The Mark SOC 3333 · Crime Lab · Midwestern State University
Every theory you have met so far explains why people break rules. Labeling theory asks a different question, and it is the question Chapter 1 named as the field's third domain: what does our reaction to rule-breakers do? The provocative version of the claim is that the justice system can manufacture the criminality it exists to fight, by taking someone who did an ordinary bad thing and turning that into who they are. That claim is easy to state and notoriously hard to test, because you cannot randomly assign people criminal records. In 2001 a graduate student found a way around that, and this page starts with what she found.
Devah Pager's experiment
Pager hired four young men, matched them carefully on appearance, age, speech, and self-presentation, and gave them identical fabricated resumes with the same education and work history. She then sent them to apply for real entry-level jobs advertised in Milwaukee, in pairs, over seven months in 2001. Within each pair, one man reported a criminal record and one did not, and which man carried the record rotated week to week so it could not be confounded with anything about the individual.
The record was an eighteen-month sentence for a first-time drug possession offense, signaled on the application form where employers asked, and otherwise through a work assignment obtained in a correctional facility and a parole officer listed as a reference. Three hundred fifty employers were audited in total. The outcome was simply whether the employer called back.
Before you look, commit to all four numbers. Set each slider to the callback rate you expect. Guessing badly here is normal and the point of the exercise is to have a prediction on record before the answer arrives.
Act one: what the record did
Now the second half of the design, which Pager included for a reason and which produced the finding that her reviewers found hardest to believe.
Pager, D. (2003). The mark of a criminal record. American Journal of Sociology 108(5), 937 to 975. Fieldwork in Milwaukee, June through December 2001. Three hundred fifty employers audited, 150 in the white pair and 200 in the Black pair. Retrieved August 2026.
There is a second finding stacked on top of that one. The record did not just lower everyone's chances equally. It hurt the Black testers more. Callbacks fell by about half for the white pair and by roughly two thirds for the Black pair, a penalty Pager reported as about 40 percent larger for Black applicants. So a criminal record is not one obstacle applied uniformly. It is an obstacle whose size depends on who is carrying it.
Why an audit and not a survey
This is a Chapter 2 slide wearing different clothes. The reason Pager's study settled an argument that decades of survey research could not is entirely about design, and it is worth being precise about what the design buys.
That last point is the one that matters most. Ask employers on a survey whether they would hold a five-year-old drug conviction against an applicant and many will say no, sincerely. Watch what they do with an actual application and you get a different number. Attitudes and behavior come apart, and audits measure the half that has consequences.
Now the limits, stated as carefully as the findings
The sentence after the sentence
Pager measured one gate. There are a great many gates. Criminologists call the whole set collateral consequences: the legal and practical penalties that attach to a conviction and continue long after the formal sentence is finished. Labeling theory has a name for what these amount to. They are a master status written into law and into databases, and a master status is a label that answers every question about a person before it is asked.
Labeling theory makes a prediction about what should follow when the legitimate doors close. Before you see the number, guess it.
Durose, M. R., and Antenangeli, L. (2021). Recidivism of prisoners released in 34 states in 2012: A 5-year follow-up period, 2012 to 2017. Bureau of Justice Statistics. Return to prison includes both new sentences and parole or probation violations. Retrieved August 2026.
Here is where you have to be disciplined. These numbers are perfectly consistent with labeling theory, and they are equally consistent with almost every other theory in this course. Control theory would point to the bonds that prison severs. Strain theory would point to blocked means. Learning theory would point to who you spent those years with. Self-control theory would say the same people who offended before were always going to offend again. Recidivism data cannot arbitrate between them, because it is an outcome that every theory predicts.
Labeling theory with a hall pass
The school-to-prison pipeline is labeling theory operating inside a public institution. The claimed sequence runs from zero-tolerance discipline, through suspension and exclusion, to falling behind and dropping out, to police contact, to a record. The hard question is causal. Are the suspended students simply the ones who were already headed for trouble, or does the exclusion itself do damage?
