Organizations always involve numerous staff members whose behavior has the potential for creating significant risk for individuals and the organization but who are only loosely supervised. This situation unavoidably raises principal-agent problems. Let's assume that the great majority of staff members are motivated by good intentions and ethical standards. That means that there are a small number of individuals whose behavior is not ethical and well intentioned. What arrangements can an organization put in place to prevent bad behavior and protect individuals and the integrity of the organization?
For certain kinds of bad behavior there are well understood institutional arrangements that work well to detect and deter the wrong actions. This is especially true for business transactions, purchasing, control of cash, expense reporting and reimbursement, and other financial processes within the organization. The audit and accounting functions within almost every sophisticated organization permit a reasonably high level of confidence in the likelihood of detection of fraud, theft, and misreporting. This doesn't mean that corrupt financial behavior does not occur; but audits make it much more difficult to succeed in persistent dishonest behavior. So an organization with an effective audit function is likely to have a reasonably high level of compliance in the areas where standard audits can be effectively conducted.
A second kind of compliance effort has to do with the culture and practice of observer reporting of misbehavior. Compliance hotlines allow individuals who have observed (or suspected) bad behavior to report that behavior to responsible agents who are obligated to investigate these allegations. Policies that require reporting of certain kinds of bad behavior to responsible officers of the organization -- sexual harassment, racial discrimination, or fraudulent actions, for example -- should have the effect of revealing some kinds of misbehavior, and deterring others from engaging in bad behavior. So a culture and expectation of reporting is helpful in controlling bad behavior.
A third approach that some organizations take to compliance is to place a great deal of emphasis the moral culture of the organization -- shared values, professional duty, and role responsibilities. Leaders can support and facilitate a culture of voluntary adherence to the values and policies of the organization, so that virtually all members of the organization fall in the "well-intentioned" category. The thrust off this approach is to make large efforts at eliciting voluntary good behavior. Business professor David Hess has done a substantial amount of research on these final two topics (link, link).
Each of these organizational mechanisms has some efficacy. But unfortunately they do not suffice to create an environment where we can be highly confident that serious forms of misconduct do not occur. In particular, reporting and culture are only partially efficacious when it comes to private and covert behavior like sexual assault, bullying, and discriminatory speech and behavior in the workplace. This leads to an important question: are there more intrusive mechanisms of supervision and observation that would permit organizations to discover patterns of misconduct even if they remain unreported by observers and victims? Are there better ways for an organization to ensure that no one is subject to the harmful actions of a predator or harasser?
A more active strategy for an organization committed to eliminating sexual assault is to attempt to predict the environments where inappropriate interpersonal behavior is possible and to redesign the setting so the behavior is substantially less likely. For example, a hospital may require that any physical examinations of minors must be conducted in the presence of a chaperone or other health professional. A school of music or art may require that after-hours private lessons are conducted in semi-public locations. These rules would deprive a potential predator of the seclusion needed for the bad behavior. And the practitioner who is observed violating the rule would then be suspect and subject to further investigation and disciplinary action.
Here is perhaps a farfetched idea: a "behavior audit" that is periodically performed in settings where inappropriate covert behavior is possible. Here we might imagine a process in which a random set of people are periodically selected for interview who might have been in a position to have been subject to inappropriate behavior. These individuals would then be interviewed with an eye to helping to surface possible negative or harmful experiences that they have had. This process might be carried out for groups of patients, students, athletes, performers, or auditioners in the entertainment industry. And the goal would be to uncover traces of the kinds of behavior involving sexual harassment and assault that are at the heart of recent revelations in a myriad of industries and organizations. The results of such an audit would occasionally reveal a pattern of previously unknown behavior requiring additional investigation, while the more frequent results would be negative. This process would lead to a higher level of confidence that the organization has reasonably good knowledge of the frequency and scope of bad behavior and a better system for putting in place a plan of remediation.
All of these organizational strategies serve fundamentally as attempts to solve principal-agent problems within the organization. The principals of the organization have expectations about the norms that ought to govern behavior within the organization. These mechanisms are intended to increase the likelihood that there is conformance between the principal's expectations and the agent's behavior. And, when they fail, several of these mechanisms are intended to make it more likely that bad behavior is identified and corrected.
(Here is an earlier post treating scientific misconduct as a principal-agent problem; link.)