Researching Stratification, Crime and Inequality

Applying Methods to Crime and Stratification

Research methods in sociology are not studied in the abstract. For this paper, methods knowledge must be applied to the research contexts of crime and deviance and social stratification. This means being able to explain not just what a method is, but why a particular method is or is not appropriate for researching offenders, the wealthy, the poor, or subcultural groups — and what specific practical and ethical challenges arise in those contexts. A generic list of “advantages and disadvantages of questionnaires” will not earn high marks; credit goes to contextualised evaluation.

Key Fact

The key methods applied to crime and stratification include: official crime statistics (police-recorded crime), the Crime Survey for England and Wales (a victim survey), self-report studies, ethnographic observation, surveys (as used by Townsend to study poverty), and longitudinal studies (to track life chances over time).

Official Statistics and the Dark Figure

Official crime statistics are produced by the government from crimes recorded by the police. They have been collected in the UK for over 150 years and provide a consistent long-term data source that allows trends to be identified over time. They are a form of secondary data — sociologists analyse figures produced by the police and Home Office rather than collecting them directly.

However, official crime statistics significantly under-represent the actual level of crime through what sociologists call the dark figure of crime. Many crimes are never reported to the police (victims may fear reprisals, not trust the police, consider the crime too minor, or not recognise that they are victims). Others are reported but not recorded. Some crimes — such as domestic violence, sexual assault and corporate crime — are systematically under-represented because of victim reluctance, police discretion and the relative invisibility of white-collar offending. From an interactionist perspective (following Becker), official statistics reflect the outcome of labelling processes — who gets stopped, arrested and convicted — rather than the actual distribution of offending. Marxists argue that official statistics are ideological: they focus attention on street crime (disproportionately associated with the working class) and under-count corporate crime committed by the powerful.

Victim Surveys and Self-Report Studies

The Crime Survey for England and Wales (CSEW) is the major victim survey. It interviews a random sample of about 30,000 adults each year, plus a smaller sample of children aged 10 to 15, asking whether they have been the victim of any crime in the past twelve months, whether or not they reported it to the police. This captures a wider range of offending than official statistics and provides an indication of the dark figure. Sociologists assess its validity as higher than official statistics for crimes such as burglary and vehicle theft, but it still under-represents crimes with no individual victim (corporate crime, drug supply) and crimes where victims are unable or unwilling to disclose (domestic violence, sexual offences, crimes against very young children).

Self-report studies ask participants to anonymously disclose their own offending. They have consistently shown that offending is more evenly distributed across social classes than official statistics suggest — middle-class crime is systematically under-detected and under-prosecuted. This finding challenges the view (implicit in official statistics) that crime is predominantly a working-class problem, supporting the Marxist and interactionist critique of selective enforcement.

Ethnographic research involves the researcher immersing themselves in a social group over a sustained period. Paul Willis’s Learning to Labour (1977) is a key example applied to education and subcultural behaviour. Willis used participant observation and interviews with twelve working-class boys in a Midlands school. His extended engagement with “the lads” produced rich qualitative data about their counter-school culture, their masculine identity and their rejection of educational authority — data that surveys or official statistics could not have captured.

Pat Carlen’s study of women, crime and poverty (1988) used in-depth, unstructured interviews with 39 women who had been convicted of criminal offences. The small, purposive sample allowed Carlen to explore the class and gender dimensions of women’s criminal careers in depth — the class deal and the gender deal that had broken down for each participant. Unstructured interviews generate high-validity qualitative data about individual experiences and motivations, but the small sample size limits generalisability. Both Willis and Carlen illustrate how qualitative methods are particularly valuable for research into marginalised or hard-to-reach groups.

Researching Poverty and Stratification

Townsend’s Poverty in the United Kingdom (1979) illustrates how survey methods can be applied to stratification research. Townsend used questionnaires distributed to over 2,000 households and 6,000 individuals to measure poverty through his deprivation index. The large sample size and standardised questions produced quantitative data with high reliability — the same deprivation index could be applied consistently across all respondents. However, critics questioned whether the absence of certain foods or activities in a household necessarily indicated deprivation or whether it might reflect personal choice or cultural difference, raising questions about validity.

Longitudinal studies track the same individuals over time and are particularly valuable for studying social mobility and life chances. Fiona Devine’s return to Luton is a restudy rather than a true longitudinal design. She interviewed a fresh sample of workers in the same town and compared her findings with the earlier study by Goldthorpe and Lockwood. A restudy shows change over time without following the same people. However, longitudinal studies are expensive and face the problem of sample attrition — respondents dropping out over time.

Access: Researching active offenders, gang members or criminal subcultures involves significant access problems. Potential participants may distrust researchers, fear criminal exposure or be geographically dispersed. Snowball sampling — using initial contacts to recruit further participants — is commonly used in this context but produces unrepresentative samples. Researching the very wealthy presents the opposite problem: elite individuals can use social closure and lawyers to restrict access to researchers.

Researcher safety: Fieldwork among violent or criminal groups raises safety concerns. Researchers must weigh the value of participant observation against risks to themselves and their participants. Going covert (not disclosing researcher identity) can improve access but raises serious ethical concerns.

Ethical issues — informed consent: Gaining informed consent from offenders or vulnerable populations requires care. Participants must understand what they are consenting to, but in covert observation this is impossible by definition. Researchers must decide whether the knowledge gained justifies methods that bypass consent — a decision that must be weighed against professional ethical guidelines.

Harm to participants: Research that involves offenders, victims of domestic violence or members of disadvantaged communities may expose participants to harm if they are identified — for example, a gang member whose participation in research becomes known to rivals. Confidentiality and anonymisation of data are essential, but cannot always guarantee complete protection. Researchers also have conflicting obligations if they witness or are told about serious criminal activity.

Researching disadvantaged groups: People experiencing poverty or social exclusion may be harder to reach through standard sampling methods (which often rely on postal addresses or electoral registers that exclude the most marginalised). Townsend’s use of questionnaires distributed across various UK areas addressed some of these problems, but even large-scale surveys can under-sample the homeless or those in temporary accommodation.