Compared to operations within three public prisons run by the Federal Bureau of Prisons (BOP), the operations within the private prison were found to be within the lower range of performance for low-security prisons. Statistical models for all misconduct, violent misconduct, and drug misconduct over a 36-month period found that the private prison did not perform as well overall as the three public prisons. Data were drawn from BOP records on all sentenced inmates incarcerated between January 1999 and December 2001. The analytic technique involved the use of sophisticated statistical models, such as the estimation of Empirical Bayes residuals as measures of prison performance. The authors note that all prisons in the analysis were similar in architectural design, size, age, and money spent per inmate. While the findings are useful for the guidance of criminal justice policy concerning the efficacy of using private prison models, the study is also important for demonstrating research methods for comparing the relative performance of prisons. Footnotes, tables, figures, references