Science for Everybody

Contents

• Introduction
• Sutherland and the Origins of Differential Association Theory
• Crime as Learned Behaviour
• Pro-Criminal and Anti-Criminal Attitudes
• Frequency, Duration, Priority and Intensity
• Learning Criminal Techniques
• Why the Theory Can Explain White-Collar Crime
• Can Offending Be Predicted Mathematically?
• Farrington et al. (2006)
• What the Theory Predicts in Real Life
• Evaluation
• Strengths
• Weaknesses
• Summary
• Quick Questions
• Key Terms

---

Introduction

Differential association theory explains offending as behaviour that people learn through interaction with others. Edwin Sutherland argued that people can learn both pro-criminal attitudes and the practical techniques of offending, especially from groups that matter to them.

AQA includes differential association theory within the psychological explanations of offending behaviour in Forensic Psychology, alongside cognitive explanations such as moral reasoning and cognitive distortions.

Sutherland and the Origins of Differential Association Theory

Edwin H. Sutherland first proposed differential association theory in 1939. He wanted a general theory that could explain offending across different social classes, cultures and types of crime. This was important because many earlier explanations focused heavily on biological abnormality, poverty or a supposed “criminal type”.

Sutherland argued that offending is not inherited as a ready-made behaviour. Instead, people learn it through the same broad social processes that they use to learn non-criminal behaviour. A person can learn how to obey rules from family and friends, but they can also learn attitudes and skills that support offending.

The word differential is important. People experience different patterns of social influence. One person may spend most of their time with people who strongly disapprove of offending. Another may spend time with people who normalise theft, violence, fraud or drug dealing. These different associations can produce different behavioural outcomes.

Crime as Learned Behaviour

The central claim is that criminal behaviour is learned. Learning takes place through interaction and communication with other people, particularly within intimate personal groups.

Two main things are learned:

• Attitudes, values and motives that support offending.
• Techniques that make offending possible.

For example, a young person may repeatedly hear friends describe shoplifting as harmless because “big companies can afford it”. At the same time, those friends may show the person where cameras are placed, how security tags are removed or how stolen goods are sold. The individual is therefore learning both a way of thinking about the offence and a way of carrying it out.

This approach does not require formal teaching. Learning can happen through observation, imitation, conversation, encouragement and repeated exposure to group norms.

Pro-Criminal and Anti-Criminal Attitudes

Sutherland argued that people are exposed to different definitions of the law. Some definitions favour obeying the law, while others favour breaking it.

Offending becomes more likely when a person experiences an excess of definitions favourable to law-breaking over definitions unfavourable to law-breaking. In simpler terms, the person receives more messages that support offending than messages that reject it.

These messages might include ideas such as:

• “Everyone does it.”
• “The victim deserved it.”
• “It is not really stealing if the company is rich.”
• “You have to fight or people will think you are weak.”
• “Breaking this rule is the only way to get ahead.”

The theory therefore links offending to the norms and values of the individual’s social environment. However, it does not claim that one criminal friend automatically causes crime. The overall balance, strength and history of associations matters.

Reflection question: If two people are exposed to the same pro-criminal peer group, why might one begin offending while the other does not? This highlights one of the theory’s central evaluation issues: individual differences may affect how strongly social influence changes behaviour.

Frequency, Duration, Priority and Intensity

Sutherland suggested that associations vary in their influence. Four factors are especially useful for explaining this:

Frequency refers to how often the individual comes into contact with a particular set of attitudes. Daily exposure may be more influential than occasional contact.

Duration refers to how long the association continues. A friendship that lasts for years may have more influence than a brief acquaintance.

Priority refers to when the association begins. Experiences earlier in development may have a stronger effect because important attitudes and habits are still being formed.

Intensity refers to how important or respected the person or group is. A close sibling, parent, best friend or admired group leader may exert more influence than someone the individual barely knows.

These ideas help explain why simply counting how many offenders someone knows is not enough. A single highly valued relationship could have more influence than several weak contacts.

Learning Criminal Techniques

Differential association theory also explains how people may learn the practical skills needed to offend. These skills can be simple or complex.

A person may learn how to conceal stolen items, bypass a basic lock or avoid detection. In more complex crimes, an individual may learn how to manipulate financial systems, commit digital fraud, launder money or exploit professional access.

This part of the theory helps explain why some forms of offending are concentrated within particular groups or occupations. People often need access to specialised knowledge before they can carry out certain crimes effectively.

Why the Theory Can Explain White-Collar Crime

One strength of Sutherland’s approach is that it does not limit offending to deprived communities. He was particularly interested in white-collar crime, which includes offences committed by people in professional, business or organisational settings.

