Testing the Expertise Hypothesis

Joshua D. Wright & Angela M. Diveley, Do Expert Agencies Outperform Generalist Judges?  Some Preliminary Evidence from the Federal Trade Commission (2012), available at SSRN.

Proponents of administrative agencies have long touted the expertise that specialized agencies enjoy. Indeed, perceived agency expertise helps to explain Congress’s willingness to delegate to agencies the authority to set policy through rulemaking, to render adjudicatory decisions, and to conduct other activities. Yet, as Joshua D. Wright and Angela M. Diveley point out in a study posted this past January to SSRN, the so-called “expertise hypothesis”—which posits that expert agencies will consistently produce higher quality outputs than generalists—lacks empirical support. In their recent study, Wright and Diveley seek to fill the void by conducting an empirical study that examines whether the Federal Trade Commission (FTC) performs as well as generalist judges in its adjudicatory antitrust decision-making role.

Specifically, Wright and Diveley’s study tests the expertise hypothesis by comparing antitrust decisions before the FTC with those issued by Article III courts. They use “appeal” as their primary measure of quality performance—comparing the appeal rates of federal district court judges and FTC Commissioners. They explain that appeals are a “useful indicator for whether the initial court made an error” because “a higher appeal rate implies the decision-maker has issued more opinions that leave at least one party feeling strongly enough to invest in the opportunity for another decision-maker to decide that he has committed reversible error.” (P. 12) Nonetheless, because they acknowledge that reversal rates also can contain some information on the quality of the underlying decision, Wright and Diveley also report their results about the differences between the FTC and generalist judges using reversal rates. 

In terms of the study’s conclusions, Wright and Diveley ultimately find evidence that suggests—contrary to the expertise hypothesis—that the FTC’s decisions are more likely to be appealed and reversed than those of Article III judges and that the FTC does not perform as well as generalist judges in its adjudicatory antitrust decision-making role. Wright and Diveley acknowledge that these differences in appeal rates may be the result of various factors that could influence the decision to appeal from an FTC decision differently than from a district court judgment. For example, an overwhelming majority of Commission decisions favor the plaintiff (i.e., the FTC) and thus Wright and Diveley note that “appeals from Commission decisions may be systematically different in quality or other dimensions from the distribution of cases from which appeals from district court opinions are drawn.” (P. 21) Yet after Wright and Diveley attempt to control for such factors, they consistently observe higher appeal and reversal rates for the Commission. Hence, they claim that their basic results remain unchanged.

In terms of its overall contribution to the literature, the study—as the authors acknowledge—is fairly narrow in scope. It analyzes just one agency (the FTC), it examines only one aspect of the agency’s activities (agency decision-making in administrative litigation), and it presents what is described as merely “preliminary evidence” on the question studied. Nonetheless, despite its limits, the study is quite notable for the dialogue that it seeks to open on some very important questions of institutional competency and design. Specifically, the study is well worth a read because it highlights the need for more research on the relationship between institutional design and agency expertise, and for more analysis of the “expertise” hypothesis. This additional research is needed not just in the antitrust context where Wright and Diveley focus their attention, but also in many other areas that involve a choice of delegating to courts or agencies, such as bankruptcy administration, which Professor Rafael Pardo and I recently argued is in need of more careful study and attention.

In short, Wright and Diveley’s preliminary study gives some much-needed attention to the performance of courts versus agencies. Hopefully more scholars—inspired by Wright and Diveley’s work—will engage in additional empirical research about the relationship between institutional design and agency expertise. Such future research is needed to shed more light on the optimal roles of courts and agencies and on whether Congress should delegate certain powers to generalist courts or to specialized agencies. Scholars, such as Margaret Lemos, have recently focused on some of the theoretical and doctrinal implications of Congress’s decision to delegate to agencies versus courts in the policymaking context, and additional empirical work along the lines of what Wright and Diveley have produced would help to inform these broader theoretical and doctrinal discussions.