3 Patterns of Crime

What crime patterns describe

A crime pattern describes a population-level tendency across time, places, or groups. It can help researchers understand when and where offenses occur, who is more exposed to victimization, and how patterns differ by offense. A group pattern does not predict what any particular person will do.

Reading counts and rates

Counts report how many incidents occurred. A rate relates incidents to a population or amount of exposure, so a city with more incidents than a small town could still have a lower crime rate per resident. Comparisons are most meaningful when they use consistent offense definitions, time periods, and data sources.

What crime data sources capture

Two major U.S. data sources capture different parts of the picture.

The National Incident-Based Reporting System (NIBRS), maintained by the FBI, records incidents reported to law enforcement. Its incident details can include offense, location, time, and victim and offender characteristics. Missing or incomplete agency reporting can affect estimates, and police records exclude crimes that were not reported or recorded.

The National Crime Victimization Survey (NCVS), maintained by the Bureau of Justice Statistics, asks people age $12$ and older about selected nonfatal crimes, whether or not those crimes were reported to police. Its coverage is limited to the offenses and populations included in the survey. Neither source captures every offense, so their results should be interpreted in light of their different coverage and limitations.

How patterns change over time

Crime rates can change from year to year, and short-term changes may differ from longer-term patterns. In Bureau of Justice Statistics estimates of offenses reported to police, the violent-offense rate fell from $393.9$ per $100{,}000$ people in 2023 to $370.8$ in 2024. The property-offense rate fell from $2{,}019.7$ to $1{,}835.1$ over the same period. These figures describe offenses reported to police, not every crime that occurred.

Patterns can also vary within a day, week, or year. The timing of an offense can reflect its opportunities and setting: theft, for example, may be tied to when property or businesses are accessible, while violence may be linked to particular social situations. Incident-level data allow researchers to compare timing by offense instead of treating all crime as having one schedule. Such comparisons should account for differences in reporting and recording.

The age–crime curve

The age–crime curve describes the tendency for recorded offending to rise during adolescence, peak in the teenage years, and decline in early adulthood. The pattern is not identical for every offense: violence tends to peak later than property crime. The curve describes group averages; most young people do not become persistent offenders, and individual pathways vary.

Patterns across places

Crime is geographically concentrated: incidents are often clustered in particular small areas, sometimes called hot spots. Local patterns can reflect differences in routine activity, available targets, guardianship, land use, social conditions, and exposure to risk. The relevant setting depends on the offense; residential areas, commercial locations, roads, and online spaces create different opportunities. A high-crime area does not imply that all of its residents are involved in crime.

Geographic comparisons require care because a neighborhood’s raw count is affected by how many people live, work, travel, or gather there. Police-recorded maps also reflect where crimes are reported and how agencies classify and submit incidents. NIBRS supports analysis by incident location and offense, but missing or incomplete agency reporting can affect estimates.

Demographic patterns and disparities

Age, sex, and other demographic characteristics can be associated with differences in offending or victimization, but those associations do not explain the differences on their own. They can reflect unequal exposure to violence, social and economic conditions, access to protection, and differences in enforcement or reporting. Researchers distinguish rates from counts and examine the context behind group differences.

Bureau of Justice Statistics estimates indicate that $1.95\%$ of people ages $12$–$17$ experienced one or more violent victimizations in 2024, compared with $1.45\%$ in 2023. The overall share of people age $12$ or older who experienced at least one violent victimization was $1.45\%$ in 2024, similar to 2023. These are National Crime Victimization Survey estimates using its legacy instrument. The redesigned instrument was administered alongside it in 2024, and estimates from that instrument were to be released separately.

Homicide victimization rates also differed across groups in 2023. The rate was $9.3$ per $100{,}000$ for males and $2.6$ for females. By race, the rate for Black people was $21.3$ per $100{,}000$, compared with $3.2$ for White people. These disparities describe unequal outcomes; they do not mean that race itself causes crime. Understanding them requires attention to social conditions, exposure, and other relevant factors.

Why offense type and coverage matter

Crime does not follow one universal pattern. Homicide, assault, robbery, burglary, motor vehicle theft, and fraud involve different situations, targets, and reporting practices, so a trend in one offense may rise while another falls. From 2023 to 2024, both violent and property offenses known to police decreased, but by different amounts. Examining specific offenses alongside broad categories gives a more informative picture.

Data sources also differ by the crimes they cover. The NCVS covers selected nonfatal violent and household property crimes, while homicide data come from law-enforcement and death-record systems. Comparing results without considering differences in definitions and coverage can lead to misleading conclusions. Sound interpretation also requires attention to exposure, reporting, and other data limitations.