The most prominent theories for explaining the incidence and prevalence of misconduct in prison are deprivation (Clemmer, 1940; Sykes, 1958; Colvin, 1992), importation (Irwin and Cressey, 1962; Harer and Steffensmeier, 1996; Cao Zhao, and Van Dine, 1997), and administrative control (DiIulio, 1987; Useem and Kimball, 1989; Useem and Reisig, 1999). Administrative control does not supersede deprivation and importation theories, but rather adds to them by asserting quality management is essential for the maintenance of an orderly environment jeopardized by the effects of deprivation and importation. Even though research has supported administrative control, researchers have disagreed over which aspects of management are most effective for mitigating disorder, and have thus far only proven that poor management leads to administrative breakdowns in policies and practices that predicate disorder. This dissertation introduces the concept of positive administrative control, based on social exchange theory, as the mechanism prison administrators should use to induce staff to act in accordance with agency desires and avoid administrative breakdowns. The results show that when prison employees experience quality relationships with their supervisors, they are more likely engage with inmates in a similar fashion, which is associated with reductions in misconduct rates. At the same time, when prison employees are supported and rewarded for their good behavior, they are more likely to actively monitor inmates, which is associated with increases in misconduct rates. Additionally, the results support importation theory by demonstrating that the aggregated criminogenic characteristics of inmates in a prison are representative of cultural influences on prison misconduct.

Based on these findings, recommendations are made to restructure leadership training to emphasize relational leadership skills that positively influence staff to act in accordance with agency desires and more closely follow policy guidelines. It is also recommended that future studies consider including aggregated demographic variables in studies of misconduct in order to capture cultural/environmental influences that may otherwise be missed, and should avoid overuse of composite variables, in particular institution security level.