One study got unusually close to an answer by using something that had nothing to do with any individual student. Between 1999 and 2008 the federal government awarded competitive grants to school districts to hire police officers. A researcher compared discipline outcomes within the same Texas districts across years with and without that grant funding, covering more than two and a half million students.
Weisburst, E. K. (2019). Patrolling public schools: The impact of funding for school police on student discipline and long-term education outcomes. Journal of Policy Analysis and Management 38(2). Texas districts, COPS grants 1999 to 2008, over 2.5 million students. The graduation and enrollment results are described by the author as suggestive. Retrieved August 2026.
Two honest qualifications. First, the graduation and college enrollment findings are described by the author herself as suggestive rather than definitive, and you should repeat that word if you cite them. Second, this is one state over one period, and school policing has changed since. The discipline finding is the solid one. The long-term education findings are the interesting one.
Braithwaite: shaming that reintegrates, and shaming that casts out
If labels cause secondary deviance, the obvious conclusion is that societies should stop condemning wrongdoing, which is absurd and which no labeling theorist has ever argued. John Braithwaite proposed the resolution in 1989. What matters is not whether a society shames but how. Reintegrative shaming condemns the act and then deliberately readmits the person. Stigmatizing shaming condemns the person and leaves them outside, which is precisely how you manufacture a master status.
Eight responses to the same offense are below. Sort each one.
So does it work
The main policy expression of reintegrative shaming is restorative justice conferencing, in which the person who caused harm meets the people affected, hears what it did, and agrees to repair it. The National Institute of Justice has graded the practice for juveniles, outcome by outcome, and the pattern is strange enough to be worth sitting with.
All three are rated Promising. Now the outcomes rated Ineffective, meaning no statistically significant effect was found: offender satisfaction, victim satisfaction, remorse, and recognition of wrongdoing.
CrimeSolutions, National Institute of Justice, practice profile for Restorative Justice Programs for Juveniles. Delinquency effect size 0.23 across 90 comparisons in Wilson et al. (2017), with a second meta-analysis reporting a log odds ratio of 0.248 across 21 comparisons and a third finding no significant effect. Retrieved August 2026. Ratings are periodically re-reviewed.
Three possibilities, and you should be able to argue all three. The mechanism may be real and badly measured, since remorse is self-reported at a single moment and people are poor reporters of it. The mechanism may be wrong while the practice still works, perhaps because conferencing simply keeps young people out of court and therefore out of the labeling machinery, which would be an even purer confirmation of labeling theory. Or the reoffending effect may be smaller than it looks, since one of the three meta-analyses found nothing. This is what a live research question looks like, and Chapter 13 will return to it.
Forum: argue both sides of ban-the-box
Ban-the-box policies prohibit employers from asking about criminal history until late in the hiring process, typically after a first interview. Post your initial response and then reply to two classmates. Length and deadlines are in that week's module in the LMS.
- Argue FOR ban-the-box using labeling theory. Name the concepts you are using, and use Pager's actual numbers rather than the idea of them.
- Now argue against it, using any other theory from this course. Make it the strongest version of the objection you can construct, not a version you can easily knock down.
- Say which side you actually find more persuasive and what specific evidence would change your mind. Not what argument would change your mind. What evidence.
Where to look for the objection. A serious argument against ban-the-box exists and it comes from within this course. If employers cannot ask about records, and they still want to avoid hiring people with them, they have to guess. Some research has found that removing the question increased racial gaps in callbacks, on the theory that employers fell back on statistical assumptions about who is likely to have a record. Whether that finding holds up is contested. Steel-manning that argument, rather than dismissing it, is what step 2 is asking for.
On step 3. Naming the evidence that would change your mind is the hardest sentence in the post and the one worth the most. If you cannot name any, that is worth noticing about your own position.