Differential association theory, developed by criminologist Edwin Sutherland in 1939, is the theory that postulates deviance is learned. Its central claim is straightforward: criminal and deviant behavior is not inherited, not invented by the individual, and not the product of mental illness. It is learned through interaction with other people, in exactly the same way that law-abiding behavior is learned.
How Differential Association Theory Works
Sutherland laid out his theory in nine formal principles. The first and most important is simply that criminal behavior is learned. From there, the principles explain how that learning happens: through communication with other people, primarily within close personal groups like family, friends, and peers. What gets learned isn’t just the physical techniques of committing a crime. It’s the motives, attitudes, and rationalizations that make breaking the law feel acceptable or even desirable.
The key mechanism is what Sutherland called “definitions.” In this context, a definition is an attitude toward the law: either favorable (it’s wrong to steal) or unfavorable (stealing from a big corporation doesn’t really hurt anyone). A person becomes deviant when their exposure to unfavorable definitions outweighs their exposure to favorable ones. Everyone encounters both types throughout life. The tipping point is which type dominates.
Not all exposure counts equally. Sutherland identified four variables that determine how much weight any particular association carries:
- Frequency: Associations that happen often have more influence than rare encounters.
- Duration: Long-lasting relationships shape attitudes more than brief ones.
- Priority: Associations formed early in life, particularly in childhood, carry more weight than those formed later.
- Intensity: Relationships with people you respect or admire matter more than those with people you don’t.
This framework explains why two people in the same neighborhood can turn out very differently. One might have a close, long-standing relationship with a mentor who models law-abiding values. The other might spend years in a tight-knit peer group where rule-breaking is normalized and even rewarded. The theory predicts that the balance of these associations, filtered through frequency, duration, priority, and intensity, determines the outcome.
What Makes This Theory Different
Before Sutherland, most explanations for deviance focused on individual traits: poverty, psychological problems, low intelligence, or biological predisposition. Differential association theory broke from all of that by treating deviance as a normal social process. The learning happens through ordinary communication, not through some special defect in the person.
This also sets it apart from control theories, which take the opposite starting assumption. Control theories ask why most people don’t commit crimes, and answer that social bonds, institutions, and self-control hold people in check. When those bonds weaken, deviance emerges. Sutherland’s theory flips this: it assumes people are essentially blank slates who must be taught to be deviant (or taught to be conformist) through their social environment.
Strain theories offer yet another contrast. They argue that deviance arises when people experience a gap between their goals and legitimate means of achieving them, pushing them toward rule-breaking out of frustration or desperation. Differential association doesn’t require any strain or frustration. A person can learn deviant behavior while perfectly content, simply because the people around them model and reinforce it.
How Social Learning Theory Expanded the Idea
In 1966, Robert Burgess and Ronald Akers took Sutherland’s framework and merged it with behavioral psychology. Their social learning theory kept the core idea that deviance is learned through association but added specific mechanisms from operant conditioning: reinforcement, punishment, and imitation. In Sutherland’s original version, the learning process was somewhat vague. Akers made it concrete by arguing that people repeat behaviors that are rewarded and avoid behaviors that are punished, and that they also learn by observing what happens to others.
Social learning theory defines criminal behavior as something learned through association with people involved in crime, exposure to models of delinquent behavior, and the expectation that rewards will outweigh punishments. This addition helped explain not just where deviant attitudes come from, but why some people act on them and others don’t. If a teenager sees a friend shoplift and get away with it (positive reinforcement through acquiring goods, no punishment), social learning theory predicts the teenager is more likely to try it. If the friend gets caught and faces serious consequences, the opposite effect kicks in.
Critics have pointed out that the central mechanism of social learning theory, differential reinforcement, is difficult to observe and measure in real-world settings. Still, the broader idea that peer associations shape deviant behavior has held up well across decades of research.
What Research Shows About Peer Influence
The link between peer associations and deviance is one of the most consistently supported findings in criminology. Studies repeatedly show that having close friends who engage in delinquent behavior is among the strongest predictors of a person’s own involvement in deviance, across categories including substance use, theft, and violence.
Research has also revealed a more complicated picture than Sutherland originally described. A study published in PMC examined how deviant behavior within friendships affects the quality of those relationships. The findings showed that when both a person and their friend engaged in theft or violence, the person’s perception of their friendship quality dropped significantly. In other words, the very associations that differential association theory says transmit deviant behavior also tend to erode the social bonds involved. Friends who steal together or fight together report weaker, less satisfying friendships. This suggests that while deviance may be learned through close relationships, those relationships often suffer as a result.
This creates something of a paradox for the theory. Differential association predicts that close, high-quality relationships are the most powerful vehicles for learning deviance (because they score high on intensity and duration). But deviance within those relationships appears to undermine their quality over time, potentially weakening the very mechanism that sustains the behavior.
Why This Theory Still Matters
Differential association theory reshaped how criminologists, educators, and policymakers think about deviance. If rule-breaking is learned through social interaction rather than caused by individual defects, then prevention strategies look very different. Programs that focus on changing peer environments, mentorship, and redirecting social groups all draw on Sutherland’s core insight. So do interventions that try to shift the balance of “definitions” a young person encounters by increasing exposure to pro-social role models and reducing exposure to groups where deviance is normalized.
The theory also helps explain patterns that individual-level explanations struggle with, like why crime clusters in certain neighborhoods, why it runs in families without any genetic mechanism, and why people with no apparent psychological problems or economic disadvantage still end up involved in criminal behavior. The answer, according to Sutherland, is always the same: they learned it from the people closest to them.