INTRODUCTION 157 I. CONCEPTS AND TRENDS 162 A. Lexical Ordering 162 B. Widespread Endorsement 163 1. Legislative History 164 2. Agency Interpretations 170 II. TRADE-OFFS IN THEORY 173 A. Ideal Decision Situations 174 1. Against Lexical Ordering 174 2. Principled Beginnings 177 B. Nonideal Constraints 180 1. Decision Costs 181 2. Indecision Risks 184 C. Convenient Compromise 187 III. IMPLEMENTATION CHALLENGES 190 A. Prior Research 190 B. Clarity Survey 194 C. Vignette Experiment 196 1. Participants and Procedure 196 2. Hypotheses 199 3. Results 201 D. Implications 209 1. Patterns and Limitations 209 2. Failure's Types and Success's Burdens 212 3. Toward Reform 213 CONCLUSION 216 APPENDIX A 217 APPENDIX B 218 Where the language is plain... the rules which are to aid doubtful meanings need no discussion. (**)
And you must strike it from your minds. (***)
Abandoned at the Supreme Court for decades, lexical ordering returned to the core conventions of statutory interpretation by the late twentieth century. (1) That is, today most judges are supposed to decide whether a statute is clear using a limited set of top-tier sources and, if so, apply this meaning; if the statute remains unclear, lower-tier sources may or must be considered. Lower-tier sources are held aside just in case. Examples familiar to lawyers involve plain meaning rules that demote without condemning considerations such as legislative history, deference to administrative agencies, and the rule of lenity. Some version of this analytical strut is now planted in the approved method for deciding statutory cases in nearly every court in the country. A popular judicial turn of phrase nowadays is that legislative history "is meant to clear up ambiguity, not create it," (2) and agency positions usually are assigned a comparable tiebreaking role. (3)
Actually, the idea of lexically ordered interpretation is no younger than Blackstone's Commentaries. In 1765, Blackstone listed secondary considerations for statutory cases that were supposed to matter only if "words happen to be still dubious." (4) His lower-tier sources included statutes on the same subject, along with the reason and spirit of the statute at issue. (5) Lexically ordered statutory interpretation, which can take many forms, was endorsed at the Supreme Court by 1920. (6) But that approach always had competitors. Sutherland's Statutes countered with an apparently wide-open aggregation of valid sources. The 1943 edition of the treatise pronounced that "statutory interpretation... is a fact issue. Where available, the courts should never exclude relevant evidence on that issue of fact." (7) In this regard the treatise was less Blackstone and more Holmes, (8) as were manyjudicial opinions in the 1940s through the 1970s. (9) Equally important, simplification through outright exclusion of interpretive sources has been recognized as an option for centuries. (10)
Blackstone did not evaluate the trade-offs and implementation challenges for lexical ordering, of course. Nor have the rest of us done much better since then, even as the idea returned to prominence by the 1990s. (11) Indeed, judges might have become all too familiar with lexical ordering as a methodological compromise. (12) One side's persistent support for using certain sources is joined with another side's persistent objection to those sources, and a compromise is announced in which the controversial sources are not repudiated, not embraced, and not reweighted to calibrate their influence. Instead, all interpreters are asked to demote the controversial sources into a lower tier of the decision tree--with the hinge to demoted sources turning on each interpreter's sense of clarity, which is supposed to develop while demoted sources are ignored.
We ought to reconsider, now, whether this complex doctrinal architecture is justified and whether judges can follow it. Without satisfying either the judge who prefers to mix together or the judge who prefers to flatly exclude sources, the lexical ordering of sources lacks a simple and principled defense under ideal conditions, thus far. (13) Adding practical concerns about decision costs and judicial decisiveness tends to strengthen the argument for lexical ordering, as we shall see. But the upsides still might not be worth the effort after considering the downside risk that useful information will be ignored. (14) The trade-offs should be troubling.
