This post describes some ideas that have developed while working on a project with Justin Ready. We have been examining data collected from police officers regarding their attitudes toward and use of on-officer cameras. In what follows, I examine the principal agent problem with respect to the public and social control organizations, such as the police.
A principal agent relation is simply a form of exchange where a principal seeks to have some task performed by an agent. Put simply, I want you to do something, so I pay you to do it. For example, a principal needs the oil changed in his car, but does not know how to perform this task. Rather than fumble around and try to accomplish the task through trial-and-error, the principal can simply pay a mechanic (agent) to do it for him. Both are now better off than before the exchange.
The principal agent problem occurs when the market mechanism does not function properly due to asymmetric information. Let's return to the oil-change example. Since the principal does not know how to change the oil in his car, the mechanic (agent) could simply take the money and not do any work. The principal's lack of information about whether an oil change has been performed is a problem of asymmetric information. The agent knows if the task has been performed, but the principal does not. As a consequence, the principal cannot make the agent account for his actions. Under these conditions (i.e. market friction), a rational principal would not engage in exchange. The fact that many people still exchange under these conditions points to the importance of trust as a mechanism for greasing the wheels to reduce frictions (but this is a separate issue of which the current discussion does not pursue further). In principal agent relations that have minimal transaction difficulties, the principal can hold the agent accountable. In other words, accountability concerns the ability of a principal to determine whether the agent has done what is expected of him.
What does this have to do with the public and the police?
If we think of a principal as a collective interest (e.g. the public), then we can apply this same form of interaction to exchanges between the public and formal organizations of social control. This is exactly what Thomas Hobbes did nearly four centuries ago who argued that, to ensure public safety, the commonwealth establishes a formal entity of social control, in exchange for freedom and capital. In other words, the public tasks an organization with ensuring social control. Though Hobbes emphasized that only a despotic and terrifying organization could accomplish the task (a Leviathan), formal organizations of social control in the U.S. are much less powerful and more democratic then the institution Hobbes envisioned.
With this in mind, we can return to our original problem of how a principal holds an agent to account for his actions. In the case of policing organizations, a primary concern is the manner in which interactions with the public are conducted. Research indicates that most individuals have little knowledge of what actually occurs during police initiated contact. Thus, there is a lack of symmetry between the public regarding police conduct during such interactions (i.e. lack of knowledge about task performance). Despite such information asymmetry, the prevalence of formal social control organizations indicates that trust is an important device for influencing perceptions of the police among the public.
Recently, however, information symmetry between the public and the police has fundamentally changed. The information revolution and the diffusion of technological innovations (e.g. cell-phones) has altered the ability for individuals to create, distribute, and access information. Of greatest relevance here, is the spread of third party recordings of police initiated contact ("Don't taze me bro!" comes immediately to mind). This shift in the public's ability to access information about police initiated contact is directly relevant to the principal agent problem previously described.
The availability of this information changes the relationship such that the public is in a better position to hold the police more accountable with respect to their behavior. The problem with this change is the assumption that the newly available information is accurate. That is, if third party recordings reduce information asymmetry, then the validity and representativeness of the recorded content is crucial for making decisions about police accountability. The diffusion of unrepresentative information about police conduct represents an organizational dilemma for police departments. That is, information available to a principal that does not accurately reflect the agent's task performance may create friction in the relation. In this case, unrepresentative information may incorrectly undermine police legitimacy among the public. As many scholars have recognized, transparency in the organization is one means of demonstrating accountability. With respect to policing organizations, on-officer cameras represent a technological innovation which can facilitate adaptation to this dilemma and foster transparency in the organization.
We are still working on this issue with the data we are examining. When we are ready to release findings on our analysis I will be sure to post them here.