The Yale Law Journal

VOLUME
116
2006-2007
Forum

Chevron Should Not Be Converted into a Voting Rule: A Response to Gersen and Vermeule's Proposal

26 Jan 2007

Jacob Gersen and Adrian Vermeule urge the Supreme Court to convert the Chevron doctrine into a voting rule. In Chevron, the Supreme Court instructed reviewing courts to reject an agency interpretation of an agency-administered statute if the court concludes that Congress clearly resolved the question in a manner inconsistent with the agency’s interpretation, but to uphold any reasonable agency interpretation of an ambiguous statutory provision. Gersen and Vermeule propose to replace the two-step Chevron test with a single-step test—de novo review of all agency statutory interpretations—coupled with a super-majority requirement to overturn an agency interpretation.

Under Chevron, a judge is required to ask two questions. First, the judge must determine whether Congress resolved the specific issue before the agency. If so, the statute has a clear meaning as a matter of law, and the judge’s task is to reject any agency interpretation that is inconsistent with the congressional resolution of the issue. If the judge determines instead that the statute is ambiguous, i.e., that Congress did not resolve the issue presented, the judge then knows that he is reviewing a policy decision, rather than a legal decision, and the judge must uphold the agency’s resolution of the policy dispute as long as it is reasonable. In contrast, under Gersen and Vermeule’s proposal, a judge asks only one question—what is the best interpretation of the statute? If the judge disagrees with the agency’s interpretation, he votes to reject it. Under their proposal, however, a circuit court cannot reject an agency interpretation unless all three members of a panel vote to reject it, and the Supreme Court cannot reject an agency interpretation unless six of the nine Justices vote to reject it.

The Gersen and Vermeule proposal reflects the fertile imagination of these two bright young scholars, but I see no merit in it. My response to their proposal is three-fold. The proposal is not needed; it is not desirable; and it is not going to happen.

I. The Gersen and Vermeule Proposal Is Not Needed

I agree with Gersen and Vermeule’s characterization of the goals of Chevron. It has the potential to further three closely related goals—“to open up a ‘policy space’ that gives agencies breathing room to pursue policies based on technocratic judgments or democratic politics,” to reduce the effects of the political biases of judges when they are called upon to review agency policy decisions that take the form of interpretations of ambiguous statutes, and to increase the consistency and predictability of judicial review of agency policy decisions. The Gersen and Vermeule proposal is premised on their belief that Chevron has failed to further those goals. That is the point on which we disagree. I believe that Chevron has succeeded. Of course, part of our disagreement is definitional. Thus, for instance, if Gersen and Vermeule define success in this context as eliminating all effects of judicial bias in reviewing agency actions, I agree that Chevron has been a failure. If instead, one defined success as significantly reducing bias in judicial review of agency decisions, all of the available evidence supports my belief.

A few years after the Court decided Chevron, Peter Schuck and Donald Elliott published a study in which they found that Chevron had reduced the rate of judicial rejection of agency statutory interpretations by thirty-nine percent. It is hard to imagine stronger evidence that Chevron has achieved its first goal—to increase the space in which agencies can make policy decisions. We are unlikely ever to have access to better evidence of this effect than the Schuck and Elliott study because lawyers adjust their patterns of behavior over time to reflect changes in patterns of judicial behavior. In other words, as lawyers realize that their probability of prevailing by making a particular argument in a review proceeding has declined, they make the argument less frequently, thereby producing a gradual return to approximately the rate of success of the argument that existed before the change in the pattern of judicial behavior. Thus, any more recent study of the effects of Chevron would have less validity than the Schuck and Elliott study.

Ricky Revesz’s 1997 empirical study of D.C. Circuit decision-making provides powerful evidence that Chevron has been a success in furthering its other goals. Revesz analyzed 250 environmental cases decided by the D.C. Circuit. He found “staggering differences” between the rates of reversal of EPA by Republican and Democratic judges when they applied the State Farm doctrine. Thus, for instance, he found that Republican judges detected fatal flaws in EPA reasoning in eighty-nine percent of cases in which regulatees challenged an EPA action as arbitrary and capricious under State Farm, while Democratic judges detected fatal errors in EPA reasoning in only thirteen percent of such cases. Revesz also found no statistically significant difference between the results of Democratic and Republican judges’ applications of Chevron in the same population of cases, however. Revesz’s study provides powerful support for four important propositions: (1) judges experience powerful temptation to vote in accordance with their personal political biases in administrative law disputes with high political salience; (2) the probability that a judge will succumb to that temptation depends largely on whether the applicable doctrine is relatively objective and determinate or is instead subjective and malleable; (3) the State Farm doctrine is subjective and malleable; and, (4) the Chevron doctrine is relatively objective and determinate. The Revesz study also provides powerful support for my belief that Chevron is furthering all three of its goals.

As Gersen and Vermeule note, some studies have detected political bias in judicial applications of Chevron. In a 2006 study, Thomas Miles and Cass Sunstein found that “conservative justices are 30 percentage points more likely to vote to validate agency interpretations that are coded as conservative than to validate agency interpretations coded as liberal. By contrast, the more liberal justices are 27 percentage points more likely to vote to validate agency interpretations coded as liberal than to validate those coded as conservative.” In circuit courts, Miles and Sunstein found that "[w]hen the agency decision is liberal, the average Democratic appointee . . . is about 14 percentage points more likely to vote to validate the agency than the average Republican appointee. But when the agency decision is conservative, this pattern flips, and the average Democratic appointee is 19 percentage points less likely to vote to validate the agency decision than the average Republican appointee.”

