The United States Supreme Court heard oral argument last week in a critical if obscure administrative law case: Kisor v. Wilkie asked the simple question of whether the courts should be highly deferential to federal administrative officials in the interpretation of their own statutes. Rebuffing agencies—like the Department of Veterans Affairs, in this case—could reshape the world of modern administrative law. The specific question at issue in Kisor is how to determine the date at which a veteran becomes eligible for disability benefits. Few people seem interested in the particulars of the case, not even the nine Justices, but a clear understanding of them and other key cases is needed to orient the basic theoretical discussion. Ultimately, giving judicial deference is either unnecessary or mischievous. It is best to make agencies defend their legal position like any other party in the system.

In this case, James Kisor applied to the VA for a disability claim for post-traumatic stress disorder. No one doubted that he had the condition. The dispute was over when it started. Kisor claimed an earlier date than the VA allowed, and he sought to introduce evidence from his file to support his contention. The VA disregarded his new evidence. On its view, the governing statutory provision requires that claims for disability benefits be reconsidered only if the VA gets “relevant official service documents” that pertain to his claim. Kisor thought that he should be able to introduce evidence already in the record when his initial claim date was determined. The VA refused to consider that new evidence from the record because it predated the “last denial” of Kisor’s claim. Kisor responded that this earlier evidence should be reviewed in any event so long as it has “any tendency” to tip the outcome in the case. Kisor wanted, therefore, a broader reading of the term “relevant” than the VA allowed, and he claimed that the Federal Rules of Evidence supported him because it used the same broad “any tendency” definition of “relevant” that he endorsed.

In ordinary litigation, Kisor’s argument would carry a lot of weight. But not here. Unfortunately, the Circuit Court for the Federal Circuit punted deciding the issue on the merits, saying in effect that it owed extreme deference to the federal administrator in the construction of his own statute. Eyebrows should raise because the VA is an interested party—but the Federal Circuit invoked two Supreme Court cases, decided over 50 years apart, Bowles v. Seminole Rock (1945) and Auer v. Robbins (1997), to bolster its decision. Both those decisions have come under spirited attack from the conservative side of the Court, because they are said to represent an abnegation of the judicial duty to decide all questions of law that come before a court. The liberal response has been equally emphatic, insisting that Congress wants agencies that use their greater expertise on these specialized issues to determine coherent public policy. So who’s right? To answer that question, it is important to see how judicial deference works in practice.