observed across the shared variables. User and researcher discretion is strongly advised.



Do crime victims perceive crime or interact with police differently than non-victims?



How does fear relate to experience?


Do those who experience police contact perceive or experience crime differently than those who do not?






About this data tool

This interactive data tool is an entry into the 2023 Joint Statistical Meetings Data Challenge Expo, and is aimed at letting viewers explore data on crime from multiple survey sources in one location. 'Crime, Fear, and Police Contact' presents data on six measures from the National Crime Victimization Survey (NCVS), three measures from the General Social Survey (GSS), and three measures from the Police-Public Contact Survey (PPCS). The measures from the NCVS focus on crime victimization, those from GSS focus on beliefs surrounding crime and policing, and the PPCS measures provide details on the public's interaction with law enforcement.

Users can explore each of these measures by age groupings, gender, race/ethnicity, and geographic region using the interactive data tools on this site.

NCVS
The NCVS is an annual, nationally representative survey of about 240,000 people which collects information on nonfatal personal and household property crimes. NCVS survey estimates were obtained using 2020 data within the concatenated household file, concatenated person file, and the concatenated incident file. The following demographic and outcomes variables were used from NCVS with estimates generated using appropriate survey design and weighting variables.

Demographic Variables

Demographic

Grouping

Crime Type

NCVS File

NCVS Variable Name

Age

18-24, 25-44, 45-64, 65+

Property

Household

V2042

Personal

Person

V3014

Sex

Male, Female

Property

Household

V2045

Personal

Person

V3018

Race

White, Black, Hispanic, Other

Property

Household

V2049A & V2050

Personal

Person

V3024A & V3023A

Region

Midwest, Northeast, South, West

Property

Household

V2127B

Personal

Person

Outcome Variables

Outcome

Definition

Operationalization

Personal rate

The rate of personal crime victimization (per 100 people) which includes rape/sexual assault, robbery, aggravated assault, simple assault, and/or personal larceny.

V4529 <= 23

Personal reported rate

The rate of personal crime victimization (per 100 people) which includes rape/sexual assault, robbery, aggravated assault, simple assault, and/or personal larceny that was reported to police.

V4529 <= 23 and V4399 = 1

Personal unreported rate

The rate of personal crime victimization (per 100 people) which includes rape/sexual assault, robbery, aggravated assault, simple assault, and/or personal larceny that was not reported to police.

V4529 <= 23 and V4399 = 2

Property rate

The rate of burglary/ trespassing, motor vehicle theft, and/or theft (per 100 households).

V4529 > 23

Property reported rate

The rate of burglary/ trespassing, motor vehicle theft, and/or theft (per 100 households) that was reported to police.

V4529 > 23 and V4399 = 1

Property unreported rate

The rate of burglary/ trespassing, motor vehicle theft, and/or theft (per 100 households) that was not reported to police.

V4529 > 23 and V4399 = 2

Read more about the NCVS here.

GSS
The GSS is a nationally representative survey of adults in the United States conducted since 1972 that collects data on opinions, attitudes, and behaviors of contemporary American society. GSS survey estimates were obtained using data from the 2021 cross-section study. The following demographic and outcomes variables were used from GSS with estimates generated using appropriate survey design and weighting variables.

Demographic Variables

Demographic

Grouping

GSS Variable Name

Age

18-24, 25-44, 45-64, 65+

AGE

Sex

Male, Female

SEX

Race

White, Black, Hispanic, Other

RACE & HISPANIC

Region

Midwest, Northeast, South, West

REGION

Outcome Variables

Outcome

Definition

Operationalization

Afraid to walk alone at night

Percent of people who indicated they would be afraid to walk alone at night within a mile of their location.

FEAR = 1

Too little spent on crime reduction

Percent of people who indicated we are spending too little on halting the rising crime rate.

NATCRIME = 1

Too little spent on law enforcement

Percent of people who indicated we are spending too little on law enforcement.

NATCRIMY = 1

Read more about the GSS here.

PPCS
The PPCS is a nationally representative survey of people who had some type of contact with police during the past year, including those who contacted the police to report a crime or were pulled over in a traffic stop. The PPCS is a supplement to the NCVS. PPCS survey estimates were obtained from the 2020 Police-Public Contact Survey. The following demographic and outcomes variables were used from PPCS with estimates generated using appropriate survey design and weighting variables.

Demographic Variables

Demographic

Grouping

PPCS Variable Name

Age

18-24, 25-44, 45-64, 65+

AGE

Sex

Male, Female

SEX

Race

White, Black, Hispanic, Other

HISP

Outcome Variables

Outcome

Definition

Operationalization

Any contact with police

Percentage of people who indicated they had any contact with the police within the past year.

V10A = 1 or V10B = 1 or V10C = 1 or V10D = 1 or V10E = 1 or V10F = 1 or V10G = 1 or V10H = 1 or V10I = 1 or V10J = 1 or V10K = 1 or V10M = 1 or V10N = 1 or V22A = 1 or V22B = 1 or V22C = 1 or V22D = 1 or V22E = 1 or V22F = 1 or V22G = 1 or V22H = 1 or V22I = 1 or V22J = 1 or V22K = 1 or V22M = 1 or V22N = 1 or V220 = 1 or V22P = 1 or V23A = 1 or V23E = 1 or V23I = 1 or V23M = 1 or V23Q = 1 or V23T = 1 or V23W = 1 or V23Z = 1 or V24C = 1 or (V24J = 1 and V24H = 1 and V22A = 8 and V22B = 8 and V22C = 8 and V22D = 8 and V22E = 8 and V22F = 8 and V22G = 8 and V22H = 8 and V22I = 8 and V22J = 8 and V22K = 8 and V22L = 8 and V22M = 8 and V22N = 8 and V22O = 8)

Resident-initiated contact with police

Percentage of people who indicated they had contact with the police that they initiated within the past year.

V10A = 1 or V10B = 1 or V10C = 1 or V10G = 1 or V10H = 1 or V10J = 1 or V10M = 1 or V10N = 1 or V22A = 1 or V22B = 1 or V22I = 1 or V22P = 1 or V23A = 1 or V23E = 1 or V23I = 1 or V23M = 1

Police-initiated contact

Percentage of people who indicated they had contact with the police that the police initiated within the past year.

V10E = 1 or V10F = 1 or V10I = 1 or V10K = 1 or V22D = 1 or V22E = 1 or V22F = 1 or V22G = 1 or V22H = 1 or V22J = 1 or V22K = 1 or V22M = 1 or V22N = 1 or V220 = 1 or V23T = 1 or V23W = 1 or V23Z = 1 or V24C = 1 or (V24J = 1 and V24H = 1 and V22A = 8 and V22B = 8 and V22C = 8 and V22D = 8 and V22E = 8 and V22F = 8 and V22G = 8 and V22H = 8 and V22I = 8 and V22J = 8 and V22K = 8 and V22L = 8 and V22M = 8 and V22N = 8 and V22O = 8)

Read more about the PPCS here.

Contact us

Anthony Washburn
Research Methodologist
Project Lead
washburn-anthony@norc.org

Kiegan Rice
Senior Statistician
Lead Developer
rice-kiegan@norc.org

This tool was developed by Taylor Wing, Henry Beimers, Anthony Washburn, and Kiegan Rice at NORC at the University of Chicago. Read more about NORC here.