A context-sensitive voting protocol paradigm for multimember courts.

• Author: Nash, Jonathan Remy

INTRODUCTION I. OVERVIEW OF THE DOCTRINAL PARADOX II. CURRENT PRACTICE A. Forays into Issue-Based Voting B. Presumption Against Interlocutory Appeals C. Statement of Reasoning D. Resolution of Cases Under Current Practice is Not Purely Outcome-Based III. PROPOSALS TO ADDRESS THE DOCTRINAL PARADOX A. Uniform Rule Proposals B. The Metavote Proposal IV. SHORTCOMINGS OF EXISTING PROPOSALS TO ADDRESS THE DOCTRINAL PARADOX A. Shortcomings of Outcome-Based and Issue-Based Voting 1. Asserted shortcomings of outcome-based voting as opposed to issue-based voting 2. Asserted shortcomings of issue-based voting as compared to outcome-based voting B. Shortcomings of Kornhauser and Sager's Approach 1. Brief overview of Kornhauser and Sager's metavote proposal 2. Problem of conferring special treatment on particular classes of eases 3. Conceptual problems with the metavote V. A NEW PROPOSAL FOR A CONTEXT-SENSITIVE VOTING PROTOCOL PARADIGM A. Guiding Principles B. The Substantive Proposal C. Application VI. PROCEDURAL PROPOSALS TO MINIMIZE THE IMPACT OF DOCTRINAL PARADOXES CONCLUSION INTRODUCTION

Consider an appeal of a criminal conviction pending before a typical three-judge panel of a federal court of appeals. The appeal presents the court with two issues: first, whether it was error for the trial judge to have allowed into evidence a confession that the defendant claims was coerced, and second, whether, assuming the answer to the first question is affirmative, the error was harmless or not. One judge, Judge Alvin, is of the opinion that the admission of the confession was error, but harmless error. Another judge, Judge Brooks, would rule that the admission of the confession was error, and that the error was not harmless. The third judge, Judge Carne, believes that the admission of the confession was entirely proper but also is of the view that, were the admission assumed to be error, the error would not be harmless. In other words, while Judge Carne would hold that the admission of the confession does not constitute reversible error, had the prosecution conceded that it was error and thus restricted the issue on appeal solely to whether the error was harmless, Judge Carne would rule that it was harmful.

How should the court resolve this appeal? While the answer to this question at first blush might seem apparent, in fact it is not. Observe that the final result will turn upon whether the judges cast their votes on the basis of final outcome--at is, whether the defendant's conviction should be affirmed or reversed--or on an issue-by-issue basis. Under the first approach, there would be a 2-1 vote to affirm the conviction: Judges Alvin and Carne would vote to affirm (Judge Alvin because the error was harmless, and Judge Carne because there was no error), while Judge Brooks would vote to reverse. Under the second approach, the judges would vote first on the question of whether there was error, and second (if the first vote produced the result that there was error) on whether the error was harmless. In the first vote, only Judge Carne would vote that there was no error; Judges Alvin and Brooks would vote that there was. Accordingly, the second vote--on whether the error was harmless--would be necessary, and would result in a 2-1 vote that it was not harmless (with Judges Brooks and Carne constituting the majority). Under the second approach, then, the defendant's conviction would be reversed, since there was error that was not harmless.

Lewis Kornhauser and Lawrence Sager have identified situations such as this one as cases of "doctrinal paradox." (1) A doctrinal paradox arises when the outcome of an appeal varies depending upon the choice of "voting protocol"-that is, whether the court uses issue-based or outcome-based voting as a means of securing the result. (2)

The doctrinal paradox does not arise merely in hypothetical situations but indeed has arisen in practice. (3) Although judges in this country have traditionally adhered to outcome-based voting, (4) they sometimes opt to engage in issue-based voting. To the consternation of those who are aware of the issue, judges' forays into the "choice" of voting protocol appear to be standardless (5)--indeed, the courts generally do not even acknowledge the issue (6)--and piecemeal--that is, with different judges using different voting protocols in the same case. (7) This has led some to lament that "[c]urrent appellate practice with regard to paradoxical cases is in shambles." (8)

