Economists have developed a large literature on optimal contracting in a principal-agent setting, illustrating how a labor contract can overcome informational asymmetries. Authors such as Tirole (1986) have expanded upon the standard principal-agent framework to consider the possibility of collusion within firms. For instance, Tirole shows that in a three-tier hierarchy where the principal hires a supervisor to report on the agent, possibilities for corruption exist wherein the supervisor might become an advocate for the agent. The supervisor could be bribed by the agent to provide preferential performance evaluations. Such activity generates a surplus for both the supervisor and the agent at the principal's expense.
Such corrupt and collusive behavior can take many forms. Robinson and Bennett (1995) categorize deviant behavior as, for example, unjustifiably terminating an employee, working unnecessary overtime, accepting kickbacks, and lying about hours worked, and they note that consistent punishment, proportional to the seriousness of the deviant action, is crucial to preventing such behavior and maintaining organizational morale. Kacmar and Carlson (1998) focus on deviant political behavior in organizations. They document employee complaints about preferential promotions, demotions, job assignments, and performance evaluations. Some examples cited specifically portray collusive cliques at work within the firm. For example, one respondent indicated that he was asked by the V.P. of Operations to spy on another department. When he refused, his mail was opened and he was prevented from working with members of that VP's "team." As Kacmar and Carlson note, such collusion creates serious problems for the organization, ranging from inefficient use of resources to lowered morale. Raelin (1986) hypothesizes that such a political environment encourages additional deviant behavior.
Tirole and others such as Holmstrom and Milgrom (1991) and Bac (1996) speculate that bureaucratization and formalization can prevent some types of intrafirm collusion.  By restricting the discretion of workers, the firm can reduce the opportunities for and potential gains from corruption. Sundstrom (1988) documents one instance of such behavior by studying the wage-setting policies of an explosives factory in the WWI era. Prior to reform, wages for most workers were set by shop foremen with little or no regulation from central authority. The result was a "foreman's empire" rife with the predictable corruption and kickbacks. Such a system of wage setting would allow a corrupt foreman to sell jobs, reward workers who cooperated in other types of corruption, or simply buy the silence of potential whistle-blowers. Standardization of wages reduced this form of collusion; foremen still made most hiring decisions, but their ability to grant preferential or inflated wages was eliminated. As Tirole (1986) notes, s uch corruption was reduced even further when firms took hiring and promotion responsibilities away from shop foremen and transferred them to a central human resources department.
Similarly, other types of formalization, such as formal labor contracts, job descriptions, written evaluations, or policies and procedures manuals can reduce the autonomy of any particular employee, thereby reducing the opportunities for corruption. According to Demski et al. (1999), such formalization is useful in consulting and law firms, as well as other organizations that deal with confidential information about their clients. In the absence of codified standards, deviant employees might use this proprietary information for their own private ends, but bureaucratization can check this impulse and allow such information to flow safely and efficiently through the firm. Likewise, Miceli (1996) argues that deviant managers often are in a position to reward compliance or punish dissent, and that consistent application of rules within the firm can control these tendencies and deny such managers the opportunity to abuse their authority. …