Recent developments in American public policy include a strong movement toward the devolution of federal policy to states. Under the Personal Responsibility and Work Opportunity Act of 1996, states were given discretionary power to implement the specifics of the program within their own states. This article will analyze Mississippi's approach to welfare reform and the effects of its implementation choices on administrative functions and program goals. Mississippi's program relies heavily on "second-order devolution" (Nathan 1998), devolving programmatic functions--specifically, "point-of-contact" functions--to private and nonprofit organizations.
Mississippi has privatized two components of a three-component process. However, the social contract between government and citizens places accountability for government programs entirely on the shoulders of the government. Citizens, whether or not they are recipients of a specific government program, have an expectation in a democracy that their government will be accountable to them. Privatization, and in this case multiple-component privatization, dilutes accountability and threatens the social contract.
The theoretical lens employed in this article is principal-agent theory. In its traditional formulation, there is one principal and one agent. We will argue that the standard two-actor conception fails to capture the richness and complexity of democratic governance, and that privatization complicates the principal-agent relationship by adding additional actors and roles into the mix, such that elements of government become both principal and agent. When more than one element of the same program is privatized by different entities, the principal-agent theory becomes more complicated until the roles of principal and agent are even further removed from the original principal, the citizenry. The issue that arises is how these new roles affect programmatic functions when the agent is ultimately responsible to the principal. These issues raise specific questions relevant to this article: Does principal-agent theory, in its standard two-actor configuration, capture the complexity of the privatization arrangement? How does a more robust conception of a principal-agent relationship affect our understanding of monitoring and accountability in privatization?
These issues and questions are of current relevance to fully understand the impact of privatization as a solution for smaller, more efficient government. If privatization is to continue to be a policy solution, we must analyze its impact on policy outcomes. The effects on government accountability have often been analyzed, mainly from a contractual perspective (Kettl 1988; Savas 1987); this article will examine the larger implications of accountability from constitutional expectations of government responsibility.
From a practitioner's perspective, research into the programmatic effects of privatization, and especially privatization of multiple elements of a single program, will provide additional information about the manageability of government programs. We contend that complications of accountability and manageability of programmatic functions negatively impact the entire program and the ability of agencies to implement welfare reform.
The traditional conception of the principal-agent relationship consists of a single principal and a single agent (Donaldson 1990; Moe 1984; Kettl 1991; Donahue 1989; Jensen and Meckling 1976). The principal, who wishes to accomplish a goal, hires an agent to act as his surrogate and to carry out his wishes as they relate to that goal. Thus, the agent is bound (typically through a contract) to carry out the wishes of the principal and to help the principal achieve the relevant goal. In return, the agent is compensated by the principal.
Accountability is a hallmark of the principal-agent relationship (Donahue 1989). …