Victimisation Patterns and Repeat Victimisation

One of the most consistent findings in modern criminology is that victimisation concentrates sharply. In most national surveys, roughly 1 to 3 percent of respondents account for between 30 and 50 percent of all incidents reported. This concentration exists for property crime, violent crime, and harassment. It holds across different countries, different survey instruments, and different reference periods.

The Crime Survey for England and Wales, conducted annually since 1982, has documented this pattern over four decades. In any given survey year, most respondents report no victimisation. A significant minority report a single incident. A small group reports multiple incidents, and this group drives the headline rate. The same pattern appears in the US National Crime Victimization Survey, in Statistics Canada's General Social Survey on victimisation, and in India's National Crime Records Bureau data, though the Indian administrative data captures only reported offences.

Place-level concentration mirrors individual-level concentration. Research in Minneapolis in the late 1980s by Lawrence Sherman, Patrick Gartin, and Michael Buerger found that 3.3 percent of addresses generated 50 percent of all calls for service to police. Subsequent replications in Kansas City, Boston, Seattle, and cities in the Netherlands, Australia, and Sweden have consistently found that between 3 and 10 percent of places generate 50 percent or more of crime incidents. This finding gave rise to the concept of crime hot spots and hot-spot policing.

Victimisation risk is not uniform across social groups, and the pattern varies by crime type. For violent crime, young males aged 16 to 24 consistently show the highest risk in British, American, and Australian surveys. The British Crime Survey has shown this demographic peak for violent victimisation since its first wave in 1982. In the US, Black and Hispanic Americans face higher homicide victimisation rates than white Americans, a disparity linked to structural factors including neighbourhood poverty and segregation rather than individual characteristics.

For domestic violence and sexual assault, the gender pattern is reversed: women face substantially higher risk than men across all countries surveyed, though male victimisation is significantly under-reported. The Council of Europe's GREVIO monitoring body and the United Nations Office on Drugs and Crime (UNODC) publish comparative data across jurisdictions confirming this pattern. In India, the National Family Health Survey has documented intimate partner violence rates, with rural women and those with less education facing higher risk.

Socioeconomic factors shape property crime risk strongly. Renters, people in lower-income households, and residents of densely populated urban areas face higher burglary and vehicle theft risk. The explanation, consistent with routine activity theory, is that wealthier households tend to have better security, more capable guardianship (neighbours at home, alarm systems), and live in areas with lower motivated offender density. Older adults, despite public perception, face lower risk for most crime types than younger adults, but they face disproportionate risk from telephone and online fraud.

Two competing explanations for repeat victimisation have dominated the literature since Graham Farrell and Ken Pease introduced the flag versus boost distinction in the 1990s. Flag theory holds that a target has stable characteristics that make it independently attractive to different offenders. A poorly secured terraced house in a high-crime street will attract burglars repeatedly because the features that made it an easy target, weak lock, side access, low natural surveillance, remain unchanged after the first burglary. Multiple different offenders, acting independently, select it for the same reasons.

Boost theory holds that the original offender returns. Having already assessed the target, identified the security weaknesses, learned the occupant's routines, and successfully completed one offence, the same offender revisits. Each successful offence updates the offender's information about the target and raises the expected return on a second visit. Applied to domestic violence, boost theory maps onto the established pattern that the same perpetrator re-offends against the same victim, often escalating in severity.

The time-course of repeat victimisation provides some evidence for boost over flag. Studies of residential burglary in the UK, the Netherlands, and Australia consistently find that risk is highest in the first few days to weeks after the initial offence, then declines toward baseline over roughly two to three months. This decay curve is consistent with boost: an offender returns quickly while knowledge is fresh and then either exhausts the target or moves on. Flag alone would predict a more uniform elevated risk across time. In practice, both mechanisms operate, and the two theories predict the same outcome from different directions.

Near repeat victimisation extends the concept spatially. Properties close to a victimised address face elevated risk immediately after a burglary, because an offender who has identified a productive street may work nearby targets before leaving the area. Research by Shane Johnson and Kate Bowers in London and Liverpool, and replicated in studies across the US and the Netherlands, shows that within roughly 200 metres and two weeks of a burglary, risk for nearby properties is measurably elevated. This pattern supports tactical deployment of crime prevention resources to the immediate neighbourhood after a first offence.

Survey measurement of repeat victimisation is complicated by the problem of series incidents. A victim of domestic abuse who experiences weekly violence over six months cannot meaningfully recall each incident separately during a survey interview. Early versions of the National Crime Victimization Survey in the US asked respondents to count incidents but applied no cap, leading to the suspicion that a small number of high-count respondents were inflating national estimates. The NCVS was revised from 2006 to address this.