Commentators have offered different answers to the doctrinal paradox. John Rogers urges adherence to a uniform rule of outcome-based voting, (9) while Maxwell Stearns provides a positivist account of outcome-based voting as the result of a natural evolution in the decisionmaking process. (10) David Post and Steven Salop propose a uniform issue-based voting rule. (11) Kornhauser and Sager reject the propriety of any uniform rule of voting protocol, arguing instead in favor of a context-sensitive approach. (12)

Kornhauser and Sager, and Post and Salop, have exposed many shortcomings in outcome-based voting. At the same time, Kornhauser and Sager, now aligned with Rogers, have explored shortcomings with issue-based voting. It is precisely because Kornhauser and Sager find both voting protocols problematic in different situations--and both protocols valuable in different situations--that they argue in favor of a context-sensitive approach to the choice of voting protocol.

Kornhauser and Sager are correct to eschew a uniform voting protocol rule in favor of a context-sensitive approach. At the same time, however, the approach they outline is (as they themselves state) incomplete. Moreover, it is flawed in several respects. First, their approach erroneously affords special treatment to particular subject-matter classes of cases. Second, their approach fails to address the possibility that their context-sensitive approach itself might generate considerable disagreement among judges. Third, the substantive content of their approach is likely to increase occurrences of such disagreements.

I offer a competing context-sensitive approach to the choice of voting protocol. Unlike Kornhauser and Sager's approach, my proposal treats all cases alike regardless of subject matter. It relies more heavily on bright-line distinctions than does Kornhauser and Sager's approach; for that reason, it is less likely to give rise to disputes over the proper choice of voting protocol in a given case. Moreover, the proposal includes a methodology for addressing such disputes to the extent that they do arise.

I also offer a few procedural reforms that seek to limit the deleterious effects of doctrinal paradoxes when they do arise. I envision these reforms as supplementary to the primary substantive reform, but they might equally well be enacted regardless of the governing substantive voting protocol paradigm--even the current amorphous one.

This Article proceeds as follows. Part I presents an overview of the doctrinal paradox. Part II discusses the interplay between current voting practices on appellate courts and occurrences of the doctrinal paradox. Part III presents commentators' proposals to address the doctrinal paradox--uniform outcome- and issue-based voting, and Kornhauser and Sager's metavote approach. Part IV canvasses various problems (some more troubling than others) that commentators have identified with the various proposals discussed in Part III.

Parts V and VI present my proposal for occurrences of the doctrinal paradox. Part V outlines a context-sensitive voting protocol paradigm. Part VI suggests some modest procedural reforms that should accompany the substantive proposal in Part V, and that, in the alternative, should be enactedeven if the proposal in Part V is not. The short Conclusion includes some final reflections on the doctrinal paradox, the choice of voting protocol, and the contours of law.

1. OVERVIEW OF THE DOCTRINAL PARADOX

   What is the doctrinal paradox, and when might it occur? In this Part, I undertake to explain the doctrinal paradox in terms of appellate court (13) voting.

   As elucidated by the example in the Introduction, the doctrinal paradox arises wherever a panel of judges would reach one result in a case based upon outcome-based voting, and another differing result using issue-based voting (14) Kornhauser and Sager explain the conditions under which the paradox will arise in a simple, but not uncommon, class of cases:

   In simple disputes that present only one cause of action (or ground for recovery), the doctrinal paradox will arise if two conditions are met: (1) a majority of judges would each rule against the plaintiff on at least one issue, and (2) there is no one issue on which a majority of judges would join in ruling against the plaintiff. Condition (1) ensures that, in case-by-case adjudication, the court finds against the plaintiff. Condition (2) ensures that, in issue-by-issue adjudication, the court finds for the plaintiff. (15) Steams describes the conditions under which the paradox will arise using the concepts of dimensionality and symmetricality of preferences, explaining that the paradox will arise only in cases that present multidimensional judgment choice and asymmetrical preferences. (16) Dimensionality represents the number of choices relevant to ultimate disposition of a case. While a one-dimensional case may present more than two voting options (and thus may give rise to plurality opinions), it cannot give rise to a paradox. (17) This condition matches one's intuitive expectations: Roughly speaking, a case that presents only one issue could not produce one result using issue-based voting and another using outcome-based voting.

   Stearns employs a useful graphic form to demonstrate the symmetricality of preferences, or absence thereof. (18) Assuming that a case is two-dimensional, presenting issues A and B, the court's voting pattern can be represented in a...