Who Taught You That? SOC 3333 · Crime Lab · Midwestern State University

Last chapter argued that crime needs no teaching, only a failure of restraint. This chapter argues the opposite: that crime is learned the way everything else is learned, from the people closest to you, and that the pushing is the whole story. Both camps agree on one stubborn fact. People who offend have friends who offend, and that association is one of the strongest correlates in the entire field. They disagree completely about what it means. This page works through that disagreement, then shows you what happened when someone took learning theory seriously enough to build programs on it. The last two tabs are a review aid for the exam.

One fact, two theories, no agreement

Take an ordinary finding: adolescents who report delinquent friends report more delinquency themselves. Nobody disputes it. Here is what the two camps say it means.

Control theory says
The bond broke first. Weak attachment to parents and school left these kids free to drift toward each other. The friendship is a symptom, not a cause, and the causal work was already done before they met.
Learning theory says
The friendship is the cause. Techniques, opportunities, and above all definitions favorable to law-breaking are transmitted in intimate groups. The friends taught them, exactly as friends teach everything else.

Criminologists call this the selection versus socialization problem, and it is the single most important methodological argument in this chapter. Before we get to the answer, commit to a number.

A meta-analysis pooled 133 studies of social learning theory covering 118,403 people. It reported the mean correlation between delinquent peer association and offending. Where do you think that number landed?
.30

The four components of Akers's theory, tested separately

Differential association (who you are with)
.225
Definitions (attitudes toward the law)
.218
Modeling and imitation
.103
Differential reinforcement
.097

Pratt, T. C., Cullen, F. T., Sellers, C. S., Winfree, L. T. Jr., Madensen, T. D., Daigle, L. E., Fearn, N. E., and Gau, J. M. (2010). The empirical status of social learning theory: A meta-analysis. Justice Quarterly 27(6), 765 to 802. One hundred thirty-three studies from 1974 to 2003, 704 effect size estimates, 118,403 cases across 85 independent datasets. Retrieved August 2026.

Read this the way you read the Chapter 8 scorecard. The two components Sutherland gave the theory in 1947, association and definitions, are the ones carrying it. The two components Akers added in the 1960s, reinforcement and imitation, are less than half as strong. The authors said outright that the imitation portion may not be well supported by the existing literature. So the theory works, and it works mostly because of its oldest parts.

How this compares. Chapter 8 gave you self-control at mean correlations of about .26 attitudinal and .28 behavioral. Differential association sits at .225. A 2019 review of the peer literature put it this way: the effects of delinquent peer associations are at least comparable to, if not larger than, the effects of other well-known correlates of crime such as self-control. Two theories that call each other wrong have produced predictors of roughly the same size, measured by different people over fifty years. That is worth sitting with.

So which is it, selection or socialization

Pick the position you find most defensible before you look at the answer. There is no trick here and the honest answer is not the boring one.

What happens when you take learning theory seriously

Here is a genuine test. If crime is learned in intimate groups through definitions favorable to law-breaking, then deliberately gathering delinquent young people into a group should teach them more crime, not less. That is an uncomfortable prediction, it is unambiguous, and it has been tested many times, because a great deal of juvenile programming does exactly that.

Researchers named the mechanism deviancy training: within a group of high-risk peers, talk about rule-breaking gets laughter, attention, and status, which is textbook differential reinforcement. Take the most famous example, the programs in which young people are brought into a prison to be confronted by incarcerated adults. Guess what the randomized trials found before you look.

1.68
odds ratio for reoffending among participants compared with controls, 95 percent confidence interval 1.20 to 2.36
9
randomized controlled trials, 7 of them pooled in the meta-analysis
946
young people across all nine trials

Petrosino, A., Turpin-Petrosino, C., Hollis-Peel, M. E., and Lavenberg, J. G., Scared Straight and other juvenile awareness programs for preventing juvenile delinquency, Campbell Systematic Reviews. The National Institute of Justice rates this practice Ineffective at CrimeSolutions and states that participation increases the odds that youth will commit offenses in the future. Retrieved August 2026.

An odds ratio above 1 means the program group did worse. Every analysis the reviewers ran pointed the same direction. Their conclusion was that jurisdictions already running these programs should reevaluate them, because they are at risk of causing harm to the very citizens they aim to help. This is one of the clearest cases in all of criminal justice of an intervention that is popular, intuitive, cheap, and actively counterproductive.

Be careful about what this does and does not prove. These trials show that the programs made things worse. They do not by themselves prove that deviancy training was the mechanism, because a confrontational prison visit could backfire for several reasons at once. What makes the learning explanation credible is that the same harmful pattern shows up across a wider family of interventions that share one feature: they put high-risk young people in a room together. That is the pattern a 1999 review in American Psychologist called attention to under the heading of when interventions harm.

Now put this beside last week. Chapter 8 showed you a theory whose weakest strand, involvement, predicted correctly that keeping young people busy would not reduce crime. Chapter 9 shows you a theory that predicted correctly that grouping them would increase it. Two rival theories, two successful predictions, both about programs that sound obviously good. Theory is not decoration. It is the thing that would have told you, in advance and for free, that these were bad bets.

Dishion, T. J., McCord, J., and Poulin, F. (1999). When interventions harm: Peer groups and problem behavior. American Psychologist 54(9), 755 to 764. Retrieved August 2026.

The five techniques of neutralization

Gresham Sykes and David Matza argued in 1957 that most people who break rules do not reject the rules. They excuse themselves from them, in advance, using a small and surprisingly standard set of moves. Matza called the resulting state drift: not committed to crime, not committed to conformity, floating between them and released episodically by whichever excuse is available.

This is also the exam-relevant bridge back to Sutherland, because these five are what definitions favorable to law-breaking look like when you catch them in the act. Sort each statement into the technique it uses.

Sykes, G. M., and Matza, D. (1957). Techniques of neutralization: A theory of delinquency. American Sociological Review 22(6), 664 to 670.

Exam 2 self-check: 0 of 0 answered correctly. Fourteen questions across Chapters 6 through 9.

These are written in the style of the exam. Each one names the chapter it comes from, so if you miss several from the same chapter you know exactly where to spend tonight. Nothing here is recorded and you can be wrong as many times as you like.

Chapters 6 through 9 in one page

The exam asks you to connect a theorist to a theory to a concept. Here is every triple you need, in order. Read down the list and say each one out loud. Anything you cannot finish from memory is where your studying should go.

One structural thing worth noticing before you close this. Chapters 6 and 7 are about what criminology thinks crime is. Chapters 8 and 9 are two answers to the same question about what makes an individual offend, and they are direct opposites: restraint failed, or motivation was taught. If a question gives you a scenario and asks which theory explains it, the fastest route is to ask whether the scenario is telling you about something that broke or something that was learned.