Among the many reasons for mourning Justice Scalia’s untimely passing (on which I’ve written at length elsewhere) is the fact that his death abruptly cut short his late-career reconsiderations of the administrative state.

As Aaron Nielson observes, recent years had seen Justice Scalia expressing serious doubts about judicial deference to agency interpretations of their own rules — that is, doubts about the Seminole Rock–Auer doctrine that he had expounded for so long. Beginning with his Talk America concurrence in 2011, Scalia newly questioned whether allowing agencies to both write and interpret the law clashed with the Framers’ fundamental view, per Montesquieu, that the tasks of lawmaking, execution, and adjudication be kept separate.

By the time that he issued his concurrence last Term in Perez v. Mortgage Bankers, Scalia was long past the point of asking questions. Wholly embracing this critique of Auer, which he had drawn from his former clerk John Manning, Scalia announced that “I would . . . restore the balance originally struck by the APA with respect to an agency’s interpretation of its own regulations, not by rewriting the Act in order to make up for Auer, but by abandoning Auer and applying the Act as written. The agency is free to interpret its own regulations with or without notice and comment; but courts will decide—with no deference to the agency—whether that interpretation is correct.”

But his Perez concurrence contained an even more surprising attack: on Chevron itself. The “problem” of judicial deference “is bad enough,” Scalia wrote, “and perhaps insoluble if Chevron is not to be uprooted, with respect to interpretive rules setting forth agency interpretation of statutes.”

Scalia made at least a passing attempt to distinguish Chevron from Auer: “As I have described elsewhere, the rule of Chevron, if it did not comport with the APA, at least was in conformity with the long history of judicial review of executive action, where ‘[s]tatutory ambiguities . . . were left to reasonable resolution by the Executive.’ United States v. Mead Corp., 533 U. S. 218, 243 (2001) ( Scalia, J., dissenting ). I am unaware of any such history justifying deference to agency interpretations of its own regulations.”

But no matter how he nominally limited it, even this restrained criticism of Chevron surprised those who had watched Scalia defend Chevron deference, more than any of his colleagues, from the very beginning. If his belated acceptance of Mead’s narrowing of Chevron, in his opinion for the Court in City of Arlington, could be explained away as the price Scalia paid to write the Arlington majority, there could be no similar rationalization for his gratuitous criticism of Chevron in the Perez concurrence.

And in fact Scalia was seriously reconsidering Chevron deference — or so he said in conversations in recent months, word of which spread quickly, if quietly, in legal circles. It was a remarkable turn of events, and his death robs us of the opportunity to follow his reconsideration of Chevron deference to its proper conclusion.

In the last week, there has been some discussion of Scalia and Chevron. Such reflections tend to trace the story back to Scalia’s 1989 article in the Duke Law Journal, “ Judicial Deference to Administrative Interpretations of Law .” There Scalia, still rather newly arrived on the Supreme Court, defendedChevron in terms of congressional intent. Where Congress “had no particular intent” as to the interpretation of a given statute, “but meant to leave its resolution to the agency,” then “what we have is the conferral of discretion upon the agency, and the only question of law presented to the courts is whether the agency has acted within the scope of its discretion—i.e., whether [the agency’s] resolution of the ambiguity is reasonable.”

It is a very interesting article, and indispensible to those focused specifically on Scalia’s evolution on Chevron. But it would be a mistake not to look further into this article, and the writings that preceded it. Because only then can we see the more clearly the roots of his original support of Chevron—and, today, his late reconsideration on the subject.

In the Duke Law Journal article, Scalia made clear that his embrace of deference was not based on abstract principle so much as a practical judgment. On the presumption that Congress intended to confer interpretative discretion to agencies instead of courts, Scalia noted that “[i]t is beyond the scope of these remarks to defend that presumption.” But he noted that “[s]urely, however, it is a more rational presumption today than it would have been thirty years ago—which explains the change in the law [in favor of deference].” In other words, Scalia believed that Chevron reflected well enough the practical reality of modern governance. He re-emphasized this point in his closing words: “I tend to think . . . that in the long run Chevron will endure and be given its full scope—not so much because it represents a rule that is easier to follow and thus easier to predict (though that is true enough), but because it more accurately reflects the reality of government, and thus more adequately serves its needs.”

In fact, these are themes that he repeated several times before his appointment to the federal bench. But because these writings were primarily in essays, not in law reviews, they have attracted much less attention.

Beginning in 1977, when he left government to join the American Enterprise Institute (a think-tank in Washington), Scalia wrote for and later edited Regulation, AEI’s house organ. He continued to edit and write for Regulation long after he moved to the University of Chicago.

