Notice & Comment

Replacing Agency Adjudication with Article III Administrative Courts, by Michael B. Rappaport

*This is the third post in a symposium on the decisional independence of administrative adjudicators. For other posts in the series, click here.

One of the troubling aspects of the modern administrative state is that administrative adjudications violate the separation of powers—either understood as the Constitution’s original meaning or as a political principle. Yet, this violation is tolerated and defended on pragmatic grounds. It is thought that administrative adjudication by independent Article III courts would be undesirable because these courts would lack expertise and would incur high decision-making costs. But I argue here (relying on previous scholarship) that it is possible to establish independent Article III administrative courts that possess expertise and reach decisions expeditiously.

Combined Rather Than Separated Powers

At present, formal administrative adjudications are conducted by administrative agencies rather than Article III courts. Consequently, a single agency both decides whether to bring an enforcement action against a private actor and adjudicates whether that actor has violated the law. By contrast, under the separation of powers where adjudications are conducted by Article III courts, judges decide questions of fact and law, and their decisions can only be reviewed by other Article III judges.

The combination of executive and judicial power under administrative adjudication distorts the exercise of both of these powers. Because the agency has made the decision to bring an enforcement action and has partially staked its reputation on that action, its adjudication is likely to be biased in its own favor. The agency will be the judge in its own case—a classic example of improper government procedure. The combination of executive and judicial power also undermines the prosecutorial or enforcement function. Prosecutors are often constrained by the requirement that they win their case in a neutral forum. But if the agency prosecuting the case also adjudicates it, the agency can bring less supported cases and still expect to win. 

It is true that administrative law imposes some checks on the agency to guard against these problematic incentives, but these checks are not adequate to generate the level of fair and accurate decision-making produced by genuinely independent judges. First, administrative law judges (ALJs), who generally preside at formal adjudications, are protected by good cause removal provisions and are not supervised by agency investigatory and enforcement personnel. But this independence, while important, is nonetheless limited.

Most significantly, ALJ decisions are typically subject to review by higher agency officials who are largely free to reverse the decisions. This power to review ALJ decisions not only gives the agency formal authority as to adjudications, but also gives the agency indirect influence over ALJs since they are likely to anticipate the agency’s position to the extent that they have the normal human tendency to avoid reversal. Agency heads also retain the authority to decide the adjudication on their own, so they can bypass ALJs in cases of their choosing, such as when they want to avoid the possibility of an adverse decision below.

Second, administrative adjudications are subject to judicial review by Article III courts, but such review is limited by a strong set of deference rules, such as Chevron and substantial evidence review. Thus, so long as an agency does not stray too far, it has significant discretion to enforce whatever decision it seeks to impose.  

These defects of administrative adjudication are compounded by Supreme Court decisions that have threatened to undermine administrative adjudication. For example, a Supreme Court decision holding unconstitutional dual for-cause removal restrictions threatens to eliminate the removal protections of the ALJs of independent agencies. 

Despite the serious problems with administrative adjudication, these defects have largely been accepted on the ground that strictly following the separation of powers would not be feasible for our existing administrative state. It is thought that adjudication by Article III courts would involve both a serious lack of expertise and high decision-making costs from adjudications that are both lengthy and expensive.

But this objection to following the separation of powers is mistaken. While employing Article III courts on the model of U.S. District Courts—with generalist judges following strict procedural rules—would suffer from these problems, Article III administrative courts could be restructured to avoid them.


Rather than employ generalist Article III judges, Article III administrative courts should be staffed with people having one of three types of expertise: medical, scientific, or economic. The administrative judges should be divided into three groups, each of which has expert knowledge in one of these areas. For example, persons in the medical area might have developed medical knowledge either through additional education, such as a master’s degree, or through career experience.

The judges in each of these groups would not hear cases from only a single agency. Instead, they would be assigned cases implicating their expertise from any of the agencies. For example, the administrative judges with medical expertise would hear cases involving medical issues from OSHA, EPA, MSHA, or other agencies.

While these judges would have expert knowledge as to non-legal matters, they would also develop legal expertise in their specific area. Some of the judges might have this legal knowledge from their prior years as a lawyer. But even if they did not initially have this knowledge, the Article III administrative judges in a specific group would regularly hear cases concerning certain statutes and regulations. Over time, these administrative judges would develop legal expertise as to a range of statutory and regulatory provisions.

The expertise of these judges would differ a bit from the expertise—both legal and non-legal—of agency adjudicators under the present system. Under the present system, adjudicators enjoy agency specific expertise. They know the agency’s organic statute and non-legal subject matter extremely well, but have less knowledge of other areas and of general skills, such as statutory interpretation. By contrast, the Article III administrative judges would have broader knowledge of both legal and non-legal matters since they would hear cases from a range of agencies in their subject matter. There are advantages and disadvantages to these different arrangements. But one significant advantage of the Article III arrangement is that it would avoid the tunnel vision that often afflicts specialized administrative agencies.

These Article III administrative judges should be appointed by the President with the advice and consent of the Senate. Thus, the judges would no longer be selected by the agencies, which can now use their appointment power as to ALJs to select persons whom they view as favorable to their interests or ideology. Moreover, the President’s selection would be subject to the significant check of senatorial consent.

These appointment methods would also be feasible. At present, there are approximately 270 ALJs (apart from those that hear cases involving Social Security or Medicare which I would treat under a different arrangement). A similar number of Article III administrative judges would not be too large, as there are now more than 650 authorized district court judges.

Decision-making Costs

The Article III administrative courts could also be designed to have low decision-making costs. There is no reason why such courts could not employ many of the same streamlined procedures that the existing system of administrative adjudication currently uses. Moreover, the expertise of these Article III administrative judges would also allow their decisions to be made more expeditiously.

The Article III administrative courts would have other features that would lower decision-making costs. Administrative courts would eliminate the multiple levels of appeals within an agency, as the administrative court would be the only adjudicator below the federal circuit courts. In addition, subpoenas would not need to be enforced by a separate court, as they often are now, but could be enforced by the Article III administrative court.

In conclusion, our system of administrative adjudication, which violates the basic principle that one should not be the judge in one’s own case, should be eliminated. We should replace it with a system of Article III administrative courts that respects the separation of powers, and does so without losing the benefits of expert and expeditious decision-making.

Michael B. Rappaport is the Hugh and Hazel Darling Foundation Professor of Law at the University of San Diego School of Law.

Print Friendly, PDF & Email