Williamson applies a transaction costs framework to evaluate the relative strengths of the “public agency” as a mode of governance and compares it to private and semi-private (regulated) alternatives. His central premise is that public agencies are “well-suited for some purposes but poorly suited for others” and the article lays out his rationale for why public agencies have a comparative advantage when it comes to “sovereign transactions.” Williamson focuses on foreign affairs as an illustrative example of a sovereign transaction that helps to flesh out his ideas.

In Section 1, he reviews some of the literature on transaction cost economics:

•  Wilson (1989): It is difficult to evaluate public agencies due to multiple objectives, and equity and accountability concerns.
•  Moe (1990, 1997): Bureaucracy is over-formalized and inefficient in democracies because of insecure property rights and separation of powers.
• North (1990): Issues with high transaction costs gravitate toward the public sphere.
•  Dixit (1996): Analyzes public agencies as one of several flawed but feasible alternatives rather than as a theoretical construct.

Section 2 characterizes the nature of “transaction cost economics.” Under this framework, transactions are the relevant unit of analysis. Williamson views all contracts as incomplete because of bounded rationality and opportunism. In particular, he describes problems arising from this reality as “hazards” that modes of governance (or organization) attempt to overcome. In order to evaluate modes of governance for a given purpose, one must: (1) identify the key attributes of transactions; (2) state the core purposes/trade-offs of organizations, and (3) identify the structural attributes of governance structures and how they differ. He uses private firms (markets) and public agencies (hierarchies) as alternatives that can be evaluated based on how well they address hazards associated with a given transaction type. For him, markets are characterized by “high-powered incentives with little administrative control and a legalistic dispute settling mechanism.” Hierarchies, on the other hand, use low-powered incentives, extensive administrative control, and internal dispute resolution.

In Section 3, Williamson explains that his analysis is concerned with evaluating feasible alternatives rather the theoretical optimum (remediableness). For him, if no better feasible alternative can be described, then the existing mode is presumed to be efficient. He defends this premise but mentions four exceptions under which it does not apply, including: lack of political legitimacy, predatory/strategic actions with no social benefit, a mistake in original efficiency calculations, or changes to an organization that were unintended by its creators. Section 4 goes on to briefly introduce the six types of public transactions: procurement, redistribution, regulatory, sovereign, judiciary and infrastructure. However, the rest of the article is only concerned with sovereign transactions.  

According to Williamson (Section 5), Sovereign transactions are those that are considered essential functions of the State as the embodiment of public authority. These functions are typically administered by the executive branch, and include activities such as national defense, managing the money supply, and foreign affairs. Sovereign transactions have three hazards associated with them: asset specificity, probity and cost excesses. Asset specificity refers to the idea that the skills and equipment necessary for carrying out these transactions are highly specialized and not very portable. To remedy this, an organization often emphasizes retaining and socializing staff. Probity refers to the integrity, cohesiveness and responsiveness of the organization that is necessary to carry out sovereign transactions. Probity has three elements to it: vertical (relationship between organization and executive), horizontal (relationship with counterparts) and internal (relations within the organization). Finally, cost excesses refer to the ability of the organization to carry out the transaction at the lowest cost. In the context of his chosen example, foreign affairs, he claims that probity is highly important (strong need to be responsive to the president and act with integrity), asset specificity is moderately important (Foreign Service officers have specialized training and only some ability to transfer skills) and cost excesses are relatively minor. He claims that public agencies are well suited to addressing sovereign transaction hazards because they employ low-powered incentives that do not encourage unwanted creativity, and enhanced job security that helps retain employees. He then goes on to describe how the actual U.S. State Department demonstrates these characteristics as further proof that they are positive adaptations and not merely signs of inefficiency.

In Section 6, Williamson conducts a thought experiment by asking the question: could foreign affairs be privatized or partially privatized? He contends that a contract for managing foreign affairs would be highly incomplete because there are so many contingencies to plan for. This would impose high transaction costs if the contract had to be continually renegotiated and provide less responsiveness to the President. High powered incentives and concern for the bottom line, which are the cornerstones of efficient private sector performance, are at odds with the tow-the-line atmosphere desired in foreign affairs. The incentives might also affect executive appointment, staff retention and administrative controls in ways considered adverse to the mission of a foreign affairs organization. Overall, Williamson finds that it is impossible for a purely private firm to copy the positive aspects of public agencies without sacrificing the profit incentive that is by definition a part of private firms. He also considers the case of a regulated private firm and concludes that the same trade-offs are operating here too but to a lesser extent. A regulated firm that gives the President more control over executive appointments, provides increased job security and otherwise follows regulations designed to introduce positive aspects of bureaucracy, will also put an extra layer between the President and the firm, and introduce inefficiency since the firm has less flexibility. Table 2 on page 336 provides an overview of the trade-offs.

Sections 7 and 8 provide further discussion of his results and extends his analysis to other transaction types in Table 3 (page 339). Overall he finds that public agencies have a comparative advantage when it comes to sovereign transactions and potentially others.