Restrictive housing for prison rule violators: Specific deterrence or defiance?

Objectives

To determine whether segregation (restrictive housing/RH) for first-time prison rule violators in Ohio shapes his/her odds of violations thereafter in a manner consistent with specific deterrence or criminogenic effects.

Methods

Four types of propensity score analyses were conducted to assess placement in and length of RH for 14,880 rule violators: Matching without replacement, matching with replacement, propensity score stratification, and inverse probability of treatment weighting. Group differences were also compared using Abadie-Imbens robust standard errors.

Results

Findings were robust across methods. RH placement corresponded with significantly but modestly lower odds of any misconduct within 12 months after release from RH, but effects on more serious offenses (particularly violence) were weak. Longer RH terms corresponded with modestly higher odds of any misconduct.

Conclusions

Lower odds of misconduct post-RH is consistent with specific deterrence, but the higher odds associated with longer RH terms is consistent with a criminogenic effect.