A biological explanation based on a visible “criminal type” struggles to explain why apparently successful professionals may commit fraud, corruption or insider offences. Differential association theory can explain this because employees may learn workplace norms that excuse or reward illegal behaviour.

For example, a new employee may enter a team where manipulating figures is treated as normal and where experienced colleagues teach methods for hiding it.

This gives the theory broad explanatory scope because the same learning principles can be applied to street crime, organised crime and occupational crime.

Can Offending Be Predicted Mathematically?

Sutherland hoped that the theory could eventually produce scientific predictions. In principle, if researchers could measure the number and strength of pro-criminal and anti-criminal associations, they might estimate the likelihood of offending.

However, this creates a major problem. Concepts such as the intensity of a relationship or the exact balance of pro-criminal definitions are difficult to measure objectively. Researchers cannot easily assign a precise numerical value to how important a parent, friend or peer group is to someone.

As a result, the theory sounds scientific because it proposes measurable social processes, but in practice the key variables can be difficult to operationalise and test.

Farrington et al. (2006): Cambridge Study in Delinquent Development

Aim

To examine criminal careers across the lifespan and identify childhood factors linked with later offending.

Procedure or Method

The researchers used a prospective longitudinal study of 411 males from South London. The participants were first studied at around age eight in 1961 and were followed for decades. Researchers used interviews, information from families and schools, and official conviction records to examine offending and life outcomes.

Findings

By age 50, 41% of the men had received at least one conviction. Those convicted had an average of about five convictions across an average conviction career of around nine years. A small group, about 7%, accounted for more than half of all convictions. Important childhood risk factors included family criminality, daring or risk-taking, low school attainment, poverty and poor parenting.

Conclusion

Later offending was associated with childhood family and social factors. The findings are consistent with differential association theory because they show that social environments and close relationships are linked with later offending. However, the study does not directly prove that criminal attitudes or techniques were learned through association, so the evidence remains correlational.

What the Theory Predicts in Real Life

Differential association theory predicts that offending should become more likely when people spend frequent and meaningful time with others who support offending. It also predicts that changes in social networks may change behaviour.

For example, if a person leaves a gang and develops stronger relationships with law-abiding peers, employers, coaches or mentors, they may receive fewer pro-criminal definitions and more anti-criminal definitions. This suggests a practical route for prevention and rehabilitation: change the social environment as well as the individual.

However, social influence does not operate in isolation. Personality, biological vulnerability, cognitive processes, opportunity and wider social conditions may also contribute. Differential association theory is therefore best understood as a strong social explanation rather than a complete account of every offender.

---

Evaluation

Strengths

Broad Explanatory Power

Differential association theory has strong explanatory power because it can account for many different forms of offending. Sutherland did not restrict his explanation to violent or street crime. The same learning processes can explain theft, gang offending, organised crime and white-collar crime. For example, a worker may learn fraudulent practices from experienced colleagues just as a young person may learn shoplifting techniques from peers. This breadth is useful because it avoids the assumption that offenders form one biologically distinct group. Therefore, the theory offers a more flexible explanation than approaches that rely on a single criminal personality or physical type.

Focus on Social and Environmental Influences

The theory shifted attention towards social and environmental influences on crime. Earlier explanations often focused on biological abnormality or individual defect. In contrast, differential association theory asks what the person has learned and from whom. This has practical value because social environments can potentially be changed. Mentoring, pro-social peer groups, education and rehabilitation programmes may reduce exposure to pro-criminal attitudes and provide alternative norms. As a result, the theory has useful applications for prevention and rehabilitation rather than simply labelling individuals as naturally criminal.

Longitudinal Research Support

Longitudinal evidence offers some support for the importance of close social and family factors. Farrington et al. (2006) followed 411 South London males and found that family criminality and poor parenting were among the childhood factors associated with later convictions. This supports the general claim that a person’s social environment matters. The study also followed participants for decades, which increases confidence that early risk factors came before many later offences. However, the evidence is supportive rather than a direct test of Sutherland’s learning mechanisms, so the theory should not claim more support than the data provide.

Explains the Learning of Specific Skills

The approach can explain how offenders acquire specific skills as well as motives. Some offences require knowledge that is unlikely to appear spontaneously. Fraud, cybercrime, organised theft and other specialised offences often involve techniques that can be passed between people. Differential association theory explicitly includes this learning of techniques, which gives it explanatory detail. Therefore, the theory can explain not only why offending may become acceptable to an individual, but also how the individual becomes capable of carrying it out.

