Differential Association Theory

Defining Crime

The most widely used definition of crime is a legal one: an act that violates the criminal law of a society and is punishable by the state. This definition is straightforward in principle but raises several problems in practice. Crime is not a fixed, universal category — it varies across jurisdiction (an act legal in one country may be criminal in another), time (homosexuality was criminalised in England and Wales until 1967), and culture. Not all harmful acts are crimes (e.g. emotional abuse may cause severe harm but is not always criminal), and not all crimes are harmful in any obvious sense (e.g. some regulatory offences). Crime also differs from deviance: deviance refers to behaviour that violates social norms but is not necessarily illegal (e.g. extreme rudeness).

Measuring Crime

Three main methods are used to measure the extent of crime in society. Each captures a different portion of criminal activity and has its own strengths and limitations.

1. Official Statistics (Police-Recorded Crime)

Official statistics are compiled from crimes reported to and recorded by the police. They are published annually by the Home Office. A major limitation is the dark figure of crime — the substantial gap between crimes committed and crimes recorded. Offences go unreported for many reasons: embarrassment (sexual offences, domestic violence), distrust of the police, the victim considers the offence too trivial to report, or lack of insurance removes any practical incentive. Even reported crimes may not be recorded if officers exercise discretion. Changes in recording practices and legal definitions also make historical comparisons unreliable.

2. Victim Surveys (Crime Survey for England and Wales)

The Crime Survey for England and Wales (CSEW) — formerly the British Crime Survey — asks a random sample of approximately 35,000 households about their experience of victimisation over the previous year, regardless of whether they reported incidents to the police. This captures some of the dark figure. However, it still misses crimes without individual victims (drug possession, fraud against organisations), crimes against those not living in private households (homeless, institutionalised), crimes the respondent does not identify as crimes, and incidents that have been forgotten.

Farrington and Dowds (1985) demonstrated that regional variations in official crime statistics could be partly attributed to differences in reporting behaviour rather than genuine differences in crime rates — highlighting a systematic bias in police-recorded data.

3. Self-Report Studies

Self-report studies ask individuals to anonymously disclose crimes they have committed, whether or not they were caught. This captures offending that would never appear in official records. Graham and Bowling (1995) surveyed 1,721 young people aged 14 to 25 and found that 55% of males and 31% of females admitted to committing at least one criminal offence — rates far higher than official statistics suggest. Limitations include social desirability and deliberate exaggeration or minimisation, reliance on accurate memory, biased sampling (volunteers and those accessible through community organisations), and serious or violent offences are typically underrepresented.

Evaluation

No single method provides a complete picture of crime. Official statistics are biased towards visible street crime committed by those without resources to avoid detection. Victim surveys capture more offences but rely on recall and exclude corporate and victimless crime. Self-report studies reveal hidden offending but suffer from demand characteristics. Most criminologists recommend triangulation — using all three methods together — to obtain the most reliable estimate of crime prevalence and trends.