The Crime Survey for England and Wales historically applied a cap of five incidents per victim for any series of similar incidents. Critics, including Farrell and Pease, argued this cap substantially understated the true volume of repeat victimisation, particularly domestic violence. The Office for National Statistics reviewed the cap rule and modified the approach; the survey now records series incidents separately and applies more nuanced treatment. Regardless of the specific instrument, researchers working with victimisation survey data must understand how the instrument handles series incidents before making cross-survey comparisons.

Police-recorded data also understates repeat victimisation. Many victims of repeated domestic violence do not call the police for every incident, and when they do, recording practices differ between forces. Some police databases link incidents to the same victim address or individual, allowing repeat victimisation rates to be calculated from administrative data. Others do not. The dark figure of crime, the gap between victimisation survey estimates and police records, is disproportionately large for repeat victimised groups, especially domestic violence victims.

The operational implication of repeat victimisation research is precise: resources should be targeted at prior victims in the short window of elevated post-incident risk. This is the basis of targeted repeat victimisation prevention programmes, which began in the UK in the 1990s and have since been adopted in various forms in the US, Australia, and the Netherlands. The Kirkholt Burglary Prevention Project in Rochdale, England, is the most cited early example. Following a survey showing that a significant proportion of burglaries were repeats at previously victimised addresses, the project delivered security upgrades, removed electricity and gas coin meters (which were being stolen), and established a neighbourhood cocoon watch. Burglary fell by 75 percent over three years.

Hot-spot policing uses the place-level equivalent of the same logic. Directing additional patrol to crime-concentrated street segments has a well-evidenced suppression effect. A 2014 systematic review by Weisburd, Telep, Hinkle, and Eck found consistent crime reductions from hot-spot patrol across multiple evaluations, with limited evidence of displacement to adjacent areas. The US Department of Justice's Office of Community Oriented Policing Services has funded hot-spot policing programmes in multiple cities. Similar programmes have operated in the UK under the National Policing Improvement Agency guidance and in Australia under state police force directives.

For domestic violence, the DASH (Domestic Abuse, Stalking and Honour-based Violence) risk assessment tool used across England and Wales identifies victims at high risk of repeat assault or serious harm. High-risk cases are referred to Multi-Agency Risk Assessment Conferences (MARACs), where police, health, housing, and social services coordinate protection. India's Protection of Women from Domestic Violence Act 2005 and its amendments provide a comparable framework for identifying and supporting repeat domestic violence victims, as does the Violence Against Women Act (VAWA) in the United States and the Domestic Abuse Act 2021 in England and Wales.

Situational crime prevention, drawing on routine activity theory, addresses the environmental factors that flag a target as attractive. Security improvements, lighting upgrades, target hardening, and altering routines to reduce exposure can cut repeat victimisation for property crime. For personal crime, reducing routine exposure to settings with motivated offenders and increasing capable guardianship are the core mechanisms. The Crime Reduction Programme in England and Wales from 1999 to 2002 funded many repeat victimisation reduction initiatives, and its evaluations produced a body of evidence on what works.

The International Crime Victims Survey (ICVS), conducted in waves from 1989 to 2005 across more than 70 countries, provided the first systematic comparative data on victimisation rates and patterns. Its findings showed that while crime type distributions varied, the concentration of victimisation on a small minority was a consistent cross-national pattern. Countries with higher inequality, measured by the Gini coefficient, tended to show higher victimisation rates, particularly for property crime, consistent with routine activity and strain theory predictions.

In India, victimisation research has been constrained by the dominance of police-recorded data from the National Crime Records Bureau (NCRB). The NCRB Crime in India reports document registered offences but do not capture the dark figure. Victimisation surveys in India, including regional studies in urban areas and the National Family Health Survey (NFHS) for domestic violence, indicate substantial under-reporting, particularly for violence against women, sexual offences, and crimes in tribal and rural communities. The Bharatiya Nagarik Suraksha Sanhita 2023 (BNSS) introduced procedural reforms to encourage reporting, but changing recording practices is a long-term process.

In the European Union, the EU Agency for Fundamental Rights (FRA) conducts large-scale surveys on violence against women and on minority group victimisation. The FRA's survey on violence against women, published in 2014 with a follow-up in 2021, found that one in three women in the EU had experienced physical or sexual violence since the age of 15, with substantial variation across member states. These cross-national datasets have become the evidence base for EU crime prevention policy and for Directive 2012/29/EU, which established minimum standards for crime victims across the bloc. The comparable framework in the US is the Victims of Crime Act (VOCA) of 1984 and its subsequent amendments.

Forensic psychology contributes to understanding repeat victimisation through research on victim-offender relationships, trauma responses that affect victim behaviour and reporting, and the psychology of perpetrators who select the same targets repeatedly. For a deeper treatment of these psychological dimensions, see the forensic psychology subject index.