Data on Crime and the Dark Figure
What Do We Actually Know About Crime?
Before sociologists can analyse crime, they must grapple with a fundamental methodological question: how is crime measured, and how reliable are those measurements? Different sources of crime data paint different pictures, and sociological perspectives disagree about what official statistics reveal, conceal and distort. This lesson examines the main sources of crime data, the patterns they reveal and their significant limitations.
The three main sources of crime data in England and Wales are: (1) police-recorded crime, (2) the Crime Survey for England and Wales (a large-scale victim survey), and (3) self-report studies. Each source captures different dimensions of crime and each has distinct strengths and limitations.
Police-Recorded Crime
Police-recorded crime refers to incidents that are reported to the police and recorded as crimes in their official statistics, published by the Home Office. This source has the longest historical run of data and allows analysis of trends over decades.
However, police-recorded crime has well-documented limitations. First, it only captures crimes that are reported to the police by victims or witnesses. Many crimes — particularly domestic violence, sexual offences, many thefts and hate crimes — are never reported. Second, even among reported crimes, the police have discretion about whether to record an incident as a crime: policing priorities, officer judgements and administrative recording practices all influence what enters the official figures. Third, changes in recording practices over time make comparisons difficult — a rise in recorded sexual offences, for example, may reflect increased willingness to report rather than increased incidence.
From a Marxist perspective, police-recorded statistics reflect the priorities and interests of those with power to define and enforce the law. Crimes committed by the powerful (corporate fraud, tax evasion, health and safety offences) receive less police attention and thus less recording, while street crime committed by the poorest groups is disproportionately pursued. The official statistics therefore tell us as much about policing choices as about crime.
The Crime Survey for England and Wales
The Crime Survey for England and Wales (CSEW) was established in 1982 and asks a large representative sample of households in England and Wales about crimes they have experienced in the previous year, regardless of whether those crimes were reported to the police. By surveying victims directly, it provides a measure of crime that is independent of police recording decisions.
The CSEW consistently records higher crime levels than police-recorded figures because it captures many incidents that go unreported. It is therefore regarded by sociologists as a more comprehensive measure of crime experienced by the population. It also provides consistent data over time unaffected by changes in police recording practices.
The CSEW has its own limitations, however. It relies on self-report by respondents, who may misremember events, under-report sensitive offences (domestic abuse, sexual violence) out of shame or fear, or fail to recognise that an experience constituted a crime. It also excludes crimes against commercial premises, murder (where there is no living victim to survey), and crimes experienced by those living in institutions such as prisons or care homes. Crimes without a direct personal victim — such as drug possession or fraud against organisations — are also outside its scope.
Police-recorded crime: based on reports to the police; affected by policing priorities and recording decisions; shows long-term historical trends; captures crimes against businesses and public property; sensitive to changes in police practice.
Crime Survey for England and Wales: based on victim self-report; independent of policing decisions; more consistent over time; captures unreported crime; excludes murder, victimless crimes, crimes against organisations and under-reporting of sensitive offences.
When the two measures diverge (as they often do), sociologists ask which factors explain the gap. A rise in police-recorded crime during a period of stable CSEW figures, for example, is more likely to reflect improved recording practices than a real rise in offending.
Self-Report Studies
Self-report studies ask individuals to disclose crimes they have committed, often anonymously, providing a third source of data that challenges official crime statistics. Self-report studies have consistently shown that offending is far more widespread across social classes and ethnic groups than official statistics suggest — middle-class and white individuals commit offences at rates much closer to working-class and Black individuals than police-recorded crime implies. This finding supports the interactionist argument that official statistics reflect selective policing and prosecution rather than actual offending behaviour.
Self-report studies have their own limitations: respondents may under-report serious crimes (rape, grievous bodily harm) while over-reporting minor offences; they rely on accurate recall; and they cannot capture offences the respondent does not recognise as crimes or is unwilling to disclose even anonymously.
The Dark Figure of Crime
The dark figure of crime refers to the substantial gap between the amount of crime that actually occurs and the amount that appears in official statistics. This gap exists for two main reasons: crimes are not reported to the police, and crimes that are reported are not always recorded.
Victims may not report crime because: they consider it too trivial; they fear they will not be believed; they are intimidated by the offender; they feel embarrassed or ashamed; they distrust the police; or they resolve the matter informally. Police may not record a reported crime if they judge it insufficiently serious, if the victim withdraws the complaint, or if recording practices are inconsistent.
The dark figure is not evenly distributed across all crime types. Domestic violence and sexual offences have particularly large dark figures because of shame, fear of retaliation and the intimate relationship between victim and perpetrator. Fraud and cybercrime are significantly under-recorded because victims often do not recognise they have been defrauded, report to a bank rather than the police, or assume the police will not act. Minor property crime may go unreported simply because victims feel it is not worth the effort of reporting.
A very common error in sociology exams is to cite official crime statistics as if they were objective facts about the true level of crime in society. The spec is explicit that official statistics have both strengths and weaknesses and that sociological perspectives differ on their usefulness. The dark figure concept is the spec’s core illustration of why official statistics under-represent crime. Always evaluate statistics by asking: what is the source? Who is likely to be missing from this data? What policing or reporting decisions might affect these figures?
Sociological Perspectives on Crime Data
Sociological perspectives disagree about the value and meaning of official crime statistics. Functionalists broadly accept official statistics as a useful measure of crime trends, arguing that they allow governments to monitor social order and direct policing resources. Functionalists treat crime data as a reasonable approximation to reality, while acknowledging recording limitations.
Marxists are far more sceptical. They argue that crime statistics are socially constructed in the interests of the ruling class. The law itself is made by and for the bourgeoisie; enforcement focuses on working-class crime; and the statistics therefore reflect class power rather than objective harm. Official data systematically conceals corporate and white-collar crime while inflating working-class street crime.
Feminist perspectives point out that crime statistics consistently under-represent crimes against women, particularly domestic violence and sexual assault. Patriarchal norms discourage women from reporting these crimes; police response has historically been inadequate; and victim-blaming attitudes in courts further reduce the likelihood of prosecution. Feminists argue that the dark figure of gendered violence is particularly large and that official statistics should always be read critically in this light.
Interactionists argue that crime statistics measure policing activity rather than criminal behaviour. Because the police have discretion about whom to stop, search, arrest and charge, the statistics reflect which groups have been labelled as criminal rather than which groups actually offend more. Self-report studies, which show offending rates that are more evenly spread across social groups, support this critical reading of official data.