Those essays—which I describe and list in a blog post this week at the Weekly Standard —contain many examples of Scalia grappling with the administrative state’s role in modern federal government. One sees Scalia advancing judicial deference not because he sees it commanded by constitutional principle, but rather because he was attempting to make a prudential judgment in an area not susceptible to bright-line rules.

As Scalia explained in his famous Mistretta dissent, the difference congressional delegations of power to the executive branch and such delegations to the judicial branch is not always self-evident, and thus such lines must eventually be drawn, if at all, “according to common sense and the inherent necessities of the governmental co-ordination.” We see a similar, practically minded approach in his Regulation essays.

In his contribution to its inaugural issue in 1977, Scalia wrote of the interplay between agencies, congressional delegations of power, and judicial imposition of procedural and substantive requirements on the agencies. Because “[m]odern delegations include standards no more precise (and no more limiting) than the injunction to adopt rules ‘in the public interest,’” one might expect agencies to run amok. But in fact, he further explained, “[t]he system has functioned reasonably well, however, because—either consciously or through the instinctive capacity for fashioning workable arrangements that characterizes Anglo-American politics—the second method for achieving rough compliance with the legislative will,” namely the focusing of political forces within and upon the agencies, “has in fact been adopted.”

In other articles (including one that preluded his Mistretta opinion nearly verbatim), Scalia counseled against further burdening administrative agencies with judicial impediments, because he saw the agencies as largely succeeding in both respecting the rule of law and heeding popular will. In an essay published for Reagan’s first inauguration, Scalia warned his fellow conservatives that “the game has changed” on questions of regulatory oversight. “Executive-enfeebling measures” long pursued by Congress-focused conservatives “do not specifically deter regulation. What they deter is change,” and when “imposed upon an executive that is seeking to dissolve the encrusted regulation of past decades . . . will impede the dissolution.” He wrote bluntly: “Regulatory reformers who do not recognize this fact, and who continue to support the unmodified proposals of the pat as though the fundamental game had not been altered, will be scoring points for the other team.”

A few years later, in his 1983 preface to the Journal of Law and Politics, Judge Scalia (by then appointed to the D.C. Circuit) neatly summarized the basic balance to be struck on such questions:

Resolving the tension between the rule of law and the will of the people — between law and politics — is the supreme task of our government system. We sometimes tend to forget that it is more a matter of resolving tensions than of drawing lines, for there is no clear demarcation between the two. Laws are made, and even interpreted and applied (by administrative agencies), through a political process; and politics are conducted under the constitution and statutory constraints of the law.

Deference doctrines seemed to vindicate both the rule of law and the need for political accountability, as he explained six years later in the Duke Law Journal. But it would be a mistake to conclude that Scalia assumed the doctrines would always maintain that balance. (Recall that even in his Duke Law Journal article he stressed that the presumption of congressional intent toward agency interpretation, justifiable in 1989, would have been a much less “rational” presumption 30 years earlier.) As Scalia would warn similarly in 1982 on the question of drawing lines in federalism issues, there may be “understandable tactical reason[s]” for striking a particular balance at the outset, but one must be vigilant lest a tactic “develop into a philosophy.”

In his later years, it is not hard to see Scalia concluding that the regulatory game had indeed changed once again. And not for partisan reasons—his commitment to deference did not falter during the Clinton Administration. But because he saw as much as anyone the tectonic shifts occurring in the administrative state, with agencies acting ever more aggressively in lieu of statutes, or against them altogether. In cases like King v. Burwell and Utility Air Regulatory Group v. EPA, Scalia criticized what he saw (rightly, in my opinion) as agencies asserting unprecedented power over American people, companies, and organizations, with ever-decreasing commitments to following statutory direction or constitutional constraints. One can add to that list still more cases in which Scalia joined the opinion of the Court pushing back strongly against agencies, such as Burwell v. Hobby Lobby and Hosanna-Tabor v. EEOC. One suspects that Scalia would have eventually added U.S. v. Texas, the immigration policy case, to that list. Given a couple more years, perhaps he would also have added to that list the cases challenging the EPA’s so-called “Clean Power Plan”— which he voted to enjoin pending litigation, in one of his last acts as a Justice —and the FCC’s so-called “Open Internet Order.”

In other words, Justice Scalia may have surveyed recent years’ developments and concluded that the state of affairs justifying judicial deference twenty-five years ago no longer held, and that it was time to strike a new balance between the courts and agencies.

This is not to say that he would have completely undone Chevron, or that it is obvious what new model he would have adopted in lieu of it. But whatever choice he would ultimately have made, he would have struck a balance intended to “resolv[e] the tension between the rule of law and the will of the people—between law and politics.”

On such questions of administrative law, that had been his work all along.

*Adam J. White is a visiting fellow at the Hoover Institution in Washington, D.C.