Weaknesses

Difficult to Test Scientifically

A major weakness is that the central concepts are difficult to test scientifically. Sutherland suggested that associations differ in frequency, duration, priority and intensity, but these are hard to measure objectively. Researchers could count contact hours, yet this would not necessarily measure emotional importance or the strength of a learned belief. This makes precise prediction difficult and weakens falsifiability. Therefore, although the theory aims to be scientific, some of its key processes are difficult to operationalise reliably.

Correlation and Causation

Much of the supporting evidence is correlational, so cause and effect remain unclear. People with offending behaviour may spend time with criminal peers because those peers caused their offending, but the direction could run the other way. Someone already willing to offend may actively seek friends with similar values. A third variable, such as family disruption, poverty or impulsivity, could also increase both criminal friendships and offending. Therefore, an association between criminal peers and offending does not prove that differential association caused the behaviour.

Individual Differences

The theory can be criticised for underestimating individual differences. Not everyone exposed to pro-criminal attitudes becomes an offender. Two siblings may grow up in the same family and have similar peer groups but behave very differently. Temperament, self-control, moral reasoning, fear of punishment and biological factors may change how a person responds to social influence. This means differential association theory can appear socially deterministic because it places heavy emphasis on learned environments while giving less attention to personal choice and dispositional differences.

Alternative Explanations

Alternative explanations may account for apparent family transmission. If offending runs in families, differential association theory interprets this as learning from close relatives. However, biological explanations could argue that relatives also share genes linked with traits such as impulsivity or aggression. Family members also share environments, so separating genetic and social effects is difficult. Therefore, evidence that criminal parents are associated with criminal behaviour in their children does not uniquely support differential association theory.

Sample Bias and Population Validity

Some evidence has limited population validity and may contain sample bias. Farrington et al.’s Cambridge study focused on 411 males from a largely working-class area of South London. The long follow-up is a strength, but the all-male and geographically specific sample limits generalisation to women, different social classes and other cultures. Research on offending can also over-represent people who enter the criminal justice system while missing undetected offending. This sample bias may distort conclusions about who learns criminal behaviour and in what circumstances.

Ethical Limitations

There are ethical limits on how directly researchers can test the theory. It would be unethical to deliberately place participants into groups that teach criminal attitudes or techniques and then observe whether they offend. Researchers must therefore rely on naturalistic, longitudinal or correlational evidence. Studies involving self-reported offending also create confidentiality and disclosure issues. These ethical constraints protect participants, but they make strong experimental tests of the theory extremely difficult.

Temporal Validity

Temporal validity is mixed because the social world has changed since Sutherland developed the theory. The original theory emphasised interaction within intimate personal groups. Today, people can encounter pro-criminal attitudes, extremist communities, fraud techniques or criminal subcultures through online networks without meeting members face to face. This could reduce the theory’s temporal validity if “intimate groups” is interpreted narrowly. However, the broader learning principle can still apply if online communities are treated as meaningful sources of association and communication.

Quick Questions

Who proposed differential association theory, and in what year was it first published?

Edwin H. Sutherland first proposed differential association theory in 1939.

What two broad types of information does a person learn according to differential association theory?

A person learns pro-criminal attitudes, values, motives or justifications, and also practical techniques for committing offences.

What does Sutherland mean by an “excess of definitions favourable to law-breaking”?

It means the person is exposed to more or stronger messages supporting law-breaking than messages supporting law-abiding behaviour.

What four features affect the strength of a differential association?

Frequency, duration, priority and intensity.

Give one finding from Farrington et al. (2006) that is relevant to explanations of offending.

For example, 41% of the 411 males had at least one conviction by age 50; around 7% accounted for more than half of all convictions; or family criminality and poor parenting were among childhood risk factors linked with later offending.

Key Terms

Differential association theory – Sutherland’s explanation that offending behaviour is learned through interaction and communication with other people, especially close social groups.

Pro-criminal attitudes – Beliefs, values or justifications that support, excuse or encourage offending.

Anti-criminal attitudes – Beliefs and values that support obeying the law and discourage offending.

Definitions favourable to law-breaking – Messages or beliefs that make breaking the law appear acceptable, justified or rewarding.

Intimate personal groups – Close groups such as family members, friendship groups or close peers whose attitudes and behaviour have a strong influence.

Criminal techniques – Practical skills or methods used to carry out an offence, ranging from simple actions to highly specialised knowledge.

Frequency, duration, priority and intensity – Features that affect the strength of an association: how often it occurs, how long it lasts, how early it begins and how important the people involved are to the individual.

Socialisation – The process through which people learn norms, values and expected behaviour from others.

White-collar crime – Financial or occupational crime committed in professional or business settings, such as fraud or insider dealing.