On the other hand, perhaps the instructions do not matter. Judges and others have suggested repeatedly that there is nothing clear about clarity tests for statutes. (15) Perhaps judges have convenient understandings of "clarity" that allow each judge to consider lower-tier sources when they want to and not when they do not. Another possibility is that judges are unable to follow the instructions regardless of their preferences. "When the reading is done and the case has been analyzed and argued," Judge Randolph asked twenty years ago, "how can it be said that the judge turned to the legislative history only after finding the statutory language ambiguous?" (16) The same question should be asked about every lower-tier source. (17)
Before another twenty years pass, we need to make additional theoretical and empirical progress on all of the above ideas. This Article aims to help by exploring multiple aspects of lexical ordering in statutory cases: its resurgence, its unique trade-offs when followed, and its implementation challenges, alongside new experimental evidence. These matters are tightly connected. Lexical ordering's spread increases the urgency of revisiting its core trade-offs, which are shaped by its ground-level implementation. There will be much good work left over, in statutory interpretation and in other fields where lexical ordering might be adopted. (18) But the work below contributes on multiple fronts. More broadly, efforts like this support thoughtful evaluation of the intertwined architectures of interpretation--doctrinal, cognitive, institutional, physical--by taking theory and implementation seriously, and together. There can be no simple and definitive take on lexical ordering for now. We should recognize this, even advertise it, as motivation to continue rethinking the foundations of statutory interpretation within realistic decision environments.
Part I reviews the logic of lexical ordering and shows how far it has spread into the conventions for deciding statutory cases. The discussion focuses on the position of legislative history and agency interpretations. Joining working majorities at the U.S. Supreme Court, more than forty state supreme courts now endorse lexical inferiority for state-level legislative history, and approximately thirty-five do so for state agency interpretations. But official positions on interpretive method have not been stable over long periods in the past, whatever effect they have on behavior.
Part II identifies hard trade-offs, assuming that lexical ordering instructions are followed and that the clarity test is diagnostic. (19) Compared to aggregating all relevant source inferences, lexically ordering sources threatens decision quality, but it can reduce decision costs and probably will increase decisiveness. Compared to flatly excluding lower-tier sources, lexical ordering probably yields higher quality decisions and increased decisiveness, but also higher decision costs. Effects on the probability of a decisive outcome, in particular, are not widely recognized today. Adding indecision risks to decision costs does generate plausible arguments for demoting sources such as agency interpretations and legislative history. Perhaps the overall compromise is tolerable to methodological moderates. But these debatable judgment calls should be recognized for what they are. Moreover, the actual trade-offs depend on whether judges lexically order sources in their decisionmaking, not only in their opinion writing. To date, we lack on-point evidence either way.
Part III turns to implementation and evidence. It presents the results of a new survey and experiment conducted with approximately one hundred appellate judges. The survey offers only limited evidence that judges' propensities to find statutory text clear are aligned with their general views about the usefulness of legislative history. (20) Instead, the survey suggests that lexical ordering entails sacrifice from many judges. In the vignette experiment, though, judges showed curiously mixed success at ignoring lower-tier sources. (21) In a trade name case, we find little evidence that judges were improperly influenced by legislative history. In an election law case, by contrast, we do find evidence that judges were improperly influenced by an agency's position. Even in the election law case, however, we cannot confirm that judges conveniently bent the clarity test to change the legal relevance of attractive or unattractive sources. Other kinds of failure are possible--including that exposure to a lower-tier source will, distractingly, boost the judge's assessment of statutory clarity with or without final judgments flipping. (22)
There is more to learn. But our results do suggest the outlines of actionable lessons. Sometimes judges will indeed succeed at lexical ordering (perhaps when the case is run-of-the-mill and the lower-tier source presents a complication) in which case the instructions will bite and the hard trade-offs of quality, cost, and decisiveness should be confronted. At other times, judges will fail (perhaps when the case is ideologically charged and the lower-tier source category is attractive) in which case lexical ordering instructions are at best wasteful distractions. As we build theory and evidence to better identify domains of success and failure, we should generally expect mixed judicial success at achieving the mixed advantages and disadvantages of lexically ordered statutory interpretation. Maybe that much...