Chevron deference should be replaced with a system that accords weight to both contemporaneous and customary interpretations.
With the changes in the Supreme Court, the Chevron and Auer doctrines are both in jeopardy of being significantly cut back or eliminated. I generally favor eliminating both doctrines. While there are arguments on both sides of the matter, I find the defenders of these doctrines fail to address one of the key disadvantages of judicial deference to administrative agencies.
The Chevron doctrine holds that when an agency interprets a statute it administers, its interpretation is often entitled to deference. In practice, this means that the agency’s interpretation will prevail in court so long as it is not unreasonable. The Auer doctrine is similar, but it applies to an agency’s interpretation of its own regulation. To oversimplify, the agency will prevail on its interpretation unless its interpretation was unreasonable.
The most common argument for providing Chevron and Auer deference is that the agency has expertise. It knows more about its statutes and regulations than courts do. Moreover, it has more policy expertise about the area than courts do. Thus, the agency should be not be reversed unless it is unreasonable. (Another argument involves the lower decisionmaking costs of agencies, but let’s put to the side for this post.)
This argument from expertise is not without weight. Other things being equal, we certainly want more expert people interpreting statutes and regulations than less expert people. But defenders of Chevron and Auer usually argue as if this is the main consideration, or even the only relevant one. But this is not true. There is also the issue of agency bias and the way these forms of judicial deference reduce incentives to interpret the law well. I call the failure to acknowledge these concerns “the agency lack of bias fallacy.”
The interpretation of a statute or a regulation is normally thought to be judicial power. The traditional separation of powers recognized that combining that power with executive power (or legislative power) led to bias and bad results. For example, if the same entity both brings an enforcement action and adjudicates the meaning of the law, then it has incentives to be biased as to both enforcement and adjudicatory incentives.
First, consider the bias that is created for enforcers. The enforcers will not have to ask whether their interpretation of the law is one that will persuade an impartial judge. Instead, they can merely ask whether they can get away with an interpretation that allows them to prosecute the defendant they are going after.
Second, consider the bias that is created for adjudicators. The adjudicators will not simply ask what is the best interpretation of the law. Instead, they will ask what interpretation will further the prosecutions they are conducting.
The basic point is that conferring judicial power on an executive agency makes it worse at doing its job. (The same point applies to the combining of other powers, such as legislative and judicial power.) Thus, even if one believes that expertise (or low decisionmaking costs) is a benefit of Chevron deference, one must consider the increased bias that it permits and encourages. Most defenders omit this consideration, which renders their analysis seriously incomplete.