The Miles and Sunstein study supports two important propositions: Chevron has been less effective in furthering its goals in the Supreme Court than in circuit courts, and Chevron has not completely eliminated political bias even in circuit courts. The Miles and Sunstein study also confirms my belief that Chevron has succeeded in furthering its goals significantly. One way of measuring Chevron’s relative success is to compare Revesz’s findings with respect to circuit court applications of State Farm with Miles and Sunstein’s findings with respect to circuit court applications of Chevron. If you restate Revesz’s findings in a manner consistent with Miles and Sunstein’s description of their findings, a Democratic judge is seventy-six percentage points more likely than a Republican judge to validate a liberal agency decision when judges apply the State Farm doctrine. By contrast, a Democratic judge is only fourteen percentage points more likely than a Republican judge to validate a liberal agency decision when judges apply Chevron. By my standards, that eighty percent reduction in the effects of political bias and eighty percent increase in the consistency and predictability of judicial decision-making makes Chevron one of the greatest success stories in U.S. legal history. Thus, Chevron has succeeded, and the Gersen and Vermeule proposal is not necessary.

II. The Gersen and Vermeule Proposal Is Not Desirable

Gersen and Vermeule draw from psychology and political science theories to support their belief that their proposed voting rule will do a better job of furthering the goals of Chevron than has the doctrinally based Chevron test. They are appropriately cautious with respect to the degree of confidence they express. Thus, they use language like “we may plausibly conjecture.” Stated in those terms, I have difficulty quarreling with their assertions. Frankly, I lack confidence that I can predict accurately all of the effects of judicial adoption of their proposal. I am confident that the ironclad rule of unintended consequences would manifest itself in some ways in this situation as it does in all others, however. I fear that the consequences of adopting the Gersen and Vermeule proposal could be far worse than the results of the imperfect Chevron doctrine. To use the Gersen and Vermeule method of describing one of my many concerns, it is “not altogether implausible” that judges asked to vote based on their respective beliefs about the best interpretation of an ambiguous statute would indulge their tendency to vote in a manner consistent with their political biases to the same “staggering” extent that Revesz found that they do when they apply the open-ended State Farm test.

When such a voting pattern is combined with the supermajority requirement Gersen and Vermeule propose, it might yield about the same rate of judicial validation of agency policy decisions that Chevron now produces. It would produce a much worse world, however, when it is viewed through a broader prism. Courts should be doing what they can to reduce the effects of political bias on the voting patterns of judges. The Gersen and Vermeule proposal would move in the opposite direction. It would be an implicit invitation to judges and Justices to vote in accordance with their political biases. That, in turn, would increase the already high and growing level of public cynicism about the behavior of judges and Justices.

III. The Gersen and Vermeule Proposal Is Not Going To Happen

I am not going to lose any sleep over the possibility that the Supreme Court might adopt the Gersen and Vermeule proposal. That will not happen. I rarely rate anything a zero percent probability event, but this proposal meets my strict criteria for doing so. In the 1930s, supporters of the New Deal proposed a supermajority requirement to invalidate statutes on constitutional grounds. The proposal went nowhere even in an environment in which the supporters of New Deal legislation controlled the White House and both houses of Congress, and in which there was widespread anger about the perceived arrogance of the Supreme Court for resisting the will of the people by striking down popular legislation.

No Anglo-American court has ever adopted a supermajority requirement applicable to the merits of any class of disputes, for good reasons. A supermajority requirement would leave a policy in effect even though it has been delegitimated by judicial review and would privilege the votes of some Justices and judges and dilute the votes of others. Imagine the political campaign that an opponent of a policy could mount against a supermajority requirement when a policy remains in effect after five Justices vote to invalidate it. A supermajority requirement would increase the present disturbing level of antipathy toward the judiciary. The Gersen and Vermeule proposal provides the occasion for an interesting academic debate, but that is all it will ever be.

Professor Richard J. Pierce is Lyle T. Alverson Professor of Law, George Washington University. He wishes to thank Michael Abramowicz, Donald Clarke, John Duffy, David Fontana, Richard Freer, Orin Kerr, Fred Lawrence, Chip Lupu, Greg Maggs, Susan Martyn, and Sean Murphy for providing helpful comments on an earlier version of this essay.

Preferred citation: Richard J. Pierce, Jr., Chevron Should Not Be Converted into a Voting Rule: A Response to Gersen and Vermeule’s Proposal, 116 Yale L.J. Pocket Part 248 (2007), http://yalelawjournal.org/forum/chevron-should-not-be-converted-into-a-voting-rule-a-response-to-gersen-and-vermeuleas-proposal.

Related Content:

  • Article: Jacob E. Gersen & Adrian Vermeule, Chevron as a Voting Rule, 116 Yale L.J. 676 (2007).
  • Response: Richard J. Pierce, Jr., Chevron Should Not Be Converted into a Voting Rule: A Response to Gersen and Vermeule’s Proposal, 116 Yale L.J. Pocket Part 248 (2007), http://thepocketpart.org/2007/01/25/pierce.html.
  • Response: Matthew C. Stephenson, The Costs of Voting Rule Chevron: A Comment on Gersen and Vermeule’s Proposal, 116 Yale L.J. Pocket Part 238 (2007), http://thepocketpart.org/2007/01/25/stephenson.html.