Repairing Broken Windows
Perry, Frank L., The FBI Law Enforcement Bulletin
Preventing Corruption within Our Ranks
In the last two decades, research and commentary regarding the causes and effects of law enforcement corruption have intensified and diversified. Efforts in Australia, Canada, Great Britain, and the United States have effectively identified symptoms and remedies in those countries, as emerging democracies in Africa, Eastern Europe, and the Pacific Rim face the more immediate and stark realities of self-governance and the police role. Comparative reviews of problems and best practices, as well as academic research, suggest that corruption follows certain predictable routes and that precursory signs occur prior to any actual quid pro quo corrupt activity.
Three organizational failures can foster a resentful, cynical, and demoralized work force leading to individual and collective acts of corruption. These failures are: little or ineffective discipline and deselection of trainees (a commitment to fairly but firmly graduate only those individuals who truly demonstrate performance and integrity standards); ignorance of the nature and effects of the goal-gradient phenomenon (the farther away individuals remain from their goal, the less the tendency to remain passionately interested in its attainment); and the allowance of a double standard within the organization, thereby decreasing moral accountability as professional responsibility increases. All of these factors represent instances of what sociologists have referred to for many years as the "broken window theory"--if enough broken windows in a neighborhood go unattended, the neighborhood falls into a moral and material malaise. Law enforcement applications of this theory are addressed rarely. 
Corruption can include an abuse of position, although not all abuses of position constitute corrupt acts.  Committing a criminal act under color of law  represents one example of corruption, while using one's law enforcement position for a de minimus, or insignificant, private gain may not necessarily rise to what reasonable persons will call a corrupt act, though it may be corrupting.  All self-interested or potentially corrupt acts are not completely corrupt.  In fact, these acts can constitute police deviance,  which best captures the nature of the precursory signs of corruption, as opposed to actual corruption.
Precursory signs, or instances of police deviance, may be agency--specific, or generic and found in law enforcement as a profession. Unprofessional on- and off-duty misconduct, isolated instances of misuse of position, improper relationships with informants or criminals, sexual harassment, disparaging racial or sexual comments, embellished/falsified reporting, time and attendance abuse, insubordination, nepotism, cronyism, and noncriminal unauthorized disclosure of information all represent precursory signs of police deviance that inspection and internal affairs components must monitor. When agencies determine a trend of increasing frequency and egregiousness of such deviance, they must take steps before classic or quid pro quo corruption occurs. An organization with an increase in such deviance becomes a "rotten barrel," even without completely "rotten apples."
Literature on the rotten-barrel concept has become more sophisticated. One study surmises that most of the major inquiries into police corruption reject the "bad-apple" theory: "The rotten-apple theory won't work any longer. Corrupt police officers are not natural-born criminals, nor morally wicked men, constitutionally different from their honest colleagues. The task of corruption control is to examine the barrel, not just the apples, the organization, not just the individual in it, because corrupt police are made, not born." 
How, then, can agencies examine the barrel? They must analyze the increasing frequency and egregiousness of precursory signs, then assess their department's training. …